rus-cyrl-10mb-100mb_seed3407/checkpoint-21000/trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.8351830813597831,
  "eval_steps": 3000,
  "global_step": 21000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 10.742575645446777,
      "epoch": 0.0004369483527047103,
      "grad_norm": 4.875,
      "learning_rate": 2e-06,
      "loss": 10.7635,
      "mean_token_accuracy": 0.00012376237427815794,
      "num_tokens": 10478.0,
      "step": 5
    },
    {
      "entropy": 10.742580890655518,
      "epoch": 0.0008738967054094206,
      "grad_norm": 5.25,
      "learning_rate": 4.5e-06,
      "loss": 10.7564,
      "mean_token_accuracy": 5.527915782295167e-05,
      "num_tokens": 21152.0,
      "step": 10
    },
    {
      "entropy": 10.74257001876831,
      "epoch": 0.0013108450581141308,
      "grad_norm": 5.15625,
      "learning_rate": 7e-06,
      "loss": 10.7163,
      "mean_token_accuracy": 6.680026999674737e-05,
      "num_tokens": 31246.0,
      "step": 15
    },
    {
      "entropy": 10.742575645446777,
      "epoch": 0.0017477934108188413,
      "grad_norm": 4.75,
      "learning_rate": 9.5e-06,
      "loss": 10.6955,
      "mean_token_accuracy": 0.0014378628693521023,
      "num_tokens": 41934.0,
      "step": 20
    },
    {
      "entropy": 10.742451190948486,
      "epoch": 0.0021847417635235517,
      "grad_norm": 4.09375,
      "learning_rate": 1.2e-05,
      "loss": 10.5979,
      "mean_token_accuracy": 0.01751632974483073,
      "num_tokens": 52830.0,
      "step": 25
    },
    {
      "entropy": 10.741966247558594,
      "epoch": 0.0026216901162282617,
      "grad_norm": 3.875,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 10.4324,
      "mean_token_accuracy": 0.046310391277074814,
      "num_tokens": 62706.0,
      "step": 30
    },
    {
      "entropy": 10.739739894866943,
      "epoch": 0.003058638468932972,
      "grad_norm": 3.15625,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 10.3067,
      "mean_token_accuracy": 0.046505602076649666,
      "num_tokens": 72619.0,
      "step": 35
    },
    {
      "entropy": 10.732788467407227,
      "epoch": 0.0034955868216376825,
      "grad_norm": 2.90625,
      "learning_rate": 1.95e-05,
      "loss": 10.198,
      "mean_token_accuracy": 0.04933134503662586,
      "num_tokens": 82903.0,
      "step": 40
    },
    {
      "entropy": 10.720102882385254,
      "epoch": 0.003932535174342393,
      "grad_norm": 2.21875,
      "learning_rate": 2.2e-05,
      "loss": 10.0464,
      "mean_token_accuracy": 0.06326089426875114,
      "num_tokens": 94243.0,
      "step": 45
    },
    {
      "entropy": 10.703330898284912,
      "epoch": 0.004369483527047103,
      "grad_norm": 2.125,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 10.0094,
      "mean_token_accuracy": 0.05352335274219513,
      "num_tokens": 103618.0,
      "step": 50
    },
    {
      "entropy": 10.69249677658081,
      "epoch": 0.004806431879751813,
      "grad_norm": 1.8671875,
      "learning_rate": 2.7e-05,
      "loss": 9.9487,
      "mean_token_accuracy": 0.048276785761117935,
      "num_tokens": 112180.0,
      "step": 55
    },
    {
      "entropy": 10.693581295013427,
      "epoch": 0.005243380232456523,
      "grad_norm": 2.140625,
      "learning_rate": 2.95e-05,
      "loss": 9.9131,
      "mean_token_accuracy": 0.05066986009478569,
      "num_tokens": 122462.0,
      "step": 60
    },
    {
      "entropy": 10.694738101959228,
      "epoch": 0.005680328585161234,
      "grad_norm": 1.8671875,
      "learning_rate": 3.2e-05,
      "loss": 9.8716,
      "mean_token_accuracy": 0.05316077768802643,
      "num_tokens": 132741.0,
      "step": 65
    },
    {
      "entropy": 10.687106227874756,
      "epoch": 0.006117276937865944,
      "grad_norm": 2.015625,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 9.7556,
      "mean_token_accuracy": 0.055888672918081285,
      "num_tokens": 142973.0,
      "step": 70
    },
    {
      "entropy": 10.674335384368897,
      "epoch": 0.006554225290570655,
      "grad_norm": 1.8046875,
      "learning_rate": 3.7e-05,
      "loss": 9.7947,
      "mean_token_accuracy": 0.04946387633681297,
      "num_tokens": 153199.0,
      "step": 75
    },
    {
      "entropy": 10.660592651367187,
      "epoch": 0.006991173643275365,
      "grad_norm": 1.8203125,
      "learning_rate": 3.95e-05,
      "loss": 9.6937,
      "mean_token_accuracy": 0.05424195118248463,
      "num_tokens": 163079.0,
      "step": 80
    },
    {
      "entropy": 10.647120380401612,
      "epoch": 0.0074281219959800755,
      "grad_norm": 1.7578125,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 9.6379,
      "mean_token_accuracy": 0.05559157766401768,
      "num_tokens": 173308.0,
      "step": 85
    },
    {
      "entropy": 10.629187202453613,
      "epoch": 0.007865070348684786,
      "grad_norm": 1.65625,
      "learning_rate": 4.45e-05,
      "loss": 9.5576,
      "mean_token_accuracy": 0.0650589868426323,
      "num_tokens": 183856.0,
      "step": 90
    },
    {
      "entropy": 10.605023765563965,
      "epoch": 0.008302018701389496,
      "grad_norm": 1.796875,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 9.4922,
      "mean_token_accuracy": 0.055145786702632905,
      "num_tokens": 193368.0,
      "step": 95
    },
    {
      "entropy": 10.602028179168702,
      "epoch": 0.008738967054094207,
      "grad_norm": 1.78125,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 9.4365,
      "mean_token_accuracy": 0.05791991353034973,
      "num_tokens": 202686.0,
      "step": 100
    },
    {
      "entropy": 10.589914703369141,
      "epoch": 0.009175915406798917,
      "grad_norm": 1.7578125,
      "learning_rate": 5.2e-05,
      "loss": 9.3332,
      "mean_token_accuracy": 0.06775561273097992,
      "num_tokens": 213237.0,
      "step": 105
    },
    {
      "entropy": 10.540931034088135,
      "epoch": 0.009612863759503626,
      "grad_norm": 1.59375,
      "learning_rate": 5.45e-05,
      "loss": 9.2707,
      "mean_token_accuracy": 0.0691582515835762,
      "num_tokens": 222406.0,
      "step": 110
    },
    {
      "entropy": 10.49779634475708,
      "epoch": 0.010049812112208336,
      "grad_norm": 1.671875,
      "learning_rate": 5.7e-05,
      "loss": 9.2158,
      "mean_token_accuracy": 0.06427372209727764,
      "num_tokens": 232139.0,
      "step": 115
    },
    {
      "entropy": 10.439636135101319,
      "epoch": 0.010486760464913047,
      "grad_norm": 1.40625,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 9.1331,
      "mean_token_accuracy": 0.06220344454050064,
      "num_tokens": 242438.0,
      "step": 120
    },
    {
      "entropy": 10.37880210876465,
      "epoch": 0.010923708817617757,
      "grad_norm": 1.4375,
      "learning_rate": 6.2e-05,
      "loss": 9.0192,
      "mean_token_accuracy": 0.06697547994554043,
      "num_tokens": 253082.0,
      "step": 125
    },
    {
      "entropy": 10.290444278717041,
      "epoch": 0.011360657170322468,
      "grad_norm": 1.4140625,
      "learning_rate": 6.450000000000001e-05,
      "loss": 8.9381,
      "mean_token_accuracy": 0.07001210227608681,
      "num_tokens": 264067.0,
      "step": 130
    },
    {
      "entropy": 10.230575561523438,
      "epoch": 0.011797605523027178,
      "grad_norm": 1.34375,
      "learning_rate": 6.7e-05,
      "loss": 8.8679,
      "mean_token_accuracy": 0.06120241619646549,
      "num_tokens": 274571.0,
      "step": 135
    },
    {
      "entropy": 10.212278461456298,
      "epoch": 0.012234553875731888,
      "grad_norm": 1.40625,
      "learning_rate": 6.950000000000001e-05,
      "loss": 8.7477,
      "mean_token_accuracy": 0.06530353389680385,
      "num_tokens": 284561.0,
      "step": 140
    },
    {
      "entropy": 10.070102214813232,
      "epoch": 0.012671502228436599,
      "grad_norm": 1.1796875,
      "learning_rate": 7.2e-05,
      "loss": 8.6078,
      "mean_token_accuracy": 0.06655799858272075,
      "num_tokens": 295366.0,
      "step": 145
    },
    {
      "entropy": 9.980097579956055,
      "epoch": 0.01310845058114131,
      "grad_norm": 1.0546875,
      "learning_rate": 7.45e-05,
      "loss": 8.6389,
      "mean_token_accuracy": 0.06603652313351631,
      "num_tokens": 305423.0,
      "step": 150
    },
    {
      "entropy": 9.907401084899902,
      "epoch": 0.01354539893384602,
      "grad_norm": 1.046875,
      "learning_rate": 7.7e-05,
      "loss": 8.4744,
      "mean_token_accuracy": 0.06612432450056076,
      "num_tokens": 315182.0,
      "step": 155
    },
    {
      "entropy": 9.672210597991944,
      "epoch": 0.01398234728655073,
      "grad_norm": 1.0,
      "learning_rate": 7.950000000000001e-05,
      "loss": 8.4101,
      "mean_token_accuracy": 0.07041444033384323,
      "num_tokens": 324780.0,
      "step": 160
    },
    {
      "entropy": 9.59818925857544,
      "epoch": 0.01441929563925544,
      "grad_norm": 0.8515625,
      "learning_rate": 8.2e-05,
      "loss": 8.3853,
      "mean_token_accuracy": 0.06835086569190026,
      "num_tokens": 335104.0,
      "step": 165
    },
    {
      "entropy": 9.408317279815673,
      "epoch": 0.014856243991960151,
      "grad_norm": 0.99609375,
      "learning_rate": 8.450000000000001e-05,
      "loss": 8.3064,
      "mean_token_accuracy": 0.06708988398313523,
      "num_tokens": 345073.0,
      "step": 170
    },
    {
      "entropy": 9.27857847213745,
      "epoch": 0.015293192344664861,
      "grad_norm": 0.87109375,
      "learning_rate": 8.7e-05,
      "loss": 8.2728,
      "mean_token_accuracy": 0.06384357884526252,
      "num_tokens": 354342.0,
      "step": 175
    },
    {
      "entropy": 9.09966974258423,
      "epoch": 0.015730140697369572,
      "grad_norm": 0.8203125,
      "learning_rate": 8.95e-05,
      "loss": 8.1513,
      "mean_token_accuracy": 0.07314991652965545,
      "num_tokens": 364441.0,
      "step": 180
    },
    {
      "entropy": 8.942261791229248,
      "epoch": 0.01616708905007428,
      "grad_norm": 0.7421875,
      "learning_rate": 9.2e-05,
      "loss": 8.1527,
      "mean_token_accuracy": 0.06924629844725132,
      "num_tokens": 373918.0,
      "step": 185
    },
    {
      "entropy": 8.88406400680542,
      "epoch": 0.016604037402778993,
      "grad_norm": 0.75,
      "learning_rate": 9.45e-05,
      "loss": 8.1632,
      "mean_token_accuracy": 0.0712675966322422,
      "num_tokens": 383273.0,
      "step": 190
    },
    {
      "entropy": 8.708986186981202,
      "epoch": 0.0170409857554837,
      "grad_norm": 0.796875,
      "learning_rate": 9.7e-05,
      "loss": 8.0308,
      "mean_token_accuracy": 0.08221853896975517,
      "num_tokens": 393262.0,
      "step": 195
    },
    {
      "entropy": 8.639312553405762,
      "epoch": 0.017477934108188414,
      "grad_norm": 0.73046875,
      "learning_rate": 9.95e-05,
      "loss": 8.1392,
      "mean_token_accuracy": 0.06822756379842758,
      "num_tokens": 402389.0,
      "step": 200
    },
    {
      "entropy": 8.639290618896485,
      "epoch": 0.017914882460893122,
      "grad_norm": 0.77734375,
      "learning_rate": 0.000102,
      "loss": 8.0721,
      "mean_token_accuracy": 0.06872168891131877,
      "num_tokens": 412997.0,
      "step": 205
    },
    {
      "entropy": 8.622207546234131,
      "epoch": 0.018351830813597834,
      "grad_norm": 0.8203125,
      "learning_rate": 0.00010449999999999999,
      "loss": 8.0805,
      "mean_token_accuracy": 0.07161624394357205,
      "num_tokens": 422222.0,
      "step": 210
    },
    {
      "entropy": 8.481709098815918,
      "epoch": 0.018788779166302543,
      "grad_norm": 0.71875,
      "learning_rate": 0.000107,
      "loss": 8.0159,
      "mean_token_accuracy": 0.07780317105352878,
      "num_tokens": 432539.0,
      "step": 215
    },
    {
      "entropy": 8.539759254455566,
      "epoch": 0.019225727519007252,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0001095,
      "loss": 8.018,
      "mean_token_accuracy": 0.07488755434751511,
      "num_tokens": 442311.0,
      "step": 220
    },
    {
      "entropy": 8.473109245300293,
      "epoch": 0.019662675871711964,
      "grad_norm": 0.7421875,
      "learning_rate": 0.000112,
      "loss": 8.0033,
      "mean_token_accuracy": 0.07176320515573024,
      "num_tokens": 452727.0,
      "step": 225
    },
    {
      "entropy": 8.49320011138916,
      "epoch": 0.020099624224416673,
      "grad_norm": 0.84375,
      "learning_rate": 0.0001145,
      "loss": 8.0269,
      "mean_token_accuracy": 0.0723928865045309,
      "num_tokens": 462461.0,
      "step": 230
    },
    {
      "entropy": 8.517394256591796,
      "epoch": 0.020536572577121385,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00011700000000000001,
      "loss": 7.9988,
      "mean_token_accuracy": 0.07492853850126266,
      "num_tokens": 472362.0,
      "step": 235
    },
    {
      "entropy": 8.411276721954346,
      "epoch": 0.020973520929826094,
      "grad_norm": 0.79296875,
      "learning_rate": 0.00011949999999999999,
      "loss": 7.9901,
      "mean_token_accuracy": 0.07653542906045914,
      "num_tokens": 482718.0,
      "step": 240
    },
    {
      "entropy": 8.451101970672607,
      "epoch": 0.021410469282530806,
      "grad_norm": 0.76953125,
      "learning_rate": 0.000122,
      "loss": 7.9973,
      "mean_token_accuracy": 0.07303892523050308,
      "num_tokens": 492947.0,
      "step": 245
    },
    {
      "entropy": 8.451656246185303,
      "epoch": 0.021847417635235514,
      "grad_norm": 0.875,
      "learning_rate": 0.0001245,
      "loss": 8.0293,
      "mean_token_accuracy": 0.07184793651103974,
      "num_tokens": 502957.0,
      "step": 250
    },
    {
      "entropy": 8.451603507995605,
      "epoch": 0.022284365987940227,
      "grad_norm": 0.90625,
      "learning_rate": 0.000127,
      "loss": 7.9786,
      "mean_token_accuracy": 0.07364074736833573,
      "num_tokens": 513220.0,
      "step": 255
    },
    {
      "entropy": 8.423027420043946,
      "epoch": 0.022721314340644935,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0001295,
      "loss": 7.927,
      "mean_token_accuracy": 0.07086114585399628,
      "num_tokens": 523217.0,
      "step": 260
    },
    {
      "entropy": 8.423391246795655,
      "epoch": 0.023158262693349647,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000132,
      "loss": 7.9515,
      "mean_token_accuracy": 0.0709399688988924,
      "num_tokens": 533789.0,
      "step": 265
    },
    {
      "entropy": 8.354116344451905,
      "epoch": 0.023595211046054356,
      "grad_norm": 0.80078125,
      "learning_rate": 0.00013450000000000002,
      "loss": 7.9324,
      "mean_token_accuracy": 0.07723044976592064,
      "num_tokens": 544146.0,
      "step": 270
    },
    {
      "entropy": 8.401555442810059,
      "epoch": 0.02403215939875907,
      "grad_norm": 0.84375,
      "learning_rate": 0.00013700000000000002,
      "loss": 7.9685,
      "mean_token_accuracy": 0.07366384789347649,
      "num_tokens": 554215.0,
      "step": 275
    },
    {
      "entropy": 8.307286357879638,
      "epoch": 0.024469107751463777,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0001395,
      "loss": 7.9334,
      "mean_token_accuracy": 0.0711975559592247,
      "num_tokens": 564587.0,
      "step": 280
    },
    {
      "entropy": 8.396804809570312,
      "epoch": 0.024906056104168486,
      "grad_norm": 0.73046875,
      "learning_rate": 0.00014199999999999998,
      "loss": 7.9794,
      "mean_token_accuracy": 0.07162599861621857,
      "num_tokens": 574819.0,
      "step": 285
    },
    {
      "entropy": 8.358118438720703,
      "epoch": 0.025343004456873198,
      "grad_norm": 0.796875,
      "learning_rate": 0.0001445,
      "loss": 7.8989,
      "mean_token_accuracy": 0.07322670742869378,
      "num_tokens": 585178.0,
      "step": 290
    },
    {
      "entropy": 8.294723510742188,
      "epoch": 0.025779952809577907,
      "grad_norm": 0.90625,
      "learning_rate": 0.000147,
      "loss": 7.8845,
      "mean_token_accuracy": 0.07678839713335037,
      "num_tokens": 595086.0,
      "step": 295
    },
    {
      "entropy": 8.285739517211914,
      "epoch": 0.02621690116228262,
      "grad_norm": 0.8125,
      "learning_rate": 0.0001495,
      "loss": 7.89,
      "mean_token_accuracy": 0.08111602552235127,
      "num_tokens": 604942.0,
      "step": 300
    },
    {
      "entropy": 8.310109424591065,
      "epoch": 0.026653849514987327,
      "grad_norm": 0.765625,
      "learning_rate": 0.000152,
      "loss": 7.874,
      "mean_token_accuracy": 0.07880265414714813,
      "num_tokens": 616082.0,
      "step": 305
    },
    {
      "entropy": 8.292809677124023,
      "epoch": 0.02709079786769204,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00015450000000000001,
      "loss": 7.9016,
      "mean_token_accuracy": 0.07674356997013092,
      "num_tokens": 626375.0,
      "step": 310
    },
    {
      "entropy": 8.272511959075928,
      "epoch": 0.027527746220396748,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000157,
      "loss": 7.8627,
      "mean_token_accuracy": 0.07318086177110672,
      "num_tokens": 637563.0,
      "step": 315
    },
    {
      "entropy": 8.222749423980712,
      "epoch": 0.02796469457310146,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0001595,
      "loss": 7.8692,
      "mean_token_accuracy": 0.07608433291316033,
      "num_tokens": 647256.0,
      "step": 320
    },
    {
      "entropy": 8.308216381072999,
      "epoch": 0.02840164292580617,
      "grad_norm": 0.75,
      "learning_rate": 0.000162,
      "loss": 7.9245,
      "mean_token_accuracy": 0.0743330918252468,
      "num_tokens": 657247.0,
      "step": 325
    },
    {
      "entropy": 8.2763596534729,
      "epoch": 0.02883859127851088,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00016450000000000001,
      "loss": 7.8367,
      "mean_token_accuracy": 0.07250913083553315,
      "num_tokens": 667613.0,
      "step": 330
    },
    {
      "entropy": 8.169933319091797,
      "epoch": 0.02927553963121559,
      "grad_norm": 1.0625,
      "learning_rate": 0.00016700000000000002,
      "loss": 7.8732,
      "mean_token_accuracy": 0.07495946660637856,
      "num_tokens": 678048.0,
      "step": 335
    },
    {
      "entropy": 8.294293117523193,
      "epoch": 0.029712487983920302,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00016950000000000003,
      "loss": 7.865,
      "mean_token_accuracy": 0.07813518717885018,
      "num_tokens": 688468.0,
      "step": 340
    },
    {
      "entropy": 8.193762683868409,
      "epoch": 0.03014943633662501,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00017199999999999998,
      "loss": 7.84,
      "mean_token_accuracy": 0.07896131426095962,
      "num_tokens": 699809.0,
      "step": 345
    },
    {
      "entropy": 8.181721496582032,
      "epoch": 0.030586384689329723,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00017449999999999999,
      "loss": 7.7969,
      "mean_token_accuracy": 0.08212044797837734,
      "num_tokens": 709227.0,
      "step": 350
    },
    {
      "entropy": 8.286689472198486,
      "epoch": 0.03102333304203443,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000177,
      "loss": 7.904,
      "mean_token_accuracy": 0.0743263080716133,
      "num_tokens": 720178.0,
      "step": 355
    },
    {
      "entropy": 8.163193130493164,
      "epoch": 0.031460281394739144,
      "grad_norm": 0.78125,
      "learning_rate": 0.0001795,
      "loss": 7.8202,
      "mean_token_accuracy": 0.0798851165920496,
      "num_tokens": 730514.0,
      "step": 360
    },
    {
      "entropy": 8.259191036224365,
      "epoch": 0.03189722974744385,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000182,
      "loss": 7.8692,
      "mean_token_accuracy": 0.07668052241206169,
      "num_tokens": 741345.0,
      "step": 365
    },
    {
      "entropy": 8.17900333404541,
      "epoch": 0.03233417810014856,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0001845,
      "loss": 7.7991,
      "mean_token_accuracy": 0.08136411309242249,
      "num_tokens": 751163.0,
      "step": 370
    },
    {
      "entropy": 8.175147342681885,
      "epoch": 0.03277112645285327,
      "grad_norm": 0.81640625,
      "learning_rate": 0.000187,
      "loss": 7.7993,
      "mean_token_accuracy": 0.0767936147749424,
      "num_tokens": 762090.0,
      "step": 375
    },
    {
      "entropy": 8.196157550811767,
      "epoch": 0.033208074805557986,
      "grad_norm": 1.125,
      "learning_rate": 0.0001895,
      "loss": 7.7674,
      "mean_token_accuracy": 0.08601637631654739,
      "num_tokens": 772049.0,
      "step": 380
    },
    {
      "entropy": 8.117916297912597,
      "epoch": 0.03364502315826269,
      "grad_norm": 1.109375,
      "learning_rate": 0.000192,
      "loss": 7.7365,
      "mean_token_accuracy": 0.08381113037467003,
      "num_tokens": 781280.0,
      "step": 385
    },
    {
      "entropy": 8.152852725982665,
      "epoch": 0.0340819715109674,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0001945,
      "loss": 7.8156,
      "mean_token_accuracy": 0.07903518378734589,
      "num_tokens": 791642.0,
      "step": 390
    },
    {
      "entropy": 8.175984954833984,
      "epoch": 0.034518919863672115,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00019700000000000002,
      "loss": 7.7973,
      "mean_token_accuracy": 0.07584034129977227,
      "num_tokens": 801224.0,
      "step": 395
    },
    {
      "entropy": 8.157032155990601,
      "epoch": 0.03495586821637683,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00019950000000000002,
      "loss": 7.7379,
      "mean_token_accuracy": 0.08061860464513301,
      "num_tokens": 810727.0,
      "step": 400
    },
    {
      "entropy": 8.220977401733398,
      "epoch": 0.03539281656908153,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000202,
      "loss": 7.8159,
      "mean_token_accuracy": 0.07617722004652024,
      "num_tokens": 819943.0,
      "step": 405
    },
    {
      "entropy": 8.027306842803956,
      "epoch": 0.035829764921786245,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00020449999999999998,
      "loss": 7.744,
      "mean_token_accuracy": 0.07786332815885544,
      "num_tokens": 829474.0,
      "step": 410
    },
    {
      "entropy": 8.135616159439087,
      "epoch": 0.03626671327449096,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000207,
      "loss": 7.7154,
      "mean_token_accuracy": 0.07687773257493973,
      "num_tokens": 839288.0,
      "step": 415
    },
    {
      "entropy": 8.123063278198241,
      "epoch": 0.03670366162719567,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0002095,
      "loss": 7.7375,
      "mean_token_accuracy": 0.08089107796549797,
      "num_tokens": 850064.0,
      "step": 420
    },
    {
      "entropy": 8.010475349426269,
      "epoch": 0.037140609979900374,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000212,
      "loss": 7.7137,
      "mean_token_accuracy": 0.08113521486520767,
      "num_tokens": 860732.0,
      "step": 425
    },
    {
      "entropy": 8.06641664505005,
      "epoch": 0.037577558332605086,
      "grad_norm": 0.9375,
      "learning_rate": 0.0002145,
      "loss": 7.7097,
      "mean_token_accuracy": 0.08033490926027298,
      "num_tokens": 870076.0,
      "step": 430
    },
    {
      "entropy": 7.975166702270508,
      "epoch": 0.0380145066853098,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00021700000000000002,
      "loss": 7.6087,
      "mean_token_accuracy": 0.08692155629396439,
      "num_tokens": 880201.0,
      "step": 435
    },
    {
      "entropy": 8.02914743423462,
      "epoch": 0.038451455038014504,
      "grad_norm": 0.78125,
      "learning_rate": 0.0002195,
      "loss": 7.6725,
      "mean_token_accuracy": 0.08405178114771843,
      "num_tokens": 890344.0,
      "step": 440
    },
    {
      "entropy": 8.042040014266968,
      "epoch": 0.038888403390719216,
      "grad_norm": 0.921875,
      "learning_rate": 0.000222,
      "loss": 7.6622,
      "mean_token_accuracy": 0.0840606488287449,
      "num_tokens": 900443.0,
      "step": 445
    },
    {
      "entropy": 8.088546133041381,
      "epoch": 0.03932535174342393,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0002245,
      "loss": 7.6868,
      "mean_token_accuracy": 0.07948172986507415,
      "num_tokens": 910284.0,
      "step": 450
    },
    {
      "entropy": 8.075697088241578,
      "epoch": 0.03976230009612864,
      "grad_norm": 0.984375,
      "learning_rate": 0.00022700000000000002,
      "loss": 7.7751,
      "mean_token_accuracy": 0.07789854928851128,
      "num_tokens": 920986.0,
      "step": 455
    },
    {
      "entropy": 7.980392408370972,
      "epoch": 0.040199248448833345,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00022950000000000002,
      "loss": 7.6705,
      "mean_token_accuracy": 0.08652443066239357,
      "num_tokens": 930481.0,
      "step": 460
    },
    {
      "entropy": 8.108790636062622,
      "epoch": 0.04063619680153806,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00023200000000000003,
      "loss": 7.6989,
      "mean_token_accuracy": 0.08597837761044502,
      "num_tokens": 940554.0,
      "step": 465
    },
    {
      "entropy": 7.986668729782105,
      "epoch": 0.04107314515424277,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00023449999999999998,
      "loss": 7.6399,
      "mean_token_accuracy": 0.08571613058447838,
      "num_tokens": 950005.0,
      "step": 470
    },
    {
      "entropy": 7.950281715393066,
      "epoch": 0.04151009350694748,
      "grad_norm": 0.8125,
      "learning_rate": 0.000237,
      "loss": 7.6361,
      "mean_token_accuracy": 0.0789151020348072,
      "num_tokens": 959414.0,
      "step": 475
    },
    {
      "entropy": 8.029509019851684,
      "epoch": 0.04194704185965219,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0002395,
      "loss": 7.6712,
      "mean_token_accuracy": 0.07935962677001954,
      "num_tokens": 969352.0,
      "step": 480
    },
    {
      "entropy": 8.051045751571655,
      "epoch": 0.0423839902123569,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000242,
      "loss": 7.6912,
      "mean_token_accuracy": 0.07945066690444946,
      "num_tokens": 980294.0,
      "step": 485
    },
    {
      "entropy": 7.899024248123169,
      "epoch": 0.04282093856506161,
      "grad_norm": 0.921875,
      "learning_rate": 0.0002445,
      "loss": 7.5774,
      "mean_token_accuracy": 0.08797174990177155,
      "num_tokens": 990453.0,
      "step": 490
    },
    {
      "entropy": 8.07900128364563,
      "epoch": 0.04325788691776632,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000247,
      "loss": 7.5947,
      "mean_token_accuracy": 0.08882696852087975,
      "num_tokens": 1000509.0,
      "step": 495
    },
    {
      "entropy": 7.993640470504761,
      "epoch": 0.04369483527047103,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002495,
      "loss": 7.7357,
      "mean_token_accuracy": 0.07507672794163227,
      "num_tokens": 1011747.0,
      "step": 500
    },
    {
      "entropy": 7.9472064018249515,
      "epoch": 0.04413178362317574,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000252,
      "loss": 7.6713,
      "mean_token_accuracy": 0.08197397291660309,
      "num_tokens": 1021511.0,
      "step": 505
    },
    {
      "entropy": 8.02770299911499,
      "epoch": 0.04456873197588045,
      "grad_norm": 0.90625,
      "learning_rate": 0.0002545,
      "loss": 7.6761,
      "mean_token_accuracy": 0.0799028217792511,
      "num_tokens": 1032116.0,
      "step": 510
    },
    {
      "entropy": 7.903038883209229,
      "epoch": 0.04500568032858516,
      "grad_norm": 1.015625,
      "learning_rate": 0.000257,
      "loss": 7.5751,
      "mean_token_accuracy": 0.09080960974097252,
      "num_tokens": 1041356.0,
      "step": 515
    },
    {
      "entropy": 7.906881475448609,
      "epoch": 0.04544262868128987,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002595,
      "loss": 7.6368,
      "mean_token_accuracy": 0.08153252303600311,
      "num_tokens": 1051314.0,
      "step": 520
    },
    {
      "entropy": 8.004989957809448,
      "epoch": 0.04587957703399458,
      "grad_norm": 0.859375,
      "learning_rate": 0.000262,
      "loss": 7.6159,
      "mean_token_accuracy": 0.08653398081660271,
      "num_tokens": 1060857.0,
      "step": 525
    },
    {
      "entropy": 7.971346616744995,
      "epoch": 0.046316525386699295,
      "grad_norm": 0.78515625,
      "learning_rate": 0.00026450000000000003,
      "loss": 7.624,
      "mean_token_accuracy": 0.08459379523992538,
      "num_tokens": 1071744.0,
      "step": 530
    },
    {
      "entropy": 7.969670820236206,
      "epoch": 0.046753473739404,
      "grad_norm": 1.046875,
      "learning_rate": 0.00026700000000000004,
      "loss": 7.5816,
      "mean_token_accuracy": 0.08811679482460022,
      "num_tokens": 1080777.0,
      "step": 535
    },
    {
      "entropy": 7.867853498458862,
      "epoch": 0.04719042209210871,
      "grad_norm": 1.09375,
      "learning_rate": 0.00026950000000000005,
      "loss": 7.5781,
      "mean_token_accuracy": 0.0892080545425415,
      "num_tokens": 1090951.0,
      "step": 540
    },
    {
      "entropy": 7.928713369369507,
      "epoch": 0.047627370444813424,
      "grad_norm": 1.09375,
      "learning_rate": 0.00027200000000000005,
      "loss": 7.5405,
      "mean_token_accuracy": 0.08424938097596169,
      "num_tokens": 1101667.0,
      "step": 545
    },
    {
      "entropy": 7.9725940227508545,
      "epoch": 0.04806431879751814,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0002745,
      "loss": 7.6535,
      "mean_token_accuracy": 0.0810074783861637,
      "num_tokens": 1112652.0,
      "step": 550
    },
    {
      "entropy": 7.903535270690918,
      "epoch": 0.04850126715022284,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000277,
      "loss": 7.6086,
      "mean_token_accuracy": 0.08384414836764335,
      "num_tokens": 1122518.0,
      "step": 555
    },
    {
      "entropy": 7.999388360977173,
      "epoch": 0.048938215502927554,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002795,
      "loss": 7.6081,
      "mean_token_accuracy": 0.08308766409754753,
      "num_tokens": 1132867.0,
      "step": 560
    },
    {
      "entropy": 7.961488819122314,
      "epoch": 0.049375163855632266,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00028199999999999997,
      "loss": 7.6766,
      "mean_token_accuracy": 0.07999438382685184,
      "num_tokens": 1142524.0,
      "step": 565
    },
    {
      "entropy": 8.025870513916015,
      "epoch": 0.04981211220833697,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002845,
      "loss": 7.5489,
      "mean_token_accuracy": 0.08902617059648037,
      "num_tokens": 1152983.0,
      "step": 570
    },
    {
      "entropy": 7.944514799118042,
      "epoch": 0.050249060561041684,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000287,
      "loss": 7.6243,
      "mean_token_accuracy": 0.08356612995266914,
      "num_tokens": 1164271.0,
      "step": 575
    },
    {
      "entropy": 7.8507874488830565,
      "epoch": 0.050686008913746396,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002895,
      "loss": 7.4777,
      "mean_token_accuracy": 0.08888518437743187,
      "num_tokens": 1173567.0,
      "step": 580
    },
    {
      "entropy": 7.87242693901062,
      "epoch": 0.05112295726645111,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000292,
      "loss": 7.5272,
      "mean_token_accuracy": 0.08922602906823159,
      "num_tokens": 1183619.0,
      "step": 585
    },
    {
      "entropy": 7.838789176940918,
      "epoch": 0.05155990561915581,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002945,
      "loss": 7.4864,
      "mean_token_accuracy": 0.08544198125600815,
      "num_tokens": 1193922.0,
      "step": 590
    },
    {
      "entropy": 7.948909044265747,
      "epoch": 0.051996853971860525,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000297,
      "loss": 7.5839,
      "mean_token_accuracy": 0.08625762760639191,
      "num_tokens": 1203665.0,
      "step": 595
    },
    {
      "entropy": 7.866623973846435,
      "epoch": 0.05243380232456524,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0002995,
      "loss": 7.5713,
      "mean_token_accuracy": 0.0907437652349472,
      "num_tokens": 1214119.0,
      "step": 600
    },
    {
      "entropy": 7.910204887390137,
      "epoch": 0.05287075067726995,
      "grad_norm": 1.15625,
      "learning_rate": 0.000302,
      "loss": 7.5954,
      "mean_token_accuracy": 0.08006279319524764,
      "num_tokens": 1225275.0,
      "step": 605
    },
    {
      "entropy": 7.817413377761841,
      "epoch": 0.053307699029974655,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003045,
      "loss": 7.4647,
      "mean_token_accuracy": 0.08570721000432968,
      "num_tokens": 1235199.0,
      "step": 610
    },
    {
      "entropy": 7.833220481872559,
      "epoch": 0.05374464738267937,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000307,
      "loss": 7.4878,
      "mean_token_accuracy": 0.08858786523342133,
      "num_tokens": 1245648.0,
      "step": 615
    },
    {
      "entropy": 7.811017656326294,
      "epoch": 0.05418159573538408,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003095,
      "loss": 7.4341,
      "mean_token_accuracy": 0.09172307997941971,
      "num_tokens": 1254793.0,
      "step": 620
    },
    {
      "entropy": 7.904883337020874,
      "epoch": 0.05461854408808879,
      "grad_norm": 0.90625,
      "learning_rate": 0.000312,
      "loss": 7.4908,
      "mean_token_accuracy": 0.09243529438972473,
      "num_tokens": 1264429.0,
      "step": 625
    },
    {
      "entropy": 7.71052098274231,
      "epoch": 0.055055492440793496,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003145,
      "loss": 7.4554,
      "mean_token_accuracy": 0.0870548702776432,
      "num_tokens": 1274449.0,
      "step": 630
    },
    {
      "entropy": 7.8729815006256105,
      "epoch": 0.05549244079349821,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000317,
      "loss": 7.5543,
      "mean_token_accuracy": 0.08215527608990669,
      "num_tokens": 1284987.0,
      "step": 635
    },
    {
      "entropy": 7.923571491241455,
      "epoch": 0.05592938914620292,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003195,
      "loss": 7.5696,
      "mean_token_accuracy": 0.08023575767874717,
      "num_tokens": 1295984.0,
      "step": 640
    },
    {
      "entropy": 7.778188037872314,
      "epoch": 0.056366337498907626,
      "grad_norm": 1.0,
      "learning_rate": 0.000322,
      "loss": 7.4517,
      "mean_token_accuracy": 0.08863993138074874,
      "num_tokens": 1305738.0,
      "step": 645
    },
    {
      "entropy": 7.7744215965271,
      "epoch": 0.05680328585161234,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032450000000000003,
      "loss": 7.4031,
      "mean_token_accuracy": 0.08852944150567055,
      "num_tokens": 1315691.0,
      "step": 650
    },
    {
      "entropy": 7.780942630767822,
      "epoch": 0.05724023420431705,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00032700000000000003,
      "loss": 7.4193,
      "mean_token_accuracy": 0.0951805904507637,
      "num_tokens": 1326195.0,
      "step": 655
    },
    {
      "entropy": 7.719406795501709,
      "epoch": 0.05767718255702176,
      "grad_norm": 1.0625,
      "learning_rate": 0.00032950000000000004,
      "loss": 7.4457,
      "mean_token_accuracy": 0.08878713175654411,
      "num_tokens": 1336916.0,
      "step": 660
    },
    {
      "entropy": 7.77421612739563,
      "epoch": 0.05811413090972647,
      "grad_norm": 1.0625,
      "learning_rate": 0.00033200000000000005,
      "loss": 7.4653,
      "mean_token_accuracy": 0.08794940188527108,
      "num_tokens": 1347386.0,
      "step": 665
    },
    {
      "entropy": 7.852012014389038,
      "epoch": 0.05855107926243118,
      "grad_norm": 1.171875,
      "learning_rate": 0.00033450000000000005,
      "loss": 7.4342,
      "mean_token_accuracy": 0.08515654727816582,
      "num_tokens": 1357959.0,
      "step": 670
    },
    {
      "entropy": 7.772040128707886,
      "epoch": 0.05898802761513589,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000337,
      "loss": 7.4385,
      "mean_token_accuracy": 0.08945433050394058,
      "num_tokens": 1368135.0,
      "step": 675
    },
    {
      "entropy": 7.773286199569702,
      "epoch": 0.059424975967840604,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003395,
      "loss": 7.4691,
      "mean_token_accuracy": 0.08570723086595536,
      "num_tokens": 1378702.0,
      "step": 680
    },
    {
      "entropy": 7.67702841758728,
      "epoch": 0.05986192432054531,
      "grad_norm": 1.0625,
      "learning_rate": 0.000342,
      "loss": 7.3237,
      "mean_token_accuracy": 0.09294968172907829,
      "num_tokens": 1387902.0,
      "step": 685
    },
    {
      "entropy": 7.789173412322998,
      "epoch": 0.06029887267325002,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00034449999999999997,
      "loss": 7.3411,
      "mean_token_accuracy": 0.09882994741201401,
      "num_tokens": 1397288.0,
      "step": 690
    },
    {
      "entropy": 7.668393468856811,
      "epoch": 0.060735821025954734,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000347,
      "loss": 7.3877,
      "mean_token_accuracy": 0.09437731429934501,
      "num_tokens": 1407416.0,
      "step": 695
    },
    {
      "entropy": 7.834095811843872,
      "epoch": 0.061172769378659446,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003495,
      "loss": 7.443,
      "mean_token_accuracy": 0.08851219043135643,
      "num_tokens": 1417733.0,
      "step": 700
    },
    {
      "entropy": 7.720336055755615,
      "epoch": 0.06160971773136415,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000352,
      "loss": 7.3365,
      "mean_token_accuracy": 0.09252023622393608,
      "num_tokens": 1426723.0,
      "step": 705
    },
    {
      "entropy": 7.701070976257324,
      "epoch": 0.06204666608406886,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003545,
      "loss": 7.3991,
      "mean_token_accuracy": 0.09218236356973648,
      "num_tokens": 1437290.0,
      "step": 710
    },
    {
      "entropy": 7.631638383865356,
      "epoch": 0.062483614436773575,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000357,
      "loss": 7.3274,
      "mean_token_accuracy": 0.09394551664590836,
      "num_tokens": 1446756.0,
      "step": 715
    },
    {
      "entropy": 7.714466619491577,
      "epoch": 0.06292056278947829,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003595,
      "loss": 7.4124,
      "mean_token_accuracy": 0.08511779010295868,
      "num_tokens": 1457074.0,
      "step": 720
    },
    {
      "entropy": 7.746994829177856,
      "epoch": 0.06335751114218299,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000362,
      "loss": 7.4538,
      "mean_token_accuracy": 0.08377371467649937,
      "num_tokens": 1467785.0,
      "step": 725
    },
    {
      "entropy": 7.678491926193237,
      "epoch": 0.0637944594948877,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0003645,
      "loss": 7.3443,
      "mean_token_accuracy": 0.09074902832508087,
      "num_tokens": 1477532.0,
      "step": 730
    },
    {
      "entropy": 7.6045910835266115,
      "epoch": 0.06423140784759242,
      "grad_norm": 1.0,
      "learning_rate": 0.000367,
      "loss": 7.2766,
      "mean_token_accuracy": 0.09526442065834999,
      "num_tokens": 1487102.0,
      "step": 735
    },
    {
      "entropy": 7.650178909301758,
      "epoch": 0.06466835620029712,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003695,
      "loss": 7.3517,
      "mean_token_accuracy": 0.08858921527862548,
      "num_tokens": 1497459.0,
      "step": 740
    },
    {
      "entropy": 7.654647779464722,
      "epoch": 0.06510530455300184,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000372,
      "loss": 7.3427,
      "mean_token_accuracy": 0.08997647315263749,
      "num_tokens": 1508756.0,
      "step": 745
    },
    {
      "entropy": 7.6317675590515135,
      "epoch": 0.06554225290570655,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003745,
      "loss": 7.2637,
      "mean_token_accuracy": 0.09751128628849984,
      "num_tokens": 1518735.0,
      "step": 750
    },
    {
      "entropy": 7.698947954177856,
      "epoch": 0.06597920125841125,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000377,
      "loss": 7.2964,
      "mean_token_accuracy": 0.09505418241024018,
      "num_tokens": 1528632.0,
      "step": 755
    },
    {
      "entropy": 7.500899124145508,
      "epoch": 0.06641614961111597,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003795,
      "loss": 7.295,
      "mean_token_accuracy": 0.09173346012830734,
      "num_tokens": 1538180.0,
      "step": 760
    },
    {
      "entropy": 7.681208324432373,
      "epoch": 0.06685309796382068,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000382,
      "loss": 7.3137,
      "mean_token_accuracy": 0.09170436710119248,
      "num_tokens": 1547788.0,
      "step": 765
    },
    {
      "entropy": 7.601843214035034,
      "epoch": 0.06729004631652538,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003845,
      "loss": 7.2644,
      "mean_token_accuracy": 0.09326299503445626,
      "num_tokens": 1558008.0,
      "step": 770
    },
    {
      "entropy": 7.5049737930297855,
      "epoch": 0.0677269946692301,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038700000000000003,
      "loss": 7.3246,
      "mean_token_accuracy": 0.08936878740787506,
      "num_tokens": 1567728.0,
      "step": 775
    },
    {
      "entropy": 7.653160285949707,
      "epoch": 0.0681639430219348,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00038950000000000003,
      "loss": 7.2736,
      "mean_token_accuracy": 0.09446480795741082,
      "num_tokens": 1578488.0,
      "step": 780
    },
    {
      "entropy": 7.6056342124938965,
      "epoch": 0.06860089137463951,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00039200000000000004,
      "loss": 7.2499,
      "mean_token_accuracy": 0.09750332310795784,
      "num_tokens": 1588070.0,
      "step": 785
    },
    {
      "entropy": 7.5712121486663815,
      "epoch": 0.06903783972734423,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00039450000000000005,
      "loss": 7.3194,
      "mean_token_accuracy": 0.09099319577217102,
      "num_tokens": 1598572.0,
      "step": 790
    },
    {
      "entropy": 7.714870977401733,
      "epoch": 0.06947478808004894,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039700000000000005,
      "loss": 7.3286,
      "mean_token_accuracy": 0.09093371108174324,
      "num_tokens": 1607982.0,
      "step": 795
    },
    {
      "entropy": 7.541324853897095,
      "epoch": 0.06991173643275365,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003995,
      "loss": 7.3322,
      "mean_token_accuracy": 0.0973043292760849,
      "num_tokens": 1618785.0,
      "step": 800
    },
    {
      "entropy": 7.621754837036133,
      "epoch": 0.07034868478545836,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000402,
      "loss": 7.2548,
      "mean_token_accuracy": 0.09845281392335892,
      "num_tokens": 1627988.0,
      "step": 805
    },
    {
      "entropy": 7.560144567489624,
      "epoch": 0.07078563313816306,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004045,
      "loss": 7.3355,
      "mean_token_accuracy": 0.08645970746874809,
      "num_tokens": 1638024.0,
      "step": 810
    },
    {
      "entropy": 7.686397504806519,
      "epoch": 0.07122258149086778,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040699999999999997,
      "loss": 7.3101,
      "mean_token_accuracy": 0.0890956200659275,
      "num_tokens": 1649215.0,
      "step": 815
    },
    {
      "entropy": 7.442864656448364,
      "epoch": 0.07165952984357249,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004095,
      "loss": 7.1978,
      "mean_token_accuracy": 0.09686766788363457,
      "num_tokens": 1659198.0,
      "step": 820
    },
    {
      "entropy": 7.594072866439819,
      "epoch": 0.0720964781962772,
      "grad_norm": 1.0625,
      "learning_rate": 0.000412,
      "loss": 7.2452,
      "mean_token_accuracy": 0.0892924964427948,
      "num_tokens": 1668972.0,
      "step": 825
    },
    {
      "entropy": 7.439077568054199,
      "epoch": 0.07253342654898191,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004145,
      "loss": 7.2232,
      "mean_token_accuracy": 0.09631062857806683,
      "num_tokens": 1679807.0,
      "step": 830
    },
    {
      "entropy": 7.600879383087158,
      "epoch": 0.07297037490168662,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000417,
      "loss": 7.2853,
      "mean_token_accuracy": 0.09247459471225739,
      "num_tokens": 1690567.0,
      "step": 835
    },
    {
      "entropy": 7.469448566436768,
      "epoch": 0.07340732325439134,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004195,
      "loss": 7.1689,
      "mean_token_accuracy": 0.09535667225718499,
      "num_tokens": 1700939.0,
      "step": 840
    },
    {
      "entropy": 7.388382434844971,
      "epoch": 0.07384427160709604,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000422,
      "loss": 7.2393,
      "mean_token_accuracy": 0.099522565305233,
      "num_tokens": 1711181.0,
      "step": 845
    },
    {
      "entropy": 7.600938224792481,
      "epoch": 0.07428121995980075,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004245,
      "loss": 7.2539,
      "mean_token_accuracy": 0.09072699621319771,
      "num_tokens": 1721392.0,
      "step": 850
    },
    {
      "entropy": 7.495168161392212,
      "epoch": 0.07471816831250547,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000427,
      "loss": 7.1767,
      "mean_token_accuracy": 0.09786228016018868,
      "num_tokens": 1731942.0,
      "step": 855
    },
    {
      "entropy": 7.469488048553467,
      "epoch": 0.07515511666521017,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004295,
      "loss": 7.2758,
      "mean_token_accuracy": 0.09157036766409873,
      "num_tokens": 1743409.0,
      "step": 860
    },
    {
      "entropy": 7.502536010742188,
      "epoch": 0.07559206501791488,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000432,
      "loss": 7.1633,
      "mean_token_accuracy": 0.09798691645264626,
      "num_tokens": 1753425.0,
      "step": 865
    },
    {
      "entropy": 7.4638996601104735,
      "epoch": 0.0760290133706196,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004345,
      "loss": 7.1934,
      "mean_token_accuracy": 0.09164758920669555,
      "num_tokens": 1762722.0,
      "step": 870
    },
    {
      "entropy": 7.642696189880371,
      "epoch": 0.0764659617233243,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000437,
      "loss": 7.2749,
      "mean_token_accuracy": 0.0860434666275978,
      "num_tokens": 1771895.0,
      "step": 875
    },
    {
      "entropy": 7.42297739982605,
      "epoch": 0.07690291007602901,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004395,
      "loss": 7.1698,
      "mean_token_accuracy": 0.09690217077732086,
      "num_tokens": 1782175.0,
      "step": 880
    },
    {
      "entropy": 7.478855276107788,
      "epoch": 0.07733985842873373,
      "grad_norm": 1.015625,
      "learning_rate": 0.000442,
      "loss": 7.2251,
      "mean_token_accuracy": 0.09283168092370034,
      "num_tokens": 1792409.0,
      "step": 885
    },
    {
      "entropy": 7.5595293045043945,
      "epoch": 0.07777680678143843,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004445,
      "loss": 7.2736,
      "mean_token_accuracy": 0.08964755833148956,
      "num_tokens": 1802721.0,
      "step": 890
    },
    {
      "entropy": 7.382868146896362,
      "epoch": 0.07821375513414315,
      "grad_norm": 1.03125,
      "learning_rate": 0.000447,
      "loss": 7.1126,
      "mean_token_accuracy": 0.09378557428717613,
      "num_tokens": 1812064.0,
      "step": 895
    },
    {
      "entropy": 7.530679559707641,
      "epoch": 0.07865070348684786,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044950000000000003,
      "loss": 7.215,
      "mean_token_accuracy": 0.08957840651273727,
      "num_tokens": 1822284.0,
      "step": 900
    },
    {
      "entropy": 7.388239574432373,
      "epoch": 0.07908765183955256,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045200000000000004,
      "loss": 7.1062,
      "mean_token_accuracy": 0.10064681470394135,
      "num_tokens": 1832150.0,
      "step": 905
    },
    {
      "entropy": 7.545283079147339,
      "epoch": 0.07952460019225728,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045450000000000004,
      "loss": 7.2334,
      "mean_token_accuracy": 0.0913564309477806,
      "num_tokens": 1842870.0,
      "step": 910
    },
    {
      "entropy": 7.420826816558838,
      "epoch": 0.07996154854496199,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045700000000000005,
      "loss": 7.0721,
      "mean_token_accuracy": 0.10421185195446014,
      "num_tokens": 1852771.0,
      "step": 915
    },
    {
      "entropy": 7.382372570037842,
      "epoch": 0.08039849689766669,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045950000000000006,
      "loss": 7.1698,
      "mean_token_accuracy": 0.09922162219882011,
      "num_tokens": 1863095.0,
      "step": 920
    },
    {
      "entropy": 7.418209648132324,
      "epoch": 0.08083544525037141,
      "grad_norm": 1.09375,
      "learning_rate": 0.000462,
      "loss": 7.171,
      "mean_token_accuracy": 0.09901941046118737,
      "num_tokens": 1873060.0,
      "step": 925
    },
    {
      "entropy": 7.4146593570709225,
      "epoch": 0.08127239360307612,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004645,
      "loss": 7.1089,
      "mean_token_accuracy": 0.10169215872883797,
      "num_tokens": 1882574.0,
      "step": 930
    },
    {
      "entropy": 7.437223148345947,
      "epoch": 0.08170934195578082,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000467,
      "loss": 7.1725,
      "mean_token_accuracy": 0.09495756924152374,
      "num_tokens": 1892227.0,
      "step": 935
    },
    {
      "entropy": 7.439269733428955,
      "epoch": 0.08214629030848554,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004695,
      "loss": 7.1934,
      "mean_token_accuracy": 0.09354586526751518,
      "num_tokens": 1902868.0,
      "step": 940
    },
    {
      "entropy": 7.387948656082154,
      "epoch": 0.08258323866119024,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000472,
      "loss": 7.1448,
      "mean_token_accuracy": 0.09095190614461898,
      "num_tokens": 1912781.0,
      "step": 945
    },
    {
      "entropy": 7.391385173797607,
      "epoch": 0.08302018701389496,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004745,
      "loss": 7.1389,
      "mean_token_accuracy": 0.0988696813583374,
      "num_tokens": 1923638.0,
      "step": 950
    },
    {
      "entropy": 7.388245868682861,
      "epoch": 0.08345713536659967,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000477,
      "loss": 7.1097,
      "mean_token_accuracy": 0.10090045481920243,
      "num_tokens": 1933498.0,
      "step": 955
    },
    {
      "entropy": 7.285632038116455,
      "epoch": 0.08389408371930437,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004795,
      "loss": 7.1227,
      "mean_token_accuracy": 0.09693519324064255,
      "num_tokens": 1943798.0,
      "step": 960
    },
    {
      "entropy": 7.4096606254577635,
      "epoch": 0.0843310320720091,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000482,
      "loss": 7.1564,
      "mean_token_accuracy": 0.09475521519780158,
      "num_tokens": 1954235.0,
      "step": 965
    },
    {
      "entropy": 7.368670797348022,
      "epoch": 0.0847679804247138,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004845,
      "loss": 7.152,
      "mean_token_accuracy": 0.09138476997613906,
      "num_tokens": 1964421.0,
      "step": 970
    },
    {
      "entropy": 7.36945948600769,
      "epoch": 0.0852049287774185,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000487,
      "loss": 7.1236,
      "mean_token_accuracy": 0.09725102558732032,
      "num_tokens": 1975862.0,
      "step": 975
    },
    {
      "entropy": 7.358737993240356,
      "epoch": 0.08564187713012322,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004895,
      "loss": 7.1205,
      "mean_token_accuracy": 0.09733417257666588,
      "num_tokens": 1985758.0,
      "step": 980
    },
    {
      "entropy": 7.362492322921753,
      "epoch": 0.08607882548282793,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000492,
      "loss": 7.0829,
      "mean_token_accuracy": 0.09430268853902816,
      "num_tokens": 1996904.0,
      "step": 985
    },
    {
      "entropy": 7.448752498626709,
      "epoch": 0.08651577383553263,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004945,
      "loss": 7.1749,
      "mean_token_accuracy": 0.09713461846113206,
      "num_tokens": 2006114.0,
      "step": 990
    },
    {
      "entropy": 7.365045785903931,
      "epoch": 0.08695272218823735,
      "grad_norm": 0.9375,
      "learning_rate": 0.000497,
      "loss": 7.1111,
      "mean_token_accuracy": 0.09818708300590515,
      "num_tokens": 2016349.0,
      "step": 995
    },
    {
      "entropy": 7.291970348358154,
      "epoch": 0.08738967054094206,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995,
      "loss": 7.061,
      "mean_token_accuracy": 0.09953764006495476,
      "num_tokens": 2026701.0,
      "step": 1000
    },
    {
      "entropy": 7.302675056457519,
      "epoch": 0.08782661889364678,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999999986190023,
      "loss": 7.0537,
      "mean_token_accuracy": 0.09898199439048767,
      "num_tokens": 2036950.0,
      "step": 1005
    },
    {
      "entropy": 7.409235525131225,
      "epoch": 0.08826356724635148,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999999930086991,
      "loss": 7.0489,
      "mean_token_accuracy": 0.10240035951137542,
      "num_tokens": 2047655.0,
      "step": 1010
    },
    {
      "entropy": 7.260290336608887,
      "epoch": 0.08870051559905619,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499999983082778,
      "loss": 7.0274,
      "mean_token_accuracy": 0.10356488525867462,
      "num_tokens": 2056800.0,
      "step": 1015
    },
    {
      "entropy": 7.276587057113647,
      "epoch": 0.0891374639517609,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999999688412394,
      "loss": 7.07,
      "mean_token_accuracy": 0.09646826907992363,
      "num_tokens": 2066862.0,
      "step": 1020
    },
    {
      "entropy": 7.327617454528808,
      "epoch": 0.08957441230446561,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999999502840836,
      "loss": 7.0896,
      "mean_token_accuracy": 0.10041970610618592,
      "num_tokens": 2076585.0,
      "step": 1025
    },
    {
      "entropy": 7.3417980670928955,
      "epoch": 0.09001136065717032,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999999274113107,
      "loss": 7.1548,
      "mean_token_accuracy": 0.09449673742055893,
      "num_tokens": 2087768.0,
      "step": 1030
    },
    {
      "entropy": 7.343121814727783,
      "epoch": 0.09044830900987504,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999999002229214,
      "loss": 7.0622,
      "mean_token_accuracy": 0.09806089103221893,
      "num_tokens": 2097027.0,
      "step": 1035
    },
    {
      "entropy": 7.191218137741089,
      "epoch": 0.09088525736257974,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999998687189161,
      "loss": 6.9901,
      "mean_token_accuracy": 0.10912425816059113,
      "num_tokens": 2107055.0,
      "step": 1040
    },
    {
      "entropy": 7.271266222000122,
      "epoch": 0.09132220571528446,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999998328992954,
      "loss": 6.9198,
      "mean_token_accuracy": 0.10159473046660424,
      "num_tokens": 2116771.0,
      "step": 1045
    },
    {
      "entropy": 7.177909517288208,
      "epoch": 0.09175915406798917,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00049999979276406,
      "loss": 7.0603,
      "mean_token_accuracy": 0.10223758444190026,
      "num_tokens": 2126881.0,
      "step": 1050
    },
    {
      "entropy": 7.254304838180542,
      "epoch": 0.09219610242069387,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999997483132108,
      "loss": 7.0248,
      "mean_token_accuracy": 0.101267521828413,
      "num_tokens": 2136964.0,
      "step": 1055
    },
    {
      "entropy": 7.332895040512085,
      "epoch": 0.09263305077339859,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999996995467483,
      "loss": 7.0764,
      "mean_token_accuracy": 0.09840327501296997,
      "num_tokens": 2147263.0,
      "step": 1060
    },
    {
      "entropy": 7.248973989486695,
      "epoch": 0.0930699991261033,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999996464646738,
      "loss": 7.089,
      "mean_token_accuracy": 0.09868241250514984,
      "num_tokens": 2157474.0,
      "step": 1065
    },
    {
      "entropy": 7.242979240417481,
      "epoch": 0.093506947478808,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999995890669882,
      "loss": 7.0217,
      "mean_token_accuracy": 0.10121526122093201,
      "num_tokens": 2167701.0,
      "step": 1070
    },
    {
      "entropy": 7.1951741695404055,
      "epoch": 0.09394389583151272,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999995273536925,
      "loss": 6.9862,
      "mean_token_accuracy": 0.10409378558397293,
      "num_tokens": 2178581.0,
      "step": 1075
    },
    {
      "entropy": 7.250149822235107,
      "epoch": 0.09438084418421742,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499999461324788,
      "loss": 6.9949,
      "mean_token_accuracy": 0.10199672728776932,
      "num_tokens": 2189529.0,
      "step": 1080
    },
    {
      "entropy": 7.254613780975342,
      "epoch": 0.09481779253692213,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999993909802759,
      "loss": 7.0541,
      "mean_token_accuracy": 0.09610985442996026,
      "num_tokens": 2199805.0,
      "step": 1085
    },
    {
      "entropy": 7.232923269271851,
      "epoch": 0.09525474088962685,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999993163201577,
      "loss": 7.0206,
      "mean_token_accuracy": 0.09527453482151031,
      "num_tokens": 2210108.0,
      "step": 1090
    },
    {
      "entropy": 7.358099699020386,
      "epoch": 0.09569168924233155,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999992373444347,
      "loss": 7.01,
      "mean_token_accuracy": 0.10141385495662689,
      "num_tokens": 2220325.0,
      "step": 1095
    },
    {
      "entropy": 7.133218908309937,
      "epoch": 0.09612863759503627,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999991540531084,
      "loss": 7.0098,
      "mean_token_accuracy": 0.0968212477862835,
      "num_tokens": 2230521.0,
      "step": 1100
    },
    {
      "entropy": 7.322919178009033,
      "epoch": 0.09656558594774098,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004999990664461804,
      "loss": 7.0377,
      "mean_token_accuracy": 0.09720605388283729,
      "num_tokens": 2240372.0,
      "step": 1105
    },
    {
      "entropy": 7.171909475326538,
      "epoch": 0.09700253430044568,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999989745236524,
      "loss": 6.992,
      "mean_token_accuracy": 0.10702305659651756,
      "num_tokens": 2250522.0,
      "step": 1110
    },
    {
      "entropy": 7.190585279464722,
      "epoch": 0.0974394826531504,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999988782855261,
      "loss": 6.9584,
      "mean_token_accuracy": 0.09914659708738327,
      "num_tokens": 2260142.0,
      "step": 1115
    },
    {
      "entropy": 7.227755498886109,
      "epoch": 0.09787643100585511,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999987777318035,
      "loss": 6.9664,
      "mean_token_accuracy": 0.10190813839435578,
      "num_tokens": 2270372.0,
      "step": 1120
    },
    {
      "entropy": 7.071720218658447,
      "epoch": 0.09831337935855981,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999986728624865,
      "loss": 6.9076,
      "mean_token_accuracy": 0.11096167787909508,
      "num_tokens": 2280517.0,
      "step": 1125
    },
    {
      "entropy": 7.284544038772583,
      "epoch": 0.09875032771126453,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499998563677577,
      "loss": 6.9999,
      "mean_token_accuracy": 0.09647511318325996,
      "num_tokens": 2290778.0,
      "step": 1130
    },
    {
      "entropy": 7.183425855636597,
      "epoch": 0.09918727606396924,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999984501770771,
      "loss": 7.0455,
      "mean_token_accuracy": 0.09870154932141303,
      "num_tokens": 2301692.0,
      "step": 1135
    },
    {
      "entropy": 7.21610689163208,
      "epoch": 0.09962422441667394,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999983323609891,
      "loss": 6.9337,
      "mean_token_accuracy": 0.10188546925783157,
      "num_tokens": 2312759.0,
      "step": 1140
    },
    {
      "entropy": 7.203960323333741,
      "epoch": 0.10006117276937866,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999982102293153,
      "loss": 6.9185,
      "mean_token_accuracy": 0.10292895063757897,
      "num_tokens": 2322016.0,
      "step": 1145
    },
    {
      "entropy": 7.224305295944214,
      "epoch": 0.10049812112208337,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999980837820578,
      "loss": 6.9399,
      "mean_token_accuracy": 0.10199728310108185,
      "num_tokens": 2332476.0,
      "step": 1150
    },
    {
      "entropy": 7.190159654617309,
      "epoch": 0.10093506947478809,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999979530192192,
      "loss": 6.9518,
      "mean_token_accuracy": 0.10138825252652169,
      "num_tokens": 2342914.0,
      "step": 1155
    },
    {
      "entropy": 7.146100616455078,
      "epoch": 0.10137201782749279,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999978179408019,
      "loss": 6.9927,
      "mean_token_accuracy": 0.0981225997209549,
      "num_tokens": 2353381.0,
      "step": 1160
    },
    {
      "entropy": 7.201415205001831,
      "epoch": 0.1018089661801975,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004999976785468085,
      "loss": 6.982,
      "mean_token_accuracy": 0.10512297451496125,
      "num_tokens": 2364207.0,
      "step": 1165
    },
    {
      "entropy": 7.082000207901001,
      "epoch": 0.10224591453290222,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999975348372418,
      "loss": 6.8996,
      "mean_token_accuracy": 0.1049137569963932,
      "num_tokens": 2374985.0,
      "step": 1170
    },
    {
      "entropy": 7.163199710845947,
      "epoch": 0.10268286288560692,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999973868121044,
      "loss": 6.9749,
      "mean_token_accuracy": 0.1049682468175888,
      "num_tokens": 2385684.0,
      "step": 1175
    },
    {
      "entropy": 7.2960728168487545,
      "epoch": 0.10311981123831163,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999972344713994,
      "loss": 6.8993,
      "mean_token_accuracy": 0.09997786954045296,
      "num_tokens": 2395177.0,
      "step": 1180
    },
    {
      "entropy": 7.023489665985108,
      "epoch": 0.10355675959101635,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999970778151295,
      "loss": 6.9248,
      "mean_token_accuracy": 0.109105683863163,
      "num_tokens": 2405154.0,
      "step": 1185
    },
    {
      "entropy": 7.2286418914794925,
      "epoch": 0.10399370794372105,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999969168432976,
      "loss": 6.9182,
      "mean_token_accuracy": 0.10396889075636864,
      "num_tokens": 2415818.0,
      "step": 1190
    },
    {
      "entropy": 7.086079740524292,
      "epoch": 0.10443065629642576,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999967515559071,
      "loss": 6.8955,
      "mean_token_accuracy": 0.1030263639986515,
      "num_tokens": 2425072.0,
      "step": 1195
    },
    {
      "entropy": 7.1413719177246096,
      "epoch": 0.10486760464913047,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499996581952961,
      "loss": 6.9562,
      "mean_token_accuracy": 0.10766926109790802,
      "num_tokens": 2435055.0,
      "step": 1200
    },
    {
      "entropy": 7.168883657455444,
      "epoch": 0.10530455300183518,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999964080344625,
      "loss": 6.8422,
      "mean_token_accuracy": 0.11093364953994751,
      "num_tokens": 2444434.0,
      "step": 1205
    },
    {
      "entropy": 7.101231050491333,
      "epoch": 0.1057415013545399,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999962298004151,
      "loss": 6.8734,
      "mean_token_accuracy": 0.10424663722515107,
      "num_tokens": 2453860.0,
      "step": 1210
    },
    {
      "entropy": 7.190755844116211,
      "epoch": 0.1061784497072446,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999960472508219,
      "loss": 6.9733,
      "mean_token_accuracy": 0.09948998987674713,
      "num_tokens": 2464235.0,
      "step": 1215
    },
    {
      "entropy": 7.079598331451416,
      "epoch": 0.10661539805994931,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999958603856869,
      "loss": 6.8763,
      "mean_token_accuracy": 0.10838079452514648,
      "num_tokens": 2474561.0,
      "step": 1220
    },
    {
      "entropy": 7.070672941207886,
      "epoch": 0.10705234641265403,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999956692050132,
      "loss": 6.8808,
      "mean_token_accuracy": 0.10815191268920898,
      "num_tokens": 2484574.0,
      "step": 1225
    },
    {
      "entropy": 7.15245943069458,
      "epoch": 0.10748929476535873,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999954737088048,
      "loss": 6.8478,
      "mean_token_accuracy": 0.11184538155794144,
      "num_tokens": 2494146.0,
      "step": 1230
    },
    {
      "entropy": 6.9708349227905275,
      "epoch": 0.10792624311806344,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999952738970652,
      "loss": 6.9382,
      "mean_token_accuracy": 0.10538864210247993,
      "num_tokens": 2504157.0,
      "step": 1235
    },
    {
      "entropy": 7.157891273498535,
      "epoch": 0.10836319147076816,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999950697697983,
      "loss": 6.8751,
      "mean_token_accuracy": 0.10475126504898072,
      "num_tokens": 2512979.0,
      "step": 1240
    },
    {
      "entropy": 7.113284254074097,
      "epoch": 0.10880013982347286,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999948613270081,
      "loss": 6.9312,
      "mean_token_accuracy": 0.10144072771072388,
      "num_tokens": 2523202.0,
      "step": 1245
    },
    {
      "entropy": 7.073151969909668,
      "epoch": 0.10923708817617758,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999946485686987,
      "loss": 6.8758,
      "mean_token_accuracy": 0.10646733790636062,
      "num_tokens": 2532790.0,
      "step": 1250
    },
    {
      "entropy": 7.101035165786743,
      "epoch": 0.10967403652888229,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000499994431494874,
      "loss": 6.8643,
      "mean_token_accuracy": 0.10860786586999893,
      "num_tokens": 2542337.0,
      "step": 1255
    },
    {
      "entropy": 7.028286504745483,
      "epoch": 0.11011098488158699,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999942101055381,
      "loss": 6.8175,
      "mean_token_accuracy": 0.10841023772954941,
      "num_tokens": 2552762.0,
      "step": 1260
    },
    {
      "entropy": 7.066199254989624,
      "epoch": 0.11054793323429171,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999939844006955,
      "loss": 6.858,
      "mean_token_accuracy": 0.10801806300878525,
      "num_tokens": 2563261.0,
      "step": 1265
    },
    {
      "entropy": 7.134337902069092,
      "epoch": 0.11098488158699642,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999937543803503,
      "loss": 6.8252,
      "mean_token_accuracy": 0.1146186038851738,
      "num_tokens": 2573089.0,
      "step": 1270
    },
    {
      "entropy": 7.0005029201507565,
      "epoch": 0.11142182993970112,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499993520044507,
      "loss": 6.9056,
      "mean_token_accuracy": 0.10176412537693977,
      "num_tokens": 2584358.0,
      "step": 1275
    },
    {
      "entropy": 7.139933633804321,
      "epoch": 0.11185877829240584,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999932813931702,
      "loss": 6.8709,
      "mean_token_accuracy": 0.1116573728621006,
      "num_tokens": 2594981.0,
      "step": 1280
    },
    {
      "entropy": 7.084459447860718,
      "epoch": 0.11229572664511055,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999930384263441,
      "loss": 6.8717,
      "mean_token_accuracy": 0.1069402813911438,
      "num_tokens": 2603508.0,
      "step": 1285
    },
    {
      "entropy": 7.084566974639893,
      "epoch": 0.11273267499781525,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499992791144034,
      "loss": 6.8535,
      "mean_token_accuracy": 0.10854770466685296,
      "num_tokens": 2613666.0,
      "step": 1290
    },
    {
      "entropy": 7.203876972198486,
      "epoch": 0.11316962335051997,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999925395462439,
      "loss": 6.9042,
      "mean_token_accuracy": 0.10472902953624726,
      "num_tokens": 2624441.0,
      "step": 1295
    },
    {
      "entropy": 7.066425085067749,
      "epoch": 0.11360657170322468,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999922836329793,
      "loss": 6.8576,
      "mean_token_accuracy": 0.10966517627239228,
      "num_tokens": 2635261.0,
      "step": 1300
    },
    {
      "entropy": 6.97562608718872,
      "epoch": 0.1140435200559294,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999920234042446,
      "loss": 6.8321,
      "mean_token_accuracy": 0.11323759481310844,
      "num_tokens": 2645493.0,
      "step": 1305
    },
    {
      "entropy": 7.0524732112884525,
      "epoch": 0.1144804684086341,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999917588600451,
      "loss": 6.7956,
      "mean_token_accuracy": 0.11606079787015915,
      "num_tokens": 2655430.0,
      "step": 1310
    },
    {
      "entropy": 7.048898983001709,
      "epoch": 0.1149174167613388,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999914900003857,
      "loss": 6.8529,
      "mean_token_accuracy": 0.10828904882073402,
      "num_tokens": 2665675.0,
      "step": 1315
    },
    {
      "entropy": 7.087590456008911,
      "epoch": 0.11535436511404352,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999912168252717,
      "loss": 6.8204,
      "mean_token_accuracy": 0.10795803815126419,
      "num_tokens": 2675962.0,
      "step": 1320
    },
    {
      "entropy": 7.021550798416138,
      "epoch": 0.11579131346674823,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999909393347083,
      "loss": 6.8477,
      "mean_token_accuracy": 0.10889611020684242,
      "num_tokens": 2687015.0,
      "step": 1325
    },
    {
      "entropy": 7.038705635070801,
      "epoch": 0.11622826181945294,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004999906575287006,
      "loss": 6.8956,
      "mean_token_accuracy": 0.10390607044100761,
      "num_tokens": 2698065.0,
      "step": 1330
    },
    {
      "entropy": 7.041507291793823,
      "epoch": 0.11666521017215765,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999903714072544,
      "loss": 6.7831,
      "mean_token_accuracy": 0.11800387650728225,
      "num_tokens": 2709046.0,
      "step": 1335
    },
    {
      "entropy": 7.065563631057739,
      "epoch": 0.11710215852486236,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999900809703749,
      "loss": 6.8565,
      "mean_token_accuracy": 0.11037417724728585,
      "num_tokens": 2719705.0,
      "step": 1340
    },
    {
      "entropy": 7.038755893707275,
      "epoch": 0.11753910687756706,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004999897862180678,
      "loss": 6.7328,
      "mean_token_accuracy": 0.11823359355330468,
      "num_tokens": 2729556.0,
      "step": 1345
    },
    {
      "entropy": 6.916531419754028,
      "epoch": 0.11797605523027178,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999894871503386,
      "loss": 6.8563,
      "mean_token_accuracy": 0.10672288462519645,
      "num_tokens": 2741144.0,
      "step": 1350
    },
    {
      "entropy": 7.106829738616943,
      "epoch": 0.11841300358297649,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999891837671932,
      "loss": 6.7757,
      "mean_token_accuracy": 0.11289330124855042,
      "num_tokens": 2751384.0,
      "step": 1355
    },
    {
      "entropy": 6.979836463928223,
      "epoch": 0.11884995193568121,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999888760686375,
      "loss": 6.7905,
      "mean_token_accuracy": 0.10638567060232162,
      "num_tokens": 2761044.0,
      "step": 1360
    },
    {
      "entropy": 6.934152507781983,
      "epoch": 0.11928690028838591,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499988564054677,
      "loss": 6.7447,
      "mean_token_accuracy": 0.11002607867121697,
      "num_tokens": 2770946.0,
      "step": 1365
    },
    {
      "entropy": 7.068941926956176,
      "epoch": 0.11972384864109062,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999882477253182,
      "loss": 6.8113,
      "mean_token_accuracy": 0.1063067264854908,
      "num_tokens": 2781247.0,
      "step": 1370
    },
    {
      "entropy": 6.94280161857605,
      "epoch": 0.12016079699379534,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999879270805667,
      "loss": 6.6546,
      "mean_token_accuracy": 0.12014054954051971,
      "num_tokens": 2792078.0,
      "step": 1375
    },
    {
      "entropy": 6.9622045993804935,
      "epoch": 0.12059774534650004,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999876021204291,
      "loss": 6.7449,
      "mean_token_accuracy": 0.11271467059850693,
      "num_tokens": 2802458.0,
      "step": 1380
    },
    {
      "entropy": 6.967449569702149,
      "epoch": 0.12103469369920475,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999872728449112,
      "loss": 6.886,
      "mean_token_accuracy": 0.10394820868968964,
      "num_tokens": 2813352.0,
      "step": 1385
    },
    {
      "entropy": 7.045512008666992,
      "epoch": 0.12147164205190947,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999869392540195,
      "loss": 6.7739,
      "mean_token_accuracy": 0.10725075080990791,
      "num_tokens": 2822954.0,
      "step": 1390
    },
    {
      "entropy": 7.02110652923584,
      "epoch": 0.12190859040461417,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999866013477605,
      "loss": 6.8006,
      "mean_token_accuracy": 0.10808226242661476,
      "num_tokens": 2833375.0,
      "step": 1395
    },
    {
      "entropy": 6.927406930923462,
      "epoch": 0.12234553875731889,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999862591261407,
      "loss": 6.7472,
      "mean_token_accuracy": 0.10812352672219276,
      "num_tokens": 2843103.0,
      "step": 1400
    },
    {
      "entropy": 7.004934597015381,
      "epoch": 0.1227824871100236,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999859125891664,
      "loss": 6.8174,
      "mean_token_accuracy": 0.10998742878437043,
      "num_tokens": 2853801.0,
      "step": 1405
    },
    {
      "entropy": 7.010592699050903,
      "epoch": 0.1232194354627283,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999855617368445,
      "loss": 6.7962,
      "mean_token_accuracy": 0.10651249438524246,
      "num_tokens": 2863851.0,
      "step": 1410
    },
    {
      "entropy": 6.922228670120239,
      "epoch": 0.12365638381543302,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999852065691816,
      "loss": 6.7389,
      "mean_token_accuracy": 0.11268523782491684,
      "num_tokens": 2875073.0,
      "step": 1415
    },
    {
      "entropy": 7.0324512958526615,
      "epoch": 0.12409333216813773,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999848470861846,
      "loss": 6.7683,
      "mean_token_accuracy": 0.11101210638880729,
      "num_tokens": 2885676.0,
      "step": 1420
    },
    {
      "entropy": 6.929794263839722,
      "epoch": 0.12453028052084243,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999844832878602,
      "loss": 6.7645,
      "mean_token_accuracy": 0.11524223163723946,
      "num_tokens": 2896663.0,
      "step": 1425
    },
    {
      "entropy": 7.003040027618408,
      "epoch": 0.12496722887354715,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999841151742156,
      "loss": 6.7606,
      "mean_token_accuracy": 0.11545668840408325,
      "num_tokens": 2907555.0,
      "step": 1430
    },
    {
      "entropy": 6.966159200668335,
      "epoch": 0.12540417722625186,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999837427452578,
      "loss": 6.7548,
      "mean_token_accuracy": 0.1095295563340187,
      "num_tokens": 2917173.0,
      "step": 1435
    },
    {
      "entropy": 6.940775394439697,
      "epoch": 0.12584112557895658,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499983366000994,
      "loss": 6.7623,
      "mean_token_accuracy": 0.1120285913348198,
      "num_tokens": 2928268.0,
      "step": 1440
    },
    {
      "entropy": 7.006380558013916,
      "epoch": 0.12627807393166127,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999829849414312,
      "loss": 6.7183,
      "mean_token_accuracy": 0.11839316189289092,
      "num_tokens": 2938203.0,
      "step": 1445
    },
    {
      "entropy": 6.905020713806152,
      "epoch": 0.12671502228436599,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499982599566577,
      "loss": 6.743,
      "mean_token_accuracy": 0.11780567914247513,
      "num_tokens": 2948094.0,
      "step": 1450
    },
    {
      "entropy": 6.976879024505616,
      "epoch": 0.1271519706370707,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999822098764385,
      "loss": 6.7402,
      "mean_token_accuracy": 0.10916817337274551,
      "num_tokens": 2957865.0,
      "step": 1455
    },
    {
      "entropy": 6.912422466278076,
      "epoch": 0.1275889189897754,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999818158710234,
      "loss": 6.6597,
      "mean_token_accuracy": 0.11891222149133682,
      "num_tokens": 2967880.0,
      "step": 1460
    },
    {
      "entropy": 6.932916593551636,
      "epoch": 0.12802586734248012,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999814175503392,
      "loss": 6.7246,
      "mean_token_accuracy": 0.11173751205205917,
      "num_tokens": 2979537.0,
      "step": 1465
    },
    {
      "entropy": 6.902735710144043,
      "epoch": 0.12846281569518483,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999810149143935,
      "loss": 6.7097,
      "mean_token_accuracy": 0.11276535987854004,
      "num_tokens": 2989700.0,
      "step": 1470
    },
    {
      "entropy": 6.83418345451355,
      "epoch": 0.12889976404788953,
      "grad_norm": 1.0,
      "learning_rate": 0.000499980607963194,
      "loss": 6.7178,
      "mean_token_accuracy": 0.11464745998382568,
      "num_tokens": 2999141.0,
      "step": 1475
    },
    {
      "entropy": 6.9692826747894285,
      "epoch": 0.12933671240059424,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999801966967486,
      "loss": 6.6815,
      "mean_token_accuracy": 0.11767361015081405,
      "num_tokens": 3008259.0,
      "step": 1480
    },
    {
      "entropy": 6.939387321472168,
      "epoch": 0.12977366075329896,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999797811150652,
      "loss": 6.7097,
      "mean_token_accuracy": 0.11343906447291374,
      "num_tokens": 3018281.0,
      "step": 1485
    },
    {
      "entropy": 6.963279438018799,
      "epoch": 0.13021060910600368,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999793612181516,
      "loss": 6.7186,
      "mean_token_accuracy": 0.11935974359512329,
      "num_tokens": 3029062.0,
      "step": 1490
    },
    {
      "entropy": 6.882570552825928,
      "epoch": 0.13064755745870837,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999789370060161,
      "loss": 6.6863,
      "mean_token_accuracy": 0.12017739489674568,
      "num_tokens": 3038142.0,
      "step": 1495
    },
    {
      "entropy": 6.990508508682251,
      "epoch": 0.1310845058114131,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999785084786667,
      "loss": 6.7944,
      "mean_token_accuracy": 0.1040409967303276,
      "num_tokens": 3049123.0,
      "step": 1500
    },
    {
      "entropy": 6.859305667877197,
      "epoch": 0.1315214541641178,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999780756361114,
      "loss": 6.6964,
      "mean_token_accuracy": 0.11377365663647651,
      "num_tokens": 3059031.0,
      "step": 1505
    },
    {
      "entropy": 6.957183074951172,
      "epoch": 0.1319584025168225,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999776384783589,
      "loss": 6.7387,
      "mean_token_accuracy": 0.11297168284654617,
      "num_tokens": 3068830.0,
      "step": 1510
    },
    {
      "entropy": 6.836016607284546,
      "epoch": 0.13239535086952722,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004999771970054174,
      "loss": 6.7098,
      "mean_token_accuracy": 0.11699092388153076,
      "num_tokens": 3080124.0,
      "step": 1515
    },
    {
      "entropy": 6.9651100635528564,
      "epoch": 0.13283229922223194,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999767512172953,
      "loss": 6.72,
      "mean_token_accuracy": 0.10890761166810989,
      "num_tokens": 3089607.0,
      "step": 1520
    },
    {
      "entropy": 6.871801853179932,
      "epoch": 0.13326924757493663,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999763011140013,
      "loss": 6.7972,
      "mean_token_accuracy": 0.10377966463565827,
      "num_tokens": 3099363.0,
      "step": 1525
    },
    {
      "entropy": 7.077904081344604,
      "epoch": 0.13370619592764135,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999758466955439,
      "loss": 6.7009,
      "mean_token_accuracy": 0.11966917067766189,
      "num_tokens": 3109552.0,
      "step": 1530
    },
    {
      "entropy": 6.822536611557007,
      "epoch": 0.13414314428034607,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999753879619319,
      "loss": 6.6813,
      "mean_token_accuracy": 0.11428188979625702,
      "num_tokens": 3119536.0,
      "step": 1535
    },
    {
      "entropy": 6.9331159591674805,
      "epoch": 0.13458009263305076,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999749249131741,
      "loss": 6.7238,
      "mean_token_accuracy": 0.11595344543457031,
      "num_tokens": 3130552.0,
      "step": 1540
    },
    {
      "entropy": 6.9094908237457275,
      "epoch": 0.13501704098575548,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999744575492793,
      "loss": 6.6219,
      "mean_token_accuracy": 0.11641833782196045,
      "num_tokens": 3140723.0,
      "step": 1545
    },
    {
      "entropy": 6.898475456237793,
      "epoch": 0.1354539893384602,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999739858702567,
      "loss": 6.6648,
      "mean_token_accuracy": 0.12024381458759308,
      "num_tokens": 3151031.0,
      "step": 1550
    },
    {
      "entropy": 6.850230026245117,
      "epoch": 0.1358909376911649,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499973509876115,
      "loss": 6.6687,
      "mean_token_accuracy": 0.11914398446679116,
      "num_tokens": 3161373.0,
      "step": 1555
    },
    {
      "entropy": 6.880463647842407,
      "epoch": 0.1363278860438696,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999730295668634,
      "loss": 6.6831,
      "mean_token_accuracy": 0.11570274606347083,
      "num_tokens": 3171094.0,
      "step": 1560
    },
    {
      "entropy": 6.895293521881103,
      "epoch": 0.13676483439657433,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999725449425114,
      "loss": 6.6283,
      "mean_token_accuracy": 0.1216744564473629,
      "num_tokens": 3181646.0,
      "step": 1565
    },
    {
      "entropy": 6.8003284454345705,
      "epoch": 0.13720178274927902,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499972056003068,
      "loss": 6.603,
      "mean_token_accuracy": 0.12167230620980263,
      "num_tokens": 3190727.0,
      "step": 1570
    },
    {
      "entropy": 6.9626048564910885,
      "epoch": 0.13763873110198374,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999715627485426,
      "loss": 6.7085,
      "mean_token_accuracy": 0.10727820545434952,
      "num_tokens": 3200316.0,
      "step": 1575
    },
    {
      "entropy": 6.841419315338134,
      "epoch": 0.13807567945468846,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000499971065178945,
      "loss": 6.6663,
      "mean_token_accuracy": 0.11115863621234894,
      "num_tokens": 3211830.0,
      "step": 1580
    },
    {
      "entropy": 6.864039182662964,
      "epoch": 0.13851262780739318,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999705632942843,
      "loss": 6.6778,
      "mean_token_accuracy": 0.11532776281237603,
      "num_tokens": 3222711.0,
      "step": 1585
    },
    {
      "entropy": 6.850924634933472,
      "epoch": 0.13894957616009787,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999700570945704,
      "loss": 6.5875,
      "mean_token_accuracy": 0.1223737932741642,
      "num_tokens": 3232810.0,
      "step": 1590
    },
    {
      "entropy": 6.806768321990967,
      "epoch": 0.1393865245128026,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999695465798128,
      "loss": 6.6716,
      "mean_token_accuracy": 0.11834604367613792,
      "num_tokens": 3243366.0,
      "step": 1595
    },
    {
      "entropy": 6.887040901184082,
      "epoch": 0.1398234728655073,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999690317500215,
      "loss": 6.6644,
      "mean_token_accuracy": 0.1145846463739872,
      "num_tokens": 3254000.0,
      "step": 1600
    },
    {
      "entropy": 6.924470329284668,
      "epoch": 0.140260421218212,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999685126052063,
      "loss": 6.6944,
      "mean_token_accuracy": 0.11372155621647835,
      "num_tokens": 3263601.0,
      "step": 1605
    },
    {
      "entropy": 6.830214357376098,
      "epoch": 0.14069736957091672,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999679891453772,
      "loss": 6.7183,
      "mean_token_accuracy": 0.11202776506543159,
      "num_tokens": 3273243.0,
      "step": 1610
    },
    {
      "entropy": 6.92011456489563,
      "epoch": 0.14113431792362144,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000499967461370544,
      "loss": 6.6652,
      "mean_token_accuracy": 0.11686113476753235,
      "num_tokens": 3284208.0,
      "step": 1615
    },
    {
      "entropy": 6.883434915542603,
      "epoch": 0.14157126627632613,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999669292807172,
      "loss": 6.646,
      "mean_token_accuracy": 0.11703974828124046,
      "num_tokens": 3294216.0,
      "step": 1620
    },
    {
      "entropy": 6.908010053634643,
      "epoch": 0.14200821462903085,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999663928759066,
      "loss": 6.6971,
      "mean_token_accuracy": 0.11337215676903725,
      "num_tokens": 3304576.0,
      "step": 1625
    },
    {
      "entropy": 6.85626482963562,
      "epoch": 0.14244516298173557,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999658521561229,
      "loss": 6.6682,
      "mean_token_accuracy": 0.10966120883822442,
      "num_tokens": 3314214.0,
      "step": 1630
    },
    {
      "entropy": 6.887621068954468,
      "epoch": 0.14288211133444026,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999653071213763,
      "loss": 6.6773,
      "mean_token_accuracy": 0.11030459403991699,
      "num_tokens": 3325527.0,
      "step": 1635
    },
    {
      "entropy": 6.751001405715942,
      "epoch": 0.14331905968714498,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999647577716771,
      "loss": 6.5835,
      "mean_token_accuracy": 0.11718848198652268,
      "num_tokens": 3334511.0,
      "step": 1640
    },
    {
      "entropy": 6.8687464714050295,
      "epoch": 0.1437560080398497,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999642041070361,
      "loss": 6.5728,
      "mean_token_accuracy": 0.12193055376410485,
      "num_tokens": 3343922.0,
      "step": 1645
    },
    {
      "entropy": 6.801642322540284,
      "epoch": 0.1441929563925544,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999636461274637,
      "loss": 6.6072,
      "mean_token_accuracy": 0.11523821577429771,
      "num_tokens": 3353895.0,
      "step": 1650
    },
    {
      "entropy": 6.818548059463501,
      "epoch": 0.1446299047452591,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004999630838329706,
      "loss": 6.6567,
      "mean_token_accuracy": 0.11228160336613655,
      "num_tokens": 3364637.0,
      "step": 1655
    },
    {
      "entropy": 6.882724046707153,
      "epoch": 0.14506685309796383,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999625172235679,
      "loss": 6.6184,
      "mean_token_accuracy": 0.11534506604075431,
      "num_tokens": 3374308.0,
      "step": 1660
    },
    {
      "entropy": 6.775670385360717,
      "epoch": 0.14550380145066852,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499961946299266,
      "loss": 6.6,
      "mean_token_accuracy": 0.11504803895950318,
      "num_tokens": 3383637.0,
      "step": 1665
    },
    {
      "entropy": 6.865798473358154,
      "epoch": 0.14594074980337324,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999613710600763,
      "loss": 6.6789,
      "mean_token_accuracy": 0.11600840091705322,
      "num_tokens": 3394150.0,
      "step": 1670
    },
    {
      "entropy": 6.889613914489746,
      "epoch": 0.14637769815607796,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999607915060094,
      "loss": 6.6368,
      "mean_token_accuracy": 0.11584951207041741,
      "num_tokens": 3404445.0,
      "step": 1675
    },
    {
      "entropy": 6.7219658374786375,
      "epoch": 0.14681464650878268,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004999602076370768,
      "loss": 6.541,
      "mean_token_accuracy": 0.12608898878097535,
      "num_tokens": 3415183.0,
      "step": 1680
    },
    {
      "entropy": 6.776702976226806,
      "epoch": 0.14725159486148737,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999596194532895,
      "loss": 6.5253,
      "mean_token_accuracy": 0.12415588200092316,
      "num_tokens": 3424824.0,
      "step": 1685
    },
    {
      "entropy": 6.871752786636352,
      "epoch": 0.1476885432141921,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999590269546588,
      "loss": 6.6382,
      "mean_token_accuracy": 0.11570465192198753,
      "num_tokens": 3434789.0,
      "step": 1690
    },
    {
      "entropy": 6.707127237319947,
      "epoch": 0.1481254915668968,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999584301411961,
      "loss": 6.5214,
      "mean_token_accuracy": 0.1381271928548813,
      "num_tokens": 3445810.0,
      "step": 1695
    },
    {
      "entropy": 6.895666456222534,
      "epoch": 0.1485624399196015,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999578290129129,
      "loss": 6.661,
      "mean_token_accuracy": 0.11268147677183152,
      "num_tokens": 3455510.0,
      "step": 1700
    },
    {
      "entropy": 6.774993848800659,
      "epoch": 0.14899938827230622,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999572235698206,
      "loss": 6.6274,
      "mean_token_accuracy": 0.11752628609538078,
      "num_tokens": 3465785.0,
      "step": 1705
    },
    {
      "entropy": 6.814649772644043,
      "epoch": 0.14943633662501093,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999566138119308,
      "loss": 6.6843,
      "mean_token_accuracy": 0.1124463550746441,
      "num_tokens": 3477568.0,
      "step": 1710
    },
    {
      "entropy": 6.754849433898926,
      "epoch": 0.14987328497771563,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999559997392554,
      "loss": 6.5138,
      "mean_token_accuracy": 0.13679737225174904,
      "num_tokens": 3487703.0,
      "step": 1715
    },
    {
      "entropy": 6.805766534805298,
      "epoch": 0.15031023333042035,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999553813518061,
      "loss": 6.5906,
      "mean_token_accuracy": 0.1175839364528656,
      "num_tokens": 3497757.0,
      "step": 1720
    },
    {
      "entropy": 6.736756134033203,
      "epoch": 0.15074718168312506,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999547586495946,
      "loss": 6.5553,
      "mean_token_accuracy": 0.1203412413597107,
      "num_tokens": 3507594.0,
      "step": 1725
    },
    {
      "entropy": 6.828785610198975,
      "epoch": 0.15118413003582976,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999541316326329,
      "loss": 6.6505,
      "mean_token_accuracy": 0.11356838271021844,
      "num_tokens": 3517889.0,
      "step": 1730
    },
    {
      "entropy": 6.788142585754395,
      "epoch": 0.15162107838853447,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999535003009332,
      "loss": 6.621,
      "mean_token_accuracy": 0.11979384943842888,
      "num_tokens": 3528158.0,
      "step": 1735
    },
    {
      "entropy": 6.820278310775757,
      "epoch": 0.1520580267412392,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999528646545074,
      "loss": 6.6304,
      "mean_token_accuracy": 0.1174927718937397,
      "num_tokens": 3539099.0,
      "step": 1740
    },
    {
      "entropy": 6.8182226657867435,
      "epoch": 0.15249497509394389,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999522246933678,
      "loss": 6.631,
      "mean_token_accuracy": 0.11622772663831711,
      "num_tokens": 3549819.0,
      "step": 1745
    },
    {
      "entropy": 6.893373823165893,
      "epoch": 0.1529319234466486,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999515804175268,
      "loss": 6.632,
      "mean_token_accuracy": 0.11604494601488113,
      "num_tokens": 3560022.0,
      "step": 1750
    },
    {
      "entropy": 6.761956310272216,
      "epoch": 0.15336887179935332,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999509318269964,
      "loss": 6.5647,
      "mean_token_accuracy": 0.1173597402870655,
      "num_tokens": 3569681.0,
      "step": 1755
    },
    {
      "entropy": 6.857851600646972,
      "epoch": 0.15380582015205801,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999502789217894,
      "loss": 6.6579,
      "mean_token_accuracy": 0.11668316721916198,
      "num_tokens": 3580731.0,
      "step": 1760
    },
    {
      "entropy": 6.75460410118103,
      "epoch": 0.15424276850476273,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999496217019181,
      "loss": 6.6264,
      "mean_token_accuracy": 0.12062862291932105,
      "num_tokens": 3590482.0,
      "step": 1765
    },
    {
      "entropy": 6.796440553665161,
      "epoch": 0.15467971685746745,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999489601673952,
      "loss": 6.6159,
      "mean_token_accuracy": 0.11088166609406472,
      "num_tokens": 3600988.0,
      "step": 1770
    },
    {
      "entropy": 6.9524415016174315,
      "epoch": 0.15511666521017214,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999482943182334,
      "loss": 6.6848,
      "mean_token_accuracy": 0.1131463699042797,
      "num_tokens": 3611226.0,
      "step": 1775
    },
    {
      "entropy": 6.716507911682129,
      "epoch": 0.15555361356287686,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999476241544454,
      "loss": 6.5993,
      "mean_token_accuracy": 0.11544200181961059,
      "num_tokens": 3621324.0,
      "step": 1780
    },
    {
      "entropy": 6.776705932617188,
      "epoch": 0.15599056191558158,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999469496760441,
      "loss": 6.5311,
      "mean_token_accuracy": 0.11900553330779076,
      "num_tokens": 3631650.0,
      "step": 1785
    },
    {
      "entropy": 6.790343856811523,
      "epoch": 0.1564275102682863,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999462708830425,
      "loss": 6.5413,
      "mean_token_accuracy": 0.13263114020228386,
      "num_tokens": 3641173.0,
      "step": 1790
    },
    {
      "entropy": 6.7471825122833256,
      "epoch": 0.156864458620991,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999455877754535,
      "loss": 6.5587,
      "mean_token_accuracy": 0.1188740074634552,
      "num_tokens": 3651846.0,
      "step": 1795
    },
    {
      "entropy": 6.735989761352539,
      "epoch": 0.1573014069736957,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999449003532903,
      "loss": 6.5573,
      "mean_token_accuracy": 0.116403878480196,
      "num_tokens": 3661174.0,
      "step": 1800
    },
    {
      "entropy": 6.798489284515381,
      "epoch": 0.15773835532640043,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999442086165661,
      "loss": 6.5437,
      "mean_token_accuracy": 0.11836761459708214,
      "num_tokens": 3671042.0,
      "step": 1805
    },
    {
      "entropy": 6.753465700149536,
      "epoch": 0.15817530367910512,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499943512565294,
      "loss": 6.5084,
      "mean_token_accuracy": 0.11789297163486481,
      "num_tokens": 3681206.0,
      "step": 1810
    },
    {
      "entropy": 6.780641365051269,
      "epoch": 0.15861225203180984,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999428121994874,
      "loss": 6.6077,
      "mean_token_accuracy": 0.12206789627671241,
      "num_tokens": 3691748.0,
      "step": 1815
    },
    {
      "entropy": 6.706904029846191,
      "epoch": 0.15904920038451456,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999421075191599,
      "loss": 6.505,
      "mean_token_accuracy": 0.12005864605307578,
      "num_tokens": 3701730.0,
      "step": 1820
    },
    {
      "entropy": 6.79358229637146,
      "epoch": 0.15948614873721925,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999413985243249,
      "loss": 6.6198,
      "mean_token_accuracy": 0.11459683477878571,
      "num_tokens": 3712525.0,
      "step": 1825
    },
    {
      "entropy": 6.8429272174835205,
      "epoch": 0.15992309708992397,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499940685214996,
      "loss": 6.5678,
      "mean_token_accuracy": 0.11963910087943078,
      "num_tokens": 3722016.0,
      "step": 1830
    },
    {
      "entropy": 6.786110496520996,
      "epoch": 0.1603600454426287,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999399675911869,
      "loss": 6.476,
      "mean_token_accuracy": 0.12395038083195686,
      "num_tokens": 3731299.0,
      "step": 1835
    },
    {
      "entropy": 6.662138032913208,
      "epoch": 0.16079699379533338,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999392456529113,
      "loss": 6.5372,
      "mean_token_accuracy": 0.1260882593691349,
      "num_tokens": 3740922.0,
      "step": 1840
    },
    {
      "entropy": 6.7078369617462155,
      "epoch": 0.1612339421480381,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999385194001831,
      "loss": 6.5398,
      "mean_token_accuracy": 0.11944171339273453,
      "num_tokens": 3750339.0,
      "step": 1845
    },
    {
      "entropy": 6.79173173904419,
      "epoch": 0.16167089050074282,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999377888330164,
      "loss": 6.5166,
      "mean_token_accuracy": 0.12448024600744248,
      "num_tokens": 3761379.0,
      "step": 1850
    },
    {
      "entropy": 6.675159645080567,
      "epoch": 0.1621078388534475,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999370539514249,
      "loss": 6.5731,
      "mean_token_accuracy": 0.11835632175207138,
      "num_tokens": 3771649.0,
      "step": 1855
    },
    {
      "entropy": 6.754816055297852,
      "epoch": 0.16254478720615223,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999363147554229,
      "loss": 6.4983,
      "mean_token_accuracy": 0.1242656834423542,
      "num_tokens": 3782393.0,
      "step": 1860
    },
    {
      "entropy": 6.758715486526489,
      "epoch": 0.16298173555885695,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999355712450244,
      "loss": 6.5619,
      "mean_token_accuracy": 0.12336914837360383,
      "num_tokens": 3792849.0,
      "step": 1865
    },
    {
      "entropy": 6.747960758209229,
      "epoch": 0.16341868391156164,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999348234202438,
      "loss": 6.5009,
      "mean_token_accuracy": 0.1251760870218277,
      "num_tokens": 3803848.0,
      "step": 1870
    },
    {
      "entropy": 6.647700023651123,
      "epoch": 0.16385563226426636,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999340712810956,
      "loss": 6.4194,
      "mean_token_accuracy": 0.13567735776305198,
      "num_tokens": 3814704.0,
      "step": 1875
    },
    {
      "entropy": 6.708374786376953,
      "epoch": 0.16429258061697108,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999333148275939,
      "loss": 6.5592,
      "mean_token_accuracy": 0.11908423453569413,
      "num_tokens": 3824955.0,
      "step": 1880
    },
    {
      "entropy": 6.752207708358765,
      "epoch": 0.1647295289696758,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999325540597535,
      "loss": 6.5372,
      "mean_token_accuracy": 0.12644217535853386,
      "num_tokens": 3836540.0,
      "step": 1885
    },
    {
      "entropy": 6.604652452468872,
      "epoch": 0.1651664773223805,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999317889775887,
      "loss": 6.431,
      "mean_token_accuracy": 0.13235808461904525,
      "num_tokens": 3846472.0,
      "step": 1890
    },
    {
      "entropy": 6.771966743469238,
      "epoch": 0.1656034256750852,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999310195811143,
      "loss": 6.5333,
      "mean_token_accuracy": 0.11936069577932358,
      "num_tokens": 3857187.0,
      "step": 1895
    },
    {
      "entropy": 6.751552248001099,
      "epoch": 0.16604037402778993,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999302458703451,
      "loss": 6.5008,
      "mean_token_accuracy": 0.12447258234024047,
      "num_tokens": 3866259.0,
      "step": 1900
    },
    {
      "entropy": 6.657085752487182,
      "epoch": 0.16647732238049462,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499929467845296,
      "loss": 6.5587,
      "mean_token_accuracy": 0.11343137621879577,
      "num_tokens": 3876869.0,
      "step": 1905
    },
    {
      "entropy": 6.7602463722229,
      "epoch": 0.16691427073319934,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999286855059818,
      "loss": 6.5404,
      "mean_token_accuracy": 0.12126922458410264,
      "num_tokens": 3888278.0,
      "step": 1910
    },
    {
      "entropy": 6.7652349948883055,
      "epoch": 0.16735121908590406,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999278988524177,
      "loss": 6.6004,
      "mean_token_accuracy": 0.12169176936149598,
      "num_tokens": 3898648.0,
      "step": 1915
    },
    {
      "entropy": 6.759922075271606,
      "epoch": 0.16778816743860875,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999271078846184,
      "loss": 6.5229,
      "mean_token_accuracy": 0.11797695755958557,
      "num_tokens": 3908282.0,
      "step": 1920
    },
    {
      "entropy": 6.548764705657959,
      "epoch": 0.16822511579131347,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999263126025994,
      "loss": 6.5006,
      "mean_token_accuracy": 0.12412167713046074,
      "num_tokens": 3918333.0,
      "step": 1925
    },
    {
      "entropy": 6.82745590209961,
      "epoch": 0.1686620641440182,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999255130063759,
      "loss": 6.51,
      "mean_token_accuracy": 0.1215559110045433,
      "num_tokens": 3927403.0,
      "step": 1930
    },
    {
      "entropy": 6.691656970977784,
      "epoch": 0.16909901249672288,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999247090959633,
      "loss": 6.5286,
      "mean_token_accuracy": 0.12117488756775856,
      "num_tokens": 3937352.0,
      "step": 1935
    },
    {
      "entropy": 6.6749176502227785,
      "epoch": 0.1695359608494276,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999239008713767,
      "loss": 6.3616,
      "mean_token_accuracy": 0.12652190998196602,
      "num_tokens": 3947164.0,
      "step": 1940
    },
    {
      "entropy": 6.7179159164428714,
      "epoch": 0.16997290920213232,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999230883326319,
      "loss": 6.5359,
      "mean_token_accuracy": 0.1260188527405262,
      "num_tokens": 3956887.0,
      "step": 1945
    },
    {
      "entropy": 6.743377208709717,
      "epoch": 0.170409857554837,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999222714797445,
      "loss": 6.5084,
      "mean_token_accuracy": 0.1210441879928112,
      "num_tokens": 3966542.0,
      "step": 1950
    },
    {
      "entropy": 6.723532009124756,
      "epoch": 0.17084680590754173,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00049992145031273,
      "loss": 6.4706,
      "mean_token_accuracy": 0.13221301287412643,
      "num_tokens": 3976127.0,
      "step": 1955
    },
    {
      "entropy": 6.577977323532105,
      "epoch": 0.17128375426024645,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999206248316043,
      "loss": 6.4402,
      "mean_token_accuracy": 0.12495635226368904,
      "num_tokens": 3987371.0,
      "step": 1960
    },
    {
      "entropy": 6.643271446228027,
      "epoch": 0.17172070261295114,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499919795036383,
      "loss": 6.441,
      "mean_token_accuracy": 0.12783539444208145,
      "num_tokens": 3997924.0,
      "step": 1965
    },
    {
      "entropy": 6.687552833557129,
      "epoch": 0.17215765096565586,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999189609270823,
      "loss": 6.5346,
      "mean_token_accuracy": 0.12447880506515503,
      "num_tokens": 4009150.0,
      "step": 1970
    },
    {
      "entropy": 6.702529954910278,
      "epoch": 0.17259459931836058,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499918122503718,
      "loss": 6.5461,
      "mean_token_accuracy": 0.12213218584656715,
      "num_tokens": 4018899.0,
      "step": 1975
    },
    {
      "entropy": 6.684345149993897,
      "epoch": 0.17303154767106527,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999172797663064,
      "loss": 6.489,
      "mean_token_accuracy": 0.12457576245069504,
      "num_tokens": 4028652.0,
      "step": 1980
    },
    {
      "entropy": 6.704040813446045,
      "epoch": 0.17346849602376999,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999164327148634,
      "loss": 6.53,
      "mean_token_accuracy": 0.11930816993117332,
      "num_tokens": 4039819.0,
      "step": 1985
    },
    {
      "entropy": 6.703826856613159,
      "epoch": 0.1739054443764747,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999155813494053,
      "loss": 6.4394,
      "mean_token_accuracy": 0.12949882224202156,
      "num_tokens": 4050029.0,
      "step": 1990
    },
    {
      "entropy": 6.5715407371521,
      "epoch": 0.17434239272917942,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999147256699487,
      "loss": 6.4687,
      "mean_token_accuracy": 0.1293756529688835,
      "num_tokens": 4060497.0,
      "step": 1995
    },
    {
      "entropy": 6.748078489303589,
      "epoch": 0.17477934108188412,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999138656765096,
      "loss": 6.45,
      "mean_token_accuracy": 0.1262899488210678,
      "num_tokens": 4071477.0,
      "step": 2000
    },
    {
      "entropy": 6.660907793045044,
      "epoch": 0.17521628943458883,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999130013691049,
      "loss": 6.5565,
      "mean_token_accuracy": 0.12222278639674186,
      "num_tokens": 4081869.0,
      "step": 2005
    },
    {
      "entropy": 6.7544700622558596,
      "epoch": 0.17565323778729355,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999121327477509,
      "loss": 6.528,
      "mean_token_accuracy": 0.12365535125136376,
      "num_tokens": 4092797.0,
      "step": 2010
    },
    {
      "entropy": 6.605335807800293,
      "epoch": 0.17609018613999824,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999112598124642,
      "loss": 6.4776,
      "mean_token_accuracy": 0.12264428436756133,
      "num_tokens": 4103688.0,
      "step": 2015
    },
    {
      "entropy": 6.75774884223938,
      "epoch": 0.17652713449270296,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999103825632618,
      "loss": 6.4325,
      "mean_token_accuracy": 0.12662287577986717,
      "num_tokens": 4114037.0,
      "step": 2020
    },
    {
      "entropy": 6.603244209289551,
      "epoch": 0.17696408284540768,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999095010001605,
      "loss": 6.4215,
      "mean_token_accuracy": 0.12197433784604073,
      "num_tokens": 4124164.0,
      "step": 2025
    },
    {
      "entropy": 6.765624523162842,
      "epoch": 0.17740103119811237,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499908615123177,
      "loss": 6.5396,
      "mean_token_accuracy": 0.11871138587594032,
      "num_tokens": 4134225.0,
      "step": 2030
    },
    {
      "entropy": 6.675165271759033,
      "epoch": 0.1778379795508171,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999077249323284,
      "loss": 6.5901,
      "mean_token_accuracy": 0.11778633445501327,
      "num_tokens": 4145962.0,
      "step": 2035
    },
    {
      "entropy": 6.702001953125,
      "epoch": 0.1782749279035218,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999068304276319,
      "loss": 6.4709,
      "mean_token_accuracy": 0.12179390788078308,
      "num_tokens": 4156572.0,
      "step": 2040
    },
    {
      "entropy": 6.660705947875977,
      "epoch": 0.1787118762562265,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999059316091045,
      "loss": 6.5431,
      "mean_token_accuracy": 0.12079031616449357,
      "num_tokens": 4166960.0,
      "step": 2045
    },
    {
      "entropy": 6.70681414604187,
      "epoch": 0.17914882460893122,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999050284767634,
      "loss": 6.4874,
      "mean_token_accuracy": 0.12624370977282523,
      "num_tokens": 4177833.0,
      "step": 2050
    },
    {
      "entropy": 6.578025674819946,
      "epoch": 0.17958577296163594,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999041210306261,
      "loss": 6.4109,
      "mean_token_accuracy": 0.12834745794534683,
      "num_tokens": 4187858.0,
      "step": 2055
    },
    {
      "entropy": 6.776542139053345,
      "epoch": 0.18002272131434063,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999032092707098,
      "loss": 6.545,
      "mean_token_accuracy": 0.12043375000357628,
      "num_tokens": 4198207.0,
      "step": 2060
    },
    {
      "entropy": 6.6667626857757565,
      "epoch": 0.18045966966704535,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999022931970323,
      "loss": 6.425,
      "mean_token_accuracy": 0.12915648147463799,
      "num_tokens": 4207594.0,
      "step": 2065
    },
    {
      "entropy": 6.629980421066284,
      "epoch": 0.18089661801975007,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999013728096107,
      "loss": 6.431,
      "mean_token_accuracy": 0.12719237804412842,
      "num_tokens": 4217334.0,
      "step": 2070
    },
    {
      "entropy": 6.644939041137695,
      "epoch": 0.18133356637245476,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499900448108463,
      "loss": 6.4226,
      "mean_token_accuracy": 0.13288431987166405,
      "num_tokens": 4226407.0,
      "step": 2075
    },
    {
      "entropy": 6.604313325881958,
      "epoch": 0.18177051472515948,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499899519093607,
      "loss": 6.376,
      "mean_token_accuracy": 0.13293503150343894,
      "num_tokens": 4237818.0,
      "step": 2080
    },
    {
      "entropy": 6.664392614364624,
      "epoch": 0.1822074630778642,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004998985857650603,
      "loss": 6.4837,
      "mean_token_accuracy": 0.12130877822637558,
      "num_tokens": 4247783.0,
      "step": 2085
    },
    {
      "entropy": 6.606699132919312,
      "epoch": 0.18264441143056892,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998976481228409,
      "loss": 6.4436,
      "mean_token_accuracy": 0.12809643447399138,
      "num_tokens": 4258049.0,
      "step": 2090
    },
    {
      "entropy": 6.618226480484009,
      "epoch": 0.1830813597832736,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004998967061669668,
      "loss": 6.4403,
      "mean_token_accuracy": 0.11963641420006751,
      "num_tokens": 4269357.0,
      "step": 2095
    },
    {
      "entropy": 6.6365663528442385,
      "epoch": 0.18351830813597833,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998957598974559,
      "loss": 6.3581,
      "mean_token_accuracy": 0.13459884151816368,
      "num_tokens": 4279576.0,
      "step": 2100
    },
    {
      "entropy": 6.598727512359619,
      "epoch": 0.18395525648868305,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998948093143266,
      "loss": 6.4112,
      "mean_token_accuracy": 0.1286292240023613,
      "num_tokens": 4289761.0,
      "step": 2105
    },
    {
      "entropy": 6.582360744476318,
      "epoch": 0.18439220484138774,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499893854417597,
      "loss": 6.355,
      "mean_token_accuracy": 0.12943675965070725,
      "num_tokens": 4299231.0,
      "step": 2110
    },
    {
      "entropy": 6.613821411132813,
      "epoch": 0.18482915319409246,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998928952072854,
      "loss": 6.5108,
      "mean_token_accuracy": 0.12351166307926179,
      "num_tokens": 4308687.0,
      "step": 2115
    },
    {
      "entropy": 6.6865287780761715,
      "epoch": 0.18526610154679718,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998919316834102,
      "loss": 6.4583,
      "mean_token_accuracy": 0.12498997375369073,
      "num_tokens": 4318183.0,
      "step": 2120
    },
    {
      "entropy": 6.668766021728516,
      "epoch": 0.18570304989950187,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00049989096384599,
      "loss": 6.5085,
      "mean_token_accuracy": 0.12343138679862023,
      "num_tokens": 4328482.0,
      "step": 2125
    },
    {
      "entropy": 6.633816528320312,
      "epoch": 0.1861399982522066,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004998899916950431,
      "loss": 6.3824,
      "mean_token_accuracy": 0.1267977774143219,
      "num_tokens": 4339313.0,
      "step": 2130
    },
    {
      "entropy": 6.633962535858155,
      "epoch": 0.1865769466049113,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998890152305885,
      "loss": 6.4954,
      "mean_token_accuracy": 0.12780801132321357,
      "num_tokens": 4350299.0,
      "step": 2135
    },
    {
      "entropy": 6.715720367431641,
      "epoch": 0.187013894957616,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998880344526447,
      "loss": 6.4273,
      "mean_token_accuracy": 0.13316971361637114,
      "num_tokens": 4361424.0,
      "step": 2140
    },
    {
      "entropy": 6.532921981811524,
      "epoch": 0.18745084331032072,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998870493612306,
      "loss": 6.3943,
      "mean_token_accuracy": 0.1273239456117153,
      "num_tokens": 4371480.0,
      "step": 2145
    },
    {
      "entropy": 6.555461359024048,
      "epoch": 0.18788779166302544,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004998860599563649,
      "loss": 6.4256,
      "mean_token_accuracy": 0.12563225403428077,
      "num_tokens": 4381951.0,
      "step": 2150
    },
    {
      "entropy": 6.6292718887329105,
      "epoch": 0.18832474001573013,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004998850662380668,
      "loss": 6.4814,
      "mean_token_accuracy": 0.12124204561114311,
      "num_tokens": 4392290.0,
      "step": 2155
    },
    {
      "entropy": 6.614552593231201,
      "epoch": 0.18876168836843485,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998840682063553,
      "loss": 6.2488,
      "mean_token_accuracy": 0.14245351031422615,
      "num_tokens": 4403350.0,
      "step": 2160
    },
    {
      "entropy": 6.56493353843689,
      "epoch": 0.18919863672113957,
      "grad_norm": 0.875,
      "learning_rate": 0.0004998830658612496,
      "loss": 6.4447,
      "mean_token_accuracy": 0.12840750589966773,
      "num_tokens": 4413171.0,
      "step": 2165
    },
    {
      "entropy": 6.563513231277466,
      "epoch": 0.18963558507384426,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998820592027688,
      "loss": 6.3957,
      "mean_token_accuracy": 0.1307518906891346,
      "num_tokens": 4424371.0,
      "step": 2170
    },
    {
      "entropy": 6.649606466293335,
      "epoch": 0.19007253342654898,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004998810482309322,
      "loss": 6.3667,
      "mean_token_accuracy": 0.12607548013329506,
      "num_tokens": 4434040.0,
      "step": 2175
    },
    {
      "entropy": 6.6798676490783695,
      "epoch": 0.1905094817792537,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998800329457595,
      "loss": 6.4636,
      "mean_token_accuracy": 0.12016530409455299,
      "num_tokens": 4444663.0,
      "step": 2180
    },
    {
      "entropy": 6.602048730850219,
      "epoch": 0.1909464301319584,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998790133472697,
      "loss": 6.4146,
      "mean_token_accuracy": 0.1262466661632061,
      "num_tokens": 4454733.0,
      "step": 2185
    },
    {
      "entropy": 6.578805065155029,
      "epoch": 0.1913833784846631,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998779894354827,
      "loss": 6.3965,
      "mean_token_accuracy": 0.13079310059547425,
      "num_tokens": 4464286.0,
      "step": 2190
    },
    {
      "entropy": 6.633804512023926,
      "epoch": 0.19182032683736783,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000499876961210418,
      "loss": 6.389,
      "mean_token_accuracy": 0.13454255908727647,
      "num_tokens": 4475005.0,
      "step": 2195
    },
    {
      "entropy": 6.588158702850341,
      "epoch": 0.19225727519007255,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004998759286720953,
      "loss": 6.336,
      "mean_token_accuracy": 0.13137971311807634,
      "num_tokens": 4484606.0,
      "step": 2200
    },
    {
      "entropy": 6.507455158233642,
      "epoch": 0.19269422354277724,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998748918205345,
      "loss": 6.4482,
      "mean_token_accuracy": 0.12594832628965377,
      "num_tokens": 4494930.0,
      "step": 2205
    },
    {
      "entropy": 6.659773206710815,
      "epoch": 0.19313117189548196,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998738506557554,
      "loss": 6.4699,
      "mean_token_accuracy": 0.12103351801633835,
      "num_tokens": 4505304.0,
      "step": 2210
    },
    {
      "entropy": 6.693742942810059,
      "epoch": 0.19356812024818668,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499872805177778,
      "loss": 6.392,
      "mean_token_accuracy": 0.1304948538541794,
      "num_tokens": 4514333.0,
      "step": 2215
    },
    {
      "entropy": 6.483904218673706,
      "epoch": 0.19400506860089137,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998717553866225,
      "loss": 6.4429,
      "mean_token_accuracy": 0.13353654071688653,
      "num_tokens": 4525175.0,
      "step": 2220
    },
    {
      "entropy": 6.648404788970947,
      "epoch": 0.1944420169535961,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998707012823087,
      "loss": 6.3151,
      "mean_token_accuracy": 0.13554888740181922,
      "num_tokens": 4534212.0,
      "step": 2225
    },
    {
      "entropy": 6.492750024795532,
      "epoch": 0.1948789653063008,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004998696428648573,
      "loss": 6.4957,
      "mean_token_accuracy": 0.12358262240886689,
      "num_tokens": 4545141.0,
      "step": 2230
    },
    {
      "entropy": 6.6611635208129885,
      "epoch": 0.1953159136590055,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998685801342881,
      "loss": 6.3623,
      "mean_token_accuracy": 0.13657205328345298,
      "num_tokens": 4554410.0,
      "step": 2235
    },
    {
      "entropy": 6.515309810638428,
      "epoch": 0.19575286201171022,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998675130906217,
      "loss": 6.3629,
      "mean_token_accuracy": 0.1314949244260788,
      "num_tokens": 4564397.0,
      "step": 2240
    },
    {
      "entropy": 6.506466865539551,
      "epoch": 0.19618981036441493,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998664417338787,
      "loss": 6.3565,
      "mean_token_accuracy": 0.13614993840456008,
      "num_tokens": 4573986.0,
      "step": 2245
    },
    {
      "entropy": 6.616925239562988,
      "epoch": 0.19662675871711963,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004998653660640794,
      "loss": 6.3557,
      "mean_token_accuracy": 0.1368450790643692,
      "num_tokens": 4583449.0,
      "step": 2250
    },
    {
      "entropy": 6.583423757553101,
      "epoch": 0.19706370706982435,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998642860812447,
      "loss": 6.393,
      "mean_token_accuracy": 0.12896343395113946,
      "num_tokens": 4593637.0,
      "step": 2255
    },
    {
      "entropy": 6.550680303573609,
      "epoch": 0.19750065542252906,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998632017853951,
      "loss": 6.3727,
      "mean_token_accuracy": 0.13199751377105712,
      "num_tokens": 4603865.0,
      "step": 2260
    },
    {
      "entropy": 6.628088903427124,
      "epoch": 0.19793760377523376,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998621131765514,
      "loss": 6.4037,
      "mean_token_accuracy": 0.1369698889553547,
      "num_tokens": 4613502.0,
      "step": 2265
    },
    {
      "entropy": 6.536065626144409,
      "epoch": 0.19837455212793847,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998610202547345,
      "loss": 6.3718,
      "mean_token_accuracy": 0.1352036103606224,
      "num_tokens": 4623163.0,
      "step": 2270
    },
    {
      "entropy": 6.6453217506408695,
      "epoch": 0.1988115004806432,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004998599230199654,
      "loss": 6.3806,
      "mean_token_accuracy": 0.13569082021713258,
      "num_tokens": 4632850.0,
      "step": 2275
    },
    {
      "entropy": 6.540207004547119,
      "epoch": 0.19924844883334789,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998588214722653,
      "loss": 6.39,
      "mean_token_accuracy": 0.13032061085104943,
      "num_tokens": 4642858.0,
      "step": 2280
    },
    {
      "entropy": 6.648379802703857,
      "epoch": 0.1996853971860526,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004998577156116551,
      "loss": 6.4224,
      "mean_token_accuracy": 0.12049075067043305,
      "num_tokens": 4653083.0,
      "step": 2285
    },
    {
      "entropy": 6.599198913574218,
      "epoch": 0.20012234553875732,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998566054381561,
      "loss": 6.4329,
      "mean_token_accuracy": 0.12729515209794046,
      "num_tokens": 4663353.0,
      "step": 2290
    },
    {
      "entropy": 6.545134258270264,
      "epoch": 0.20055929389146204,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004998554909517895,
      "loss": 6.2739,
      "mean_token_accuracy": 0.13422200456261635,
      "num_tokens": 4673878.0,
      "step": 2295
    },
    {
      "entropy": 6.544426107406617,
      "epoch": 0.20099624224416673,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998543721525769,
      "loss": 6.3686,
      "mean_token_accuracy": 0.13058247044682503,
      "num_tokens": 4683661.0,
      "step": 2300
    },
    {
      "entropy": 6.487976789474487,
      "epoch": 0.20143319059687145,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998532490405397,
      "loss": 6.4136,
      "mean_token_accuracy": 0.1305684819817543,
      "num_tokens": 4693817.0,
      "step": 2305
    },
    {
      "entropy": 6.587650442123413,
      "epoch": 0.20187013894957617,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998521216156991,
      "loss": 6.3358,
      "mean_token_accuracy": 0.12866818010807038,
      "num_tokens": 4703550.0,
      "step": 2310
    },
    {
      "entropy": 6.52202959060669,
      "epoch": 0.20230708730228086,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998509898780771,
      "loss": 6.4504,
      "mean_token_accuracy": 0.12608920335769652,
      "num_tokens": 4714872.0,
      "step": 2315
    },
    {
      "entropy": 6.569240665435791,
      "epoch": 0.20274403565498558,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004998498538276954,
      "loss": 6.3384,
      "mean_token_accuracy": 0.13502655774354935,
      "num_tokens": 4725745.0,
      "step": 2320
    },
    {
      "entropy": 6.534421777725219,
      "epoch": 0.2031809840076903,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998487134645755,
      "loss": 6.4199,
      "mean_token_accuracy": 0.1276063710451126,
      "num_tokens": 4735560.0,
      "step": 2325
    },
    {
      "entropy": 6.615751838684082,
      "epoch": 0.203617932360395,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998475687887396,
      "loss": 6.4198,
      "mean_token_accuracy": 0.12353309392929077,
      "num_tokens": 4746304.0,
      "step": 2330
    },
    {
      "entropy": 6.603571605682373,
      "epoch": 0.2040548807130997,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998464198002095,
      "loss": 6.3777,
      "mean_token_accuracy": 0.1305947408080101,
      "num_tokens": 4756200.0,
      "step": 2335
    },
    {
      "entropy": 6.568876934051514,
      "epoch": 0.20449182906580443,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998452664990072,
      "loss": 6.3913,
      "mean_token_accuracy": 0.12563365921378136,
      "num_tokens": 4766249.0,
      "step": 2340
    },
    {
      "entropy": 6.6453080654144285,
      "epoch": 0.20492877741850912,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998441088851548,
      "loss": 6.4154,
      "mean_token_accuracy": 0.12701233252882957,
      "num_tokens": 4776510.0,
      "step": 2345
    },
    {
      "entropy": 6.529261922836303,
      "epoch": 0.20536572577121384,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004998429469586748,
      "loss": 6.383,
      "mean_token_accuracy": 0.1331693261861801,
      "num_tokens": 4786798.0,
      "step": 2350
    },
    {
      "entropy": 6.568332815170288,
      "epoch": 0.20580267412391856,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000499841780719589,
      "loss": 6.3123,
      "mean_token_accuracy": 0.12657533809542656,
      "num_tokens": 4796001.0,
      "step": 2355
    },
    {
      "entropy": 6.49063138961792,
      "epoch": 0.20623962247662325,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004998406101679202,
      "loss": 6.2915,
      "mean_token_accuracy": 0.13698873668909073,
      "num_tokens": 4807634.0,
      "step": 2360
    },
    {
      "entropy": 6.5027268409729,
      "epoch": 0.20667657082932797,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998394353036906,
      "loss": 6.3005,
      "mean_token_accuracy": 0.1387121118605137,
      "num_tokens": 4817840.0,
      "step": 2365
    },
    {
      "entropy": 6.520350217819214,
      "epoch": 0.2071135191820327,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998382561269229,
      "loss": 6.2813,
      "mean_token_accuracy": 0.13542777374386789,
      "num_tokens": 4827186.0,
      "step": 2370
    },
    {
      "entropy": 6.492212581634521,
      "epoch": 0.20755046753473738,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998370726376395,
      "loss": 6.3373,
      "mean_token_accuracy": 0.12861326411366464,
      "num_tokens": 4837140.0,
      "step": 2375
    },
    {
      "entropy": 6.579593086242676,
      "epoch": 0.2079874158874421,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998358848358634,
      "loss": 6.359,
      "mean_token_accuracy": 0.13045982643961906,
      "num_tokens": 4846907.0,
      "step": 2380
    },
    {
      "entropy": 6.4922003746032715,
      "epoch": 0.20842436424014682,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499834692721617,
      "loss": 6.3765,
      "mean_token_accuracy": 0.1315027058124542,
      "num_tokens": 4856165.0,
      "step": 2385
    },
    {
      "entropy": 6.563181781768799,
      "epoch": 0.2088613125928515,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998334962949237,
      "loss": 6.2397,
      "mean_token_accuracy": 0.14191331565380097,
      "num_tokens": 4865308.0,
      "step": 2390
    },
    {
      "entropy": 6.4551702499389645,
      "epoch": 0.20929826094555623,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004998322955558059,
      "loss": 6.311,
      "mean_token_accuracy": 0.14102340936660768,
      "num_tokens": 4876286.0,
      "step": 2395
    },
    {
      "entropy": 6.538469457626343,
      "epoch": 0.20973520929826095,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004998310905042869,
      "loss": 6.3287,
      "mean_token_accuracy": 0.13503796905279158,
      "num_tokens": 4887227.0,
      "step": 2400
    },
    {
      "entropy": 6.518013572692871,
      "epoch": 0.21017215765096567,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004998298811403898,
      "loss": 6.3721,
      "mean_token_accuracy": 0.13244589641690255,
      "num_tokens": 4896994.0,
      "step": 2405
    },
    {
      "entropy": 6.554981422424317,
      "epoch": 0.21060910600367036,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998286674641378,
      "loss": 6.3652,
      "mean_token_accuracy": 0.1293633349239826,
      "num_tokens": 4907623.0,
      "step": 2410
    },
    {
      "entropy": 6.4484740734100345,
      "epoch": 0.21104605435637508,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499827449475554,
      "loss": 6.2625,
      "mean_token_accuracy": 0.133323784917593,
      "num_tokens": 4917040.0,
      "step": 2415
    },
    {
      "entropy": 6.612685298919677,
      "epoch": 0.2114830027090798,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499826227174662,
      "loss": 6.406,
      "mean_token_accuracy": 0.12849230915308,
      "num_tokens": 4928390.0,
      "step": 2420
    },
    {
      "entropy": 6.526340484619141,
      "epoch": 0.2119199510617845,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004998250005614852,
      "loss": 6.3792,
      "mean_token_accuracy": 0.130333261936903,
      "num_tokens": 4939040.0,
      "step": 2425
    },
    {
      "entropy": 6.447541618347168,
      "epoch": 0.2123568994144892,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000499823769636047,
      "loss": 6.3954,
      "mean_token_accuracy": 0.1286868393421173,
      "num_tokens": 4949915.0,
      "step": 2430
    },
    {
      "entropy": 6.531896543502808,
      "epoch": 0.21279384776719393,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998225343983712,
      "loss": 6.2937,
      "mean_token_accuracy": 0.13460685759782792,
      "num_tokens": 4960127.0,
      "step": 2435
    },
    {
      "entropy": 6.602890491485596,
      "epoch": 0.21323079611989862,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998212948484814,
      "loss": 6.3519,
      "mean_token_accuracy": 0.12345878928899764,
      "num_tokens": 4970152.0,
      "step": 2440
    },
    {
      "entropy": 6.47441291809082,
      "epoch": 0.21366774447260334,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998200509864012,
      "loss": 6.3047,
      "mean_token_accuracy": 0.13682056069374085,
      "num_tokens": 4979745.0,
      "step": 2445
    },
    {
      "entropy": 6.590406131744385,
      "epoch": 0.21410469282530806,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998188028121547,
      "loss": 6.3249,
      "mean_token_accuracy": 0.14112883135676385,
      "num_tokens": 4989924.0,
      "step": 2450
    },
    {
      "entropy": 6.459373140335083,
      "epoch": 0.21454164117801275,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998175503257658,
      "loss": 6.2795,
      "mean_token_accuracy": 0.1387402407824993,
      "num_tokens": 4999698.0,
      "step": 2455
    },
    {
      "entropy": 6.507392549514771,
      "epoch": 0.21497858953071747,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998162935272585,
      "loss": 6.374,
      "mean_token_accuracy": 0.13244377076625824,
      "num_tokens": 5010625.0,
      "step": 2460
    },
    {
      "entropy": 6.44816575050354,
      "epoch": 0.2154155378834222,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998150324166568,
      "loss": 6.2678,
      "mean_token_accuracy": 0.14076413363218307,
      "num_tokens": 5019740.0,
      "step": 2465
    },
    {
      "entropy": 6.608205986022949,
      "epoch": 0.21585248623612688,
      "grad_norm": 0.78515625,
      "learning_rate": 0.000499813766993985,
      "loss": 6.3888,
      "mean_token_accuracy": 0.12639941275119781,
      "num_tokens": 5030601.0,
      "step": 2470
    },
    {
      "entropy": 6.60090184211731,
      "epoch": 0.2162894345888316,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998124972592674,
      "loss": 6.316,
      "mean_token_accuracy": 0.13324218690395356,
      "num_tokens": 5040567.0,
      "step": 2475
    },
    {
      "entropy": 6.433722639083863,
      "epoch": 0.21672638294153632,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998112232125283,
      "loss": 6.2152,
      "mean_token_accuracy": 0.13335614651441574,
      "num_tokens": 5049512.0,
      "step": 2480
    },
    {
      "entropy": 6.4912238121032715,
      "epoch": 0.217163331294241,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000499809944853792,
      "loss": 6.4084,
      "mean_token_accuracy": 0.12704317271709442,
      "num_tokens": 5060986.0,
      "step": 2485
    },
    {
      "entropy": 6.486314105987549,
      "epoch": 0.21760027964694573,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998086621830833,
      "loss": 6.3174,
      "mean_token_accuracy": 0.13398604840040207,
      "num_tokens": 5071492.0,
      "step": 2490
    },
    {
      "entropy": 6.531875467300415,
      "epoch": 0.21803722799965045,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004998073752004267,
      "loss": 6.3477,
      "mean_token_accuracy": 0.1316291257739067,
      "num_tokens": 5081722.0,
      "step": 2495
    },
    {
      "entropy": 6.555519723892212,
      "epoch": 0.21847417635235516,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998060839058467,
      "loss": 6.2966,
      "mean_token_accuracy": 0.13120478838682176,
      "num_tokens": 5091906.0,
      "step": 2500
    },
    {
      "entropy": 6.486591482162476,
      "epoch": 0.21891112470505986,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998047882993683,
      "loss": 6.2947,
      "mean_token_accuracy": 0.13860670775175093,
      "num_tokens": 5102025.0,
      "step": 2505
    },
    {
      "entropy": 6.444411563873291,
      "epoch": 0.21934807305776458,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004998034883810162,
      "loss": 6.2987,
      "mean_token_accuracy": 0.13405173942446708,
      "num_tokens": 5112354.0,
      "step": 2510
    },
    {
      "entropy": 6.513107395172119,
      "epoch": 0.2197850214104693,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004998021841508155,
      "loss": 6.2366,
      "mean_token_accuracy": 0.14100577235221862,
      "num_tokens": 5121695.0,
      "step": 2515
    },
    {
      "entropy": 6.501870727539062,
      "epoch": 0.22022196976317399,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004998008756087911,
      "loss": 6.3494,
      "mean_token_accuracy": 0.13254786729812623,
      "num_tokens": 5132561.0,
      "step": 2520
    },
    {
      "entropy": 6.552183055877686,
      "epoch": 0.2206589181158787,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499799562754968,
      "loss": 6.3747,
      "mean_token_accuracy": 0.12984378933906554,
      "num_tokens": 5141981.0,
      "step": 2525
    },
    {
      "entropy": 6.45419282913208,
      "epoch": 0.22109586646858342,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997982455893716,
      "loss": 6.2503,
      "mean_token_accuracy": 0.13417460918426513,
      "num_tokens": 5151550.0,
      "step": 2530
    },
    {
      "entropy": 6.523354196548462,
      "epoch": 0.22153281482128812,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997969241120271,
      "loss": 6.3171,
      "mean_token_accuracy": 0.12860540449619293,
      "num_tokens": 5161679.0,
      "step": 2535
    },
    {
      "entropy": 6.444218873977661,
      "epoch": 0.22196976317399283,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997955983229597,
      "loss": 6.3836,
      "mean_token_accuracy": 0.13248728290200235,
      "num_tokens": 5171575.0,
      "step": 2540
    },
    {
      "entropy": 6.49551568031311,
      "epoch": 0.22240671152669755,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499794268222195,
      "loss": 6.2317,
      "mean_token_accuracy": 0.13764370903372763,
      "num_tokens": 5180152.0,
      "step": 2545
    },
    {
      "entropy": 6.458237552642823,
      "epoch": 0.22284365987940224,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997929338097583,
      "loss": 6.3363,
      "mean_token_accuracy": 0.13028131797909737,
      "num_tokens": 5190518.0,
      "step": 2550
    },
    {
      "entropy": 6.557815790176392,
      "epoch": 0.22328060823210696,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997915950856755,
      "loss": 6.3577,
      "mean_token_accuracy": 0.12511950880289077,
      "num_tokens": 5201729.0,
      "step": 2555
    },
    {
      "entropy": 6.497034549713135,
      "epoch": 0.22371755658481168,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997902520499721,
      "loss": 6.3337,
      "mean_token_accuracy": 0.1315643459558487,
      "num_tokens": 5211940.0,
      "step": 2560
    },
    {
      "entropy": 6.42897891998291,
      "epoch": 0.22415450493751637,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997889047026738,
      "loss": 6.2651,
      "mean_token_accuracy": 0.13462026566267013,
      "num_tokens": 5221028.0,
      "step": 2565
    },
    {
      "entropy": 6.471619081497193,
      "epoch": 0.2245914532902211,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997875530438066,
      "loss": 6.2416,
      "mean_token_accuracy": 0.13898374885320663,
      "num_tokens": 5230498.0,
      "step": 2570
    },
    {
      "entropy": 6.477230548858643,
      "epoch": 0.2250284016429258,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004997861970733963,
      "loss": 6.2929,
      "mean_token_accuracy": 0.13128770291805267,
      "num_tokens": 5240112.0,
      "step": 2575
    },
    {
      "entropy": 6.502867269515991,
      "epoch": 0.2254653499956305,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499784836791469,
      "loss": 6.2492,
      "mean_token_accuracy": 0.13940685987472534,
      "num_tokens": 5249201.0,
      "step": 2580
    },
    {
      "entropy": 6.488351058959961,
      "epoch": 0.22590229834833522,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997834721980507,
      "loss": 6.2683,
      "mean_token_accuracy": 0.1345277301967144,
      "num_tokens": 5258541.0,
      "step": 2585
    },
    {
      "entropy": 6.57556848526001,
      "epoch": 0.22633924670103994,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004997821032931676,
      "loss": 6.3723,
      "mean_token_accuracy": 0.1347046300768852,
      "num_tokens": 5268524.0,
      "step": 2590
    },
    {
      "entropy": 6.418022775650025,
      "epoch": 0.22677619505374463,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000499780730076846,
      "loss": 6.2883,
      "mean_token_accuracy": 0.13537805154919624,
      "num_tokens": 5280702.0,
      "step": 2595
    },
    {
      "entropy": 6.522402095794678,
      "epoch": 0.22721314340644935,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997793525491122,
      "loss": 6.2628,
      "mean_token_accuracy": 0.1282507784664631,
      "num_tokens": 5291214.0,
      "step": 2600
    },
    {
      "entropy": 6.40453028678894,
      "epoch": 0.22765009175915407,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997779707099926,
      "loss": 6.2218,
      "mean_token_accuracy": 0.14659756794571877,
      "num_tokens": 5300729.0,
      "step": 2605
    },
    {
      "entropy": 6.47467451095581,
      "epoch": 0.2280870401118588,
      "grad_norm": 0.875,
      "learning_rate": 0.0004997765845595138,
      "loss": 6.3621,
      "mean_token_accuracy": 0.13119488582015038,
      "num_tokens": 5310551.0,
      "step": 2610
    },
    {
      "entropy": 6.483618068695068,
      "epoch": 0.22852398846456348,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997751940977024,
      "loss": 6.2198,
      "mean_token_accuracy": 0.13728965371847152,
      "num_tokens": 5320038.0,
      "step": 2615
    },
    {
      "entropy": 6.453027200698853,
      "epoch": 0.2289609368172682,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997737993245848,
      "loss": 6.2905,
      "mean_token_accuracy": 0.1325733795762062,
      "num_tokens": 5329753.0,
      "step": 2620
    },
    {
      "entropy": 6.422470378875732,
      "epoch": 0.22939788516997292,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997724002401881,
      "loss": 6.2727,
      "mean_token_accuracy": 0.13691200613975524,
      "num_tokens": 5340546.0,
      "step": 2625
    },
    {
      "entropy": 6.40334906578064,
      "epoch": 0.2298348335226776,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997709968445391,
      "loss": 6.3444,
      "mean_token_accuracy": 0.13128848671913146,
      "num_tokens": 5352097.0,
      "step": 2630
    },
    {
      "entropy": 6.561565923690796,
      "epoch": 0.23027178187538233,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997695891376644,
      "loss": 6.2735,
      "mean_token_accuracy": 0.13817958086729049,
      "num_tokens": 5361646.0,
      "step": 2635
    },
    {
      "entropy": 6.418984413146973,
      "epoch": 0.23070873022808705,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997681771195912,
      "loss": 6.2515,
      "mean_token_accuracy": 0.14123460948467254,
      "num_tokens": 5371695.0,
      "step": 2640
    },
    {
      "entropy": 6.445744562149048,
      "epoch": 0.23114567858079174,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004997667607903466,
      "loss": 6.3225,
      "mean_token_accuracy": 0.13134846240282058,
      "num_tokens": 5381978.0,
      "step": 2645
    },
    {
      "entropy": 6.554618072509766,
      "epoch": 0.23158262693349646,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997653401499578,
      "loss": 6.3257,
      "mean_token_accuracy": 0.1353774681687355,
      "num_tokens": 5391891.0,
      "step": 2650
    },
    {
      "entropy": 6.483434438705444,
      "epoch": 0.23201957528620118,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499763915198452,
      "loss": 6.2978,
      "mean_token_accuracy": 0.13159872964024544,
      "num_tokens": 5403172.0,
      "step": 2655
    },
    {
      "entropy": 6.40730848312378,
      "epoch": 0.23245652363890587,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997624859358565,
      "loss": 6.2971,
      "mean_token_accuracy": 0.13805141896009446,
      "num_tokens": 5413130.0,
      "step": 2660
    },
    {
      "entropy": 6.454382991790771,
      "epoch": 0.2328934719916106,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997610523621988,
      "loss": 6.332,
      "mean_token_accuracy": 0.13280672580003738,
      "num_tokens": 5423510.0,
      "step": 2665
    },
    {
      "entropy": 6.487873840332031,
      "epoch": 0.2333304203443153,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997596144775063,
      "loss": 6.1581,
      "mean_token_accuracy": 0.14178016260266305,
      "num_tokens": 5433707.0,
      "step": 2670
    },
    {
      "entropy": 6.299017810821534,
      "epoch": 0.23376736869702,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004997581722818067,
      "loss": 6.2586,
      "mean_token_accuracy": 0.137728713452816,
      "num_tokens": 5444381.0,
      "step": 2675
    },
    {
      "entropy": 6.607342195510864,
      "epoch": 0.23420431704972472,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997567257751275,
      "loss": 6.2167,
      "mean_token_accuracy": 0.13631812259554862,
      "num_tokens": 5453905.0,
      "step": 2680
    },
    {
      "entropy": 6.388966417312622,
      "epoch": 0.23464126540242944,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004997552749574965,
      "loss": 6.2393,
      "mean_token_accuracy": 0.13831414729356767,
      "num_tokens": 5464839.0,
      "step": 2685
    },
    {
      "entropy": 6.426655292510986,
      "epoch": 0.23507821375513413,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997538198289416,
      "loss": 6.2768,
      "mean_token_accuracy": 0.12901907563209533,
      "num_tokens": 5475459.0,
      "step": 2690
    },
    {
      "entropy": 6.431175565719604,
      "epoch": 0.23551516210783885,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997523603894907,
      "loss": 6.285,
      "mean_token_accuracy": 0.1379141092300415,
      "num_tokens": 5484886.0,
      "step": 2695
    },
    {
      "entropy": 6.4651857852935795,
      "epoch": 0.23595211046054357,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997508966391717,
      "loss": 6.271,
      "mean_token_accuracy": 0.13728889897465707,
      "num_tokens": 5494803.0,
      "step": 2700
    },
    {
      "entropy": 6.40422306060791,
      "epoch": 0.2363890588132483,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997494285780129,
      "loss": 6.2203,
      "mean_token_accuracy": 0.13441247269511222,
      "num_tokens": 5504997.0,
      "step": 2705
    },
    {
      "entropy": 6.3673491954803465,
      "epoch": 0.23682600716595298,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997479562060421,
      "loss": 6.2201,
      "mean_token_accuracy": 0.14114237129688262,
      "num_tokens": 5514492.0,
      "step": 2710
    },
    {
      "entropy": 6.491820907592773,
      "epoch": 0.2372629555186577,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997464795232877,
      "loss": 6.2568,
      "mean_token_accuracy": 0.13033155351877213,
      "num_tokens": 5524419.0,
      "step": 2715
    },
    {
      "entropy": 6.461743450164795,
      "epoch": 0.23769990387136242,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997449985297782,
      "loss": 6.282,
      "mean_token_accuracy": 0.13783859759569167,
      "num_tokens": 5534289.0,
      "step": 2720
    },
    {
      "entropy": 6.40246434211731,
      "epoch": 0.2381368522240671,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004997435132255418,
      "loss": 6.2304,
      "mean_token_accuracy": 0.13647418320178986,
      "num_tokens": 5543034.0,
      "step": 2725
    },
    {
      "entropy": 6.480632209777832,
      "epoch": 0.23857380057677183,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997420236106071,
      "loss": 6.2934,
      "mean_token_accuracy": 0.1379154346883297,
      "num_tokens": 5553291.0,
      "step": 2730
    },
    {
      "entropy": 6.460249185562134,
      "epoch": 0.23901074892947655,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004997405296850026,
      "loss": 6.2744,
      "mean_token_accuracy": 0.13468145504593848,
      "num_tokens": 5563772.0,
      "step": 2735
    },
    {
      "entropy": 6.447478199005127,
      "epoch": 0.23944769728218124,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997390314487569,
      "loss": 6.3068,
      "mean_token_accuracy": 0.13267375007271767,
      "num_tokens": 5574348.0,
      "step": 2740
    },
    {
      "entropy": 6.4441228866577145,
      "epoch": 0.23988464563488596,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997375289018989,
      "loss": 6.1637,
      "mean_token_accuracy": 0.13476863205432893,
      "num_tokens": 5583491.0,
      "step": 2745
    },
    {
      "entropy": 6.34928069114685,
      "epoch": 0.24032159398759068,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997360220444574,
      "loss": 6.324,
      "mean_token_accuracy": 0.12789161056280135,
      "num_tokens": 5594273.0,
      "step": 2750
    },
    {
      "entropy": 6.478325939178466,
      "epoch": 0.24075854234029537,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997345108764611,
      "loss": 6.2209,
      "mean_token_accuracy": 0.13871316835284234,
      "num_tokens": 5604529.0,
      "step": 2755
    },
    {
      "entropy": 6.397528219223022,
      "epoch": 0.2411954906930001,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997329953979391,
      "loss": 6.2102,
      "mean_token_accuracy": 0.14513446986675263,
      "num_tokens": 5614036.0,
      "step": 2760
    },
    {
      "entropy": 6.458429145812988,
      "epoch": 0.2416324390457048,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997314756089206,
      "loss": 6.2591,
      "mean_token_accuracy": 0.13734069988131523,
      "num_tokens": 5624036.0,
      "step": 2765
    },
    {
      "entropy": 6.424458265304565,
      "epoch": 0.2420693873984095,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997299515094347,
      "loss": 6.2456,
      "mean_token_accuracy": 0.1391109459102154,
      "num_tokens": 5633987.0,
      "step": 2770
    },
    {
      "entropy": 6.383187675476075,
      "epoch": 0.24250633575111422,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997284230995105,
      "loss": 6.1199,
      "mean_token_accuracy": 0.14982096254825591,
      "num_tokens": 5643905.0,
      "step": 2775
    },
    {
      "entropy": 6.436825084686279,
      "epoch": 0.24294328410381894,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997268903791773,
      "loss": 6.3289,
      "mean_token_accuracy": 0.1315071016550064,
      "num_tokens": 5654922.0,
      "step": 2780
    },
    {
      "entropy": 6.39974889755249,
      "epoch": 0.24338023245652363,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004997253533484647,
      "loss": 6.1586,
      "mean_token_accuracy": 0.1425160974264145,
      "num_tokens": 5664407.0,
      "step": 2785
    },
    {
      "entropy": 6.417563819885254,
      "epoch": 0.24381718080922835,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997238120074021,
      "loss": 6.2272,
      "mean_token_accuracy": 0.13456432819366454,
      "num_tokens": 5674524.0,
      "step": 2790
    },
    {
      "entropy": 6.452868366241455,
      "epoch": 0.24425412916193306,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004997222663560188,
      "loss": 6.2615,
      "mean_token_accuracy": 0.1321194626390934,
      "num_tokens": 5684691.0,
      "step": 2795
    },
    {
      "entropy": 6.3742493152618405,
      "epoch": 0.24469107751463778,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997207163943449,
      "loss": 6.1785,
      "mean_token_accuracy": 0.13999853581190108,
      "num_tokens": 5693683.0,
      "step": 2800
    },
    {
      "entropy": 6.406622314453125,
      "epoch": 0.24512802586734248,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004997191621224098,
      "loss": 6.2671,
      "mean_token_accuracy": 0.13562537357211113,
      "num_tokens": 5704394.0,
      "step": 2805
    },
    {
      "entropy": 6.394683933258056,
      "epoch": 0.2455649742200472,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004997176035402435,
      "loss": 6.1658,
      "mean_token_accuracy": 0.1439206138253212,
      "num_tokens": 5715480.0,
      "step": 2810
    },
    {
      "entropy": 6.465519857406616,
      "epoch": 0.2460019225727519,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004997160406478758,
      "loss": 6.1749,
      "mean_token_accuracy": 0.14369347989559172,
      "num_tokens": 5725973.0,
      "step": 2815
    },
    {
      "entropy": 6.26147494316101,
      "epoch": 0.2464388709254566,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997144734453367,
      "loss": 6.2044,
      "mean_token_accuracy": 0.1463428631424904,
      "num_tokens": 5735798.0,
      "step": 2820
    },
    {
      "entropy": 6.491905450820923,
      "epoch": 0.24687581927816132,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004997129019326562,
      "loss": 6.2764,
      "mean_token_accuracy": 0.12938766777515412,
      "num_tokens": 5746623.0,
      "step": 2825
    },
    {
      "entropy": 6.469847059249878,
      "epoch": 0.24731276763086604,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997113261098645,
      "loss": 6.2206,
      "mean_token_accuracy": 0.14192579537630082,
      "num_tokens": 5756526.0,
      "step": 2830
    },
    {
      "entropy": 6.384304094314575,
      "epoch": 0.24774971598357073,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004997097459769919,
      "loss": 6.2083,
      "mean_token_accuracy": 0.13960244581103326,
      "num_tokens": 5766670.0,
      "step": 2835
    },
    {
      "entropy": 6.459981393814087,
      "epoch": 0.24818666433627545,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004997081615340685,
      "loss": 6.2636,
      "mean_token_accuracy": 0.13670481294393538,
      "num_tokens": 5776473.0,
      "step": 2840
    },
    {
      "entropy": 6.503982400894165,
      "epoch": 0.24862361268898017,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997065727811249,
      "loss": 6.2949,
      "mean_token_accuracy": 0.13849538564682007,
      "num_tokens": 5787059.0,
      "step": 2845
    },
    {
      "entropy": 6.414534044265747,
      "epoch": 0.24906056104168486,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004997049797181915,
      "loss": 6.2884,
      "mean_token_accuracy": 0.1313557967543602,
      "num_tokens": 5797367.0,
      "step": 2850
    },
    {
      "entropy": 6.3486549854278564,
      "epoch": 0.24949750939438958,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997033823452988,
      "loss": 6.1751,
      "mean_token_accuracy": 0.13914304301142694,
      "num_tokens": 5806704.0,
      "step": 2855
    },
    {
      "entropy": 6.40527114868164,
      "epoch": 0.2499344577470943,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004997017806624774,
      "loss": 6.2468,
      "mean_token_accuracy": 0.1393037997186184,
      "num_tokens": 5817797.0,
      "step": 2860
    },
    {
      "entropy": 6.466370344161987,
      "epoch": 0.250371406099799,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997001746697582,
      "loss": 6.2764,
      "mean_token_accuracy": 0.14098547995090485,
      "num_tokens": 5829270.0,
      "step": 2865
    },
    {
      "entropy": 6.405104541778565,
      "epoch": 0.2508083544525037,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004996985643671719,
      "loss": 6.2277,
      "mean_token_accuracy": 0.1410501480102539,
      "num_tokens": 5839214.0,
      "step": 2870
    },
    {
      "entropy": 6.361758279800415,
      "epoch": 0.2512453028052084,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004996969497547495,
      "loss": 6.2165,
      "mean_token_accuracy": 0.13901832550764084,
      "num_tokens": 5848171.0,
      "step": 2875
    },
    {
      "entropy": 6.408371257781982,
      "epoch": 0.25168225115791315,
      "grad_norm": 0.875,
      "learning_rate": 0.0004996953308325217,
      "loss": 6.1341,
      "mean_token_accuracy": 0.14583304151892662,
      "num_tokens": 5858399.0,
      "step": 2880
    },
    {
      "entropy": 6.343305540084839,
      "epoch": 0.25211919951061784,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004996937076005198,
      "loss": 6.1379,
      "mean_token_accuracy": 0.14553209990262986,
      "num_tokens": 5868212.0,
      "step": 2885
    },
    {
      "entropy": 6.380623054504395,
      "epoch": 0.25255614786332253,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004996920800587749,
      "loss": 6.207,
      "mean_token_accuracy": 0.13883035629987717,
      "num_tokens": 5877883.0,
      "step": 2890
    },
    {
      "entropy": 6.433513402938843,
      "epoch": 0.2529930962160273,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004996904482073181,
      "loss": 6.2114,
      "mean_token_accuracy": 0.14031296521425246,
      "num_tokens": 5887090.0,
      "step": 2895
    },
    {
      "entropy": 6.32197265625,
      "epoch": 0.25343004456873197,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996888120461808,
      "loss": 6.2558,
      "mean_token_accuracy": 0.1309097707271576,
      "num_tokens": 5897491.0,
      "step": 2900
    },
    {
      "entropy": 6.420869779586792,
      "epoch": 0.25386699292143666,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004996871715753943,
      "loss": 6.2715,
      "mean_token_accuracy": 0.13663778081536293,
      "num_tokens": 5908494.0,
      "step": 2905
    },
    {
      "entropy": 6.50685076713562,
      "epoch": 0.2543039412741414,
      "grad_norm": 0.875,
      "learning_rate": 0.00049968552679499,
      "loss": 6.3246,
      "mean_token_accuracy": 0.12492352053523063,
      "num_tokens": 5919708.0,
      "step": 2910
    },
    {
      "entropy": 6.412569522857666,
      "epoch": 0.2547408896268461,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004996838777049998,
      "loss": 6.1817,
      "mean_token_accuracy": 0.1387775108218193,
      "num_tokens": 5930122.0,
      "step": 2915
    },
    {
      "entropy": 6.335930967330933,
      "epoch": 0.2551778379795508,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499682224305455,
      "loss": 6.1921,
      "mean_token_accuracy": 0.1372269608080387,
      "num_tokens": 5940033.0,
      "step": 2920
    },
    {
      "entropy": 6.3563535690307615,
      "epoch": 0.25561478633225554,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004996805665963874,
      "loss": 6.1798,
      "mean_token_accuracy": 0.13515137359499932,
      "num_tokens": 5949335.0,
      "step": 2925
    },
    {
      "entropy": 6.468244028091431,
      "epoch": 0.25605173468496023,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004996789045778287,
      "loss": 6.2149,
      "mean_token_accuracy": 0.1392026238143444,
      "num_tokens": 5959299.0,
      "step": 2930
    },
    {
      "entropy": 6.408330249786377,
      "epoch": 0.2564886830376649,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004996772382498111,
      "loss": 6.1833,
      "mean_token_accuracy": 0.13642103523015975,
      "num_tokens": 5970114.0,
      "step": 2935
    },
    {
      "entropy": 6.343743133544922,
      "epoch": 0.25692563139036967,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004996755676123662,
      "loss": 6.281,
      "mean_token_accuracy": 0.13536796644330024,
      "num_tokens": 5979630.0,
      "step": 2940
    },
    {
      "entropy": 6.489995050430298,
      "epoch": 0.25736257974307436,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996738926655262,
      "loss": 6.2612,
      "mean_token_accuracy": 0.13175804391503335,
      "num_tokens": 5990190.0,
      "step": 2945
    },
    {
      "entropy": 6.450542163848877,
      "epoch": 0.25779952809577905,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004996722134093231,
      "loss": 6.2849,
      "mean_token_accuracy": 0.1331914894282818,
      "num_tokens": 6000706.0,
      "step": 2950
    },
    {
      "entropy": 6.346250867843628,
      "epoch": 0.2582364764484838,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004996705298437894,
      "loss": 6.2405,
      "mean_token_accuracy": 0.13917525857686996,
      "num_tokens": 6011274.0,
      "step": 2955
    },
    {
      "entropy": 6.465360450744629,
      "epoch": 0.2586734248011885,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004996688419689572,
      "loss": 6.1539,
      "mean_token_accuracy": 0.148239216953516,
      "num_tokens": 6021481.0,
      "step": 2960
    },
    {
      "entropy": 6.345787000656128,
      "epoch": 0.25911037315389324,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996671497848589,
      "loss": 6.2279,
      "mean_token_accuracy": 0.13937210515141488,
      "num_tokens": 6032091.0,
      "step": 2965
    },
    {
      "entropy": 6.386220216751099,
      "epoch": 0.25954732150659793,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004996654532915269,
      "loss": 6.2324,
      "mean_token_accuracy": 0.13179687783122063,
      "num_tokens": 6043371.0,
      "step": 2970
    },
    {
      "entropy": 6.471800041198731,
      "epoch": 0.2599842698593026,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996637524889937,
      "loss": 6.1863,
      "mean_token_accuracy": 0.13913432955741883,
      "num_tokens": 6052812.0,
      "step": 2975
    },
    {
      "entropy": 6.313665294647217,
      "epoch": 0.26042121821200737,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004996620473772921,
      "loss": 6.1687,
      "mean_token_accuracy": 0.14615661054849624,
      "num_tokens": 6064033.0,
      "step": 2980
    },
    {
      "entropy": 6.398783588409424,
      "epoch": 0.26085816656471206,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004996603379564546,
      "loss": 6.3042,
      "mean_token_accuracy": 0.13552941903471946,
      "num_tokens": 6075837.0,
      "step": 2985
    },
    {
      "entropy": 6.529281282424927,
      "epoch": 0.26129511491741675,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004996586242265142,
      "loss": 6.1871,
      "mean_token_accuracy": 0.14371934831142424,
      "num_tokens": 6085514.0,
      "step": 2990
    },
    {
      "entropy": 6.355500984191894,
      "epoch": 0.2617320632701215,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004996569061875036,
      "loss": 6.1057,
      "mean_token_accuracy": 0.14477267414331435,
      "num_tokens": 6094992.0,
      "step": 2995
    },
    {
      "entropy": 6.338720464706421,
      "epoch": 0.2621690116228262,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004996551838394559,
      "loss": 6.1787,
      "mean_token_accuracy": 0.13958787322044372,
      "num_tokens": 6104841.0,
      "step": 3000
    },
    {
      "epoch": 0.2621690116228262,
      "eval_entropy": 6.140728340206636,
      "eval_loss": 6.207319259643555,
      "eval_mean_token_accuracy": 0.1469711780999768,
      "eval_num_tokens": 6104841.0,
      "eval_runtime": 21.6443,
      "eval_samples_per_second": 1710.102,
      "eval_steps_per_second": 213.774,
      "step": 3000
    },
    {
      "entropy": 6.2954404830932615,
      "epoch": 0.2626059599755309,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499653457182404,
      "loss": 6.0721,
      "mean_token_accuracy": 0.14866723716259003,
      "num_tokens": 6114830.0,
      "step": 3005
    },
    {
      "entropy": 6.330233383178711,
      "epoch": 0.2630429083282356,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004996517262163811,
      "loss": 6.2002,
      "mean_token_accuracy": 0.1417041540145874,
      "num_tokens": 6125165.0,
      "step": 3010
    },
    {
      "entropy": 6.360240745544433,
      "epoch": 0.2634798566809403,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996499909414204,
      "loss": 6.1454,
      "mean_token_accuracy": 0.14034123420715333,
      "num_tokens": 6135670.0,
      "step": 3015
    },
    {
      "entropy": 6.297566843032837,
      "epoch": 0.263916805033645,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004996482513575551,
      "loss": 6.1294,
      "mean_token_accuracy": 0.14982950687408447,
      "num_tokens": 6146055.0,
      "step": 3020
    },
    {
      "entropy": 6.299193429946899,
      "epoch": 0.26435375338634975,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996465074648186,
      "loss": 6.1479,
      "mean_token_accuracy": 0.14521702080965043,
      "num_tokens": 6155941.0,
      "step": 3025
    },
    {
      "entropy": 6.452114486694336,
      "epoch": 0.26479070173905445,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004996447592632444,
      "loss": 6.2433,
      "mean_token_accuracy": 0.1405558854341507,
      "num_tokens": 6166291.0,
      "step": 3030
    },
    {
      "entropy": 6.352116060256958,
      "epoch": 0.26522765009175914,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996430067528661,
      "loss": 6.1996,
      "mean_token_accuracy": 0.13908485621213912,
      "num_tokens": 6175294.0,
      "step": 3035
    },
    {
      "entropy": 6.441894578933716,
      "epoch": 0.2656645984444639,
      "grad_norm": 0.875,
      "learning_rate": 0.0004996412499337172,
      "loss": 6.1799,
      "mean_token_accuracy": 0.14599951803684236,
      "num_tokens": 6185273.0,
      "step": 3040
    },
    {
      "entropy": 6.258168268203735,
      "epoch": 0.2661015467971686,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004996394888058314,
      "loss": 6.1415,
      "mean_token_accuracy": 0.1480218470096588,
      "num_tokens": 6195075.0,
      "step": 3045
    },
    {
      "entropy": 6.221267890930176,
      "epoch": 0.26653849514987327,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004996377233692425,
      "loss": 6.0302,
      "mean_token_accuracy": 0.15916952788829802,
      "num_tokens": 6206018.0,
      "step": 3050
    },
    {
      "entropy": 6.406425666809082,
      "epoch": 0.266975443502578,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004996359536239843,
      "loss": 6.1703,
      "mean_token_accuracy": 0.13967156410217285,
      "num_tokens": 6216577.0,
      "step": 3055
    },
    {
      "entropy": 6.416021108627319,
      "epoch": 0.2674123918552827,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996341795700908,
      "loss": 6.1967,
      "mean_token_accuracy": 0.14080278426408768,
      "num_tokens": 6226521.0,
      "step": 3060
    },
    {
      "entropy": 6.317079877853393,
      "epoch": 0.2678493402079874,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004996324012075961,
      "loss": 6.1758,
      "mean_token_accuracy": 0.14496348276734353,
      "num_tokens": 6236509.0,
      "step": 3065
    },
    {
      "entropy": 6.361893129348755,
      "epoch": 0.26828628856069214,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004996306185365341,
      "loss": 6.161,
      "mean_token_accuracy": 0.14295196756720543,
      "num_tokens": 6246584.0,
      "step": 3070
    },
    {
      "entropy": 6.38335657119751,
      "epoch": 0.26872323691339683,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004996288315569393,
      "loss": 6.184,
      "mean_token_accuracy": 0.1432151563465595,
      "num_tokens": 6256881.0,
      "step": 3075
    },
    {
      "entropy": 6.398205709457398,
      "epoch": 0.2691601852661015,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004996270402688457,
      "loss": 6.146,
      "mean_token_accuracy": 0.14504260420799256,
      "num_tokens": 6266380.0,
      "step": 3080
    },
    {
      "entropy": 6.373401737213134,
      "epoch": 0.2695971336188063,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996252446722877,
      "loss": 6.1964,
      "mean_token_accuracy": 0.1460059978067875,
      "num_tokens": 6274872.0,
      "step": 3085
    },
    {
      "entropy": 6.350259590148926,
      "epoch": 0.27003408197151096,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004996234447673,
      "loss": 6.2075,
      "mean_token_accuracy": 0.14173548221588134,
      "num_tokens": 6286167.0,
      "step": 3090
    },
    {
      "entropy": 6.369588041305542,
      "epoch": 0.27047103032421566,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004996216405539169,
      "loss": 6.156,
      "mean_token_accuracy": 0.1371704325079918,
      "num_tokens": 6295898.0,
      "step": 3095
    },
    {
      "entropy": 6.381024503707886,
      "epoch": 0.2709079786769204,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004996198320321728,
      "loss": 6.1818,
      "mean_token_accuracy": 0.14304329603910446,
      "num_tokens": 6306543.0,
      "step": 3100
    },
    {
      "entropy": 6.351856899261475,
      "epoch": 0.2713449270296251,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004996180192021029,
      "loss": 6.1697,
      "mean_token_accuracy": 0.13672457933425902,
      "num_tokens": 6317709.0,
      "step": 3105
    },
    {
      "entropy": 6.351310873031617,
      "epoch": 0.2717818753823298,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996162020637415,
      "loss": 6.1198,
      "mean_token_accuracy": 0.14543213993310927,
      "num_tokens": 6326874.0,
      "step": 3110
    },
    {
      "entropy": 6.359648990631103,
      "epoch": 0.27221882373503453,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004996143806171238,
      "loss": 6.2203,
      "mean_token_accuracy": 0.13802286610007286,
      "num_tokens": 6337943.0,
      "step": 3115
    },
    {
      "entropy": 6.3505360126495365,
      "epoch": 0.2726557720877392,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996125548622845,
      "loss": 6.2814,
      "mean_token_accuracy": 0.1384304866194725,
      "num_tokens": 6348442.0,
      "step": 3120
    },
    {
      "entropy": 6.348107242584229,
      "epoch": 0.2730927204404439,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004996107247992587,
      "loss": 6.1768,
      "mean_token_accuracy": 0.14265852496027948,
      "num_tokens": 6358675.0,
      "step": 3125
    },
    {
      "entropy": 6.382030820846557,
      "epoch": 0.27352966879314866,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004996088904280815,
      "loss": 6.2168,
      "mean_token_accuracy": 0.13871825635433196,
      "num_tokens": 6368631.0,
      "step": 3130
    },
    {
      "entropy": 6.371310091018676,
      "epoch": 0.27396661714585335,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004996070517487881,
      "loss": 6.0773,
      "mean_token_accuracy": 0.14500803351402283,
      "num_tokens": 6378709.0,
      "step": 3135
    },
    {
      "entropy": 6.318156099319458,
      "epoch": 0.27440356549855804,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004996052087614138,
      "loss": 6.2322,
      "mean_token_accuracy": 0.13356006070971488,
      "num_tokens": 6389686.0,
      "step": 3140
    },
    {
      "entropy": 6.3309718608856205,
      "epoch": 0.2748405138512628,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004996033614659938,
      "loss": 6.1964,
      "mean_token_accuracy": 0.13752286732196808,
      "num_tokens": 6399447.0,
      "step": 3145
    },
    {
      "entropy": 6.345420885086059,
      "epoch": 0.2752774622039675,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996015098625638,
      "loss": 6.1715,
      "mean_token_accuracy": 0.14288139641284942,
      "num_tokens": 6408891.0,
      "step": 3150
    },
    {
      "entropy": 6.360024499893188,
      "epoch": 0.2757144105566722,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499599653951159,
      "loss": 6.144,
      "mean_token_accuracy": 0.1456197828054428,
      "num_tokens": 6419471.0,
      "step": 3155
    },
    {
      "entropy": 6.319950771331787,
      "epoch": 0.2761513589093769,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004995977937318153,
      "loss": 6.1237,
      "mean_token_accuracy": 0.14010287821292877,
      "num_tokens": 6430445.0,
      "step": 3160
    },
    {
      "entropy": 6.3929389953613285,
      "epoch": 0.2765883072620816,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000499595929204568,
      "loss": 6.1615,
      "mean_token_accuracy": 0.14165643975138664,
      "num_tokens": 6440821.0,
      "step": 3165
    },
    {
      "entropy": 6.299115943908691,
      "epoch": 0.27702525561478636,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004995940603694533,
      "loss": 6.0938,
      "mean_token_accuracy": 0.14386325478553771,
      "num_tokens": 6450958.0,
      "step": 3170
    },
    {
      "entropy": 6.301024341583252,
      "epoch": 0.27746220396749105,
      "grad_norm": 0.875,
      "learning_rate": 0.0004995921872265067,
      "loss": 6.1834,
      "mean_token_accuracy": 0.14242049753665925,
      "num_tokens": 6462106.0,
      "step": 3175
    },
    {
      "entropy": 6.328290796279907,
      "epoch": 0.27789915232019574,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995903097757643,
      "loss": 6.0674,
      "mean_token_accuracy": 0.15150198489427566,
      "num_tokens": 6471890.0,
      "step": 3180
    },
    {
      "entropy": 6.29628005027771,
      "epoch": 0.2783361006729005,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004995884280172622,
      "loss": 6.0887,
      "mean_token_accuracy": 0.1439910836517811,
      "num_tokens": 6481959.0,
      "step": 3185
    },
    {
      "entropy": 6.357741498947144,
      "epoch": 0.2787730490256052,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995865419510362,
      "loss": 6.1264,
      "mean_token_accuracy": 0.14735689684748648,
      "num_tokens": 6492086.0,
      "step": 3190
    },
    {
      "entropy": 6.274436140060425,
      "epoch": 0.27920999737830987,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004995846515771227,
      "loss": 6.16,
      "mean_token_accuracy": 0.14143534004688263,
      "num_tokens": 6502530.0,
      "step": 3195
    },
    {
      "entropy": 6.367669486999512,
      "epoch": 0.2796469457310146,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499582756895558,
      "loss": 6.161,
      "mean_token_accuracy": 0.14963102713227272,
      "num_tokens": 6511600.0,
      "step": 3200
    },
    {
      "entropy": 6.2996594429016115,
      "epoch": 0.2800838940837193,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004995808579063782,
      "loss": 6.1345,
      "mean_token_accuracy": 0.13933146744966507,
      "num_tokens": 6522712.0,
      "step": 3205
    },
    {
      "entropy": 6.15234899520874,
      "epoch": 0.280520842436424,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00049957895460962,
      "loss": 6.0024,
      "mean_token_accuracy": 0.1613503262400627,
      "num_tokens": 6532472.0,
      "step": 3210
    },
    {
      "entropy": 6.329534006118775,
      "epoch": 0.28095779078912875,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004995770470053197,
      "loss": 6.0545,
      "mean_token_accuracy": 0.15718746110796927,
      "num_tokens": 6541778.0,
      "step": 3215
    },
    {
      "entropy": 6.323871088027954,
      "epoch": 0.28139473914183344,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004995751350935139,
      "loss": 6.0655,
      "mean_token_accuracy": 0.14491002410650253,
      "num_tokens": 6552368.0,
      "step": 3220
    },
    {
      "entropy": 6.281246185302734,
      "epoch": 0.28183168749453813,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004995732188742395,
      "loss": 6.0917,
      "mean_token_accuracy": 0.14650973826646804,
      "num_tokens": 6562593.0,
      "step": 3225
    },
    {
      "entropy": 6.327645111083984,
      "epoch": 0.2822686358472429,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004995712983475329,
      "loss": 6.1427,
      "mean_token_accuracy": 0.14498995393514633,
      "num_tokens": 6572695.0,
      "step": 3230
    },
    {
      "entropy": 6.225468492507934,
      "epoch": 0.28270558419994757,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004995693735134313,
      "loss": 6.1144,
      "mean_token_accuracy": 0.1426160916686058,
      "num_tokens": 6581671.0,
      "step": 3235
    },
    {
      "entropy": 6.335493373870849,
      "epoch": 0.28314253255265226,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995674443719713,
      "loss": 6.1229,
      "mean_token_accuracy": 0.14055858105421065,
      "num_tokens": 6590460.0,
      "step": 3240
    },
    {
      "entropy": 6.303600740432739,
      "epoch": 0.283579480905357,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00049956551092319,
      "loss": 6.1244,
      "mean_token_accuracy": 0.1433852382004261,
      "num_tokens": 6599892.0,
      "step": 3245
    },
    {
      "entropy": 6.354545307159424,
      "epoch": 0.2840164292580617,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995635731671247,
      "loss": 6.2012,
      "mean_token_accuracy": 0.13369778245687486,
      "num_tokens": 6610530.0,
      "step": 3250
    },
    {
      "entropy": 6.2983636379241945,
      "epoch": 0.2844533776107664,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995616311038124,
      "loss": 6.1237,
      "mean_token_accuracy": 0.14690847247838973,
      "num_tokens": 6621117.0,
      "step": 3255
    },
    {
      "entropy": 6.379088354110718,
      "epoch": 0.28489032596347114,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004995596847332901,
      "loss": 6.1595,
      "mean_token_accuracy": 0.14022180885076524,
      "num_tokens": 6631521.0,
      "step": 3260
    },
    {
      "entropy": 6.423089265823364,
      "epoch": 0.2853272743161758,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995577340555956,
      "loss": 6.247,
      "mean_token_accuracy": 0.13523154854774475,
      "num_tokens": 6642180.0,
      "step": 3265
    },
    {
      "entropy": 6.334853219985962,
      "epoch": 0.2857642226688805,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004995557790707661,
      "loss": 6.1713,
      "mean_token_accuracy": 0.1466689735651016,
      "num_tokens": 6652735.0,
      "step": 3270
    },
    {
      "entropy": 6.3520190715789795,
      "epoch": 0.28620117102158527,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499553819778839,
      "loss": 6.16,
      "mean_token_accuracy": 0.14624010920524597,
      "num_tokens": 6663135.0,
      "step": 3275
    },
    {
      "entropy": 6.2731767177581785,
      "epoch": 0.28663811937428996,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499551856179852,
      "loss": 6.1412,
      "mean_token_accuracy": 0.1431369364261627,
      "num_tokens": 6673375.0,
      "step": 3280
    },
    {
      "entropy": 6.350570297241211,
      "epoch": 0.28707506772699465,
      "grad_norm": 0.875,
      "learning_rate": 0.0004995498882738428,
      "loss": 6.1158,
      "mean_token_accuracy": 0.14772647768259048,
      "num_tokens": 6683101.0,
      "step": 3285
    },
    {
      "entropy": 6.329085493087769,
      "epoch": 0.2875120160796994,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004995479160608491,
      "loss": 6.173,
      "mean_token_accuracy": 0.14756054282188416,
      "num_tokens": 6692763.0,
      "step": 3290
    },
    {
      "entropy": 6.351175737380982,
      "epoch": 0.2879489644324041,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995459395409085,
      "loss": 6.155,
      "mean_token_accuracy": 0.1399838872253895,
      "num_tokens": 6702833.0,
      "step": 3295
    },
    {
      "entropy": 6.3768007278442385,
      "epoch": 0.2883859127851088,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004995439587140593,
      "loss": 6.1495,
      "mean_token_accuracy": 0.13922087401151656,
      "num_tokens": 6713083.0,
      "step": 3300
    },
    {
      "entropy": 6.256941509246826,
      "epoch": 0.2888228611378135,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004995419735803392,
      "loss": 6.0994,
      "mean_token_accuracy": 0.14441604316234588,
      "num_tokens": 6722501.0,
      "step": 3305
    },
    {
      "entropy": 6.345928049087524,
      "epoch": 0.2892598094905182,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004995399841397865,
      "loss": 6.1313,
      "mean_token_accuracy": 0.1453465297818184,
      "num_tokens": 6731961.0,
      "step": 3310
    },
    {
      "entropy": 6.358537864685059,
      "epoch": 0.2896967578432229,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004995379903924392,
      "loss": 6.1214,
      "mean_token_accuracy": 0.1412414386868477,
      "num_tokens": 6742576.0,
      "step": 3315
    },
    {
      "entropy": 6.313660049438477,
      "epoch": 0.29013370619592765,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004995359923383356,
      "loss": 6.1247,
      "mean_token_accuracy": 0.15020241886377333,
      "num_tokens": 6753066.0,
      "step": 3320
    },
    {
      "entropy": 6.3176545143127445,
      "epoch": 0.29057065454863235,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499533989977514,
      "loss": 6.2618,
      "mean_token_accuracy": 0.1356256164610386,
      "num_tokens": 6764455.0,
      "step": 3325
    },
    {
      "entropy": 6.375495719909668,
      "epoch": 0.29100760290133704,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004995319833100128,
      "loss": 6.1293,
      "mean_token_accuracy": 0.1450370132923126,
      "num_tokens": 6774655.0,
      "step": 3330
    },
    {
      "entropy": 6.2986647605896,
      "epoch": 0.2914445512540418,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004995299723358705,
      "loss": 6.1021,
      "mean_token_accuracy": 0.15024321302771568,
      "num_tokens": 6785942.0,
      "step": 3335
    },
    {
      "entropy": 6.265730381011963,
      "epoch": 0.2918814996067465,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004995279570551257,
      "loss": 6.0624,
      "mean_token_accuracy": 0.1497375935316086,
      "num_tokens": 6795355.0,
      "step": 3340
    },
    {
      "entropy": 6.296811914443969,
      "epoch": 0.29231844795945117,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499525937467817,
      "loss": 6.1472,
      "mean_token_accuracy": 0.14276159554719925,
      "num_tokens": 6805579.0,
      "step": 3345
    },
    {
      "entropy": 6.325125026702881,
      "epoch": 0.2927553963121559,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004995239135739833,
      "loss": 6.1636,
      "mean_token_accuracy": 0.14122274965047837,
      "num_tokens": 6814878.0,
      "step": 3350
    },
    {
      "entropy": 6.344474601745605,
      "epoch": 0.2931923446648606,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004995218853736631,
      "loss": 6.055,
      "mean_token_accuracy": 0.15251176059246063,
      "num_tokens": 6824533.0,
      "step": 3355
    },
    {
      "entropy": 6.125179576873779,
      "epoch": 0.29362929301756535,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995198528668956,
      "loss": 5.9683,
      "mean_token_accuracy": 0.159256049990654,
      "num_tokens": 6835529.0,
      "step": 3360
    },
    {
      "entropy": 6.3220854759216305,
      "epoch": 0.29406624137027004,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004995178160537197,
      "loss": 6.1517,
      "mean_token_accuracy": 0.14569269567728044,
      "num_tokens": 6845679.0,
      "step": 3365
    },
    {
      "entropy": 6.354899740219116,
      "epoch": 0.29450318972297473,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004995157749341744,
      "loss": 6.1334,
      "mean_token_accuracy": 0.13885997533798217,
      "num_tokens": 6855876.0,
      "step": 3370
    },
    {
      "entropy": 6.297861623764038,
      "epoch": 0.2949401380756795,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004995137295082989,
      "loss": 6.2052,
      "mean_token_accuracy": 0.1401164412498474,
      "num_tokens": 6866200.0,
      "step": 3375
    },
    {
      "entropy": 6.359179210662842,
      "epoch": 0.2953770864283842,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004995116797761324,
      "loss": 6.0778,
      "mean_token_accuracy": 0.14605300948023797,
      "num_tokens": 6875191.0,
      "step": 3380
    },
    {
      "entropy": 6.273428821563721,
      "epoch": 0.29581403478108886,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004995096257377143,
      "loss": 6.0599,
      "mean_token_accuracy": 0.14512585699558259,
      "num_tokens": 6885352.0,
      "step": 3385
    },
    {
      "entropy": 6.30303840637207,
      "epoch": 0.2962509831337936,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004995075673930839,
      "loss": 6.1311,
      "mean_token_accuracy": 0.14580143094062806,
      "num_tokens": 6895585.0,
      "step": 3390
    },
    {
      "entropy": 6.382786750793457,
      "epoch": 0.2966879314864983,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004995055047422808,
      "loss": 6.1014,
      "mean_token_accuracy": 0.14060337394475936,
      "num_tokens": 6905532.0,
      "step": 3395
    },
    {
      "entropy": 6.158539295196533,
      "epoch": 0.297124879839203,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004995034377853443,
      "loss": 6.1029,
      "mean_token_accuracy": 0.14624154269695283,
      "num_tokens": 6915835.0,
      "step": 3400
    },
    {
      "entropy": 6.280689668655396,
      "epoch": 0.29756182819190774,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004995013665223143,
      "loss": 6.1525,
      "mean_token_accuracy": 0.14352616518735886,
      "num_tokens": 6925964.0,
      "step": 3405
    },
    {
      "entropy": 6.314380741119384,
      "epoch": 0.29799877654461243,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004994992909532305,
      "loss": 6.091,
      "mean_token_accuracy": 0.14589581415057182,
      "num_tokens": 6936145.0,
      "step": 3410
    },
    {
      "entropy": 6.29896092414856,
      "epoch": 0.2984357248973171,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004994972110781326,
      "loss": 6.0656,
      "mean_token_accuracy": 0.14432404488325118,
      "num_tokens": 6945557.0,
      "step": 3415
    },
    {
      "entropy": 6.251769351959228,
      "epoch": 0.29887267325002187,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004994951268970606,
      "loss": 6.091,
      "mean_token_accuracy": 0.1475279785692692,
      "num_tokens": 6955147.0,
      "step": 3420
    },
    {
      "entropy": 6.284211874008179,
      "epoch": 0.29930962160272656,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004994930384100545,
      "loss": 6.059,
      "mean_token_accuracy": 0.14195767417550087,
      "num_tokens": 6964024.0,
      "step": 3425
    },
    {
      "entropy": 6.276056480407715,
      "epoch": 0.29974656995543125,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004994909456171541,
      "loss": 6.1159,
      "mean_token_accuracy": 0.1480628401041031,
      "num_tokens": 6974399.0,
      "step": 3430
    },
    {
      "entropy": 6.257720851898194,
      "epoch": 0.300183518308136,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004994888485183998,
      "loss": 6.0676,
      "mean_token_accuracy": 0.14755637869238852,
      "num_tokens": 6984324.0,
      "step": 3435
    },
    {
      "entropy": 6.278302526473999,
      "epoch": 0.3006204666608407,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499486747113832,
      "loss": 6.0759,
      "mean_token_accuracy": 0.15024591833353043,
      "num_tokens": 6995659.0,
      "step": 3440
    },
    {
      "entropy": 6.34593620300293,
      "epoch": 0.3010574150135454,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004994846414034905,
      "loss": 6.1757,
      "mean_token_accuracy": 0.13790618106722832,
      "num_tokens": 7006463.0,
      "step": 3445
    },
    {
      "entropy": 6.315510845184326,
      "epoch": 0.30149436336625013,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499482531387416,
      "loss": 6.0778,
      "mean_token_accuracy": 0.1473211646080017,
      "num_tokens": 7016188.0,
      "step": 3450
    },
    {
      "entropy": 6.284571266174316,
      "epoch": 0.3019313117189548,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004994804170656489,
      "loss": 6.1437,
      "mean_token_accuracy": 0.14328745901584625,
      "num_tokens": 7028435.0,
      "step": 3455
    },
    {
      "entropy": 6.376664781570435,
      "epoch": 0.3023682600716595,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004994782984382299,
      "loss": 6.1132,
      "mean_token_accuracy": 0.14584433510899544,
      "num_tokens": 7037992.0,
      "step": 3460
    },
    {
      "entropy": 6.252527952194214,
      "epoch": 0.30280520842436426,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004994761755051995,
      "loss": 6.1248,
      "mean_token_accuracy": 0.14423885941505432,
      "num_tokens": 7048743.0,
      "step": 3465
    },
    {
      "entropy": 6.2551322937011715,
      "epoch": 0.30324215677706895,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004994740482665982,
      "loss": 6.0851,
      "mean_token_accuracy": 0.14877668246626854,
      "num_tokens": 7059524.0,
      "step": 3470
    },
    {
      "entropy": 6.315831565856934,
      "epoch": 0.30367910512977364,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004994719167224672,
      "loss": 6.0897,
      "mean_token_accuracy": 0.1395334392786026,
      "num_tokens": 7069742.0,
      "step": 3475
    },
    {
      "entropy": 6.26900954246521,
      "epoch": 0.3041160534824784,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004994697808728472,
      "loss": 6.1423,
      "mean_token_accuracy": 0.1453932225704193,
      "num_tokens": 7079909.0,
      "step": 3480
    },
    {
      "entropy": 6.348114776611328,
      "epoch": 0.3045530018351831,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004994676407177791,
      "loss": 6.1526,
      "mean_token_accuracy": 0.14140928015112877,
      "num_tokens": 7089136.0,
      "step": 3485
    },
    {
      "entropy": 6.288559532165527,
      "epoch": 0.30498995018788777,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004994654962573043,
      "loss": 6.1544,
      "mean_token_accuracy": 0.14411381259560585,
      "num_tokens": 7099621.0,
      "step": 3490
    },
    {
      "entropy": 6.291727304458618,
      "epoch": 0.3054268985405925,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004994633474914634,
      "loss": 6.0123,
      "mean_token_accuracy": 0.15279096812009813,
      "num_tokens": 7109599.0,
      "step": 3495
    },
    {
      "entropy": 6.285858392715454,
      "epoch": 0.3058638468932972,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004994611944202979,
      "loss": 6.0568,
      "mean_token_accuracy": 0.15079958140850067,
      "num_tokens": 7120733.0,
      "step": 3500
    },
    {
      "entropy": 6.276545238494873,
      "epoch": 0.3063007952460019,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004994590370438492,
      "loss": 6.136,
      "mean_token_accuracy": 0.1393118381500244,
      "num_tokens": 7129871.0,
      "step": 3505
    },
    {
      "entropy": 6.256414079666138,
      "epoch": 0.30673774359870665,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004994568753621585,
      "loss": 6.0527,
      "mean_token_accuracy": 0.14701383113861083,
      "num_tokens": 7139167.0,
      "step": 3510
    },
    {
      "entropy": 6.227274465560913,
      "epoch": 0.30717469195141134,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004994547093752673,
      "loss": 6.1219,
      "mean_token_accuracy": 0.13770676925778388,
      "num_tokens": 7150534.0,
      "step": 3515
    },
    {
      "entropy": 6.320562839508057,
      "epoch": 0.30761164030411603,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004994525390832171,
      "loss": 6.0774,
      "mean_token_accuracy": 0.14651177376508712,
      "num_tokens": 7160472.0,
      "step": 3520
    },
    {
      "entropy": 6.295433425903321,
      "epoch": 0.3080485886568208,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004994503644860497,
      "loss": 6.103,
      "mean_token_accuracy": 0.1437145009636879,
      "num_tokens": 7171636.0,
      "step": 3525
    },
    {
      "entropy": 6.27422308921814,
      "epoch": 0.30848553700952547,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004994481855838067,
      "loss": 6.0513,
      "mean_token_accuracy": 0.14762744009494783,
      "num_tokens": 7182163.0,
      "step": 3530
    },
    {
      "entropy": 6.206560754776001,
      "epoch": 0.30892248536223016,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004994460023765298,
      "loss": 6.0918,
      "mean_token_accuracy": 0.14782901704311371,
      "num_tokens": 7191694.0,
      "step": 3535
    },
    {
      "entropy": 6.269797325134277,
      "epoch": 0.3093594337149349,
      "grad_norm": 0.75390625,
      "learning_rate": 0.000499443814864261,
      "loss": 6.1028,
      "mean_token_accuracy": 0.14703719019889833,
      "num_tokens": 7202813.0,
      "step": 3540
    },
    {
      "entropy": 6.328323841094971,
      "epoch": 0.3097963820676396,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004994416230470422,
      "loss": 6.0943,
      "mean_token_accuracy": 0.1497291296720505,
      "num_tokens": 7213226.0,
      "step": 3545
    },
    {
      "entropy": 6.286290502548217,
      "epoch": 0.3102333304203443,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004994394269249155,
      "loss": 6.1446,
      "mean_token_accuracy": 0.13995824977755547,
      "num_tokens": 7223520.0,
      "step": 3550
    },
    {
      "entropy": 6.266398096084595,
      "epoch": 0.31067027877304904,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004994372264979231,
      "loss": 6.0356,
      "mean_token_accuracy": 0.15077984631061553,
      "num_tokens": 7232651.0,
      "step": 3555
    },
    {
      "entropy": 6.315041875839233,
      "epoch": 0.3111072271257537,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004994350217661069,
      "loss": 6.1171,
      "mean_token_accuracy": 0.14419228360056877,
      "num_tokens": 7242064.0,
      "step": 3560
    },
    {
      "entropy": 6.287273693084717,
      "epoch": 0.3115441754784585,
      "grad_norm": 0.875,
      "learning_rate": 0.0004994328127295095,
      "loss": 6.136,
      "mean_token_accuracy": 0.1410224989056587,
      "num_tokens": 7253048.0,
      "step": 3565
    },
    {
      "entropy": 6.305558681488037,
      "epoch": 0.31198112383116317,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004994305993881731,
      "loss": 6.079,
      "mean_token_accuracy": 0.14993986338377,
      "num_tokens": 7262516.0,
      "step": 3570
    },
    {
      "entropy": 6.22655839920044,
      "epoch": 0.31241807218386786,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004994283817421402,
      "loss": 6.0066,
      "mean_token_accuracy": 0.15258399695158004,
      "num_tokens": 7272079.0,
      "step": 3575
    },
    {
      "entropy": 6.220291090011597,
      "epoch": 0.3128550205365726,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004994261597914534,
      "loss": 5.9878,
      "mean_token_accuracy": 0.15598128288984298,
      "num_tokens": 7282150.0,
      "step": 3580
    },
    {
      "entropy": 6.26607666015625,
      "epoch": 0.3132919688892773,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004994239335361552,
      "loss": 6.0999,
      "mean_token_accuracy": 0.14056497439742088,
      "num_tokens": 7291782.0,
      "step": 3585
    },
    {
      "entropy": 6.293278360366822,
      "epoch": 0.313728917241982,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004994217029762884,
      "loss": 6.0599,
      "mean_token_accuracy": 0.14566777795553207,
      "num_tokens": 7301472.0,
      "step": 3590
    },
    {
      "entropy": 6.261723375320434,
      "epoch": 0.31416586559468673,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004994194681118957,
      "loss": 6.0321,
      "mean_token_accuracy": 0.1512453258037567,
      "num_tokens": 7311456.0,
      "step": 3595
    },
    {
      "entropy": 6.295667886734009,
      "epoch": 0.3146028139473914,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00049941722894302,
      "loss": 6.1232,
      "mean_token_accuracy": 0.14058786481618882,
      "num_tokens": 7321106.0,
      "step": 3600
    },
    {
      "entropy": 6.266748237609863,
      "epoch": 0.3150397623000961,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004994149854697042,
      "loss": 6.1111,
      "mean_token_accuracy": 0.14695683643221855,
      "num_tokens": 7331232.0,
      "step": 3605
    },
    {
      "entropy": 6.314934539794922,
      "epoch": 0.31547671065280086,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004994127376919916,
      "loss": 6.0965,
      "mean_token_accuracy": 0.14339785426855087,
      "num_tokens": 7341560.0,
      "step": 3610
    },
    {
      "entropy": 6.326571130752564,
      "epoch": 0.31591365900550555,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004994104856099249,
      "loss": 6.0868,
      "mean_token_accuracy": 0.14662182778120042,
      "num_tokens": 7350923.0,
      "step": 3615
    },
    {
      "entropy": 6.197239208221435,
      "epoch": 0.31635060735821025,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004994082292235476,
      "loss": 5.9859,
      "mean_token_accuracy": 0.15532282441854478,
      "num_tokens": 7360121.0,
      "step": 3620
    },
    {
      "entropy": 6.209159564971924,
      "epoch": 0.316787555710915,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004994059685329027,
      "loss": 6.0856,
      "mean_token_accuracy": 0.14521137177944182,
      "num_tokens": 7370823.0,
      "step": 3625
    },
    {
      "entropy": 6.269903326034546,
      "epoch": 0.3172245040636197,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000499403703538034,
      "loss": 6.0541,
      "mean_token_accuracy": 0.1409251481294632,
      "num_tokens": 7380102.0,
      "step": 3630
    },
    {
      "entropy": 6.256821346282959,
      "epoch": 0.3176614524163244,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004994014342389845,
      "loss": 6.0329,
      "mean_token_accuracy": 0.14819468557834625,
      "num_tokens": 7389909.0,
      "step": 3635
    },
    {
      "entropy": 6.277596092224121,
      "epoch": 0.3180984007690291,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004993991606357979,
      "loss": 6.1144,
      "mean_token_accuracy": 0.1460643008351326,
      "num_tokens": 7401092.0,
      "step": 3640
    },
    {
      "entropy": 6.189544677734375,
      "epoch": 0.3185353491217338,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004993968827285179,
      "loss": 6.026,
      "mean_token_accuracy": 0.14508207887411118,
      "num_tokens": 7412155.0,
      "step": 3645
    },
    {
      "entropy": 6.2149989128112795,
      "epoch": 0.3189722974744385,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004993946005171881,
      "loss": 5.9921,
      "mean_token_accuracy": 0.153763547539711,
      "num_tokens": 7421981.0,
      "step": 3650
    },
    {
      "entropy": 6.267820215225219,
      "epoch": 0.31940924582714325,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004993923140018523,
      "loss": 6.0067,
      "mean_token_accuracy": 0.15336575210094452,
      "num_tokens": 7432298.0,
      "step": 3655
    },
    {
      "entropy": 6.185194301605224,
      "epoch": 0.31984619417984794,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004993900231825542,
      "loss": 6.0497,
      "mean_token_accuracy": 0.15190931037068367,
      "num_tokens": 7442575.0,
      "step": 3660
    },
    {
      "entropy": 6.248135757446289,
      "epoch": 0.32028314253255263,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004993877280593381,
      "loss": 6.0386,
      "mean_token_accuracy": 0.15646653175354003,
      "num_tokens": 7453072.0,
      "step": 3665
    },
    {
      "entropy": 6.271560096740723,
      "epoch": 0.3207200908852574,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004993854286322477,
      "loss": 5.9975,
      "mean_token_accuracy": 0.1494731992483139,
      "num_tokens": 7463173.0,
      "step": 3670
    },
    {
      "entropy": 6.215144538879395,
      "epoch": 0.32115703923796207,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004993831249013272,
      "loss": 6.1129,
      "mean_token_accuracy": 0.1444126322865486,
      "num_tokens": 7473751.0,
      "step": 3675
    },
    {
      "entropy": 6.231976127624511,
      "epoch": 0.32159398759066676,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004993808168666207,
      "loss": 5.9103,
      "mean_token_accuracy": 0.15974051058292388,
      "num_tokens": 7483233.0,
      "step": 3680
    },
    {
      "entropy": 6.229567003250122,
      "epoch": 0.3220309359433715,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004993785045281726,
      "loss": 6.0487,
      "mean_token_accuracy": 0.14229854047298432,
      "num_tokens": 7493152.0,
      "step": 3685
    },
    {
      "entropy": 6.293207740783691,
      "epoch": 0.3224678842960762,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004993761878860272,
      "loss": 6.008,
      "mean_token_accuracy": 0.15029292106628417,
      "num_tokens": 7502463.0,
      "step": 3690
    },
    {
      "entropy": 6.283467102050781,
      "epoch": 0.3229048326487809,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000499373866940229,
      "loss": 6.0655,
      "mean_token_accuracy": 0.15074712336063384,
      "num_tokens": 7512301.0,
      "step": 3695
    },
    {
      "entropy": 6.14256911277771,
      "epoch": 0.32334178100148564,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004993715416908225,
      "loss": 6.0948,
      "mean_token_accuracy": 0.14658709317445756,
      "num_tokens": 7522591.0,
      "step": 3700
    },
    {
      "entropy": 6.322413969039917,
      "epoch": 0.32377872935419033,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004993692121378521,
      "loss": 6.0879,
      "mean_token_accuracy": 0.1442805364727974,
      "num_tokens": 7532400.0,
      "step": 3705
    },
    {
      "entropy": 6.263008069992066,
      "epoch": 0.324215677706895,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004993668782813629,
      "loss": 6.0059,
      "mean_token_accuracy": 0.14632104933261872,
      "num_tokens": 7542163.0,
      "step": 3710
    },
    {
      "entropy": 6.202136135101318,
      "epoch": 0.32465262605959977,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004993645401213992,
      "loss": 5.9957,
      "mean_token_accuracy": 0.1582685887813568,
      "num_tokens": 7551638.0,
      "step": 3715
    },
    {
      "entropy": 6.236076211929321,
      "epoch": 0.32508957441230446,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004993621976580061,
      "loss": 6.0695,
      "mean_token_accuracy": 0.1474739357829094,
      "num_tokens": 7561230.0,
      "step": 3720
    },
    {
      "entropy": 6.2406653881073,
      "epoch": 0.32552652276500915,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004993598508912285,
      "loss": 5.9482,
      "mean_token_accuracy": 0.15776676088571548,
      "num_tokens": 7570962.0,
      "step": 3725
    },
    {
      "entropy": 6.198595762252808,
      "epoch": 0.3259634711177139,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004993574998211114,
      "loss": 6.13,
      "mean_token_accuracy": 0.1489112690091133,
      "num_tokens": 7581454.0,
      "step": 3730
    },
    {
      "entropy": 6.233174753189087,
      "epoch": 0.3264004194704186,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004993551444476998,
      "loss": 6.001,
      "mean_token_accuracy": 0.15145405381917953,
      "num_tokens": 7591497.0,
      "step": 3735
    },
    {
      "entropy": 6.271373987197876,
      "epoch": 0.3268373678231233,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004993527847710392,
      "loss": 6.008,
      "mean_token_accuracy": 0.15311285704374314,
      "num_tokens": 7600719.0,
      "step": 3740
    },
    {
      "entropy": 6.20383505821228,
      "epoch": 0.32727431617582803,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004993504207911744,
      "loss": 6.04,
      "mean_token_accuracy": 0.15261640399694443,
      "num_tokens": 7610550.0,
      "step": 3745
    },
    {
      "entropy": 6.275401020050049,
      "epoch": 0.3277112645285327,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004993480525081511,
      "loss": 6.0498,
      "mean_token_accuracy": 0.15051608979701997,
      "num_tokens": 7620472.0,
      "step": 3750
    },
    {
      "entropy": 6.2765693187713625,
      "epoch": 0.3281482128812374,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004993456799220145,
      "loss": 5.9904,
      "mean_token_accuracy": 0.146634604036808,
      "num_tokens": 7630354.0,
      "step": 3755
    },
    {
      "entropy": 6.198559522628784,
      "epoch": 0.32858516123394216,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004993433030328103,
      "loss": 6.0309,
      "mean_token_accuracy": 0.14934386909008027,
      "num_tokens": 7639689.0,
      "step": 3760
    },
    {
      "entropy": 6.244138765335083,
      "epoch": 0.32902210958664685,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004993409218405839,
      "loss": 6.0687,
      "mean_token_accuracy": 0.1417417049407959,
      "num_tokens": 7649319.0,
      "step": 3765
    },
    {
      "entropy": 6.326660203933716,
      "epoch": 0.3294590579393516,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004993385363453812,
      "loss": 6.0627,
      "mean_token_accuracy": 0.1466643713414669,
      "num_tokens": 7659008.0,
      "step": 3770
    },
    {
      "entropy": 6.22724666595459,
      "epoch": 0.3298960062920563,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004993361465472477,
      "loss": 6.0084,
      "mean_token_accuracy": 0.15965658724308013,
      "num_tokens": 7669064.0,
      "step": 3775
    },
    {
      "entropy": 6.2213051319122314,
      "epoch": 0.330332954644761,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004993337524462295,
      "loss": 6.0794,
      "mean_token_accuracy": 0.1521337777376175,
      "num_tokens": 7679122.0,
      "step": 3780
    },
    {
      "entropy": 6.1721875190734865,
      "epoch": 0.3307699029974657,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004993313540423722,
      "loss": 6.0814,
      "mean_token_accuracy": 0.14944267868995667,
      "num_tokens": 7688373.0,
      "step": 3785
    },
    {
      "entropy": 6.261203241348267,
      "epoch": 0.3312068513501704,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004993289513357221,
      "loss": 6.0095,
      "mean_token_accuracy": 0.15269952714443208,
      "num_tokens": 7697879.0,
      "step": 3790
    },
    {
      "entropy": 6.238248348236084,
      "epoch": 0.3316437997028751,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004993265443263251,
      "loss": 6.0476,
      "mean_token_accuracy": 0.14502775445580482,
      "num_tokens": 7708104.0,
      "step": 3795
    },
    {
      "entropy": 6.258405303955078,
      "epoch": 0.33208074805557986,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004993241330142274,
      "loss": 6.0196,
      "mean_token_accuracy": 0.15770744234323503,
      "num_tokens": 7717626.0,
      "step": 3800
    },
    {
      "entropy": 6.172434234619141,
      "epoch": 0.33251769640828455,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004993217173994754,
      "loss": 6.0529,
      "mean_token_accuracy": 0.15278787910938263,
      "num_tokens": 7728221.0,
      "step": 3805
    },
    {
      "entropy": 6.2589140892028805,
      "epoch": 0.33295464476098924,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004993192974821154,
      "loss": 6.0592,
      "mean_token_accuracy": 0.15035932809114455,
      "num_tokens": 7739083.0,
      "step": 3810
    },
    {
      "entropy": 6.2666344165802,
      "epoch": 0.333391593113694,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004993168732621935,
      "loss": 6.0896,
      "mean_token_accuracy": 0.144187543541193,
      "num_tokens": 7748573.0,
      "step": 3815
    },
    {
      "entropy": 6.239906644821167,
      "epoch": 0.3338285414663987,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004993144447397565,
      "loss": 6.036,
      "mean_token_accuracy": 0.15224629193544387,
      "num_tokens": 7758741.0,
      "step": 3820
    },
    {
      "entropy": 6.211666822433472,
      "epoch": 0.33426548981910337,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499312011914851,
      "loss": 6.0719,
      "mean_token_accuracy": 0.14373058900237085,
      "num_tokens": 7768504.0,
      "step": 3825
    },
    {
      "entropy": 6.250473356246948,
      "epoch": 0.3347024381718081,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004993095747875234,
      "loss": 6.1943,
      "mean_token_accuracy": 0.14370191246271133,
      "num_tokens": 7779017.0,
      "step": 3830
    },
    {
      "entropy": 6.270117139816284,
      "epoch": 0.3351393865245128,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004993071333578207,
      "loss": 5.9947,
      "mean_token_accuracy": 0.15620245635509492,
      "num_tokens": 7787630.0,
      "step": 3835
    },
    {
      "entropy": 6.185167360305786,
      "epoch": 0.3355763348772175,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004993046876257898,
      "loss": 6.0063,
      "mean_token_accuracy": 0.1563827708363533,
      "num_tokens": 7797105.0,
      "step": 3840
    },
    {
      "entropy": 6.144869089126587,
      "epoch": 0.33601328322992224,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004993022375914773,
      "loss": 6.0018,
      "mean_token_accuracy": 0.15615751594305038,
      "num_tokens": 7807070.0,
      "step": 3845
    },
    {
      "entropy": 6.301814270019531,
      "epoch": 0.33645023158262694,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004992997832549303,
      "loss": 6.0356,
      "mean_token_accuracy": 0.1477530762553215,
      "num_tokens": 7818090.0,
      "step": 3850
    },
    {
      "entropy": 6.200486707687378,
      "epoch": 0.3368871799353316,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004992973246161961,
      "loss": 5.9792,
      "mean_token_accuracy": 0.1545109361410141,
      "num_tokens": 7827562.0,
      "step": 3855
    },
    {
      "entropy": 6.198225593566894,
      "epoch": 0.3373241282880364,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004992948616753215,
      "loss": 6.038,
      "mean_token_accuracy": 0.14645904898643494,
      "num_tokens": 7837581.0,
      "step": 3860
    },
    {
      "entropy": 6.237083196640015,
      "epoch": 0.33776107664074106,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004992923944323541,
      "loss": 6.0914,
      "mean_token_accuracy": 0.14635520726442336,
      "num_tokens": 7848729.0,
      "step": 3865
    },
    {
      "entropy": 6.258805942535401,
      "epoch": 0.33819802499344576,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499289922887341,
      "loss": 6.0141,
      "mean_token_accuracy": 0.15104980170726776,
      "num_tokens": 7857379.0,
      "step": 3870
    },
    {
      "entropy": 6.239930486679077,
      "epoch": 0.3386349733461505,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004992874470403296,
      "loss": 5.9866,
      "mean_token_accuracy": 0.15481197834014893,
      "num_tokens": 7867200.0,
      "step": 3875
    },
    {
      "entropy": 6.176450443267822,
      "epoch": 0.3390719216988552,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004992849668913674,
      "loss": 5.9995,
      "mean_token_accuracy": 0.1490779235959053,
      "num_tokens": 7876709.0,
      "step": 3880
    },
    {
      "entropy": 6.228991889953614,
      "epoch": 0.3395088700515599,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004992824824405021,
      "loss": 6.0927,
      "mean_token_accuracy": 0.14840275645256043,
      "num_tokens": 7886256.0,
      "step": 3885
    },
    {
      "entropy": 6.205230760574341,
      "epoch": 0.33994581840426463,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004992799936877813,
      "loss": 6.029,
      "mean_token_accuracy": 0.1480804704129696,
      "num_tokens": 7896497.0,
      "step": 3890
    },
    {
      "entropy": 6.236300754547119,
      "epoch": 0.3403827667569693,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004992775006332526,
      "loss": 6.0244,
      "mean_token_accuracy": 0.14987484216690064,
      "num_tokens": 7906608.0,
      "step": 3895
    },
    {
      "entropy": 6.279505777359009,
      "epoch": 0.340819715109674,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499275003276964,
      "loss": 6.1049,
      "mean_token_accuracy": 0.1473986364901066,
      "num_tokens": 7916359.0,
      "step": 3900
    },
    {
      "entropy": 6.21343584060669,
      "epoch": 0.34125666346237876,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004992725016189632,
      "loss": 5.9918,
      "mean_token_accuracy": 0.15183644890785217,
      "num_tokens": 7925800.0,
      "step": 3905
    },
    {
      "entropy": 6.26988959312439,
      "epoch": 0.34169361181508345,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004992699956592984,
      "loss": 6.0665,
      "mean_token_accuracy": 0.14467344135046006,
      "num_tokens": 7935424.0,
      "step": 3910
    },
    {
      "entropy": 6.236717939376831,
      "epoch": 0.34213056016778814,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004992674853980175,
      "loss": 6.0533,
      "mean_token_accuracy": 0.14863845109939575,
      "num_tokens": 7945766.0,
      "step": 3915
    },
    {
      "entropy": 6.269533777236939,
      "epoch": 0.3425675085204929,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004992649708351689,
      "loss": 6.0645,
      "mean_token_accuracy": 0.14135592579841613,
      "num_tokens": 7955622.0,
      "step": 3920
    },
    {
      "entropy": 6.248694849014282,
      "epoch": 0.3430044568731976,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004992624519708004,
      "loss": 6.127,
      "mean_token_accuracy": 0.14197314158082008,
      "num_tokens": 7965750.0,
      "step": 3925
    },
    {
      "entropy": 6.266123914718628,
      "epoch": 0.3434414052259023,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004992599288049607,
      "loss": 6.0537,
      "mean_token_accuracy": 0.15157223641872405,
      "num_tokens": 7977568.0,
      "step": 3930
    },
    {
      "entropy": 6.255645704269409,
      "epoch": 0.343878353578607,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499257401337698,
      "loss": 6.1047,
      "mean_token_accuracy": 0.14600171893835068,
      "num_tokens": 7988411.0,
      "step": 3935
    },
    {
      "entropy": 6.17456579208374,
      "epoch": 0.3443153019313117,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004992548695690608,
      "loss": 5.9695,
      "mean_token_accuracy": 0.1559397757053375,
      "num_tokens": 7998315.0,
      "step": 3940
    },
    {
      "entropy": 6.207716941833496,
      "epoch": 0.3447522502840164,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004992523334990977,
      "loss": 6.0146,
      "mean_token_accuracy": 0.14814926981925963,
      "num_tokens": 8007133.0,
      "step": 3945
    },
    {
      "entropy": 6.179518938064575,
      "epoch": 0.34518919863672115,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004992497931278574,
      "loss": 5.9608,
      "mean_token_accuracy": 0.1595747470855713,
      "num_tokens": 8016882.0,
      "step": 3950
    },
    {
      "entropy": 6.179951095581055,
      "epoch": 0.34562614698942584,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004992472484553886,
      "loss": 5.9728,
      "mean_token_accuracy": 0.15136277675628662,
      "num_tokens": 8026893.0,
      "step": 3955
    },
    {
      "entropy": 6.2377852439880375,
      "epoch": 0.34606309534213053,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00049924469948174,
      "loss": 6.0043,
      "mean_token_accuracy": 0.15436646193265915,
      "num_tokens": 8036240.0,
      "step": 3960
    },
    {
      "entropy": 6.265040302276612,
      "epoch": 0.3465000436948353,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004992421462069606,
      "loss": 5.9976,
      "mean_token_accuracy": 0.14397231116890907,
      "num_tokens": 8046086.0,
      "step": 3965
    },
    {
      "entropy": 6.084886646270752,
      "epoch": 0.34693699204753997,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004992395886310993,
      "loss": 5.9995,
      "mean_token_accuracy": 0.1564622104167938,
      "num_tokens": 8056294.0,
      "step": 3970
    },
    {
      "entropy": 6.250084590911865,
      "epoch": 0.3473739404002447,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004992370267542052,
      "loss": 6.0348,
      "mean_token_accuracy": 0.1465321108698845,
      "num_tokens": 8067329.0,
      "step": 3975
    },
    {
      "entropy": 6.234533977508545,
      "epoch": 0.3478108887529494,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992344605763275,
      "loss": 5.99,
      "mean_token_accuracy": 0.15098554342985154,
      "num_tokens": 8077106.0,
      "step": 3980
    },
    {
      "entropy": 6.200424528121948,
      "epoch": 0.3482478371056541,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004992318900975153,
      "loss": 5.9058,
      "mean_token_accuracy": 0.15573831498622895,
      "num_tokens": 8087766.0,
      "step": 3985
    },
    {
      "entropy": 6.1102227687835695,
      "epoch": 0.34868478545835885,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004992293153178179,
      "loss": 6.012,
      "mean_token_accuracy": 0.1529572382569313,
      "num_tokens": 8098649.0,
      "step": 3990
    },
    {
      "entropy": 6.219414758682251,
      "epoch": 0.34912173381106354,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004992267362372848,
      "loss": 6.0228,
      "mean_token_accuracy": 0.14785912334918977,
      "num_tokens": 8109237.0,
      "step": 3995
    },
    {
      "entropy": 6.186442136764526,
      "epoch": 0.34955868216376823,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004992241528559653,
      "loss": 6.0131,
      "mean_token_accuracy": 0.15060167163610458,
      "num_tokens": 8119731.0,
      "step": 4000
    },
    {
      "entropy": 6.268817949295044,
      "epoch": 0.349995630516473,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004992215651739092,
      "loss": 6.0347,
      "mean_token_accuracy": 0.14275750368833542,
      "num_tokens": 8129448.0,
      "step": 4005
    },
    {
      "entropy": 6.165897274017334,
      "epoch": 0.35043257886917767,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499218973191166,
      "loss": 5.9221,
      "mean_token_accuracy": 0.15644895881414414,
      "num_tokens": 8138345.0,
      "step": 4010
    },
    {
      "entropy": 6.224948310852051,
      "epoch": 0.35086952722188236,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004992163769077853,
      "loss": 5.9829,
      "mean_token_accuracy": 0.15417408645153047,
      "num_tokens": 8148380.0,
      "step": 4015
    },
    {
      "entropy": 6.180328798294068,
      "epoch": 0.3513064755745871,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499213776323817,
      "loss": 6.0055,
      "mean_token_accuracy": 0.15456762313842773,
      "num_tokens": 8159359.0,
      "step": 4020
    },
    {
      "entropy": 6.2428892135620115,
      "epoch": 0.3517434239272918,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499211171439311,
      "loss": 6.0872,
      "mean_token_accuracy": 0.14517719000577928,
      "num_tokens": 8168586.0,
      "step": 4025
    },
    {
      "entropy": 6.204936695098877,
      "epoch": 0.3521803722799965,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004992085622543172,
      "loss": 5.9854,
      "mean_token_accuracy": 0.1576502501964569,
      "num_tokens": 8178836.0,
      "step": 4030
    },
    {
      "entropy": 6.168943452835083,
      "epoch": 0.35261732063270124,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004992059487688859,
      "loss": 6.008,
      "mean_token_accuracy": 0.15231774598360062,
      "num_tokens": 8188933.0,
      "step": 4035
    },
    {
      "entropy": 6.144512033462524,
      "epoch": 0.35305426898540593,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004992033309830668,
      "loss": 5.9847,
      "mean_token_accuracy": 0.16302024722099304,
      "num_tokens": 8199366.0,
      "step": 4040
    },
    {
      "entropy": 6.314675426483154,
      "epoch": 0.3534912173381106,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004992007088969105,
      "loss": 5.9634,
      "mean_token_accuracy": 0.1587460979819298,
      "num_tokens": 8209806.0,
      "step": 4045
    },
    {
      "entropy": 6.2659971714019775,
      "epoch": 0.35392816569081537,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499198082510467,
      "loss": 6.0697,
      "mean_token_accuracy": 0.14272622764110565,
      "num_tokens": 8219367.0,
      "step": 4050
    },
    {
      "entropy": 6.154402875900269,
      "epoch": 0.35436511404352006,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004991954518237868,
      "loss": 5.9533,
      "mean_token_accuracy": 0.151335009932518,
      "num_tokens": 8228918.0,
      "step": 4055
    },
    {
      "entropy": 6.1810705184936525,
      "epoch": 0.35480206239622475,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004991928168369205,
      "loss": 6.039,
      "mean_token_accuracy": 0.1519352450966835,
      "num_tokens": 8239380.0,
      "step": 4060
    },
    {
      "entropy": 6.269686651229859,
      "epoch": 0.3552390107489295,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004991901775499183,
      "loss": 6.1006,
      "mean_token_accuracy": 0.14787229001522065,
      "num_tokens": 8249131.0,
      "step": 4065
    },
    {
      "entropy": 6.278018474578857,
      "epoch": 0.3556759591016342,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004991875339628311,
      "loss": 5.984,
      "mean_token_accuracy": 0.15121260285377502,
      "num_tokens": 8258550.0,
      "step": 4070
    },
    {
      "entropy": 6.158737659454346,
      "epoch": 0.3561129074543389,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004991848860757095,
      "loss": 6.0121,
      "mean_token_accuracy": 0.15460924804210663,
      "num_tokens": 8269089.0,
      "step": 4075
    },
    {
      "entropy": 6.170330953598023,
      "epoch": 0.3565498558070436,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004991822338886044,
      "loss": 5.9981,
      "mean_token_accuracy": 0.1478424534201622,
      "num_tokens": 8279735.0,
      "step": 4080
    },
    {
      "entropy": 6.221038961410523,
      "epoch": 0.3569868041597483,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004991795774015665,
      "loss": 5.9616,
      "mean_token_accuracy": 0.15689391344785691,
      "num_tokens": 8289693.0,
      "step": 4085
    },
    {
      "entropy": 6.19352879524231,
      "epoch": 0.357423752512453,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004991769166146469,
      "loss": 6.0344,
      "mean_token_accuracy": 0.1523745611310005,
      "num_tokens": 8300909.0,
      "step": 4090
    },
    {
      "entropy": 6.205101537704468,
      "epoch": 0.35786070086515775,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004991742515278965,
      "loss": 6.0008,
      "mean_token_accuracy": 0.15361023545265198,
      "num_tokens": 8311479.0,
      "step": 4095
    },
    {
      "entropy": 6.265884780883789,
      "epoch": 0.35829764921786245,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004991715821413665,
      "loss": 5.9999,
      "mean_token_accuracy": 0.15454096496105194,
      "num_tokens": 8321267.0,
      "step": 4100
    },
    {
      "entropy": 6.146849822998047,
      "epoch": 0.35873459757056714,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004991689084551082,
      "loss": 5.9561,
      "mean_token_accuracy": 0.15801454037427903,
      "num_tokens": 8331445.0,
      "step": 4105
    },
    {
      "entropy": 6.163304042816162,
      "epoch": 0.3591715459232719,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004991662304691727,
      "loss": 6.0427,
      "mean_token_accuracy": 0.14610980674624444,
      "num_tokens": 8340519.0,
      "step": 4110
    },
    {
      "entropy": 6.259560871124267,
      "epoch": 0.3596084942759766,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004991635481836114,
      "loss": 6.0199,
      "mean_token_accuracy": 0.14872232228517532,
      "num_tokens": 8350467.0,
      "step": 4115
    },
    {
      "entropy": 6.172140598297119,
      "epoch": 0.36004544262868127,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004991608615984757,
      "loss": 5.9613,
      "mean_token_accuracy": 0.1463010773062706,
      "num_tokens": 8360067.0,
      "step": 4120
    },
    {
      "entropy": 6.2325794219970705,
      "epoch": 0.360482390981386,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004991581707138175,
      "loss": 6.0362,
      "mean_token_accuracy": 0.15214309245347976,
      "num_tokens": 8370509.0,
      "step": 4125
    },
    {
      "entropy": 6.223120260238647,
      "epoch": 0.3609193393340907,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004991554755296878,
      "loss": 5.982,
      "mean_token_accuracy": 0.15292935967445373,
      "num_tokens": 8381489.0,
      "step": 4130
    },
    {
      "entropy": 6.122712326049805,
      "epoch": 0.3613562876867954,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004991527760461389,
      "loss": 6.0008,
      "mean_token_accuracy": 0.15350275188684465,
      "num_tokens": 8392287.0,
      "step": 4135
    },
    {
      "entropy": 6.170266437530517,
      "epoch": 0.36179323603950014,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004991500722632223,
      "loss": 5.9812,
      "mean_token_accuracy": 0.15097640305757523,
      "num_tokens": 8402433.0,
      "step": 4140
    },
    {
      "entropy": 6.1707926273345945,
      "epoch": 0.36223018439220483,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004991473641809896,
      "loss": 6.0225,
      "mean_token_accuracy": 0.14983665943145752,
      "num_tokens": 8412944.0,
      "step": 4145
    },
    {
      "entropy": 6.2134088516235355,
      "epoch": 0.3626671327449095,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004991446517994932,
      "loss": 5.9384,
      "mean_token_accuracy": 0.15240545570850372,
      "num_tokens": 8422511.0,
      "step": 4150
    },
    {
      "entropy": 6.168175792694091,
      "epoch": 0.3631040810976143,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004991419351187849,
      "loss": 5.9653,
      "mean_token_accuracy": 0.15691232234239577,
      "num_tokens": 8433094.0,
      "step": 4155
    },
    {
      "entropy": 6.125294160842896,
      "epoch": 0.36354102945031896,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004991392141389169,
      "loss": 6.0623,
      "mean_token_accuracy": 0.15332620590925217,
      "num_tokens": 8444430.0,
      "step": 4160
    },
    {
      "entropy": 6.231081438064575,
      "epoch": 0.36397797780302366,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004991364888599412,
      "loss": 6.1176,
      "mean_token_accuracy": 0.1470529094338417,
      "num_tokens": 8455171.0,
      "step": 4165
    },
    {
      "entropy": 6.221288394927979,
      "epoch": 0.3644149261557284,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004991337592819103,
      "loss": 5.921,
      "mean_token_accuracy": 0.1584060475230217,
      "num_tokens": 8464724.0,
      "step": 4170
    },
    {
      "entropy": 6.2204231262207035,
      "epoch": 0.3648518745084331,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004991310254048763,
      "loss": 6.0201,
      "mean_token_accuracy": 0.15029047429561615,
      "num_tokens": 8474732.0,
      "step": 4175
    },
    {
      "entropy": 6.133815145492553,
      "epoch": 0.36528882286113784,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004991282872288919,
      "loss": 5.9576,
      "mean_token_accuracy": 0.15694391280412673,
      "num_tokens": 8484731.0,
      "step": 4180
    },
    {
      "entropy": 6.150207233428955,
      "epoch": 0.36572577121384253,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004991255447540095,
      "loss": 5.9559,
      "mean_token_accuracy": 0.15796650052070618,
      "num_tokens": 8494509.0,
      "step": 4185
    },
    {
      "entropy": 6.219148397445679,
      "epoch": 0.3661627195665472,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004991227979802817,
      "loss": 6.0406,
      "mean_token_accuracy": 0.15501723140478135,
      "num_tokens": 8504260.0,
      "step": 4190
    },
    {
      "entropy": 6.221963882446289,
      "epoch": 0.36659966791925197,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004991200469077612,
      "loss": 6.0644,
      "mean_token_accuracy": 0.14710740000009537,
      "num_tokens": 8514222.0,
      "step": 4195
    },
    {
      "entropy": 6.22112398147583,
      "epoch": 0.36703661627195666,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004991172915365008,
      "loss": 6.0062,
      "mean_token_accuracy": 0.15218688696622848,
      "num_tokens": 8523979.0,
      "step": 4200
    },
    {
      "entropy": 6.199363327026367,
      "epoch": 0.36747356462466135,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004991145318665532,
      "loss": 5.9521,
      "mean_token_accuracy": 0.1610599249601364,
      "num_tokens": 8533516.0,
      "step": 4205
    },
    {
      "entropy": 6.191833066940307,
      "epoch": 0.3679105129773661,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004991117678979716,
      "loss": 6.0551,
      "mean_token_accuracy": 0.1441046938300133,
      "num_tokens": 8544444.0,
      "step": 4210
    },
    {
      "entropy": 6.20907039642334,
      "epoch": 0.3683474613300708,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004991089996308089,
      "loss": 5.9794,
      "mean_token_accuracy": 0.15260274708271027,
      "num_tokens": 8555571.0,
      "step": 4215
    },
    {
      "entropy": 6.278723573684692,
      "epoch": 0.3687844096827755,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004991062270651179,
      "loss": 6.0342,
      "mean_token_accuracy": 0.14726278856396674,
      "num_tokens": 8566486.0,
      "step": 4220
    },
    {
      "entropy": 6.21203727722168,
      "epoch": 0.36922135803548023,
      "grad_norm": 1.0,
      "learning_rate": 0.0004991034502009522,
      "loss": 5.9621,
      "mean_token_accuracy": 0.15388412922620773,
      "num_tokens": 8575324.0,
      "step": 4225
    },
    {
      "entropy": 6.1372088432312015,
      "epoch": 0.3696583063881849,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004991006690383649,
      "loss": 5.9502,
      "mean_token_accuracy": 0.15232342332601548,
      "num_tokens": 8586015.0,
      "step": 4230
    },
    {
      "entropy": 6.132140016555786,
      "epoch": 0.3700952547408896,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004990978835774093,
      "loss": 5.9306,
      "mean_token_accuracy": 0.155072882771492,
      "num_tokens": 8594748.0,
      "step": 4235
    },
    {
      "entropy": 6.13446011543274,
      "epoch": 0.37053220309359436,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004990950938181389,
      "loss": 6.0013,
      "mean_token_accuracy": 0.14853542819619178,
      "num_tokens": 8605805.0,
      "step": 4240
    },
    {
      "entropy": 6.186568546295166,
      "epoch": 0.37096915144629905,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004990922997606071,
      "loss": 5.9595,
      "mean_token_accuracy": 0.15177000910043717,
      "num_tokens": 8616591.0,
      "step": 4245
    },
    {
      "entropy": 6.214903497695923,
      "epoch": 0.37140609979900374,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004990895014048677,
      "loss": 5.987,
      "mean_token_accuracy": 0.1489131972193718,
      "num_tokens": 8627063.0,
      "step": 4250
    },
    {
      "entropy": 6.300046157836914,
      "epoch": 0.3718430481517085,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004990866987509741,
      "loss": 6.0661,
      "mean_token_accuracy": 0.14470021352171897,
      "num_tokens": 8637041.0,
      "step": 4255
    },
    {
      "entropy": 6.190048170089722,
      "epoch": 0.3722799965044132,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004990838917989802,
      "loss": 5.9773,
      "mean_token_accuracy": 0.15778657048940659,
      "num_tokens": 8647485.0,
      "step": 4260
    },
    {
      "entropy": 6.16148829460144,
      "epoch": 0.37271694485711787,
      "grad_norm": 0.921875,
      "learning_rate": 0.00049908108054894,
      "loss": 5.9516,
      "mean_token_accuracy": 0.15104584991931916,
      "num_tokens": 8657185.0,
      "step": 4265
    },
    {
      "entropy": 6.20255479812622,
      "epoch": 0.3731538932098226,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004990782650009071,
      "loss": 5.9613,
      "mean_token_accuracy": 0.15804735124111174,
      "num_tokens": 8667639.0,
      "step": 4270
    },
    {
      "entropy": 6.188499164581299,
      "epoch": 0.3735908415625273,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004990754451549358,
      "loss": 5.9794,
      "mean_token_accuracy": 0.15283505171537398,
      "num_tokens": 8678003.0,
      "step": 4275
    },
    {
      "entropy": 6.210866785049438,
      "epoch": 0.374027789915232,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004990726210110798,
      "loss": 6.0443,
      "mean_token_accuracy": 0.15008555352687836,
      "num_tokens": 8688837.0,
      "step": 4280
    },
    {
      "entropy": 6.155147314071655,
      "epoch": 0.37446473826793675,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004990697925693938,
      "loss": 5.8289,
      "mean_token_accuracy": 0.16587514132261277,
      "num_tokens": 8698395.0,
      "step": 4285
    },
    {
      "entropy": 6.157758569717407,
      "epoch": 0.37490168662064144,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004990669598299316,
      "loss": 5.9545,
      "mean_token_accuracy": 0.15433542281389237,
      "num_tokens": 8708031.0,
      "step": 4290
    },
    {
      "entropy": 6.12908787727356,
      "epoch": 0.37533863497334613,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004990641227927478,
      "loss": 5.9331,
      "mean_token_accuracy": 0.15757068544626235,
      "num_tokens": 8717825.0,
      "step": 4295
    },
    {
      "entropy": 6.257285165786743,
      "epoch": 0.3757755833260509,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004990612814578967,
      "loss": 6.0194,
      "mean_token_accuracy": 0.14891362860798835,
      "num_tokens": 8728873.0,
      "step": 4300
    },
    {
      "entropy": 6.150735092163086,
      "epoch": 0.37621253167875557,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990584358254327,
      "loss": 5.9536,
      "mean_token_accuracy": 0.15229746103286743,
      "num_tokens": 8737917.0,
      "step": 4305
    },
    {
      "entropy": 6.098759412765503,
      "epoch": 0.37664948003146026,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004990555858954107,
      "loss": 5.9038,
      "mean_token_accuracy": 0.1548462837934494,
      "num_tokens": 8747466.0,
      "step": 4310
    },
    {
      "entropy": 6.17538890838623,
      "epoch": 0.377086428384165,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000499052731667885,
      "loss": 5.9409,
      "mean_token_accuracy": 0.1568036288022995,
      "num_tokens": 8757037.0,
      "step": 4315
    },
    {
      "entropy": 6.196196222305298,
      "epoch": 0.3775233767368697,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990498731429105,
      "loss": 5.9771,
      "mean_token_accuracy": 0.1564203307032585,
      "num_tokens": 8766780.0,
      "step": 4320
    },
    {
      "entropy": 6.169270992279053,
      "epoch": 0.3779603250895744,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004990470103205421,
      "loss": 5.9469,
      "mean_token_accuracy": 0.1535429298877716,
      "num_tokens": 8777063.0,
      "step": 4325
    },
    {
      "entropy": 6.119955730438233,
      "epoch": 0.37839727344227914,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004990441432008347,
      "loss": 5.9802,
      "mean_token_accuracy": 0.15425399094820022,
      "num_tokens": 8786479.0,
      "step": 4330
    },
    {
      "entropy": 6.275782108306885,
      "epoch": 0.3788342217949838,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004990412717838431,
      "loss": 5.9646,
      "mean_token_accuracy": 0.15385251939296724,
      "num_tokens": 8795663.0,
      "step": 4335
    },
    {
      "entropy": 6.1232274055480955,
      "epoch": 0.3792711701476885,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004990383960696226,
      "loss": 5.9272,
      "mean_token_accuracy": 0.15729418247938157,
      "num_tokens": 8805854.0,
      "step": 4340
    },
    {
      "entropy": 6.137463521957398,
      "epoch": 0.37970811850039327,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004990355160582283,
      "loss": 5.8988,
      "mean_token_accuracy": 0.15592976957559584,
      "num_tokens": 8816068.0,
      "step": 4345
    },
    {
      "entropy": 6.1173069953918455,
      "epoch": 0.38014506685309796,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004990326317497153,
      "loss": 5.9099,
      "mean_token_accuracy": 0.15910188257694244,
      "num_tokens": 8826301.0,
      "step": 4350
    },
    {
      "entropy": 6.201882314682007,
      "epoch": 0.38058201520580265,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004990297431441391,
      "loss": 5.9579,
      "mean_token_accuracy": 0.16176767647266388,
      "num_tokens": 8837482.0,
      "step": 4355
    },
    {
      "entropy": 6.178161764144898,
      "epoch": 0.3810189635585074,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499026850241555,
      "loss": 5.9815,
      "mean_token_accuracy": 0.15647393688559533,
      "num_tokens": 8847041.0,
      "step": 4360
    },
    {
      "entropy": 6.138125228881836,
      "epoch": 0.3814559119112121,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004990239530420185,
      "loss": 5.971,
      "mean_token_accuracy": 0.1511300563812256,
      "num_tokens": 8856963.0,
      "step": 4365
    },
    {
      "entropy": 6.18984546661377,
      "epoch": 0.3818928602639168,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004990210515455853,
      "loss": 5.9517,
      "mean_token_accuracy": 0.1531401291489601,
      "num_tokens": 8867318.0,
      "step": 4370
    },
    {
      "entropy": 6.149747371673584,
      "epoch": 0.3823298086166215,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004990181457523109,
      "loss": 5.9576,
      "mean_token_accuracy": 0.15716585367918015,
      "num_tokens": 8877634.0,
      "step": 4375
    },
    {
      "entropy": 6.152999687194824,
      "epoch": 0.3827667569693262,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499015235662251,
      "loss": 5.9975,
      "mean_token_accuracy": 0.15242273956537247,
      "num_tokens": 8887708.0,
      "step": 4380
    },
    {
      "entropy": 6.169225645065308,
      "epoch": 0.38320370532203096,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004990123212754616,
      "loss": 5.925,
      "mean_token_accuracy": 0.1560312107205391,
      "num_tokens": 8897510.0,
      "step": 4385
    },
    {
      "entropy": 6.145167112350464,
      "epoch": 0.38364065367473565,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004990094025919985,
      "loss": 5.9577,
      "mean_token_accuracy": 0.1475707024335861,
      "num_tokens": 8908282.0,
      "step": 4390
    },
    {
      "entropy": 6.280637264251709,
      "epoch": 0.38407760202744035,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004990064796119175,
      "loss": 5.9905,
      "mean_token_accuracy": 0.14701349660754204,
      "num_tokens": 8918332.0,
      "step": 4395
    },
    {
      "entropy": 6.1294549942016605,
      "epoch": 0.3845145503801451,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004990035523352751,
      "loss": 5.957,
      "mean_token_accuracy": 0.15417761653661727,
      "num_tokens": 8929138.0,
      "step": 4400
    },
    {
      "entropy": 6.135020399093628,
      "epoch": 0.3849514987328498,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000499000620762127,
      "loss": 5.9958,
      "mean_token_accuracy": 0.14873793125152587,
      "num_tokens": 8940871.0,
      "step": 4405
    },
    {
      "entropy": 6.231491804122925,
      "epoch": 0.3853884470855545,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004989976848925297,
      "loss": 5.9361,
      "mean_token_accuracy": 0.15769388526678085,
      "num_tokens": 8951199.0,
      "step": 4410
    },
    {
      "entropy": 6.145180130004883,
      "epoch": 0.3858253954382592,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004989947447265394,
      "loss": 6.0565,
      "mean_token_accuracy": 0.14883989095687866,
      "num_tokens": 8961482.0,
      "step": 4415
    },
    {
      "entropy": 6.272399139404297,
      "epoch": 0.3862623437909639,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004989918002642126,
      "loss": 5.966,
      "mean_token_accuracy": 0.1529775321483612,
      "num_tokens": 8971884.0,
      "step": 4420
    },
    {
      "entropy": 6.130031728744507,
      "epoch": 0.3866992921436686,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004989888515056057,
      "loss": 5.9206,
      "mean_token_accuracy": 0.1541355162858963,
      "num_tokens": 8982086.0,
      "step": 4425
    },
    {
      "entropy": 6.135330867767334,
      "epoch": 0.38713624049637335,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004989858984507752,
      "loss": 5.9874,
      "mean_token_accuracy": 0.152043230086565,
      "num_tokens": 8991555.0,
      "step": 4430
    },
    {
      "entropy": 6.160719919204712,
      "epoch": 0.38757318884907804,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004989829410997778,
      "loss": 5.9709,
      "mean_token_accuracy": 0.1504402533173561,
      "num_tokens": 9002459.0,
      "step": 4435
    },
    {
      "entropy": 6.1761571884155275,
      "epoch": 0.38801013720178273,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004989799794526702,
      "loss": 6.0036,
      "mean_token_accuracy": 0.14963382706046105,
      "num_tokens": 9012514.0,
      "step": 4440
    },
    {
      "entropy": 6.172939491271973,
      "epoch": 0.3884470855544875,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004989770135095093,
      "loss": 5.9261,
      "mean_token_accuracy": 0.1512054905295372,
      "num_tokens": 9022636.0,
      "step": 4445
    },
    {
      "entropy": 6.146451759338379,
      "epoch": 0.3888840339071922,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004989740432703519,
      "loss": 5.9381,
      "mean_token_accuracy": 0.15658134073019028,
      "num_tokens": 9032868.0,
      "step": 4450
    },
    {
      "entropy": 6.137963438034058,
      "epoch": 0.38932098225989686,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000498971068735255,
      "loss": 5.9352,
      "mean_token_accuracy": 0.14985790252685546,
      "num_tokens": 9043057.0,
      "step": 4455
    },
    {
      "entropy": 6.127811861038208,
      "epoch": 0.3897579306126016,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004989680899042756,
      "loss": 5.9517,
      "mean_token_accuracy": 0.15457415729761123,
      "num_tokens": 9053179.0,
      "step": 4460
    },
    {
      "entropy": 6.138548040390015,
      "epoch": 0.3901948789653063,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004989651067774708,
      "loss": 5.96,
      "mean_token_accuracy": 0.15666047632694244,
      "num_tokens": 9063112.0,
      "step": 4465
    },
    {
      "entropy": 6.1771155834198,
      "epoch": 0.390631827318011,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004989621193548981,
      "loss": 5.9731,
      "mean_token_accuracy": 0.15245330333709717,
      "num_tokens": 9073512.0,
      "step": 4470
    },
    {
      "entropy": 6.193724632263184,
      "epoch": 0.39106877567071574,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004989591276366145,
      "loss": 5.9841,
      "mean_token_accuracy": 0.14708113372325898,
      "num_tokens": 9083369.0,
      "step": 4475
    },
    {
      "entropy": 6.141264724731445,
      "epoch": 0.39150572402342043,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004989561316226774,
      "loss": 5.9619,
      "mean_token_accuracy": 0.15295498222112655,
      "num_tokens": 9092814.0,
      "step": 4480
    },
    {
      "entropy": 6.042321491241455,
      "epoch": 0.3919426723761251,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004989531313131444,
      "loss": 5.8825,
      "mean_token_accuracy": 0.16421083658933638,
      "num_tokens": 9102816.0,
      "step": 4485
    },
    {
      "entropy": 6.154022264480591,
      "epoch": 0.39237962072882987,
      "grad_norm": 0.90625,
      "learning_rate": 0.000498950126708073,
      "loss": 6.0144,
      "mean_token_accuracy": 0.14581276923418046,
      "num_tokens": 9114470.0,
      "step": 4490
    },
    {
      "entropy": 6.221161270141602,
      "epoch": 0.39281656908153456,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004989471178075208,
      "loss": 5.989,
      "mean_token_accuracy": 0.14692097306251525,
      "num_tokens": 9125604.0,
      "step": 4495
    },
    {
      "entropy": 6.133387279510498,
      "epoch": 0.39325351743423925,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004989441046115456,
      "loss": 5.9294,
      "mean_token_accuracy": 0.15933895260095596,
      "num_tokens": 9135088.0,
      "step": 4500
    },
    {
      "entropy": 6.188969993591309,
      "epoch": 0.393690465786944,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004989410871202051,
      "loss": 5.9832,
      "mean_token_accuracy": 0.15392538905143738,
      "num_tokens": 9145302.0,
      "step": 4505
    },
    {
      "entropy": 6.105658292770386,
      "epoch": 0.3941274141396487,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004989380653335571,
      "loss": 5.9199,
      "mean_token_accuracy": 0.15435281097888948,
      "num_tokens": 9155104.0,
      "step": 4510
    },
    {
      "entropy": 6.124040269851685,
      "epoch": 0.3945643624923534,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004989350392516598,
      "loss": 5.8527,
      "mean_token_accuracy": 0.15398550778627396,
      "num_tokens": 9165055.0,
      "step": 4515
    },
    {
      "entropy": 6.115254163742065,
      "epoch": 0.39500131084505813,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004989320088745711,
      "loss": 5.8808,
      "mean_token_accuracy": 0.16048204451799392,
      "num_tokens": 9175305.0,
      "step": 4520
    },
    {
      "entropy": 6.1552516460418705,
      "epoch": 0.3954382591977628,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000498928974202349,
      "loss": 5.9706,
      "mean_token_accuracy": 0.1471177026629448,
      "num_tokens": 9185153.0,
      "step": 4525
    },
    {
      "entropy": 6.160642766952515,
      "epoch": 0.3958752075504675,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000498925935235052,
      "loss": 5.9264,
      "mean_token_accuracy": 0.1507927320897579,
      "num_tokens": 9195385.0,
      "step": 4530
    },
    {
      "entropy": 6.099987506866455,
      "epoch": 0.39631215590317226,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004989228919727382,
      "loss": 5.8825,
      "mean_token_accuracy": 0.16080036759376526,
      "num_tokens": 9205009.0,
      "step": 4535
    },
    {
      "entropy": 6.138633680343628,
      "epoch": 0.39674910425587695,
      "grad_norm": 0.859375,
      "learning_rate": 0.000498919844415466,
      "loss": 5.9448,
      "mean_token_accuracy": 0.15411535501480103,
      "num_tokens": 9214684.0,
      "step": 4540
    },
    {
      "entropy": 6.087051582336426,
      "epoch": 0.39718605260858164,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004989167925632937,
      "loss": 5.8855,
      "mean_token_accuracy": 0.1588042065501213,
      "num_tokens": 9225721.0,
      "step": 4545
    },
    {
      "entropy": 6.1690984725952145,
      "epoch": 0.3976230009612864,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004989137364162801,
      "loss": 5.967,
      "mean_token_accuracy": 0.15482041388750076,
      "num_tokens": 9236167.0,
      "step": 4550
    },
    {
      "entropy": 6.09454870223999,
      "epoch": 0.3980599493139911,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004989106759744837,
      "loss": 5.9052,
      "mean_token_accuracy": 0.16375220119953154,
      "num_tokens": 9245716.0,
      "step": 4555
    },
    {
      "entropy": 6.142717027664185,
      "epoch": 0.39849689766669577,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004989076112379633,
      "loss": 5.9952,
      "mean_token_accuracy": 0.1506937101483345,
      "num_tokens": 9256897.0,
      "step": 4560
    },
    {
      "entropy": 6.255190420150757,
      "epoch": 0.3989338460194005,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004989045422067775,
      "loss": 6.0027,
      "mean_token_accuracy": 0.14253595024347304,
      "num_tokens": 9266080.0,
      "step": 4565
    },
    {
      "entropy": 6.113092374801636,
      "epoch": 0.3993707943721052,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004989014688809853,
      "loss": 5.9129,
      "mean_token_accuracy": 0.1528676062822342,
      "num_tokens": 9276815.0,
      "step": 4570
    },
    {
      "entropy": 6.166882181167603,
      "epoch": 0.3998077427248099,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004988983912606454,
      "loss": 5.9845,
      "mean_token_accuracy": 0.15317472666502,
      "num_tokens": 9287339.0,
      "step": 4575
    },
    {
      "entropy": 6.104081726074218,
      "epoch": 0.40024469107751465,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004988953093458173,
      "loss": 5.8525,
      "mean_token_accuracy": 0.160149547457695,
      "num_tokens": 9298002.0,
      "step": 4580
    },
    {
      "entropy": 6.127166795730591,
      "epoch": 0.40068163943021934,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004988922231365597,
      "loss": 5.9923,
      "mean_token_accuracy": 0.14690545797348023,
      "num_tokens": 9309008.0,
      "step": 4585
    },
    {
      "entropy": 6.125999641418457,
      "epoch": 0.4011185877829241,
      "grad_norm": 0.84375,
      "learning_rate": 0.000498889132632932,
      "loss": 5.9939,
      "mean_token_accuracy": 0.1510233238339424,
      "num_tokens": 9319391.0,
      "step": 4590
    },
    {
      "entropy": 6.04268856048584,
      "epoch": 0.4015555361356288,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004988860378349934,
      "loss": 5.8255,
      "mean_token_accuracy": 0.1640730954706669,
      "num_tokens": 9329808.0,
      "step": 4595
    },
    {
      "entropy": 6.1170600891113285,
      "epoch": 0.40199248448833347,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004988829387428033,
      "loss": 5.8954,
      "mean_token_accuracy": 0.158653524518013,
      "num_tokens": 9339580.0,
      "step": 4600
    },
    {
      "entropy": 6.180007410049439,
      "epoch": 0.4024294328410382,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004988798353564211,
      "loss": 5.914,
      "mean_token_accuracy": 0.16047825962305068,
      "num_tokens": 9350008.0,
      "step": 4605
    },
    {
      "entropy": 6.181416606903076,
      "epoch": 0.4028663811937429,
      "grad_norm": 1.0,
      "learning_rate": 0.0004988767276759064,
      "loss": 5.9099,
      "mean_token_accuracy": 0.15556182861328124,
      "num_tokens": 9359026.0,
      "step": 4610
    },
    {
      "entropy": 6.122279930114746,
      "epoch": 0.4033033295464476,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004988736157013186,
      "loss": 5.8758,
      "mean_token_accuracy": 0.15592683404684066,
      "num_tokens": 9369084.0,
      "step": 4615
    },
    {
      "entropy": 6.068732404708863,
      "epoch": 0.40374027789915234,
      "grad_norm": 1.0,
      "learning_rate": 0.0004988704994327178,
      "loss": 5.8703,
      "mean_token_accuracy": 0.16378343552351,
      "num_tokens": 9378273.0,
      "step": 4620
    },
    {
      "entropy": 6.097204971313476,
      "epoch": 0.40417722625185704,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004988673788701633,
      "loss": 5.9002,
      "mean_token_accuracy": 0.1529839888215065,
      "num_tokens": 9388434.0,
      "step": 4625
    },
    {
      "entropy": 6.173472833633423,
      "epoch": 0.4046141746045617,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004988642540137152,
      "loss": 5.9829,
      "mean_token_accuracy": 0.14962709844112396,
      "num_tokens": 9398484.0,
      "step": 4630
    },
    {
      "entropy": 6.070959091186523,
      "epoch": 0.4050511229572665,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004988611248634335,
      "loss": 5.8537,
      "mean_token_accuracy": 0.1586995616555214,
      "num_tokens": 9409544.0,
      "step": 4635
    },
    {
      "entropy": 6.145827198028565,
      "epoch": 0.40548807130997117,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000498857991419378,
      "loss": 5.9176,
      "mean_token_accuracy": 0.15430696606636046,
      "num_tokens": 9419391.0,
      "step": 4640
    },
    {
      "entropy": 6.167720651626587,
      "epoch": 0.40592501966267586,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004988548536816091,
      "loss": 5.8823,
      "mean_token_accuracy": 0.16215257942676545,
      "num_tokens": 9429342.0,
      "step": 4645
    },
    {
      "entropy": 6.117645025253296,
      "epoch": 0.4063619680153806,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004988517116501867,
      "loss": 5.9971,
      "mean_token_accuracy": 0.14209671840071678,
      "num_tokens": 9440643.0,
      "step": 4650
    },
    {
      "entropy": 6.208621311187744,
      "epoch": 0.4067989163680853,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004988485653251711,
      "loss": 5.9451,
      "mean_token_accuracy": 0.15437053889036179,
      "num_tokens": 9450674.0,
      "step": 4655
    },
    {
      "entropy": 6.1741351127624515,
      "epoch": 0.40723586472079,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004988454147066228,
      "loss": 5.9197,
      "mean_token_accuracy": 0.15716563463211058,
      "num_tokens": 9460075.0,
      "step": 4660
    },
    {
      "entropy": 6.099549198150635,
      "epoch": 0.40767281307349473,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004988422597946022,
      "loss": 5.896,
      "mean_token_accuracy": 0.16132495701313018,
      "num_tokens": 9469327.0,
      "step": 4665
    },
    {
      "entropy": 6.170559120178223,
      "epoch": 0.4081097614261994,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004988391005891698,
      "loss": 5.9437,
      "mean_token_accuracy": 0.1529410094022751,
      "num_tokens": 9478304.0,
      "step": 4670
    },
    {
      "entropy": 6.089248180389404,
      "epoch": 0.4085467097789041,
      "grad_norm": 0.984375,
      "learning_rate": 0.000498835937090386,
      "loss": 5.8651,
      "mean_token_accuracy": 0.15965885519981385,
      "num_tokens": 9487453.0,
      "step": 4675
    },
    {
      "entropy": 6.091420078277588,
      "epoch": 0.40898365813160886,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004988327692983116,
      "loss": 5.8263,
      "mean_token_accuracy": 0.16226141899824142,
      "num_tokens": 9496623.0,
      "step": 4680
    },
    {
      "entropy": 6.112924242019654,
      "epoch": 0.40942060648431355,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004988295972130076,
      "loss": 5.9733,
      "mean_token_accuracy": 0.15255378037691117,
      "num_tokens": 9506293.0,
      "step": 4685
    },
    {
      "entropy": 6.144012498855591,
      "epoch": 0.40985755483701825,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004988264208345346,
      "loss": 5.8609,
      "mean_token_accuracy": 0.1574437811970711,
      "num_tokens": 9516127.0,
      "step": 4690
    },
    {
      "entropy": 6.063244152069092,
      "epoch": 0.410294503189723,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004988232401629535,
      "loss": 5.8666,
      "mean_token_accuracy": 0.15879199355840684,
      "num_tokens": 9525597.0,
      "step": 4695
    },
    {
      "entropy": 6.0836179733276365,
      "epoch": 0.4107314515424277,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004988200551983254,
      "loss": 5.8454,
      "mean_token_accuracy": 0.1594383656978607,
      "num_tokens": 9535208.0,
      "step": 4700
    },
    {
      "entropy": 6.075837135314941,
      "epoch": 0.4111683998951324,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004988168659407113,
      "loss": 5.8929,
      "mean_token_accuracy": 0.16378208845853806,
      "num_tokens": 9545078.0,
      "step": 4705
    },
    {
      "entropy": 6.160520839691162,
      "epoch": 0.4116053482478371,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004988136723901725,
      "loss": 5.9045,
      "mean_token_accuracy": 0.1531546324491501,
      "num_tokens": 9555074.0,
      "step": 4710
    },
    {
      "entropy": 6.109802627563477,
      "epoch": 0.4120422966005418,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004988104745467702,
      "loss": 5.9027,
      "mean_token_accuracy": 0.15896735340356827,
      "num_tokens": 9564777.0,
      "step": 4715
    },
    {
      "entropy": 6.107122421264648,
      "epoch": 0.4124792449532465,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004988072724105657,
      "loss": 5.8435,
      "mean_token_accuracy": 0.1565806970000267,
      "num_tokens": 9574551.0,
      "step": 4720
    },
    {
      "entropy": 6.101800775527954,
      "epoch": 0.41291619330595125,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004988040659816203,
      "loss": 5.9663,
      "mean_token_accuracy": 0.14592478796839714,
      "num_tokens": 9585162.0,
      "step": 4725
    },
    {
      "entropy": 6.126856660842895,
      "epoch": 0.41335314165865594,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004988008552599958,
      "loss": 5.8922,
      "mean_token_accuracy": 0.16171422302722932,
      "num_tokens": 9595422.0,
      "step": 4730
    },
    {
      "entropy": 6.096254062652588,
      "epoch": 0.41379009001136063,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004987976402457536,
      "loss": 5.8825,
      "mean_token_accuracy": 0.165338733792305,
      "num_tokens": 9604315.0,
      "step": 4735
    },
    {
      "entropy": 6.0762707710266115,
      "epoch": 0.4142270383640654,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004987944209389555,
      "loss": 5.8958,
      "mean_token_accuracy": 0.15499772429466246,
      "num_tokens": 9614369.0,
      "step": 4740
    },
    {
      "entropy": 6.158682441711425,
      "epoch": 0.41466398671677007,
      "grad_norm": 0.96875,
      "learning_rate": 0.000498791197339663,
      "loss": 5.9443,
      "mean_token_accuracy": 0.1528699055314064,
      "num_tokens": 9624154.0,
      "step": 4745
    },
    {
      "entropy": 6.150540733337403,
      "epoch": 0.41510093506947476,
      "grad_norm": 1.0625,
      "learning_rate": 0.000498787969447938,
      "loss": 5.9644,
      "mean_token_accuracy": 0.15262885689735411,
      "num_tokens": 9634903.0,
      "step": 4750
    },
    {
      "entropy": 6.073925113677978,
      "epoch": 0.4155378834221795,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004987847372638427,
      "loss": 5.9637,
      "mean_token_accuracy": 0.1611965700984001,
      "num_tokens": 9644873.0,
      "step": 4755
    },
    {
      "entropy": 6.153850030899048,
      "epoch": 0.4159748317748842,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004987815007874388,
      "loss": 5.8692,
      "mean_token_accuracy": 0.17039722949266434,
      "num_tokens": 9654661.0,
      "step": 4760
    },
    {
      "entropy": 6.119657421112061,
      "epoch": 0.4164117801275889,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004987782600187884,
      "loss": 5.884,
      "mean_token_accuracy": 0.16364409178495407,
      "num_tokens": 9664980.0,
      "step": 4765
    },
    {
      "entropy": 6.09555344581604,
      "epoch": 0.41684872848029364,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004987750149579537,
      "loss": 5.8792,
      "mean_token_accuracy": 0.16119770854711532,
      "num_tokens": 9675117.0,
      "step": 4770
    },
    {
      "entropy": 6.113473224639892,
      "epoch": 0.41728567683299833,
      "grad_norm": 0.9375,
      "learning_rate": 0.000498771765604997,
      "loss": 5.923,
      "mean_token_accuracy": 0.15419435352087021,
      "num_tokens": 9684900.0,
      "step": 4775
    },
    {
      "entropy": 6.145103168487549,
      "epoch": 0.417722625185703,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004987685119599806,
      "loss": 5.9188,
      "mean_token_accuracy": 0.1530965194106102,
      "num_tokens": 9694586.0,
      "step": 4780
    },
    {
      "entropy": 6.1405035018920895,
      "epoch": 0.41815957353840777,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004987652540229668,
      "loss": 5.9196,
      "mean_token_accuracy": 0.15793851613998414,
      "num_tokens": 9705676.0,
      "step": 4785
    },
    {
      "entropy": 6.1464728832244875,
      "epoch": 0.41859652189111246,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004987619917940183,
      "loss": 5.9458,
      "mean_token_accuracy": 0.15899438858032228,
      "num_tokens": 9715442.0,
      "step": 4790
    },
    {
      "entropy": 6.219377660751343,
      "epoch": 0.4190334702438172,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004987587252731975,
      "loss": 6.0173,
      "mean_token_accuracy": 0.15520615503191948,
      "num_tokens": 9726621.0,
      "step": 4795
    },
    {
      "entropy": 6.143201494216919,
      "epoch": 0.4194704185965219,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004987554544605671,
      "loss": 6.0238,
      "mean_token_accuracy": 0.14427199214696884,
      "num_tokens": 9736688.0,
      "step": 4800
    },
    {
      "entropy": 6.126335477828979,
      "epoch": 0.4199073669492266,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004987521793561898,
      "loss": 5.8957,
      "mean_token_accuracy": 0.16006623208522797,
      "num_tokens": 9746478.0,
      "step": 4805
    },
    {
      "entropy": 6.120117235183716,
      "epoch": 0.42034431530193134,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004987488999601285,
      "loss": 5.9316,
      "mean_token_accuracy": 0.156678207218647,
      "num_tokens": 9756370.0,
      "step": 4810
    },
    {
      "entropy": 6.161874628067016,
      "epoch": 0.42078126365463603,
      "grad_norm": 0.9375,
      "learning_rate": 0.000498745616272446,
      "loss": 5.9347,
      "mean_token_accuracy": 0.14946951419115068,
      "num_tokens": 9766293.0,
      "step": 4815
    },
    {
      "entropy": 6.136138820648194,
      "epoch": 0.4212182120073407,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004987423282932054,
      "loss": 5.9138,
      "mean_token_accuracy": 0.15771846622228622,
      "num_tokens": 9777172.0,
      "step": 4820
    },
    {
      "entropy": 6.083975219726563,
      "epoch": 0.42165516036004547,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004987390360224697,
      "loss": 5.8839,
      "mean_token_accuracy": 0.15716514885425567,
      "num_tokens": 9787194.0,
      "step": 4825
    },
    {
      "entropy": 6.096252679824829,
      "epoch": 0.42209210871275016,
      "grad_norm": 0.875,
      "learning_rate": 0.000498735739460302,
      "loss": 5.9231,
      "mean_token_accuracy": 0.15954737961292267,
      "num_tokens": 9797528.0,
      "step": 4830
    },
    {
      "entropy": 6.119057941436767,
      "epoch": 0.42252905706545485,
      "grad_norm": 1.125,
      "learning_rate": 0.0004987324386067656,
      "loss": 5.9458,
      "mean_token_accuracy": 0.15423630625009538,
      "num_tokens": 9808412.0,
      "step": 4835
    },
    {
      "entropy": 6.164724588394165,
      "epoch": 0.4229660054181596,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004987291334619239,
      "loss": 6.0183,
      "mean_token_accuracy": 0.14648769348859786,
      "num_tokens": 9820047.0,
      "step": 4840
    },
    {
      "entropy": 6.1444111347198485,
      "epoch": 0.4234029537708643,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00049872582402584,
      "loss": 5.8427,
      "mean_token_accuracy": 0.16165557652711868,
      "num_tokens": 9828812.0,
      "step": 4845
    },
    {
      "entropy": 6.07479567527771,
      "epoch": 0.423839902123569,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004987225102985778,
      "loss": 5.8709,
      "mean_token_accuracy": 0.1626689463853836,
      "num_tokens": 9838288.0,
      "step": 4850
    },
    {
      "entropy": 6.17829852104187,
      "epoch": 0.4242768504762737,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004987191922802003,
      "loss": 5.8979,
      "mean_token_accuracy": 0.15759438425302505,
      "num_tokens": 9848536.0,
      "step": 4855
    },
    {
      "entropy": 6.081185388565063,
      "epoch": 0.4247137988289784,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004987158699707717,
      "loss": 5.9403,
      "mean_token_accuracy": 0.15936280265450478,
      "num_tokens": 9859741.0,
      "step": 4860
    },
    {
      "entropy": 6.154061698913575,
      "epoch": 0.4251507471816831,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004987125433703554,
      "loss": 5.8932,
      "mean_token_accuracy": 0.15802878588438035,
      "num_tokens": 9869173.0,
      "step": 4865
    },
    {
      "entropy": 6.140933799743652,
      "epoch": 0.42558769553438786,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004987092124790155,
      "loss": 5.9152,
      "mean_token_accuracy": 0.15708720907568932,
      "num_tokens": 9879619.0,
      "step": 4870
    },
    {
      "entropy": 6.073929786682129,
      "epoch": 0.42602464388709255,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004987058772968154,
      "loss": 5.8475,
      "mean_token_accuracy": 0.16600827127695084,
      "num_tokens": 9890214.0,
      "step": 4875
    },
    {
      "entropy": 6.079315423965454,
      "epoch": 0.42646159223979724,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004987025378238195,
      "loss": 5.8829,
      "mean_token_accuracy": 0.15276064574718476,
      "num_tokens": 9900266.0,
      "step": 4880
    },
    {
      "entropy": 6.166507816314697,
      "epoch": 0.426898540592502,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004986991940600917,
      "loss": 5.9015,
      "mean_token_accuracy": 0.15680881440639496,
      "num_tokens": 9910790.0,
      "step": 4885
    },
    {
      "entropy": 6.145452642440796,
      "epoch": 0.4273354889452067,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004986958460056961,
      "loss": 5.9196,
      "mean_token_accuracy": 0.1490568682551384,
      "num_tokens": 9921167.0,
      "step": 4890
    },
    {
      "entropy": 6.087120628356933,
      "epoch": 0.42777243729791137,
      "grad_norm": 0.90625,
      "learning_rate": 0.000498692493660697,
      "loss": 5.8626,
      "mean_token_accuracy": 0.15905486345291137,
      "num_tokens": 9931674.0,
      "step": 4895
    },
    {
      "entropy": 6.098935270309449,
      "epoch": 0.4282093856506161,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004986891370251587,
      "loss": 5.8165,
      "mean_token_accuracy": 0.15766057521104812,
      "num_tokens": 9941484.0,
      "step": 4900
    },
    {
      "entropy": 6.127752304077148,
      "epoch": 0.4286463340033208,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004986857760991454,
      "loss": 5.9676,
      "mean_token_accuracy": 0.15095995366573334,
      "num_tokens": 9951885.0,
      "step": 4905
    },
    {
      "entropy": 6.1411848068237305,
      "epoch": 0.4290832823560255,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004986824108827218,
      "loss": 5.7956,
      "mean_token_accuracy": 0.16215234696865083,
      "num_tokens": 9961658.0,
      "step": 4910
    },
    {
      "entropy": 5.999189138412476,
      "epoch": 0.42952023070873024,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004986790413759523,
      "loss": 5.9074,
      "mean_token_accuracy": 0.15364072769880294,
      "num_tokens": 9971971.0,
      "step": 4915
    },
    {
      "entropy": 6.094682407379151,
      "epoch": 0.42995717906143494,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004986756675789016,
      "loss": 5.8129,
      "mean_token_accuracy": 0.15694656521081923,
      "num_tokens": 9982211.0,
      "step": 4920
    },
    {
      "entropy": 6.112336492538452,
      "epoch": 0.4303941274141396,
      "grad_norm": 1.0,
      "learning_rate": 0.0004986722894916344,
      "loss": 5.8617,
      "mean_token_accuracy": 0.16319485306739806,
      "num_tokens": 9991935.0,
      "step": 4925
    },
    {
      "entropy": 6.052273893356324,
      "epoch": 0.4308310757668444,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004986689071142155,
      "loss": 5.8917,
      "mean_token_accuracy": 0.15127992033958435,
      "num_tokens": 10002014.0,
      "step": 4930
    },
    {
      "entropy": 6.119452953338623,
      "epoch": 0.43126802411954906,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004986655204467097,
      "loss": 5.9239,
      "mean_token_accuracy": 0.15334465056657792,
      "num_tokens": 10012030.0,
      "step": 4935
    },
    {
      "entropy": 6.091619825363159,
      "epoch": 0.43170497247225376,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004986621294891822,
      "loss": 5.8779,
      "mean_token_accuracy": 0.15746373534202576,
      "num_tokens": 10022188.0,
      "step": 4940
    },
    {
      "entropy": 6.132827997207642,
      "epoch": 0.4321419208249585,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004986587342416976,
      "loss": 5.7939,
      "mean_token_accuracy": 0.16190674751996995,
      "num_tokens": 10032131.0,
      "step": 4945
    },
    {
      "entropy": 6.1291422843933105,
      "epoch": 0.4325788691776632,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004986553347043214,
      "loss": 5.841,
      "mean_token_accuracy": 0.15926993489265442,
      "num_tokens": 10043204.0,
      "step": 4950
    },
    {
      "entropy": 6.077534294128418,
      "epoch": 0.4330158175303679,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004986519308771188,
      "loss": 5.9102,
      "mean_token_accuracy": 0.15396561622619628,
      "num_tokens": 10052863.0,
      "step": 4955
    },
    {
      "entropy": 6.091691589355468,
      "epoch": 0.43345276588307263,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004986485227601548,
      "loss": 5.8971,
      "mean_token_accuracy": 0.16426312774419785,
      "num_tokens": 10063212.0,
      "step": 4960
    },
    {
      "entropy": 6.135089302062989,
      "epoch": 0.4338897142357773,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000498645110353495,
      "loss": 5.9125,
      "mean_token_accuracy": 0.1550646737217903,
      "num_tokens": 10074180.0,
      "step": 4965
    },
    {
      "entropy": 6.034312391281128,
      "epoch": 0.434326662588482,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004986416936572049,
      "loss": 5.8586,
      "mean_token_accuracy": 0.16182521730661392,
      "num_tokens": 10084513.0,
      "step": 4970
    },
    {
      "entropy": 6.12773118019104,
      "epoch": 0.43476361094118676,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004986382726713497,
      "loss": 5.8815,
      "mean_token_accuracy": 0.15795285105705262,
      "num_tokens": 10094352.0,
      "step": 4975
    },
    {
      "entropy": 6.01217246055603,
      "epoch": 0.43520055929389145,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004986348473959955,
      "loss": 5.7399,
      "mean_token_accuracy": 0.17221405506134033,
      "num_tokens": 10103677.0,
      "step": 4980
    },
    {
      "entropy": 6.117711782455444,
      "epoch": 0.43563750764659614,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004986314178312076,
      "loss": 5.9636,
      "mean_token_accuracy": 0.14685683250427245,
      "num_tokens": 10113479.0,
      "step": 4985
    },
    {
      "entropy": 6.107839775085449,
      "epoch": 0.4360744559993009,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004986279839770519,
      "loss": 5.8576,
      "mean_token_accuracy": 0.15548283010721206,
      "num_tokens": 10123526.0,
      "step": 4990
    },
    {
      "entropy": 6.107575511932373,
      "epoch": 0.4365114043520056,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004986245458335944,
      "loss": 5.8764,
      "mean_token_accuracy": 0.15630691051483153,
      "num_tokens": 10134176.0,
      "step": 4995
    },
    {
      "entropy": 6.0929351329803465,
      "epoch": 0.43694835270471033,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004986211034009008,
      "loss": 5.8435,
      "mean_token_accuracy": 0.1651988849043846,
      "num_tokens": 10144763.0,
      "step": 5000
    },
    {
      "entropy": 6.047188711166382,
      "epoch": 0.437385301057415,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004986176566790375,
      "loss": 5.8092,
      "mean_token_accuracy": 0.16798071563243866,
      "num_tokens": 10155011.0,
      "step": 5005
    },
    {
      "entropy": 6.10062165260315,
      "epoch": 0.4378222494101197,
      "grad_norm": 0.875,
      "learning_rate": 0.0004986142056680701,
      "loss": 5.9057,
      "mean_token_accuracy": 0.15978446751832961,
      "num_tokens": 10165474.0,
      "step": 5010
    },
    {
      "entropy": 6.080620527267456,
      "epoch": 0.43825919776282446,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004986107503680652,
      "loss": 5.9249,
      "mean_token_accuracy": 0.15882788896560668,
      "num_tokens": 10175137.0,
      "step": 5015
    },
    {
      "entropy": 6.098481130599976,
      "epoch": 0.43869614611552915,
      "grad_norm": 1.0,
      "learning_rate": 0.0004986072907790889,
      "loss": 5.9203,
      "mean_token_accuracy": 0.1513497620820999,
      "num_tokens": 10185623.0,
      "step": 5020
    },
    {
      "entropy": 6.172309303283692,
      "epoch": 0.43913309446823384,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004986038269012076,
      "loss": 5.8909,
      "mean_token_accuracy": 0.1573339194059372,
      "num_tokens": 10195201.0,
      "step": 5025
    },
    {
      "entropy": 6.099797010421753,
      "epoch": 0.4395700428209386,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004986003587344877,
      "loss": 5.9272,
      "mean_token_accuracy": 0.15723302960395813,
      "num_tokens": 10205951.0,
      "step": 5030
    },
    {
      "entropy": 6.0694938659667965,
      "epoch": 0.4400069911736433,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004985968862789957,
      "loss": 5.7995,
      "mean_token_accuracy": 0.1665978655219078,
      "num_tokens": 10215665.0,
      "step": 5035
    },
    {
      "entropy": 6.0943746089935305,
      "epoch": 0.44044393952634797,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004985934095347985,
      "loss": 5.8876,
      "mean_token_accuracy": 0.1551556296646595,
      "num_tokens": 10225037.0,
      "step": 5040
    },
    {
      "entropy": 6.0550844192504885,
      "epoch": 0.4408808878790527,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004985899285019624,
      "loss": 5.8357,
      "mean_token_accuracy": 0.15570665746927262,
      "num_tokens": 10234581.0,
      "step": 5045
    },
    {
      "entropy": 6.072161102294922,
      "epoch": 0.4413178362317574,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004985864431805543,
      "loss": 5.7679,
      "mean_token_accuracy": 0.16553371846675874,
      "num_tokens": 10243716.0,
      "step": 5050
    },
    {
      "entropy": 6.05171570777893,
      "epoch": 0.4417547845844621,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004985829535706411,
      "loss": 5.8731,
      "mean_token_accuracy": 0.15933090895414354,
      "num_tokens": 10253198.0,
      "step": 5055
    },
    {
      "entropy": 6.05954155921936,
      "epoch": 0.44219173293716685,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004985794596722897,
      "loss": 5.9501,
      "mean_token_accuracy": 0.15209509134292604,
      "num_tokens": 10262912.0,
      "step": 5060
    },
    {
      "entropy": 6.088939428329468,
      "epoch": 0.44262868128987154,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000498575961485567,
      "loss": 5.8864,
      "mean_token_accuracy": 0.15810550898313522,
      "num_tokens": 10272960.0,
      "step": 5065
    },
    {
      "entropy": 6.16075382232666,
      "epoch": 0.44306562964257623,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004985724590105403,
      "loss": 5.9316,
      "mean_token_accuracy": 0.15536672621965408,
      "num_tokens": 10283034.0,
      "step": 5070
    },
    {
      "entropy": 6.0558013916015625,
      "epoch": 0.443502577995281,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004985689522472767,
      "loss": 5.8871,
      "mean_token_accuracy": 0.15896483659744262,
      "num_tokens": 10292553.0,
      "step": 5075
    },
    {
      "entropy": 6.028521251678467,
      "epoch": 0.44393952634798567,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004985654411958434,
      "loss": 5.853,
      "mean_token_accuracy": 0.16116765141487122,
      "num_tokens": 10302475.0,
      "step": 5080
    },
    {
      "entropy": 6.089235115051269,
      "epoch": 0.44437647470069036,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004985619258563078,
      "loss": 5.8595,
      "mean_token_accuracy": 0.1636977016925812,
      "num_tokens": 10311693.0,
      "step": 5085
    },
    {
      "entropy": 6.120714521408081,
      "epoch": 0.4448134230533951,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004985584062287372,
      "loss": 5.8539,
      "mean_token_accuracy": 0.16536978930234908,
      "num_tokens": 10323052.0,
      "step": 5090
    },
    {
      "entropy": 6.062410688400268,
      "epoch": 0.4452503714060998,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004985548823131994,
      "loss": 5.8291,
      "mean_token_accuracy": 0.16429725140333176,
      "num_tokens": 10332997.0,
      "step": 5095
    },
    {
      "entropy": 6.1553514957427975,
      "epoch": 0.4456873197588045,
      "grad_norm": 1.0,
      "learning_rate": 0.0004985513541097618,
      "loss": 5.9593,
      "mean_token_accuracy": 0.15591451823711394,
      "num_tokens": 10343298.0,
      "step": 5100
    },
    {
      "entropy": 6.171523571014404,
      "epoch": 0.44612426811150924,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000498547821618492,
      "loss": 5.8499,
      "mean_token_accuracy": 0.15821425765752792,
      "num_tokens": 10352485.0,
      "step": 5105
    },
    {
      "entropy": 5.98296799659729,
      "epoch": 0.44656121646421393,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004985442848394579,
      "loss": 5.9016,
      "mean_token_accuracy": 0.15563226342201233,
      "num_tokens": 10362810.0,
      "step": 5110
    },
    {
      "entropy": 6.0878349304199215,
      "epoch": 0.4469981648169186,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004985407437727273,
      "loss": 5.8386,
      "mean_token_accuracy": 0.16296190321445464,
      "num_tokens": 10372817.0,
      "step": 5115
    },
    {
      "entropy": 6.139986419677735,
      "epoch": 0.44743511316962337,
      "grad_norm": 1.0,
      "learning_rate": 0.0004985371984183681,
      "loss": 5.861,
      "mean_token_accuracy": 0.16705801486968994,
      "num_tokens": 10383046.0,
      "step": 5120
    },
    {
      "entropy": 6.054336261749268,
      "epoch": 0.44787206152232806,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004985336487764483,
      "loss": 5.9406,
      "mean_token_accuracy": 0.15005630254745483,
      "num_tokens": 10393498.0,
      "step": 5125
    },
    {
      "entropy": 6.110240030288696,
      "epoch": 0.44830900987503275,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004985300948470359,
      "loss": 5.9147,
      "mean_token_accuracy": 0.15699672549962998,
      "num_tokens": 10403595.0,
      "step": 5130
    },
    {
      "entropy": 6.111717271804809,
      "epoch": 0.4487459582277375,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004985265366301992,
      "loss": 5.8574,
      "mean_token_accuracy": 0.15583672225475312,
      "num_tokens": 10413475.0,
      "step": 5135
    },
    {
      "entropy": 6.064886283874512,
      "epoch": 0.4491829065804422,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004985229741260065,
      "loss": 5.8758,
      "mean_token_accuracy": 0.1638584241271019,
      "num_tokens": 10423512.0,
      "step": 5140
    },
    {
      "entropy": 6.083583545684815,
      "epoch": 0.4496198549331469,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004985194073345259,
      "loss": 5.8933,
      "mean_token_accuracy": 0.15554327368736268,
      "num_tokens": 10434198.0,
      "step": 5145
    },
    {
      "entropy": 6.109025287628174,
      "epoch": 0.4500568032858516,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004985158362558261,
      "loss": 5.8197,
      "mean_token_accuracy": 0.1605295419692993,
      "num_tokens": 10444077.0,
      "step": 5150
    },
    {
      "entropy": 6.064793586730957,
      "epoch": 0.4504937516385563,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004985122608899752,
      "loss": 5.8867,
      "mean_token_accuracy": 0.1586834102869034,
      "num_tokens": 10454481.0,
      "step": 5155
    },
    {
      "entropy": 6.120577573776245,
      "epoch": 0.450930699991261,
      "grad_norm": 0.875,
      "learning_rate": 0.0004985086812370422,
      "loss": 5.8943,
      "mean_token_accuracy": 0.16226042658090592,
      "num_tokens": 10464453.0,
      "step": 5160
    },
    {
      "entropy": 6.134780406951904,
      "epoch": 0.45136764834396575,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004985050972970956,
      "loss": 5.9678,
      "mean_token_accuracy": 0.15403684377670288,
      "num_tokens": 10474932.0,
      "step": 5165
    },
    {
      "entropy": 6.126642560958862,
      "epoch": 0.45180459669667045,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000498501509070204,
      "loss": 5.9591,
      "mean_token_accuracy": 0.1517454206943512,
      "num_tokens": 10484719.0,
      "step": 5170
    },
    {
      "entropy": 6.129985237121582,
      "epoch": 0.45224154504937514,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004984979165564364,
      "loss": 5.9909,
      "mean_token_accuracy": 0.14962753057479858,
      "num_tokens": 10495808.0,
      "step": 5175
    },
    {
      "entropy": 6.180895233154297,
      "epoch": 0.4526784934020799,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004984943197558616,
      "loss": 5.894,
      "mean_token_accuracy": 0.15759684890508652,
      "num_tokens": 10506528.0,
      "step": 5180
    },
    {
      "entropy": 6.100573635101318,
      "epoch": 0.4531154417547846,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004984907186685487,
      "loss": 5.8813,
      "mean_token_accuracy": 0.16605969071388244,
      "num_tokens": 10517117.0,
      "step": 5185
    },
    {
      "entropy": 6.166533660888672,
      "epoch": 0.45355239010748927,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004984871132945667,
      "loss": 5.8792,
      "mean_token_accuracy": 0.1546791136264801,
      "num_tokens": 10527394.0,
      "step": 5190
    },
    {
      "entropy": 6.110329866409302,
      "epoch": 0.453989338460194,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004984835036339848,
      "loss": 5.8734,
      "mean_token_accuracy": 0.15998434275388718,
      "num_tokens": 10536779.0,
      "step": 5195
    },
    {
      "entropy": 6.09484977722168,
      "epoch": 0.4544262868128987,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004984798896868722,
      "loss": 5.95,
      "mean_token_accuracy": 0.14918453395366668,
      "num_tokens": 10547116.0,
      "step": 5200
    },
    {
      "entropy": 6.090693998336792,
      "epoch": 0.45486323516560345,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004984762714532981,
      "loss": 5.8305,
      "mean_token_accuracy": 0.16339221000671386,
      "num_tokens": 10556396.0,
      "step": 5205
    },
    {
      "entropy": 6.060329151153565,
      "epoch": 0.45530018351830814,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000498472648933332,
      "loss": 6.017,
      "mean_token_accuracy": 0.15394461452960967,
      "num_tokens": 10568127.0,
      "step": 5210
    },
    {
      "entropy": 6.157448148727417,
      "epoch": 0.45573713187101283,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004984690221270436,
      "loss": 5.8351,
      "mean_token_accuracy": 0.163652141392231,
      "num_tokens": 10578565.0,
      "step": 5215
    },
    {
      "entropy": 6.087811136245728,
      "epoch": 0.4561740802237176,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004984653910345021,
      "loss": 5.8763,
      "mean_token_accuracy": 0.14887666031718255,
      "num_tokens": 10588633.0,
      "step": 5220
    },
    {
      "entropy": 6.125527667999267,
      "epoch": 0.4566110285764223,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004984617556557774,
      "loss": 5.8284,
      "mean_token_accuracy": 0.16143705323338509,
      "num_tokens": 10597917.0,
      "step": 5225
    },
    {
      "entropy": 6.145944738388062,
      "epoch": 0.45704797692912696,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000498458115990939,
      "loss": 5.8797,
      "mean_token_accuracy": 0.1518850490450859,
      "num_tokens": 10608396.0,
      "step": 5230
    },
    {
      "entropy": 6.045504951477051,
      "epoch": 0.4574849252818317,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004984544720400568,
      "loss": 5.9012,
      "mean_token_accuracy": 0.1609129011631012,
      "num_tokens": 10618191.0,
      "step": 5235
    },
    {
      "entropy": 6.115693616867065,
      "epoch": 0.4579218736345364,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000498450823803201,
      "loss": 5.8819,
      "mean_token_accuracy": 0.15265067145228386,
      "num_tokens": 10627594.0,
      "step": 5240
    },
    {
      "entropy": 6.037624931335449,
      "epoch": 0.4583588219872411,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004984471712804412,
      "loss": 5.8556,
      "mean_token_accuracy": 0.16447641402482988,
      "num_tokens": 10637260.0,
      "step": 5245
    },
    {
      "entropy": 6.126225137710572,
      "epoch": 0.45879577033994584,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004984435144718475,
      "loss": 5.8806,
      "mean_token_accuracy": 0.16061059087514878,
      "num_tokens": 10647968.0,
      "step": 5250
    },
    {
      "entropy": 6.047690582275391,
      "epoch": 0.45923271869265053,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004984398533774901,
      "loss": 5.8283,
      "mean_token_accuracy": 0.1624503865838051,
      "num_tokens": 10657760.0,
      "step": 5255
    },
    {
      "entropy": 6.077875328063965,
      "epoch": 0.4596696670453552,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004984361879974393,
      "loss": 5.8548,
      "mean_token_accuracy": 0.15239312797784804,
      "num_tokens": 10667577.0,
      "step": 5260
    },
    {
      "entropy": 6.071667623519898,
      "epoch": 0.46010661539805997,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004984325183317654,
      "loss": 5.8095,
      "mean_token_accuracy": 0.15807173699140548,
      "num_tokens": 10677984.0,
      "step": 5265
    },
    {
      "entropy": 6.0686726570129395,
      "epoch": 0.46054356375076466,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004984288443805386,
      "loss": 5.8646,
      "mean_token_accuracy": 0.16376810520887375,
      "num_tokens": 10687700.0,
      "step": 5270
    },
    {
      "entropy": 6.089095544815064,
      "epoch": 0.46098051210346935,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004984251661438295,
      "loss": 5.8463,
      "mean_token_accuracy": 0.16278012990951538,
      "num_tokens": 10697351.0,
      "step": 5275
    },
    {
      "entropy": 6.053745889663697,
      "epoch": 0.4614174604561741,
      "grad_norm": 1.125,
      "learning_rate": 0.0004984214836217087,
      "loss": 5.876,
      "mean_token_accuracy": 0.1601707324385643,
      "num_tokens": 10707498.0,
      "step": 5280
    },
    {
      "entropy": 6.080528354644775,
      "epoch": 0.4618544088088788,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004984177968142467,
      "loss": 5.9022,
      "mean_token_accuracy": 0.1559564083814621,
      "num_tokens": 10717724.0,
      "step": 5285
    },
    {
      "entropy": 6.118378019332885,
      "epoch": 0.4622913571615835,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004984141057215143,
      "loss": 5.864,
      "mean_token_accuracy": 0.15625895708799362,
      "num_tokens": 10727030.0,
      "step": 5290
    },
    {
      "entropy": 6.041504955291748,
      "epoch": 0.46272830551428823,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004984104103435824,
      "loss": 5.7468,
      "mean_token_accuracy": 0.16861266046762466,
      "num_tokens": 10736861.0,
      "step": 5295
    },
    {
      "entropy": 6.121583938598633,
      "epoch": 0.4631652538669929,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004984067106805217,
      "loss": 5.9697,
      "mean_token_accuracy": 0.15583069622516632,
      "num_tokens": 10748310.0,
      "step": 5300
    },
    {
      "entropy": 6.073477172851563,
      "epoch": 0.4636022022196976,
      "grad_norm": 1.0,
      "learning_rate": 0.0004984030067324032,
      "loss": 5.8063,
      "mean_token_accuracy": 0.1636292114853859,
      "num_tokens": 10758110.0,
      "step": 5305
    },
    {
      "entropy": 6.064459705352784,
      "epoch": 0.46403915057240236,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004983992984992979,
      "loss": 5.8569,
      "mean_token_accuracy": 0.16498022079467772,
      "num_tokens": 10767919.0,
      "step": 5310
    },
    {
      "entropy": 6.065229988098144,
      "epoch": 0.46447609892510705,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004983955859812771,
      "loss": 5.7786,
      "mean_token_accuracy": 0.1648388147354126,
      "num_tokens": 10776888.0,
      "step": 5315
    },
    {
      "entropy": 6.072467899322509,
      "epoch": 0.46491304727781174,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004983918691784119,
      "loss": 5.8742,
      "mean_token_accuracy": 0.1590124249458313,
      "num_tokens": 10786806.0,
      "step": 5320
    },
    {
      "entropy": 6.10413556098938,
      "epoch": 0.4653499956305165,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004983881480907736,
      "loss": 5.8477,
      "mean_token_accuracy": 0.1579391196370125,
      "num_tokens": 10796452.0,
      "step": 5325
    },
    {
      "entropy": 6.104733037948608,
      "epoch": 0.4657869439832212,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004983844227184334,
      "loss": 5.8239,
      "mean_token_accuracy": 0.15865322947502136,
      "num_tokens": 10806718.0,
      "step": 5330
    },
    {
      "entropy": 6.097746181488037,
      "epoch": 0.46622389233592587,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004983806930614631,
      "loss": 5.9065,
      "mean_token_accuracy": 0.1553451269865036,
      "num_tokens": 10817978.0,
      "step": 5335
    },
    {
      "entropy": 6.0263652324676515,
      "epoch": 0.4666608406886306,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000498376959119934,
      "loss": 5.7833,
      "mean_token_accuracy": 0.1663276508450508,
      "num_tokens": 10827818.0,
      "step": 5340
    },
    {
      "entropy": 6.1345278263092045,
      "epoch": 0.4670977890413353,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004983732208939178,
      "loss": 5.9218,
      "mean_token_accuracy": 0.1574622869491577,
      "num_tokens": 10838492.0,
      "step": 5345
    },
    {
      "entropy": 6.077729845046997,
      "epoch": 0.46753473739404,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004983694783834863,
      "loss": 5.825,
      "mean_token_accuracy": 0.15848338454961777,
      "num_tokens": 10847807.0,
      "step": 5350
    },
    {
      "entropy": 6.045288324356079,
      "epoch": 0.46797168574674475,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004983657315887109,
      "loss": 5.9006,
      "mean_token_accuracy": 0.15541983991861344,
      "num_tokens": 10858637.0,
      "step": 5355
    },
    {
      "entropy": 6.122717761993409,
      "epoch": 0.46840863409944944,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004983619805096639,
      "loss": 5.8334,
      "mean_token_accuracy": 0.1633836179971695,
      "num_tokens": 10868042.0,
      "step": 5360
    },
    {
      "entropy": 6.083197355270386,
      "epoch": 0.46884558245215413,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004983582251464171,
      "loss": 5.8429,
      "mean_token_accuracy": 0.1656041830778122,
      "num_tokens": 10877171.0,
      "step": 5365
    },
    {
      "entropy": 6.062127876281738,
      "epoch": 0.4692825308048589,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004983544654990423,
      "loss": 5.8458,
      "mean_token_accuracy": 0.15868037939071655,
      "num_tokens": 10888434.0,
      "step": 5370
    },
    {
      "entropy": 6.090228271484375,
      "epoch": 0.46971947915756357,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000498350701567612,
      "loss": 5.8758,
      "mean_token_accuracy": 0.15781670212745666,
      "num_tokens": 10898467.0,
      "step": 5375
    },
    {
      "entropy": 5.993190670013428,
      "epoch": 0.47015642751026826,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004983469333521983,
      "loss": 5.8475,
      "mean_token_accuracy": 0.16673893183469773,
      "num_tokens": 10908178.0,
      "step": 5380
    },
    {
      "entropy": 6.0736956119537355,
      "epoch": 0.470593375862973,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004983431608528732,
      "loss": 5.9307,
      "mean_token_accuracy": 0.1510281190276146,
      "num_tokens": 10919957.0,
      "step": 5385
    },
    {
      "entropy": 6.17218861579895,
      "epoch": 0.4710303242156777,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004983393840697092,
      "loss": 5.8789,
      "mean_token_accuracy": 0.1570976510643959,
      "num_tokens": 10929728.0,
      "step": 5390
    },
    {
      "entropy": 6.0762652397155765,
      "epoch": 0.4714672725683824,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004983356030027789,
      "loss": 5.9169,
      "mean_token_accuracy": 0.1537831172347069,
      "num_tokens": 10940427.0,
      "step": 5395
    },
    {
      "entropy": 5.99635124206543,
      "epoch": 0.47190422092108714,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004983318176521546,
      "loss": 5.8623,
      "mean_token_accuracy": 0.16047628670930864,
      "num_tokens": 10950641.0,
      "step": 5400
    },
    {
      "entropy": 6.131054782867432,
      "epoch": 0.4723411692737918,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004983280280179092,
      "loss": 5.828,
      "mean_token_accuracy": 0.16604136526584626,
      "num_tokens": 10959341.0,
      "step": 5405
    },
    {
      "entropy": 6.076440000534058,
      "epoch": 0.4727781176264966,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000498324234100115,
      "loss": 5.8583,
      "mean_token_accuracy": 0.16322903633117675,
      "num_tokens": 10967745.0,
      "step": 5410
    },
    {
      "entropy": 6.018764305114746,
      "epoch": 0.47321506597920127,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004983204358988451,
      "loss": 5.8205,
      "mean_token_accuracy": 0.16208201944828032,
      "num_tokens": 10978345.0,
      "step": 5415
    },
    {
      "entropy": 6.096686792373657,
      "epoch": 0.47365201433190596,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004983166334141722,
      "loss": 5.8818,
      "mean_token_accuracy": 0.15515933856368064,
      "num_tokens": 10988030.0,
      "step": 5420
    },
    {
      "entropy": 6.145278215408325,
      "epoch": 0.4740889626846107,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004983128266461693,
      "loss": 5.8527,
      "mean_token_accuracy": 0.16152588576078414,
      "num_tokens": 10999169.0,
      "step": 5425
    },
    {
      "entropy": 6.05873441696167,
      "epoch": 0.4745259110373154,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004983090155949093,
      "loss": 5.9396,
      "mean_token_accuracy": 0.1501481741666794,
      "num_tokens": 11009608.0,
      "step": 5430
    },
    {
      "entropy": 6.13284592628479,
      "epoch": 0.4749628593900201,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004983052002604653,
      "loss": 5.8727,
      "mean_token_accuracy": 0.15617259219288826,
      "num_tokens": 11020187.0,
      "step": 5435
    },
    {
      "entropy": 6.095590591430664,
      "epoch": 0.47539980774272483,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004983013806429107,
      "loss": 5.8371,
      "mean_token_accuracy": 0.16051960587501526,
      "num_tokens": 11030459.0,
      "step": 5440
    },
    {
      "entropy": 6.00195426940918,
      "epoch": 0.4758367560954295,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004982975567423185,
      "loss": 5.7936,
      "mean_token_accuracy": 0.1667706102132797,
      "num_tokens": 11040880.0,
      "step": 5445
    },
    {
      "entropy": 6.054232740402222,
      "epoch": 0.4762737044481342,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004982937285587623,
      "loss": 5.8042,
      "mean_token_accuracy": 0.16394554674625397,
      "num_tokens": 11050901.0,
      "step": 5450
    },
    {
      "entropy": 6.088509798049927,
      "epoch": 0.47671065280083896,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004982898960923152,
      "loss": 5.9762,
      "mean_token_accuracy": 0.14890825003385544,
      "num_tokens": 11061274.0,
      "step": 5455
    },
    {
      "entropy": 6.127956914901733,
      "epoch": 0.47714760115354365,
      "grad_norm": 1.0,
      "learning_rate": 0.0004982860593430509,
      "loss": 5.794,
      "mean_token_accuracy": 0.15740275830030442,
      "num_tokens": 11070644.0,
      "step": 5460
    },
    {
      "entropy": 6.042313289642334,
      "epoch": 0.47758454950624835,
      "grad_norm": 0.875,
      "learning_rate": 0.0004982822183110431,
      "loss": 5.8447,
      "mean_token_accuracy": 0.1646100953221321,
      "num_tokens": 11081154.0,
      "step": 5465
    },
    {
      "entropy": 6.081279993057251,
      "epoch": 0.4780214978589531,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004982783729963654,
      "loss": 5.9438,
      "mean_token_accuracy": 0.15628594309091567,
      "num_tokens": 11091929.0,
      "step": 5470
    },
    {
      "entropy": 6.000416278839111,
      "epoch": 0.4784584462116578,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004982745233990914,
      "loss": 5.7724,
      "mean_token_accuracy": 0.17148518860340117,
      "num_tokens": 11102075.0,
      "step": 5475
    },
    {
      "entropy": 6.0447105884552,
      "epoch": 0.4788953945643625,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004982706695192949,
      "loss": 5.8729,
      "mean_token_accuracy": 0.15767095237970352,
      "num_tokens": 11111886.0,
      "step": 5480
    },
    {
      "entropy": 6.1248678207397464,
      "epoch": 0.4793323429170672,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004982668113570503,
      "loss": 5.8597,
      "mean_token_accuracy": 0.15638172328472139,
      "num_tokens": 11122329.0,
      "step": 5485
    },
    {
      "entropy": 6.062728977203369,
      "epoch": 0.4797692912697719,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000498262948912431,
      "loss": 5.7911,
      "mean_token_accuracy": 0.1594435304403305,
      "num_tokens": 11132348.0,
      "step": 5490
    },
    {
      "entropy": 6.077106475830078,
      "epoch": 0.4802062396224766,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004982590821855115,
      "loss": 5.8575,
      "mean_token_accuracy": 0.16328706592321396,
      "num_tokens": 11141975.0,
      "step": 5495
    },
    {
      "entropy": 6.098625373840332,
      "epoch": 0.48064318797518135,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004982552111763657,
      "loss": 5.8586,
      "mean_token_accuracy": 0.15738974511623383,
      "num_tokens": 11152174.0,
      "step": 5500
    },
    {
      "entropy": 6.052024221420288,
      "epoch": 0.48108013632788604,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004982513358850681,
      "loss": 5.8605,
      "mean_token_accuracy": 0.15670732408761978,
      "num_tokens": 11162563.0,
      "step": 5505
    },
    {
      "entropy": 6.084334468841552,
      "epoch": 0.48151708468059073,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004982474563116928,
      "loss": 5.7676,
      "mean_token_accuracy": 0.16274628937244415,
      "num_tokens": 11172572.0,
      "step": 5510
    },
    {
      "entropy": 6.056139326095581,
      "epoch": 0.4819540330332955,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004982435724563143,
      "loss": 5.8529,
      "mean_token_accuracy": 0.1674552872776985,
      "num_tokens": 11183404.0,
      "step": 5515
    },
    {
      "entropy": 6.025302410125732,
      "epoch": 0.4823909813860002,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004982396843190072,
      "loss": 5.8477,
      "mean_token_accuracy": 0.16038714200258256,
      "num_tokens": 11193905.0,
      "step": 5520
    },
    {
      "entropy": 6.099229335784912,
      "epoch": 0.48282792973870486,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004982357918998459,
      "loss": 5.7562,
      "mean_token_accuracy": 0.17048548460006713,
      "num_tokens": 11203127.0,
      "step": 5525
    },
    {
      "entropy": 6.02909369468689,
      "epoch": 0.4832648780914096,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004982318951989053,
      "loss": 5.7721,
      "mean_token_accuracy": 0.16809311956167222,
      "num_tokens": 11213468.0,
      "step": 5530
    },
    {
      "entropy": 6.023222398757935,
      "epoch": 0.4837018264441143,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004982279942162597,
      "loss": 5.8596,
      "mean_token_accuracy": 0.16149110794067384,
      "num_tokens": 11224722.0,
      "step": 5535
    },
    {
      "entropy": 6.100701427459716,
      "epoch": 0.484138774796819,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004982240889519844,
      "loss": 5.8437,
      "mean_token_accuracy": 0.16304720044136048,
      "num_tokens": 11234565.0,
      "step": 5540
    },
    {
      "entropy": 6.084958171844482,
      "epoch": 0.48457572314952374,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000498220179406154,
      "loss": 5.8372,
      "mean_token_accuracy": 0.16213309168815612,
      "num_tokens": 11244496.0,
      "step": 5545
    },
    {
      "entropy": 6.096549654006958,
      "epoch": 0.48501267150222843,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004982162655788436,
      "loss": 5.8194,
      "mean_token_accuracy": 0.16341753602027892,
      "num_tokens": 11254187.0,
      "step": 5550
    },
    {
      "entropy": 6.047856283187866,
      "epoch": 0.4854496198549331,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004982123474701283,
      "loss": 5.7992,
      "mean_token_accuracy": 0.16489096730947495,
      "num_tokens": 11264580.0,
      "step": 5555
    },
    {
      "entropy": 6.067271184921265,
      "epoch": 0.48588656820763787,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004982084250800832,
      "loss": 5.9469,
      "mean_token_accuracy": 0.15531962588429452,
      "num_tokens": 11275104.0,
      "step": 5560
    },
    {
      "entropy": 6.1391197681427006,
      "epoch": 0.48632351656034256,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004982044984087836,
      "loss": 5.9635,
      "mean_token_accuracy": 0.1547545611858368,
      "num_tokens": 11285983.0,
      "step": 5565
    },
    {
      "entropy": 6.03480339050293,
      "epoch": 0.48676046491304725,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004982005674563048,
      "loss": 5.8322,
      "mean_token_accuracy": 0.16253410577774047,
      "num_tokens": 11295378.0,
      "step": 5570
    },
    {
      "entropy": 6.071707391738892,
      "epoch": 0.487197413265752,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000498196632222722,
      "loss": 5.9392,
      "mean_token_accuracy": 0.15828753113746644,
      "num_tokens": 11305500.0,
      "step": 5575
    },
    {
      "entropy": 6.089603424072266,
      "epoch": 0.4876343616184567,
      "grad_norm": 0.875,
      "learning_rate": 0.0004981926927081109,
      "loss": 5.8472,
      "mean_token_accuracy": 0.1600751742720604,
      "num_tokens": 11316455.0,
      "step": 5580
    },
    {
      "entropy": 6.074129629135132,
      "epoch": 0.4880713099711614,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004981887489125471,
      "loss": 5.8112,
      "mean_token_accuracy": 0.1575797788798809,
      "num_tokens": 11325865.0,
      "step": 5585
    },
    {
      "entropy": 5.981272888183594,
      "epoch": 0.48850825832386613,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000498184800836106,
      "loss": 5.7768,
      "mean_token_accuracy": 0.17487731128931044,
      "num_tokens": 11337368.0,
      "step": 5590
    },
    {
      "entropy": 6.0062541484832765,
      "epoch": 0.4889452066765708,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004981808484788635,
      "loss": 5.7523,
      "mean_token_accuracy": 0.17290766686201095,
      "num_tokens": 11347129.0,
      "step": 5595
    },
    {
      "entropy": 6.070178651809693,
      "epoch": 0.48938215502927557,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004981768918408955,
      "loss": 5.831,
      "mean_token_accuracy": 0.15914562493562698,
      "num_tokens": 11356641.0,
      "step": 5600
    },
    {
      "entropy": 6.019480848312378,
      "epoch": 0.48981910338198026,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004981729309222777,
      "loss": 5.7193,
      "mean_token_accuracy": 0.17006878405809403,
      "num_tokens": 11366485.0,
      "step": 5605
    },
    {
      "entropy": 6.009759902954102,
      "epoch": 0.49025605173468495,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004981689657230861,
      "loss": 5.8558,
      "mean_token_accuracy": 0.16212911680340766,
      "num_tokens": 11376305.0,
      "step": 5610
    },
    {
      "entropy": 6.054521274566651,
      "epoch": 0.4906930000873897,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000498164996243397,
      "loss": 5.8259,
      "mean_token_accuracy": 0.16426292508840562,
      "num_tokens": 11385837.0,
      "step": 5615
    },
    {
      "entropy": 6.0769446849822994,
      "epoch": 0.4911299484400944,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004981610224832861,
      "loss": 5.8674,
      "mean_token_accuracy": 0.16243572831153869,
      "num_tokens": 11396719.0,
      "step": 5620
    },
    {
      "entropy": 6.109344911575318,
      "epoch": 0.4915668967927991,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004981570444428299,
      "loss": 5.8794,
      "mean_token_accuracy": 0.15813525915145873,
      "num_tokens": 11408107.0,
      "step": 5625
    },
    {
      "entropy": 6.026615953445434,
      "epoch": 0.4920038451455038,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004981530621221048,
      "loss": 5.7906,
      "mean_token_accuracy": 0.17031390815973282,
      "num_tokens": 11417547.0,
      "step": 5630
    },
    {
      "entropy": 6.092897939682007,
      "epoch": 0.4924407934982085,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004981490755211869,
      "loss": 5.8722,
      "mean_token_accuracy": 0.16181053072214127,
      "num_tokens": 11427660.0,
      "step": 5635
    },
    {
      "entropy": 5.994324016571045,
      "epoch": 0.4928777418509132,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004981450846401528,
      "loss": 5.7465,
      "mean_token_accuracy": 0.16703419536352157,
      "num_tokens": 11438032.0,
      "step": 5640
    },
    {
      "entropy": 6.065396118164062,
      "epoch": 0.49331469020361796,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000498141089479079,
      "loss": 5.8522,
      "mean_token_accuracy": 0.15965607166290283,
      "num_tokens": 11448987.0,
      "step": 5645
    },
    {
      "entropy": 6.04461407661438,
      "epoch": 0.49375163855632265,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004981370900380423,
      "loss": 5.8294,
      "mean_token_accuracy": 0.16260455697774887,
      "num_tokens": 11458659.0,
      "step": 5650
    },
    {
      "entropy": 6.053560400009156,
      "epoch": 0.49418858690902734,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004981330863171194,
      "loss": 5.8066,
      "mean_token_accuracy": 0.16022174954414367,
      "num_tokens": 11468104.0,
      "step": 5655
    },
    {
      "entropy": 6.050134658813477,
      "epoch": 0.4946255352617321,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004981290783163867,
      "loss": 5.7662,
      "mean_token_accuracy": 0.16689103543758393,
      "num_tokens": 11479579.0,
      "step": 5660
    },
    {
      "entropy": 6.049282073974609,
      "epoch": 0.4950624836144368,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004981250660359216,
      "loss": 5.8378,
      "mean_token_accuracy": 0.1668689325451851,
      "num_tokens": 11489202.0,
      "step": 5665
    },
    {
      "entropy": 6.028315734863281,
      "epoch": 0.49549943196714147,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004981210494758007,
      "loss": 5.8735,
      "mean_token_accuracy": 0.15290959030389786,
      "num_tokens": 11499425.0,
      "step": 5670
    },
    {
      "entropy": 6.09428915977478,
      "epoch": 0.4959363803198462,
      "grad_norm": 1.0,
      "learning_rate": 0.000498117028636101,
      "loss": 5.8559,
      "mean_token_accuracy": 0.1578807756304741,
      "num_tokens": 11508983.0,
      "step": 5675
    },
    {
      "entropy": 6.036450290679932,
      "epoch": 0.4963733286725509,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004981130035169,
      "loss": 5.8957,
      "mean_token_accuracy": 0.1656028926372528,
      "num_tokens": 11519791.0,
      "step": 5680
    },
    {
      "entropy": 6.058498096466065,
      "epoch": 0.4968102770252556,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004981089741182747,
      "loss": 5.8084,
      "mean_token_accuracy": 0.1662641003727913,
      "num_tokens": 11530537.0,
      "step": 5685
    },
    {
      "entropy": 6.059791564941406,
      "epoch": 0.49724722537796034,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004981049404403023,
      "loss": 5.8354,
      "mean_token_accuracy": 0.16185742169618605,
      "num_tokens": 11540591.0,
      "step": 5690
    },
    {
      "entropy": 5.9863921165466305,
      "epoch": 0.49768417373066504,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004981009024830602,
      "loss": 5.7767,
      "mean_token_accuracy": 0.16423231661319732,
      "num_tokens": 11550719.0,
      "step": 5695
    },
    {
      "entropy": 6.011070585250854,
      "epoch": 0.4981211220833697,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000498096860246626,
      "loss": 5.8403,
      "mean_token_accuracy": 0.16285658478736878,
      "num_tokens": 11561805.0,
      "step": 5700
    },
    {
      "entropy": 6.165259504318238,
      "epoch": 0.4985580704360745,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000498092813731077,
      "loss": 5.8735,
      "mean_token_accuracy": 0.1558795154094696,
      "num_tokens": 11571380.0,
      "step": 5705
    },
    {
      "entropy": 6.012100601196289,
      "epoch": 0.49899501878877917,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000498088762936491,
      "loss": 5.8465,
      "mean_token_accuracy": 0.15960172265768052,
      "num_tokens": 11580853.0,
      "step": 5710
    },
    {
      "entropy": 5.926440191268921,
      "epoch": 0.49943196714148386,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004980847078629457,
      "loss": 5.7472,
      "mean_token_accuracy": 0.17658197581768037,
      "num_tokens": 11591966.0,
      "step": 5715
    },
    {
      "entropy": 6.049602222442627,
      "epoch": 0.4998689154941886,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004980806485105186,
      "loss": 5.8151,
      "mean_token_accuracy": 0.16555469185113908,
      "num_tokens": 11601239.0,
      "step": 5720
    },
    {
      "entropy": 6.057469320297241,
      "epoch": 0.5003058638468934,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004980765848792879,
      "loss": 5.737,
      "mean_token_accuracy": 0.1763659179210663,
      "num_tokens": 11609857.0,
      "step": 5725
    },
    {
      "entropy": 6.077443647384643,
      "epoch": 0.500742812199598,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004980725169693314,
      "loss": 5.793,
      "mean_token_accuracy": 0.16501347720623016,
      "num_tokens": 11618952.0,
      "step": 5730
    },
    {
      "entropy": 6.033372402191162,
      "epoch": 0.5011797605523027,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004980684447807272,
      "loss": 5.7241,
      "mean_token_accuracy": 0.17032039165496826,
      "num_tokens": 11627638.0,
      "step": 5735
    },
    {
      "entropy": 6.073156690597534,
      "epoch": 0.5016167089050074,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004980643683135533,
      "loss": 5.8382,
      "mean_token_accuracy": 0.165236958861351,
      "num_tokens": 11637842.0,
      "step": 5740
    },
    {
      "entropy": 6.023643064498901,
      "epoch": 0.5020536572577121,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004980602875678878,
      "loss": 5.7428,
      "mean_token_accuracy": 0.1751474380493164,
      "num_tokens": 11647217.0,
      "step": 5745
    },
    {
      "entropy": 5.952885913848877,
      "epoch": 0.5024906056104168,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004980562025438093,
      "loss": 5.7367,
      "mean_token_accuracy": 0.16680333018302917,
      "num_tokens": 11657558.0,
      "step": 5750
    },
    {
      "entropy": 6.0496491432189945,
      "epoch": 0.5029275539631216,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004980521132413958,
      "loss": 5.749,
      "mean_token_accuracy": 0.1631416290998459,
      "num_tokens": 11668054.0,
      "step": 5755
    },
    {
      "entropy": 6.065122604370117,
      "epoch": 0.5033645023158263,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000498048019660726,
      "loss": 5.7608,
      "mean_token_accuracy": 0.1645866945385933,
      "num_tokens": 11677779.0,
      "step": 5760
    },
    {
      "entropy": 6.0388609886169435,
      "epoch": 0.503801450668531,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004980439218018783,
      "loss": 5.8191,
      "mean_token_accuracy": 0.16016570776700972,
      "num_tokens": 11687577.0,
      "step": 5765
    },
    {
      "entropy": 6.143608713150025,
      "epoch": 0.5042383990212357,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004980398196649312,
      "loss": 5.9225,
      "mean_token_accuracy": 0.1562805950641632,
      "num_tokens": 11699630.0,
      "step": 5770
    },
    {
      "entropy": 6.046047735214233,
      "epoch": 0.5046753473739404,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004980357132499636,
      "loss": 5.8515,
      "mean_token_accuracy": 0.15963000506162645,
      "num_tokens": 11710139.0,
      "step": 5775
    },
    {
      "entropy": 6.0711297512054445,
      "epoch": 0.5051122957266451,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004980316025570542,
      "loss": 5.8034,
      "mean_token_accuracy": 0.16327530294656753,
      "num_tokens": 11719070.0,
      "step": 5780
    },
    {
      "entropy": 6.022385835647583,
      "epoch": 0.5055492440793499,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004980274875862816,
      "loss": 5.8419,
      "mean_token_accuracy": 0.1634025901556015,
      "num_tokens": 11728770.0,
      "step": 5785
    },
    {
      "entropy": 6.030874395370484,
      "epoch": 0.5059861924320546,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000498023368337725,
      "loss": 5.7771,
      "mean_token_accuracy": 0.16805996745824814,
      "num_tokens": 11737753.0,
      "step": 5790
    },
    {
      "entropy": 5.990839719772339,
      "epoch": 0.5064231407847593,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004980192448114635,
      "loss": 5.7601,
      "mean_token_accuracy": 0.1641977235674858,
      "num_tokens": 11747837.0,
      "step": 5795
    },
    {
      "entropy": 6.0531830310821535,
      "epoch": 0.5068600891374639,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004980151170075758,
      "loss": 5.8853,
      "mean_token_accuracy": 0.1629597380757332,
      "num_tokens": 11757911.0,
      "step": 5800
    },
    {
      "entropy": 6.1099793910980225,
      "epoch": 0.5072970374901686,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004980109849261414,
      "loss": 5.7496,
      "mean_token_accuracy": 0.16389151364564897,
      "num_tokens": 11766213.0,
      "step": 5805
    },
    {
      "entropy": 5.956259536743164,
      "epoch": 0.5077339858428733,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004980068485672394,
      "loss": 5.7053,
      "mean_token_accuracy": 0.17541376501321793,
      "num_tokens": 11775640.0,
      "step": 5810
    },
    {
      "entropy": 6.052332258224487,
      "epoch": 0.5081709341955781,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004980027079309493,
      "loss": 5.7603,
      "mean_token_accuracy": 0.16962587982416152,
      "num_tokens": 11785664.0,
      "step": 5815
    },
    {
      "entropy": 6.101211261749268,
      "epoch": 0.5086078825482828,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004979985630173503,
      "loss": 5.8708,
      "mean_token_accuracy": 0.1568834140896797,
      "num_tokens": 11795402.0,
      "step": 5820
    },
    {
      "entropy": 6.002501583099365,
      "epoch": 0.5090448309009875,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004979944138265221,
      "loss": 5.7352,
      "mean_token_accuracy": 0.16570198237895967,
      "num_tokens": 11805923.0,
      "step": 5825
    },
    {
      "entropy": 6.021250009536743,
      "epoch": 0.5094817792536922,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004979902603585442,
      "loss": 5.8499,
      "mean_token_accuracy": 0.159707410633564,
      "num_tokens": 11816097.0,
      "step": 5830
    },
    {
      "entropy": 6.048093175888061,
      "epoch": 0.5099187276063969,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004979861026134961,
      "loss": 5.7511,
      "mean_token_accuracy": 0.16532458364963531,
      "num_tokens": 11826889.0,
      "step": 5835
    },
    {
      "entropy": 6.058287286758423,
      "epoch": 0.5103556759591016,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004979819405914579,
      "loss": 5.8224,
      "mean_token_accuracy": 0.16387106627225875,
      "num_tokens": 11837935.0,
      "step": 5840
    },
    {
      "entropy": 6.050278091430664,
      "epoch": 0.5107926243118064,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004979777742925092,
      "loss": 5.8071,
      "mean_token_accuracy": 0.16187942028045654,
      "num_tokens": 11847715.0,
      "step": 5845
    },
    {
      "entropy": 6.108582592010498,
      "epoch": 0.5112295726645111,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004979736037167298,
      "loss": 5.8401,
      "mean_token_accuracy": 0.16293361485004426,
      "num_tokens": 11857841.0,
      "step": 5850
    },
    {
      "entropy": 6.041381406784057,
      "epoch": 0.5116665210172158,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004979694288642,
      "loss": 5.828,
      "mean_token_accuracy": 0.16747580021619796,
      "num_tokens": 11868135.0,
      "step": 5855
    },
    {
      "entropy": 6.044627094268799,
      "epoch": 0.5121034693699205,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004979652497349997,
      "loss": 5.7595,
      "mean_token_accuracy": 0.16045006960630417,
      "num_tokens": 11877637.0,
      "step": 5860
    },
    {
      "entropy": 5.958197832107544,
      "epoch": 0.5125404177226252,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000497961066329209,
      "loss": 5.8228,
      "mean_token_accuracy": 0.1567137211561203,
      "num_tokens": 11887128.0,
      "step": 5865
    },
    {
      "entropy": 6.115878343582153,
      "epoch": 0.5129773660753298,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004979568786469083,
      "loss": 5.8504,
      "mean_token_accuracy": 0.1540878966450691,
      "num_tokens": 11897885.0,
      "step": 5870
    },
    {
      "entropy": 6.017870473861694,
      "epoch": 0.5134143144280346,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004979526866881777,
      "loss": 5.7764,
      "mean_token_accuracy": 0.16695483028888702,
      "num_tokens": 11907919.0,
      "step": 5875
    },
    {
      "entropy": 5.91297402381897,
      "epoch": 0.5138512627807393,
      "grad_norm": 0.90625,
      "learning_rate": 0.000497948490453098,
      "loss": 5.749,
      "mean_token_accuracy": 0.16705156862735748,
      "num_tokens": 11918612.0,
      "step": 5880
    },
    {
      "entropy": 5.986278533935547,
      "epoch": 0.514288211133444,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004979442899417493,
      "loss": 5.8105,
      "mean_token_accuracy": 0.1623510330915451,
      "num_tokens": 11927759.0,
      "step": 5885
    },
    {
      "entropy": 6.073221349716187,
      "epoch": 0.5147251594861487,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004979400851542122,
      "loss": 5.7793,
      "mean_token_accuracy": 0.16810414046049119,
      "num_tokens": 11937522.0,
      "step": 5890
    },
    {
      "entropy": 5.950752019882202,
      "epoch": 0.5151621078388534,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004979358760905675,
      "loss": 5.7659,
      "mean_token_accuracy": 0.16226734668016435,
      "num_tokens": 11947787.0,
      "step": 5895
    },
    {
      "entropy": 5.993432235717774,
      "epoch": 0.5155990561915581,
      "grad_norm": 0.890625,
      "learning_rate": 0.000497931662750896,
      "loss": 5.8018,
      "mean_token_accuracy": 0.16491454988718032,
      "num_tokens": 11958929.0,
      "step": 5900
    },
    {
      "entropy": 6.028296089172363,
      "epoch": 0.5160360045442629,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004979274451352783,
      "loss": 5.7525,
      "mean_token_accuracy": 0.16111301630735397,
      "num_tokens": 11969242.0,
      "step": 5905
    },
    {
      "entropy": 6.05470871925354,
      "epoch": 0.5164729528969676,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004979232232437954,
      "loss": 5.8135,
      "mean_token_accuracy": 0.1630721300840378,
      "num_tokens": 11980149.0,
      "step": 5910
    },
    {
      "entropy": 6.071461725234985,
      "epoch": 0.5169099012496723,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004979189970765282,
      "loss": 5.8085,
      "mean_token_accuracy": 0.1617502823472023,
      "num_tokens": 11990636.0,
      "step": 5915
    },
    {
      "entropy": 6.038999891281128,
      "epoch": 0.517346849602377,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004979147666335579,
      "loss": 5.8137,
      "mean_token_accuracy": 0.1656518757343292,
      "num_tokens": 12000615.0,
      "step": 5920
    },
    {
      "entropy": 6.024601888656616,
      "epoch": 0.5177837979550817,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004979105319149656,
      "loss": 5.8525,
      "mean_token_accuracy": 0.16385390758514404,
      "num_tokens": 12011072.0,
      "step": 5925
    },
    {
      "entropy": 6.008143901824951,
      "epoch": 0.5182207463077865,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004979062929208324,
      "loss": 5.7506,
      "mean_token_accuracy": 0.16876157373189926,
      "num_tokens": 12021726.0,
      "step": 5930
    },
    {
      "entropy": 6.075357961654663,
      "epoch": 0.5186576946604912,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004979020496512398,
      "loss": 5.8845,
      "mean_token_accuracy": 0.1558486267924309,
      "num_tokens": 12031353.0,
      "step": 5935
    },
    {
      "entropy": 6.0586957931518555,
      "epoch": 0.5190946430131959,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000497897802106269,
      "loss": 5.8251,
      "mean_token_accuracy": 0.1579397052526474,
      "num_tokens": 12041234.0,
      "step": 5940
    },
    {
      "entropy": 5.9722131252288815,
      "epoch": 0.5195315913659005,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004978935502860017,
      "loss": 5.7596,
      "mean_token_accuracy": 0.16703967601060868,
      "num_tokens": 12051100.0,
      "step": 5945
    },
    {
      "entropy": 6.037038087844849,
      "epoch": 0.5199685397186052,
      "grad_norm": 0.875,
      "learning_rate": 0.0004978892941905191,
      "loss": 5.806,
      "mean_token_accuracy": 0.16735392659902573,
      "num_tokens": 12062012.0,
      "step": 5950
    },
    {
      "entropy": 6.086524915695191,
      "epoch": 0.5204054880713099,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004978850338199033,
      "loss": 5.8402,
      "mean_token_accuracy": 0.1612513154745102,
      "num_tokens": 12072763.0,
      "step": 5955
    },
    {
      "entropy": 5.9862555980682375,
      "epoch": 0.5208424364240147,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004978807691742356,
      "loss": 5.6893,
      "mean_token_accuracy": 0.1724808156490326,
      "num_tokens": 12081892.0,
      "step": 5960
    },
    {
      "entropy": 6.019426059722901,
      "epoch": 0.5212793847767194,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000497876500253598,
      "loss": 5.7704,
      "mean_token_accuracy": 0.16536674797534942,
      "num_tokens": 12091428.0,
      "step": 5965
    },
    {
      "entropy": 5.9819176197052,
      "epoch": 0.5217163331294241,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004978722270580722,
      "loss": 5.7747,
      "mean_token_accuracy": 0.16727808862924576,
      "num_tokens": 12101687.0,
      "step": 5970
    },
    {
      "entropy": 5.963487434387207,
      "epoch": 0.5221532814821288,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004978679495877405,
      "loss": 5.7282,
      "mean_token_accuracy": 0.1675598442554474,
      "num_tokens": 12111726.0,
      "step": 5975
    },
    {
      "entropy": 6.071728181838989,
      "epoch": 0.5225902298348335,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004978636678426847,
      "loss": 5.755,
      "mean_token_accuracy": 0.1716315284371376,
      "num_tokens": 12122171.0,
      "step": 5980
    },
    {
      "entropy": 6.087211227416992,
      "epoch": 0.5230271781875382,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004978593818229869,
      "loss": 5.8079,
      "mean_token_accuracy": 0.16242510229349136,
      "num_tokens": 12131414.0,
      "step": 5985
    },
    {
      "entropy": 5.968759298324585,
      "epoch": 0.523464126540243,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004978550915287294,
      "loss": 5.7627,
      "mean_token_accuracy": 0.15576607882976531,
      "num_tokens": 12141665.0,
      "step": 5990
    },
    {
      "entropy": 6.049363803863526,
      "epoch": 0.5239010748929477,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004978507969599945,
      "loss": 5.8318,
      "mean_token_accuracy": 0.15878561586141587,
      "num_tokens": 12151263.0,
      "step": 5995
    },
    {
      "entropy": 6.077195549011231,
      "epoch": 0.5243380232456524,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004978464981168645,
      "loss": 5.8145,
      "mean_token_accuracy": 0.16041911840438844,
      "num_tokens": 12161070.0,
      "step": 6000
    },
    {
      "epoch": 0.5243380232456524,
      "eval_entropy": 5.812668307543265,
      "eval_loss": 5.803792476654053,
      "eval_mean_token_accuracy": 0.17073484680949289,
      "eval_num_tokens": 12161070.0,
      "eval_runtime": 21.6682,
      "eval_samples_per_second": 1708.221,
      "eval_steps_per_second": 213.539,
      "step": 6000
    },
    {
      "entropy": 6.0304847240448,
      "epoch": 0.5247749715983571,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000497842194999422,
      "loss": 5.7829,
      "mean_token_accuracy": 0.1679076761007309,
      "num_tokens": 12171197.0,
      "step": 6005
    },
    {
      "entropy": 6.032250595092774,
      "epoch": 0.5252119199510618,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004978378876077493,
      "loss": 5.8105,
      "mean_token_accuracy": 0.16307166516780852,
      "num_tokens": 12181462.0,
      "step": 6010
    },
    {
      "entropy": 6.014563846588135,
      "epoch": 0.5256488683037664,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004978335759419293,
      "loss": 5.7857,
      "mean_token_accuracy": 0.16811424642801284,
      "num_tokens": 12191473.0,
      "step": 6015
    },
    {
      "entropy": 6.06669340133667,
      "epoch": 0.5260858166564713,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004978292600020444,
      "loss": 5.7419,
      "mean_token_accuracy": 0.16636988520622253,
      "num_tokens": 12200920.0,
      "step": 6020
    },
    {
      "entropy": 6.057852363586425,
      "epoch": 0.5265227650091759,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004978249397881776,
      "loss": 5.8126,
      "mean_token_accuracy": 0.16200290471315384,
      "num_tokens": 12210823.0,
      "step": 6025
    },
    {
      "entropy": 6.055155944824219,
      "epoch": 0.5269597133618806,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004978206153004117,
      "loss": 5.8265,
      "mean_token_accuracy": 0.16370147168636323,
      "num_tokens": 12221967.0,
      "step": 6030
    },
    {
      "entropy": 6.020875787734985,
      "epoch": 0.5273966617145853,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004978162865388297,
      "loss": 5.782,
      "mean_token_accuracy": 0.16336679309606553,
      "num_tokens": 12232519.0,
      "step": 6035
    },
    {
      "entropy": 6.05156192779541,
      "epoch": 0.52783361006729,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004978119535035144,
      "loss": 5.7621,
      "mean_token_accuracy": 0.16045322120189667,
      "num_tokens": 12243193.0,
      "step": 6040
    },
    {
      "entropy": 6.081714200973511,
      "epoch": 0.5282705584199947,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004978076161945492,
      "loss": 5.7604,
      "mean_token_accuracy": 0.16626867204904555,
      "num_tokens": 12253640.0,
      "step": 6045
    },
    {
      "entropy": 6.128384017944336,
      "epoch": 0.5287075067726995,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004978032746120173,
      "loss": 5.8846,
      "mean_token_accuracy": 0.1566349059343338,
      "num_tokens": 12265072.0,
      "step": 6050
    },
    {
      "entropy": 6.049355840682983,
      "epoch": 0.5291444551254042,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004977989287560016,
      "loss": 5.8196,
      "mean_token_accuracy": 0.1635644853115082,
      "num_tokens": 12274449.0,
      "step": 6055
    },
    {
      "entropy": 6.086436796188354,
      "epoch": 0.5295814034781089,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004977945786265857,
      "loss": 5.8203,
      "mean_token_accuracy": 0.15821414291858674,
      "num_tokens": 12284910.0,
      "step": 6060
    },
    {
      "entropy": 6.08204493522644,
      "epoch": 0.5300183518308136,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004977902242238533,
      "loss": 5.831,
      "mean_token_accuracy": 0.16470483988523482,
      "num_tokens": 12294046.0,
      "step": 6065
    },
    {
      "entropy": 5.985720920562744,
      "epoch": 0.5304553001835183,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004977858655478874,
      "loss": 5.821,
      "mean_token_accuracy": 0.1622784748673439,
      "num_tokens": 12303900.0,
      "step": 6070
    },
    {
      "entropy": 6.100087261199951,
      "epoch": 0.530892248536223,
      "grad_norm": 1.046875,
      "learning_rate": 0.000497781502598772,
      "loss": 5.759,
      "mean_token_accuracy": 0.16908952593803406,
      "num_tokens": 12313326.0,
      "step": 6075
    },
    {
      "entropy": 6.028599691390991,
      "epoch": 0.5313291968889278,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004977771353765906,
      "loss": 5.7521,
      "mean_token_accuracy": 0.16284270733594894,
      "num_tokens": 12322792.0,
      "step": 6080
    },
    {
      "entropy": 6.041930627822876,
      "epoch": 0.5317661452416325,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000497772763881427,
      "loss": 5.8635,
      "mean_token_accuracy": 0.1564528912305832,
      "num_tokens": 12333417.0,
      "step": 6085
    },
    {
      "entropy": 6.093335771560669,
      "epoch": 0.5322030935943372,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004977683881133651,
      "loss": 5.7688,
      "mean_token_accuracy": 0.1608974501490593,
      "num_tokens": 12343944.0,
      "step": 6090
    },
    {
      "entropy": 6.014216899871826,
      "epoch": 0.5326400419470418,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004977640080724888,
      "loss": 5.7729,
      "mean_token_accuracy": 0.1603822112083435,
      "num_tokens": 12353838.0,
      "step": 6095
    },
    {
      "entropy": 6.045888519287109,
      "epoch": 0.5330769902997465,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004977596237588821,
      "loss": 5.829,
      "mean_token_accuracy": 0.1591583475470543,
      "num_tokens": 12364996.0,
      "step": 6100
    },
    {
      "entropy": 6.05091347694397,
      "epoch": 0.5335139386524512,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000497755235172629,
      "loss": 5.8356,
      "mean_token_accuracy": 0.16187589168548583,
      "num_tokens": 12375111.0,
      "step": 6105
    },
    {
      "entropy": 5.986320829391479,
      "epoch": 0.533950887005156,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004977508423138139,
      "loss": 5.7983,
      "mean_token_accuracy": 0.1610079899430275,
      "num_tokens": 12384376.0,
      "step": 6110
    },
    {
      "entropy": 6.06788010597229,
      "epoch": 0.5343878353578607,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004977464451825208,
      "loss": 5.8486,
      "mean_token_accuracy": 0.15392994210124017,
      "num_tokens": 12395248.0,
      "step": 6115
    },
    {
      "entropy": 6.10393648147583,
      "epoch": 0.5348247837105654,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004977420437788341,
      "loss": 5.8288,
      "mean_token_accuracy": 0.15712782740592957,
      "num_tokens": 12405083.0,
      "step": 6120
    },
    {
      "entropy": 5.938368082046509,
      "epoch": 0.5352617320632701,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004977376381028385,
      "loss": 5.7311,
      "mean_token_accuracy": 0.17602842003107072,
      "num_tokens": 12415787.0,
      "step": 6125
    },
    {
      "entropy": 6.052881240844727,
      "epoch": 0.5356986804159748,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004977332281546181,
      "loss": 5.7894,
      "mean_token_accuracy": 0.16413515359163283,
      "num_tokens": 12425911.0,
      "step": 6130
    },
    {
      "entropy": 6.03786072731018,
      "epoch": 0.5361356287686796,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004977288139342578,
      "loss": 5.8116,
      "mean_token_accuracy": 0.16361333578824996,
      "num_tokens": 12437270.0,
      "step": 6135
    },
    {
      "entropy": 6.017053842544556,
      "epoch": 0.5365725771213843,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004977243954418422,
      "loss": 5.772,
      "mean_token_accuracy": 0.16508300602436066,
      "num_tokens": 12446778.0,
      "step": 6140
    },
    {
      "entropy": 5.9529952049255375,
      "epoch": 0.537009525474089,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004977199726774558,
      "loss": 5.7344,
      "mean_token_accuracy": 0.1696458026766777,
      "num_tokens": 12456550.0,
      "step": 6145
    },
    {
      "entropy": 6.0351649761199955,
      "epoch": 0.5374464738267937,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004977155456411837,
      "loss": 5.7378,
      "mean_token_accuracy": 0.16536538749933244,
      "num_tokens": 12465562.0,
      "step": 6150
    },
    {
      "entropy": 6.108797216415406,
      "epoch": 0.5378834221794984,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004977111143331108,
      "loss": 5.8015,
      "mean_token_accuracy": 0.1596318483352661,
      "num_tokens": 12475277.0,
      "step": 6155
    },
    {
      "entropy": 6.0047135829925535,
      "epoch": 0.538320370532203,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004977066787533219,
      "loss": 5.7292,
      "mean_token_accuracy": 0.16796960085630416,
      "num_tokens": 12485478.0,
      "step": 6160
    },
    {
      "entropy": 6.0715264797210695,
      "epoch": 0.5387573188849079,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004977022389019023,
      "loss": 5.8504,
      "mean_token_accuracy": 0.15865544825792313,
      "num_tokens": 12495714.0,
      "step": 6165
    },
    {
      "entropy": 6.102363109588623,
      "epoch": 0.5391942672376125,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004976977947789371,
      "loss": 5.7688,
      "mean_token_accuracy": 0.1704864725470543,
      "num_tokens": 12505125.0,
      "step": 6170
    },
    {
      "entropy": 6.005610132217408,
      "epoch": 0.5396312155903172,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004976933463845115,
      "loss": 5.7649,
      "mean_token_accuracy": 0.16732384115457535,
      "num_tokens": 12515051.0,
      "step": 6175
    },
    {
      "entropy": 5.99391803741455,
      "epoch": 0.5400681639430219,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004976888937187107,
      "loss": 5.7202,
      "mean_token_accuracy": 0.16938636898994447,
      "num_tokens": 12525389.0,
      "step": 6180
    },
    {
      "entropy": 6.039830923080444,
      "epoch": 0.5405051122957266,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004976844367816203,
      "loss": 5.8352,
      "mean_token_accuracy": 0.1489139124751091,
      "num_tokens": 12535454.0,
      "step": 6185
    },
    {
      "entropy": 5.988320589065552,
      "epoch": 0.5409420606484313,
      "grad_norm": 0.875,
      "learning_rate": 0.0004976799755733257,
      "loss": 5.6599,
      "mean_token_accuracy": 0.17356159836053847,
      "num_tokens": 12546092.0,
      "step": 6190
    },
    {
      "entropy": 6.021031284332276,
      "epoch": 0.5413790090011361,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004976755100939126,
      "loss": 5.7704,
      "mean_token_accuracy": 0.16135476529598236,
      "num_tokens": 12555462.0,
      "step": 6195
    },
    {
      "entropy": 6.0270583152771,
      "epoch": 0.5418159573538408,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004976710403434663,
      "loss": 5.8106,
      "mean_token_accuracy": 0.16644200831651687,
      "num_tokens": 12564506.0,
      "step": 6200
    },
    {
      "entropy": 6.019489192962647,
      "epoch": 0.5422529057065455,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000497666566322073,
      "loss": 5.776,
      "mean_token_accuracy": 0.16439228653907775,
      "num_tokens": 12574887.0,
      "step": 6205
    },
    {
      "entropy": 6.141333770751953,
      "epoch": 0.5426898540592502,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004976620880298182,
      "loss": 5.8018,
      "mean_token_accuracy": 0.16166453510522844,
      "num_tokens": 12584499.0,
      "step": 6210
    },
    {
      "entropy": 6.033131408691406,
      "epoch": 0.5431268024119549,
      "grad_norm": 1.0,
      "learning_rate": 0.0004976576054667878,
      "loss": 5.7471,
      "mean_token_accuracy": 0.16789500117301942,
      "num_tokens": 12593796.0,
      "step": 6215
    },
    {
      "entropy": 5.968105840682983,
      "epoch": 0.5435637507646596,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000497653118633068,
      "loss": 5.8051,
      "mean_token_accuracy": 0.15516676455736161,
      "num_tokens": 12603719.0,
      "step": 6220
    },
    {
      "entropy": 5.979581546783447,
      "epoch": 0.5440006991173644,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004976486275287446,
      "loss": 5.7982,
      "mean_token_accuracy": 0.16076839417219163,
      "num_tokens": 12614407.0,
      "step": 6225
    },
    {
      "entropy": 6.039636421203613,
      "epoch": 0.5444376474700691,
      "grad_norm": 0.875,
      "learning_rate": 0.0004976441321539038,
      "loss": 5.7668,
      "mean_token_accuracy": 0.17348510921001434,
      "num_tokens": 12624414.0,
      "step": 6230
    },
    {
      "entropy": 6.0514914989471436,
      "epoch": 0.5448745958227738,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004976396325086321,
      "loss": 5.7914,
      "mean_token_accuracy": 0.1653514564037323,
      "num_tokens": 12635679.0,
      "step": 6235
    },
    {
      "entropy": 6.046541833877564,
      "epoch": 0.5453115441754784,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004976351285930154,
      "loss": 5.7286,
      "mean_token_accuracy": 0.16179855465888976,
      "num_tokens": 12645253.0,
      "step": 6240
    },
    {
      "entropy": 5.910202360153198,
      "epoch": 0.5457484925281831,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004976306204071404,
      "loss": 5.6926,
      "mean_token_accuracy": 0.1789444386959076,
      "num_tokens": 12655815.0,
      "step": 6245
    },
    {
      "entropy": 5.987710189819336,
      "epoch": 0.5461854408808878,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004976261079510934,
      "loss": 5.7896,
      "mean_token_accuracy": 0.1636798232793808,
      "num_tokens": 12666173.0,
      "step": 6250
    },
    {
      "entropy": 6.016307640075683,
      "epoch": 0.5466223892335926,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000497621591224961,
      "loss": 5.731,
      "mean_token_accuracy": 0.1690751299262047,
      "num_tokens": 12676008.0,
      "step": 6255
    },
    {
      "entropy": 6.057435941696167,
      "epoch": 0.5470593375862973,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004976170702288298,
      "loss": 5.7951,
      "mean_token_accuracy": 0.16168733835220336,
      "num_tokens": 12686217.0,
      "step": 6260
    },
    {
      "entropy": 6.003119945526123,
      "epoch": 0.547496285939002,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004976125449627866,
      "loss": 5.8004,
      "mean_token_accuracy": 0.16176285594701767,
      "num_tokens": 12696809.0,
      "step": 6265
    },
    {
      "entropy": 6.08587965965271,
      "epoch": 0.5479332342917067,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004976080154269181,
      "loss": 5.8683,
      "mean_token_accuracy": 0.154628224670887,
      "num_tokens": 12708328.0,
      "step": 6270
    },
    {
      "entropy": 6.029016017913818,
      "epoch": 0.5483701826444114,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004976034816213112,
      "loss": 5.6818,
      "mean_token_accuracy": 0.16641477048397063,
      "num_tokens": 12718478.0,
      "step": 6275
    },
    {
      "entropy": 5.922139453887939,
      "epoch": 0.5488071309971161,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004975989435460528,
      "loss": 5.7728,
      "mean_token_accuracy": 0.1684291496872902,
      "num_tokens": 12728853.0,
      "step": 6280
    },
    {
      "entropy": 6.017479944229126,
      "epoch": 0.5492440793498209,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004975944012012302,
      "loss": 5.825,
      "mean_token_accuracy": 0.16401030272245407,
      "num_tokens": 12739036.0,
      "step": 6285
    },
    {
      "entropy": 6.019897413253784,
      "epoch": 0.5496810277025256,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004975898545869303,
      "loss": 5.7316,
      "mean_token_accuracy": 0.16734253764152526,
      "num_tokens": 12748908.0,
      "step": 6290
    },
    {
      "entropy": 5.968127202987671,
      "epoch": 0.5501179760552303,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004975853037032403,
      "loss": 5.7991,
      "mean_token_accuracy": 0.1596520185470581,
      "num_tokens": 12758671.0,
      "step": 6295
    },
    {
      "entropy": 6.016580581665039,
      "epoch": 0.550554924407935,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004975807485502476,
      "loss": 5.778,
      "mean_token_accuracy": 0.16463625133037568,
      "num_tokens": 12769552.0,
      "step": 6300
    },
    {
      "entropy": 6.083640289306641,
      "epoch": 0.5509918727606397,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004975761891280394,
      "loss": 5.908,
      "mean_token_accuracy": 0.1530622199177742,
      "num_tokens": 12781032.0,
      "step": 6305
    },
    {
      "entropy": 6.088249063491821,
      "epoch": 0.5514288211133443,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004975716254367033,
      "loss": 5.8265,
      "mean_token_accuracy": 0.1623299963772297,
      "num_tokens": 12790818.0,
      "step": 6310
    },
    {
      "entropy": 5.977339124679565,
      "epoch": 0.5518657694660491,
      "grad_norm": 0.953125,
      "learning_rate": 0.000497567057476327,
      "loss": 5.7079,
      "mean_token_accuracy": 0.1669737681746483,
      "num_tokens": 12799701.0,
      "step": 6315
    },
    {
      "entropy": 5.981634950637817,
      "epoch": 0.5523027178187538,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004975624852469976,
      "loss": 5.8427,
      "mean_token_accuracy": 0.16832137554883958,
      "num_tokens": 12809575.0,
      "step": 6320
    },
    {
      "entropy": 6.051725721359253,
      "epoch": 0.5527396661714585,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004975579087488033,
      "loss": 5.7485,
      "mean_token_accuracy": 0.17082757502794266,
      "num_tokens": 12820278.0,
      "step": 6325
    },
    {
      "entropy": 6.0816059589385985,
      "epoch": 0.5531766145241632,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004975533279818317,
      "loss": 5.7762,
      "mean_token_accuracy": 0.16289230436086655,
      "num_tokens": 12830774.0,
      "step": 6330
    },
    {
      "entropy": 5.980174160003662,
      "epoch": 0.5536135628768679,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004975487429461706,
      "loss": 5.7413,
      "mean_token_accuracy": 0.16966917216777802,
      "num_tokens": 12839487.0,
      "step": 6335
    },
    {
      "entropy": 5.946662950515747,
      "epoch": 0.5540505112295727,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000497544153641908,
      "loss": 5.6807,
      "mean_token_accuracy": 0.1687507912516594,
      "num_tokens": 12849614.0,
      "step": 6340
    },
    {
      "entropy": 6.020115947723388,
      "epoch": 0.5544874595822774,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004975395600691319,
      "loss": 5.806,
      "mean_token_accuracy": 0.1653425872325897,
      "num_tokens": 12859428.0,
      "step": 6345
    },
    {
      "entropy": 5.9838159561157225,
      "epoch": 0.5549244079349821,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004975349622279305,
      "loss": 5.81,
      "mean_token_accuracy": 0.16461237370967866,
      "num_tokens": 12869167.0,
      "step": 6350
    },
    {
      "entropy": 6.014318132400513,
      "epoch": 0.5553613562876868,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004975303601183919,
      "loss": 5.7457,
      "mean_token_accuracy": 0.1653507873415947,
      "num_tokens": 12878410.0,
      "step": 6355
    },
    {
      "entropy": 6.017940855026245,
      "epoch": 0.5557983046403915,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004975257537406043,
      "loss": 5.7615,
      "mean_token_accuracy": 0.16263557225465775,
      "num_tokens": 12888372.0,
      "step": 6360
    },
    {
      "entropy": 6.019524955749512,
      "epoch": 0.5562352529930962,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004975211430946563,
      "loss": 5.808,
      "mean_token_accuracy": 0.16446324735879897,
      "num_tokens": 12897958.0,
      "step": 6365
    },
    {
      "entropy": 6.00267653465271,
      "epoch": 0.556672201345801,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000497516528180636,
      "loss": 5.792,
      "mean_token_accuracy": 0.16702653765678405,
      "num_tokens": 12907890.0,
      "step": 6370
    },
    {
      "entropy": 6.021846103668213,
      "epoch": 0.5571091496985057,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004975119089986322,
      "loss": 5.7423,
      "mean_token_accuracy": 0.16592501401901244,
      "num_tokens": 12917914.0,
      "step": 6375
    },
    {
      "entropy": 6.040018033981323,
      "epoch": 0.5575460980512104,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004975072855487333,
      "loss": 5.7876,
      "mean_token_accuracy": 0.16546638309955597,
      "num_tokens": 12927090.0,
      "step": 6380
    },
    {
      "entropy": 6.006717729568481,
      "epoch": 0.557983046403915,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004975026578310282,
      "loss": 5.7675,
      "mean_token_accuracy": 0.16308177411556243,
      "num_tokens": 12937236.0,
      "step": 6385
    },
    {
      "entropy": 6.001976537704468,
      "epoch": 0.5584199947566197,
      "grad_norm": 0.875,
      "learning_rate": 0.0004974980258456055,
      "loss": 5.7191,
      "mean_token_accuracy": 0.16658366918563844,
      "num_tokens": 12947635.0,
      "step": 6390
    },
    {
      "entropy": 6.0470386981964115,
      "epoch": 0.5588569431093244,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000497493389592554,
      "loss": 5.6743,
      "mean_token_accuracy": 0.16891563683748245,
      "num_tokens": 12957236.0,
      "step": 6395
    },
    {
      "entropy": 6.027141094207764,
      "epoch": 0.5592938914620292,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004974887490719627,
      "loss": 5.8016,
      "mean_token_accuracy": 0.16280270516872405,
      "num_tokens": 12967497.0,
      "step": 6400
    },
    {
      "entropy": 5.95338077545166,
      "epoch": 0.5597308398147339,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004974841042839207,
      "loss": 5.7219,
      "mean_token_accuracy": 0.1686733514070511,
      "num_tokens": 12977467.0,
      "step": 6405
    },
    {
      "entropy": 6.025985336303711,
      "epoch": 0.5601677881674386,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000497479455228517,
      "loss": 5.788,
      "mean_token_accuracy": 0.16235905438661574,
      "num_tokens": 12987867.0,
      "step": 6410
    },
    {
      "entropy": 6.037254858016968,
      "epoch": 0.5606047365201433,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004974748019058407,
      "loss": 5.7889,
      "mean_token_accuracy": 0.15933752059936523,
      "num_tokens": 12998081.0,
      "step": 6415
    },
    {
      "entropy": 5.971248435974121,
      "epoch": 0.561041684872848,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004974701443159812,
      "loss": 5.7786,
      "mean_token_accuracy": 0.1640427976846695,
      "num_tokens": 13008047.0,
      "step": 6420
    },
    {
      "entropy": 6.030923986434937,
      "epoch": 0.5614786332255527,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004974654824590277,
      "loss": 5.7432,
      "mean_token_accuracy": 0.16663393825292588,
      "num_tokens": 13018048.0,
      "step": 6425
    },
    {
      "entropy": 5.987953567504883,
      "epoch": 0.5619155815782575,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004974608163350697,
      "loss": 5.7499,
      "mean_token_accuracy": 0.16480678021907808,
      "num_tokens": 13028156.0,
      "step": 6430
    },
    {
      "entropy": 6.047450256347656,
      "epoch": 0.5623525299309622,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004974561459441967,
      "loss": 5.7745,
      "mean_token_accuracy": 0.16245208978652953,
      "num_tokens": 13038191.0,
      "step": 6435
    },
    {
      "entropy": 5.97278470993042,
      "epoch": 0.5627894782836669,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004974514712864981,
      "loss": 5.76,
      "mean_token_accuracy": 0.1672798439860344,
      "num_tokens": 13049085.0,
      "step": 6440
    },
    {
      "entropy": 5.996664142608642,
      "epoch": 0.5632264266363716,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004974467923620639,
      "loss": 5.7837,
      "mean_token_accuracy": 0.1627726063132286,
      "num_tokens": 13060010.0,
      "step": 6445
    },
    {
      "entropy": 6.094581937789917,
      "epoch": 0.5636633749890763,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004974421091709836,
      "loss": 5.7758,
      "mean_token_accuracy": 0.1608213871717453,
      "num_tokens": 13069768.0,
      "step": 6450
    },
    {
      "entropy": 5.993890810012817,
      "epoch": 0.564100323341781,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004974374217133471,
      "loss": 5.7769,
      "mean_token_accuracy": 0.16323172599077224,
      "num_tokens": 13079842.0,
      "step": 6455
    },
    {
      "entropy": 6.026523542404175,
      "epoch": 0.5645372716944858,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004974327299892443,
      "loss": 5.7005,
      "mean_token_accuracy": 0.16720663458108903,
      "num_tokens": 13089235.0,
      "step": 6460
    },
    {
      "entropy": 6.010472583770752,
      "epoch": 0.5649742200471904,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004974280339987651,
      "loss": 5.7701,
      "mean_token_accuracy": 0.15765927135944366,
      "num_tokens": 13098364.0,
      "step": 6465
    },
    {
      "entropy": 6.048176717758179,
      "epoch": 0.5654111683998951,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004974233337419997,
      "loss": 5.9009,
      "mean_token_accuracy": 0.15627993047237396,
      "num_tokens": 13108876.0,
      "step": 6470
    },
    {
      "entropy": 6.096752119064331,
      "epoch": 0.5658481167525998,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004974186292190383,
      "loss": 5.8764,
      "mean_token_accuracy": 0.16218368113040924,
      "num_tokens": 13119591.0,
      "step": 6475
    },
    {
      "entropy": 6.0366997718811035,
      "epoch": 0.5662850651053045,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000497413920429971,
      "loss": 5.7708,
      "mean_token_accuracy": 0.17020626217126847,
      "num_tokens": 13129796.0,
      "step": 6480
    },
    {
      "entropy": 5.998263025283814,
      "epoch": 0.5667220134580092,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004974092073748882,
      "loss": 5.7174,
      "mean_token_accuracy": 0.17187049686908723,
      "num_tokens": 13139302.0,
      "step": 6485
    },
    {
      "entropy": 5.9880951881408695,
      "epoch": 0.567158961810714,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004974044900538802,
      "loss": 5.7877,
      "mean_token_accuracy": 0.1646083414554596,
      "num_tokens": 13148983.0,
      "step": 6490
    },
    {
      "entropy": 6.032616806030274,
      "epoch": 0.5675959101634187,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004973997684670375,
      "loss": 5.7633,
      "mean_token_accuracy": 0.16235041618347168,
      "num_tokens": 13159016.0,
      "step": 6495
    },
    {
      "entropy": 6.028706121444702,
      "epoch": 0.5680328585161234,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004973950426144508,
      "loss": 5.8021,
      "mean_token_accuracy": 0.15465459674596788,
      "num_tokens": 13169705.0,
      "step": 6500
    },
    {
      "entropy": 6.060849809646607,
      "epoch": 0.5684698068688281,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004973903124962106,
      "loss": 5.7879,
      "mean_token_accuracy": 0.15404265373945236,
      "num_tokens": 13181372.0,
      "step": 6505
    },
    {
      "entropy": 6.021795177459717,
      "epoch": 0.5689067552215328,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004973855781124077,
      "loss": 5.8341,
      "mean_token_accuracy": 0.1632041722536087,
      "num_tokens": 13192581.0,
      "step": 6510
    },
    {
      "entropy": 6.0210487842559814,
      "epoch": 0.5693437035742375,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004973808394631329,
      "loss": 5.8481,
      "mean_token_accuracy": 0.16023556515574455,
      "num_tokens": 13203269.0,
      "step": 6515
    },
    {
      "entropy": 6.017901849746704,
      "epoch": 0.5697806519269423,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000497376096548477,
      "loss": 5.7178,
      "mean_token_accuracy": 0.17219983637332917,
      "num_tokens": 13213183.0,
      "step": 6520
    },
    {
      "entropy": 6.032643032073975,
      "epoch": 0.570217600279647,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004973713493685311,
      "loss": 5.8149,
      "mean_token_accuracy": 0.16332971751689912,
      "num_tokens": 13223196.0,
      "step": 6525
    },
    {
      "entropy": 6.023701143264771,
      "epoch": 0.5706545486323517,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004973665979233861,
      "loss": 5.7303,
      "mean_token_accuracy": 0.16850854307413102,
      "num_tokens": 13233915.0,
      "step": 6530
    },
    {
      "entropy": 6.000444078445435,
      "epoch": 0.5710914969850563,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004973618422131334,
      "loss": 5.7641,
      "mean_token_accuracy": 0.16293680369853974,
      "num_tokens": 13245108.0,
      "step": 6535
    },
    {
      "entropy": 5.978904819488525,
      "epoch": 0.571528445337761,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000497357082237864,
      "loss": 5.7876,
      "mean_token_accuracy": 0.16653110235929489,
      "num_tokens": 13255508.0,
      "step": 6540
    },
    {
      "entropy": 6.0873438835144045,
      "epoch": 0.5719653936904658,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004973523179976692,
      "loss": 5.793,
      "mean_token_accuracy": 0.1653340458869934,
      "num_tokens": 13265118.0,
      "step": 6545
    },
    {
      "entropy": 6.046351051330566,
      "epoch": 0.5724023420431705,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004973475494926404,
      "loss": 5.7061,
      "mean_token_accuracy": 0.1634884238243103,
      "num_tokens": 13275947.0,
      "step": 6550
    },
    {
      "entropy": 5.989783716201782,
      "epoch": 0.5728392903958752,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004973427767228692,
      "loss": 5.7294,
      "mean_token_accuracy": 0.1725679337978363,
      "num_tokens": 13284662.0,
      "step": 6555
    },
    {
      "entropy": 6.008610582351684,
      "epoch": 0.5732762387485799,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000497337999688447,
      "loss": 5.7986,
      "mean_token_accuracy": 0.16211242079734803,
      "num_tokens": 13295627.0,
      "step": 6560
    },
    {
      "entropy": 6.040422105789185,
      "epoch": 0.5737131871012846,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004973332183894653,
      "loss": 5.6971,
      "mean_token_accuracy": 0.17034139931201936,
      "num_tokens": 13305205.0,
      "step": 6565
    },
    {
      "entropy": 5.958159112930298,
      "epoch": 0.5741501354539893,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004973284328260162,
      "loss": 5.7344,
      "mean_token_accuracy": 0.1664531111717224,
      "num_tokens": 13315464.0,
      "step": 6570
    },
    {
      "entropy": 5.932192039489746,
      "epoch": 0.5745870838066941,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004973236429981912,
      "loss": 5.6577,
      "mean_token_accuracy": 0.17045878171920775,
      "num_tokens": 13325468.0,
      "step": 6575
    },
    {
      "entropy": 6.00683069229126,
      "epoch": 0.5750240321593988,
      "grad_norm": 1.0,
      "learning_rate": 0.0004973188489060821,
      "loss": 5.7744,
      "mean_token_accuracy": 0.16734251528978347,
      "num_tokens": 13334719.0,
      "step": 6580
    },
    {
      "entropy": 5.985804796218872,
      "epoch": 0.5754609805121035,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004973140505497812,
      "loss": 5.7194,
      "mean_token_accuracy": 0.16590756624937059,
      "num_tokens": 13344545.0,
      "step": 6585
    },
    {
      "entropy": 6.034628629684448,
      "epoch": 0.5758979288648082,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004973092479293802,
      "loss": 5.8077,
      "mean_token_accuracy": 0.16228435784578324,
      "num_tokens": 13355212.0,
      "step": 6590
    },
    {
      "entropy": 5.993490314483642,
      "epoch": 0.5763348772175129,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004973044410449713,
      "loss": 5.7088,
      "mean_token_accuracy": 0.16826409548521043,
      "num_tokens": 13364869.0,
      "step": 6595
    },
    {
      "entropy": 5.986246824264526,
      "epoch": 0.5767718255702176,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004972996298966468,
      "loss": 5.7534,
      "mean_token_accuracy": 0.1636641129851341,
      "num_tokens": 13376511.0,
      "step": 6600
    },
    {
      "entropy": 6.011943817138672,
      "epoch": 0.5772087739229224,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000497294814484499,
      "loss": 5.78,
      "mean_token_accuracy": 0.16534585058689116,
      "num_tokens": 13385470.0,
      "step": 6605
    },
    {
      "entropy": 6.029364728927613,
      "epoch": 0.577645722275627,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00049728999480862,
      "loss": 5.8449,
      "mean_token_accuracy": 0.16150766462087632,
      "num_tokens": 13395328.0,
      "step": 6610
    },
    {
      "entropy": 6.0173345565795895,
      "epoch": 0.5780826706283317,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004972851708691025,
      "loss": 5.7402,
      "mean_token_accuracy": 0.17220347374677658,
      "num_tokens": 13405319.0,
      "step": 6615
    },
    {
      "entropy": 5.919049072265625,
      "epoch": 0.5785196189810364,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004972803426660389,
      "loss": 5.7413,
      "mean_token_accuracy": 0.16287230849266052,
      "num_tokens": 13416119.0,
      "step": 6620
    },
    {
      "entropy": 6.01291675567627,
      "epoch": 0.5789565673337411,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004972755101995219,
      "loss": 5.7354,
      "mean_token_accuracy": 0.16355575770139694,
      "num_tokens": 13425824.0,
      "step": 6625
    },
    {
      "entropy": 5.973971843719482,
      "epoch": 0.5793935156864458,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004972706734696442,
      "loss": 5.6503,
      "mean_token_accuracy": 0.1747452959418297,
      "num_tokens": 13435506.0,
      "step": 6630
    },
    {
      "entropy": 5.949373197555542,
      "epoch": 0.5798304640391506,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004972658324764984,
      "loss": 5.6919,
      "mean_token_accuracy": 0.1741135373711586,
      "num_tokens": 13445313.0,
      "step": 6635
    },
    {
      "entropy": 5.952116394042969,
      "epoch": 0.5802674123918553,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004972609872201775,
      "loss": 5.6519,
      "mean_token_accuracy": 0.17122422009706498,
      "num_tokens": 13455312.0,
      "step": 6640
    },
    {
      "entropy": 6.0242777347564695,
      "epoch": 0.58070436074456,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004972561377007744,
      "loss": 5.7644,
      "mean_token_accuracy": 0.16550102084875107,
      "num_tokens": 13465721.0,
      "step": 6645
    },
    {
      "entropy": 6.040964269638062,
      "epoch": 0.5811413090972647,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004972512839183822,
      "loss": 5.7989,
      "mean_token_accuracy": 0.16431246548891068,
      "num_tokens": 13477185.0,
      "step": 6650
    },
    {
      "entropy": 5.933858585357666,
      "epoch": 0.5815782574499694,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004972464258730939,
      "loss": 5.7197,
      "mean_token_accuracy": 0.17679304927587508,
      "num_tokens": 13487846.0,
      "step": 6655
    },
    {
      "entropy": 5.950643730163574,
      "epoch": 0.5820152058026741,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004972415635650027,
      "loss": 5.7915,
      "mean_token_accuracy": 0.1747171014547348,
      "num_tokens": 13498175.0,
      "step": 6660
    },
    {
      "entropy": 6.070603322982788,
      "epoch": 0.5824521541553789,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004972366969942017,
      "loss": 5.7203,
      "mean_token_accuracy": 0.16188783943653107,
      "num_tokens": 13506864.0,
      "step": 6665
    },
    {
      "entropy": 5.98835825920105,
      "epoch": 0.5828891025080836,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004972318261607845,
      "loss": 5.7427,
      "mean_token_accuracy": 0.16818013042211533,
      "num_tokens": 13516208.0,
      "step": 6670
    },
    {
      "entropy": 6.019196271896362,
      "epoch": 0.5833260508607883,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004972269510648444,
      "loss": 5.8097,
      "mean_token_accuracy": 0.1604367598891258,
      "num_tokens": 13526902.0,
      "step": 6675
    },
    {
      "entropy": 6.014893054962158,
      "epoch": 0.583762999213493,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000497222071706475,
      "loss": 5.7147,
      "mean_token_accuracy": 0.17692699879407883,
      "num_tokens": 13537001.0,
      "step": 6680
    },
    {
      "entropy": 5.924656534194947,
      "epoch": 0.5841999475661976,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004972171880857698,
      "loss": 5.6992,
      "mean_token_accuracy": 0.1721874475479126,
      "num_tokens": 13546971.0,
      "step": 6685
    },
    {
      "entropy": 5.916549968719482,
      "epoch": 0.5846368959189023,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004972123002028226,
      "loss": 5.6763,
      "mean_token_accuracy": 0.17543130367994308,
      "num_tokens": 13557202.0,
      "step": 6690
    },
    {
      "entropy": 5.917145872116089,
      "epoch": 0.5850738442716071,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004972074080577269,
      "loss": 5.6819,
      "mean_token_accuracy": 0.17548966258764268,
      "num_tokens": 13567291.0,
      "step": 6695
    },
    {
      "entropy": 6.012505102157593,
      "epoch": 0.5855107926243118,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004972025116505767,
      "loss": 5.7497,
      "mean_token_accuracy": 0.17616625428199767,
      "num_tokens": 13576864.0,
      "step": 6700
    },
    {
      "entropy": 5.99484281539917,
      "epoch": 0.5859477409770165,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004971976109814658,
      "loss": 5.675,
      "mean_token_accuracy": 0.17459282279014587,
      "num_tokens": 13586728.0,
      "step": 6705
    },
    {
      "entropy": 5.936760187149048,
      "epoch": 0.5863846893297212,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004971927060504885,
      "loss": 5.7382,
      "mean_token_accuracy": 0.17031691819429398,
      "num_tokens": 13596875.0,
      "step": 6710
    },
    {
      "entropy": 6.0446747779846195,
      "epoch": 0.5868216376824259,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004971877968577386,
      "loss": 5.7556,
      "mean_token_accuracy": 0.15463765263557433,
      "num_tokens": 13606483.0,
      "step": 6715
    },
    {
      "entropy": 6.028912448883057,
      "epoch": 0.5872585860351307,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004971828834033102,
      "loss": 5.8013,
      "mean_token_accuracy": 0.17006202936172485,
      "num_tokens": 13616780.0,
      "step": 6720
    },
    {
      "entropy": 5.974723291397095,
      "epoch": 0.5876955343878354,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004971779656872978,
      "loss": 5.7591,
      "mean_token_accuracy": 0.16448667496442795,
      "num_tokens": 13626418.0,
      "step": 6725
    },
    {
      "entropy": 6.070558834075928,
      "epoch": 0.5881324827405401,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004971730437097955,
      "loss": 5.8179,
      "mean_token_accuracy": 0.1648862823843956,
      "num_tokens": 13636956.0,
      "step": 6730
    },
    {
      "entropy": 6.010634517669677,
      "epoch": 0.5885694310932448,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000497168117470898,
      "loss": 5.7454,
      "mean_token_accuracy": 0.16742426604032518,
      "num_tokens": 13647458.0,
      "step": 6735
    },
    {
      "entropy": 5.918993425369263,
      "epoch": 0.5890063794459495,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004971631869706995,
      "loss": 5.6982,
      "mean_token_accuracy": 0.1731320098042488,
      "num_tokens": 13657655.0,
      "step": 6740
    },
    {
      "entropy": 6.0385987758636475,
      "epoch": 0.5894433277986542,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004971582522092946,
      "loss": 5.778,
      "mean_token_accuracy": 0.15799714177846907,
      "num_tokens": 13668472.0,
      "step": 6745
    },
    {
      "entropy": 5.993940210342407,
      "epoch": 0.589880276151359,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000497153313186778,
      "loss": 5.6529,
      "mean_token_accuracy": 0.17010727673768997,
      "num_tokens": 13677319.0,
      "step": 6750
    },
    {
      "entropy": 5.921839809417724,
      "epoch": 0.5903172245040637,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004971483699032445,
      "loss": 5.6889,
      "mean_token_accuracy": 0.1712235063314438,
      "num_tokens": 13688037.0,
      "step": 6755
    },
    {
      "entropy": 6.050872182846069,
      "epoch": 0.5907541728567683,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004971434223587888,
      "loss": 5.8146,
      "mean_token_accuracy": 0.16486726552248002,
      "num_tokens": 13698441.0,
      "step": 6760
    },
    {
      "entropy": 6.0213282108306885,
      "epoch": 0.591191121209473,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000497138470553506,
      "loss": 5.7335,
      "mean_token_accuracy": 0.17158997505903245,
      "num_tokens": 13708140.0,
      "step": 6765
    },
    {
      "entropy": 6.00905213356018,
      "epoch": 0.5916280695621777,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004971335144874909,
      "loss": 5.7303,
      "mean_token_accuracy": 0.17032813280820847,
      "num_tokens": 13718069.0,
      "step": 6770
    },
    {
      "entropy": 5.950638628005981,
      "epoch": 0.5920650179148824,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004971285541608385,
      "loss": 5.7022,
      "mean_token_accuracy": 0.18172554522752762,
      "num_tokens": 13727932.0,
      "step": 6775
    },
    {
      "entropy": 6.071123218536377,
      "epoch": 0.5925019662675872,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004971235895736441,
      "loss": 5.7852,
      "mean_token_accuracy": 0.1578560031950474,
      "num_tokens": 13738239.0,
      "step": 6780
    },
    {
      "entropy": 6.022912979125977,
      "epoch": 0.5929389146202919,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004971186207260029,
      "loss": 5.7314,
      "mean_token_accuracy": 0.1627219721674919,
      "num_tokens": 13748622.0,
      "step": 6785
    },
    {
      "entropy": 5.86298098564148,
      "epoch": 0.5933758629729966,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00049711364761801,
      "loss": 5.6662,
      "mean_token_accuracy": 0.17138780355453492,
      "num_tokens": 13759221.0,
      "step": 6790
    },
    {
      "entropy": 5.993860387802124,
      "epoch": 0.5938128113257013,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000497108670249761,
      "loss": 5.6741,
      "mean_token_accuracy": 0.16800866276025772,
      "num_tokens": 13768343.0,
      "step": 6795
    },
    {
      "entropy": 5.968444156646728,
      "epoch": 0.594249759678406,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004971036886213515,
      "loss": 5.6618,
      "mean_token_accuracy": 0.1711151823401451,
      "num_tokens": 13778225.0,
      "step": 6800
    },
    {
      "entropy": 5.989865016937256,
      "epoch": 0.5946867080311107,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004970987027328767,
      "loss": 5.7574,
      "mean_token_accuracy": 0.1645011216402054,
      "num_tokens": 13788633.0,
      "step": 6805
    },
    {
      "entropy": 6.018437385559082,
      "epoch": 0.5951236563838155,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004970937125844325,
      "loss": 5.6685,
      "mean_token_accuracy": 0.17222141996026039,
      "num_tokens": 13798172.0,
      "step": 6810
    },
    {
      "entropy": 5.882524394989014,
      "epoch": 0.5955606047365202,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004970887181761144,
      "loss": 5.6492,
      "mean_token_accuracy": 0.1793105944991112,
      "num_tokens": 13807284.0,
      "step": 6815
    },
    {
      "entropy": 6.011940145492554,
      "epoch": 0.5959975530892249,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004970837195080184,
      "loss": 5.7324,
      "mean_token_accuracy": 0.17324329912662506,
      "num_tokens": 13816807.0,
      "step": 6820
    },
    {
      "entropy": 6.0003620147705075,
      "epoch": 0.5964345014419296,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004970787165802403,
      "loss": 5.6857,
      "mean_token_accuracy": 0.1669215276837349,
      "num_tokens": 13827635.0,
      "step": 6825
    },
    {
      "entropy": 5.969858884811401,
      "epoch": 0.5968714497946342,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000497073709392876,
      "loss": 5.7672,
      "mean_token_accuracy": 0.166155669093132,
      "num_tokens": 13838060.0,
      "step": 6830
    },
    {
      "entropy": 6.004700422286987,
      "epoch": 0.5973083981473389,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004970686979460216,
      "loss": 5.7486,
      "mean_token_accuracy": 0.1617456153035164,
      "num_tokens": 13848669.0,
      "step": 6835
    },
    {
      "entropy": 5.972333765029907,
      "epoch": 0.5977453465000437,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004970636822397733,
      "loss": 5.7145,
      "mean_token_accuracy": 0.1715695306658745,
      "num_tokens": 13857949.0,
      "step": 6840
    },
    {
      "entropy": 5.976265954971313,
      "epoch": 0.5981822948527484,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004970586622742272,
      "loss": 5.6997,
      "mean_token_accuracy": 0.17573309689760208,
      "num_tokens": 13867726.0,
      "step": 6845
    },
    {
      "entropy": 6.067394828796386,
      "epoch": 0.5986192432054531,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004970536380494795,
      "loss": 5.7803,
      "mean_token_accuracy": 0.16161804348230363,
      "num_tokens": 13878072.0,
      "step": 6850
    },
    {
      "entropy": 5.8850842952728275,
      "epoch": 0.5990561915581578,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004970486095656268,
      "loss": 5.6456,
      "mean_token_accuracy": 0.1785709634423256,
      "num_tokens": 13887771.0,
      "step": 6855
    },
    {
      "entropy": 6.007570028305054,
      "epoch": 0.5994931399108625,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004970435768227652,
      "loss": 5.6863,
      "mean_token_accuracy": 0.17100532501935958,
      "num_tokens": 13897867.0,
      "step": 6860
    },
    {
      "entropy": 6.00260329246521,
      "epoch": 0.5999300882635672,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004970385398209916,
      "loss": 5.7547,
      "mean_token_accuracy": 0.16340159624814987,
      "num_tokens": 13909171.0,
      "step": 6865
    },
    {
      "entropy": 5.984949827194214,
      "epoch": 0.600367036616272,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004970334985604024,
      "loss": 5.7453,
      "mean_token_accuracy": 0.16914733052253722,
      "num_tokens": 13920374.0,
      "step": 6870
    },
    {
      "entropy": 5.9857048988342285,
      "epoch": 0.6008039849689767,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004970284530410945,
      "loss": 5.7588,
      "mean_token_accuracy": 0.16747318506240844,
      "num_tokens": 13930150.0,
      "step": 6875
    },
    {
      "entropy": 6.0073535442352295,
      "epoch": 0.6012409333216814,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004970234032631644,
      "loss": 5.7082,
      "mean_token_accuracy": 0.1711632117629051,
      "num_tokens": 13940553.0,
      "step": 6880
    },
    {
      "entropy": 5.9754478454589846,
      "epoch": 0.6016778816743861,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004970183492267091,
      "loss": 5.805,
      "mean_token_accuracy": 0.1727660581469536,
      "num_tokens": 13951086.0,
      "step": 6885
    },
    {
      "entropy": 5.961571931838989,
      "epoch": 0.6021148300270908,
      "grad_norm": 1.0,
      "learning_rate": 0.0004970132909318254,
      "loss": 5.7207,
      "mean_token_accuracy": 0.16675912588834763,
      "num_tokens": 13960684.0,
      "step": 6890
    },
    {
      "entropy": 5.857917404174804,
      "epoch": 0.6025517783797955,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004970082283786107,
      "loss": 5.5578,
      "mean_token_accuracy": 0.19067661017179488,
      "num_tokens": 13971409.0,
      "step": 6895
    },
    {
      "entropy": 6.0183985233306885,
      "epoch": 0.6029887267325003,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004970031615671617,
      "loss": 5.7572,
      "mean_token_accuracy": 0.1718629330396652,
      "num_tokens": 13981185.0,
      "step": 6900
    },
    {
      "entropy": 5.964128637313843,
      "epoch": 0.603425675085205,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004969980904975758,
      "loss": 5.7576,
      "mean_token_accuracy": 0.17039245516061782,
      "num_tokens": 13991670.0,
      "step": 6905
    },
    {
      "entropy": 5.9098163604736325,
      "epoch": 0.6038626234379096,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004969930151699501,
      "loss": 5.6952,
      "mean_token_accuracy": 0.17063284963369368,
      "num_tokens": 14002343.0,
      "step": 6910
    },
    {
      "entropy": 5.9980532169342045,
      "epoch": 0.6042995717906143,
      "grad_norm": 0.859375,
      "learning_rate": 0.000496987935584382,
      "loss": 5.7721,
      "mean_token_accuracy": 0.16801706999540328,
      "num_tokens": 14013373.0,
      "step": 6915
    },
    {
      "entropy": 6.0484741687774655,
      "epoch": 0.604736520143319,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004969828517409693,
      "loss": 5.68,
      "mean_token_accuracy": 0.1707855999469757,
      "num_tokens": 14023233.0,
      "step": 6920
    },
    {
      "entropy": 5.916602516174317,
      "epoch": 0.6051734684960238,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000496977763639809,
      "loss": 5.7453,
      "mean_token_accuracy": 0.16715220361948013,
      "num_tokens": 14033054.0,
      "step": 6925
    },
    {
      "entropy": 5.978852033615112,
      "epoch": 0.6056104168487285,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004969726712809988,
      "loss": 5.6656,
      "mean_token_accuracy": 0.16202201098203659,
      "num_tokens": 14042039.0,
      "step": 6930
    },
    {
      "entropy": 6.060223627090454,
      "epoch": 0.6060473652014332,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004969675746646364,
      "loss": 5.8214,
      "mean_token_accuracy": 0.16090283840894698,
      "num_tokens": 14051957.0,
      "step": 6935
    },
    {
      "entropy": 6.020773935317993,
      "epoch": 0.6064843135541379,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004969624737908197,
      "loss": 5.7052,
      "mean_token_accuracy": 0.16751313656568528,
      "num_tokens": 14062438.0,
      "step": 6940
    },
    {
      "entropy": 6.024255847930908,
      "epoch": 0.6069212619068426,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004969573686596464,
      "loss": 5.717,
      "mean_token_accuracy": 0.1610104188323021,
      "num_tokens": 14073196.0,
      "step": 6945
    },
    {
      "entropy": 5.965788459777832,
      "epoch": 0.6073582102595473,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004969522592712146,
      "loss": 5.7141,
      "mean_token_accuracy": 0.16826278418302537,
      "num_tokens": 14083462.0,
      "step": 6950
    },
    {
      "entropy": 5.944077491760254,
      "epoch": 0.6077951586122521,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004969471456256221,
      "loss": 5.7771,
      "mean_token_accuracy": 0.16373893171548842,
      "num_tokens": 14094194.0,
      "step": 6955
    },
    {
      "entropy": 6.031118488311767,
      "epoch": 0.6082321069649568,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000496942027722967,
      "loss": 5.7164,
      "mean_token_accuracy": 0.1701880842447281,
      "num_tokens": 14104553.0,
      "step": 6960
    },
    {
      "entropy": 6.02759370803833,
      "epoch": 0.6086690553176615,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004969369055633476,
      "loss": 5.661,
      "mean_token_accuracy": 0.1713284134864807,
      "num_tokens": 14114140.0,
      "step": 6965
    },
    {
      "entropy": 5.967182445526123,
      "epoch": 0.6091060036703662,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000496931779146862,
      "loss": 5.8016,
      "mean_token_accuracy": 0.16551539301872253,
      "num_tokens": 14124723.0,
      "step": 6970
    },
    {
      "entropy": 5.97725477218628,
      "epoch": 0.6095429520230708,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004969266484736085,
      "loss": 5.6536,
      "mean_token_accuracy": 0.17271650284528733,
      "num_tokens": 14134030.0,
      "step": 6975
    },
    {
      "entropy": 5.949508476257324,
      "epoch": 0.6099799003757755,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004969215135436858,
      "loss": 5.6236,
      "mean_token_accuracy": 0.17391377240419387,
      "num_tokens": 14144691.0,
      "step": 6980
    },
    {
      "entropy": 5.963813638687133,
      "epoch": 0.6104168487284803,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004969163743571921,
      "loss": 5.7908,
      "mean_token_accuracy": 0.167827071249485,
      "num_tokens": 14155114.0,
      "step": 6985
    },
    {
      "entropy": 6.012882041931152,
      "epoch": 0.610853797081185,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000496911230914226,
      "loss": 5.802,
      "mean_token_accuracy": 0.15911106914281845,
      "num_tokens": 14166097.0,
      "step": 6990
    },
    {
      "entropy": 5.985409212112427,
      "epoch": 0.6112907454338897,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004969060832148862,
      "loss": 5.6586,
      "mean_token_accuracy": 0.1634743556380272,
      "num_tokens": 14176538.0,
      "step": 6995
    },
    {
      "entropy": 5.978002119064331,
      "epoch": 0.6117276937865944,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004969009312592715,
      "loss": 5.6815,
      "mean_token_accuracy": 0.16739267259836196,
      "num_tokens": 14187912.0,
      "step": 7000
    },
    {
      "entropy": 5.808555603027344,
      "epoch": 0.6121646421392991,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004968957750474806,
      "loss": 5.5699,
      "mean_token_accuracy": 0.18329621702432633,
      "num_tokens": 14198446.0,
      "step": 7005
    },
    {
      "entropy": 6.046875381469727,
      "epoch": 0.6126015904920038,
      "grad_norm": 1.0,
      "learning_rate": 0.0004968906145796126,
      "loss": 5.8062,
      "mean_token_accuracy": 0.1677398756146431,
      "num_tokens": 14207723.0,
      "step": 7010
    },
    {
      "entropy": 6.0319489479064945,
      "epoch": 0.6130385388447086,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004968854498557662,
      "loss": 5.6793,
      "mean_token_accuracy": 0.16880306154489516,
      "num_tokens": 14217590.0,
      "step": 7015
    },
    {
      "entropy": 5.9797279834747314,
      "epoch": 0.6134754871974133,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004968802808760407,
      "loss": 5.763,
      "mean_token_accuracy": 0.16556700319051743,
      "num_tokens": 14226629.0,
      "step": 7020
    },
    {
      "entropy": 6.016149425506592,
      "epoch": 0.613912435550118,
      "grad_norm": 1.0,
      "learning_rate": 0.0004968751076405351,
      "loss": 5.8257,
      "mean_token_accuracy": 0.15700202584266662,
      "num_tokens": 14235782.0,
      "step": 7025
    },
    {
      "entropy": 5.9611303329467775,
      "epoch": 0.6143493839028227,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004968699301493488,
      "loss": 5.7045,
      "mean_token_accuracy": 0.17367437481880188,
      "num_tokens": 14246687.0,
      "step": 7030
    },
    {
      "entropy": 6.020163822174072,
      "epoch": 0.6147863322555274,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004968647484025808,
      "loss": 5.7259,
      "mean_token_accuracy": 0.1696771651506424,
      "num_tokens": 14255961.0,
      "step": 7035
    },
    {
      "entropy": 5.984710216522217,
      "epoch": 0.6152232806082321,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004968595624003307,
      "loss": 5.7216,
      "mean_token_accuracy": 0.17268447428941727,
      "num_tokens": 14266024.0,
      "step": 7040
    },
    {
      "entropy": 5.9635638236999515,
      "epoch": 0.6156602289609369,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000496854372142698,
      "loss": 5.7495,
      "mean_token_accuracy": 0.16340765804052354,
      "num_tokens": 14276590.0,
      "step": 7045
    },
    {
      "entropy": 5.983493900299072,
      "epoch": 0.6160971773136416,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004968491776297821,
      "loss": 5.773,
      "mean_token_accuracy": 0.16818772554397582,
      "num_tokens": 14285992.0,
      "step": 7050
    },
    {
      "entropy": 5.954173469543457,
      "epoch": 0.6165341256663462,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004968439788616829,
      "loss": 5.6774,
      "mean_token_accuracy": 0.1721647411584854,
      "num_tokens": 14295843.0,
      "step": 7055
    },
    {
      "entropy": 5.957034683227539,
      "epoch": 0.6169710740190509,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004968387758384999,
      "loss": 5.7017,
      "mean_token_accuracy": 0.1720426008105278,
      "num_tokens": 14305643.0,
      "step": 7060
    },
    {
      "entropy": 5.983202171325684,
      "epoch": 0.6174080223717556,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004968335685603328,
      "loss": 5.6917,
      "mean_token_accuracy": 0.1712380602955818,
      "num_tokens": 14315907.0,
      "step": 7065
    },
    {
      "entropy": 5.938024091720581,
      "epoch": 0.6178449707244603,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004968283570272817,
      "loss": 5.758,
      "mean_token_accuracy": 0.15925931185483932,
      "num_tokens": 14325893.0,
      "step": 7070
    },
    {
      "entropy": 5.982112646102905,
      "epoch": 0.6182819190771651,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004968231412394464,
      "loss": 5.7551,
      "mean_token_accuracy": 0.16040001809597015,
      "num_tokens": 14337510.0,
      "step": 7075
    },
    {
      "entropy": 6.023133087158203,
      "epoch": 0.6187188674298698,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004968179211969272,
      "loss": 5.6846,
      "mean_token_accuracy": 0.1758490562438965,
      "num_tokens": 14346373.0,
      "step": 7080
    },
    {
      "entropy": 5.864609003067017,
      "epoch": 0.6191558157825745,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004968126968998239,
      "loss": 5.5653,
      "mean_token_accuracy": 0.19222135245800018,
      "num_tokens": 14356303.0,
      "step": 7085
    },
    {
      "entropy": 5.955215978622436,
      "epoch": 0.6195927641352792,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004968074683482369,
      "loss": 5.7254,
      "mean_token_accuracy": 0.1715029686689377,
      "num_tokens": 14366171.0,
      "step": 7090
    },
    {
      "entropy": 6.020372629165649,
      "epoch": 0.6200297124879839,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004968022355422663,
      "loss": 5.8169,
      "mean_token_accuracy": 0.16561906188726425,
      "num_tokens": 14376133.0,
      "step": 7095
    },
    {
      "entropy": 6.021739482879639,
      "epoch": 0.6204666608406886,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004967969984820128,
      "loss": 5.7082,
      "mean_token_accuracy": 0.16554543226957322,
      "num_tokens": 14385670.0,
      "step": 7100
    },
    {
      "entropy": 5.956287050247193,
      "epoch": 0.6209036091933934,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004967917571675765,
      "loss": 5.71,
      "mean_token_accuracy": 0.1719312623143196,
      "num_tokens": 14396002.0,
      "step": 7105
    },
    {
      "entropy": 6.006824254989624,
      "epoch": 0.6213405575460981,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004967865115990582,
      "loss": 5.8128,
      "mean_token_accuracy": 0.16002215892076493,
      "num_tokens": 14406718.0,
      "step": 7110
    },
    {
      "entropy": 6.047213792800903,
      "epoch": 0.6217775058988028,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004967812617765583,
      "loss": 5.7189,
      "mean_token_accuracy": 0.1653324469923973,
      "num_tokens": 14417027.0,
      "step": 7115
    },
    {
      "entropy": 5.9284779071807865,
      "epoch": 0.6222144542515075,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004967760077001775,
      "loss": 5.6998,
      "mean_token_accuracy": 0.17354241609573365,
      "num_tokens": 14426885.0,
      "step": 7120
    },
    {
      "entropy": 5.932647180557251,
      "epoch": 0.6226514026042121,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004967707493700169,
      "loss": 5.6787,
      "mean_token_accuracy": 0.16786099672317506,
      "num_tokens": 14437743.0,
      "step": 7125
    },
    {
      "entropy": 5.908795928955078,
      "epoch": 0.623088350956917,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004967654867861769,
      "loss": 5.6956,
      "mean_token_accuracy": 0.1769107073545456,
      "num_tokens": 14448808.0,
      "step": 7130
    },
    {
      "entropy": 5.948978567123413,
      "epoch": 0.6235252993096216,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004967602199487588,
      "loss": 5.7392,
      "mean_token_accuracy": 0.17278732657432555,
      "num_tokens": 14458560.0,
      "step": 7135
    },
    {
      "entropy": 6.0007603645324705,
      "epoch": 0.6239622476623263,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004967549488578635,
      "loss": 5.7908,
      "mean_token_accuracy": 0.16461694687604905,
      "num_tokens": 14469375.0,
      "step": 7140
    },
    {
      "entropy": 6.048993444442749,
      "epoch": 0.624399196015031,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000496749673513592,
      "loss": 5.7172,
      "mean_token_accuracy": 0.16761179268360138,
      "num_tokens": 14478444.0,
      "step": 7145
    },
    {
      "entropy": 5.916772747039795,
      "epoch": 0.6248361443677357,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004967443939160456,
      "loss": 5.6042,
      "mean_token_accuracy": 0.17784847766160966,
      "num_tokens": 14487776.0,
      "step": 7150
    },
    {
      "entropy": 5.966445207595825,
      "epoch": 0.6252730927204404,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004967391100653255,
      "loss": 5.7261,
      "mean_token_accuracy": 0.17165634036064148,
      "num_tokens": 14497879.0,
      "step": 7155
    },
    {
      "entropy": 5.960116243362426,
      "epoch": 0.6257100410731452,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004967338219615331,
      "loss": 5.6545,
      "mean_token_accuracy": 0.17122041583061218,
      "num_tokens": 14507753.0,
      "step": 7160
    },
    {
      "entropy": 5.936701536178589,
      "epoch": 0.6261469894258499,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004967285296047699,
      "loss": 5.6224,
      "mean_token_accuracy": 0.1743711367249489,
      "num_tokens": 14517086.0,
      "step": 7165
    },
    {
      "entropy": 5.944527530670166,
      "epoch": 0.6265839377785546,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004967232329951372,
      "loss": 5.7048,
      "mean_token_accuracy": 0.1624232068657875,
      "num_tokens": 14527097.0,
      "step": 7170
    },
    {
      "entropy": 5.955741024017334,
      "epoch": 0.6270208861312593,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004967179321327367,
      "loss": 5.7381,
      "mean_token_accuracy": 0.16680745780467987,
      "num_tokens": 14537924.0,
      "step": 7175
    },
    {
      "entropy": 5.996904563903809,
      "epoch": 0.627457834483964,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004967126270176702,
      "loss": 5.7423,
      "mean_token_accuracy": 0.16276608407497406,
      "num_tokens": 14547542.0,
      "step": 7180
    },
    {
      "entropy": 5.957862281799317,
      "epoch": 0.6278947828366687,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004967073176500392,
      "loss": 5.6543,
      "mean_token_accuracy": 0.1741482809185982,
      "num_tokens": 14557150.0,
      "step": 7185
    },
    {
      "entropy": 6.020190811157226,
      "epoch": 0.6283317311893735,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004967020040299458,
      "loss": 5.6802,
      "mean_token_accuracy": 0.1717903807759285,
      "num_tokens": 14567242.0,
      "step": 7190
    },
    {
      "entropy": 5.861844778060913,
      "epoch": 0.6287686795420782,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004966966861574917,
      "loss": 5.6527,
      "mean_token_accuracy": 0.17167726755142212,
      "num_tokens": 14576631.0,
      "step": 7195
    },
    {
      "entropy": 5.9593912124633786,
      "epoch": 0.6292056278947828,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000496691364032779,
      "loss": 5.7546,
      "mean_token_accuracy": 0.16851683855056762,
      "num_tokens": 14587731.0,
      "step": 7200
    },
    {
      "entropy": 6.070434951782227,
      "epoch": 0.6296425762474875,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004966860376559097,
      "loss": 5.7535,
      "mean_token_accuracy": 0.16793397217988967,
      "num_tokens": 14598136.0,
      "step": 7205
    },
    {
      "entropy": 5.919837808609008,
      "epoch": 0.6300795246001922,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004966807070269862,
      "loss": 5.715,
      "mean_token_accuracy": 0.16734849512577057,
      "num_tokens": 14607243.0,
      "step": 7210
    },
    {
      "entropy": 5.925194597244262,
      "epoch": 0.6305164729528969,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004966753721461105,
      "loss": 5.717,
      "mean_token_accuracy": 0.16665616780519485,
      "num_tokens": 14617094.0,
      "step": 7215
    },
    {
      "entropy": 6.044453096389771,
      "epoch": 0.6309534213056017,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004966700330133849,
      "loss": 5.7428,
      "mean_token_accuracy": 0.16312877386808394,
      "num_tokens": 14626486.0,
      "step": 7220
    },
    {
      "entropy": 5.854386329650879,
      "epoch": 0.6313903696583064,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004966646896289121,
      "loss": 5.606,
      "mean_token_accuracy": 0.18634792119264604,
      "num_tokens": 14637170.0,
      "step": 7225
    },
    {
      "entropy": 5.967525482177734,
      "epoch": 0.6318273180110111,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004966593419927944,
      "loss": 5.6754,
      "mean_token_accuracy": 0.1700664222240448,
      "num_tokens": 14646396.0,
      "step": 7230
    },
    {
      "entropy": 6.0127198696136475,
      "epoch": 0.6322642663637158,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004966539901051343,
      "loss": 5.7426,
      "mean_token_accuracy": 0.16377434134483337,
      "num_tokens": 14656097.0,
      "step": 7235
    },
    {
      "entropy": 5.944140195846558,
      "epoch": 0.6327012147164205,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004966486339660345,
      "loss": 5.6917,
      "mean_token_accuracy": 0.17073993235826493,
      "num_tokens": 14667124.0,
      "step": 7240
    },
    {
      "entropy": 5.9437713623046875,
      "epoch": 0.6331381630691252,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004966432735755978,
      "loss": 5.786,
      "mean_token_accuracy": 0.16578828692436218,
      "num_tokens": 14677427.0,
      "step": 7245
    },
    {
      "entropy": 5.923660945892334,
      "epoch": 0.63357511142183,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000496637908933927,
      "loss": 5.6141,
      "mean_token_accuracy": 0.17330324798822402,
      "num_tokens": 14686429.0,
      "step": 7250
    },
    {
      "entropy": 5.961338138580322,
      "epoch": 0.6340120597745347,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000496632540041125,
      "loss": 5.6996,
      "mean_token_accuracy": 0.17854312360286712,
      "num_tokens": 14696264.0,
      "step": 7255
    },
    {
      "entropy": 5.997078227996826,
      "epoch": 0.6344490081272394,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004966271668972946,
      "loss": 5.7015,
      "mean_token_accuracy": 0.16481100469827653,
      "num_tokens": 14705935.0,
      "step": 7260
    },
    {
      "entropy": 5.9779621124267575,
      "epoch": 0.6348859564799441,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004966217895025391,
      "loss": 5.6807,
      "mean_token_accuracy": 0.1679082229733467,
      "num_tokens": 14715978.0,
      "step": 7265
    },
    {
      "entropy": 6.02035870552063,
      "epoch": 0.6353229048326487,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004966164078569616,
      "loss": 5.6797,
      "mean_token_accuracy": 0.16719867885112763,
      "num_tokens": 14726170.0,
      "step": 7270
    },
    {
      "entropy": 5.915210962295532,
      "epoch": 0.6357598531853534,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004966110219606652,
      "loss": 5.7012,
      "mean_token_accuracy": 0.16877488121390344,
      "num_tokens": 14735775.0,
      "step": 7275
    },
    {
      "entropy": 5.944903802871704,
      "epoch": 0.6361968015380582,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004966056318137532,
      "loss": 5.6554,
      "mean_token_accuracy": 0.17061079442501068,
      "num_tokens": 14746143.0,
      "step": 7280
    },
    {
      "entropy": 6.000142526626587,
      "epoch": 0.6366337498907629,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004966002374163293,
      "loss": 5.7132,
      "mean_token_accuracy": 0.17375547885894777,
      "num_tokens": 14756707.0,
      "step": 7285
    },
    {
      "entropy": 5.9694061279296875,
      "epoch": 0.6370706982434676,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004965948387684964,
      "loss": 5.6979,
      "mean_token_accuracy": 0.17061454355716704,
      "num_tokens": 14765778.0,
      "step": 7290
    },
    {
      "entropy": 5.941481018066407,
      "epoch": 0.6375076465961723,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004965894358703587,
      "loss": 5.7343,
      "mean_token_accuracy": 0.16467194259166718,
      "num_tokens": 14775264.0,
      "step": 7295
    },
    {
      "entropy": 5.961967849731446,
      "epoch": 0.637944594948877,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004965840287220193,
      "loss": 5.6908,
      "mean_token_accuracy": 0.17460215091705322,
      "num_tokens": 14785754.0,
      "step": 7300
    },
    {
      "entropy": 5.958018779754639,
      "epoch": 0.6383815433015817,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004965786173235821,
      "loss": 5.7366,
      "mean_token_accuracy": 0.16654301583766937,
      "num_tokens": 14795936.0,
      "step": 7305
    },
    {
      "entropy": 6.029556846618652,
      "epoch": 0.6388184916542865,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004965732016751511,
      "loss": 5.7296,
      "mean_token_accuracy": 0.16779305338859557,
      "num_tokens": 14805721.0,
      "step": 7310
    },
    {
      "entropy": 5.952966499328613,
      "epoch": 0.6392554400069912,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004965677817768299,
      "loss": 5.7361,
      "mean_token_accuracy": 0.16711219102144242,
      "num_tokens": 14816610.0,
      "step": 7315
    },
    {
      "entropy": 6.017750930786133,
      "epoch": 0.6396923883596959,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004965623576287225,
      "loss": 5.7169,
      "mean_token_accuracy": 0.1665639027953148,
      "num_tokens": 14826126.0,
      "step": 7320
    },
    {
      "entropy": 5.945874404907227,
      "epoch": 0.6401293367124006,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004965569292309331,
      "loss": 5.612,
      "mean_token_accuracy": 0.1770862326025963,
      "num_tokens": 14835657.0,
      "step": 7325
    },
    {
      "entropy": 5.951558065414429,
      "epoch": 0.6405662850651053,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004965514965835655,
      "loss": 5.705,
      "mean_token_accuracy": 0.1689366951584816,
      "num_tokens": 14845430.0,
      "step": 7330
    },
    {
      "entropy": 5.972777223587036,
      "epoch": 0.6410032334178101,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004965460596867243,
      "loss": 5.6732,
      "mean_token_accuracy": 0.17379135340452195,
      "num_tokens": 14855459.0,
      "step": 7335
    },
    {
      "entropy": 6.005117988586425,
      "epoch": 0.6414401817705148,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004965406185405135,
      "loss": 5.6915,
      "mean_token_accuracy": 0.17110253721475602,
      "num_tokens": 14865375.0,
      "step": 7340
    },
    {
      "entropy": 5.967947340011596,
      "epoch": 0.6418771301232195,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004965351731450376,
      "loss": 5.6691,
      "mean_token_accuracy": 0.17488979697227477,
      "num_tokens": 14875211.0,
      "step": 7345
    },
    {
      "entropy": 5.924662351608276,
      "epoch": 0.6423140784759241,
      "grad_norm": 1.0,
      "learning_rate": 0.000496529723500401,
      "loss": 5.7326,
      "mean_token_accuracy": 0.16935643553733826,
      "num_tokens": 14886221.0,
      "step": 7350
    },
    {
      "entropy": 6.023074054718018,
      "epoch": 0.6427510268286288,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004965242696067081,
      "loss": 5.6894,
      "mean_token_accuracy": 0.17177527248859406,
      "num_tokens": 14896002.0,
      "step": 7355
    },
    {
      "entropy": 6.034801292419433,
      "epoch": 0.6431879751813335,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004965188114640637,
      "loss": 5.7565,
      "mean_token_accuracy": 0.16670249700546264,
      "num_tokens": 14906311.0,
      "step": 7360
    },
    {
      "entropy": 6.023222208023071,
      "epoch": 0.6436249235340383,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004965133490725725,
      "loss": 5.7061,
      "mean_token_accuracy": 0.171078884601593,
      "num_tokens": 14915890.0,
      "step": 7365
    },
    {
      "entropy": 5.990642356872558,
      "epoch": 0.644061871886743,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004965078824323391,
      "loss": 5.6912,
      "mean_token_accuracy": 0.169252710044384,
      "num_tokens": 14925891.0,
      "step": 7370
    },
    {
      "entropy": 5.931860303878784,
      "epoch": 0.6444988202394477,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004965024115434685,
      "loss": 5.7136,
      "mean_token_accuracy": 0.169284325838089,
      "num_tokens": 14937272.0,
      "step": 7375
    },
    {
      "entropy": 5.962608432769775,
      "epoch": 0.6449357685921524,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004964969364060656,
      "loss": 5.7291,
      "mean_token_accuracy": 0.16876414567232131,
      "num_tokens": 14948135.0,
      "step": 7380
    },
    {
      "entropy": 6.05620551109314,
      "epoch": 0.6453727169448571,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004964914570202353,
      "loss": 5.7946,
      "mean_token_accuracy": 0.16149679869413375,
      "num_tokens": 14959254.0,
      "step": 7385
    },
    {
      "entropy": 5.972436618804932,
      "epoch": 0.6458096652975618,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004964859733860829,
      "loss": 5.7003,
      "mean_token_accuracy": 0.17190984040498733,
      "num_tokens": 14969313.0,
      "step": 7390
    },
    {
      "entropy": 5.944139289855957,
      "epoch": 0.6462466136502666,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004964804855037134,
      "loss": 5.704,
      "mean_token_accuracy": 0.16546256095170975,
      "num_tokens": 14978963.0,
      "step": 7395
    },
    {
      "entropy": 6.032906818389892,
      "epoch": 0.6466835620029713,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004964749933732322,
      "loss": 5.7251,
      "mean_token_accuracy": 0.17265428006649017,
      "num_tokens": 14989355.0,
      "step": 7400
    },
    {
      "entropy": 5.979280424118042,
      "epoch": 0.647120510355676,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004964694969947445,
      "loss": 5.689,
      "mean_token_accuracy": 0.16724557429552078,
      "num_tokens": 14998995.0,
      "step": 7405
    },
    {
      "entropy": 5.959960031509399,
      "epoch": 0.6475574587083807,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004964639963683558,
      "loss": 5.7312,
      "mean_token_accuracy": 0.16624291241168976,
      "num_tokens": 15008762.0,
      "step": 7410
    },
    {
      "entropy": 6.036027145385742,
      "epoch": 0.6479944070610854,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004964584914941716,
      "loss": 5.6947,
      "mean_token_accuracy": 0.16754455715417862,
      "num_tokens": 15020062.0,
      "step": 7415
    },
    {
      "entropy": 6.0214667320251465,
      "epoch": 0.64843135541379,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004964529823722975,
      "loss": 5.8579,
      "mean_token_accuracy": 0.15754903256893157,
      "num_tokens": 15030440.0,
      "step": 7420
    },
    {
      "entropy": 5.9521252632141115,
      "epoch": 0.6488683037664948,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004964474690028393,
      "loss": 5.6844,
      "mean_token_accuracy": 0.16686192154884338,
      "num_tokens": 15041709.0,
      "step": 7425
    },
    {
      "entropy": 5.962891483306885,
      "epoch": 0.6493052521191995,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004964419513859025,
      "loss": 5.7987,
      "mean_token_accuracy": 0.16418152600526809,
      "num_tokens": 15052277.0,
      "step": 7430
    },
    {
      "entropy": 5.970396852493286,
      "epoch": 0.6497422004719042,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000496436429521593,
      "loss": 5.7051,
      "mean_token_accuracy": 0.16340623795986176,
      "num_tokens": 15061870.0,
      "step": 7435
    },
    {
      "entropy": 5.873222017288208,
      "epoch": 0.6501791488246089,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004964309034100168,
      "loss": 5.5788,
      "mean_token_accuracy": 0.17656823098659516,
      "num_tokens": 15071403.0,
      "step": 7440
    },
    {
      "entropy": 5.921652936935425,
      "epoch": 0.6506160971773136,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004964253730512799,
      "loss": 5.6667,
      "mean_token_accuracy": 0.17216823548078536,
      "num_tokens": 15081651.0,
      "step": 7445
    },
    {
      "entropy": 6.012656211853027,
      "epoch": 0.6510530455300183,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004964198384454882,
      "loss": 5.7224,
      "mean_token_accuracy": 0.16656941026449204,
      "num_tokens": 15093403.0,
      "step": 7450
    },
    {
      "entropy": 5.979521942138672,
      "epoch": 0.6514899938827231,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000496414299592748,
      "loss": 5.8,
      "mean_token_accuracy": 0.16117063611745835,
      "num_tokens": 15103618.0,
      "step": 7455
    },
    {
      "entropy": 5.9783618450164795,
      "epoch": 0.6519269422354278,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004964087564931656,
      "loss": 5.7365,
      "mean_token_accuracy": 0.17271816730499268,
      "num_tokens": 15113926.0,
      "step": 7460
    },
    {
      "entropy": 6.0650975704193115,
      "epoch": 0.6523638905881325,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004964032091468472,
      "loss": 5.7653,
      "mean_token_accuracy": 0.16449286490678788,
      "num_tokens": 15125302.0,
      "step": 7465
    },
    {
      "entropy": 5.9772899627685545,
      "epoch": 0.6528008389408372,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004963976575538991,
      "loss": 5.65,
      "mean_token_accuracy": 0.17406051754951476,
      "num_tokens": 15135943.0,
      "step": 7470
    },
    {
      "entropy": 5.910923480987549,
      "epoch": 0.6532377872935419,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000496392101714428,
      "loss": 5.7273,
      "mean_token_accuracy": 0.17380317896604539,
      "num_tokens": 15145001.0,
      "step": 7475
    },
    {
      "entropy": 5.94571738243103,
      "epoch": 0.6536747356462466,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004963865416285405,
      "loss": 5.6372,
      "mean_token_accuracy": 0.16757314205169677,
      "num_tokens": 15154567.0,
      "step": 7480
    },
    {
      "entropy": 6.059688234329224,
      "epoch": 0.6541116839989514,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004963809772963431,
      "loss": 5.7867,
      "mean_token_accuracy": 0.16612908989191055,
      "num_tokens": 15164311.0,
      "step": 7485
    },
    {
      "entropy": 5.990157508850098,
      "epoch": 0.6545486323516561,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004963754087179425,
      "loss": 5.6817,
      "mean_token_accuracy": 0.1657267153263092,
      "num_tokens": 15175217.0,
      "step": 7490
    },
    {
      "entropy": 5.968144369125366,
      "epoch": 0.6549855807043607,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004963698358934455,
      "loss": 5.6836,
      "mean_token_accuracy": 0.16955721974372864,
      "num_tokens": 15184895.0,
      "step": 7495
    },
    {
      "entropy": 5.897038841247559,
      "epoch": 0.6554225290570654,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004963642588229591,
      "loss": 5.672,
      "mean_token_accuracy": 0.16963538974523545,
      "num_tokens": 15194380.0,
      "step": 7500
    },
    {
      "entropy": 5.973154020309448,
      "epoch": 0.6558594774097701,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004963586775065902,
      "loss": 5.7043,
      "mean_token_accuracy": 0.16656977236270903,
      "num_tokens": 15204364.0,
      "step": 7505
    },
    {
      "entropy": 5.9842935562133786,
      "epoch": 0.6562964257624748,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004963530919444459,
      "loss": 5.6752,
      "mean_token_accuracy": 0.16770085990428923,
      "num_tokens": 15214485.0,
      "step": 7510
    },
    {
      "entropy": 6.022993755340576,
      "epoch": 0.6567333741151796,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004963475021366333,
      "loss": 5.8068,
      "mean_token_accuracy": 0.16371995359659194,
      "num_tokens": 15224764.0,
      "step": 7515
    },
    {
      "entropy": 6.007327842712402,
      "epoch": 0.6571703224678843,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004963419080832597,
      "loss": 5.8052,
      "mean_token_accuracy": 0.16424482613801955,
      "num_tokens": 15235481.0,
      "step": 7520
    },
    {
      "entropy": 6.07501974105835,
      "epoch": 0.657607270820589,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004963363097844323,
      "loss": 5.7606,
      "mean_token_accuracy": 0.16616501659154892,
      "num_tokens": 15245332.0,
      "step": 7525
    },
    {
      "entropy": 6.035985136032105,
      "epoch": 0.6580442191732937,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004963307072402584,
      "loss": 5.7581,
      "mean_token_accuracy": 0.16567514836788177,
      "num_tokens": 15255562.0,
      "step": 7530
    },
    {
      "entropy": 5.985346889495849,
      "epoch": 0.6584811675259984,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004963251004508456,
      "loss": 5.6967,
      "mean_token_accuracy": 0.16489960700273515,
      "num_tokens": 15265655.0,
      "step": 7535
    },
    {
      "entropy": 5.908224296569824,
      "epoch": 0.6589181158787032,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004963194894163015,
      "loss": 5.6232,
      "mean_token_accuracy": 0.17775764912366868,
      "num_tokens": 15276311.0,
      "step": 7540
    },
    {
      "entropy": 5.973880863189697,
      "epoch": 0.6593550642314079,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004963138741367334,
      "loss": 5.7759,
      "mean_token_accuracy": 0.16482124328613282,
      "num_tokens": 15287142.0,
      "step": 7545
    },
    {
      "entropy": 6.0848041534423825,
      "epoch": 0.6597920125841126,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004963082546122495,
      "loss": 5.7635,
      "mean_token_accuracy": 0.1659230560064316,
      "num_tokens": 15297484.0,
      "step": 7550
    },
    {
      "entropy": 6.021333360671997,
      "epoch": 0.6602289609368173,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004963026308429571,
      "loss": 5.7151,
      "mean_token_accuracy": 0.16891633421182634,
      "num_tokens": 15307468.0,
      "step": 7555
    },
    {
      "entropy": 5.90835165977478,
      "epoch": 0.660665909289522,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004962970028289644,
      "loss": 5.666,
      "mean_token_accuracy": 0.17503694742918013,
      "num_tokens": 15318487.0,
      "step": 7560
    },
    {
      "entropy": 5.9384973526000975,
      "epoch": 0.6611028576422266,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004962913705703791,
      "loss": 5.7251,
      "mean_token_accuracy": 0.1709133878350258,
      "num_tokens": 15329193.0,
      "step": 7565
    },
    {
      "entropy": 5.999801683425903,
      "epoch": 0.6615398059949315,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004962857340673095,
      "loss": 5.6783,
      "mean_token_accuracy": 0.16653147339820862,
      "num_tokens": 15338886.0,
      "step": 7570
    },
    {
      "entropy": 5.904755640029907,
      "epoch": 0.6619767543476361,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004962800933198635,
      "loss": 5.5952,
      "mean_token_accuracy": 0.18010661602020264,
      "num_tokens": 15348629.0,
      "step": 7575
    },
    {
      "entropy": 5.951264762878418,
      "epoch": 0.6624137027003408,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004962744483281495,
      "loss": 5.7272,
      "mean_token_accuracy": 0.17247799038887024,
      "num_tokens": 15358004.0,
      "step": 7580
    },
    {
      "entropy": 5.987465286254883,
      "epoch": 0.6628506510530455,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004962687990922754,
      "loss": 5.7075,
      "mean_token_accuracy": 0.16978215128183366,
      "num_tokens": 15368018.0,
      "step": 7585
    },
    {
      "entropy": 6.071308708190918,
      "epoch": 0.6632875994057502,
      "grad_norm": 1.03125,
      "learning_rate": 0.00049626314561235,
      "loss": 5.6882,
      "mean_token_accuracy": 0.16783523112535476,
      "num_tokens": 15377443.0,
      "step": 7590
    },
    {
      "entropy": 6.008503246307373,
      "epoch": 0.6637245477584549,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004962574878884814,
      "loss": 5.7121,
      "mean_token_accuracy": 0.16673815995454788,
      "num_tokens": 15388278.0,
      "step": 7595
    },
    {
      "entropy": 5.95421199798584,
      "epoch": 0.6641614961111597,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004962518259207782,
      "loss": 5.7117,
      "mean_token_accuracy": 0.1696106806397438,
      "num_tokens": 15398619.0,
      "step": 7600
    },
    {
      "entropy": 5.911576128005981,
      "epoch": 0.6645984444638644,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004962461597093491,
      "loss": 5.614,
      "mean_token_accuracy": 0.1788553386926651,
      "num_tokens": 15408635.0,
      "step": 7605
    },
    {
      "entropy": 6.027652788162231,
      "epoch": 0.6650353928165691,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004962404892543028,
      "loss": 5.7208,
      "mean_token_accuracy": 0.1637152314186096,
      "num_tokens": 15419113.0,
      "step": 7610
    },
    {
      "entropy": 6.052756309509277,
      "epoch": 0.6654723411692738,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000496234814555748,
      "loss": 5.7004,
      "mean_token_accuracy": 0.16512987166643142,
      "num_tokens": 15428959.0,
      "step": 7615
    },
    {
      "entropy": 5.984261083602905,
      "epoch": 0.6659092895219785,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004962291356137934,
      "loss": 5.7735,
      "mean_token_accuracy": 0.1598317191004753,
      "num_tokens": 15439116.0,
      "step": 7620
    },
    {
      "entropy": 5.963285446166992,
      "epoch": 0.6663462378746832,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004962234524285481,
      "loss": 5.7651,
      "mean_token_accuracy": 0.17018457502126694,
      "num_tokens": 15449715.0,
      "step": 7625
    },
    {
      "entropy": 5.94545431137085,
      "epoch": 0.666783186227388,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004962177650001212,
      "loss": 5.5425,
      "mean_token_accuracy": 0.18361233919858932,
      "num_tokens": 15459170.0,
      "step": 7630
    },
    {
      "entropy": 5.911886310577392,
      "epoch": 0.6672201345800927,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004962120733286215,
      "loss": 5.6346,
      "mean_token_accuracy": 0.17728826105594636,
      "num_tokens": 15469497.0,
      "step": 7635
    },
    {
      "entropy": 5.926416397094727,
      "epoch": 0.6676570829327974,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004962063774141583,
      "loss": 5.7829,
      "mean_token_accuracy": 0.16903098970651625,
      "num_tokens": 15481298.0,
      "step": 7640
    },
    {
      "entropy": 6.028058767318726,
      "epoch": 0.668094031285502,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004962006772568409,
      "loss": 5.7293,
      "mean_token_accuracy": 0.1712724357843399,
      "num_tokens": 15491696.0,
      "step": 7645
    },
    {
      "entropy": 6.05077862739563,
      "epoch": 0.6685309796382067,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004961949728567787,
      "loss": 5.7292,
      "mean_token_accuracy": 0.16523629873991014,
      "num_tokens": 15502412.0,
      "step": 7650
    },
    {
      "entropy": 5.9615202903747555,
      "epoch": 0.6689679279909114,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004961892642140809,
      "loss": 5.713,
      "mean_token_accuracy": 0.16547127962112426,
      "num_tokens": 15513436.0,
      "step": 7655
    },
    {
      "entropy": 5.967373895645141,
      "epoch": 0.6694048763436162,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004961835513288572,
      "loss": 5.7173,
      "mean_token_accuracy": 0.17350703328847886,
      "num_tokens": 15524564.0,
      "step": 7660
    },
    {
      "entropy": 5.969564437866211,
      "epoch": 0.6698418246963209,
      "grad_norm": 0.921875,
      "learning_rate": 0.000496177834201217,
      "loss": 5.6795,
      "mean_token_accuracy": 0.16774281710386277,
      "num_tokens": 15534539.0,
      "step": 7665
    },
    {
      "entropy": 6.004529523849487,
      "epoch": 0.6702787730490256,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004961721128312702,
      "loss": 5.7629,
      "mean_token_accuracy": 0.15742171108722686,
      "num_tokens": 15544571.0,
      "step": 7670
    },
    {
      "entropy": 6.029096603393555,
      "epoch": 0.6707157214017303,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004961663872191263,
      "loss": 5.6988,
      "mean_token_accuracy": 0.17978453636169434,
      "num_tokens": 15555066.0,
      "step": 7675
    },
    {
      "entropy": 5.968546438217163,
      "epoch": 0.671152669754435,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004961606573648951,
      "loss": 5.7204,
      "mean_token_accuracy": 0.16951085776090621,
      "num_tokens": 15565128.0,
      "step": 7680
    },
    {
      "entropy": 5.947306394577026,
      "epoch": 0.6715896181071397,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004961549232686867,
      "loss": 5.7654,
      "mean_token_accuracy": 0.16702262312173843,
      "num_tokens": 15576394.0,
      "step": 7685
    },
    {
      "entropy": 6.061140441894532,
      "epoch": 0.6720265664598445,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000496149184930611,
      "loss": 5.7009,
      "mean_token_accuracy": 0.16518072932958602,
      "num_tokens": 15585806.0,
      "step": 7690
    },
    {
      "entropy": 5.983309602737426,
      "epoch": 0.6724635148125492,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004961434423507779,
      "loss": 5.6915,
      "mean_token_accuracy": 0.17046522945165635,
      "num_tokens": 15596471.0,
      "step": 7695
    },
    {
      "entropy": 5.953005027770996,
      "epoch": 0.6729004631652539,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004961376955292979,
      "loss": 5.6088,
      "mean_token_accuracy": 0.18013895750045777,
      "num_tokens": 15605720.0,
      "step": 7700
    },
    {
      "entropy": 5.965842151641846,
      "epoch": 0.6733374115179586,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004961319444662809,
      "loss": 5.6848,
      "mean_token_accuracy": 0.17525060772895812,
      "num_tokens": 15616167.0,
      "step": 7705
    },
    {
      "entropy": 6.017074680328369,
      "epoch": 0.6737743598706633,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004961261891618374,
      "loss": 5.7816,
      "mean_token_accuracy": 0.15856820940971375,
      "num_tokens": 15627033.0,
      "step": 7710
    },
    {
      "entropy": 5.9991613864898685,
      "epoch": 0.6742113082233679,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004961204296160777,
      "loss": 5.7273,
      "mean_token_accuracy": 0.1706279642879963,
      "num_tokens": 15637460.0,
      "step": 7715
    },
    {
      "entropy": 5.982107162475586,
      "epoch": 0.6746482565760727,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004961146658291123,
      "loss": 5.654,
      "mean_token_accuracy": 0.17517194747924805,
      "num_tokens": 15646027.0,
      "step": 7720
    },
    {
      "entropy": 5.976183080673218,
      "epoch": 0.6750852049287774,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004961088978010518,
      "loss": 5.67,
      "mean_token_accuracy": 0.17170514911413193,
      "num_tokens": 15655241.0,
      "step": 7725
    },
    {
      "entropy": 5.938190698623657,
      "epoch": 0.6755221532814821,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004961031255320067,
      "loss": 5.6261,
      "mean_token_accuracy": 0.168367463350296,
      "num_tokens": 15666213.0,
      "step": 7730
    },
    {
      "entropy": 6.035438871383667,
      "epoch": 0.6759591016341868,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004960973490220878,
      "loss": 5.7081,
      "mean_token_accuracy": 0.16741962134838104,
      "num_tokens": 15676985.0,
      "step": 7735
    },
    {
      "entropy": 6.012124919891358,
      "epoch": 0.6763960499868915,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004960915682714059,
      "loss": 5.753,
      "mean_token_accuracy": 0.1617443323135376,
      "num_tokens": 15687519.0,
      "step": 7740
    },
    {
      "entropy": 6.010191535949707,
      "epoch": 0.6768329983395963,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004960857832800719,
      "loss": 5.7121,
      "mean_token_accuracy": 0.16730067878961563,
      "num_tokens": 15697401.0,
      "step": 7745
    },
    {
      "entropy": 5.94223952293396,
      "epoch": 0.677269946692301,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004960799940481968,
      "loss": 5.7411,
      "mean_token_accuracy": 0.16217271834611893,
      "num_tokens": 15708159.0,
      "step": 7750
    },
    {
      "entropy": 6.0108153343200685,
      "epoch": 0.6777068950450057,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004960742005758914,
      "loss": 5.6916,
      "mean_token_accuracy": 0.16727320700883866,
      "num_tokens": 15717749.0,
      "step": 7755
    },
    {
      "entropy": 5.946748733520508,
      "epoch": 0.6781438433977104,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004960684028632671,
      "loss": 5.6441,
      "mean_token_accuracy": 0.17235903888940812,
      "num_tokens": 15726868.0,
      "step": 7760
    },
    {
      "entropy": 5.993214893341064,
      "epoch": 0.6785807917504151,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004960626009104349,
      "loss": 5.6578,
      "mean_token_accuracy": 0.17183670103549958,
      "num_tokens": 15736038.0,
      "step": 7765
    },
    {
      "entropy": 5.977568769454956,
      "epoch": 0.6790177401031198,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004960567947175064,
      "loss": 5.6915,
      "mean_token_accuracy": 0.16744136661291123,
      "num_tokens": 15746676.0,
      "step": 7770
    },
    {
      "entropy": 6.004445934295655,
      "epoch": 0.6794546884558246,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004960509842845926,
      "loss": 5.6759,
      "mean_token_accuracy": 0.1715102717280388,
      "num_tokens": 15756064.0,
      "step": 7775
    },
    {
      "entropy": 5.96464352607727,
      "epoch": 0.6798916368085293,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004960451696118049,
      "loss": 5.6824,
      "mean_token_accuracy": 0.1720867395401001,
      "num_tokens": 15767414.0,
      "step": 7780
    },
    {
      "entropy": 5.98832745552063,
      "epoch": 0.680328585161234,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004960393506992553,
      "loss": 5.6596,
      "mean_token_accuracy": 0.17125210613012315,
      "num_tokens": 15778276.0,
      "step": 7785
    },
    {
      "entropy": 5.922569847106933,
      "epoch": 0.6807655335139386,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004960335275470551,
      "loss": 5.6013,
      "mean_token_accuracy": 0.17457276582717896,
      "num_tokens": 15788365.0,
      "step": 7790
    },
    {
      "entropy": 5.954694223403931,
      "epoch": 0.6812024818666433,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004960277001553159,
      "loss": 5.6011,
      "mean_token_accuracy": 0.17465912103652953,
      "num_tokens": 15799021.0,
      "step": 7795
    },
    {
      "entropy": 5.878193044662476,
      "epoch": 0.681639430219348,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004960218685241498,
      "loss": 5.6892,
      "mean_token_accuracy": 0.17144405841827393,
      "num_tokens": 15808779.0,
      "step": 7800
    },
    {
      "entropy": 6.01234016418457,
      "epoch": 0.6820763785720528,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004960160326536681,
      "loss": 5.6972,
      "mean_token_accuracy": 0.17568242847919463,
      "num_tokens": 15818515.0,
      "step": 7805
    },
    {
      "entropy": 6.002443742752075,
      "epoch": 0.6825133269247575,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004960101925439834,
      "loss": 5.6666,
      "mean_token_accuracy": 0.17127903699874877,
      "num_tokens": 15828843.0,
      "step": 7810
    },
    {
      "entropy": 5.9726309299469,
      "epoch": 0.6829502752774622,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004960043481952072,
      "loss": 5.6615,
      "mean_token_accuracy": 0.17104052752256393,
      "num_tokens": 15839323.0,
      "step": 7815
    },
    {
      "entropy": 5.954789066314698,
      "epoch": 0.6833872236301669,
      "grad_norm": 0.9375,
      "learning_rate": 0.000495998499607452,
      "loss": 5.694,
      "mean_token_accuracy": 0.1714374989271164,
      "num_tokens": 15848399.0,
      "step": 7820
    },
    {
      "entropy": 5.944324016571045,
      "epoch": 0.6838241719828716,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004959926467808296,
      "loss": 5.5734,
      "mean_token_accuracy": 0.17026680558919907,
      "num_tokens": 15858137.0,
      "step": 7825
    },
    {
      "entropy": 5.910632944107055,
      "epoch": 0.6842611203355763,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004959867897154526,
      "loss": 5.6857,
      "mean_token_accuracy": 0.16693269461393356,
      "num_tokens": 15867643.0,
      "step": 7830
    },
    {
      "entropy": 5.987176895141602,
      "epoch": 0.6846980686882811,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000495980928411433,
      "loss": 5.577,
      "mean_token_accuracy": 0.17323559671640396,
      "num_tokens": 15877523.0,
      "step": 7835
    },
    {
      "entropy": 5.984577989578247,
      "epoch": 0.6851350170409858,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004959750628688834,
      "loss": 5.7627,
      "mean_token_accuracy": 0.16247362345457078,
      "num_tokens": 15886946.0,
      "step": 7840
    },
    {
      "entropy": 5.919642972946167,
      "epoch": 0.6855719653936905,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004959691930879164,
      "loss": 5.6032,
      "mean_token_accuracy": 0.17992403656244277,
      "num_tokens": 15896646.0,
      "step": 7845
    },
    {
      "entropy": 5.924122858047485,
      "epoch": 0.6860089137463952,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004959633190686445,
      "loss": 5.6271,
      "mean_token_accuracy": 0.17496545761823654,
      "num_tokens": 15907204.0,
      "step": 7850
    },
    {
      "entropy": 5.943504285812378,
      "epoch": 0.6864458620990999,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004959574408111804,
      "loss": 5.6821,
      "mean_token_accuracy": 0.1758973315358162,
      "num_tokens": 15917634.0,
      "step": 7855
    },
    {
      "entropy": 5.947372913360596,
      "epoch": 0.6868828104518045,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004959515583156367,
      "loss": 5.6523,
      "mean_token_accuracy": 0.17229314893484116,
      "num_tokens": 15926584.0,
      "step": 7860
    },
    {
      "entropy": 5.925803089141846,
      "epoch": 0.6873197588045094,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004959456715821264,
      "loss": 5.6293,
      "mean_token_accuracy": 0.1727978378534317,
      "num_tokens": 15936636.0,
      "step": 7865
    },
    {
      "entropy": 5.952526473999024,
      "epoch": 0.687756707157214,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004959397806107624,
      "loss": 5.6401,
      "mean_token_accuracy": 0.1804833695292473,
      "num_tokens": 15946113.0,
      "step": 7870
    },
    {
      "entropy": 5.986904287338257,
      "epoch": 0.6881936555099187,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004959338854016574,
      "loss": 5.6878,
      "mean_token_accuracy": 0.17136859595775605,
      "num_tokens": 15956630.0,
      "step": 7875
    },
    {
      "entropy": 5.895862674713134,
      "epoch": 0.6886306038626234,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004959279859549249,
      "loss": 5.6105,
      "mean_token_accuracy": 0.1797366663813591,
      "num_tokens": 15967129.0,
      "step": 7880
    },
    {
      "entropy": 5.953169870376587,
      "epoch": 0.6890675522153281,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004959220822706779,
      "loss": 5.6235,
      "mean_token_accuracy": 0.17317614555358887,
      "num_tokens": 15976663.0,
      "step": 7885
    },
    {
      "entropy": 5.932521772384644,
      "epoch": 0.6895045005680328,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004959161743490295,
      "loss": 5.6293,
      "mean_token_accuracy": 0.17501546293497086,
      "num_tokens": 15986666.0,
      "step": 7890
    },
    {
      "entropy": 5.86325216293335,
      "epoch": 0.6899414489207376,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004959102621900931,
      "loss": 5.5636,
      "mean_token_accuracy": 0.18833111673593522,
      "num_tokens": 15996233.0,
      "step": 7895
    },
    {
      "entropy": 5.9696207523345945,
      "epoch": 0.6903783972734423,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004959043457939822,
      "loss": 5.7564,
      "mean_token_accuracy": 0.16811695843935012,
      "num_tokens": 16008188.0,
      "step": 7900
    },
    {
      "entropy": 5.987939405441284,
      "epoch": 0.690815345626147,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004958984251608101,
      "loss": 5.6628,
      "mean_token_accuracy": 0.1699799805879593,
      "num_tokens": 16018494.0,
      "step": 7905
    },
    {
      "entropy": 5.952067232131958,
      "epoch": 0.6912522939788517,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004958925002906905,
      "loss": 5.6747,
      "mean_token_accuracy": 0.17613188177347183,
      "num_tokens": 16028531.0,
      "step": 7910
    },
    {
      "entropy": 5.965585947036743,
      "epoch": 0.6916892423315564,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000495886571183737,
      "loss": 5.6508,
      "mean_token_accuracy": 0.17149328589439392,
      "num_tokens": 16039050.0,
      "step": 7915
    },
    {
      "entropy": 5.934993314743042,
      "epoch": 0.6921261906842611,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004958806378400634,
      "loss": 5.7484,
      "mean_token_accuracy": 0.16022841036319732,
      "num_tokens": 16050751.0,
      "step": 7920
    },
    {
      "entropy": 5.99234676361084,
      "epoch": 0.6925631390369659,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004958747002597834,
      "loss": 5.6706,
      "mean_token_accuracy": 0.16815691888332368,
      "num_tokens": 16060758.0,
      "step": 7925
    },
    {
      "entropy": 5.990365743637085,
      "epoch": 0.6930000873896706,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004958687584430109,
      "loss": 5.6769,
      "mean_token_accuracy": 0.17218579202890397,
      "num_tokens": 16070852.0,
      "step": 7930
    },
    {
      "entropy": 5.902193450927735,
      "epoch": 0.6934370357423753,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004958628123898597,
      "loss": 5.607,
      "mean_token_accuracy": 0.17413050234317778,
      "num_tokens": 16080081.0,
      "step": 7935
    },
    {
      "entropy": 5.95090913772583,
      "epoch": 0.6938739840950799,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004958568621004442,
      "loss": 5.6358,
      "mean_token_accuracy": 0.17013378888368608,
      "num_tokens": 16090483.0,
      "step": 7940
    },
    {
      "entropy": 6.035020446777343,
      "epoch": 0.6943109324477846,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004958509075748785,
      "loss": 5.7261,
      "mean_token_accuracy": 0.16586334854364396,
      "num_tokens": 16101136.0,
      "step": 7945
    },
    {
      "entropy": 5.943944025039673,
      "epoch": 0.6947478808004894,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004958449488132764,
      "loss": 5.6302,
      "mean_token_accuracy": 0.17929710447788239,
      "num_tokens": 16110124.0,
      "step": 7950
    },
    {
      "entropy": 5.879774522781372,
      "epoch": 0.6951848291531941,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004958389858157526,
      "loss": 5.5834,
      "mean_token_accuracy": 0.17821687757968901,
      "num_tokens": 16119581.0,
      "step": 7955
    },
    {
      "entropy": 5.925380611419678,
      "epoch": 0.6956217775058988,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004958330185824213,
      "loss": 5.6804,
      "mean_token_accuracy": 0.16888240724802017,
      "num_tokens": 16130119.0,
      "step": 7960
    },
    {
      "entropy": 5.960365533828735,
      "epoch": 0.6960587258586035,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000495827047113397,
      "loss": 5.7072,
      "mean_token_accuracy": 0.1691547855734825,
      "num_tokens": 16140702.0,
      "step": 7965
    },
    {
      "entropy": 5.970318078994751,
      "epoch": 0.6964956742113082,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004958210714087942,
      "loss": 5.6769,
      "mean_token_accuracy": 0.17426644265651703,
      "num_tokens": 16151107.0,
      "step": 7970
    },
    {
      "entropy": 5.97598557472229,
      "epoch": 0.6969326225640129,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004958150914687276,
      "loss": 5.6182,
      "mean_token_accuracy": 0.17455643117427827,
      "num_tokens": 16160779.0,
      "step": 7975
    },
    {
      "entropy": 5.882073831558228,
      "epoch": 0.6973695709167177,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004958091072933118,
      "loss": 5.5635,
      "mean_token_accuracy": 0.17653561234474183,
      "num_tokens": 16169624.0,
      "step": 7980
    },
    {
      "entropy": 5.829104328155518,
      "epoch": 0.6978065192694224,
      "grad_norm": 1.0,
      "learning_rate": 0.0004958031188826616,
      "loss": 5.565,
      "mean_token_accuracy": 0.1844409167766571,
      "num_tokens": 16179192.0,
      "step": 7985
    },
    {
      "entropy": 5.883768320083618,
      "epoch": 0.6982434676221271,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004957971262368919,
      "loss": 5.6353,
      "mean_token_accuracy": 0.17475892156362532,
      "num_tokens": 16188847.0,
      "step": 7990
    },
    {
      "entropy": 6.02799882888794,
      "epoch": 0.6986804159748318,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004957911293561176,
      "loss": 5.7442,
      "mean_token_accuracy": 0.16867593675851822,
      "num_tokens": 16199480.0,
      "step": 7995
    },
    {
      "entropy": 5.967816305160523,
      "epoch": 0.6991173643275365,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004957851282404538,
      "loss": 5.6343,
      "mean_token_accuracy": 0.17969929426908493,
      "num_tokens": 16208818.0,
      "step": 8000
    },
    {
      "entropy": 5.96125020980835,
      "epoch": 0.6995543126802412,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004957791228900154,
      "loss": 5.6159,
      "mean_token_accuracy": 0.17456261813640594,
      "num_tokens": 16219298.0,
      "step": 8005
    },
    {
      "entropy": 5.88961877822876,
      "epoch": 0.699991261032946,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004957731133049179,
      "loss": 5.6304,
      "mean_token_accuracy": 0.17814383208751677,
      "num_tokens": 16228170.0,
      "step": 8010
    },
    {
      "entropy": 5.996907711029053,
      "epoch": 0.7004282093856506,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004957670994852764,
      "loss": 5.6858,
      "mean_token_accuracy": 0.1678822547197342,
      "num_tokens": 16237552.0,
      "step": 8015
    },
    {
      "entropy": 6.010057783126831,
      "epoch": 0.7008651577383553,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004957610814312062,
      "loss": 5.7307,
      "mean_token_accuracy": 0.16357045322656633,
      "num_tokens": 16248157.0,
      "step": 8020
    },
    {
      "entropy": 5.973263216018677,
      "epoch": 0.70130210609106,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004957550591428229,
      "loss": 5.6051,
      "mean_token_accuracy": 0.17548278719186783,
      "num_tokens": 16258266.0,
      "step": 8025
    },
    {
      "entropy": 5.934809589385987,
      "epoch": 0.7017390544437647,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004957490326202418,
      "loss": 5.7068,
      "mean_token_accuracy": 0.17345640808343887,
      "num_tokens": 16268769.0,
      "step": 8030
    },
    {
      "entropy": 5.971967029571533,
      "epoch": 0.7021760027964694,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004957430018635785,
      "loss": 5.6702,
      "mean_token_accuracy": 0.17018100023269653,
      "num_tokens": 16279688.0,
      "step": 8035
    },
    {
      "entropy": 5.952615356445312,
      "epoch": 0.7026129511491742,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004957369668729488,
      "loss": 5.6441,
      "mean_token_accuracy": 0.17465924769639968,
      "num_tokens": 16289441.0,
      "step": 8040
    },
    {
      "entropy": 5.97790150642395,
      "epoch": 0.7030498995018789,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004957309276484685,
      "loss": 5.6272,
      "mean_token_accuracy": 0.1699447065591812,
      "num_tokens": 16299802.0,
      "step": 8045
    },
    {
      "entropy": 5.903774547576904,
      "epoch": 0.7034868478545836,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004957248841902533,
      "loss": 5.6132,
      "mean_token_accuracy": 0.17849778681993483,
      "num_tokens": 16309577.0,
      "step": 8050
    },
    {
      "entropy": 5.937518739700318,
      "epoch": 0.7039237962072883,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004957188364984191,
      "loss": 5.5595,
      "mean_token_accuracy": 0.18190144300460814,
      "num_tokens": 16319205.0,
      "step": 8055
    },
    {
      "entropy": 5.845996618270874,
      "epoch": 0.704360744559993,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000495712784573082,
      "loss": 5.6384,
      "mean_token_accuracy": 0.17703121900558472,
      "num_tokens": 16329650.0,
      "step": 8060
    },
    {
      "entropy": 5.89699010848999,
      "epoch": 0.7047976929126977,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004957067284143581,
      "loss": 5.6553,
      "mean_token_accuracy": 0.17381427735090255,
      "num_tokens": 16339242.0,
      "step": 8065
    },
    {
      "entropy": 5.977160310745239,
      "epoch": 0.7052346412654025,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004957006680223634,
      "loss": 5.6203,
      "mean_token_accuracy": 0.17956723272800446,
      "num_tokens": 16348441.0,
      "step": 8070
    },
    {
      "entropy": 5.9417959690094,
      "epoch": 0.7056715896181072,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004956946033972143,
      "loss": 5.6302,
      "mean_token_accuracy": 0.17256979197263717,
      "num_tokens": 16357425.0,
      "step": 8075
    },
    {
      "entropy": 5.917374229431152,
      "epoch": 0.7061085379708119,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000495688534539027,
      "loss": 5.5929,
      "mean_token_accuracy": 0.17252266108989717,
      "num_tokens": 16366180.0,
      "step": 8080
    },
    {
      "entropy": 5.9063619613647464,
      "epoch": 0.7065454863235165,
      "grad_norm": 1.0,
      "learning_rate": 0.0004956824614479181,
      "loss": 5.6502,
      "mean_token_accuracy": 0.16984034031629563,
      "num_tokens": 16375714.0,
      "step": 8085
    },
    {
      "entropy": 6.003971529006958,
      "epoch": 0.7069824346762212,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004956763841240038,
      "loss": 5.6559,
      "mean_token_accuracy": 0.1706181660294533,
      "num_tokens": 16386051.0,
      "step": 8090
    },
    {
      "entropy": 5.98858757019043,
      "epoch": 0.7074193830289259,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000495670302567401,
      "loss": 5.699,
      "mean_token_accuracy": 0.16644042283296584,
      "num_tokens": 16396616.0,
      "step": 8095
    },
    {
      "entropy": 5.918686771392823,
      "epoch": 0.7078563313816307,
      "grad_norm": 0.984375,
      "learning_rate": 0.000495664216778226,
      "loss": 5.6283,
      "mean_token_accuracy": 0.17941041737794877,
      "num_tokens": 16406301.0,
      "step": 8100
    },
    {
      "entropy": 5.9636773586273195,
      "epoch": 0.7082932797343354,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004956581267565957,
      "loss": 5.6977,
      "mean_token_accuracy": 0.1668062165379524,
      "num_tokens": 16417107.0,
      "step": 8105
    },
    {
      "entropy": 5.971851348876953,
      "epoch": 0.7087302280870401,
      "grad_norm": 0.90625,
      "learning_rate": 0.000495652032502627,
      "loss": 5.6211,
      "mean_token_accuracy": 0.17662839442491532,
      "num_tokens": 16427292.0,
      "step": 8110
    },
    {
      "entropy": 5.911182641983032,
      "epoch": 0.7091671764397448,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004956459340164366,
      "loss": 5.6238,
      "mean_token_accuracy": 0.17797663062810898,
      "num_tokens": 16438191.0,
      "step": 8115
    },
    {
      "entropy": 5.9499287605285645,
      "epoch": 0.7096041247924495,
      "grad_norm": 1.0,
      "learning_rate": 0.0004956398312981417,
      "loss": 5.6012,
      "mean_token_accuracy": 0.17193090468645095,
      "num_tokens": 16448237.0,
      "step": 8120
    },
    {
      "entropy": 5.9065721988677975,
      "epoch": 0.7100410731451542,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004956337243478592,
      "loss": 5.6841,
      "mean_token_accuracy": 0.17263902276754378,
      "num_tokens": 16458947.0,
      "step": 8125
    },
    {
      "entropy": 6.013359117507934,
      "epoch": 0.710478021497859,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004956276131657061,
      "loss": 5.6228,
      "mean_token_accuracy": 0.17274061888456343,
      "num_tokens": 16467464.0,
      "step": 8130
    },
    {
      "entropy": 5.960533332824707,
      "epoch": 0.7109149698505637,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004956214977518,
      "loss": 5.6225,
      "mean_token_accuracy": 0.1793978363275528,
      "num_tokens": 16477747.0,
      "step": 8135
    },
    {
      "entropy": 5.9507862567901615,
      "epoch": 0.7113519182032684,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004956153781062578,
      "loss": 5.7565,
      "mean_token_accuracy": 0.16770981550216674,
      "num_tokens": 16488111.0,
      "step": 8140
    },
    {
      "entropy": 5.960958242416382,
      "epoch": 0.7117888665559731,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004956092542291971,
      "loss": 5.6302,
      "mean_token_accuracy": 0.17233161479234696,
      "num_tokens": 16497689.0,
      "step": 8145
    },
    {
      "entropy": 5.914216613769531,
      "epoch": 0.7122258149086778,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004956031261207353,
      "loss": 5.7459,
      "mean_token_accuracy": 0.17005863040685654,
      "num_tokens": 16508515.0,
      "step": 8150
    },
    {
      "entropy": 5.835678148269653,
      "epoch": 0.7126627632613826,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00049559699378099,
      "loss": 5.6356,
      "mean_token_accuracy": 0.1790412947535515,
      "num_tokens": 16519184.0,
      "step": 8155
    },
    {
      "entropy": 6.008635711669922,
      "epoch": 0.7130997116140873,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004955908572100787,
      "loss": 5.7261,
      "mean_token_accuracy": 0.16656320840120314,
      "num_tokens": 16529917.0,
      "step": 8160
    },
    {
      "entropy": 5.964855909347534,
      "epoch": 0.7135366599667919,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004955847164081192,
      "loss": 5.6273,
      "mean_token_accuracy": 0.16868760585784912,
      "num_tokens": 16539687.0,
      "step": 8165
    },
    {
      "entropy": 5.9487100601196286,
      "epoch": 0.7139736083194966,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004955785713752292,
      "loss": 5.704,
      "mean_token_accuracy": 0.1707751929759979,
      "num_tokens": 16550273.0,
      "step": 8170
    },
    {
      "entropy": 6.0051532745361325,
      "epoch": 0.7144105566722013,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004955724221115267,
      "loss": 5.7664,
      "mean_token_accuracy": 0.1642640709877014,
      "num_tokens": 16562215.0,
      "step": 8175
    },
    {
      "entropy": 6.008783388137817,
      "epoch": 0.714847505024906,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004955662686171295,
      "loss": 5.7144,
      "mean_token_accuracy": 0.1742344543337822,
      "num_tokens": 16573218.0,
      "step": 8180
    },
    {
      "entropy": 6.035821437835693,
      "epoch": 0.7152844533776108,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004955601108921557,
      "loss": 5.6931,
      "mean_token_accuracy": 0.16520745158195496,
      "num_tokens": 16582740.0,
      "step": 8185
    },
    {
      "entropy": 5.920021009445191,
      "epoch": 0.7157214017303155,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004955539489367235,
      "loss": 5.6468,
      "mean_token_accuracy": 0.17994729578495025,
      "num_tokens": 16592169.0,
      "step": 8190
    },
    {
      "entropy": 5.946580839157105,
      "epoch": 0.7161583500830202,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004955477827509509,
      "loss": 5.6334,
      "mean_token_accuracy": 0.16887528598308563,
      "num_tokens": 16602706.0,
      "step": 8195
    },
    {
      "entropy": 5.935957336425782,
      "epoch": 0.7165952984357249,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004955416123349562,
      "loss": 5.6083,
      "mean_token_accuracy": 0.1749313458800316,
      "num_tokens": 16612012.0,
      "step": 8200
    },
    {
      "entropy": 5.931778335571289,
      "epoch": 0.7170322467884296,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000495535437688858,
      "loss": 5.7415,
      "mean_token_accuracy": 0.16622963547706604,
      "num_tokens": 16621798.0,
      "step": 8205
    },
    {
      "entropy": 5.996038341522217,
      "epoch": 0.7174691951411343,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004955292588127744,
      "loss": 5.6431,
      "mean_token_accuracy": 0.17456454187631607,
      "num_tokens": 16631792.0,
      "step": 8210
    },
    {
      "entropy": 5.979377889633179,
      "epoch": 0.7179061434938391,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000495523075706824,
      "loss": 5.6037,
      "mean_token_accuracy": 0.17521053552627563,
      "num_tokens": 16641655.0,
      "step": 8215
    },
    {
      "entropy": 5.965241956710815,
      "epoch": 0.7183430918465438,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004955168883711257,
      "loss": 5.6983,
      "mean_token_accuracy": 0.16480444073677064,
      "num_tokens": 16651210.0,
      "step": 8220
    },
    {
      "entropy": 6.017394733428955,
      "epoch": 0.7187800401992485,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004955106968057977,
      "loss": 5.6479,
      "mean_token_accuracy": 0.16759678572416306,
      "num_tokens": 16660078.0,
      "step": 8225
    },
    {
      "entropy": 5.967160749435425,
      "epoch": 0.7192169885519532,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004955045010109592,
      "loss": 5.6748,
      "mean_token_accuracy": 0.17211651653051377,
      "num_tokens": 16671211.0,
      "step": 8230
    },
    {
      "entropy": 5.963998603820801,
      "epoch": 0.7196539369046578,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004954983009867286,
      "loss": 5.5941,
      "mean_token_accuracy": 0.17463987320661545,
      "num_tokens": 16680850.0,
      "step": 8235
    },
    {
      "entropy": 5.933062887191772,
      "epoch": 0.7200908852573625,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004954920967332252,
      "loss": 5.5788,
      "mean_token_accuracy": 0.17373492419719697,
      "num_tokens": 16690447.0,
      "step": 8240
    },
    {
      "entropy": 5.925530004501343,
      "epoch": 0.7205278336100673,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004954858882505678,
      "loss": 5.6994,
      "mean_token_accuracy": 0.16813771277666092,
      "num_tokens": 16701061.0,
      "step": 8245
    },
    {
      "entropy": 5.973330020904541,
      "epoch": 0.720964781962772,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004954796755388755,
      "loss": 5.5854,
      "mean_token_accuracy": 0.18063774555921555,
      "num_tokens": 16711139.0,
      "step": 8250
    },
    {
      "entropy": 5.969666957855225,
      "epoch": 0.7214017303154767,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004954734585982675,
      "loss": 5.7272,
      "mean_token_accuracy": 0.16368738412857056,
      "num_tokens": 16722343.0,
      "step": 8255
    },
    {
      "entropy": 5.964052438735962,
      "epoch": 0.7218386786681814,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004954672374288631,
      "loss": 5.7047,
      "mean_token_accuracy": 0.16662293672561646,
      "num_tokens": 16733475.0,
      "step": 8260
    },
    {
      "entropy": 5.977066707611084,
      "epoch": 0.7222756270208861,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004954610120307815,
      "loss": 5.6928,
      "mean_token_accuracy": 0.16628620028495789,
      "num_tokens": 16744356.0,
      "step": 8265
    },
    {
      "entropy": 5.9376500129699705,
      "epoch": 0.7227125753735908,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004954547824041422,
      "loss": 5.7076,
      "mean_token_accuracy": 0.17259007096290588,
      "num_tokens": 16755275.0,
      "step": 8270
    },
    {
      "entropy": 5.91647162437439,
      "epoch": 0.7231495237262956,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004954485485490647,
      "loss": 5.6251,
      "mean_token_accuracy": 0.17619731575250625,
      "num_tokens": 16766114.0,
      "step": 8275
    },
    {
      "entropy": 5.948045015335083,
      "epoch": 0.7235864720790003,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004954423104656685,
      "loss": 5.6184,
      "mean_token_accuracy": 0.176860573887825,
      "num_tokens": 16776391.0,
      "step": 8280
    },
    {
      "entropy": 5.86934118270874,
      "epoch": 0.724023420431705,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004954360681540732,
      "loss": 5.5789,
      "mean_token_accuracy": 0.18258280009031297,
      "num_tokens": 16784845.0,
      "step": 8285
    },
    {
      "entropy": 5.877782821655273,
      "epoch": 0.7244603687844097,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004954298216143986,
      "loss": 5.607,
      "mean_token_accuracy": 0.18136582970619203,
      "num_tokens": 16794735.0,
      "step": 8290
    },
    {
      "entropy": 6.031446743011474,
      "epoch": 0.7248973171371144,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004954235708467647,
      "loss": 5.6934,
      "mean_token_accuracy": 0.17234368175268172,
      "num_tokens": 16803643.0,
      "step": 8295
    },
    {
      "entropy": 5.946221971511841,
      "epoch": 0.725334265489819,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000495417315851291,
      "loss": 5.6829,
      "mean_token_accuracy": 0.1700431853532791,
      "num_tokens": 16813064.0,
      "step": 8300
    },
    {
      "entropy": 5.941778373718262,
      "epoch": 0.7257712138425239,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004954110566280978,
      "loss": 5.6836,
      "mean_token_accuracy": 0.17075045108795167,
      "num_tokens": 16823166.0,
      "step": 8305
    },
    {
      "entropy": 5.993211936950684,
      "epoch": 0.7262081621952285,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000495404793177305,
      "loss": 5.7052,
      "mean_token_accuracy": 0.1682531714439392,
      "num_tokens": 16833546.0,
      "step": 8310
    },
    {
      "entropy": 5.957877254486084,
      "epoch": 0.7266451105479332,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004953985254990328,
      "loss": 5.5657,
      "mean_token_accuracy": 0.1767212361097336,
      "num_tokens": 16843794.0,
      "step": 8315
    },
    {
      "entropy": 5.9451807022094725,
      "epoch": 0.7270820589006379,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004953922535934014,
      "loss": 5.7202,
      "mean_token_accuracy": 0.16943780332803726,
      "num_tokens": 16855677.0,
      "step": 8320
    },
    {
      "entropy": 5.951779365539551,
      "epoch": 0.7275190072533426,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000495385977460531,
      "loss": 5.6544,
      "mean_token_accuracy": 0.1674034848809242,
      "num_tokens": 16866137.0,
      "step": 8325
    },
    {
      "entropy": 5.988984298706055,
      "epoch": 0.7279559556060473,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004953796971005422,
      "loss": 5.5899,
      "mean_token_accuracy": 0.17629415392875672,
      "num_tokens": 16875997.0,
      "step": 8330
    },
    {
      "entropy": 5.83278169631958,
      "epoch": 0.7283929039587521,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004953734125135552,
      "loss": 5.63,
      "mean_token_accuracy": 0.1850111022591591,
      "num_tokens": 16885838.0,
      "step": 8335
    },
    {
      "entropy": 5.874803495407105,
      "epoch": 0.7288298523114568,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004953671236996909,
      "loss": 5.5488,
      "mean_token_accuracy": 0.18516989201307296,
      "num_tokens": 16895533.0,
      "step": 8340
    },
    {
      "entropy": 6.0033347606658936,
      "epoch": 0.7292668006641615,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004953608306590696,
      "loss": 5.6963,
      "mean_token_accuracy": 0.17182930558919907,
      "num_tokens": 16906258.0,
      "step": 8345
    },
    {
      "entropy": 5.959201097488403,
      "epoch": 0.7297037490168662,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000495354533391812,
      "loss": 5.6583,
      "mean_token_accuracy": 0.17517121285200118,
      "num_tokens": 16917123.0,
      "step": 8350
    },
    {
      "entropy": 5.942789506912232,
      "epoch": 0.7301406973695709,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004953482318980389,
      "loss": 5.667,
      "mean_token_accuracy": 0.17443449348211287,
      "num_tokens": 16927533.0,
      "step": 8355
    },
    {
      "entropy": 5.994970703125,
      "epoch": 0.7305776457222757,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004953419261778714,
      "loss": 5.6382,
      "mean_token_accuracy": 0.1767793118953705,
      "num_tokens": 16937763.0,
      "step": 8360
    },
    {
      "entropy": 5.900545072555542,
      "epoch": 0.7310145940749804,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004953356162314304,
      "loss": 5.5725,
      "mean_token_accuracy": 0.1796363353729248,
      "num_tokens": 16947173.0,
      "step": 8365
    },
    {
      "entropy": 5.949201345443726,
      "epoch": 0.7314515424276851,
      "grad_norm": 1.0,
      "learning_rate": 0.0004953293020588368,
      "loss": 5.6707,
      "mean_token_accuracy": 0.17861867547035218,
      "num_tokens": 16956856.0,
      "step": 8370
    },
    {
      "entropy": 5.96291217803955,
      "epoch": 0.7318884907803898,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004953229836602115,
      "loss": 5.6537,
      "mean_token_accuracy": 0.17417480796575546,
      "num_tokens": 16966705.0,
      "step": 8375
    },
    {
      "entropy": 5.932378673553467,
      "epoch": 0.7323254391330944,
      "grad_norm": 0.90625,
      "learning_rate": 0.000495316661035676,
      "loss": 5.6567,
      "mean_token_accuracy": 0.17184317409992217,
      "num_tokens": 16977026.0,
      "step": 8380
    },
    {
      "entropy": 5.929072713851928,
      "epoch": 0.7327623874857991,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004953103341853517,
      "loss": 5.5838,
      "mean_token_accuracy": 0.1785085156559944,
      "num_tokens": 16986800.0,
      "step": 8385
    },
    {
      "entropy": 5.926044464111328,
      "epoch": 0.7331993358385039,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004953040031093596,
      "loss": 5.5587,
      "mean_token_accuracy": 0.17587348371744155,
      "num_tokens": 16996392.0,
      "step": 8390
    },
    {
      "entropy": 5.928092956542969,
      "epoch": 0.7336362841912086,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004952976678078213,
      "loss": 5.6079,
      "mean_token_accuracy": 0.1730113670229912,
      "num_tokens": 17006430.0,
      "step": 8395
    },
    {
      "entropy": 5.9935118675231935,
      "epoch": 0.7340732325439133,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004952913282808584,
      "loss": 5.6585,
      "mean_token_accuracy": 0.16798997819423675,
      "num_tokens": 17017168.0,
      "step": 8400
    },
    {
      "entropy": 5.8380444049835205,
      "epoch": 0.734510180896618,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004952849845285922,
      "loss": 5.5731,
      "mean_token_accuracy": 0.17897841334342957,
      "num_tokens": 17027337.0,
      "step": 8405
    },
    {
      "entropy": 5.930065536499024,
      "epoch": 0.7349471292493227,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004952786365511447,
      "loss": 5.5451,
      "mean_token_accuracy": 0.18138835281133653,
      "num_tokens": 17036839.0,
      "step": 8410
    },
    {
      "entropy": 5.974288272857666,
      "epoch": 0.7353840776020274,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004952722843486375,
      "loss": 5.7503,
      "mean_token_accuracy": 0.16606501787900924,
      "num_tokens": 17048221.0,
      "step": 8415
    },
    {
      "entropy": 5.897845888137818,
      "epoch": 0.7358210259547322,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004952659279211924,
      "loss": 5.5866,
      "mean_token_accuracy": 0.17471111416816712,
      "num_tokens": 17057694.0,
      "step": 8420
    },
    {
      "entropy": 5.954220867156982,
      "epoch": 0.7362579743074369,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004952595672689315,
      "loss": 5.663,
      "mean_token_accuracy": 0.16486438661813735,
      "num_tokens": 17066492.0,
      "step": 8425
    },
    {
      "entropy": 5.986512517929077,
      "epoch": 0.7366949226601416,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004952532023919766,
      "loss": 5.6654,
      "mean_token_accuracy": 0.1708378478884697,
      "num_tokens": 17077211.0,
      "step": 8430
    },
    {
      "entropy": 5.948093557357788,
      "epoch": 0.7371318710128463,
      "grad_norm": 0.875,
      "learning_rate": 0.0004952468332904498,
      "loss": 5.6428,
      "mean_token_accuracy": 0.17409518659114837,
      "num_tokens": 17088665.0,
      "step": 8435
    },
    {
      "entropy": 5.918775844573974,
      "epoch": 0.737568819365551,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004952404599644735,
      "loss": 5.6294,
      "mean_token_accuracy": 0.1793772757053375,
      "num_tokens": 17099767.0,
      "step": 8440
    },
    {
      "entropy": 5.904738998413086,
      "epoch": 0.7380057677182557,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004952340824141698,
      "loss": 5.5441,
      "mean_token_accuracy": 0.18302232027053833,
      "num_tokens": 17109414.0,
      "step": 8445
    },
    {
      "entropy": 5.902750873565674,
      "epoch": 0.7384427160709605,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004952277006396609,
      "loss": 5.6369,
      "mean_token_accuracy": 0.17203447520732879,
      "num_tokens": 17119107.0,
      "step": 8450
    },
    {
      "entropy": 5.952141332626343,
      "epoch": 0.7388796644236652,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004952213146410693,
      "loss": 5.6729,
      "mean_token_accuracy": 0.16804802119731904,
      "num_tokens": 17129010.0,
      "step": 8455
    },
    {
      "entropy": 5.987590789794922,
      "epoch": 0.7393166127763698,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004952149244185174,
      "loss": 5.695,
      "mean_token_accuracy": 0.1697276145219803,
      "num_tokens": 17138951.0,
      "step": 8460
    },
    {
      "entropy": 5.99385986328125,
      "epoch": 0.7397535611290745,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004952085299721279,
      "loss": 5.6194,
      "mean_token_accuracy": 0.17129565328359603,
      "num_tokens": 17149416.0,
      "step": 8465
    },
    {
      "entropy": 5.908122539520264,
      "epoch": 0.7401905094817792,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004952021313020235,
      "loss": 5.5734,
      "mean_token_accuracy": 0.18438582569360734,
      "num_tokens": 17160068.0,
      "step": 8470
    },
    {
      "entropy": 5.9084144115448,
      "epoch": 0.7406274578344839,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004951957284083269,
      "loss": 5.6139,
      "mean_token_accuracy": 0.17089245319366456,
      "num_tokens": 17168972.0,
      "step": 8475
    },
    {
      "entropy": 5.907905721664429,
      "epoch": 0.7410644061871887,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004951893212911607,
      "loss": 5.5991,
      "mean_token_accuracy": 0.1764339029788971,
      "num_tokens": 17178239.0,
      "step": 8480
    },
    {
      "entropy": 5.947232103347778,
      "epoch": 0.7415013545398934,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000495182909950648,
      "loss": 5.7591,
      "mean_token_accuracy": 0.1692235067486763,
      "num_tokens": 17190619.0,
      "step": 8485
    },
    {
      "entropy": 5.875487899780273,
      "epoch": 0.7419383028925981,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004951764943869117,
      "loss": 5.5344,
      "mean_token_accuracy": 0.17977420389652252,
      "num_tokens": 17201041.0,
      "step": 8490
    },
    {
      "entropy": 5.904624509811401,
      "epoch": 0.7423752512453028,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004951700746000749,
      "loss": 5.6612,
      "mean_token_accuracy": 0.16948151737451553,
      "num_tokens": 17211779.0,
      "step": 8495
    },
    {
      "entropy": 5.956327342987061,
      "epoch": 0.7428121995980075,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004951636505902606,
      "loss": 5.614,
      "mean_token_accuracy": 0.1806664362549782,
      "num_tokens": 17221249.0,
      "step": 8500
    },
    {
      "entropy": 5.9729905128479,
      "epoch": 0.7432491479507122,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004951572223575922,
      "loss": 5.6748,
      "mean_token_accuracy": 0.1681697502732277,
      "num_tokens": 17231634.0,
      "step": 8505
    },
    {
      "entropy": 5.934487342834473,
      "epoch": 0.743686096303417,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004951507899021929,
      "loss": 5.6952,
      "mean_token_accuracy": 0.17017512768507004,
      "num_tokens": 17241751.0,
      "step": 8510
    },
    {
      "entropy": 5.9679512023925785,
      "epoch": 0.7441230446561217,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000495144353224186,
      "loss": 5.6665,
      "mean_token_accuracy": 0.17047256082296372,
      "num_tokens": 17251962.0,
      "step": 8515
    },
    {
      "entropy": 5.9460207462310795,
      "epoch": 0.7445599930088264,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004951379123236952,
      "loss": 5.5615,
      "mean_token_accuracy": 0.17725244611501695,
      "num_tokens": 17261895.0,
      "step": 8520
    },
    {
      "entropy": 5.908595371246338,
      "epoch": 0.744996941361531,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004951314672008439,
      "loss": 5.6335,
      "mean_token_accuracy": 0.1761414587497711,
      "num_tokens": 17272025.0,
      "step": 8525
    },
    {
      "entropy": 5.9498762607574465,
      "epoch": 0.7454338897142357,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004951250178557556,
      "loss": 5.6301,
      "mean_token_accuracy": 0.1812431961297989,
      "num_tokens": 17282705.0,
      "step": 8530
    },
    {
      "entropy": 5.943796634674072,
      "epoch": 0.7458708380669404,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004951185642885542,
      "loss": 5.5576,
      "mean_token_accuracy": 0.18232963532209395,
      "num_tokens": 17293045.0,
      "step": 8535
    },
    {
      "entropy": 5.955872583389282,
      "epoch": 0.7463077864196452,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004951121064993633,
      "loss": 5.6676,
      "mean_token_accuracy": 0.16826506704092026,
      "num_tokens": 17303893.0,
      "step": 8540
    },
    {
      "entropy": 5.937839698791504,
      "epoch": 0.7467447347723499,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004951056444883069,
      "loss": 5.6312,
      "mean_token_accuracy": 0.16645151227712632,
      "num_tokens": 17314250.0,
      "step": 8545
    },
    {
      "entropy": 5.912049579620361,
      "epoch": 0.7471816831250546,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000495099178255509,
      "loss": 5.5659,
      "mean_token_accuracy": 0.17440910190343856,
      "num_tokens": 17324268.0,
      "step": 8550
    },
    {
      "entropy": 5.98675570487976,
      "epoch": 0.7476186314777593,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004950927078010934,
      "loss": 5.6871,
      "mean_token_accuracy": 0.16338615119457245,
      "num_tokens": 17334512.0,
      "step": 8555
    },
    {
      "entropy": 5.974498844146728,
      "epoch": 0.748055579830464,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004950862331251845,
      "loss": 5.6237,
      "mean_token_accuracy": 0.17254247963428498,
      "num_tokens": 17346244.0,
      "step": 8560
    },
    {
      "entropy": 5.898865652084351,
      "epoch": 0.7484925281831688,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004950797542279061,
      "loss": 5.6241,
      "mean_token_accuracy": 0.17641711235046387,
      "num_tokens": 17355832.0,
      "step": 8565
    },
    {
      "entropy": 5.944173192977905,
      "epoch": 0.7489294765358735,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004950732711093828,
      "loss": 5.6174,
      "mean_token_accuracy": 0.1755294233560562,
      "num_tokens": 17366871.0,
      "step": 8570
    },
    {
      "entropy": 6.005934429168701,
      "epoch": 0.7493664248885782,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004950667837697388,
      "loss": 5.6947,
      "mean_token_accuracy": 0.16597645729780197,
      "num_tokens": 17377242.0,
      "step": 8575
    },
    {
      "entropy": 5.996667337417603,
      "epoch": 0.7498033732412829,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004950602922090988,
      "loss": 5.6574,
      "mean_token_accuracy": 0.16958146542310715,
      "num_tokens": 17387942.0,
      "step": 8580
    },
    {
      "entropy": 5.963387203216553,
      "epoch": 0.7502403215939876,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004950537964275868,
      "loss": 5.6133,
      "mean_token_accuracy": 0.17067599296569824,
      "num_tokens": 17397413.0,
      "step": 8585
    },
    {
      "entropy": 5.945202732086182,
      "epoch": 0.7506772699466923,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004950472964253278,
      "loss": 5.5736,
      "mean_token_accuracy": 0.1749727711081505,
      "num_tokens": 17408456.0,
      "step": 8590
    },
    {
      "entropy": 5.952192878723144,
      "epoch": 0.7511142182993971,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004950407922024463,
      "loss": 5.5989,
      "mean_token_accuracy": 0.17616791725158693,
      "num_tokens": 17417805.0,
      "step": 8595
    },
    {
      "entropy": 5.9006147384643555,
      "epoch": 0.7515511666521018,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004950342837590671,
      "loss": 5.5354,
      "mean_token_accuracy": 0.18208165913820268,
      "num_tokens": 17426932.0,
      "step": 8600
    },
    {
      "entropy": 5.873894262313843,
      "epoch": 0.7519881150048064,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004950277710953152,
      "loss": 5.6461,
      "mean_token_accuracy": 0.1728186398744583,
      "num_tokens": 17436770.0,
      "step": 8605
    },
    {
      "entropy": 5.922198629379272,
      "epoch": 0.7524250633575111,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004950212542113153,
      "loss": 5.5523,
      "mean_token_accuracy": 0.17483914494514466,
      "num_tokens": 17447852.0,
      "step": 8610
    },
    {
      "entropy": 5.882093095779419,
      "epoch": 0.7528620117102158,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004950147331071924,
      "loss": 5.5556,
      "mean_token_accuracy": 0.17202127277851104,
      "num_tokens": 17458020.0,
      "step": 8615
    },
    {
      "entropy": 5.95059061050415,
      "epoch": 0.7532989600629205,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004950082077830717,
      "loss": 5.6799,
      "mean_token_accuracy": 0.16885688304901122,
      "num_tokens": 17468604.0,
      "step": 8620
    },
    {
      "entropy": 5.948571109771729,
      "epoch": 0.7537359084156253,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004950016782390782,
      "loss": 5.612,
      "mean_token_accuracy": 0.17877668291330337,
      "num_tokens": 17478374.0,
      "step": 8625
    },
    {
      "entropy": 5.9679481983184814,
      "epoch": 0.75417285676833,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004949951444753372,
      "loss": 5.6038,
      "mean_token_accuracy": 0.17430623173713683,
      "num_tokens": 17487894.0,
      "step": 8630
    },
    {
      "entropy": 5.938228893280029,
      "epoch": 0.7546098051210347,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004949886064919741,
      "loss": 5.5789,
      "mean_token_accuracy": 0.17357986122369767,
      "num_tokens": 17498473.0,
      "step": 8635
    },
    {
      "entropy": 5.93488302230835,
      "epoch": 0.7550467534737394,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004949820642891142,
      "loss": 5.6043,
      "mean_token_accuracy": 0.17156960368156432,
      "num_tokens": 17507796.0,
      "step": 8640
    },
    {
      "entropy": 5.8759801387786865,
      "epoch": 0.7554837018264441,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004949755178668832,
      "loss": 5.6144,
      "mean_token_accuracy": 0.17645444422960282,
      "num_tokens": 17517629.0,
      "step": 8645
    },
    {
      "entropy": 5.959586477279663,
      "epoch": 0.7559206501791488,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004949689672254064,
      "loss": 5.6884,
      "mean_token_accuracy": 0.16384072303771974,
      "num_tokens": 17528726.0,
      "step": 8650
    },
    {
      "entropy": 5.936110591888427,
      "epoch": 0.7563575985318536,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004949624123648096,
      "loss": 5.5721,
      "mean_token_accuracy": 0.17706723511219025,
      "num_tokens": 17539022.0,
      "step": 8655
    },
    {
      "entropy": 5.958128595352173,
      "epoch": 0.7567945468845583,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004949558532852185,
      "loss": 5.6228,
      "mean_token_accuracy": 0.17900796085596085,
      "num_tokens": 17549293.0,
      "step": 8660
    },
    {
      "entropy": 5.94136905670166,
      "epoch": 0.757231495237263,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004949492899867587,
      "loss": 5.609,
      "mean_token_accuracy": 0.1712702065706253,
      "num_tokens": 17559072.0,
      "step": 8665
    },
    {
      "entropy": 5.922577810287476,
      "epoch": 0.7576684435899677,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004949427224695564,
      "loss": 5.5939,
      "mean_token_accuracy": 0.17388864010572433,
      "num_tokens": 17568757.0,
      "step": 8670
    },
    {
      "entropy": 5.957227945327759,
      "epoch": 0.7581053919426723,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004949361507337374,
      "loss": 5.5858,
      "mean_token_accuracy": 0.1751673549413681,
      "num_tokens": 17578288.0,
      "step": 8675
    },
    {
      "entropy": 5.928276538848877,
      "epoch": 0.758542340295377,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004949295747794277,
      "loss": 5.692,
      "mean_token_accuracy": 0.17010658532381057,
      "num_tokens": 17588864.0,
      "step": 8680
    },
    {
      "entropy": 5.931891679763794,
      "epoch": 0.7589792886480818,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004949229946067537,
      "loss": 5.711,
      "mean_token_accuracy": 0.16884150207042695,
      "num_tokens": 17599103.0,
      "step": 8685
    },
    {
      "entropy": 5.8696009635925295,
      "epoch": 0.7594162370007865,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004949164102158413,
      "loss": 5.5912,
      "mean_token_accuracy": 0.18062739074230194,
      "num_tokens": 17608608.0,
      "step": 8690
    },
    {
      "entropy": 5.990211439132691,
      "epoch": 0.7598531853534912,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000494909821606817,
      "loss": 5.6626,
      "mean_token_accuracy": 0.17643883526325227,
      "num_tokens": 17618604.0,
      "step": 8695
    },
    {
      "entropy": 6.002125072479248,
      "epoch": 0.7602901337061959,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004949032287798069,
      "loss": 5.6283,
      "mean_token_accuracy": 0.17779159247875215,
      "num_tokens": 17628479.0,
      "step": 8700
    },
    {
      "entropy": 5.914983654022217,
      "epoch": 0.7607270820589006,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004948966317349379,
      "loss": 5.6312,
      "mean_token_accuracy": 0.1803823009133339,
      "num_tokens": 17637519.0,
      "step": 8705
    },
    {
      "entropy": 5.926572465896607,
      "epoch": 0.7611640304116053,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004948900304723361,
      "loss": 5.5732,
      "mean_token_accuracy": 0.18241428434848786,
      "num_tokens": 17647902.0,
      "step": 8710
    },
    {
      "entropy": 5.966745138168335,
      "epoch": 0.7616009787643101,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004948834249921285,
      "loss": 5.5945,
      "mean_token_accuracy": 0.17514690458774568,
      "num_tokens": 17658576.0,
      "step": 8715
    },
    {
      "entropy": 5.9518383026123045,
      "epoch": 0.7620379271170148,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004948768152944414,
      "loss": 5.6592,
      "mean_token_accuracy": 0.1725882187485695,
      "num_tokens": 17667059.0,
      "step": 8720
    },
    {
      "entropy": 6.010783100128174,
      "epoch": 0.7624748754697195,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004948702013794018,
      "loss": 5.7433,
      "mean_token_accuracy": 0.16454955711960792,
      "num_tokens": 17677447.0,
      "step": 8725
    },
    {
      "entropy": 5.924250936508178,
      "epoch": 0.7629118238224242,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004948635832471366,
      "loss": 5.626,
      "mean_token_accuracy": 0.18001531362533568,
      "num_tokens": 17687893.0,
      "step": 8730
    },
    {
      "entropy": 5.90333399772644,
      "epoch": 0.7633487721751289,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004948569608977726,
      "loss": 5.6307,
      "mean_token_accuracy": 0.1783311888575554,
      "num_tokens": 17698599.0,
      "step": 8735
    },
    {
      "entropy": 5.98647632598877,
      "epoch": 0.7637857205278336,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004948503343314369,
      "loss": 5.6472,
      "mean_token_accuracy": 0.171012781560421,
      "num_tokens": 17709711.0,
      "step": 8740
    },
    {
      "entropy": 5.943402576446533,
      "epoch": 0.7642226688805384,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004948437035482566,
      "loss": 5.6148,
      "mean_token_accuracy": 0.17859706580638884,
      "num_tokens": 17719613.0,
      "step": 8745
    },
    {
      "entropy": 5.937295436859131,
      "epoch": 0.764659617233243,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004948370685483588,
      "loss": 5.5922,
      "mean_token_accuracy": 0.18119112998247147,
      "num_tokens": 17729829.0,
      "step": 8750
    },
    {
      "entropy": 5.931298923492432,
      "epoch": 0.7650965655859477,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004948304293318709,
      "loss": 5.5968,
      "mean_token_accuracy": 0.17523320615291596,
      "num_tokens": 17739254.0,
      "step": 8755
    },
    {
      "entropy": 5.941666460037231,
      "epoch": 0.7655335139386524,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004948237858989201,
      "loss": 5.6228,
      "mean_token_accuracy": 0.17343551069498062,
      "num_tokens": 17749294.0,
      "step": 8760
    },
    {
      "entropy": 5.900462532043457,
      "epoch": 0.7659704622913571,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000494817138249634,
      "loss": 5.5484,
      "mean_token_accuracy": 0.180706587433815,
      "num_tokens": 17758866.0,
      "step": 8765
    },
    {
      "entropy": 6.026937580108642,
      "epoch": 0.7664074106440619,
      "grad_norm": 1.0,
      "learning_rate": 0.00049481048638414,
      "loss": 5.7098,
      "mean_token_accuracy": 0.1623545467853546,
      "num_tokens": 17769202.0,
      "step": 8770
    },
    {
      "entropy": 5.916517448425293,
      "epoch": 0.7668443589967666,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004948038303025657,
      "loss": 5.6888,
      "mean_token_accuracy": 0.1753716841340065,
      "num_tokens": 17779125.0,
      "step": 8775
    },
    {
      "entropy": 5.941812896728516,
      "epoch": 0.7672813073494713,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004947971700050387,
      "loss": 5.6423,
      "mean_token_accuracy": 0.17530664205551147,
      "num_tokens": 17788696.0,
      "step": 8780
    },
    {
      "entropy": 5.934955692291259,
      "epoch": 0.767718255702176,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004947905054916869,
      "loss": 5.6186,
      "mean_token_accuracy": 0.17460120618343353,
      "num_tokens": 17799867.0,
      "step": 8785
    },
    {
      "entropy": 6.032731819152832,
      "epoch": 0.7681552040548807,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004947838367626379,
      "loss": 5.6407,
      "mean_token_accuracy": 0.17032813280820847,
      "num_tokens": 17809418.0,
      "step": 8790
    },
    {
      "entropy": 5.888445949554443,
      "epoch": 0.7685921524075854,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004947771638180198,
      "loss": 5.5351,
      "mean_token_accuracy": 0.18036646842956544,
      "num_tokens": 17818554.0,
      "step": 8795
    },
    {
      "entropy": 5.854133081436157,
      "epoch": 0.7690291007602902,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004947704866579605,
      "loss": 5.5719,
      "mean_token_accuracy": 0.18015825301408767,
      "num_tokens": 17828918.0,
      "step": 8800
    },
    {
      "entropy": 5.988088464736938,
      "epoch": 0.7694660491129949,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004947638052825883,
      "loss": 5.729,
      "mean_token_accuracy": 0.1645599514245987,
      "num_tokens": 17840791.0,
      "step": 8805
    },
    {
      "entropy": 6.005149841308594,
      "epoch": 0.7699029974656996,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004947571196920309,
      "loss": 5.6227,
      "mean_token_accuracy": 0.16883825063705443,
      "num_tokens": 17850675.0,
      "step": 8810
    },
    {
      "entropy": 5.8768534660339355,
      "epoch": 0.7703399458184043,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004947504298864169,
      "loss": 5.6024,
      "mean_token_accuracy": 0.18053377121686937,
      "num_tokens": 17860684.0,
      "step": 8815
    },
    {
      "entropy": 5.900746393203735,
      "epoch": 0.770776894171109,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004947437358658746,
      "loss": 5.6197,
      "mean_token_accuracy": 0.16997503042221068,
      "num_tokens": 17870394.0,
      "step": 8820
    },
    {
      "entropy": 5.914662027359009,
      "epoch": 0.7712138425238136,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004947370376305321,
      "loss": 5.5876,
      "mean_token_accuracy": 0.17377581298351288,
      "num_tokens": 17881256.0,
      "step": 8825
    },
    {
      "entropy": 5.928634405136108,
      "epoch": 0.7716507908765184,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004947303351805183,
      "loss": 5.6264,
      "mean_token_accuracy": 0.17449136972427368,
      "num_tokens": 17890888.0,
      "step": 8830
    },
    {
      "entropy": 5.9774586200714115,
      "epoch": 0.7720877392292231,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004947236285159615,
      "loss": 5.5973,
      "mean_token_accuracy": 0.1747547447681427,
      "num_tokens": 17900507.0,
      "step": 8835
    },
    {
      "entropy": 5.9315934658050535,
      "epoch": 0.7725246875819278,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004947169176369902,
      "loss": 5.6733,
      "mean_token_accuracy": 0.17690840661525725,
      "num_tokens": 17911333.0,
      "step": 8840
    },
    {
      "entropy": 5.989700031280518,
      "epoch": 0.7729616359346325,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004947102025437334,
      "loss": 5.6118,
      "mean_token_accuracy": 0.17218692004680633,
      "num_tokens": 17921540.0,
      "step": 8845
    },
    {
      "entropy": 5.963503742218018,
      "epoch": 0.7733985842873372,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004947034832363199,
      "loss": 5.7084,
      "mean_token_accuracy": 0.16823608428239822,
      "num_tokens": 17931820.0,
      "step": 8850
    },
    {
      "entropy": 5.9549096584320065,
      "epoch": 0.7738355326400419,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004946967597148782,
      "loss": 5.6185,
      "mean_token_accuracy": 0.17164172381162643,
      "num_tokens": 17942717.0,
      "step": 8855
    },
    {
      "entropy": 5.958958578109741,
      "epoch": 0.7742724809927467,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004946900319795377,
      "loss": 5.6091,
      "mean_token_accuracy": 0.18062109351158143,
      "num_tokens": 17953775.0,
      "step": 8860
    },
    {
      "entropy": 5.956230211257934,
      "epoch": 0.7747094293454514,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004946833000304274,
      "loss": 5.5882,
      "mean_token_accuracy": 0.1748532995581627,
      "num_tokens": 17963877.0,
      "step": 8865
    },
    {
      "entropy": 5.931340122222901,
      "epoch": 0.7751463776981561,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004946765638676761,
      "loss": 5.5973,
      "mean_token_accuracy": 0.17589059621095657,
      "num_tokens": 17974015.0,
      "step": 8870
    },
    {
      "entropy": 5.91488208770752,
      "epoch": 0.7755833260508608,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004946698234914132,
      "loss": 5.6097,
      "mean_token_accuracy": 0.17845676094293594,
      "num_tokens": 17984189.0,
      "step": 8875
    },
    {
      "entropy": 5.912623023986816,
      "epoch": 0.7760202744035655,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000494663078901768,
      "loss": 5.5632,
      "mean_token_accuracy": 0.1845254510641098,
      "num_tokens": 17994615.0,
      "step": 8880
    },
    {
      "entropy": 5.959689283370972,
      "epoch": 0.7764572227562702,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004946563300988699,
      "loss": 5.6057,
      "mean_token_accuracy": 0.17086736112833023,
      "num_tokens": 18004727.0,
      "step": 8885
    },
    {
      "entropy": 5.928523778915405,
      "epoch": 0.776894171108975,
      "grad_norm": 0.875,
      "learning_rate": 0.0004946495770828482,
      "loss": 5.6753,
      "mean_token_accuracy": 0.17195741981267929,
      "num_tokens": 18014576.0,
      "step": 8890
    },
    {
      "entropy": 5.914179468154908,
      "epoch": 0.7773311194616797,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004946428198538326,
      "loss": 5.6565,
      "mean_token_accuracy": 0.17667228281497954,
      "num_tokens": 18024709.0,
      "step": 8895
    },
    {
      "entropy": 6.010562515258789,
      "epoch": 0.7777680678143843,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004946360584119526,
      "loss": 5.5896,
      "mean_token_accuracy": 0.17290861606597902,
      "num_tokens": 18034673.0,
      "step": 8900
    },
    {
      "entropy": 5.941742753982544,
      "epoch": 0.778205016167089,
      "grad_norm": 0.890625,
      "learning_rate": 0.000494629292757338,
      "loss": 5.5893,
      "mean_token_accuracy": 0.1752185344696045,
      "num_tokens": 18044774.0,
      "step": 8905
    },
    {
      "entropy": 5.887195730209351,
      "epoch": 0.7786419645197937,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004946225228901184,
      "loss": 5.5975,
      "mean_token_accuracy": 0.1718719258904457,
      "num_tokens": 18055244.0,
      "step": 8910
    },
    {
      "entropy": 5.95854959487915,
      "epoch": 0.7790789128724984,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004946157488104236,
      "loss": 5.6827,
      "mean_token_accuracy": 0.1662286639213562,
      "num_tokens": 18066051.0,
      "step": 8915
    },
    {
      "entropy": 5.976449489593506,
      "epoch": 0.7795158612252032,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004946089705183839,
      "loss": 5.7001,
      "mean_token_accuracy": 0.16312077790498733,
      "num_tokens": 18077693.0,
      "step": 8920
    },
    {
      "entropy": 5.961237621307373,
      "epoch": 0.7799528095779079,
      "grad_norm": 0.875,
      "learning_rate": 0.0004946021880141289,
      "loss": 5.6253,
      "mean_token_accuracy": 0.17560173124074935,
      "num_tokens": 18088032.0,
      "step": 8925
    },
    {
      "entropy": 5.952757263183594,
      "epoch": 0.7803897579306126,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000494595401297789,
      "loss": 5.5901,
      "mean_token_accuracy": 0.1764475256204605,
      "num_tokens": 18097998.0,
      "step": 8930
    },
    {
      "entropy": 5.917514276504517,
      "epoch": 0.7808267062833173,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004945886103694941,
      "loss": 5.5478,
      "mean_token_accuracy": 0.1780862033367157,
      "num_tokens": 18107538.0,
      "step": 8935
    },
    {
      "entropy": 5.93975248336792,
      "epoch": 0.781263654636022,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004945818152293747,
      "loss": 5.6417,
      "mean_token_accuracy": 0.1771465867757797,
      "num_tokens": 18117792.0,
      "step": 8940
    },
    {
      "entropy": 5.96364459991455,
      "epoch": 0.7817006029887267,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004945750158775609,
      "loss": 5.6456,
      "mean_token_accuracy": 0.17407113909721375,
      "num_tokens": 18128005.0,
      "step": 8945
    },
    {
      "entropy": 5.90098876953125,
      "epoch": 0.7821375513414315,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004945682123141833,
      "loss": 5.6165,
      "mean_token_accuracy": 0.17737158089876176,
      "num_tokens": 18138398.0,
      "step": 8950
    },
    {
      "entropy": 5.969440984725952,
      "epoch": 0.7825744996941362,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004945614045393724,
      "loss": 5.5926,
      "mean_token_accuracy": 0.17322202473878862,
      "num_tokens": 18147676.0,
      "step": 8955
    },
    {
      "entropy": 5.838773965835571,
      "epoch": 0.7830114480468409,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004945545925532587,
      "loss": 5.5149,
      "mean_token_accuracy": 0.18581560254096985,
      "num_tokens": 18156842.0,
      "step": 8960
    },
    {
      "entropy": 5.885096263885498,
      "epoch": 0.7834483963995456,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004945477763559728,
      "loss": 5.6396,
      "mean_token_accuracy": 0.1713474690914154,
      "num_tokens": 18166902.0,
      "step": 8965
    },
    {
      "entropy": 5.947230291366577,
      "epoch": 0.7838853447522502,
      "grad_norm": 1.0,
      "learning_rate": 0.0004945409559476455,
      "loss": 5.6073,
      "mean_token_accuracy": 0.17449768632650375,
      "num_tokens": 18177561.0,
      "step": 8970
    },
    {
      "entropy": 5.9118547439575195,
      "epoch": 0.784322293104955,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004945341313284077,
      "loss": 5.5968,
      "mean_token_accuracy": 0.18023753464221953,
      "num_tokens": 18187739.0,
      "step": 8975
    },
    {
      "entropy": 5.895757436752319,
      "epoch": 0.7847592414576597,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004945273024983902,
      "loss": 5.531,
      "mean_token_accuracy": 0.17939130663871766,
      "num_tokens": 18198435.0,
      "step": 8980
    },
    {
      "entropy": 5.935294914245605,
      "epoch": 0.7851961898103644,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000494520469457724,
      "loss": 5.5432,
      "mean_token_accuracy": 0.18033124804496764,
      "num_tokens": 18208374.0,
      "step": 8985
    },
    {
      "entropy": 5.851685857772827,
      "epoch": 0.7856331381630691,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004945136322065402,
      "loss": 5.6233,
      "mean_token_accuracy": 0.17926690131425857,
      "num_tokens": 18219229.0,
      "step": 8990
    },
    {
      "entropy": 5.921083879470825,
      "epoch": 0.7860700865157738,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004945067907449699,
      "loss": 5.5727,
      "mean_token_accuracy": 0.17503368258476257,
      "num_tokens": 18229208.0,
      "step": 8995
    },
    {
      "entropy": 6.04007773399353,
      "epoch": 0.7865070348684785,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004944999450731442,
      "loss": 5.7214,
      "mean_token_accuracy": 0.16046712398529053,
      "num_tokens": 18238454.0,
      "step": 9000
    },
    {
      "epoch": 0.7865070348684785,
      "eval_entropy": 5.799095045651992,
      "eval_loss": 5.61815881729126,
      "eval_mean_token_accuracy": 0.1829171128774729,
      "eval_num_tokens": 18238454.0,
      "eval_runtime": 21.4388,
      "eval_samples_per_second": 1726.492,
      "eval_steps_per_second": 215.823,
      "step": 9000
    },
    {
      "entropy": 5.9690662860870365,
      "epoch": 0.7869439832211833,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004944930951911947,
      "loss": 5.5945,
      "mean_token_accuracy": 0.1766627162694931,
      "num_tokens": 18248845.0,
      "step": 9005
    },
    {
      "entropy": 5.979356288909912,
      "epoch": 0.787380931573888,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004944862410992526,
      "loss": 5.6091,
      "mean_token_accuracy": 0.17292969673871994,
      "num_tokens": 18258912.0,
      "step": 9010
    },
    {
      "entropy": 5.946245431900024,
      "epoch": 0.7878178799265927,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004944793827974494,
      "loss": 5.6417,
      "mean_token_accuracy": 0.16844676285982133,
      "num_tokens": 18269330.0,
      "step": 9015
    },
    {
      "entropy": 5.932359981536865,
      "epoch": 0.7882548282792974,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004944725202859165,
      "loss": 5.6259,
      "mean_token_accuracy": 0.17528000622987747,
      "num_tokens": 18279721.0,
      "step": 9020
    },
    {
      "entropy": 5.88718090057373,
      "epoch": 0.7886917766320021,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004944656535647858,
      "loss": 5.5506,
      "mean_token_accuracy": 0.18009503781795502,
      "num_tokens": 18288581.0,
      "step": 9025
    },
    {
      "entropy": 5.902565622329712,
      "epoch": 0.7891287249847068,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004944587826341888,
      "loss": 5.6054,
      "mean_token_accuracy": 0.1746378481388092,
      "num_tokens": 18298179.0,
      "step": 9030
    },
    {
      "entropy": 5.961957025527954,
      "epoch": 0.7895656733374116,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004944519074942574,
      "loss": 5.6087,
      "mean_token_accuracy": 0.17048639208078384,
      "num_tokens": 18307957.0,
      "step": 9035
    },
    {
      "entropy": 5.842776727676392,
      "epoch": 0.7900026216901163,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004944450281451234,
      "loss": 5.5813,
      "mean_token_accuracy": 0.17345141768455505,
      "num_tokens": 18317799.0,
      "step": 9040
    },
    {
      "entropy": 5.9876110553741455,
      "epoch": 0.790439570042821,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004944381445869188,
      "loss": 5.6453,
      "mean_token_accuracy": 0.17759709656238556,
      "num_tokens": 18327889.0,
      "step": 9045
    },
    {
      "entropy": 5.995506715774536,
      "epoch": 0.7908765183955256,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004944312568197756,
      "loss": 5.6178,
      "mean_token_accuracy": 0.18062808215618134,
      "num_tokens": 18338288.0,
      "step": 9050
    },
    {
      "entropy": 5.91752347946167,
      "epoch": 0.7913134667482303,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004944243648438258,
      "loss": 5.6735,
      "mean_token_accuracy": 0.16826575249433517,
      "num_tokens": 18349070.0,
      "step": 9055
    },
    {
      "entropy": 5.99967098236084,
      "epoch": 0.791750415100935,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004944174686592019,
      "loss": 5.5919,
      "mean_token_accuracy": 0.1786806493997574,
      "num_tokens": 18359683.0,
      "step": 9060
    },
    {
      "entropy": 5.974533224105835,
      "epoch": 0.7921873634536398,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004944105682660358,
      "loss": 5.5815,
      "mean_token_accuracy": 0.17683615982532502,
      "num_tokens": 18371355.0,
      "step": 9065
    },
    {
      "entropy": 5.7761462211608885,
      "epoch": 0.7926243118063445,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004944036636644601,
      "loss": 5.6231,
      "mean_token_accuracy": 0.17879786640405654,
      "num_tokens": 18381591.0,
      "step": 9070
    },
    {
      "entropy": 5.961842966079712,
      "epoch": 0.7930612601590492,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004943967548546071,
      "loss": 5.6018,
      "mean_token_accuracy": 0.17397417575120927,
      "num_tokens": 18390760.0,
      "step": 9075
    },
    {
      "entropy": 5.976833868026733,
      "epoch": 0.7934982085117539,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004943898418366096,
      "loss": 5.6516,
      "mean_token_accuracy": 0.16976188570261003,
      "num_tokens": 18400643.0,
      "step": 9080
    },
    {
      "entropy": 5.9301385402679445,
      "epoch": 0.7939351568644586,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004943829246105997,
      "loss": 5.6799,
      "mean_token_accuracy": 0.17168610244989396,
      "num_tokens": 18411559.0,
      "step": 9085
    },
    {
      "entropy": 5.945601987838745,
      "epoch": 0.7943721052171633,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004943760031767106,
      "loss": 5.6218,
      "mean_token_accuracy": 0.17511293292045593,
      "num_tokens": 18422645.0,
      "step": 9090
    },
    {
      "entropy": 5.955140399932861,
      "epoch": 0.7948090535698681,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004943690775350746,
      "loss": 5.5973,
      "mean_token_accuracy": 0.17491759359836578,
      "num_tokens": 18432369.0,
      "step": 9095
    },
    {
      "entropy": 5.936253213882447,
      "epoch": 0.7952460019225728,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004943621476858249,
      "loss": 5.56,
      "mean_token_accuracy": 0.18125092685222627,
      "num_tokens": 18442143.0,
      "step": 9100
    },
    {
      "entropy": 5.91200737953186,
      "epoch": 0.7956829502752775,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004943552136290942,
      "loss": 5.6122,
      "mean_token_accuracy": 0.1770312026143074,
      "num_tokens": 18452161.0,
      "step": 9105
    },
    {
      "entropy": 5.962855434417724,
      "epoch": 0.7961198986279822,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004943482753650154,
      "loss": 5.71,
      "mean_token_accuracy": 0.171943961083889,
      "num_tokens": 18461873.0,
      "step": 9110
    },
    {
      "entropy": 5.940937089920044,
      "epoch": 0.7965568469806869,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004943413328937221,
      "loss": 5.5971,
      "mean_token_accuracy": 0.17751961052417756,
      "num_tokens": 18472584.0,
      "step": 9115
    },
    {
      "entropy": 5.9574967384338375,
      "epoch": 0.7969937953333915,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004943343862153467,
      "loss": 5.549,
      "mean_token_accuracy": 0.17541657537221908,
      "num_tokens": 18482174.0,
      "step": 9120
    },
    {
      "entropy": 5.973337268829345,
      "epoch": 0.7974307436860963,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004943274353300231,
      "loss": 5.7211,
      "mean_token_accuracy": 0.16992397606372833,
      "num_tokens": 18492861.0,
      "step": 9125
    },
    {
      "entropy": 5.991778421401977,
      "epoch": 0.797867692038801,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004943204802378843,
      "loss": 5.6055,
      "mean_token_accuracy": 0.17771029323339463,
      "num_tokens": 18502830.0,
      "step": 9130
    },
    {
      "entropy": 5.965285396575927,
      "epoch": 0.7983046403915057,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004943135209390638,
      "loss": 5.6094,
      "mean_token_accuracy": 0.17244573086500167,
      "num_tokens": 18513063.0,
      "step": 9135
    },
    {
      "entropy": 5.870566463470459,
      "epoch": 0.7987415887442104,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004943065574336949,
      "loss": 5.5935,
      "mean_token_accuracy": 0.18023619502782823,
      "num_tokens": 18523773.0,
      "step": 9140
    },
    {
      "entropy": 5.953956270217896,
      "epoch": 0.7991785370969151,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004942995897219115,
      "loss": 5.6109,
      "mean_token_accuracy": 0.1785210371017456,
      "num_tokens": 18533101.0,
      "step": 9145
    },
    {
      "entropy": 5.976411914825439,
      "epoch": 0.7996154854496198,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004942926178038469,
      "loss": 5.6388,
      "mean_token_accuracy": 0.17758730351924895,
      "num_tokens": 18543863.0,
      "step": 9150
    },
    {
      "entropy": 5.905736923217773,
      "epoch": 0.8000524338023246,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004942856416796352,
      "loss": 5.6306,
      "mean_token_accuracy": 0.17494232654571534,
      "num_tokens": 18554249.0,
      "step": 9155
    },
    {
      "entropy": 5.919330739974976,
      "epoch": 0.8004893821550293,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004942786613494098,
      "loss": 5.5982,
      "mean_token_accuracy": 0.17685798406600953,
      "num_tokens": 18563525.0,
      "step": 9160
    },
    {
      "entropy": 5.933303117752075,
      "epoch": 0.800926330507734,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004942716768133048,
      "loss": 5.5885,
      "mean_token_accuracy": 0.17487515807151793,
      "num_tokens": 18572734.0,
      "step": 9165
    },
    {
      "entropy": 5.8907208919525145,
      "epoch": 0.8013632788604387,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004942646880714543,
      "loss": 5.651,
      "mean_token_accuracy": 0.17267553210258485,
      "num_tokens": 18581939.0,
      "step": 9170
    },
    {
      "entropy": 5.963696718215942,
      "epoch": 0.8018002272131434,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000494257695123992,
      "loss": 5.6388,
      "mean_token_accuracy": 0.17973453998565675,
      "num_tokens": 18591635.0,
      "step": 9175
    },
    {
      "entropy": 5.932835388183594,
      "epoch": 0.8022371755658482,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004942506979710524,
      "loss": 5.5329,
      "mean_token_accuracy": 0.17816112488508223,
      "num_tokens": 18602060.0,
      "step": 9180
    },
    {
      "entropy": 5.9098142147064205,
      "epoch": 0.8026741239185529,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004942436966127693,
      "loss": 5.5788,
      "mean_token_accuracy": 0.17823428064584732,
      "num_tokens": 18611187.0,
      "step": 9185
    },
    {
      "entropy": 5.932813262939453,
      "epoch": 0.8031110722712576,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004942366910492775,
      "loss": 5.5216,
      "mean_token_accuracy": 0.18350881487131118,
      "num_tokens": 18620329.0,
      "step": 9190
    },
    {
      "entropy": 5.962358045578003,
      "epoch": 0.8035480206239622,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004942296812807109,
      "loss": 5.6116,
      "mean_token_accuracy": 0.1757429212331772,
      "num_tokens": 18629600.0,
      "step": 9195
    },
    {
      "entropy": 5.938585519790649,
      "epoch": 0.8039849689766669,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004942226673072041,
      "loss": 5.6712,
      "mean_token_accuracy": 0.17116753458976747,
      "num_tokens": 18640771.0,
      "step": 9200
    },
    {
      "entropy": 5.886959505081177,
      "epoch": 0.8044219173293716,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004942156491288917,
      "loss": 5.5673,
      "mean_token_accuracy": 0.1778600424528122,
      "num_tokens": 18650606.0,
      "step": 9205
    },
    {
      "entropy": 5.935275506973267,
      "epoch": 0.8048588656820764,
      "grad_norm": 1.0,
      "learning_rate": 0.0004942086267459083,
      "loss": 5.6165,
      "mean_token_accuracy": 0.1745525613427162,
      "num_tokens": 18660487.0,
      "step": 9210
    },
    {
      "entropy": 5.928958177566528,
      "epoch": 0.8052958140347811,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004942016001583886,
      "loss": 5.6057,
      "mean_token_accuracy": 0.17327710539102553,
      "num_tokens": 18670987.0,
      "step": 9215
    },
    {
      "entropy": 5.925167226791382,
      "epoch": 0.8057327623874858,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004941945693664674,
      "loss": 5.6766,
      "mean_token_accuracy": 0.17558243423700332,
      "num_tokens": 18682560.0,
      "step": 9220
    },
    {
      "entropy": 5.987202215194702,
      "epoch": 0.8061697107401905,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004941875343702795,
      "loss": 5.6185,
      "mean_token_accuracy": 0.17353916466236113,
      "num_tokens": 18692743.0,
      "step": 9225
    },
    {
      "entropy": 5.914751863479614,
      "epoch": 0.8066066590928952,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004941804951699597,
      "loss": 5.5924,
      "mean_token_accuracy": 0.1785692796111107,
      "num_tokens": 18703039.0,
      "step": 9230
    },
    {
      "entropy": 5.8650001049041744,
      "epoch": 0.8070436074455999,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004941734517656432,
      "loss": 5.6079,
      "mean_token_accuracy": 0.17731017321348191,
      "num_tokens": 18713777.0,
      "step": 9235
    },
    {
      "entropy": 5.937385320663452,
      "epoch": 0.8074805557983047,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004941664041574651,
      "loss": 5.6354,
      "mean_token_accuracy": 0.17793867290019988,
      "num_tokens": 18723422.0,
      "step": 9240
    },
    {
      "entropy": 5.956562900543213,
      "epoch": 0.8079175041510094,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004941593523455605,
      "loss": 5.5627,
      "mean_token_accuracy": 0.1763332635164261,
      "num_tokens": 18734369.0,
      "step": 9245
    },
    {
      "entropy": 5.982712888717652,
      "epoch": 0.8083544525037141,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004941522963300647,
      "loss": 5.6363,
      "mean_token_accuracy": 0.17443077713251115,
      "num_tokens": 18744647.0,
      "step": 9250
    },
    {
      "entropy": 5.977589130401611,
      "epoch": 0.8087914008564188,
      "grad_norm": 1.015625,
      "learning_rate": 0.000494145236111113,
      "loss": 5.5846,
      "mean_token_accuracy": 0.1750393584370613,
      "num_tokens": 18753929.0,
      "step": 9255
    },
    {
      "entropy": 5.9497112274169925,
      "epoch": 0.8092283492091235,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004941381716888409,
      "loss": 5.6118,
      "mean_token_accuracy": 0.17641770094633102,
      "num_tokens": 18764103.0,
      "step": 9260
    },
    {
      "entropy": 5.978851366043091,
      "epoch": 0.8096652975618281,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004941311030633838,
      "loss": 5.6436,
      "mean_token_accuracy": 0.17244498878717424,
      "num_tokens": 18773756.0,
      "step": 9265
    },
    {
      "entropy": 5.898691606521607,
      "epoch": 0.810102245914533,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004941240302348773,
      "loss": 5.5,
      "mean_token_accuracy": 0.19330843836069106,
      "num_tokens": 18783778.0,
      "step": 9270
    },
    {
      "entropy": 5.917062091827392,
      "epoch": 0.8105391942672376,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004941169532034571,
      "loss": 5.6237,
      "mean_token_accuracy": 0.17767008692026137,
      "num_tokens": 18794664.0,
      "step": 9275
    },
    {
      "entropy": 5.913171100616455,
      "epoch": 0.8109761426199423,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004941098719692589,
      "loss": 5.5922,
      "mean_token_accuracy": 0.17082895636558532,
      "num_tokens": 18803773.0,
      "step": 9280
    },
    {
      "entropy": 5.8794793605804445,
      "epoch": 0.811413090972647,
      "grad_norm": 1.0,
      "learning_rate": 0.0004941027865324186,
      "loss": 5.6503,
      "mean_token_accuracy": 0.1752942144870758,
      "num_tokens": 18814419.0,
      "step": 9285
    },
    {
      "entropy": 6.010300731658935,
      "epoch": 0.8118500393253517,
      "grad_norm": 0.921875,
      "learning_rate": 0.000494095696893072,
      "loss": 5.6018,
      "mean_token_accuracy": 0.1744083970785141,
      "num_tokens": 18825011.0,
      "step": 9290
    },
    {
      "entropy": 5.914761400222778,
      "epoch": 0.8122869876780564,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004940886030513551,
      "loss": 5.52,
      "mean_token_accuracy": 0.1820392206311226,
      "num_tokens": 18835003.0,
      "step": 9295
    },
    {
      "entropy": 5.876133394241333,
      "epoch": 0.8127239360307612,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004940815050074041,
      "loss": 5.6108,
      "mean_token_accuracy": 0.17587236911058426,
      "num_tokens": 18845072.0,
      "step": 9300
    },
    {
      "entropy": 5.992804908752442,
      "epoch": 0.8131608843834659,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004940744027613549,
      "loss": 5.605,
      "mean_token_accuracy": 0.17620026618242263,
      "num_tokens": 18854583.0,
      "step": 9305
    },
    {
      "entropy": 5.983105134963989,
      "epoch": 0.8135978327361706,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004940672963133439,
      "loss": 5.6138,
      "mean_token_accuracy": 0.17938638031482695,
      "num_tokens": 18865460.0,
      "step": 9310
    },
    {
      "entropy": 5.8887717723846436,
      "epoch": 0.8140347810888753,
      "grad_norm": 1.0,
      "learning_rate": 0.0004940601856635074,
      "loss": 5.619,
      "mean_token_accuracy": 0.1770030990242958,
      "num_tokens": 18875568.0,
      "step": 9315
    },
    {
      "entropy": 5.855496025085449,
      "epoch": 0.81447172944158,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004940530708119818,
      "loss": 5.5224,
      "mean_token_accuracy": 0.18190036863088607,
      "num_tokens": 18884894.0,
      "step": 9320
    },
    {
      "entropy": 5.913312244415283,
      "epoch": 0.8149086777942847,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004940459517589034,
      "loss": 5.5743,
      "mean_token_accuracy": 0.17621856480836867,
      "num_tokens": 18895353.0,
      "step": 9325
    },
    {
      "entropy": 5.921319198608399,
      "epoch": 0.8153456261469895,
      "grad_norm": 1.015625,
      "learning_rate": 0.000494038828504409,
      "loss": 5.6143,
      "mean_token_accuracy": 0.1800361305475235,
      "num_tokens": 18905474.0,
      "step": 9330
    },
    {
      "entropy": 5.946507215499878,
      "epoch": 0.8157825744996942,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004940317010486351,
      "loss": 5.5443,
      "mean_token_accuracy": 0.18105872869491577,
      "num_tokens": 18914797.0,
      "step": 9335
    },
    {
      "entropy": 5.933585262298584,
      "epoch": 0.8162195228523988,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004940245693917183,
      "loss": 5.6293,
      "mean_token_accuracy": 0.17232362627983094,
      "num_tokens": 18924642.0,
      "step": 9340
    },
    {
      "entropy": 5.856119632720947,
      "epoch": 0.8166564712051035,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004940174335337955,
      "loss": 5.5044,
      "mean_token_accuracy": 0.1879820853471756,
      "num_tokens": 18934026.0,
      "step": 9345
    },
    {
      "entropy": 5.924860429763794,
      "epoch": 0.8170934195578082,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004940102934750037,
      "loss": 5.5183,
      "mean_token_accuracy": 0.1843946471810341,
      "num_tokens": 18943262.0,
      "step": 9350
    },
    {
      "entropy": 5.888956689834595,
      "epoch": 0.8175303679105129,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004940031492154795,
      "loss": 5.5619,
      "mean_token_accuracy": 0.17023331820964813,
      "num_tokens": 18953082.0,
      "step": 9355
    },
    {
      "entropy": 5.887869834899902,
      "epoch": 0.8179673162632177,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004939960007553603,
      "loss": 5.5891,
      "mean_token_accuracy": 0.18181607723236085,
      "num_tokens": 18962408.0,
      "step": 9360
    },
    {
      "entropy": 6.022705364227295,
      "epoch": 0.8184042646159224,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000493988848094783,
      "loss": 5.6722,
      "mean_token_accuracy": 0.1641612872481346,
      "num_tokens": 18972918.0,
      "step": 9365
    },
    {
      "entropy": 5.96536602973938,
      "epoch": 0.8188412129686271,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004939816912338848,
      "loss": 5.6217,
      "mean_token_accuracy": 0.1735097736120224,
      "num_tokens": 18983152.0,
      "step": 9370
    },
    {
      "entropy": 5.845053005218506,
      "epoch": 0.8192781613213318,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004939745301728031,
      "loss": 5.6157,
      "mean_token_accuracy": 0.17242834717035294,
      "num_tokens": 18993613.0,
      "step": 9375
    },
    {
      "entropy": 5.9107067584991455,
      "epoch": 0.8197151096740365,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004939673649116752,
      "loss": 5.5296,
      "mean_token_accuracy": 0.17893265038728715,
      "num_tokens": 19004079.0,
      "step": 9380
    },
    {
      "entropy": 5.92305736541748,
      "epoch": 0.8201520580267413,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004939601954506383,
      "loss": 5.6701,
      "mean_token_accuracy": 0.1768793821334839,
      "num_tokens": 19013994.0,
      "step": 9385
    },
    {
      "entropy": 5.9138116359710695,
      "epoch": 0.820589006379446,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004939530217898303,
      "loss": 5.6514,
      "mean_token_accuracy": 0.1697582945227623,
      "num_tokens": 19024674.0,
      "step": 9390
    },
    {
      "entropy": 5.9818462371826175,
      "epoch": 0.8210259547321507,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004939458439293886,
      "loss": 5.6166,
      "mean_token_accuracy": 0.17188004702329635,
      "num_tokens": 19035333.0,
      "step": 9395
    },
    {
      "entropy": 6.030829858779907,
      "epoch": 0.8214629030848554,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004939386618694508,
      "loss": 5.6707,
      "mean_token_accuracy": 0.17278372943401338,
      "num_tokens": 19046279.0,
      "step": 9400
    },
    {
      "entropy": 5.885382175445557,
      "epoch": 0.8218998514375601,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004939314756101547,
      "loss": 5.6061,
      "mean_token_accuracy": 0.1733551025390625,
      "num_tokens": 19056506.0,
      "step": 9405
    },
    {
      "entropy": 5.954209423065185,
      "epoch": 0.8223367997902647,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004939242851516384,
      "loss": 5.5881,
      "mean_token_accuracy": 0.17764620929956437,
      "num_tokens": 19067217.0,
      "step": 9410
    },
    {
      "entropy": 5.952820682525635,
      "epoch": 0.8227737481429696,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004939170904940395,
      "loss": 5.652,
      "mean_token_accuracy": 0.17765115052461625,
      "num_tokens": 19079028.0,
      "step": 9415
    },
    {
      "entropy": 5.865020751953125,
      "epoch": 0.8232106964956742,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000493909891637496,
      "loss": 5.6292,
      "mean_token_accuracy": 0.1764183074235916,
      "num_tokens": 19088236.0,
      "step": 9420
    },
    {
      "entropy": 5.900875997543335,
      "epoch": 0.8236476448483789,
      "grad_norm": 1.0,
      "learning_rate": 0.0004939026885821461,
      "loss": 5.5852,
      "mean_token_accuracy": 0.18227041512727737,
      "num_tokens": 19099598.0,
      "step": 9425
    },
    {
      "entropy": 5.965689086914063,
      "epoch": 0.8240845932010836,
      "grad_norm": 0.953125,
      "learning_rate": 0.000493895481328128,
      "loss": 5.6377,
      "mean_token_accuracy": 0.18035191297531128,
      "num_tokens": 19110347.0,
      "step": 9430
    },
    {
      "entropy": 6.006217813491821,
      "epoch": 0.8245215415537883,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004938882698755798,
      "loss": 5.6131,
      "mean_token_accuracy": 0.1704188570380211,
      "num_tokens": 19120643.0,
      "step": 9435
    },
    {
      "entropy": 5.928470420837402,
      "epoch": 0.824958489906493,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00049388105422464,
      "loss": 5.573,
      "mean_token_accuracy": 0.18047409504652023,
      "num_tokens": 19130209.0,
      "step": 9440
    },
    {
      "entropy": 5.9285194873809814,
      "epoch": 0.8253954382591978,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004938738343754467,
      "loss": 5.5589,
      "mean_token_accuracy": 0.1820627197623253,
      "num_tokens": 19139196.0,
      "step": 9445
    },
    {
      "entropy": 5.942254447937012,
      "epoch": 0.8258323866119025,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004938666103281389,
      "loss": 5.5738,
      "mean_token_accuracy": 0.17661122381687164,
      "num_tokens": 19149618.0,
      "step": 9450
    },
    {
      "entropy": 5.972391939163208,
      "epoch": 0.8262693349646072,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004938593820828545,
      "loss": 5.554,
      "mean_token_accuracy": 0.18163108676671982,
      "num_tokens": 19159875.0,
      "step": 9455
    },
    {
      "entropy": 5.923751926422119,
      "epoch": 0.8267062833173119,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004938521496397326,
      "loss": 5.6046,
      "mean_token_accuracy": 0.1769805744290352,
      "num_tokens": 19170403.0,
      "step": 9460
    },
    {
      "entropy": 5.920116996765136,
      "epoch": 0.8271432316700166,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004938449129989117,
      "loss": 5.5629,
      "mean_token_accuracy": 0.18427390456199647,
      "num_tokens": 19181187.0,
      "step": 9465
    },
    {
      "entropy": 5.930638170242309,
      "epoch": 0.8275801800227213,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004938376721605308,
      "loss": 5.6227,
      "mean_token_accuracy": 0.1697840616106987,
      "num_tokens": 19191328.0,
      "step": 9470
    },
    {
      "entropy": 5.907755994796753,
      "epoch": 0.8280171283754261,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004938304271247286,
      "loss": 5.5768,
      "mean_token_accuracy": 0.17677581906318665,
      "num_tokens": 19202041.0,
      "step": 9475
    },
    {
      "entropy": 5.95797381401062,
      "epoch": 0.8284540767281308,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004938231778916442,
      "loss": 5.5654,
      "mean_token_accuracy": 0.17643034756183623,
      "num_tokens": 19212016.0,
      "step": 9480
    },
    {
      "entropy": 5.939561080932617,
      "epoch": 0.8288910250808355,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004938159244614167,
      "loss": 5.5558,
      "mean_token_accuracy": 0.17979393601417543,
      "num_tokens": 19222111.0,
      "step": 9485
    },
    {
      "entropy": 5.946797275543213,
      "epoch": 0.8293279734335401,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004938086668341849,
      "loss": 5.5486,
      "mean_token_accuracy": 0.17822526395320892,
      "num_tokens": 19231865.0,
      "step": 9490
    },
    {
      "entropy": 5.903191566467285,
      "epoch": 0.8297649217862448,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004938014050100884,
      "loss": 5.5505,
      "mean_token_accuracy": 0.18404348492622374,
      "num_tokens": 19242993.0,
      "step": 9495
    },
    {
      "entropy": 5.909541845321655,
      "epoch": 0.8302018701389495,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004937941389892662,
      "loss": 5.6134,
      "mean_token_accuracy": 0.1751212701201439,
      "num_tokens": 19252776.0,
      "step": 9500
    },
    {
      "entropy": 5.926231002807617,
      "epoch": 0.8306388184916543,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004937868687718577,
      "loss": 5.5545,
      "mean_token_accuracy": 0.17665850967168809,
      "num_tokens": 19263022.0,
      "step": 9505
    },
    {
      "entropy": 5.957667827606201,
      "epoch": 0.831075766844359,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004937795943580026,
      "loss": 5.6043,
      "mean_token_accuracy": 0.17926428616046905,
      "num_tokens": 19271859.0,
      "step": 9510
    },
    {
      "entropy": 5.860514974594116,
      "epoch": 0.8315127151970637,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00049377231574784,
      "loss": 5.5303,
      "mean_token_accuracy": 0.1858970895409584,
      "num_tokens": 19282771.0,
      "step": 9515
    },
    {
      "entropy": 5.905759620666504,
      "epoch": 0.8319496635497684,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004937650329415099,
      "loss": 5.5904,
      "mean_token_accuracy": 0.17410718947649,
      "num_tokens": 19292985.0,
      "step": 9520
    },
    {
      "entropy": 5.90703444480896,
      "epoch": 0.8323866119024731,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004937577459391519,
      "loss": 5.6029,
      "mean_token_accuracy": 0.18420210033655166,
      "num_tokens": 19303409.0,
      "step": 9525
    },
    {
      "entropy": 5.9438241481781,
      "epoch": 0.8328235602551778,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004937504547409056,
      "loss": 5.5946,
      "mean_token_accuracy": 0.17865188419818878,
      "num_tokens": 19312632.0,
      "step": 9530
    },
    {
      "entropy": 5.906849336624146,
      "epoch": 0.8332605086078826,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004937431593469109,
      "loss": 5.4901,
      "mean_token_accuracy": 0.1811808541417122,
      "num_tokens": 19322759.0,
      "step": 9535
    },
    {
      "entropy": 5.908121156692505,
      "epoch": 0.8336974569605873,
      "grad_norm": 1.0,
      "learning_rate": 0.0004937358597573079,
      "loss": 5.5358,
      "mean_token_accuracy": 0.18670269548892976,
      "num_tokens": 19333307.0,
      "step": 9540
    },
    {
      "entropy": 5.905427026748657,
      "epoch": 0.834134405313292,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004937285559722364,
      "loss": 5.6418,
      "mean_token_accuracy": 0.17266591787338256,
      "num_tokens": 19342528.0,
      "step": 9545
    },
    {
      "entropy": 5.942676401138305,
      "epoch": 0.8345713536659967,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004937212479918367,
      "loss": 5.6018,
      "mean_token_accuracy": 0.17580595910549163,
      "num_tokens": 19352640.0,
      "step": 9550
    },
    {
      "entropy": 5.931405305862427,
      "epoch": 0.8350083020187014,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004937139358162488,
      "loss": 5.5678,
      "mean_token_accuracy": 0.17985041439533234,
      "num_tokens": 19363285.0,
      "step": 9555
    },
    {
      "entropy": 5.920910215377807,
      "epoch": 0.835445250371406,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000493706619445613,
      "loss": 5.5573,
      "mean_token_accuracy": 0.1780620649456978,
      "num_tokens": 19373141.0,
      "step": 9560
    },
    {
      "entropy": 5.844042348861694,
      "epoch": 0.8358821987241108,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004936992988800697,
      "loss": 5.6072,
      "mean_token_accuracy": 0.17690202444791794,
      "num_tokens": 19383091.0,
      "step": 9565
    },
    {
      "entropy": 5.9238934993743895,
      "epoch": 0.8363191470768155,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004936919741197591,
      "loss": 5.5449,
      "mean_token_accuracy": 0.17819783538579942,
      "num_tokens": 19393163.0,
      "step": 9570
    },
    {
      "entropy": 5.911790561676026,
      "epoch": 0.8367560954295202,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004936846451648219,
      "loss": 5.5106,
      "mean_token_accuracy": 0.18414548933506011,
      "num_tokens": 19403475.0,
      "step": 9575
    },
    {
      "entropy": 5.833174037933349,
      "epoch": 0.8371930437822249,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004936773120153987,
      "loss": 5.5339,
      "mean_token_accuracy": 0.18004835844039918,
      "num_tokens": 19413302.0,
      "step": 9580
    },
    {
      "entropy": 5.911049461364746,
      "epoch": 0.8376299921349296,
      "grad_norm": 1.03125,
      "learning_rate": 0.00049366997467163,
      "loss": 5.5244,
      "mean_token_accuracy": 0.18608558773994446,
      "num_tokens": 19423473.0,
      "step": 9585
    },
    {
      "entropy": 5.980356693267822,
      "epoch": 0.8380669404876344,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004936626331336567,
      "loss": 5.6436,
      "mean_token_accuracy": 0.16722328066825867,
      "num_tokens": 19433289.0,
      "step": 9590
    },
    {
      "entropy": 5.9684771537780765,
      "epoch": 0.8385038888403391,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004936552874016194,
      "loss": 5.6244,
      "mean_token_accuracy": 0.17390552014112473,
      "num_tokens": 19442188.0,
      "step": 9595
    },
    {
      "entropy": 5.9207535743713375,
      "epoch": 0.8389408371930438,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004936479374756592,
      "loss": 5.6259,
      "mean_token_accuracy": 0.1746540606021881,
      "num_tokens": 19452971.0,
      "step": 9600
    },
    {
      "entropy": 5.999142456054687,
      "epoch": 0.8393777855457485,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004936405833559169,
      "loss": 5.6067,
      "mean_token_accuracy": 0.1703120142221451,
      "num_tokens": 19463263.0,
      "step": 9605
    },
    {
      "entropy": 5.984088993072509,
      "epoch": 0.8398147338984532,
      "grad_norm": 1.0,
      "learning_rate": 0.0004936332250425337,
      "loss": 5.6213,
      "mean_token_accuracy": 0.17154912799596786,
      "num_tokens": 19473234.0,
      "step": 9610
    },
    {
      "entropy": 5.861338901519775,
      "epoch": 0.8402516822511579,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004936258625356507,
      "loss": 5.5962,
      "mean_token_accuracy": 0.17355925887823104,
      "num_tokens": 19483714.0,
      "step": 9615
    },
    {
      "entropy": 5.887535572052002,
      "epoch": 0.8406886306038627,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004936184958354091,
      "loss": 5.4906,
      "mean_token_accuracy": 0.1899486318230629,
      "num_tokens": 19493128.0,
      "step": 9620
    },
    {
      "entropy": 5.997431707382202,
      "epoch": 0.8411255789565674,
      "grad_norm": 1.71875,
      "learning_rate": 0.00049361112494195,
      "loss": 5.6428,
      "mean_token_accuracy": 0.16934374421834947,
      "num_tokens": 19503401.0,
      "step": 9625
    },
    {
      "entropy": 5.956140804290771,
      "epoch": 0.8415625273092721,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004936037498554153,
      "loss": 5.6733,
      "mean_token_accuracy": 0.17632997184991836,
      "num_tokens": 19513050.0,
      "step": 9630
    },
    {
      "entropy": 5.92626805305481,
      "epoch": 0.8419994756619767,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000493596370575946,
      "loss": 5.5663,
      "mean_token_accuracy": 0.1808543473482132,
      "num_tokens": 19524534.0,
      "step": 9635
    },
    {
      "entropy": 5.945484256744384,
      "epoch": 0.8424364240146814,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004935889871036837,
      "loss": 5.5392,
      "mean_token_accuracy": 0.17970518320798873,
      "num_tokens": 19534863.0,
      "step": 9640
    },
    {
      "entropy": 5.956271553039551,
      "epoch": 0.8428733723673861,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004935815994387701,
      "loss": 5.6373,
      "mean_token_accuracy": 0.17608590722084044,
      "num_tokens": 19545037.0,
      "step": 9645
    },
    {
      "entropy": 5.9264758110046385,
      "epoch": 0.8433103207200909,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000493574207581347,
      "loss": 5.6258,
      "mean_token_accuracy": 0.16774329990148545,
      "num_tokens": 19555377.0,
      "step": 9650
    },
    {
      "entropy": 5.994196891784668,
      "epoch": 0.8437472690727956,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004935668115315559,
      "loss": 5.6187,
      "mean_token_accuracy": 0.17211553156375886,
      "num_tokens": 19565874.0,
      "step": 9655
    },
    {
      "entropy": 5.789281892776489,
      "epoch": 0.8441842174255003,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000493559411289539,
      "loss": 5.4657,
      "mean_token_accuracy": 0.19241530001163482,
      "num_tokens": 19575002.0,
      "step": 9660
    },
    {
      "entropy": 5.882970190048217,
      "epoch": 0.844621165778205,
      "grad_norm": 0.9375,
      "learning_rate": 0.000493552006855438,
      "loss": 5.5728,
      "mean_token_accuracy": 0.1817576602101326,
      "num_tokens": 19585863.0,
      "step": 9665
    },
    {
      "entropy": 5.93998236656189,
      "epoch": 0.8450581141309097,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004935445982293949,
      "loss": 5.5331,
      "mean_token_accuracy": 0.1773715063929558,
      "num_tokens": 19596787.0,
      "step": 9670
    },
    {
      "entropy": 5.914847278594971,
      "epoch": 0.8454950624836144,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004935371854115519,
      "loss": 5.6201,
      "mean_token_accuracy": 0.1713850364089012,
      "num_tokens": 19607114.0,
      "step": 9675
    },
    {
      "entropy": 5.896970176696778,
      "epoch": 0.8459320108363192,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004935297684020513,
      "loss": 5.497,
      "mean_token_accuracy": 0.18718392848968507,
      "num_tokens": 19617334.0,
      "step": 9680
    },
    {
      "entropy": 5.91257963180542,
      "epoch": 0.8463689591890239,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004935223472010352,
      "loss": 5.5911,
      "mean_token_accuracy": 0.17462791502475739,
      "num_tokens": 19626844.0,
      "step": 9685
    },
    {
      "entropy": 5.898280000686645,
      "epoch": 0.8468059075417286,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004935149218086459,
      "loss": 5.5954,
      "mean_token_accuracy": 0.17506770491600038,
      "num_tokens": 19637897.0,
      "step": 9690
    },
    {
      "entropy": 5.9259556293487545,
      "epoch": 0.8472428558944333,
      "grad_norm": 0.984375,
      "learning_rate": 0.000493507492225026,
      "loss": 5.6617,
      "mean_token_accuracy": 0.175199656188488,
      "num_tokens": 19649161.0,
      "step": 9695
    },
    {
      "entropy": 5.96745023727417,
      "epoch": 0.847679804247138,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004935000584503178,
      "loss": 5.5815,
      "mean_token_accuracy": 0.17617280334234237,
      "num_tokens": 19659328.0,
      "step": 9700
    },
    {
      "entropy": 5.928172636032104,
      "epoch": 0.8481167525998426,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004934926204846641,
      "loss": 5.6588,
      "mean_token_accuracy": 0.17468160688877105,
      "num_tokens": 19669196.0,
      "step": 9705
    },
    {
      "entropy": 5.891255331039429,
      "epoch": 0.8485537009525475,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004934851783282073,
      "loss": 5.5447,
      "mean_token_accuracy": 0.1750273048877716,
      "num_tokens": 19680019.0,
      "step": 9710
    },
    {
      "entropy": 5.927026224136353,
      "epoch": 0.8489906493052521,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004934777319810904,
      "loss": 5.6716,
      "mean_token_accuracy": 0.1778111532330513,
      "num_tokens": 19689391.0,
      "step": 9715
    },
    {
      "entropy": 5.978184032440185,
      "epoch": 0.8494275976579568,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004934702814434561,
      "loss": 5.645,
      "mean_token_accuracy": 0.16932921782135962,
      "num_tokens": 19699683.0,
      "step": 9720
    },
    {
      "entropy": 6.018511533737183,
      "epoch": 0.8498645460106615,
      "grad_norm": 1.0,
      "learning_rate": 0.0004934628267154474,
      "loss": 5.6835,
      "mean_token_accuracy": 0.1683285966515541,
      "num_tokens": 19709844.0,
      "step": 9725
    },
    {
      "entropy": 5.997361373901367,
      "epoch": 0.8503014943633662,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000493455367797207,
      "loss": 5.6354,
      "mean_token_accuracy": 0.1725834056735039,
      "num_tokens": 19719926.0,
      "step": 9730
    },
    {
      "entropy": 5.896339845657349,
      "epoch": 0.8507384427160709,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004934479046888784,
      "loss": 5.5601,
      "mean_token_accuracy": 0.17677287459373475,
      "num_tokens": 19730548.0,
      "step": 9735
    },
    {
      "entropy": 5.963084316253662,
      "epoch": 0.8511753910687757,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004934404373906045,
      "loss": 5.6337,
      "mean_token_accuracy": 0.17841481566429138,
      "num_tokens": 19740471.0,
      "step": 9740
    },
    {
      "entropy": 5.965726613998413,
      "epoch": 0.8516123394214804,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004934329659025283,
      "loss": 5.5721,
      "mean_token_accuracy": 0.17959243804216385,
      "num_tokens": 19749395.0,
      "step": 9745
    },
    {
      "entropy": 5.9348077297210695,
      "epoch": 0.8520492877741851,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004934254902247936,
      "loss": 5.637,
      "mean_token_accuracy": 0.16899807751178741,
      "num_tokens": 19759203.0,
      "step": 9750
    },
    {
      "entropy": 5.937801790237427,
      "epoch": 0.8524862361268898,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004934180103575436,
      "loss": 5.492,
      "mean_token_accuracy": 0.18962620943784714,
      "num_tokens": 19768977.0,
      "step": 9755
    },
    {
      "entropy": 5.8845103740692135,
      "epoch": 0.8529231844795945,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004934105263009216,
      "loss": 5.4957,
      "mean_token_accuracy": 0.18022862821817398,
      "num_tokens": 19778569.0,
      "step": 9760
    },
    {
      "entropy": 5.9163658142089846,
      "epoch": 0.8533601328322992,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004934030380550713,
      "loss": 5.5932,
      "mean_token_accuracy": 0.18088656514883042,
      "num_tokens": 19787939.0,
      "step": 9765
    },
    {
      "entropy": 5.926901626586914,
      "epoch": 0.853797081185004,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004933955456201362,
      "loss": 5.5754,
      "mean_token_accuracy": 0.17841752469539643,
      "num_tokens": 19798407.0,
      "step": 9770
    },
    {
      "entropy": 5.987814092636109,
      "epoch": 0.8542340295377087,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004933880489962602,
      "loss": 5.6209,
      "mean_token_accuracy": 0.17635353952646254,
      "num_tokens": 19808495.0,
      "step": 9775
    },
    {
      "entropy": 5.896914148330689,
      "epoch": 0.8546709778904134,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004933805481835868,
      "loss": 5.555,
      "mean_token_accuracy": 0.17709617614746093,
      "num_tokens": 19818709.0,
      "step": 9780
    },
    {
      "entropy": 5.9542567253112795,
      "epoch": 0.855107926243118,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004933730431822603,
      "loss": 5.64,
      "mean_token_accuracy": 0.17609432935714722,
      "num_tokens": 19828865.0,
      "step": 9785
    },
    {
      "entropy": 5.937848234176636,
      "epoch": 0.8555448745958227,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004933655339924243,
      "loss": 5.5567,
      "mean_token_accuracy": 0.17561343759298326,
      "num_tokens": 19838757.0,
      "step": 9790
    },
    {
      "entropy": 5.846037673950195,
      "epoch": 0.8559818229485275,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000493358020614223,
      "loss": 5.6041,
      "mean_token_accuracy": 0.17844699919223786,
      "num_tokens": 19848967.0,
      "step": 9795
    },
    {
      "entropy": 5.958090353012085,
      "epoch": 0.8564187713012322,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004933505030478003,
      "loss": 5.5972,
      "mean_token_accuracy": 0.1763786181807518,
      "num_tokens": 19859487.0,
      "step": 9800
    },
    {
      "entropy": 5.977826404571533,
      "epoch": 0.8568557196539369,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004933429812933005,
      "loss": 5.6666,
      "mean_token_accuracy": 0.1732749953866005,
      "num_tokens": 19870170.0,
      "step": 9805
    },
    {
      "entropy": 5.858450984954834,
      "epoch": 0.8572926680066416,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004933354553508681,
      "loss": 5.5328,
      "mean_token_accuracy": 0.1778392732143402,
      "num_tokens": 19881031.0,
      "step": 9810
    },
    {
      "entropy": 5.948590469360352,
      "epoch": 0.8577296163593463,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004933279252206472,
      "loss": 5.5947,
      "mean_token_accuracy": 0.1749552756547928,
      "num_tokens": 19891082.0,
      "step": 9815
    },
    {
      "entropy": 5.99031982421875,
      "epoch": 0.858166564712051,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004933203909027823,
      "loss": 5.577,
      "mean_token_accuracy": 0.17932410389184952,
      "num_tokens": 19900626.0,
      "step": 9820
    },
    {
      "entropy": 5.900529432296753,
      "epoch": 0.8586035130647558,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004933128523974179,
      "loss": 5.567,
      "mean_token_accuracy": 0.1801237240433693,
      "num_tokens": 19910660.0,
      "step": 9825
    },
    {
      "entropy": 5.870386362075806,
      "epoch": 0.8590404614174605,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004933053097046986,
      "loss": 5.5413,
      "mean_token_accuracy": 0.17730663865804672,
      "num_tokens": 19920476.0,
      "step": 9830
    },
    {
      "entropy": 6.042214488983154,
      "epoch": 0.8594774097701652,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000493297762824769,
      "loss": 5.6751,
      "mean_token_accuracy": 0.16873888224363326,
      "num_tokens": 19931151.0,
      "step": 9835
    },
    {
      "entropy": 5.950340986251831,
      "epoch": 0.8599143581228699,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000493290211757774,
      "loss": 5.5632,
      "mean_token_accuracy": 0.17167053520679473,
      "num_tokens": 19942469.0,
      "step": 9840
    },
    {
      "entropy": 5.86607255935669,
      "epoch": 0.8603513064755746,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004932826565038584,
      "loss": 5.6417,
      "mean_token_accuracy": 0.17524933815002441,
      "num_tokens": 19953055.0,
      "step": 9845
    },
    {
      "entropy": 5.868257761001587,
      "epoch": 0.8607882548282793,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004932750970631669,
      "loss": 5.5932,
      "mean_token_accuracy": 0.17575591206550598,
      "num_tokens": 19963944.0,
      "step": 9850
    },
    {
      "entropy": 5.932425832748413,
      "epoch": 0.8612252031809841,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004932675334358448,
      "loss": 5.5082,
      "mean_token_accuracy": 0.1848415121436119,
      "num_tokens": 19973913.0,
      "step": 9855
    },
    {
      "entropy": 5.8965537548065186,
      "epoch": 0.8616621515336887,
      "grad_norm": 1.078125,
      "learning_rate": 0.000493259965622037,
      "loss": 5.5844,
      "mean_token_accuracy": 0.17997561544179916,
      "num_tokens": 19984130.0,
      "step": 9860
    },
    {
      "entropy": 5.886917304992676,
      "epoch": 0.8620990998863934,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004932523936218887,
      "loss": 5.4944,
      "mean_token_accuracy": 0.1824835553765297,
      "num_tokens": 19993840.0,
      "step": 9865
    },
    {
      "entropy": 5.888387250900268,
      "epoch": 0.8625360482390981,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000493244817435545,
      "loss": 5.5787,
      "mean_token_accuracy": 0.17853519767522813,
      "num_tokens": 20004228.0,
      "step": 9870
    },
    {
      "entropy": 5.935756349563599,
      "epoch": 0.8629729965918028,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004932372370631515,
      "loss": 5.593,
      "mean_token_accuracy": 0.17891252189874648,
      "num_tokens": 20012620.0,
      "step": 9875
    },
    {
      "entropy": 5.890320587158203,
      "epoch": 0.8634099449445075,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004932296525048535,
      "loss": 5.6654,
      "mean_token_accuracy": 0.17731177061796188,
      "num_tokens": 20022892.0,
      "step": 9880
    },
    {
      "entropy": 5.969244480133057,
      "epoch": 0.8638468932972123,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004932220637607962,
      "loss": 5.5586,
      "mean_token_accuracy": 0.18241119831800462,
      "num_tokens": 20032788.0,
      "step": 9885
    },
    {
      "entropy": 5.898443508148193,
      "epoch": 0.864283841649917,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004932144708311256,
      "loss": 5.5834,
      "mean_token_accuracy": 0.18007792532444,
      "num_tokens": 20043176.0,
      "step": 9890
    },
    {
      "entropy": 5.920622873306274,
      "epoch": 0.8647207900026217,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000493206873715987,
      "loss": 5.5872,
      "mean_token_accuracy": 0.17288162410259247,
      "num_tokens": 20052298.0,
      "step": 9895
    },
    {
      "entropy": 5.924004554748535,
      "epoch": 0.8651577383553264,
      "grad_norm": 1.0,
      "learning_rate": 0.0004931992724155262,
      "loss": 5.5301,
      "mean_token_accuracy": 0.18130597919225694,
      "num_tokens": 20063163.0,
      "step": 9900
    },
    {
      "entropy": 5.93014669418335,
      "epoch": 0.8655946867080311,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000493191666929889,
      "loss": 5.6225,
      "mean_token_accuracy": 0.17594730108976364,
      "num_tokens": 20073644.0,
      "step": 9905
    },
    {
      "entropy": 5.931488943099976,
      "epoch": 0.8660316350607358,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004931840572592214,
      "loss": 5.5526,
      "mean_token_accuracy": 0.18023417592048646,
      "num_tokens": 20084457.0,
      "step": 9910
    },
    {
      "entropy": 5.846243810653687,
      "epoch": 0.8664685834134406,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004931764434036692,
      "loss": 5.4587,
      "mean_token_accuracy": 0.18632205277681352,
      "num_tokens": 20094153.0,
      "step": 9915
    },
    {
      "entropy": 5.895436859130859,
      "epoch": 0.8669055317661453,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004931688253633784,
      "loss": 5.6418,
      "mean_token_accuracy": 0.17238984107971192,
      "num_tokens": 20104031.0,
      "step": 9920
    },
    {
      "entropy": 5.931737613677979,
      "epoch": 0.86734248011885,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004931612031384953,
      "loss": 5.5567,
      "mean_token_accuracy": 0.18101924061775207,
      "num_tokens": 20114431.0,
      "step": 9925
    },
    {
      "entropy": 5.889099216461181,
      "epoch": 0.8677794284715546,
      "grad_norm": 1.0,
      "learning_rate": 0.000493153576729166,
      "loss": 5.4933,
      "mean_token_accuracy": 0.18208606243133546,
      "num_tokens": 20124161.0,
      "step": 9930
    },
    {
      "entropy": 5.868448257446289,
      "epoch": 0.8682163768242593,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004931459461355368,
      "loss": 5.5274,
      "mean_token_accuracy": 0.18656356781721115,
      "num_tokens": 20134227.0,
      "step": 9935
    },
    {
      "entropy": 5.913783168792724,
      "epoch": 0.868653325176964,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004931383113577541,
      "loss": 5.5429,
      "mean_token_accuracy": 0.17693062871694565,
      "num_tokens": 20143912.0,
      "step": 9940
    },
    {
      "entropy": 5.91628885269165,
      "epoch": 0.8690902735296688,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004931306723959641,
      "loss": 5.4764,
      "mean_token_accuracy": 0.18312112241983414,
      "num_tokens": 20153858.0,
      "step": 9945
    },
    {
      "entropy": 5.959555339813233,
      "epoch": 0.8695272218823735,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004931230292503137,
      "loss": 5.5922,
      "mean_token_accuracy": 0.17543545514345169,
      "num_tokens": 20164309.0,
      "step": 9950
    },
    {
      "entropy": 5.895211124420166,
      "epoch": 0.8699641702350782,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004931153819209492,
      "loss": 5.5729,
      "mean_token_accuracy": 0.1823590099811554,
      "num_tokens": 20173949.0,
      "step": 9955
    },
    {
      "entropy": 5.909511232376099,
      "epoch": 0.8704011185877829,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004931077304080175,
      "loss": 5.5564,
      "mean_token_accuracy": 0.17889275997877122,
      "num_tokens": 20184206.0,
      "step": 9960
    },
    {
      "entropy": 5.996935129165649,
      "epoch": 0.8708380669404876,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004931000747116651,
      "loss": 5.6792,
      "mean_token_accuracy": 0.17189015150070192,
      "num_tokens": 20194430.0,
      "step": 9965
    },
    {
      "entropy": 5.9704235076904295,
      "epoch": 0.8712750152931923,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000493092414832039,
      "loss": 5.5444,
      "mean_token_accuracy": 0.17991552799940108,
      "num_tokens": 20205004.0,
      "step": 9970
    },
    {
      "entropy": 5.846359443664551,
      "epoch": 0.8717119636458971,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004930847507692862,
      "loss": 5.5204,
      "mean_token_accuracy": 0.1819006696343422,
      "num_tokens": 20214698.0,
      "step": 9975
    },
    {
      "entropy": 5.922533845901489,
      "epoch": 0.8721489119986018,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004930770825235536,
      "loss": 5.6174,
      "mean_token_accuracy": 0.17570054084062575,
      "num_tokens": 20226534.0,
      "step": 9980
    },
    {
      "entropy": 5.9892055034637455,
      "epoch": 0.8725858603513065,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004930694100949882,
      "loss": 5.5976,
      "mean_token_accuracy": 0.17287162095308303,
      "num_tokens": 20237414.0,
      "step": 9985
    },
    {
      "entropy": 5.907688570022583,
      "epoch": 0.8730228087040112,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004930617334837373,
      "loss": 5.427,
      "mean_token_accuracy": 0.1857387363910675,
      "num_tokens": 20246948.0,
      "step": 9990
    },
    {
      "entropy": 5.883991241455078,
      "epoch": 0.8734597570567159,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000493054052689948,
      "loss": 5.6457,
      "mean_token_accuracy": 0.17481672763824463,
      "num_tokens": 20257281.0,
      "step": 9995
    },
    {
      "entropy": 5.927980947494507,
      "epoch": 0.8738967054094207,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004930463677137679,
      "loss": 5.6303,
      "mean_token_accuracy": 0.1736347869038582,
      "num_tokens": 20269041.0,
      "step": 10000
    },
    {
      "entropy": 5.9450188159942625,
      "epoch": 0.8743336537621254,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004930386785553441,
      "loss": 5.6666,
      "mean_token_accuracy": 0.16809693425893785,
      "num_tokens": 20279688.0,
      "step": 10005
    },
    {
      "entropy": 5.931889438629151,
      "epoch": 0.87477060211483,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004930309852148242,
      "loss": 5.5659,
      "mean_token_accuracy": 0.18223938047885896,
      "num_tokens": 20288913.0,
      "step": 10010
    },
    {
      "entropy": 5.951949977874756,
      "epoch": 0.8752075504675347,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004930232876923558,
      "loss": 5.6028,
      "mean_token_accuracy": 0.17948952913284302,
      "num_tokens": 20299009.0,
      "step": 10015
    },
    {
      "entropy": 5.933259201049805,
      "epoch": 0.8756444988202394,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004930155859880865,
      "loss": 5.5869,
      "mean_token_accuracy": 0.16984495967626573,
      "num_tokens": 20309478.0,
      "step": 10020
    },
    {
      "entropy": 5.9248357772827145,
      "epoch": 0.8760814471729441,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000493007880102164,
      "loss": 5.5961,
      "mean_token_accuracy": 0.17758634090423583,
      "num_tokens": 20319451.0,
      "step": 10025
    },
    {
      "entropy": 5.928604030609131,
      "epoch": 0.8765183955256489,
      "grad_norm": 0.96875,
      "learning_rate": 0.000493000170034736,
      "loss": 5.5585,
      "mean_token_accuracy": 0.18144280910491944,
      "num_tokens": 20328497.0,
      "step": 10030
    },
    {
      "entropy": 5.871942186355591,
      "epoch": 0.8769553438783536,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004929924557859507,
      "loss": 5.4785,
      "mean_token_accuracy": 0.19280209988355637,
      "num_tokens": 20338584.0,
      "step": 10035
    },
    {
      "entropy": 5.887005281448364,
      "epoch": 0.8773922922310583,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004929847373559557,
      "loss": 5.5414,
      "mean_token_accuracy": 0.18173762410879135,
      "num_tokens": 20348808.0,
      "step": 10040
    },
    {
      "entropy": 5.882996940612793,
      "epoch": 0.877829240583763,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004929770147448992,
      "loss": 5.547,
      "mean_token_accuracy": 0.18102590888738632,
      "num_tokens": 20358301.0,
      "step": 10045
    },
    {
      "entropy": 5.935358095169067,
      "epoch": 0.8782661889364677,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004929692879529295,
      "loss": 5.5501,
      "mean_token_accuracy": 0.17166950553655624,
      "num_tokens": 20369032.0,
      "step": 10050
    },
    {
      "entropy": 5.940617084503174,
      "epoch": 0.8787031372891724,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004929615569801946,
      "loss": 5.5883,
      "mean_token_accuracy": 0.1793626368045807,
      "num_tokens": 20379101.0,
      "step": 10055
    },
    {
      "entropy": 5.851898193359375,
      "epoch": 0.8791400856418772,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004929538218268428,
      "loss": 5.5388,
      "mean_token_accuracy": 0.18224130868911742,
      "num_tokens": 20388864.0,
      "step": 10060
    },
    {
      "entropy": 6.013635587692261,
      "epoch": 0.8795770339945819,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004929460824930226,
      "loss": 5.67,
      "mean_token_accuracy": 0.1729261875152588,
      "num_tokens": 20398402.0,
      "step": 10065
    },
    {
      "entropy": 5.846195936203003,
      "epoch": 0.8800139823472866,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004929383389788821,
      "loss": 5.5613,
      "mean_token_accuracy": 0.17754771709442138,
      "num_tokens": 20408209.0,
      "step": 10070
    },
    {
      "entropy": 5.902110910415649,
      "epoch": 0.8804509306999913,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004929305912845703,
      "loss": 5.5748,
      "mean_token_accuracy": 0.17818148136138917,
      "num_tokens": 20418117.0,
      "step": 10075
    },
    {
      "entropy": 6.0484819412231445,
      "epoch": 0.8808878790526959,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004929228394102355,
      "loss": 5.6079,
      "mean_token_accuracy": 0.1754727378487587,
      "num_tokens": 20428347.0,
      "step": 10080
    },
    {
      "entropy": 5.9748650074005125,
      "epoch": 0.8813248274054006,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004929150833560263,
      "loss": 5.563,
      "mean_token_accuracy": 0.17875707298517227,
      "num_tokens": 20437804.0,
      "step": 10085
    },
    {
      "entropy": 5.940289306640625,
      "epoch": 0.8817617757581054,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004929073231220917,
      "loss": 5.576,
      "mean_token_accuracy": 0.17122239321470262,
      "num_tokens": 20447600.0,
      "step": 10090
    },
    {
      "entropy": 5.965912342071533,
      "epoch": 0.8821987241108101,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004928995587085806,
      "loss": 5.5771,
      "mean_token_accuracy": 0.1776450902223587,
      "num_tokens": 20458023.0,
      "step": 10095
    },
    {
      "entropy": 5.960345125198364,
      "epoch": 0.8826356724635148,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004928917901156416,
      "loss": 5.5142,
      "mean_token_accuracy": 0.18678879886865615,
      "num_tokens": 20468195.0,
      "step": 10100
    },
    {
      "entropy": 5.894933223724365,
      "epoch": 0.8830726208162195,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000492884017343424,
      "loss": 5.672,
      "mean_token_accuracy": 0.16808237433433532,
      "num_tokens": 20479250.0,
      "step": 10105
    },
    {
      "entropy": 5.932264232635498,
      "epoch": 0.8835095691689242,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004928762403920767,
      "loss": 5.5563,
      "mean_token_accuracy": 0.17680608928203584,
      "num_tokens": 20489587.0,
      "step": 10110
    },
    {
      "entropy": 5.93992018699646,
      "epoch": 0.8839465175216289,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000492868459261749,
      "loss": 5.5688,
      "mean_token_accuracy": 0.18320697247982026,
      "num_tokens": 20499788.0,
      "step": 10115
    },
    {
      "entropy": 5.8812901973724365,
      "epoch": 0.8843834658743337,
      "grad_norm": 1.015625,
      "learning_rate": 0.00049286067395259,
      "loss": 5.5885,
      "mean_token_accuracy": 0.17923698127269744,
      "num_tokens": 20509716.0,
      "step": 10120
    },
    {
      "entropy": 5.934923505783081,
      "epoch": 0.8848204142270384,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004928528844647493,
      "loss": 5.5355,
      "mean_token_accuracy": 0.18377536982297898,
      "num_tokens": 20519020.0,
      "step": 10125
    },
    {
      "entropy": 5.9398908615112305,
      "epoch": 0.8852573625797431,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000492845090798376,
      "loss": 5.6057,
      "mean_token_accuracy": 0.17383627891540526,
      "num_tokens": 20528997.0,
      "step": 10130
    },
    {
      "entropy": 5.92419581413269,
      "epoch": 0.8856943109324478,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004928372929536197,
      "loss": 5.6124,
      "mean_token_accuracy": 0.17855922877788544,
      "num_tokens": 20538599.0,
      "step": 10135
    },
    {
      "entropy": 5.978060674667359,
      "epoch": 0.8861312592851525,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004928294909306301,
      "loss": 5.5259,
      "mean_token_accuracy": 0.1829664319753647,
      "num_tokens": 20547664.0,
      "step": 10140
    },
    {
      "entropy": 5.939069223403931,
      "epoch": 0.8865682076378572,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004928216847295565,
      "loss": 5.646,
      "mean_token_accuracy": 0.1708741679787636,
      "num_tokens": 20557860.0,
      "step": 10145
    },
    {
      "entropy": 5.886430454254151,
      "epoch": 0.887005155990562,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000492813874350549,
      "loss": 5.5952,
      "mean_token_accuracy": 0.17978934794664383,
      "num_tokens": 20569003.0,
      "step": 10150
    },
    {
      "entropy": 5.962632703781128,
      "epoch": 0.8874421043432666,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004928060597937573,
      "loss": 5.5714,
      "mean_token_accuracy": 0.17794898450374602,
      "num_tokens": 20578402.0,
      "step": 10155
    },
    {
      "entropy": 5.966803932189942,
      "epoch": 0.8878790526959713,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004927982410593312,
      "loss": 5.4687,
      "mean_token_accuracy": 0.18631904423236847,
      "num_tokens": 20587874.0,
      "step": 10160
    },
    {
      "entropy": 5.896160459518432,
      "epoch": 0.888316001048676,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004927904181474207,
      "loss": 5.51,
      "mean_token_accuracy": 0.17632167488336564,
      "num_tokens": 20597270.0,
      "step": 10165
    },
    {
      "entropy": 5.931625890731811,
      "epoch": 0.8887529494013807,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004927825910581758,
      "loss": 5.6253,
      "mean_token_accuracy": 0.17671879827976228,
      "num_tokens": 20607898.0,
      "step": 10170
    },
    {
      "entropy": 5.931588077545166,
      "epoch": 0.8891898977540854,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004927747597917467,
      "loss": 5.4987,
      "mean_token_accuracy": 0.18518080562353134,
      "num_tokens": 20617300.0,
      "step": 10175
    },
    {
      "entropy": 5.9317995548248295,
      "epoch": 0.8896268461067902,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004927669243482836,
      "loss": 5.6085,
      "mean_token_accuracy": 0.17620816826820374,
      "num_tokens": 20626582.0,
      "step": 10180
    },
    {
      "entropy": 5.885530090332031,
      "epoch": 0.8900637944594949,
      "grad_norm": 1.0,
      "learning_rate": 0.0004927590847279369,
      "loss": 5.5602,
      "mean_token_accuracy": 0.18117612451314927,
      "num_tokens": 20635802.0,
      "step": 10185
    },
    {
      "entropy": 5.935629415512085,
      "epoch": 0.8905007428121996,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004927512409308568,
      "loss": 5.5487,
      "mean_token_accuracy": 0.18115555197000505,
      "num_tokens": 20646057.0,
      "step": 10190
    },
    {
      "entropy": 5.956423807144165,
      "epoch": 0.8909376911649043,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004927433929571937,
      "loss": 5.6738,
      "mean_token_accuracy": 0.16478917598724366,
      "num_tokens": 20656511.0,
      "step": 10195
    },
    {
      "entropy": 5.950502872467041,
      "epoch": 0.891374639517609,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004927355408070982,
      "loss": 5.5509,
      "mean_token_accuracy": 0.18070328384637832,
      "num_tokens": 20666479.0,
      "step": 10200
    },
    {
      "entropy": 5.9074845790863035,
      "epoch": 0.8918115878703138,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004927276844807211,
      "loss": 5.4866,
      "mean_token_accuracy": 0.18179930448532106,
      "num_tokens": 20676830.0,
      "step": 10205
    },
    {
      "entropy": 5.818777990341187,
      "epoch": 0.8922485362230185,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004927198239782129,
      "loss": 5.5339,
      "mean_token_accuracy": 0.18245137482881546,
      "num_tokens": 20687316.0,
      "step": 10210
    },
    {
      "entropy": 5.930769443511963,
      "epoch": 0.8926854845757232,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004927119592997242,
      "loss": 5.6346,
      "mean_token_accuracy": 0.17210864126682282,
      "num_tokens": 20697901.0,
      "step": 10215
    },
    {
      "entropy": 5.94341139793396,
      "epoch": 0.8931224329284279,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004927040904454062,
      "loss": 5.5148,
      "mean_token_accuracy": 0.18359944522380828,
      "num_tokens": 20707468.0,
      "step": 10220
    },
    {
      "entropy": 5.929954910278321,
      "epoch": 0.8935593812811325,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004926962174154097,
      "loss": 5.5794,
      "mean_token_accuracy": 0.17712156623601913,
      "num_tokens": 20717098.0,
      "step": 10225
    },
    {
      "entropy": 5.9871293067932125,
      "epoch": 0.8939963296338372,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004926883402098857,
      "loss": 5.6843,
      "mean_token_accuracy": 0.1705389901995659,
      "num_tokens": 20727217.0,
      "step": 10230
    },
    {
      "entropy": 5.956632280349732,
      "epoch": 0.894433277986542,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004926804588289852,
      "loss": 5.6579,
      "mean_token_accuracy": 0.17254393994808198,
      "num_tokens": 20737397.0,
      "step": 10235
    },
    {
      "entropy": 5.838171434402466,
      "epoch": 0.8948702263392467,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004926725732728595,
      "loss": 5.408,
      "mean_token_accuracy": 0.19047937989234925,
      "num_tokens": 20746712.0,
      "step": 10240
    },
    {
      "entropy": 5.944919776916504,
      "epoch": 0.8953071746919514,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004926646835416597,
      "loss": 5.5737,
      "mean_token_accuracy": 0.17735190242528914,
      "num_tokens": 20757161.0,
      "step": 10245
    },
    {
      "entropy": 5.926939487457275,
      "epoch": 0.8957441230446561,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004926567896355372,
      "loss": 5.5082,
      "mean_token_accuracy": 0.18624234050512314,
      "num_tokens": 20766478.0,
      "step": 10250
    },
    {
      "entropy": 5.877347230911255,
      "epoch": 0.8961810713973608,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004926488915546435,
      "loss": 5.5494,
      "mean_token_accuracy": 0.17950234413146973,
      "num_tokens": 20776317.0,
      "step": 10255
    },
    {
      "entropy": 5.911137580871582,
      "epoch": 0.8966180197500655,
      "grad_norm": 0.9375,
      "learning_rate": 0.00049264098929913,
      "loss": 5.4831,
      "mean_token_accuracy": 0.1830203801393509,
      "num_tokens": 20786534.0,
      "step": 10260
    },
    {
      "entropy": 5.8699156761169435,
      "epoch": 0.8970549681027703,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004926330828691482,
      "loss": 5.5197,
      "mean_token_accuracy": 0.1797871857881546,
      "num_tokens": 20796791.0,
      "step": 10265
    },
    {
      "entropy": 5.915482711791992,
      "epoch": 0.897491916455475,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004926251722648499,
      "loss": 5.5329,
      "mean_token_accuracy": 0.18120759278535842,
      "num_tokens": 20806834.0,
      "step": 10270
    },
    {
      "entropy": 5.867949533462524,
      "epoch": 0.8979288648081797,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004926172574863868,
      "loss": 5.5324,
      "mean_token_accuracy": 0.18182554990053176,
      "num_tokens": 20816664.0,
      "step": 10275
    },
    {
      "entropy": 5.883551931381225,
      "epoch": 0.8983658131608844,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004926093385339105,
      "loss": 5.6729,
      "mean_token_accuracy": 0.17133614271879197,
      "num_tokens": 20826936.0,
      "step": 10280
    },
    {
      "entropy": 5.9507652759552006,
      "epoch": 0.8988027615135891,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004926014154075732,
      "loss": 5.5223,
      "mean_token_accuracy": 0.18317344039678574,
      "num_tokens": 20836732.0,
      "step": 10285
    },
    {
      "entropy": 5.982305717468262,
      "epoch": 0.8992397098662938,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004925934881075268,
      "loss": 5.5729,
      "mean_token_accuracy": 0.18354698568582534,
      "num_tokens": 20848093.0,
      "step": 10290
    },
    {
      "entropy": 5.924367761611938,
      "epoch": 0.8996766582189986,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004925855566339232,
      "loss": 5.5337,
      "mean_token_accuracy": 0.17800108939409257,
      "num_tokens": 20858678.0,
      "step": 10295
    },
    {
      "entropy": 5.87822413444519,
      "epoch": 0.9001136065717033,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004925776209869146,
      "loss": 5.5176,
      "mean_token_accuracy": 0.1794927403330803,
      "num_tokens": 20869319.0,
      "step": 10300
    },
    {
      "entropy": 5.910153484344482,
      "epoch": 0.9005505549244079,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004925696811666532,
      "loss": 5.4809,
      "mean_token_accuracy": 0.1861803337931633,
      "num_tokens": 20879475.0,
      "step": 10305
    },
    {
      "entropy": 5.876563501358032,
      "epoch": 0.9009875032771126,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004925617371732914,
      "loss": 5.5213,
      "mean_token_accuracy": 0.18320977091789245,
      "num_tokens": 20889479.0,
      "step": 10310
    },
    {
      "entropy": 5.927946281433106,
      "epoch": 0.9014244516298173,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004925537890069815,
      "loss": 5.5999,
      "mean_token_accuracy": 0.17840709984302522,
      "num_tokens": 20899864.0,
      "step": 10315
    },
    {
      "entropy": 5.937209892272949,
      "epoch": 0.901861399982522,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004925458366678758,
      "loss": 5.587,
      "mean_token_accuracy": 0.18126829713582993,
      "num_tokens": 20909006.0,
      "step": 10320
    },
    {
      "entropy": 5.930752515792847,
      "epoch": 0.9022983483352268,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004925378801561269,
      "loss": 5.5335,
      "mean_token_accuracy": 0.178778800368309,
      "num_tokens": 20917998.0,
      "step": 10325
    },
    {
      "entropy": 5.873672294616699,
      "epoch": 0.9027352966879315,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004925299194718876,
      "loss": 5.5405,
      "mean_token_accuracy": 0.1809740424156189,
      "num_tokens": 20928407.0,
      "step": 10330
    },
    {
      "entropy": 5.853119802474976,
      "epoch": 0.9031722450406362,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004925219546153105,
      "loss": 5.5651,
      "mean_token_accuracy": 0.17524684071540833,
      "num_tokens": 20938016.0,
      "step": 10335
    },
    {
      "entropy": 5.971973657608032,
      "epoch": 0.9036091933933409,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004925139855865483,
      "loss": 5.6354,
      "mean_token_accuracy": 0.16885083615779878,
      "num_tokens": 20948578.0,
      "step": 10340
    },
    {
      "entropy": 5.955605316162109,
      "epoch": 0.9040461417460456,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004925060123857539,
      "loss": 5.4805,
      "mean_token_accuracy": 0.1810566797852516,
      "num_tokens": 20958193.0,
      "step": 10345
    },
    {
      "entropy": 5.908290052413941,
      "epoch": 0.9044830900987503,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004924980350130802,
      "loss": 5.5401,
      "mean_token_accuracy": 0.1826064869761467,
      "num_tokens": 20968092.0,
      "step": 10350
    },
    {
      "entropy": 5.876534652709961,
      "epoch": 0.9049200384514551,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004924900534686801,
      "loss": 5.5718,
      "mean_token_accuracy": 0.17364884316921234,
      "num_tokens": 20978303.0,
      "step": 10355
    },
    {
      "entropy": 5.955595302581787,
      "epoch": 0.9053569868041598,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004924820677527069,
      "loss": 5.5774,
      "mean_token_accuracy": 0.1774071291089058,
      "num_tokens": 20988331.0,
      "step": 10360
    },
    {
      "entropy": 5.863691139221191,
      "epoch": 0.9057939351568645,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004924740778653137,
      "loss": 5.4774,
      "mean_token_accuracy": 0.17916994988918306,
      "num_tokens": 20997785.0,
      "step": 10365
    },
    {
      "entropy": 5.942973995208741,
      "epoch": 0.9062308835095692,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004924660838066538,
      "loss": 5.6034,
      "mean_token_accuracy": 0.17749667167663574,
      "num_tokens": 21008026.0,
      "step": 10370
    },
    {
      "entropy": 5.910295295715332,
      "epoch": 0.9066678318622738,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004924580855768803,
      "loss": 5.5288,
      "mean_token_accuracy": 0.1846077099442482,
      "num_tokens": 21017993.0,
      "step": 10375
    },
    {
      "entropy": 5.893343687057495,
      "epoch": 0.9071047802149785,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004924500831761468,
      "loss": 5.5526,
      "mean_token_accuracy": 0.18337782770395278,
      "num_tokens": 21028149.0,
      "step": 10380
    },
    {
      "entropy": 6.007966756820679,
      "epoch": 0.9075417285676833,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004924420766046069,
      "loss": 5.6611,
      "mean_token_accuracy": 0.17085800170898438,
      "num_tokens": 21039229.0,
      "step": 10385
    },
    {
      "entropy": 5.963099479675293,
      "epoch": 0.907978676920388,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004924340658624139,
      "loss": 5.6161,
      "mean_token_accuracy": 0.17394722402095794,
      "num_tokens": 21049358.0,
      "step": 10390
    },
    {
      "entropy": 5.9448339462280275,
      "epoch": 0.9084156252730927,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004924260509497217,
      "loss": 5.6253,
      "mean_token_accuracy": 0.17513106465339662,
      "num_tokens": 21060385.0,
      "step": 10395
    },
    {
      "entropy": 5.980993175506592,
      "epoch": 0.9088525736257974,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004924180318666838,
      "loss": 5.6216,
      "mean_token_accuracy": 0.1736283302307129,
      "num_tokens": 21070423.0,
      "step": 10400
    },
    {
      "entropy": 6.059029293060303,
      "epoch": 0.9092895219785021,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004924100086134541,
      "loss": 5.6018,
      "mean_token_accuracy": 0.17407021373510362,
      "num_tokens": 21080505.0,
      "step": 10405
    },
    {
      "entropy": 5.915446472167969,
      "epoch": 0.9097264703312069,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004924019811901866,
      "loss": 5.5706,
      "mean_token_accuracy": 0.1762954741716385,
      "num_tokens": 21090875.0,
      "step": 10410
    },
    {
      "entropy": 5.923772859573364,
      "epoch": 0.9101634186839116,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004923939495970352,
      "loss": 5.4993,
      "mean_token_accuracy": 0.1786544367671013,
      "num_tokens": 21101427.0,
      "step": 10415
    },
    {
      "entropy": 5.988828420639038,
      "epoch": 0.9106003670366163,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004923859138341539,
      "loss": 5.655,
      "mean_token_accuracy": 0.17199515104293822,
      "num_tokens": 21112691.0,
      "step": 10420
    },
    {
      "entropy": 5.959005403518677,
      "epoch": 0.911037315389321,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004923778739016969,
      "loss": 5.6347,
      "mean_token_accuracy": 0.17281511127948762,
      "num_tokens": 21123932.0,
      "step": 10425
    },
    {
      "entropy": 5.970179224014283,
      "epoch": 0.9114742637420257,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004923698297998183,
      "loss": 5.6432,
      "mean_token_accuracy": 0.1704394280910492,
      "num_tokens": 21135380.0,
      "step": 10430
    },
    {
      "entropy": 5.9281518936157225,
      "epoch": 0.9119112120947304,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004923617815286725,
      "loss": 5.5254,
      "mean_token_accuracy": 0.1837554693222046,
      "num_tokens": 21144951.0,
      "step": 10435
    },
    {
      "entropy": 5.904066658020019,
      "epoch": 0.9123481604474352,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004923537290884138,
      "loss": 5.4493,
      "mean_token_accuracy": 0.1855260044336319,
      "num_tokens": 21155919.0,
      "step": 10440
    },
    {
      "entropy": 5.879783058166504,
      "epoch": 0.9127851088001399,
      "grad_norm": 1.0,
      "learning_rate": 0.0004923456724791967,
      "loss": 5.5191,
      "mean_token_accuracy": 0.177931010723114,
      "num_tokens": 21166186.0,
      "step": 10445
    },
    {
      "entropy": 5.844772577285767,
      "epoch": 0.9132220571528445,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004923376117011758,
      "loss": 5.4871,
      "mean_token_accuracy": 0.1870008334517479,
      "num_tokens": 21176043.0,
      "step": 10450
    },
    {
      "entropy": 5.853738451004029,
      "epoch": 0.9136590055055492,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004923295467545055,
      "loss": 5.4925,
      "mean_token_accuracy": 0.18429099172353744,
      "num_tokens": 21185677.0,
      "step": 10455
    },
    {
      "entropy": 5.953488731384278,
      "epoch": 0.9140959538582539,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004923214776393407,
      "loss": 5.6451,
      "mean_token_accuracy": 0.16692266911268233,
      "num_tokens": 21195977.0,
      "step": 10460
    },
    {
      "entropy": 6.005151224136353,
      "epoch": 0.9145329022109586,
      "grad_norm": 0.984375,
      "learning_rate": 0.000492313404355836,
      "loss": 5.5938,
      "mean_token_accuracy": 0.18121683448553086,
      "num_tokens": 21205454.0,
      "step": 10465
    },
    {
      "entropy": 5.901759815216065,
      "epoch": 0.9149698505636634,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004923053269041464,
      "loss": 5.4976,
      "mean_token_accuracy": 0.18386061787605285,
      "num_tokens": 21215700.0,
      "step": 10470
    },
    {
      "entropy": 5.876657438278198,
      "epoch": 0.9154067989163681,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004922972452844267,
      "loss": 5.5775,
      "mean_token_accuracy": 0.18323282897472382,
      "num_tokens": 21226544.0,
      "step": 10475
    },
    {
      "entropy": 5.978432989120483,
      "epoch": 0.9158437472690728,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004922891594968321,
      "loss": 5.6343,
      "mean_token_accuracy": 0.17339424788951874,
      "num_tokens": 21236792.0,
      "step": 10480
    },
    {
      "entropy": 5.938476133346557,
      "epoch": 0.9162806956217775,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004922810695415175,
      "loss": 5.5224,
      "mean_token_accuracy": 0.18169684559106827,
      "num_tokens": 21247039.0,
      "step": 10485
    },
    {
      "entropy": 5.881524562835693,
      "epoch": 0.9167176439744822,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004922729754186381,
      "loss": 5.5381,
      "mean_token_accuracy": 0.18478634506464003,
      "num_tokens": 21257008.0,
      "step": 10490
    },
    {
      "entropy": 5.907452821731567,
      "epoch": 0.9171545923271869,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004922648771283492,
      "loss": 5.6748,
      "mean_token_accuracy": 0.17625463157892227,
      "num_tokens": 21267638.0,
      "step": 10495
    },
    {
      "entropy": 5.960263299942016,
      "epoch": 0.9175915406798917,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004922567746708062,
      "loss": 5.5105,
      "mean_token_accuracy": 0.18463107645511628,
      "num_tokens": 21277327.0,
      "step": 10500
    },
    {
      "entropy": 5.910731554031372,
      "epoch": 0.9180284890325964,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004922486680461642,
      "loss": 5.6018,
      "mean_token_accuracy": 0.17155911177396774,
      "num_tokens": 21287579.0,
      "step": 10505
    },
    {
      "entropy": 5.891546583175659,
      "epoch": 0.9184654373853011,
      "grad_norm": 1.0,
      "learning_rate": 0.0004922405572545792,
      "loss": 5.5005,
      "mean_token_accuracy": 0.1801793247461319,
      "num_tokens": 21297888.0,
      "step": 10510
    },
    {
      "entropy": 5.9577436447143555,
      "epoch": 0.9189023857380058,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004922324422962063,
      "loss": 5.5904,
      "mean_token_accuracy": 0.1787476047873497,
      "num_tokens": 21308506.0,
      "step": 10515
    },
    {
      "entropy": 5.954980182647705,
      "epoch": 0.9193393340907104,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004922243231712013,
      "loss": 5.6549,
      "mean_token_accuracy": 0.1701146885752678,
      "num_tokens": 21319252.0,
      "step": 10520
    },
    {
      "entropy": 5.955730628967285,
      "epoch": 0.9197762824434151,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00049221619987972,
      "loss": 5.59,
      "mean_token_accuracy": 0.1813602030277252,
      "num_tokens": 21328829.0,
      "step": 10525
    },
    {
      "entropy": 5.902078437805176,
      "epoch": 0.9202132307961199,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004922080724219182,
      "loss": 5.3968,
      "mean_token_accuracy": 0.18949666768312454,
      "num_tokens": 21338524.0,
      "step": 10530
    },
    {
      "entropy": 5.8982140064239506,
      "epoch": 0.9206501791488246,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004921999407979518,
      "loss": 5.5525,
      "mean_token_accuracy": 0.1770140305161476,
      "num_tokens": 21348559.0,
      "step": 10535
    },
    {
      "entropy": 5.877134704589844,
      "epoch": 0.9210871275015293,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004921918050079766,
      "loss": 5.5727,
      "mean_token_accuracy": 0.1814416915178299,
      "num_tokens": 21360763.0,
      "step": 10540
    },
    {
      "entropy": 5.986933040618896,
      "epoch": 0.921524075854234,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004921836650521489,
      "loss": 5.5975,
      "mean_token_accuracy": 0.17603018134832382,
      "num_tokens": 21370048.0,
      "step": 10545
    },
    {
      "entropy": 5.8879753112792965,
      "epoch": 0.9219610242069387,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004921755209306247,
      "loss": 5.573,
      "mean_token_accuracy": 0.18297244608402252,
      "num_tokens": 21379760.0,
      "step": 10550
    },
    {
      "entropy": 5.935941219329834,
      "epoch": 0.9223979725596434,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004921673726435601,
      "loss": 5.5302,
      "mean_token_accuracy": 0.17724643051624298,
      "num_tokens": 21389712.0,
      "step": 10555
    },
    {
      "entropy": 5.963999462127686,
      "epoch": 0.9228349209123482,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004921592201911115,
      "loss": 5.5351,
      "mean_token_accuracy": 0.17720993012189865,
      "num_tokens": 21399515.0,
      "step": 10560
    },
    {
      "entropy": 5.8920739650726315,
      "epoch": 0.9232718692650529,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004921510635734353,
      "loss": 5.5055,
      "mean_token_accuracy": 0.18461298197507858,
      "num_tokens": 21409704.0,
      "step": 10565
    },
    {
      "entropy": 5.885887527465821,
      "epoch": 0.9237088176177576,
      "grad_norm": 1.015625,
      "learning_rate": 0.000492142902790688,
      "loss": 5.5068,
      "mean_token_accuracy": 0.1860102578997612,
      "num_tokens": 21419182.0,
      "step": 10570
    },
    {
      "entropy": 5.833557796478272,
      "epoch": 0.9241457659704623,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004921347378430259,
      "loss": 5.4033,
      "mean_token_accuracy": 0.18897175788879395,
      "num_tokens": 21429749.0,
      "step": 10575
    },
    {
      "entropy": 5.969550561904907,
      "epoch": 0.924582714323167,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004921265687306059,
      "loss": 5.6649,
      "mean_token_accuracy": 0.1686577558517456,
      "num_tokens": 21440514.0,
      "step": 10580
    },
    {
      "entropy": 5.930572414398194,
      "epoch": 0.9250196626758717,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004921183954535844,
      "loss": 5.5226,
      "mean_token_accuracy": 0.18524338006973268,
      "num_tokens": 21450647.0,
      "step": 10585
    },
    {
      "entropy": 5.8734088897705075,
      "epoch": 0.9254566110285765,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004921102180121184,
      "loss": 5.5656,
      "mean_token_accuracy": 0.17663733214139937,
      "num_tokens": 21460233.0,
      "step": 10590
    },
    {
      "entropy": 5.9221912860870365,
      "epoch": 0.9258935593812812,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004921020364063646,
      "loss": 5.5575,
      "mean_token_accuracy": 0.1818414017558098,
      "num_tokens": 21470207.0,
      "step": 10595
    },
    {
      "entropy": 5.893575954437256,
      "epoch": 0.9263305077339858,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004920938506364801,
      "loss": 5.5843,
      "mean_token_accuracy": 0.18130942583084106,
      "num_tokens": 21480658.0,
      "step": 10600
    },
    {
      "entropy": 5.883734607696534,
      "epoch": 0.9267674560866905,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004920856607026216,
      "loss": 5.4807,
      "mean_token_accuracy": 0.19019529223442078,
      "num_tokens": 21490566.0,
      "step": 10605
    },
    {
      "entropy": 5.933491992950439,
      "epoch": 0.9272044044393952,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004920774666049465,
      "loss": 5.5898,
      "mean_token_accuracy": 0.17951607406139375,
      "num_tokens": 21500752.0,
      "step": 10610
    },
    {
      "entropy": 5.919577741622925,
      "epoch": 0.9276413527921,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004920692683436118,
      "loss": 5.5567,
      "mean_token_accuracy": 0.17595620453357697,
      "num_tokens": 21509802.0,
      "step": 10615
    },
    {
      "entropy": 5.901606178283691,
      "epoch": 0.9280783011448047,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004920610659187748,
      "loss": 5.6537,
      "mean_token_accuracy": 0.16913686841726303,
      "num_tokens": 21520962.0,
      "step": 10620
    },
    {
      "entropy": 5.988981056213379,
      "epoch": 0.9285152494975094,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004920528593305929,
      "loss": 5.6142,
      "mean_token_accuracy": 0.17060058563947678,
      "num_tokens": 21530580.0,
      "step": 10625
    },
    {
      "entropy": 5.900165987014771,
      "epoch": 0.9289521978502141,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004920446485792233,
      "loss": 5.5431,
      "mean_token_accuracy": 0.17997768968343736,
      "num_tokens": 21541431.0,
      "step": 10630
    },
    {
      "entropy": 5.903982305526734,
      "epoch": 0.9293891462029188,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004920364336648236,
      "loss": 5.5572,
      "mean_token_accuracy": 0.17923776507377626,
      "num_tokens": 21551012.0,
      "step": 10635
    },
    {
      "entropy": 5.87684817314148,
      "epoch": 0.9298260945556235,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004920282145875514,
      "loss": 5.5274,
      "mean_token_accuracy": 0.18138542175292968,
      "num_tokens": 21560676.0,
      "step": 10640
    },
    {
      "entropy": 5.860891199111938,
      "epoch": 0.9302630429083283,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004920199913475644,
      "loss": 5.4813,
      "mean_token_accuracy": 0.19159356504678726,
      "num_tokens": 21570848.0,
      "step": 10645
    },
    {
      "entropy": 5.914395904541015,
      "epoch": 0.930699991261033,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004920117639450202,
      "loss": 5.5696,
      "mean_token_accuracy": 0.17611108720302582,
      "num_tokens": 21582358.0,
      "step": 10650
    },
    {
      "entropy": 5.947698879241943,
      "epoch": 0.9311369396137377,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004920035323800767,
      "loss": 5.5086,
      "mean_token_accuracy": 0.1795160010457039,
      "num_tokens": 21591709.0,
      "step": 10655
    },
    {
      "entropy": 5.945560026168823,
      "epoch": 0.9315738879664424,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004919952966528917,
      "loss": 5.4799,
      "mean_token_accuracy": 0.18560876697301865,
      "num_tokens": 21601092.0,
      "step": 10660
    },
    {
      "entropy": 5.886657571792602,
      "epoch": 0.932010836319147,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004919870567636233,
      "loss": 5.5091,
      "mean_token_accuracy": 0.18320509642362595,
      "num_tokens": 21611003.0,
      "step": 10665
    },
    {
      "entropy": 5.931182956695556,
      "epoch": 0.9324477846718517,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004919788127124293,
      "loss": 5.5898,
      "mean_token_accuracy": 0.17421558499336243,
      "num_tokens": 21621528.0,
      "step": 10670
    },
    {
      "entropy": 5.858817720413208,
      "epoch": 0.9328847330245565,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000491970564499468,
      "loss": 5.512,
      "mean_token_accuracy": 0.18707144856452942,
      "num_tokens": 21631431.0,
      "step": 10675
    },
    {
      "entropy": 5.900637769699097,
      "epoch": 0.9333216813772612,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004919623121248976,
      "loss": 5.5304,
      "mean_token_accuracy": 0.18193027824163438,
      "num_tokens": 21640717.0,
      "step": 10680
    },
    {
      "entropy": 5.871455669403076,
      "epoch": 0.9337586297299659,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004919540555888765,
      "loss": 5.5264,
      "mean_token_accuracy": 0.1866111695766449,
      "num_tokens": 21650493.0,
      "step": 10685
    },
    {
      "entropy": 5.903579950332642,
      "epoch": 0.9341955780826706,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004919457948915628,
      "loss": 5.5183,
      "mean_token_accuracy": 0.18695587813854217,
      "num_tokens": 21660062.0,
      "step": 10690
    },
    {
      "entropy": 5.8751038074493405,
      "epoch": 0.9346325264353753,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004919375300331151,
      "loss": 5.4985,
      "mean_token_accuracy": 0.1839493080973625,
      "num_tokens": 21669364.0,
      "step": 10695
    },
    {
      "entropy": 5.914821577072144,
      "epoch": 0.93506947478808,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004919292610136919,
      "loss": 5.5987,
      "mean_token_accuracy": 0.17242753505706787,
      "num_tokens": 21680793.0,
      "step": 10700
    },
    {
      "entropy": 5.936830282211304,
      "epoch": 0.9355064231407848,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004919209878334518,
      "loss": 5.6105,
      "mean_token_accuracy": 0.1778835117816925,
      "num_tokens": 21690862.0,
      "step": 10705
    },
    {
      "entropy": 5.9450860023498535,
      "epoch": 0.9359433714934895,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004919127104925535,
      "loss": 5.5137,
      "mean_token_accuracy": 0.17787016928195953,
      "num_tokens": 21700665.0,
      "step": 10710
    },
    {
      "entropy": 5.904830741882324,
      "epoch": 0.9363803198461942,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004919044289911557,
      "loss": 5.5462,
      "mean_token_accuracy": 0.18308489620685578,
      "num_tokens": 21711128.0,
      "step": 10715
    },
    {
      "entropy": 5.984479093551636,
      "epoch": 0.9368172681988989,
      "grad_norm": 1.0,
      "learning_rate": 0.0004918961433294173,
      "loss": 5.6002,
      "mean_token_accuracy": 0.17397161573171616,
      "num_tokens": 21720233.0,
      "step": 10720
    },
    {
      "entropy": 5.923879766464234,
      "epoch": 0.9372542165516036,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004918878535074973,
      "loss": 5.536,
      "mean_token_accuracy": 0.18505436033010483,
      "num_tokens": 21730980.0,
      "step": 10725
    },
    {
      "entropy": 5.833022737503052,
      "epoch": 0.9376911649043083,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004918795595255545,
      "loss": 5.5535,
      "mean_token_accuracy": 0.18610494136810302,
      "num_tokens": 21742802.0,
      "step": 10730
    },
    {
      "entropy": 5.91872763633728,
      "epoch": 0.9381281132570131,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004918712613837483,
      "loss": 5.5743,
      "mean_token_accuracy": 0.17411006093025208,
      "num_tokens": 21752743.0,
      "step": 10735
    },
    {
      "entropy": 5.962271070480346,
      "epoch": 0.9385650616097178,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004918629590822375,
      "loss": 5.5609,
      "mean_token_accuracy": 0.1836337238550186,
      "num_tokens": 21763198.0,
      "step": 10740
    },
    {
      "entropy": 5.908360338211059,
      "epoch": 0.9390020099624224,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004918546526211816,
      "loss": 5.5463,
      "mean_token_accuracy": 0.18422692865133286,
      "num_tokens": 21773030.0,
      "step": 10745
    },
    {
      "entropy": 5.898901271820068,
      "epoch": 0.9394389583151271,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004918463420007398,
      "loss": 5.5118,
      "mean_token_accuracy": 0.18063011020421982,
      "num_tokens": 21783796.0,
      "step": 10750
    },
    {
      "entropy": 5.950522375106812,
      "epoch": 0.9398759066678318,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004918380272210715,
      "loss": 5.5482,
      "mean_token_accuracy": 0.17867109179496765,
      "num_tokens": 21792504.0,
      "step": 10755
    },
    {
      "entropy": 5.974527788162232,
      "epoch": 0.9403128550205365,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004918297082823362,
      "loss": 5.5963,
      "mean_token_accuracy": 0.17308668792247772,
      "num_tokens": 21802764.0,
      "step": 10760
    },
    {
      "entropy": 5.959614181518555,
      "epoch": 0.9407498033732413,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004918213851846935,
      "loss": 5.5615,
      "mean_token_accuracy": 0.17458115965127946,
      "num_tokens": 21813051.0,
      "step": 10765
    },
    {
      "entropy": 5.905265474319458,
      "epoch": 0.941186751725946,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004918130579283031,
      "loss": 5.6051,
      "mean_token_accuracy": 0.1818806380033493,
      "num_tokens": 21822941.0,
      "step": 10770
    },
    {
      "entropy": 5.953912258148193,
      "epoch": 0.9416237000786507,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004918047265133246,
      "loss": 5.5325,
      "mean_token_accuracy": 0.17471159100532532,
      "num_tokens": 21833225.0,
      "step": 10775
    },
    {
      "entropy": 5.884690046310425,
      "epoch": 0.9420606484313554,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004917963909399178,
      "loss": 5.5185,
      "mean_token_accuracy": 0.18364615589380265,
      "num_tokens": 21842360.0,
      "step": 10780
    },
    {
      "entropy": 5.894737720489502,
      "epoch": 0.9424975967840601,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004917880512082427,
      "loss": 5.542,
      "mean_token_accuracy": 0.1756665289402008,
      "num_tokens": 21852515.0,
      "step": 10785
    },
    {
      "entropy": 5.974491214752197,
      "epoch": 0.9429345451367648,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004917797073184591,
      "loss": 5.6207,
      "mean_token_accuracy": 0.17383513301610948,
      "num_tokens": 21863040.0,
      "step": 10790
    },
    {
      "entropy": 5.957082796096802,
      "epoch": 0.9433714934894696,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004917713592707271,
      "loss": 5.572,
      "mean_token_accuracy": 0.18019585609436034,
      "num_tokens": 21872101.0,
      "step": 10795
    },
    {
      "entropy": 5.943897914886475,
      "epoch": 0.9438084418421743,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004917630070652069,
      "loss": 5.5442,
      "mean_token_accuracy": 0.1823272302746773,
      "num_tokens": 21882148.0,
      "step": 10800
    },
    {
      "entropy": 5.905967330932617,
      "epoch": 0.944245390194879,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004917546507020587,
      "loss": 5.5452,
      "mean_token_accuracy": 0.18321102112531662,
      "num_tokens": 21891832.0,
      "step": 10805
    },
    {
      "entropy": 5.893095541000366,
      "epoch": 0.9446823385475837,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004917462901814427,
      "loss": 5.5308,
      "mean_token_accuracy": 0.18215958923101425,
      "num_tokens": 21901544.0,
      "step": 10810
    },
    {
      "entropy": 5.939043188095093,
      "epoch": 0.9451192869002883,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004917379255035192,
      "loss": 5.5901,
      "mean_token_accuracy": 0.18096174150705338,
      "num_tokens": 21911690.0,
      "step": 10815
    },
    {
      "entropy": 5.96640100479126,
      "epoch": 0.9455562352529931,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004917295566684487,
      "loss": 5.602,
      "mean_token_accuracy": 0.17725116908550262,
      "num_tokens": 21922968.0,
      "step": 10820
    },
    {
      "entropy": 5.861636161804199,
      "epoch": 0.9459931836056978,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004917211836763919,
      "loss": 5.4398,
      "mean_token_accuracy": 0.18481767773628235,
      "num_tokens": 21932904.0,
      "step": 10825
    },
    {
      "entropy": 5.848102807998657,
      "epoch": 0.9464301319584025,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004917128065275091,
      "loss": 5.4433,
      "mean_token_accuracy": 0.19176410585641862,
      "num_tokens": 21943065.0,
      "step": 10830
    },
    {
      "entropy": 5.928055953979492,
      "epoch": 0.9468670803111072,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004917044252219612,
      "loss": 5.5472,
      "mean_token_accuracy": 0.17136322110891342,
      "num_tokens": 21953448.0,
      "step": 10835
    },
    {
      "entropy": 5.9623010635375975,
      "epoch": 0.9473040286638119,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004916960397599088,
      "loss": 5.5334,
      "mean_token_accuracy": 0.18247494846582413,
      "num_tokens": 21963374.0,
      "step": 10840
    },
    {
      "entropy": 5.92945704460144,
      "epoch": 0.9477409770165166,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004916876501415129,
      "loss": 5.573,
      "mean_token_accuracy": 0.1759304955601692,
      "num_tokens": 21973726.0,
      "step": 10845
    },
    {
      "entropy": 5.827088260650635,
      "epoch": 0.9481779253692214,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004916792563669342,
      "loss": 5.4373,
      "mean_token_accuracy": 0.18695189356803893,
      "num_tokens": 21983203.0,
      "step": 10850
    },
    {
      "entropy": 5.910234260559082,
      "epoch": 0.9486148737219261,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004916708584363339,
      "loss": 5.567,
      "mean_token_accuracy": 0.179359632730484,
      "num_tokens": 21993435.0,
      "step": 10855
    },
    {
      "entropy": 5.993073654174805,
      "epoch": 0.9490518220746308,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000491662456349873,
      "loss": 5.5974,
      "mean_token_accuracy": 0.1718185931444168,
      "num_tokens": 22003230.0,
      "step": 10860
    },
    {
      "entropy": 5.948968887329102,
      "epoch": 0.9494887704273355,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004916540501077127,
      "loss": 5.5185,
      "mean_token_accuracy": 0.18363691568374635,
      "num_tokens": 22012910.0,
      "step": 10865
    },
    {
      "entropy": 5.901968097686767,
      "epoch": 0.9499257187800402,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004916456397100141,
      "loss": 5.4987,
      "mean_token_accuracy": 0.1829710692167282,
      "num_tokens": 22022802.0,
      "step": 10870
    },
    {
      "entropy": 5.916840267181397,
      "epoch": 0.9503626671327449,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004916372251569386,
      "loss": 5.6042,
      "mean_token_accuracy": 0.17644574344158173,
      "num_tokens": 22032669.0,
      "step": 10875
    },
    {
      "entropy": 5.894199275970459,
      "epoch": 0.9507996154854497,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004916288064486477,
      "loss": 5.5061,
      "mean_token_accuracy": 0.1878250554203987,
      "num_tokens": 22042477.0,
      "step": 10880
    },
    {
      "entropy": 5.908306980133057,
      "epoch": 0.9512365638381544,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004916203835853027,
      "loss": 5.554,
      "mean_token_accuracy": 0.18295851796865464,
      "num_tokens": 22052891.0,
      "step": 10885
    },
    {
      "entropy": 5.8791405200958256,
      "epoch": 0.951673512190859,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004916119565670652,
      "loss": 5.4977,
      "mean_token_accuracy": 0.19159109890460968,
      "num_tokens": 22062413.0,
      "step": 10890
    },
    {
      "entropy": 5.985524225234985,
      "epoch": 0.9521104605435637,
      "grad_norm": 0.9375,
      "learning_rate": 0.000491603525394097,
      "loss": 5.5816,
      "mean_token_accuracy": 0.176093927025795,
      "num_tokens": 22072565.0,
      "step": 10895
    },
    {
      "entropy": 5.91963210105896,
      "epoch": 0.9525474088962684,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004915950900665596,
      "loss": 5.5593,
      "mean_token_accuracy": 0.17925776094198226,
      "num_tokens": 22082910.0,
      "step": 10900
    },
    {
      "entropy": 5.846152687072754,
      "epoch": 0.9529843572489731,
      "grad_norm": 1.0,
      "learning_rate": 0.000491586650584615,
      "loss": 5.4648,
      "mean_token_accuracy": 0.19186693280935288,
      "num_tokens": 22092636.0,
      "step": 10905
    },
    {
      "entropy": 5.9614451885223385,
      "epoch": 0.9534213056016779,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004915782069484248,
      "loss": 5.5586,
      "mean_token_accuracy": 0.17567274421453477,
      "num_tokens": 22102232.0,
      "step": 10910
    },
    {
      "entropy": 6.01728663444519,
      "epoch": 0.9538582539543826,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004915697591581511,
      "loss": 5.5979,
      "mean_token_accuracy": 0.17239460796117784,
      "num_tokens": 22113592.0,
      "step": 10915
    },
    {
      "entropy": 5.902490234375,
      "epoch": 0.9542952023070873,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000491561307213956,
      "loss": 5.5568,
      "mean_token_accuracy": 0.1808745250105858,
      "num_tokens": 22123731.0,
      "step": 10920
    },
    {
      "entropy": 5.909859609603882,
      "epoch": 0.954732150659792,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004915528511160016,
      "loss": 5.6006,
      "mean_token_accuracy": 0.1727975383400917,
      "num_tokens": 22133801.0,
      "step": 10925
    },
    {
      "entropy": 5.96689281463623,
      "epoch": 0.9551690990124967,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004915443908644501,
      "loss": 5.5613,
      "mean_token_accuracy": 0.17872824519872665,
      "num_tokens": 22143939.0,
      "step": 10930
    },
    {
      "entropy": 5.871942472457886,
      "epoch": 0.9556060473652014,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004915359264594636,
      "loss": 5.5684,
      "mean_token_accuracy": 0.17669312953948973,
      "num_tokens": 22154954.0,
      "step": 10935
    },
    {
      "entropy": 5.919883441925049,
      "epoch": 0.9560429957179062,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004915274579012045,
      "loss": 5.5696,
      "mean_token_accuracy": 0.17766329795122146,
      "num_tokens": 22165225.0,
      "step": 10940
    },
    {
      "entropy": 6.0024439811706545,
      "epoch": 0.9564799440706109,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004915189851898355,
      "loss": 5.6046,
      "mean_token_accuracy": 0.17271756380796432,
      "num_tokens": 22174891.0,
      "step": 10945
    },
    {
      "entropy": 5.842522859573364,
      "epoch": 0.9569168924233156,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004915105083255187,
      "loss": 5.558,
      "mean_token_accuracy": 0.1838728293776512,
      "num_tokens": 22185255.0,
      "step": 10950
    },
    {
      "entropy": 5.882547807693482,
      "epoch": 0.9573538407760203,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000491502027308417,
      "loss": 5.5997,
      "mean_token_accuracy": 0.17581600099802017,
      "num_tokens": 22196124.0,
      "step": 10955
    },
    {
      "entropy": 5.962963724136353,
      "epoch": 0.957790789128725,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000491493542138693,
      "loss": 5.5411,
      "mean_token_accuracy": 0.18111920952796937,
      "num_tokens": 22207563.0,
      "step": 10960
    },
    {
      "entropy": 5.941431999206543,
      "epoch": 0.9582277374814296,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004914850528165094,
      "loss": 5.5663,
      "mean_token_accuracy": 0.17812708765268326,
      "num_tokens": 22218527.0,
      "step": 10965
    },
    {
      "entropy": 5.92325267791748,
      "epoch": 0.9586646858341344,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000491476559342029,
      "loss": 5.5455,
      "mean_token_accuracy": 0.17574211657047273,
      "num_tokens": 22229538.0,
      "step": 10970
    },
    {
      "entropy": 5.944617080688476,
      "epoch": 0.9591016341868391,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004914680617154148,
      "loss": 5.4601,
      "mean_token_accuracy": 0.1898813009262085,
      "num_tokens": 22240047.0,
      "step": 10975
    },
    {
      "entropy": 5.877616739273071,
      "epoch": 0.9595385825395438,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004914595599368297,
      "loss": 5.432,
      "mean_token_accuracy": 0.18729379773139954,
      "num_tokens": 22250440.0,
      "step": 10980
    },
    {
      "entropy": 5.9117156028747555,
      "epoch": 0.9599755308922485,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004914510540064369,
      "loss": 5.5118,
      "mean_token_accuracy": 0.18013764917850494,
      "num_tokens": 22259909.0,
      "step": 10985
    },
    {
      "entropy": 5.936442565917969,
      "epoch": 0.9604124792449532,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004914425439243994,
      "loss": 5.5787,
      "mean_token_accuracy": 0.17472621500492097,
      "num_tokens": 22270358.0,
      "step": 10990
    },
    {
      "entropy": 5.924665355682373,
      "epoch": 0.9608494275976579,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004914340296908805,
      "loss": 5.6029,
      "mean_token_accuracy": 0.17540836632251738,
      "num_tokens": 22280591.0,
      "step": 10995
    },
    {
      "entropy": 5.996507358551026,
      "epoch": 0.9612863759503627,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004914255113060435,
      "loss": 5.5524,
      "mean_token_accuracy": 0.18506253361701966,
      "num_tokens": 22291067.0,
      "step": 11000
    },
    {
      "entropy": 5.892147445678711,
      "epoch": 0.9617233243030674,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004914169887700519,
      "loss": 5.5004,
      "mean_token_accuracy": 0.18513804972171782,
      "num_tokens": 22300919.0,
      "step": 11005
    },
    {
      "entropy": 5.917299604415893,
      "epoch": 0.9621602726557721,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004914084620830689,
      "loss": 5.5986,
      "mean_token_accuracy": 0.1718979075551033,
      "num_tokens": 22311562.0,
      "step": 11010
    },
    {
      "entropy": 5.962935590744019,
      "epoch": 0.9625972210084768,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004913999312452583,
      "loss": 5.5593,
      "mean_token_accuracy": 0.17729598581790923,
      "num_tokens": 22322980.0,
      "step": 11015
    },
    {
      "entropy": 5.934004878997802,
      "epoch": 0.9630341693611815,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004913913962567836,
      "loss": 5.4597,
      "mean_token_accuracy": 0.1862727537751198,
      "num_tokens": 22333242.0,
      "step": 11020
    },
    {
      "entropy": 5.9287238121032715,
      "epoch": 0.9634711177138863,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004913828571178085,
      "loss": 5.5604,
      "mean_token_accuracy": 0.17652868628501892,
      "num_tokens": 22343618.0,
      "step": 11025
    },
    {
      "entropy": 5.8868079662323,
      "epoch": 0.963908066066591,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004913743138284969,
      "loss": 5.5656,
      "mean_token_accuracy": 0.17667529433965684,
      "num_tokens": 22354249.0,
      "step": 11030
    },
    {
      "entropy": 5.983433961868286,
      "epoch": 0.9643450144192957,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004913657663890124,
      "loss": 5.5091,
      "mean_token_accuracy": 0.18353811502456666,
      "num_tokens": 22363824.0,
      "step": 11035
    },
    {
      "entropy": 5.960447835922241,
      "epoch": 0.9647819627720003,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004913572147995193,
      "loss": 5.502,
      "mean_token_accuracy": 0.1826457992196083,
      "num_tokens": 22373470.0,
      "step": 11040
    },
    {
      "entropy": 5.851026248931885,
      "epoch": 0.965218911124705,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004913486590601814,
      "loss": 5.4653,
      "mean_token_accuracy": 0.17790069282054902,
      "num_tokens": 22382999.0,
      "step": 11045
    },
    {
      "entropy": 5.858391046524048,
      "epoch": 0.9656558594774097,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004913400991711629,
      "loss": 5.5219,
      "mean_token_accuracy": 0.1896044582128525,
      "num_tokens": 22393415.0,
      "step": 11050
    },
    {
      "entropy": 5.915917873382568,
      "epoch": 0.9660928078301145,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004913315351326279,
      "loss": 5.4745,
      "mean_token_accuracy": 0.18704261332750322,
      "num_tokens": 22403040.0,
      "step": 11055
    },
    {
      "entropy": 5.8984579086303714,
      "epoch": 0.9665297561828192,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004913229669447406,
      "loss": 5.4945,
      "mean_token_accuracy": 0.18312206864356995,
      "num_tokens": 22413434.0,
      "step": 11060
    },
    {
      "entropy": 5.927411460876465,
      "epoch": 0.9669667045355239,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004913143946076656,
      "loss": 5.6303,
      "mean_token_accuracy": 0.17421640306711197,
      "num_tokens": 22424446.0,
      "step": 11065
    },
    {
      "entropy": 5.935898637771606,
      "epoch": 0.9674036528882286,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004913058181215671,
      "loss": 5.5712,
      "mean_token_accuracy": 0.18013276755809784,
      "num_tokens": 22433928.0,
      "step": 11070
    },
    {
      "entropy": 5.96121621131897,
      "epoch": 0.9678406012409333,
      "grad_norm": 1.0,
      "learning_rate": 0.0004912972374866096,
      "loss": 5.5349,
      "mean_token_accuracy": 0.1785163938999176,
      "num_tokens": 22443778.0,
      "step": 11075
    },
    {
      "entropy": 5.912304401397705,
      "epoch": 0.968277549593638,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004912886527029579,
      "loss": 5.5947,
      "mean_token_accuracy": 0.17785662710666655,
      "num_tokens": 22453423.0,
      "step": 11080
    },
    {
      "entropy": 5.930687999725341,
      "epoch": 0.9687144979463428,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004912800637707764,
      "loss": 5.5116,
      "mean_token_accuracy": 0.18519276678562163,
      "num_tokens": 22464067.0,
      "step": 11085
    },
    {
      "entropy": 5.8831874370574955,
      "epoch": 0.9691514462990475,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00049127147069023,
      "loss": 5.51,
      "mean_token_accuracy": 0.18464227318763732,
      "num_tokens": 22474332.0,
      "step": 11090
    },
    {
      "entropy": 5.934580993652344,
      "epoch": 0.9695883946517522,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004912628734614834,
      "loss": 5.5327,
      "mean_token_accuracy": 0.1794115498661995,
      "num_tokens": 22485304.0,
      "step": 11095
    },
    {
      "entropy": 5.890657520294189,
      "epoch": 0.9700253430044569,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004912542720847016,
      "loss": 5.4477,
      "mean_token_accuracy": 0.18796277791261673,
      "num_tokens": 22494447.0,
      "step": 11100
    },
    {
      "entropy": 5.915380859375,
      "epoch": 0.9704622913571616,
      "grad_norm": 0.875,
      "learning_rate": 0.0004912456665600495,
      "loss": 5.5363,
      "mean_token_accuracy": 0.18665578216314316,
      "num_tokens": 22503936.0,
      "step": 11105
    },
    {
      "entropy": 5.900666666030884,
      "epoch": 0.9708992397098662,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004912370568876923,
      "loss": 5.4376,
      "mean_token_accuracy": 0.18534047454595565,
      "num_tokens": 22513960.0,
      "step": 11110
    },
    {
      "entropy": 5.946606779098511,
      "epoch": 0.971336188062571,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004912284430677949,
      "loss": 5.5837,
      "mean_token_accuracy": 0.1780620902776718,
      "num_tokens": 22522309.0,
      "step": 11115
    },
    {
      "entropy": 5.923249673843384,
      "epoch": 0.9717731364152757,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004912198251005228,
      "loss": 5.475,
      "mean_token_accuracy": 0.18588151186704635,
      "num_tokens": 22533510.0,
      "step": 11120
    },
    {
      "entropy": 5.912736225128174,
      "epoch": 0.9722100847679804,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004912112029860411,
      "loss": 5.5005,
      "mean_token_accuracy": 0.18210818022489547,
      "num_tokens": 22543644.0,
      "step": 11125
    },
    {
      "entropy": 5.969217157363891,
      "epoch": 0.9726470331206851,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004912025767245153,
      "loss": 5.6636,
      "mean_token_accuracy": 0.1679539754986763,
      "num_tokens": 22553475.0,
      "step": 11130
    },
    {
      "entropy": 5.867061805725098,
      "epoch": 0.9730839814733898,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004911939463161108,
      "loss": 5.4606,
      "mean_token_accuracy": 0.18254905045032502,
      "num_tokens": 22563190.0,
      "step": 11135
    },
    {
      "entropy": 5.914308786392212,
      "epoch": 0.9735209298260945,
      "grad_norm": 1.015625,
      "learning_rate": 0.000491185311760993,
      "loss": 5.4895,
      "mean_token_accuracy": 0.18898215591907502,
      "num_tokens": 22572069.0,
      "step": 11140
    },
    {
      "entropy": 5.969358777999878,
      "epoch": 0.9739578781787993,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004911766730593276,
      "loss": 5.6164,
      "mean_token_accuracy": 0.1666291579604149,
      "num_tokens": 22581597.0,
      "step": 11145
    },
    {
      "entropy": 5.956762981414795,
      "epoch": 0.974394826531504,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004911680302112806,
      "loss": 5.4859,
      "mean_token_accuracy": 0.18615066409111022,
      "num_tokens": 22591486.0,
      "step": 11150
    },
    {
      "entropy": 5.948456144332885,
      "epoch": 0.9748317748842087,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004911593832170173,
      "loss": 5.5332,
      "mean_token_accuracy": 0.17513434141874312,
      "num_tokens": 22601615.0,
      "step": 11155
    },
    {
      "entropy": 5.922319650650024,
      "epoch": 0.9752687232369134,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004911507320767038,
      "loss": 5.5404,
      "mean_token_accuracy": 0.1814270481467247,
      "num_tokens": 22611397.0,
      "step": 11160
    },
    {
      "entropy": 5.864880895614624,
      "epoch": 0.9757056715896181,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000491142076790506,
      "loss": 5.5123,
      "mean_token_accuracy": 0.18684805035591126,
      "num_tokens": 22621269.0,
      "step": 11165
    },
    {
      "entropy": 5.963822507858277,
      "epoch": 0.9761426199423228,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004911334173585899,
      "loss": 5.4932,
      "mean_token_accuracy": 0.18352544009685517,
      "num_tokens": 22631782.0,
      "step": 11170
    },
    {
      "entropy": 5.919348096847534,
      "epoch": 0.9765795682950276,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004911247537811216,
      "loss": 5.5848,
      "mean_token_accuracy": 0.1794332444667816,
      "num_tokens": 22642240.0,
      "step": 11175
    },
    {
      "entropy": 5.941284656524658,
      "epoch": 0.9770165166477323,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004911160860582674,
      "loss": 5.5498,
      "mean_token_accuracy": 0.1781550019979477,
      "num_tokens": 22651977.0,
      "step": 11180
    },
    {
      "entropy": 5.978776741027832,
      "epoch": 0.977453465000437,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004911074141901933,
      "loss": 5.5606,
      "mean_token_accuracy": 0.18403149843215943,
      "num_tokens": 22662871.0,
      "step": 11185
    },
    {
      "entropy": 5.859020137786866,
      "epoch": 0.9778904133531416,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004910987381770657,
      "loss": 5.4867,
      "mean_token_accuracy": 0.18818991631269455,
      "num_tokens": 22672083.0,
      "step": 11190
    },
    {
      "entropy": 5.859014415740967,
      "epoch": 0.9783273617058463,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004910900580190512,
      "loss": 5.5133,
      "mean_token_accuracy": 0.18380077332258224,
      "num_tokens": 22681247.0,
      "step": 11195
    },
    {
      "entropy": 5.940107154846191,
      "epoch": 0.9787643100585511,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004910813737163161,
      "loss": 5.5586,
      "mean_token_accuracy": 0.18037173449993132,
      "num_tokens": 22691231.0,
      "step": 11200
    },
    {
      "entropy": 5.940366935729981,
      "epoch": 0.9792012584112558,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000491072685269027,
      "loss": 5.5326,
      "mean_token_accuracy": 0.17829181104898453,
      "num_tokens": 22700712.0,
      "step": 11205
    },
    {
      "entropy": 5.9852437496185305,
      "epoch": 0.9796382067639605,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004910639926773506,
      "loss": 5.6576,
      "mean_token_accuracy": 0.17474732398986817,
      "num_tokens": 22710961.0,
      "step": 11210
    },
    {
      "entropy": 5.905611228942871,
      "epoch": 0.9800751551166652,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004910552959414538,
      "loss": 5.5068,
      "mean_token_accuracy": 0.18612876981496812,
      "num_tokens": 22720450.0,
      "step": 11215
    },
    {
      "entropy": 5.831598901748658,
      "epoch": 0.9805121034693699,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004910465950615029,
      "loss": 5.4669,
      "mean_token_accuracy": 0.18411739766597748,
      "num_tokens": 22730337.0,
      "step": 11220
    },
    {
      "entropy": 5.925642108917236,
      "epoch": 0.9809490518220746,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004910378900376651,
      "loss": 5.5899,
      "mean_token_accuracy": 0.17937243729829788,
      "num_tokens": 22742379.0,
      "step": 11225
    },
    {
      "entropy": 6.011318683624268,
      "epoch": 0.9813860001747794,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004910291808701077,
      "loss": 5.6286,
      "mean_token_accuracy": 0.1666414186358452,
      "num_tokens": 22753176.0,
      "step": 11230
    },
    {
      "entropy": 5.928320741653442,
      "epoch": 0.9818229485274841,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004910204675589972,
      "loss": 5.5418,
      "mean_token_accuracy": 0.1813148409128189,
      "num_tokens": 22763263.0,
      "step": 11235
    },
    {
      "entropy": 5.85734748840332,
      "epoch": 0.9822598968801888,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004910117501045009,
      "loss": 5.5773,
      "mean_token_accuracy": 0.17647044807672502,
      "num_tokens": 22774630.0,
      "step": 11240
    },
    {
      "entropy": 5.99187593460083,
      "epoch": 0.9826968452328935,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004910030285067861,
      "loss": 5.5706,
      "mean_token_accuracy": 0.17803501933813096,
      "num_tokens": 22785027.0,
      "step": 11245
    },
    {
      "entropy": 5.976014471054077,
      "epoch": 0.9831337935855982,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004909943027660201,
      "loss": 5.6069,
      "mean_token_accuracy": 0.1702653631567955,
      "num_tokens": 22796399.0,
      "step": 11250
    },
    {
      "entropy": 5.891661930084228,
      "epoch": 0.9835707419383029,
      "grad_norm": 0.96875,
      "learning_rate": 0.00049098557288237,
      "loss": 5.5304,
      "mean_token_accuracy": 0.17793660014867782,
      "num_tokens": 22806352.0,
      "step": 11255
    },
    {
      "entropy": 5.958200931549072,
      "epoch": 0.9840076902910077,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004909768388560036,
      "loss": 5.6097,
      "mean_token_accuracy": 0.17253497689962388,
      "num_tokens": 22818369.0,
      "step": 11260
    },
    {
      "entropy": 5.933425045013427,
      "epoch": 0.9844446386437123,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004909681006870882,
      "loss": 5.5219,
      "mean_token_accuracy": 0.181375215947628,
      "num_tokens": 22828063.0,
      "step": 11265
    },
    {
      "entropy": 5.888935327529907,
      "epoch": 0.984881586996417,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004909593583757915,
      "loss": 5.5875,
      "mean_token_accuracy": 0.17755479812622071,
      "num_tokens": 22839265.0,
      "step": 11270
    },
    {
      "entropy": 5.905630302429199,
      "epoch": 0.9853185353491217,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004909506119222812,
      "loss": 5.5155,
      "mean_token_accuracy": 0.1882326915860176,
      "num_tokens": 22848501.0,
      "step": 11275
    },
    {
      "entropy": 5.884838390350342,
      "epoch": 0.9857554837018264,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004909418613267248,
      "loss": 5.4815,
      "mean_token_accuracy": 0.18201196789741517,
      "num_tokens": 22859176.0,
      "step": 11280
    },
    {
      "entropy": 5.883367872238159,
      "epoch": 0.9861924320545311,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004909331065892905,
      "loss": 5.4609,
      "mean_token_accuracy": 0.19040533304214477,
      "num_tokens": 22869312.0,
      "step": 11285
    },
    {
      "entropy": 5.958630037307739,
      "epoch": 0.9866293804072359,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000490924347710146,
      "loss": 5.5735,
      "mean_token_accuracy": 0.17513041198253632,
      "num_tokens": 22878699.0,
      "step": 11290
    },
    {
      "entropy": 5.914498233795166,
      "epoch": 0.9870663287599406,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004909155846894595,
      "loss": 5.6,
      "mean_token_accuracy": 0.17679975181818008,
      "num_tokens": 22889147.0,
      "step": 11295
    },
    {
      "entropy": 5.947338438034057,
      "epoch": 0.9875032771126453,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004909068175273989,
      "loss": 5.5126,
      "mean_token_accuracy": 0.18284790515899657,
      "num_tokens": 22899085.0,
      "step": 11300
    },
    {
      "entropy": 5.952857828140258,
      "epoch": 0.98794022546535,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004908980462241323,
      "loss": 5.6357,
      "mean_token_accuracy": 0.17365563958883284,
      "num_tokens": 22909924.0,
      "step": 11305
    },
    {
      "entropy": 6.019389581680298,
      "epoch": 0.9883771738180547,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004908892707798281,
      "loss": 5.6093,
      "mean_token_accuracy": 0.1737401604652405,
      "num_tokens": 22920426.0,
      "step": 11310
    },
    {
      "entropy": 5.932765150070191,
      "epoch": 0.9888141221707594,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004908804911946546,
      "loss": 5.5228,
      "mean_token_accuracy": 0.17751900255680084,
      "num_tokens": 22929929.0,
      "step": 11315
    },
    {
      "entropy": 5.9086424827575685,
      "epoch": 0.9892510705234642,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004908717074687801,
      "loss": 5.4997,
      "mean_token_accuracy": 0.1854547828435898,
      "num_tokens": 22940102.0,
      "step": 11320
    },
    {
      "entropy": 5.955116176605225,
      "epoch": 0.9896880188761689,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004908629196023732,
      "loss": 5.5428,
      "mean_token_accuracy": 0.18591099232435226,
      "num_tokens": 22949490.0,
      "step": 11325
    },
    {
      "entropy": 5.940638160705566,
      "epoch": 0.9901249672288736,
      "grad_norm": 1.0,
      "learning_rate": 0.0004908541275956024,
      "loss": 5.4823,
      "mean_token_accuracy": 0.18347375988960266,
      "num_tokens": 22958119.0,
      "step": 11330
    },
    {
      "entropy": 5.903124475479126,
      "epoch": 0.9905619155815782,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004908453314486363,
      "loss": 5.5131,
      "mean_token_accuracy": 0.17492784857749938,
      "num_tokens": 22969447.0,
      "step": 11335
    },
    {
      "entropy": 5.886647701263428,
      "epoch": 0.9909988639342829,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004908365311616437,
      "loss": 5.4083,
      "mean_token_accuracy": 0.19026709944009781,
      "num_tokens": 22978925.0,
      "step": 11340
    },
    {
      "entropy": 5.932002305984497,
      "epoch": 0.9914358122869876,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004908277267347931,
      "loss": 5.5251,
      "mean_token_accuracy": 0.18577434569597245,
      "num_tokens": 22988861.0,
      "step": 11345
    },
    {
      "entropy": 5.900867748260498,
      "epoch": 0.9918727606396924,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004908189181682538,
      "loss": 5.5695,
      "mean_token_accuracy": 0.17679559588432311,
      "num_tokens": 22998700.0,
      "step": 11350
    },
    {
      "entropy": 5.883777713775634,
      "epoch": 0.9923097089923971,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004908101054621946,
      "loss": 5.51,
      "mean_token_accuracy": 0.18317799717187883,
      "num_tokens": 23009095.0,
      "step": 11355
    },
    {
      "entropy": 5.892273616790772,
      "epoch": 0.9927466573451018,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004908012886167845,
      "loss": 5.5277,
      "mean_token_accuracy": 0.18400360643863678,
      "num_tokens": 23019484.0,
      "step": 11360
    },
    {
      "entropy": 5.887545967102051,
      "epoch": 0.9931836056978065,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004907924676321924,
      "loss": 5.4915,
      "mean_token_accuracy": 0.18223105520009994,
      "num_tokens": 23029867.0,
      "step": 11365
    },
    {
      "entropy": 5.901641321182251,
      "epoch": 0.9936205540505112,
      "grad_norm": 0.96875,
      "learning_rate": 0.000490783642508588,
      "loss": 5.4753,
      "mean_token_accuracy": 0.18627289831638336,
      "num_tokens": 23040179.0,
      "step": 11370
    },
    {
      "entropy": 5.913129615783691,
      "epoch": 0.9940575024032159,
      "grad_norm": 0.984375,
      "learning_rate": 0.00049077481324614,
      "loss": 5.4448,
      "mean_token_accuracy": 0.19312843382358552,
      "num_tokens": 23050389.0,
      "step": 11375
    },
    {
      "entropy": 5.827772331237793,
      "epoch": 0.9944944507559207,
      "grad_norm": 1.046875,
      "learning_rate": 0.000490765979845018,
      "loss": 5.4969,
      "mean_token_accuracy": 0.18963824808597565,
      "num_tokens": 23059984.0,
      "step": 11380
    },
    {
      "entropy": 5.95826358795166,
      "epoch": 0.9949313991086254,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004907571423053917,
      "loss": 5.4921,
      "mean_token_accuracy": 0.17986537218093873,
      "num_tokens": 23069828.0,
      "step": 11385
    },
    {
      "entropy": 5.928144121170044,
      "epoch": 0.9953683474613301,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004907483006274302,
      "loss": 5.5522,
      "mean_token_accuracy": 0.17848511934280395,
      "num_tokens": 23079223.0,
      "step": 11390
    },
    {
      "entropy": 5.895743465423584,
      "epoch": 0.9958052958140348,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004907394548113033,
      "loss": 5.5843,
      "mean_token_accuracy": 0.17392936646938323,
      "num_tokens": 23088540.0,
      "step": 11395
    },
    {
      "entropy": 5.965622997283935,
      "epoch": 0.9962422441667395,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004907306048571806,
      "loss": 5.6264,
      "mean_token_accuracy": 0.17307165414094924,
      "num_tokens": 23098617.0,
      "step": 11400
    },
    {
      "entropy": 5.960770750045777,
      "epoch": 0.9966791925194443,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004907217507652317,
      "loss": 5.5162,
      "mean_token_accuracy": 0.18303609937429427,
      "num_tokens": 23109702.0,
      "step": 11405
    },
    {
      "entropy": 5.940365362167358,
      "epoch": 0.997116140872149,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004907128925356268,
      "loss": 5.5356,
      "mean_token_accuracy": 0.18168227076530458,
      "num_tokens": 23120295.0,
      "step": 11410
    },
    {
      "entropy": 5.925732898712158,
      "epoch": 0.9975530892248536,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004907040301685355,
      "loss": 5.5587,
      "mean_token_accuracy": 0.1786484569311142,
      "num_tokens": 23131999.0,
      "step": 11415
    },
    {
      "entropy": 5.962254190444947,
      "epoch": 0.9979900375775583,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004906951636641279,
      "loss": 5.573,
      "mean_token_accuracy": 0.1862677350640297,
      "num_tokens": 23141060.0,
      "step": 11420
    },
    {
      "entropy": 5.8779665470123295,
      "epoch": 0.998426985930263,
      "grad_norm": 0.96875,
      "learning_rate": 0.000490686293022574,
      "loss": 5.5254,
      "mean_token_accuracy": 0.179360468685627,
      "num_tokens": 23151607.0,
      "step": 11425
    },
    {
      "entropy": 5.915169382095337,
      "epoch": 0.9988639342829677,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000490677418244044,
      "loss": 5.5857,
      "mean_token_accuracy": 0.17325692474842072,
      "num_tokens": 23161892.0,
      "step": 11430
    },
    {
      "entropy": 5.932821559906006,
      "epoch": 0.9993008826356725,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004906685393287081,
      "loss": 5.5776,
      "mean_token_accuracy": 0.17932833582162858,
      "num_tokens": 23172112.0,
      "step": 11435
    },
    {
      "entropy": 5.9388329029083256,
      "epoch": 0.9997378309883772,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004906596562767367,
      "loss": 5.457,
      "mean_token_accuracy": 0.18418536633253096,
      "num_tokens": 23181435.0,
      "step": 11440
    },
    {
      "entropy": 5.878952693939209,
      "epoch": 1.0001747793410818,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004906507690883,
      "loss": 5.4161,
      "mean_token_accuracy": 0.18378300368785858,
      "num_tokens": 23191084.0,
      "step": 11445
    },
    {
      "entropy": 5.863998603820801,
      "epoch": 1.0006117276937867,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004906418777635685,
      "loss": 5.4615,
      "mean_token_accuracy": 0.1867107167840004,
      "num_tokens": 23202372.0,
      "step": 11450
    },
    {
      "entropy": 5.950961637496948,
      "epoch": 1.0010486760464914,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004906329823027129,
      "loss": 5.5119,
      "mean_token_accuracy": 0.18363516479730607,
      "num_tokens": 23214113.0,
      "step": 11455
    },
    {
      "entropy": 5.929176616668701,
      "epoch": 1.001485624399196,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004906240827059036,
      "loss": 5.4099,
      "mean_token_accuracy": 0.1879650503396988,
      "num_tokens": 23223562.0,
      "step": 11460
    },
    {
      "entropy": 5.947253274917602,
      "epoch": 1.0019225727519008,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004906151789733114,
      "loss": 5.5037,
      "mean_token_accuracy": 0.18112244904041291,
      "num_tokens": 23232901.0,
      "step": 11465
    },
    {
      "entropy": 5.925134563446045,
      "epoch": 1.0023595211046055,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004906062711051073,
      "loss": 5.516,
      "mean_token_accuracy": 0.17965808510780334,
      "num_tokens": 23242869.0,
      "step": 11470
    },
    {
      "entropy": 5.899933528900147,
      "epoch": 1.0027964694573102,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004905973591014617,
      "loss": 5.4175,
      "mean_token_accuracy": 0.19358028173446656,
      "num_tokens": 23253653.0,
      "step": 11475
    },
    {
      "entropy": 5.867020320892334,
      "epoch": 1.0032334178100148,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004905884429625459,
      "loss": 5.4377,
      "mean_token_accuracy": 0.18545135259628295,
      "num_tokens": 23263530.0,
      "step": 11480
    },
    {
      "entropy": 5.930663251876831,
      "epoch": 1.0036703661627195,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004905795226885308,
      "loss": 5.4735,
      "mean_token_accuracy": 0.18357979208230973,
      "num_tokens": 23274946.0,
      "step": 11485
    },
    {
      "entropy": 5.970111560821533,
      "epoch": 1.0041073145154242,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004905705982795874,
      "loss": 5.4716,
      "mean_token_accuracy": 0.1872389942407608,
      "num_tokens": 23283811.0,
      "step": 11490
    },
    {
      "entropy": 5.8560370922088625,
      "epoch": 1.004544262868129,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000490561669735887,
      "loss": 5.5117,
      "mean_token_accuracy": 0.179312002658844,
      "num_tokens": 23294554.0,
      "step": 11495
    },
    {
      "entropy": 5.890764379501343,
      "epoch": 1.0049812112208336,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004905527370576009,
      "loss": 5.3525,
      "mean_token_accuracy": 0.18988146781921386,
      "num_tokens": 23304245.0,
      "step": 11500
    },
    {
      "entropy": 5.992785263061523,
      "epoch": 1.0054181595735383,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004905438002449004,
      "loss": 5.5166,
      "mean_token_accuracy": 0.17497579157352447,
      "num_tokens": 23315169.0,
      "step": 11505
    },
    {
      "entropy": 5.891326951980591,
      "epoch": 1.0058551079262432,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004905348592979567,
      "loss": 5.4084,
      "mean_token_accuracy": 0.18999587893486022,
      "num_tokens": 23325528.0,
      "step": 11510
    },
    {
      "entropy": 5.924654579162597,
      "epoch": 1.006292056278948,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004905259142169415,
      "loss": 5.5507,
      "mean_token_accuracy": 0.18072817474603653,
      "num_tokens": 23336744.0,
      "step": 11515
    },
    {
      "entropy": 5.903971719741821,
      "epoch": 1.0067290046316526,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004905169650020264,
      "loss": 5.4117,
      "mean_token_accuracy": 0.18949124217033386,
      "num_tokens": 23346092.0,
      "step": 11520
    },
    {
      "entropy": 5.865019941329956,
      "epoch": 1.0071659529843573,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004905080116533831,
      "loss": 5.4325,
      "mean_token_accuracy": 0.18078796565532684,
      "num_tokens": 23357326.0,
      "step": 11525
    },
    {
      "entropy": 5.913744258880615,
      "epoch": 1.007602901337062,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000490499054171183,
      "loss": 5.4453,
      "mean_token_accuracy": 0.1874909907579422,
      "num_tokens": 23367789.0,
      "step": 11530
    },
    {
      "entropy": 5.9354277610778805,
      "epoch": 1.0080398496897667,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004904900925555982,
      "loss": 5.451,
      "mean_token_accuracy": 0.18200158923864365,
      "num_tokens": 23377499.0,
      "step": 11535
    },
    {
      "entropy": 5.929340410232544,
      "epoch": 1.0084767980424714,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004904811268068005,
      "loss": 5.4264,
      "mean_token_accuracy": 0.1829432249069214,
      "num_tokens": 23387867.0,
      "step": 11540
    },
    {
      "entropy": 5.880160808563232,
      "epoch": 1.008913746395176,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004904721569249618,
      "loss": 5.5363,
      "mean_token_accuracy": 0.1826999753713608,
      "num_tokens": 23398179.0,
      "step": 11545
    },
    {
      "entropy": 5.9578992366790775,
      "epoch": 1.0093506947478808,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004904631829102543,
      "loss": 5.524,
      "mean_token_accuracy": 0.18542693853378295,
      "num_tokens": 23407407.0,
      "step": 11550
    },
    {
      "entropy": 5.93063907623291,
      "epoch": 1.0097876431005854,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004904542047628501,
      "loss": 5.368,
      "mean_token_accuracy": 0.19032351076602935,
      "num_tokens": 23417324.0,
      "step": 11555
    },
    {
      "entropy": 5.908033657073974,
      "epoch": 1.0102245914532901,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004904452224829212,
      "loss": 5.4248,
      "mean_token_accuracy": 0.19040575474500657,
      "num_tokens": 23427157.0,
      "step": 11560
    },
    {
      "entropy": 5.913088798522949,
      "epoch": 1.0106615398059948,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004904362360706403,
      "loss": 5.5087,
      "mean_token_accuracy": 0.18467679768800735,
      "num_tokens": 23437712.0,
      "step": 11565
    },
    {
      "entropy": 5.927874612808227,
      "epoch": 1.0110984881586997,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004904272455261793,
      "loss": 5.4363,
      "mean_token_accuracy": 0.18667925000190735,
      "num_tokens": 23447949.0,
      "step": 11570
    },
    {
      "entropy": 5.897248458862305,
      "epoch": 1.0115354365114044,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004904182508497109,
      "loss": 5.3863,
      "mean_token_accuracy": 0.18655433207750322,
      "num_tokens": 23457319.0,
      "step": 11575
    },
    {
      "entropy": 5.927804851531983,
      "epoch": 1.0119723848641091,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004904092520414075,
      "loss": 5.5081,
      "mean_token_accuracy": 0.17955856323242186,
      "num_tokens": 23468245.0,
      "step": 11580
    },
    {
      "entropy": 5.965659666061401,
      "epoch": 1.0124093332168138,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004904002491014418,
      "loss": 5.4689,
      "mean_token_accuracy": 0.1921745091676712,
      "num_tokens": 23477049.0,
      "step": 11585
    },
    {
      "entropy": 5.925404357910156,
      "epoch": 1.0128462815695185,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004903912420299866,
      "loss": 5.5589,
      "mean_token_accuracy": 0.17297181487083435,
      "num_tokens": 23488574.0,
      "step": 11590
    },
    {
      "entropy": 5.945957899093628,
      "epoch": 1.0132832299222232,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004903822308272144,
      "loss": 5.4706,
      "mean_token_accuracy": 0.1876446411013603,
      "num_tokens": 23499282.0,
      "step": 11595
    },
    {
      "entropy": 5.920729446411133,
      "epoch": 1.0137201782749279,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004903732154932982,
      "loss": 5.4989,
      "mean_token_accuracy": 0.19002268165349961,
      "num_tokens": 23508777.0,
      "step": 11600
    },
    {
      "entropy": 5.8889165878295895,
      "epoch": 1.0141571266276326,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004903641960284108,
      "loss": 5.4453,
      "mean_token_accuracy": 0.1909979984164238,
      "num_tokens": 23518818.0,
      "step": 11605
    },
    {
      "entropy": 5.9028199195861815,
      "epoch": 1.0145940749803373,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004903551724327254,
      "loss": 5.4419,
      "mean_token_accuracy": 0.18583340793848038,
      "num_tokens": 23529668.0,
      "step": 11610
    },
    {
      "entropy": 5.99611644744873,
      "epoch": 1.015031023333042,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004903461447064147,
      "loss": 5.437,
      "mean_token_accuracy": 0.1853141278028488,
      "num_tokens": 23540418.0,
      "step": 11615
    },
    {
      "entropy": 5.929587650299072,
      "epoch": 1.0154679716857467,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004903371128496523,
      "loss": 5.4218,
      "mean_token_accuracy": 0.1895105078816414,
      "num_tokens": 23550035.0,
      "step": 11620
    },
    {
      "entropy": 5.890632104873657,
      "epoch": 1.0159049200384516,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004903280768626113,
      "loss": 5.4756,
      "mean_token_accuracy": 0.18309216052293778,
      "num_tokens": 23559487.0,
      "step": 11625
    },
    {
      "entropy": 5.902059030532837,
      "epoch": 1.0163418683911563,
      "grad_norm": 1.125,
      "learning_rate": 0.0004903190367454648,
      "loss": 5.4118,
      "mean_token_accuracy": 0.19304501861333848,
      "num_tokens": 23569146.0,
      "step": 11630
    },
    {
      "entropy": 5.868683338165283,
      "epoch": 1.016778816743861,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004903099924983864,
      "loss": 5.3651,
      "mean_token_accuracy": 0.1880990371108055,
      "num_tokens": 23579458.0,
      "step": 11635
    },
    {
      "entropy": 5.88977837562561,
      "epoch": 1.0172157650965656,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004903009441215495,
      "loss": 5.4142,
      "mean_token_accuracy": 0.18474953770637512,
      "num_tokens": 23589416.0,
      "step": 11640
    },
    {
      "entropy": 5.926626396179199,
      "epoch": 1.0176527134492703,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004902918916151278,
      "loss": 5.4988,
      "mean_token_accuracy": 0.18681641072034835,
      "num_tokens": 23599588.0,
      "step": 11645
    },
    {
      "entropy": 5.964414834976196,
      "epoch": 1.018089661801975,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004902828349792947,
      "loss": 5.4864,
      "mean_token_accuracy": 0.18353132903575897,
      "num_tokens": 23610116.0,
      "step": 11650
    },
    {
      "entropy": 5.871355056762695,
      "epoch": 1.0185266101546797,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000490273774214224,
      "loss": 5.3954,
      "mean_token_accuracy": 0.1919994905591011,
      "num_tokens": 23619906.0,
      "step": 11655
    },
    {
      "entropy": 5.894396829605102,
      "epoch": 1.0189635585073844,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004902647093200896,
      "loss": 5.4439,
      "mean_token_accuracy": 0.18812565207481385,
      "num_tokens": 23630351.0,
      "step": 11660
    },
    {
      "entropy": 5.916657257080078,
      "epoch": 1.019400506860089,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004902556402970652,
      "loss": 5.5076,
      "mean_token_accuracy": 0.18387048095464706,
      "num_tokens": 23639696.0,
      "step": 11665
    },
    {
      "entropy": 5.908682107925415,
      "epoch": 1.0198374552127938,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004902465671453248,
      "loss": 5.4696,
      "mean_token_accuracy": 0.1837886780500412,
      "num_tokens": 23649388.0,
      "step": 11670
    },
    {
      "entropy": 5.948295640945434,
      "epoch": 1.0202744035654985,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004902374898650425,
      "loss": 5.5036,
      "mean_token_accuracy": 0.18383323550224304,
      "num_tokens": 23659125.0,
      "step": 11675
    },
    {
      "entropy": 6.002316570281982,
      "epoch": 1.0207113519182032,
      "grad_norm": 1.0,
      "learning_rate": 0.0004902284084563924,
      "loss": 5.4878,
      "mean_token_accuracy": 0.18038416355848313,
      "num_tokens": 23668817.0,
      "step": 11680
    },
    {
      "entropy": 5.89922776222229,
      "epoch": 1.021148300270908,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004902193229195487,
      "loss": 5.4878,
      "mean_token_accuracy": 0.1847448989748955,
      "num_tokens": 23679378.0,
      "step": 11685
    },
    {
      "entropy": 5.905912590026856,
      "epoch": 1.0215852486236128,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004902102332546855,
      "loss": 5.4481,
      "mean_token_accuracy": 0.18251367062330245,
      "num_tokens": 23689400.0,
      "step": 11690
    },
    {
      "entropy": 6.007722282409668,
      "epoch": 1.0220221969763175,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004902011394619772,
      "loss": 5.5848,
      "mean_token_accuracy": 0.17402046620845796,
      "num_tokens": 23699503.0,
      "step": 11695
    },
    {
      "entropy": 5.926589918136597,
      "epoch": 1.0224591453290222,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004901920415415985,
      "loss": 5.5071,
      "mean_token_accuracy": 0.18302029818296434,
      "num_tokens": 23710913.0,
      "step": 11700
    },
    {
      "entropy": 5.867807674407959,
      "epoch": 1.0228960936817268,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004901829394937234,
      "loss": 5.4128,
      "mean_token_accuracy": 0.1868189424276352,
      "num_tokens": 23720730.0,
      "step": 11705
    },
    {
      "entropy": 5.945699453353882,
      "epoch": 1.0233330420344315,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000490173833318527,
      "loss": 5.4774,
      "mean_token_accuracy": 0.18129601180553437,
      "num_tokens": 23730902.0,
      "step": 11710
    },
    {
      "entropy": 5.919826936721802,
      "epoch": 1.0237699903871362,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004901647230161837,
      "loss": 5.5232,
      "mean_token_accuracy": 0.17977993488311766,
      "num_tokens": 23741901.0,
      "step": 11715
    },
    {
      "entropy": 5.949140930175782,
      "epoch": 1.024206938739841,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004901556085868681,
      "loss": 5.4083,
      "mean_token_accuracy": 0.18570256531238555,
      "num_tokens": 23752029.0,
      "step": 11720
    },
    {
      "entropy": 5.871209239959716,
      "epoch": 1.0246438870925456,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004901464900307553,
      "loss": 5.4126,
      "mean_token_accuracy": 0.1901086151599884,
      "num_tokens": 23762125.0,
      "step": 11725
    },
    {
      "entropy": 5.89410514831543,
      "epoch": 1.0250808354452503,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004901373673480201,
      "loss": 5.4562,
      "mean_token_accuracy": 0.18741149306297303,
      "num_tokens": 23772056.0,
      "step": 11730
    },
    {
      "entropy": 5.907648992538452,
      "epoch": 1.025517783797955,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004901282405388375,
      "loss": 5.4043,
      "mean_token_accuracy": 0.1883814036846161,
      "num_tokens": 23781441.0,
      "step": 11735
    },
    {
      "entropy": 5.928481817245483,
      "epoch": 1.0259547321506597,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004901191096033825,
      "loss": 5.4005,
      "mean_token_accuracy": 0.1873543679714203,
      "num_tokens": 23791472.0,
      "step": 11740
    },
    {
      "entropy": 5.921793508529663,
      "epoch": 1.0263916805033646,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004901099745418302,
      "loss": 5.431,
      "mean_token_accuracy": 0.18412315845489502,
      "num_tokens": 23801940.0,
      "step": 11745
    },
    {
      "entropy": 5.854950475692749,
      "epoch": 1.0268286288560693,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004901008353543559,
      "loss": 5.3542,
      "mean_token_accuracy": 0.19171438217163086,
      "num_tokens": 23811515.0,
      "step": 11750
    },
    {
      "entropy": 5.893281698226929,
      "epoch": 1.027265577208774,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004900916920411349,
      "loss": 5.4494,
      "mean_token_accuracy": 0.18134599328041076,
      "num_tokens": 23821871.0,
      "step": 11755
    },
    {
      "entropy": 5.89902548789978,
      "epoch": 1.0277025255614787,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004900825446023426,
      "loss": 5.3913,
      "mean_token_accuracy": 0.19211844056844712,
      "num_tokens": 23831677.0,
      "step": 11760
    },
    {
      "entropy": 5.902160406112671,
      "epoch": 1.0281394739141834,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004900733930381543,
      "loss": 5.45,
      "mean_token_accuracy": 0.1878408297896385,
      "num_tokens": 23841415.0,
      "step": 11765
    },
    {
      "entropy": 5.94578971862793,
      "epoch": 1.028576422266888,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004900642373487455,
      "loss": 5.5088,
      "mean_token_accuracy": 0.18511707037687303,
      "num_tokens": 23851809.0,
      "step": 11770
    },
    {
      "entropy": 5.969323635101318,
      "epoch": 1.0290133706195927,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004900550775342921,
      "loss": 5.445,
      "mean_token_accuracy": 0.18021745681762696,
      "num_tokens": 23861981.0,
      "step": 11775
    },
    {
      "entropy": 5.895697689056396,
      "epoch": 1.0294503189722974,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004900459135949697,
      "loss": 5.4256,
      "mean_token_accuracy": 0.18650394529104233,
      "num_tokens": 23872437.0,
      "step": 11780
    },
    {
      "entropy": 5.898399496078492,
      "epoch": 1.0298872673250021,
      "grad_norm": 1.0,
      "learning_rate": 0.0004900367455309537,
      "loss": 5.4691,
      "mean_token_accuracy": 0.18484245985746384,
      "num_tokens": 23881650.0,
      "step": 11785
    },
    {
      "entropy": 5.919484996795655,
      "epoch": 1.0303242156777068,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004900275733424205,
      "loss": 5.4638,
      "mean_token_accuracy": 0.18415253460407258,
      "num_tokens": 23891631.0,
      "step": 11790
    },
    {
      "entropy": 5.8876666069030765,
      "epoch": 1.0307611640304115,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004900183970295457,
      "loss": 5.4141,
      "mean_token_accuracy": 0.19041501730680466,
      "num_tokens": 23901725.0,
      "step": 11795
    },
    {
      "entropy": 5.917621707916259,
      "epoch": 1.0311981123831164,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004900092165925053,
      "loss": 5.4444,
      "mean_token_accuracy": 0.18700506538152695,
      "num_tokens": 23911850.0,
      "step": 11800
    },
    {
      "entropy": 5.964566946029663,
      "epoch": 1.0316350607358211,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004900000320314757,
      "loss": 5.5282,
      "mean_token_accuracy": 0.17714182734489442,
      "num_tokens": 23922571.0,
      "step": 11805
    },
    {
      "entropy": 5.89772310256958,
      "epoch": 1.0320720090885258,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004899908433466326,
      "loss": 5.4387,
      "mean_token_accuracy": 0.1876651883125305,
      "num_tokens": 23932637.0,
      "step": 11810
    },
    {
      "entropy": 5.986398029327392,
      "epoch": 1.0325089574412305,
      "grad_norm": 1.0,
      "learning_rate": 0.0004899816505381525,
      "loss": 5.5427,
      "mean_token_accuracy": 0.17861911952495574,
      "num_tokens": 23943288.0,
      "step": 11815
    },
    {
      "entropy": 5.928348016738892,
      "epoch": 1.0329459057939352,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004899724536062117,
      "loss": 5.4104,
      "mean_token_accuracy": 0.18426510989665984,
      "num_tokens": 23952853.0,
      "step": 11820
    },
    {
      "entropy": 5.867315483093262,
      "epoch": 1.0333828541466399,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004899632525509867,
      "loss": 5.4739,
      "mean_token_accuracy": 0.18329878747463227,
      "num_tokens": 23963111.0,
      "step": 11825
    },
    {
      "entropy": 5.887027788162231,
      "epoch": 1.0338198024993446,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004899540473726538,
      "loss": 5.3709,
      "mean_token_accuracy": 0.19364586472511292,
      "num_tokens": 23972468.0,
      "step": 11830
    },
    {
      "entropy": 5.95546178817749,
      "epoch": 1.0342567508520493,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004899448380713897,
      "loss": 5.4254,
      "mean_token_accuracy": 0.19258967638015748,
      "num_tokens": 23982567.0,
      "step": 11835
    },
    {
      "entropy": 5.888498115539551,
      "epoch": 1.034693699204754,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004899356246473709,
      "loss": 5.4269,
      "mean_token_accuracy": 0.189495849609375,
      "num_tokens": 23993283.0,
      "step": 11840
    },
    {
      "entropy": 5.858783626556397,
      "epoch": 1.0351306475574586,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004899264071007741,
      "loss": 5.4088,
      "mean_token_accuracy": 0.18948449790477753,
      "num_tokens": 24004504.0,
      "step": 11845
    },
    {
      "entropy": 5.9289826393127445,
      "epoch": 1.0355675959101633,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004899171854317764,
      "loss": 5.4546,
      "mean_token_accuracy": 0.18899242132902144,
      "num_tokens": 24014557.0,
      "step": 11850
    },
    {
      "entropy": 5.898662757873535,
      "epoch": 1.036004544262868,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004899079596405543,
      "loss": 5.4463,
      "mean_token_accuracy": 0.17924203276634215,
      "num_tokens": 24024495.0,
      "step": 11855
    },
    {
      "entropy": 5.849420118331909,
      "epoch": 1.036441492615573,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000489898729727285,
      "loss": 5.4544,
      "mean_token_accuracy": 0.18621449917554855,
      "num_tokens": 24034984.0,
      "step": 11860
    },
    {
      "entropy": 5.903042459487915,
      "epoch": 1.0368784409682776,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004898894956921455,
      "loss": 5.4006,
      "mean_token_accuracy": 0.19150213450193404,
      "num_tokens": 24044379.0,
      "step": 11865
    },
    {
      "entropy": 5.961497592926025,
      "epoch": 1.0373153893209823,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004898802575353129,
      "loss": 5.4512,
      "mean_token_accuracy": 0.1809087723493576,
      "num_tokens": 24054133.0,
      "step": 11870
    },
    {
      "entropy": 5.838613605499267,
      "epoch": 1.037752337673687,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004898710152569642,
      "loss": 5.3809,
      "mean_token_accuracy": 0.18720731735229493,
      "num_tokens": 24064868.0,
      "step": 11875
    },
    {
      "entropy": 5.927127838134766,
      "epoch": 1.0381892860263917,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004898617688572769,
      "loss": 5.4591,
      "mean_token_accuracy": 0.18204597532749175,
      "num_tokens": 24074977.0,
      "step": 11880
    },
    {
      "entropy": 5.906432390213013,
      "epoch": 1.0386262343790964,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004898525183364284,
      "loss": 5.5113,
      "mean_token_accuracy": 0.18419696986675263,
      "num_tokens": 24085673.0,
      "step": 11885
    },
    {
      "entropy": 5.927881669998169,
      "epoch": 1.039063182731801,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004898432636945959,
      "loss": 5.4819,
      "mean_token_accuracy": 0.1859409838914871,
      "num_tokens": 24095465.0,
      "step": 11890
    },
    {
      "entropy": 5.888664722442627,
      "epoch": 1.0395001310845058,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000489834004931957,
      "loss": 5.4238,
      "mean_token_accuracy": 0.18510760068893434,
      "num_tokens": 24105576.0,
      "step": 11895
    },
    {
      "entropy": 5.943706798553467,
      "epoch": 1.0399370794372105,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004898247420486895,
      "loss": 5.4262,
      "mean_token_accuracy": 0.18240769654512407,
      "num_tokens": 24114998.0,
      "step": 11900
    },
    {
      "entropy": 5.909370374679566,
      "epoch": 1.0403740277899152,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004898154750449707,
      "loss": 5.4173,
      "mean_token_accuracy": 0.19176352322101592,
      "num_tokens": 24125059.0,
      "step": 11905
    },
    {
      "entropy": 5.901481580734253,
      "epoch": 1.0408109761426199,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004898062039209786,
      "loss": 5.4414,
      "mean_token_accuracy": 0.18806448876857756,
      "num_tokens": 24134062.0,
      "step": 11910
    },
    {
      "entropy": 5.930304288864136,
      "epoch": 1.0412479244953246,
      "grad_norm": 1.28125,
      "learning_rate": 0.000489796928676891,
      "loss": 5.4663,
      "mean_token_accuracy": 0.18391460031270981,
      "num_tokens": 24144016.0,
      "step": 11915
    },
    {
      "entropy": 5.928962373733521,
      "epoch": 1.0416848728480295,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004897876493128856,
      "loss": 5.4316,
      "mean_token_accuracy": 0.18730899691581726,
      "num_tokens": 24154337.0,
      "step": 11920
    },
    {
      "entropy": 5.884474849700927,
      "epoch": 1.0421218212007342,
      "grad_norm": 1.0,
      "learning_rate": 0.0004897783658291406,
      "loss": 5.4036,
      "mean_token_accuracy": 0.19233931005001068,
      "num_tokens": 24164440.0,
      "step": 11925
    },
    {
      "entropy": 5.921966981887818,
      "epoch": 1.0425587695534388,
      "grad_norm": 1.078125,
      "learning_rate": 0.000489769078225834,
      "loss": 5.5819,
      "mean_token_accuracy": 0.17364779710769654,
      "num_tokens": 24174388.0,
      "step": 11930
    },
    {
      "entropy": 5.883465003967285,
      "epoch": 1.0429957179061435,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004897597865031441,
      "loss": 5.3889,
      "mean_token_accuracy": 0.19058220982551574,
      "num_tokens": 24183543.0,
      "step": 11935
    },
    {
      "entropy": 5.924093389511109,
      "epoch": 1.0434326662588482,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004897504906612487,
      "loss": 5.4464,
      "mean_token_accuracy": 0.18977826535701753,
      "num_tokens": 24193847.0,
      "step": 11940
    },
    {
      "entropy": 5.936905145645142,
      "epoch": 1.043869614611553,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004897411907003265,
      "loss": 5.4326,
      "mean_token_accuracy": 0.1841369703412056,
      "num_tokens": 24203072.0,
      "step": 11945
    },
    {
      "entropy": 5.869912767410279,
      "epoch": 1.0443065629642576,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004897318866205558,
      "loss": 5.3498,
      "mean_token_accuracy": 0.19295794516801834,
      "num_tokens": 24213610.0,
      "step": 11950
    },
    {
      "entropy": 5.822430038452149,
      "epoch": 1.0447435113169623,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004897225784221149,
      "loss": 5.4298,
      "mean_token_accuracy": 0.19083942770957946,
      "num_tokens": 24223571.0,
      "step": 11955
    },
    {
      "entropy": 5.855165672302246,
      "epoch": 1.045180459669667,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004897132661051826,
      "loss": 5.4558,
      "mean_token_accuracy": 0.1849847063422203,
      "num_tokens": 24233867.0,
      "step": 11960
    },
    {
      "entropy": 5.929189825057984,
      "epoch": 1.0456174080223717,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004897039496699372,
      "loss": 5.472,
      "mean_token_accuracy": 0.18431731462478637,
      "num_tokens": 24243853.0,
      "step": 11965
    },
    {
      "entropy": 5.89468469619751,
      "epoch": 1.0460543563750764,
      "grad_norm": 1.0,
      "learning_rate": 0.0004896946291165575,
      "loss": 5.43,
      "mean_token_accuracy": 0.18318591713905336,
      "num_tokens": 24254049.0,
      "step": 11970
    },
    {
      "entropy": 5.8844475746154785,
      "epoch": 1.0464913047277813,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004896853044452224,
      "loss": 5.4104,
      "mean_token_accuracy": 0.1874416872859001,
      "num_tokens": 24265166.0,
      "step": 11975
    },
    {
      "entropy": 5.90068941116333,
      "epoch": 1.046928253080486,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004896759756561106,
      "loss": 5.4398,
      "mean_token_accuracy": 0.1856146201491356,
      "num_tokens": 24274981.0,
      "step": 11980
    },
    {
      "entropy": 5.937097835540771,
      "epoch": 1.0473652014331907,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004896666427494012,
      "loss": 5.4705,
      "mean_token_accuracy": 0.18242762982845306,
      "num_tokens": 24285009.0,
      "step": 11985
    },
    {
      "entropy": 5.987487363815307,
      "epoch": 1.0478021497858954,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000489657305725273,
      "loss": 5.5087,
      "mean_token_accuracy": 0.17873066812753677,
      "num_tokens": 24294669.0,
      "step": 11990
    },
    {
      "entropy": 5.947388792037964,
      "epoch": 1.0482390981386,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004896479645839052,
      "loss": 5.5087,
      "mean_token_accuracy": 0.17924651205539704,
      "num_tokens": 24304647.0,
      "step": 11995
    },
    {
      "entropy": 5.8552978992462155,
      "epoch": 1.0486760464913047,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004896386193254771,
      "loss": 5.3764,
      "mean_token_accuracy": 0.18976857811212539,
      "num_tokens": 24314266.0,
      "step": 12000
    },
    {
      "epoch": 1.0486760464913047,
      "eval_entropy": 5.62568598569586,
      "eval_loss": 5.515780448913574,
      "eval_mean_token_accuracy": 0.1908445608982688,
      "eval_num_tokens": 24314266.0,
      "eval_runtime": 21.5991,
      "eval_samples_per_second": 1713.683,
      "eval_steps_per_second": 214.222,
      "step": 12000
    },
    {
      "entropy": 5.842466640472412,
      "epoch": 1.0491129948440094,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004896292699501677,
      "loss": 5.4162,
      "mean_token_accuracy": 0.19099170863628387,
      "num_tokens": 24323770.0,
      "step": 12005
    },
    {
      "entropy": 5.891579341888428,
      "epoch": 1.0495499431967141,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004896199164581564,
      "loss": 5.3637,
      "mean_token_accuracy": 0.19347794353961945,
      "num_tokens": 24333438.0,
      "step": 12010
    },
    {
      "entropy": 5.898055362701416,
      "epoch": 1.0499868915494188,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004896105588496227,
      "loss": 5.4582,
      "mean_token_accuracy": 0.18265994638204575,
      "num_tokens": 24342929.0,
      "step": 12015
    },
    {
      "entropy": 5.922069406509399,
      "epoch": 1.0504238399021235,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004896011971247461,
      "loss": 5.4731,
      "mean_token_accuracy": 0.18305714726448058,
      "num_tokens": 24352617.0,
      "step": 12020
    },
    {
      "entropy": 5.920299816131592,
      "epoch": 1.0508607882548282,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004895918312837061,
      "loss": 5.4087,
      "mean_token_accuracy": 0.19081699699163437,
      "num_tokens": 24363940.0,
      "step": 12025
    },
    {
      "entropy": 5.9369792461395265,
      "epoch": 1.051297736607533,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004895824613266822,
      "loss": 5.4901,
      "mean_token_accuracy": 0.1823594555258751,
      "num_tokens": 24373168.0,
      "step": 12030
    },
    {
      "entropy": 5.922732400894165,
      "epoch": 1.0517346849602376,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004895730872538543,
      "loss": 5.5145,
      "mean_token_accuracy": 0.18285975307226182,
      "num_tokens": 24383067.0,
      "step": 12035
    },
    {
      "entropy": 5.9301313877105715,
      "epoch": 1.0521716333129425,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004895637090654021,
      "loss": 5.4022,
      "mean_token_accuracy": 0.19431754052639008,
      "num_tokens": 24392734.0,
      "step": 12040
    },
    {
      "entropy": 5.893765878677368,
      "epoch": 1.0526085816656472,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004895543267615057,
      "loss": 5.4352,
      "mean_token_accuracy": 0.18993961811065674,
      "num_tokens": 24403970.0,
      "step": 12045
    },
    {
      "entropy": 5.886629962921143,
      "epoch": 1.0530455300183519,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004895449403423447,
      "loss": 5.443,
      "mean_token_accuracy": 0.1857756108045578,
      "num_tokens": 24413614.0,
      "step": 12050
    },
    {
      "entropy": 5.855058479309082,
      "epoch": 1.0534824783710566,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004895355498080994,
      "loss": 5.4091,
      "mean_token_accuracy": 0.19264777153730392,
      "num_tokens": 24424237.0,
      "step": 12055
    },
    {
      "entropy": 5.895183658599853,
      "epoch": 1.0539194267237613,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004895261551589498,
      "loss": 5.4203,
      "mean_token_accuracy": 0.19014182835817336,
      "num_tokens": 24433220.0,
      "step": 12060
    },
    {
      "entropy": 5.995061540603638,
      "epoch": 1.054356375076466,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004895167563950763,
      "loss": 5.5544,
      "mean_token_accuracy": 0.17635063529014589,
      "num_tokens": 24444224.0,
      "step": 12065
    },
    {
      "entropy": 5.906583452224732,
      "epoch": 1.0547933234291706,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004895073535166588,
      "loss": 5.5489,
      "mean_token_accuracy": 0.17916615605354308,
      "num_tokens": 24454431.0,
      "step": 12070
    },
    {
      "entropy": 5.954584264755249,
      "epoch": 1.0552302717818753,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004894979465238779,
      "loss": 5.4365,
      "mean_token_accuracy": 0.1816106230020523,
      "num_tokens": 24464006.0,
      "step": 12075
    },
    {
      "entropy": 5.911929988861084,
      "epoch": 1.05566722013458,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004894885354169141,
      "loss": 5.3742,
      "mean_token_accuracy": 0.19352557957172395,
      "num_tokens": 24472706.0,
      "step": 12080
    },
    {
      "entropy": 5.879449510574341,
      "epoch": 1.0561041684872847,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004894791201959477,
      "loss": 5.3969,
      "mean_token_accuracy": 0.1941150978207588,
      "num_tokens": 24482785.0,
      "step": 12085
    },
    {
      "entropy": 5.921630334854126,
      "epoch": 1.0565411168399894,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004894697008611594,
      "loss": 5.4955,
      "mean_token_accuracy": 0.18199741393327712,
      "num_tokens": 24493318.0,
      "step": 12090
    },
    {
      "entropy": 5.867503499984741,
      "epoch": 1.0569780651926943,
      "grad_norm": 1.0,
      "learning_rate": 0.0004894602774127299,
      "loss": 5.4818,
      "mean_token_accuracy": 0.18077620714902878,
      "num_tokens": 24503780.0,
      "step": 12095
    },
    {
      "entropy": 5.982315158843994,
      "epoch": 1.057415013545399,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004894508498508399,
      "loss": 5.4947,
      "mean_token_accuracy": 0.17909740060567855,
      "num_tokens": 24513905.0,
      "step": 12100
    },
    {
      "entropy": 5.928594160079956,
      "epoch": 1.0578519618981037,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004894414181756702,
      "loss": 5.4456,
      "mean_token_accuracy": 0.18665642887353898,
      "num_tokens": 24524194.0,
      "step": 12105
    },
    {
      "entropy": 5.933513402938843,
      "epoch": 1.0582889102508084,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004894319823874017,
      "loss": 5.4893,
      "mean_token_accuracy": 0.18509892523288726,
      "num_tokens": 24533456.0,
      "step": 12110
    },
    {
      "entropy": 5.929219722747803,
      "epoch": 1.058725858603513,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004894225424862154,
      "loss": 5.4774,
      "mean_token_accuracy": 0.18546983748674392,
      "num_tokens": 24544191.0,
      "step": 12115
    },
    {
      "entropy": 5.842947578430175,
      "epoch": 1.0591628069562178,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004894130984722926,
      "loss": 5.4009,
      "mean_token_accuracy": 0.1968051701784134,
      "num_tokens": 24554353.0,
      "step": 12120
    },
    {
      "entropy": 5.939659976959229,
      "epoch": 1.0595997553089225,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000489403650345814,
      "loss": 5.5051,
      "mean_token_accuracy": 0.1890042394399643,
      "num_tokens": 24564313.0,
      "step": 12125
    },
    {
      "entropy": 5.933866786956787,
      "epoch": 1.0600367036616272,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004893941981069612,
      "loss": 5.5099,
      "mean_token_accuracy": 0.1837530553340912,
      "num_tokens": 24573129.0,
      "step": 12130
    },
    {
      "entropy": 5.844171476364136,
      "epoch": 1.0604736520143319,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004893847417559152,
      "loss": 5.3701,
      "mean_token_accuracy": 0.18995449990034102,
      "num_tokens": 24582324.0,
      "step": 12135
    },
    {
      "entropy": 5.942809772491455,
      "epoch": 1.0609106003670365,
      "grad_norm": 1.125,
      "learning_rate": 0.0004893752812928576,
      "loss": 5.5128,
      "mean_token_accuracy": 0.1797366052865982,
      "num_tokens": 24593360.0,
      "step": 12140
    },
    {
      "entropy": 5.917215061187744,
      "epoch": 1.0613475487197412,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004893658167179697,
      "loss": 5.4353,
      "mean_token_accuracy": 0.18614036440849305,
      "num_tokens": 24603597.0,
      "step": 12145
    },
    {
      "entropy": 5.928200387954712,
      "epoch": 1.061784497072446,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004893563480314331,
      "loss": 5.3732,
      "mean_token_accuracy": 0.18634904623031617,
      "num_tokens": 24612426.0,
      "step": 12150
    },
    {
      "entropy": 5.899135589599609,
      "epoch": 1.0622214454251508,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004893468752334294,
      "loss": 5.3996,
      "mean_token_accuracy": 0.18302166610956191,
      "num_tokens": 24623031.0,
      "step": 12155
    },
    {
      "entropy": 5.853148698806763,
      "epoch": 1.0626583937778555,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004893373983241404,
      "loss": 5.4431,
      "mean_token_accuracy": 0.18822433203458785,
      "num_tokens": 24633763.0,
      "step": 12160
    },
    {
      "entropy": 5.8758628368377686,
      "epoch": 1.0630953421305602,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004893279173037478,
      "loss": 5.4743,
      "mean_token_accuracy": 0.18334336280822755,
      "num_tokens": 24644340.0,
      "step": 12165
    },
    {
      "entropy": 5.981291103363037,
      "epoch": 1.063532290483265,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004893184321724335,
      "loss": 5.4418,
      "mean_token_accuracy": 0.1932005688548088,
      "num_tokens": 24654751.0,
      "step": 12170
    },
    {
      "entropy": 5.8671417236328125,
      "epoch": 1.0639692388359696,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004893089429303792,
      "loss": 5.4403,
      "mean_token_accuracy": 0.18770609200000762,
      "num_tokens": 24665134.0,
      "step": 12175
    },
    {
      "entropy": 5.88116307258606,
      "epoch": 1.0644061871886743,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004892994495777672,
      "loss": 5.4182,
      "mean_token_accuracy": 0.182558137178421,
      "num_tokens": 24673796.0,
      "step": 12180
    },
    {
      "entropy": 5.8726969242095945,
      "epoch": 1.064843135541379,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004892899521147794,
      "loss": 5.3832,
      "mean_token_accuracy": 0.19351694434881211,
      "num_tokens": 24684104.0,
      "step": 12185
    },
    {
      "entropy": 5.838142585754395,
      "epoch": 1.0652800838940837,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000489280450541598,
      "loss": 5.4012,
      "mean_token_accuracy": 0.18845458179712296,
      "num_tokens": 24693994.0,
      "step": 12190
    },
    {
      "entropy": 5.877148103713989,
      "epoch": 1.0657170322467884,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004892709448584054,
      "loss": 5.454,
      "mean_token_accuracy": 0.1869272008538246,
      "num_tokens": 24703815.0,
      "step": 12195
    },
    {
      "entropy": 5.927348613739014,
      "epoch": 1.066153980599493,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004892614350653837,
      "loss": 5.4067,
      "mean_token_accuracy": 0.18203674256801605,
      "num_tokens": 24713631.0,
      "step": 12200
    },
    {
      "entropy": 5.942470693588257,
      "epoch": 1.0665909289521978,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004892519211627156,
      "loss": 5.4174,
      "mean_token_accuracy": 0.19234438091516495,
      "num_tokens": 24723353.0,
      "step": 12205
    },
    {
      "entropy": 5.858028745651245,
      "epoch": 1.0670278773049025,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004892424031505832,
      "loss": 5.4158,
      "mean_token_accuracy": 0.1820172995328903,
      "num_tokens": 24733246.0,
      "step": 12210
    },
    {
      "entropy": 5.8454570293426515,
      "epoch": 1.0674648256576074,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004892328810291693,
      "loss": 5.4182,
      "mean_token_accuracy": 0.1866133615374565,
      "num_tokens": 24743097.0,
      "step": 12215
    },
    {
      "entropy": 5.904064607620239,
      "epoch": 1.067901774010312,
      "grad_norm": 1.0,
      "learning_rate": 0.0004892233547986564,
      "loss": 5.3819,
      "mean_token_accuracy": 0.18510522246360778,
      "num_tokens": 24752618.0,
      "step": 12220
    },
    {
      "entropy": 5.934146738052368,
      "epoch": 1.0683387223630167,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004892138244592274,
      "loss": 5.4555,
      "mean_token_accuracy": 0.1908336713910103,
      "num_tokens": 24763960.0,
      "step": 12225
    },
    {
      "entropy": 5.864588642120362,
      "epoch": 1.0687756707157214,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004892042900110649,
      "loss": 5.3462,
      "mean_token_accuracy": 0.18917935937643052,
      "num_tokens": 24773040.0,
      "step": 12230
    },
    {
      "entropy": 5.850897884368896,
      "epoch": 1.0692126190684261,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000489194751454352,
      "loss": 5.4468,
      "mean_token_accuracy": 0.18614864498376846,
      "num_tokens": 24783049.0,
      "step": 12235
    },
    {
      "entropy": 5.926939630508423,
      "epoch": 1.0696495674211308,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004891852087892714,
      "loss": 5.4591,
      "mean_token_accuracy": 0.18244907557964324,
      "num_tokens": 24794590.0,
      "step": 12240
    },
    {
      "entropy": 5.8359479904174805,
      "epoch": 1.0700865157738355,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004891756620160064,
      "loss": 5.3444,
      "mean_token_accuracy": 0.19076519161462785,
      "num_tokens": 24805146.0,
      "step": 12245
    },
    {
      "entropy": 5.910455131530762,
      "epoch": 1.0705234641265402,
      "grad_norm": 1.0,
      "learning_rate": 0.0004891661111347399,
      "loss": 5.4732,
      "mean_token_accuracy": 0.18025609850883484,
      "num_tokens": 24815573.0,
      "step": 12250
    },
    {
      "entropy": 5.9008032321929935,
      "epoch": 1.070960412479245,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004891565561456552,
      "loss": 5.3812,
      "mean_token_accuracy": 0.19707030802965164,
      "num_tokens": 24825417.0,
      "step": 12255
    },
    {
      "entropy": 5.816665887832642,
      "epoch": 1.0713973608319496,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004891469970489355,
      "loss": 5.3928,
      "mean_token_accuracy": 0.19169097691774367,
      "num_tokens": 24835589.0,
      "step": 12260
    },
    {
      "entropy": 5.876756381988526,
      "epoch": 1.0718343091846543,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004891374338447642,
      "loss": 5.5086,
      "mean_token_accuracy": 0.1805966541171074,
      "num_tokens": 24846603.0,
      "step": 12265
    },
    {
      "entropy": 5.987513685226441,
      "epoch": 1.0722712575373592,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004891278665333247,
      "loss": 5.4351,
      "mean_token_accuracy": 0.1874460607767105,
      "num_tokens": 24856368.0,
      "step": 12270
    },
    {
      "entropy": 5.825826120376587,
      "epoch": 1.0727082058900639,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004891182951148005,
      "loss": 5.347,
      "mean_token_accuracy": 0.19072122871875763,
      "num_tokens": 24866322.0,
      "step": 12275
    },
    {
      "entropy": 5.86176495552063,
      "epoch": 1.0731451542427686,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004891087195893753,
      "loss": 5.4369,
      "mean_token_accuracy": 0.1818415954709053,
      "num_tokens": 24876686.0,
      "step": 12280
    },
    {
      "entropy": 5.911796998977661,
      "epoch": 1.0735821025954733,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004890991399572325,
      "loss": 5.393,
      "mean_token_accuracy": 0.18578656911849975,
      "num_tokens": 24886561.0,
      "step": 12285
    },
    {
      "entropy": 5.879510545730591,
      "epoch": 1.074019050948178,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004890895562185562,
      "loss": 5.3987,
      "mean_token_accuracy": 0.18967887610197068,
      "num_tokens": 24896010.0,
      "step": 12290
    },
    {
      "entropy": 5.9468694686889645,
      "epoch": 1.0744559993008826,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004890799683735299,
      "loss": 5.4873,
      "mean_token_accuracy": 0.17694503664970399,
      "num_tokens": 24904852.0,
      "step": 12295
    },
    {
      "entropy": 5.965913581848144,
      "epoch": 1.0748929476535873,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004890703764223377,
      "loss": 5.4606,
      "mean_token_accuracy": 0.18719341456890107,
      "num_tokens": 24914828.0,
      "step": 12300
    },
    {
      "entropy": 5.958166790008545,
      "epoch": 1.075329896006292,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004890607803651635,
      "loss": 5.5031,
      "mean_token_accuracy": 0.172243432700634,
      "num_tokens": 24925602.0,
      "step": 12305
    },
    {
      "entropy": 5.898205137252807,
      "epoch": 1.0757668443589967,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004890511802021914,
      "loss": 5.468,
      "mean_token_accuracy": 0.190065136551857,
      "num_tokens": 24935175.0,
      "step": 12310
    },
    {
      "entropy": 5.936156272888184,
      "epoch": 1.0762037927117014,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004890415759336053,
      "loss": 5.4712,
      "mean_token_accuracy": 0.179387691617012,
      "num_tokens": 24945828.0,
      "step": 12315
    },
    {
      "entropy": 5.909340763092041,
      "epoch": 1.076640741064406,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004890319675595899,
      "loss": 5.4111,
      "mean_token_accuracy": 0.189553964138031,
      "num_tokens": 24955400.0,
      "step": 12320
    },
    {
      "entropy": 5.851690149307251,
      "epoch": 1.0770776894171108,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000489022355080329,
      "loss": 5.4866,
      "mean_token_accuracy": 0.18456988781690598,
      "num_tokens": 24966079.0,
      "step": 12325
    },
    {
      "entropy": 5.918713903427124,
      "epoch": 1.0775146377698157,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004890127384960073,
      "loss": 5.3735,
      "mean_token_accuracy": 0.1938445970416069,
      "num_tokens": 24975090.0,
      "step": 12330
    },
    {
      "entropy": 5.99315710067749,
      "epoch": 1.0779515861225204,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004890031178068091,
      "loss": 5.4934,
      "mean_token_accuracy": 0.17711793333292009,
      "num_tokens": 24986068.0,
      "step": 12335
    },
    {
      "entropy": 5.912244367599487,
      "epoch": 1.078388534475225,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000488993493012919,
      "loss": 5.5086,
      "mean_token_accuracy": 0.18439267873764037,
      "num_tokens": 24995697.0,
      "step": 12340
    },
    {
      "entropy": 5.919261789321899,
      "epoch": 1.0788254828279298,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004889838641145215,
      "loss": 5.4118,
      "mean_token_accuracy": 0.18033651113510132,
      "num_tokens": 25005602.0,
      "step": 12345
    },
    {
      "entropy": 5.930631971359253,
      "epoch": 1.0792624311806345,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004889742311118014,
      "loss": 5.4927,
      "mean_token_accuracy": 0.18898247927427292,
      "num_tokens": 25016407.0,
      "step": 12350
    },
    {
      "entropy": 5.981080770492554,
      "epoch": 1.0796993795333392,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004889645940049434,
      "loss": 5.4981,
      "mean_token_accuracy": 0.18142589628696443,
      "num_tokens": 25025889.0,
      "step": 12355
    },
    {
      "entropy": 5.877120780944824,
      "epoch": 1.0801363278860439,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004889549527941324,
      "loss": 5.4076,
      "mean_token_accuracy": 0.19223953038454056,
      "num_tokens": 25036383.0,
      "step": 12360
    },
    {
      "entropy": 5.944394922256469,
      "epoch": 1.0805732762387485,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004889453074795534,
      "loss": 5.5567,
      "mean_token_accuracy": 0.1752297192811966,
      "num_tokens": 25045718.0,
      "step": 12365
    },
    {
      "entropy": 5.918315362930298,
      "epoch": 1.0810102245914532,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004889356580613911,
      "loss": 5.5183,
      "mean_token_accuracy": 0.18233492970466614,
      "num_tokens": 25056244.0,
      "step": 12370
    },
    {
      "entropy": 5.920806217193603,
      "epoch": 1.081447172944158,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000488926004539831,
      "loss": 5.4571,
      "mean_token_accuracy": 0.18513164222240447,
      "num_tokens": 25065954.0,
      "step": 12375
    },
    {
      "entropy": 5.886236476898193,
      "epoch": 1.0818841212968626,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004889163469150578,
      "loss": 5.4935,
      "mean_token_accuracy": 0.18605174869298935,
      "num_tokens": 25075851.0,
      "step": 12380
    },
    {
      "entropy": 5.935222625732422,
      "epoch": 1.0823210696495673,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004889066851872571,
      "loss": 5.4628,
      "mean_token_accuracy": 0.18554012924432756,
      "num_tokens": 25085324.0,
      "step": 12385
    },
    {
      "entropy": 5.926063966751099,
      "epoch": 1.0827580180022722,
      "grad_norm": 0.9375,
      "learning_rate": 0.000488897019356614,
      "loss": 5.4875,
      "mean_token_accuracy": 0.18189099729061126,
      "num_tokens": 25096566.0,
      "step": 12390
    },
    {
      "entropy": 6.007735872268677,
      "epoch": 1.083194966354977,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004888873494233141,
      "loss": 5.5506,
      "mean_token_accuracy": 0.1758687525987625,
      "num_tokens": 25107642.0,
      "step": 12395
    },
    {
      "entropy": 5.920183706283569,
      "epoch": 1.0836319147076816,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004888776753875427,
      "loss": 5.4385,
      "mean_token_accuracy": 0.18584631085395814,
      "num_tokens": 25117488.0,
      "step": 12400
    },
    {
      "entropy": 5.861323976516724,
      "epoch": 1.0840688630603863,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004888679972494854,
      "loss": 5.4137,
      "mean_token_accuracy": 0.1879253640770912,
      "num_tokens": 25127107.0,
      "step": 12405
    },
    {
      "entropy": 5.849725341796875,
      "epoch": 1.084505811413091,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000488858315009328,
      "loss": 5.4267,
      "mean_token_accuracy": 0.18657674342393876,
      "num_tokens": 25138272.0,
      "step": 12410
    },
    {
      "entropy": 5.936537599563598,
      "epoch": 1.0849427597657957,
      "grad_norm": 1.0,
      "learning_rate": 0.0004888486286672559,
      "loss": 5.4608,
      "mean_token_accuracy": 0.18083604723215102,
      "num_tokens": 25147128.0,
      "step": 12415
    },
    {
      "entropy": 5.943241310119629,
      "epoch": 1.0853797081185004,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004888389382234551,
      "loss": 5.3894,
      "mean_token_accuracy": 0.19808750748634338,
      "num_tokens": 25157418.0,
      "step": 12420
    },
    {
      "entropy": 5.884813833236694,
      "epoch": 1.085816656471205,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004888292436781114,
      "loss": 5.4406,
      "mean_token_accuracy": 0.18642957657575607,
      "num_tokens": 25168123.0,
      "step": 12425
    },
    {
      "entropy": 5.840692663192749,
      "epoch": 1.0862536048239098,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004888195450314108,
      "loss": 5.4558,
      "mean_token_accuracy": 0.18749166429042816,
      "num_tokens": 25178445.0,
      "step": 12430
    },
    {
      "entropy": 5.909745502471924,
      "epoch": 1.0866905531766144,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004888098422835392,
      "loss": 5.4836,
      "mean_token_accuracy": 0.17849712669849396,
      "num_tokens": 25189246.0,
      "step": 12435
    },
    {
      "entropy": 5.886618518829346,
      "epoch": 1.0871275015293191,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004888001354346828,
      "loss": 5.4717,
      "mean_token_accuracy": 0.1833967611193657,
      "num_tokens": 25201277.0,
      "step": 12440
    },
    {
      "entropy": 5.949337816238403,
      "epoch": 1.087564449882024,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004887904244850279,
      "loss": 5.4519,
      "mean_token_accuracy": 0.18989188969135284,
      "num_tokens": 25211353.0,
      "step": 12445
    },
    {
      "entropy": 5.886809635162353,
      "epoch": 1.0880013982347287,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004887807094347606,
      "loss": 5.4327,
      "mean_token_accuracy": 0.18662758618593217,
      "num_tokens": 25222236.0,
      "step": 12450
    },
    {
      "entropy": 5.858541202545166,
      "epoch": 1.0884383465874334,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004887709902840672,
      "loss": 5.3686,
      "mean_token_accuracy": 0.18889901041984558,
      "num_tokens": 25232464.0,
      "step": 12455
    },
    {
      "entropy": 5.920749759674072,
      "epoch": 1.0888752949401381,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004887612670331342,
      "loss": 5.4444,
      "mean_token_accuracy": 0.18218039572238923,
      "num_tokens": 25243518.0,
      "step": 12460
    },
    {
      "entropy": 5.921897125244141,
      "epoch": 1.0893122432928428,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004887515396821481,
      "loss": 5.4516,
      "mean_token_accuracy": 0.18312579095363618,
      "num_tokens": 25253168.0,
      "step": 12465
    },
    {
      "entropy": 5.881242990493774,
      "epoch": 1.0897491916455475,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004887418082312955,
      "loss": 5.4152,
      "mean_token_accuracy": 0.18842000216245652,
      "num_tokens": 25262750.0,
      "step": 12470
    },
    {
      "entropy": 5.915745210647583,
      "epoch": 1.0901861399982522,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000488732072680763,
      "loss": 5.4328,
      "mean_token_accuracy": 0.1879446402192116,
      "num_tokens": 25273520.0,
      "step": 12475
    },
    {
      "entropy": 5.902719974517822,
      "epoch": 1.090623088350957,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004887223330307373,
      "loss": 5.4595,
      "mean_token_accuracy": 0.1875239372253418,
      "num_tokens": 25284136.0,
      "step": 12480
    },
    {
      "entropy": 5.935053586959839,
      "epoch": 1.0910600367036616,
      "grad_norm": 1.0,
      "learning_rate": 0.0004887125892814053,
      "loss": 5.3988,
      "mean_token_accuracy": 0.1833614870905876,
      "num_tokens": 25294523.0,
      "step": 12485
    },
    {
      "entropy": 5.87125334739685,
      "epoch": 1.0914969850563663,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004887028414329538,
      "loss": 5.4256,
      "mean_token_accuracy": 0.18472955077886583,
      "num_tokens": 25305050.0,
      "step": 12490
    },
    {
      "entropy": 5.8991968631744385,
      "epoch": 1.091933933409071,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048869308948557,
      "loss": 5.4407,
      "mean_token_accuracy": 0.18687492311000825,
      "num_tokens": 25314445.0,
      "step": 12495
    },
    {
      "entropy": 5.909014320373535,
      "epoch": 1.0923708817617757,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004886833334394406,
      "loss": 5.445,
      "mean_token_accuracy": 0.1861948013305664,
      "num_tokens": 25324426.0,
      "step": 12500
    },
    {
      "entropy": 5.874684572219849,
      "epoch": 1.0928078301144806,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004886735732947529,
      "loss": 5.4316,
      "mean_token_accuracy": 0.18774652928113938,
      "num_tokens": 25335051.0,
      "step": 12505
    },
    {
      "entropy": 5.862842845916748,
      "epoch": 1.0932447784671853,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004886638090516942,
      "loss": 5.3529,
      "mean_token_accuracy": 0.19270295798778533,
      "num_tokens": 25343955.0,
      "step": 12510
    },
    {
      "entropy": 5.841263008117676,
      "epoch": 1.09368172681989,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004886540407104517,
      "loss": 5.4474,
      "mean_token_accuracy": 0.18338657170534134,
      "num_tokens": 25353817.0,
      "step": 12515
    },
    {
      "entropy": 5.985517978668213,
      "epoch": 1.0941186751725946,
      "grad_norm": 1.0,
      "learning_rate": 0.0004886442682712126,
      "loss": 5.3717,
      "mean_token_accuracy": 0.18822979927062988,
      "num_tokens": 25363882.0,
      "step": 12520
    },
    {
      "entropy": 5.907884168624878,
      "epoch": 1.0945556235252993,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004886344917341646,
      "loss": 5.4063,
      "mean_token_accuracy": 0.18822278827428818,
      "num_tokens": 25373721.0,
      "step": 12525
    },
    {
      "entropy": 5.900276184082031,
      "epoch": 1.094992571878004,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004886247110994952,
      "loss": 5.4203,
      "mean_token_accuracy": 0.18445351868867874,
      "num_tokens": 25383520.0,
      "step": 12530
    },
    {
      "entropy": 5.9098286628723145,
      "epoch": 1.0954295202307087,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004886149263673916,
      "loss": 5.4129,
      "mean_token_accuracy": 0.1888999730348587,
      "num_tokens": 25394039.0,
      "step": 12535
    },
    {
      "entropy": 5.88784704208374,
      "epoch": 1.0958664685834134,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000488605137538042,
      "loss": 5.4312,
      "mean_token_accuracy": 0.18588075190782546,
      "num_tokens": 25403404.0,
      "step": 12540
    },
    {
      "entropy": 5.851065301895142,
      "epoch": 1.096303416936118,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004885953446116339,
      "loss": 5.4288,
      "mean_token_accuracy": 0.18879041969776153,
      "num_tokens": 25413800.0,
      "step": 12545
    },
    {
      "entropy": 5.868917751312256,
      "epoch": 1.0967403652888228,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004885855475883551,
      "loss": 5.4036,
      "mean_token_accuracy": 0.19364429712295533,
      "num_tokens": 25424252.0,
      "step": 12550
    },
    {
      "entropy": 5.957980489730835,
      "epoch": 1.0971773136415275,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004885757464683937,
      "loss": 5.4109,
      "mean_token_accuracy": 0.18612736612558364,
      "num_tokens": 25433381.0,
      "step": 12555
    },
    {
      "entropy": 5.880148887634277,
      "epoch": 1.0976142619942322,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004885659412519376,
      "loss": 5.3706,
      "mean_token_accuracy": 0.194099260866642,
      "num_tokens": 25443253.0,
      "step": 12560
    },
    {
      "entropy": 5.867990350723266,
      "epoch": 1.098051210346937,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004885561319391747,
      "loss": 5.4706,
      "mean_token_accuracy": 0.18637143522500993,
      "num_tokens": 25453590.0,
      "step": 12565
    },
    {
      "entropy": 5.8957620620727536,
      "epoch": 1.0984881586996418,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004885463185302934,
      "loss": 5.383,
      "mean_token_accuracy": 0.19600391238927842,
      "num_tokens": 25464065.0,
      "step": 12570
    },
    {
      "entropy": 5.886085176467896,
      "epoch": 1.0989251070523465,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004885365010254817,
      "loss": 5.4905,
      "mean_token_accuracy": 0.1792607992887497,
      "num_tokens": 25475236.0,
      "step": 12575
    },
    {
      "entropy": 5.92491135597229,
      "epoch": 1.0993620554050512,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004885266794249281,
      "loss": 5.5086,
      "mean_token_accuracy": 0.18409744799137115,
      "num_tokens": 25485200.0,
      "step": 12580
    },
    {
      "entropy": 5.962625360488891,
      "epoch": 1.0997990037577559,
      "grad_norm": 1.0,
      "learning_rate": 0.0004885168537288209,
      "loss": 5.4287,
      "mean_token_accuracy": 0.19065171331167222,
      "num_tokens": 25494775.0,
      "step": 12585
    },
    {
      "entropy": 5.964064741134644,
      "epoch": 1.1002359521104605,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004885070239373484,
      "loss": 5.4637,
      "mean_token_accuracy": 0.18463966995477676,
      "num_tokens": 25504435.0,
      "step": 12590
    },
    {
      "entropy": 5.862992858886718,
      "epoch": 1.1006729004631652,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004884971900506995,
      "loss": 5.4528,
      "mean_token_accuracy": 0.18344820737838746,
      "num_tokens": 25514085.0,
      "step": 12595
    },
    {
      "entropy": 5.950091552734375,
      "epoch": 1.10110984881587,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004884873520690626,
      "loss": 5.5052,
      "mean_token_accuracy": 0.18050521165132521,
      "num_tokens": 25523988.0,
      "step": 12600
    },
    {
      "entropy": 5.913284635543823,
      "epoch": 1.1015467971685746,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004884775099926264,
      "loss": 5.4765,
      "mean_token_accuracy": 0.18395519256591797,
      "num_tokens": 25534469.0,
      "step": 12605
    },
    {
      "entropy": 5.929515266418457,
      "epoch": 1.1019837455212793,
      "grad_norm": 1.0,
      "learning_rate": 0.0004884676638215797,
      "loss": 5.5143,
      "mean_token_accuracy": 0.17605181485414506,
      "num_tokens": 25544854.0,
      "step": 12610
    },
    {
      "entropy": 5.957728052139283,
      "epoch": 1.102420693873984,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004884578135561113,
      "loss": 5.4125,
      "mean_token_accuracy": 0.18701433390378952,
      "num_tokens": 25555465.0,
      "step": 12615
    },
    {
      "entropy": 5.858280849456787,
      "epoch": 1.102857642226689,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004884479591964103,
      "loss": 5.4742,
      "mean_token_accuracy": 0.18753819018602372,
      "num_tokens": 25565949.0,
      "step": 12620
    },
    {
      "entropy": 5.923497009277344,
      "epoch": 1.1032945905793936,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004884381007426654,
      "loss": 5.4521,
      "mean_token_accuracy": 0.18495838344097137,
      "num_tokens": 25576511.0,
      "step": 12625
    },
    {
      "entropy": 5.8607102394104,
      "epoch": 1.1037315389320983,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004884282381950661,
      "loss": 5.3554,
      "mean_token_accuracy": 0.1916596382856369,
      "num_tokens": 25587224.0,
      "step": 12630
    },
    {
      "entropy": 5.915241336822509,
      "epoch": 1.104168487284803,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004884183715538012,
      "loss": 5.4358,
      "mean_token_accuracy": 0.1798640102148056,
      "num_tokens": 25596696.0,
      "step": 12635
    },
    {
      "entropy": 5.8584977149963375,
      "epoch": 1.1046054356375077,
      "grad_norm": 1.0,
      "learning_rate": 0.0004884085008190602,
      "loss": 5.393,
      "mean_token_accuracy": 0.19510244876146315,
      "num_tokens": 25606829.0,
      "step": 12640
    },
    {
      "entropy": 5.907916736602783,
      "epoch": 1.1050423839902124,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004883986259910323,
      "loss": 5.4559,
      "mean_token_accuracy": 0.1910765215754509,
      "num_tokens": 25617122.0,
      "step": 12645
    },
    {
      "entropy": 5.902012157440185,
      "epoch": 1.105479332342917,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004883887470699069,
      "loss": 5.4045,
      "mean_token_accuracy": 0.1855984792113304,
      "num_tokens": 25626485.0,
      "step": 12650
    },
    {
      "entropy": 5.920029926300049,
      "epoch": 1.1059162806956218,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004883788640558735,
      "loss": 5.3921,
      "mean_token_accuracy": 0.19085173755884172,
      "num_tokens": 25636695.0,
      "step": 12655
    },
    {
      "entropy": 5.770718002319336,
      "epoch": 1.1063532290483264,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004883689769491218,
      "loss": 5.2663,
      "mean_token_accuracy": 0.199479903280735,
      "num_tokens": 25646169.0,
      "step": 12660
    },
    {
      "entropy": 5.892646408081054,
      "epoch": 1.1067901774010311,
      "grad_norm": 1.125,
      "learning_rate": 0.0004883590857498413,
      "loss": 5.4718,
      "mean_token_accuracy": 0.18242418318986892,
      "num_tokens": 25656728.0,
      "step": 12665
    },
    {
      "entropy": 5.93475456237793,
      "epoch": 1.1072271257537358,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004883491904582217,
      "loss": 5.4155,
      "mean_token_accuracy": 0.19014625251293182,
      "num_tokens": 25666184.0,
      "step": 12670
    },
    {
      "entropy": 5.920754146575928,
      "epoch": 1.1076640741064405,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004883392910744528,
      "loss": 5.4424,
      "mean_token_accuracy": 0.18661602586507797,
      "num_tokens": 25675571.0,
      "step": 12675
    },
    {
      "entropy": 5.942311525344849,
      "epoch": 1.1081010224591452,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004883293875987245,
      "loss": 5.4555,
      "mean_token_accuracy": 0.18740228861570357,
      "num_tokens": 25685485.0,
      "step": 12680
    },
    {
      "entropy": 5.88535099029541,
      "epoch": 1.1085379708118501,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004883194800312268,
      "loss": 5.3516,
      "mean_token_accuracy": 0.18854387253522872,
      "num_tokens": 25695278.0,
      "step": 12685
    },
    {
      "entropy": 5.880913257598877,
      "epoch": 1.1089749191645548,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004883095683721497,
      "loss": 5.4365,
      "mean_token_accuracy": 0.19489127397537231,
      "num_tokens": 25704410.0,
      "step": 12690
    },
    {
      "entropy": 5.899741888046265,
      "epoch": 1.1094118675172595,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004882996526216833,
      "loss": 5.43,
      "mean_token_accuracy": 0.18561683893203734,
      "num_tokens": 25715027.0,
      "step": 12695
    },
    {
      "entropy": 5.950862741470337,
      "epoch": 1.1098488158699642,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004882897327800178,
      "loss": 5.4519,
      "mean_token_accuracy": 0.18554727733135223,
      "num_tokens": 25723992.0,
      "step": 12700
    },
    {
      "entropy": 5.904939603805542,
      "epoch": 1.110285764222669,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004882798088473435,
      "loss": 5.4074,
      "mean_token_accuracy": 0.18359513878822326,
      "num_tokens": 25733791.0,
      "step": 12705
    },
    {
      "entropy": 5.848233318328857,
      "epoch": 1.1107227125753736,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048826988082385065,
      "loss": 5.4498,
      "mean_token_accuracy": 0.17648694664239883,
      "num_tokens": 25743258.0,
      "step": 12710
    },
    {
      "entropy": 5.929710006713867,
      "epoch": 1.1111596609280783,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048825994870972976,
      "loss": 5.4504,
      "mean_token_accuracy": 0.18817780762910843,
      "num_tokens": 25753183.0,
      "step": 12715
    },
    {
      "entropy": 5.96416482925415,
      "epoch": 1.111596609280783,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048825001250517124,
      "loss": 5.4845,
      "mean_token_accuracy": 0.18400918543338776,
      "num_tokens": 25763308.0,
      "step": 12720
    },
    {
      "entropy": 5.852188491821289,
      "epoch": 1.1120335576334877,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004882400722103658,
      "loss": 5.3834,
      "mean_token_accuracy": 0.19457238763570786,
      "num_tokens": 25774269.0,
      "step": 12725
    },
    {
      "entropy": 5.844601821899414,
      "epoch": 1.1124705059861923,
      "grad_norm": 1.03125,
      "learning_rate": 0.000488230127825504,
      "loss": 5.3562,
      "mean_token_accuracy": 0.18895249664783478,
      "num_tokens": 25785194.0,
      "step": 12730
    },
    {
      "entropy": 5.926044940948486,
      "epoch": 1.112907454338897,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048822017935077673,
      "loss": 5.432,
      "mean_token_accuracy": 0.18998707234859466,
      "num_tokens": 25793927.0,
      "step": 12735
    },
    {
      "entropy": 5.8779724597930905,
      "epoch": 1.113344402691602,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004882102267863746,
      "loss": 5.474,
      "mean_token_accuracy": 0.19029814153909683,
      "num_tokens": 25805395.0,
      "step": 12740
    },
    {
      "entropy": 5.89076452255249,
      "epoch": 1.1137813510443066,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048820027013248864,
      "loss": 5.5064,
      "mean_token_accuracy": 0.1837817072868347,
      "num_tokens": 25816594.0,
      "step": 12745
    },
    {
      "entropy": 5.946514749526978,
      "epoch": 1.1142182993970113,
      "grad_norm": 1.0,
      "learning_rate": 0.00048819030938930974,
      "loss": 5.3641,
      "mean_token_accuracy": 0.18960554748773575,
      "num_tokens": 25825943.0,
      "step": 12750
    },
    {
      "entropy": 5.8403980255126955,
      "epoch": 1.114655247749716,
      "grad_norm": 1.015625,
      "learning_rate": 0.000488180344557029,
      "loss": 5.4238,
      "mean_token_accuracy": 0.1892998546361923,
      "num_tokens": 25835928.0,
      "step": 12755
    },
    {
      "entropy": 5.866338396072388,
      "epoch": 1.1150921961024207,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004881703756358376,
      "loss": 5.4212,
      "mean_token_accuracy": 0.18710991144180297,
      "num_tokens": 25847363.0,
      "step": 12760
    },
    {
      "entropy": 5.938465547561646,
      "epoch": 1.1155291444551254,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004881604026259266,
      "loss": 5.4548,
      "mean_token_accuracy": 0.1873606786131859,
      "num_tokens": 25857246.0,
      "step": 12765
    },
    {
      "entropy": 5.966200828552246,
      "epoch": 1.11596609280783,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048815042552748745,
      "loss": 5.4714,
      "mean_token_accuracy": 0.18118826895952225,
      "num_tokens": 25866615.0,
      "step": 12770
    },
    {
      "entropy": 5.945507955551148,
      "epoch": 1.1164030411605348,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004881404443407115,
      "loss": 5.5241,
      "mean_token_accuracy": 0.18125769793987273,
      "num_tokens": 25876698.0,
      "step": 12775
    },
    {
      "entropy": 5.978520107269287,
      "epoch": 1.1168399895132395,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048813045906579,
      "loss": 5.4827,
      "mean_token_accuracy": 0.18043580353260041,
      "num_tokens": 25888029.0,
      "step": 12780
    },
    {
      "entropy": 5.9132448673248295,
      "epoch": 1.1172769378659442,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048812046970291473,
      "loss": 5.5019,
      "mean_token_accuracy": 0.18208254128694534,
      "num_tokens": 25897472.0,
      "step": 12785
    },
    {
      "entropy": 5.969565677642822,
      "epoch": 1.1177138862186489,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004881104762522772,
      "loss": 5.4921,
      "mean_token_accuracy": 0.18030413687229158,
      "num_tokens": 25908025.0,
      "step": 12790
    },
    {
      "entropy": 5.967176294326782,
      "epoch": 1.1181508345713538,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000488100478714069,
      "loss": 5.4351,
      "mean_token_accuracy": 0.1885099232196808,
      "num_tokens": 25919110.0,
      "step": 12795
    },
    {
      "entropy": 5.908502864837646,
      "epoch": 1.1185877829240585,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048809047708848194,
      "loss": 5.3732,
      "mean_token_accuracy": 0.19730443209409715,
      "num_tokens": 25928322.0,
      "step": 12800
    },
    {
      "entropy": 5.898533344268799,
      "epoch": 1.1190247312767632,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048808047137570795,
      "loss": 5.3721,
      "mean_token_accuracy": 0.18141342103481292,
      "num_tokens": 25937347.0,
      "step": 12805
    },
    {
      "entropy": 5.859088373184204,
      "epoch": 1.1194616796294679,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004880704615759388,
      "loss": 5.358,
      "mean_token_accuracy": 0.18611395806074144,
      "num_tokens": 25948408.0,
      "step": 12810
    },
    {
      "entropy": 5.882910203933716,
      "epoch": 1.1198986279821725,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048806044768936655,
      "loss": 5.4941,
      "mean_token_accuracy": 0.1840204820036888,
      "num_tokens": 25958249.0,
      "step": 12815
    },
    {
      "entropy": 5.965010023117065,
      "epoch": 1.1203355763348772,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004880504297161833,
      "loss": 5.4887,
      "mean_token_accuracy": 0.1805686190724373,
      "num_tokens": 25968849.0,
      "step": 12820
    },
    {
      "entropy": 5.919044828414917,
      "epoch": 1.120772524687582,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048804040765658115,
      "loss": 5.4708,
      "mean_token_accuracy": 0.18948085010051727,
      "num_tokens": 25978787.0,
      "step": 12825
    },
    {
      "entropy": 5.8638067722320555,
      "epoch": 1.1212094730402866,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048803038151075227,
      "loss": 5.4426,
      "mean_token_accuracy": 0.18300366997718812,
      "num_tokens": 25988619.0,
      "step": 12830
    },
    {
      "entropy": 5.904334163665771,
      "epoch": 1.1216464213929913,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004880203512788892,
      "loss": 5.4474,
      "mean_token_accuracy": 0.18635037839412688,
      "num_tokens": 25999080.0,
      "step": 12835
    },
    {
      "entropy": 5.89474720954895,
      "epoch": 1.122083369745696,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048801031696118406,
      "loss": 5.3899,
      "mean_token_accuracy": 0.18636806011199952,
      "num_tokens": 26010329.0,
      "step": 12840
    },
    {
      "entropy": 5.860585498809814,
      "epoch": 1.1225203180984007,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004880002785578295,
      "loss": 5.3562,
      "mean_token_accuracy": 0.19064641296863555,
      "num_tokens": 26019147.0,
      "step": 12845
    },
    {
      "entropy": 5.905275821685791,
      "epoch": 1.1229572664511054,
      "grad_norm": 1.015625,
      "learning_rate": 0.000487990236069018,
      "loss": 5.443,
      "mean_token_accuracy": 0.18322244435548782,
      "num_tokens": 26029212.0,
      "step": 12850
    },
    {
      "entropy": 5.912679529190063,
      "epoch": 1.12339421480381,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048798018949494214,
      "loss": 5.2924,
      "mean_token_accuracy": 0.19064309298992158,
      "num_tokens": 26038059.0,
      "step": 12855
    },
    {
      "entropy": 5.907798767089844,
      "epoch": 1.123831163156515,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004879701388357946,
      "loss": 5.5263,
      "mean_token_accuracy": 0.1794005587697029,
      "num_tokens": 26048076.0,
      "step": 12860
    },
    {
      "entropy": 5.913774538040161,
      "epoch": 1.1242681115092197,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004879600840917683,
      "loss": 5.4537,
      "mean_token_accuracy": 0.18528557419776917,
      "num_tokens": 26059006.0,
      "step": 12865
    },
    {
      "entropy": 5.9067535400390625,
      "epoch": 1.1247050598619244,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004879500252630559,
      "loss": 5.3743,
      "mean_token_accuracy": 0.19033916890621186,
      "num_tokens": 26069275.0,
      "step": 12870
    },
    {
      "entropy": 5.884539365768433,
      "epoch": 1.125142008214629,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004879399623498504,
      "loss": 5.337,
      "mean_token_accuracy": 0.19360905140638351,
      "num_tokens": 26078869.0,
      "step": 12875
    },
    {
      "entropy": 5.850255680084229,
      "epoch": 1.1255789565673338,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000487929895352345,
      "loss": 5.3772,
      "mean_token_accuracy": 0.18608586490154266,
      "num_tokens": 26088544.0,
      "step": 12880
    },
    {
      "entropy": 5.828926706314087,
      "epoch": 1.1260159049200384,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004879198242707325,
      "loss": 5.3399,
      "mean_token_accuracy": 0.19158085137605668,
      "num_tokens": 26097503.0,
      "step": 12885
    },
    {
      "entropy": 5.972708225250244,
      "epoch": 1.1264528532727431,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048790974910520627,
      "loss": 5.4804,
      "mean_token_accuracy": 0.18535036146640776,
      "num_tokens": 26107228.0,
      "step": 12890
    },
    {
      "entropy": 5.830967903137207,
      "epoch": 1.1268898016254478,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048789966985595946,
      "loss": 5.418,
      "mean_token_accuracy": 0.19176282435655595,
      "num_tokens": 26117633.0,
      "step": 12895
    },
    {
      "entropy": 5.8822988986969,
      "epoch": 1.1273267499781525,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004878895865231854,
      "loss": 5.4496,
      "mean_token_accuracy": 0.18896641582250595,
      "num_tokens": 26127781.0,
      "step": 12900
    },
    {
      "entropy": 5.9732081413269045,
      "epoch": 1.1277636983308572,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004878794991070776,
      "loss": 5.3936,
      "mean_token_accuracy": 0.19299009442329407,
      "num_tokens": 26137675.0,
      "step": 12905
    },
    {
      "entropy": 5.93358154296875,
      "epoch": 1.128200646683562,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048786940760782937,
      "loss": 5.4629,
      "mean_token_accuracy": 0.19115250557661057,
      "num_tokens": 26148572.0,
      "step": 12910
    },
    {
      "entropy": 5.844986391067505,
      "epoch": 1.1286375950362668,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048785931202563443,
      "loss": 5.4746,
      "mean_token_accuracy": 0.19139665514230728,
      "num_tokens": 26159261.0,
      "step": 12915
    },
    {
      "entropy": 5.843157529830933,
      "epoch": 1.1290745433889715,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004878492123606863,
      "loss": 5.4662,
      "mean_token_accuracy": 0.18945335000753402,
      "num_tokens": 26170072.0,
      "step": 12920
    },
    {
      "entropy": 5.897942590713501,
      "epoch": 1.1295114917416762,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048783910861317873,
      "loss": 5.3635,
      "mean_token_accuracy": 0.18605283200740813,
      "num_tokens": 26179947.0,
      "step": 12925
    },
    {
      "entropy": 5.967393159866333,
      "epoch": 1.129948440094381,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048782900078330556,
      "loss": 5.4533,
      "mean_token_accuracy": 0.18430609554052352,
      "num_tokens": 26189982.0,
      "step": 12930
    },
    {
      "entropy": 5.981409788131714,
      "epoch": 1.1303853884470856,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004878188888712606,
      "loss": 5.531,
      "mean_token_accuracy": 0.17256508022546768,
      "num_tokens": 26200832.0,
      "step": 12935
    },
    {
      "entropy": 5.88038215637207,
      "epoch": 1.1308223367997903,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004878087728772379,
      "loss": 5.4808,
      "mean_token_accuracy": 0.18193975538015367,
      "num_tokens": 26211506.0,
      "step": 12940
    },
    {
      "entropy": 5.894055891036987,
      "epoch": 1.131259285152495,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004877986528014314,
      "loss": 5.4175,
      "mean_token_accuracy": 0.18761641830205916,
      "num_tokens": 26222045.0,
      "step": 12945
    },
    {
      "entropy": 5.854353904724121,
      "epoch": 1.1316962335051997,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004877885286440353,
      "loss": 5.3425,
      "mean_token_accuracy": 0.19179090857505798,
      "num_tokens": 26231411.0,
      "step": 12950
    },
    {
      "entropy": 5.888177061080933,
      "epoch": 1.1321331818579043,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048777840040524365,
      "loss": 5.4608,
      "mean_token_accuracy": 0.18329200595617295,
      "num_tokens": 26241491.0,
      "step": 12955
    },
    {
      "entropy": 5.897140884399414,
      "epoch": 1.132570130210609,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048776826808525074,
      "loss": 5.5199,
      "mean_token_accuracy": 0.18284364640712739,
      "num_tokens": 26252037.0,
      "step": 12960
    },
    {
      "entropy": 5.9455889701843265,
      "epoch": 1.1330070785633137,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048775813168425104,
      "loss": 5.4986,
      "mean_token_accuracy": 0.17784121334552766,
      "num_tokens": 26262310.0,
      "step": 12965
    },
    {
      "entropy": 5.916824531555176,
      "epoch": 1.1334440269160186,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048774799120243883,
      "loss": 5.4152,
      "mean_token_accuracy": 0.19100407660007476,
      "num_tokens": 26273001.0,
      "step": 12970
    },
    {
      "entropy": 5.9415491104125975,
      "epoch": 1.1338809752687231,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004877378466400087,
      "loss": 5.4806,
      "mean_token_accuracy": 0.18632115572690963,
      "num_tokens": 26283500.0,
      "step": 12975
    },
    {
      "entropy": 5.876995038986206,
      "epoch": 1.134317923621428,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004877276979971552,
      "loss": 5.3761,
      "mean_token_accuracy": 0.18631767034530639,
      "num_tokens": 26294118.0,
      "step": 12980
    },
    {
      "entropy": 5.919408941268921,
      "epoch": 1.1347548719741327,
      "grad_norm": 1.015625,
      "learning_rate": 0.000487717545274073,
      "loss": 5.5238,
      "mean_token_accuracy": 0.17804068177938462,
      "num_tokens": 26303833.0,
      "step": 12985
    },
    {
      "entropy": 5.956403207778931,
      "epoch": 1.1351918203268374,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048770738847095683,
      "loss": 5.4719,
      "mean_token_accuracy": 0.1818734437227249,
      "num_tokens": 26313542.0,
      "step": 12990
    },
    {
      "entropy": 5.891927194595337,
      "epoch": 1.135628768679542,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048769722758800144,
      "loss": 5.3984,
      "mean_token_accuracy": 0.19360734075307845,
      "num_tokens": 26322783.0,
      "step": 12995
    },
    {
      "entropy": 5.951642751693726,
      "epoch": 1.1360657170322468,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004876870626254018,
      "loss": 5.4909,
      "mean_token_accuracy": 0.1808798611164093,
      "num_tokens": 26333770.0,
      "step": 13000
    },
    {
      "entropy": 5.956759071350097,
      "epoch": 1.1365026653849515,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048767689358335286,
      "loss": 5.43,
      "mean_token_accuracy": 0.19018628001213073,
      "num_tokens": 26344435.0,
      "step": 13005
    },
    {
      "entropy": 5.82670636177063,
      "epoch": 1.1369396137376562,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004876667204620497,
      "loss": 5.396,
      "mean_token_accuracy": 0.19598002582788468,
      "num_tokens": 26355410.0,
      "step": 13010
    },
    {
      "entropy": 5.895266342163086,
      "epoch": 1.1373765620903609,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004876565432616873,
      "loss": 5.464,
      "mean_token_accuracy": 0.19123718440532683,
      "num_tokens": 26365785.0,
      "step": 13015
    },
    {
      "entropy": 5.91470890045166,
      "epoch": 1.1378135104430656,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048764636198246104,
      "loss": 5.4183,
      "mean_token_accuracy": 0.19430843889713287,
      "num_tokens": 26375489.0,
      "step": 13020
    },
    {
      "entropy": 5.842823553085327,
      "epoch": 1.1382504587957702,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004876361766245661,
      "loss": 5.3901,
      "mean_token_accuracy": 0.19274840354919434,
      "num_tokens": 26385126.0,
      "step": 13025
    },
    {
      "entropy": 5.859316205978393,
      "epoch": 1.138687407148475,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004876259871881979,
      "loss": 5.3714,
      "mean_token_accuracy": 0.18795519173145295,
      "num_tokens": 26395314.0,
      "step": 13030
    },
    {
      "entropy": 5.981992483139038,
      "epoch": 1.1391243555011799,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004876157936735519,
      "loss": 5.4572,
      "mean_token_accuracy": 0.1807364195585251,
      "num_tokens": 26404660.0,
      "step": 13035
    },
    {
      "entropy": 5.967302322387695,
      "epoch": 1.1395613038538845,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048760559608082343,
      "loss": 5.4678,
      "mean_token_accuracy": 0.18867459744215012,
      "num_tokens": 26415371.0,
      "step": 13040
    },
    {
      "entropy": 5.908374071121216,
      "epoch": 1.1399982522065892,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004875953944102084,
      "loss": 5.4521,
      "mean_token_accuracy": 0.18323223888874055,
      "num_tokens": 26424896.0,
      "step": 13045
    },
    {
      "entropy": 5.855633020401001,
      "epoch": 1.140435200559294,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048758518866190214,
      "loss": 5.4148,
      "mean_token_accuracy": 0.18370802700519562,
      "num_tokens": 26434847.0,
      "step": 13050
    },
    {
      "entropy": 5.933574247360229,
      "epoch": 1.1408721489119986,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004875749788361006,
      "loss": 5.4002,
      "mean_token_accuracy": 0.1867118999361992,
      "num_tokens": 26444138.0,
      "step": 13055
    },
    {
      "entropy": 5.964991283416748,
      "epoch": 1.1413090972647033,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004875647649329997,
      "loss": 5.4071,
      "mean_token_accuracy": 0.1814004063606262,
      "num_tokens": 26453549.0,
      "step": 13060
    },
    {
      "entropy": 5.881485319137573,
      "epoch": 1.141746045617408,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004875545469527952,
      "loss": 5.4316,
      "mean_token_accuracy": 0.18597008138895035,
      "num_tokens": 26464386.0,
      "step": 13065
    },
    {
      "entropy": 5.899832105636596,
      "epoch": 1.1421829939701127,
      "grad_norm": 1.015625,
      "learning_rate": 0.000487544324895683,
      "loss": 5.3917,
      "mean_token_accuracy": 0.18745290786027907,
      "num_tokens": 26474146.0,
      "step": 13070
    },
    {
      "entropy": 5.882105827331543,
      "epoch": 1.1426199423228174,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048753409876185934,
      "loss": 5.3443,
      "mean_token_accuracy": 0.1974062889814377,
      "num_tokens": 26484408.0,
      "step": 13075
    },
    {
      "entropy": 5.886096239089966,
      "epoch": 1.143056890675522,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004875238685515203,
      "loss": 5.3593,
      "mean_token_accuracy": 0.1844846561551094,
      "num_tokens": 26494769.0,
      "step": 13080
    },
    {
      "entropy": 5.855268287658691,
      "epoch": 1.1434938390282268,
      "grad_norm": 1.0,
      "learning_rate": 0.00048751363426486206,
      "loss": 5.4434,
      "mean_token_accuracy": 0.18388566821813584,
      "num_tokens": 26505917.0,
      "step": 13085
    },
    {
      "entropy": 5.939682149887085,
      "epoch": 1.1439307873809317,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000487503395902081,
      "loss": 5.4639,
      "mean_token_accuracy": 0.17817013412714006,
      "num_tokens": 26516388.0,
      "step": 13090
    },
    {
      "entropy": 6.046122694015503,
      "epoch": 1.1443677357336364,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004874931534633735,
      "loss": 5.462,
      "mean_token_accuracy": 0.17709963172674179,
      "num_tokens": 26527180.0,
      "step": 13095
    },
    {
      "entropy": 5.874521780014038,
      "epoch": 1.144804684086341,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004874829069489359,
      "loss": 5.4267,
      "mean_token_accuracy": 0.18239623457193374,
      "num_tokens": 26537705.0,
      "step": 13100
    },
    {
      "entropy": 5.933683300018311,
      "epoch": 1.1452416324390458,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048747265635896487,
      "loss": 5.6027,
      "mean_token_accuracy": 0.17391210049390793,
      "num_tokens": 26548862.0,
      "step": 13105
    },
    {
      "entropy": 5.968393421173095,
      "epoch": 1.1456785807917504,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048746240169365683,
      "loss": 5.4611,
      "mean_token_accuracy": 0.18263427317142486,
      "num_tokens": 26559344.0,
      "step": 13110
    },
    {
      "entropy": 5.88459267616272,
      "epoch": 1.1461155291444551,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004874521429532087,
      "loss": 5.4051,
      "mean_token_accuracy": 0.18797842562198638,
      "num_tokens": 26568439.0,
      "step": 13115
    },
    {
      "entropy": 5.877806043624878,
      "epoch": 1.1465524774971598,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048744188013781714,
      "loss": 5.4075,
      "mean_token_accuracy": 0.18957428485155106,
      "num_tokens": 26578588.0,
      "step": 13120
    },
    {
      "entropy": 5.866632175445557,
      "epoch": 1.1469894258498645,
      "grad_norm": 0.96875,
      "learning_rate": 0.000487431613247679,
      "loss": 5.5001,
      "mean_token_accuracy": 0.18379813730716704,
      "num_tokens": 26588491.0,
      "step": 13125
    },
    {
      "entropy": 5.9319408416748045,
      "epoch": 1.1474263742025692,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048742134228299116,
      "loss": 5.4666,
      "mean_token_accuracy": 0.18417637944221496,
      "num_tokens": 26598945.0,
      "step": 13130
    },
    {
      "entropy": 5.874337577819825,
      "epoch": 1.147863322555274,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004874110672439508,
      "loss": 5.3721,
      "mean_token_accuracy": 0.18822521418333055,
      "num_tokens": 26609842.0,
      "step": 13135
    },
    {
      "entropy": 5.961509561538696,
      "epoch": 1.1483002709079786,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004874007881307547,
      "loss": 5.4861,
      "mean_token_accuracy": 0.18647308945655822,
      "num_tokens": 26620836.0,
      "step": 13140
    },
    {
      "entropy": 5.8764557361602785,
      "epoch": 1.1487372192606835,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048739050494360026,
      "loss": 5.4709,
      "mean_token_accuracy": 0.18343614786863327,
      "num_tokens": 26630148.0,
      "step": 13145
    },
    {
      "entropy": 5.9619384765625,
      "epoch": 1.149174167613388,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048738021768268463,
      "loss": 5.4525,
      "mean_token_accuracy": 0.18476019203662872,
      "num_tokens": 26640518.0,
      "step": 13150
    },
    {
      "entropy": 5.884023857116699,
      "epoch": 1.149611115966093,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004873699263482052,
      "loss": 5.3905,
      "mean_token_accuracy": 0.19663336127996445,
      "num_tokens": 26650156.0,
      "step": 13155
    },
    {
      "entropy": 5.906725168228149,
      "epoch": 1.1500480643187976,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048735963094035924,
      "loss": 5.4421,
      "mean_token_accuracy": 0.1822680190205574,
      "num_tokens": 26661251.0,
      "step": 13160
    },
    {
      "entropy": 5.950852251052856,
      "epoch": 1.1504850126715023,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004873493314593443,
      "loss": 5.4226,
      "mean_token_accuracy": 0.19162434041500093,
      "num_tokens": 26671851.0,
      "step": 13165
    },
    {
      "entropy": 5.901687860488892,
      "epoch": 1.150921961024207,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004873390279053579,
      "loss": 5.4413,
      "mean_token_accuracy": 0.18133897930383683,
      "num_tokens": 26682984.0,
      "step": 13170
    },
    {
      "entropy": 5.918591260910034,
      "epoch": 1.1513589093769117,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004873287202785977,
      "loss": 5.3935,
      "mean_token_accuracy": 0.18840377926826476,
      "num_tokens": 26692706.0,
      "step": 13175
    },
    {
      "entropy": 5.926123762130738,
      "epoch": 1.1517958577296163,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004873184085792614,
      "loss": 5.4319,
      "mean_token_accuracy": 0.18770415633916854,
      "num_tokens": 26702152.0,
      "step": 13180
    },
    {
      "entropy": 5.868023300170899,
      "epoch": 1.152232806082321,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048730809280754674,
      "loss": 5.3932,
      "mean_token_accuracy": 0.19349753111600876,
      "num_tokens": 26711404.0,
      "step": 13185
    },
    {
      "entropy": 5.934328317642212,
      "epoch": 1.1526697544350257,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004872977729636516,
      "loss": 5.4062,
      "mean_token_accuracy": 0.1926071509718895,
      "num_tokens": 26720776.0,
      "step": 13190
    },
    {
      "entropy": 5.904833507537842,
      "epoch": 1.1531067027877304,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048728744904777394,
      "loss": 5.441,
      "mean_token_accuracy": 0.1881423622369766,
      "num_tokens": 26730700.0,
      "step": 13195
    },
    {
      "entropy": 5.93522424697876,
      "epoch": 1.1535436511404351,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048727712106011185,
      "loss": 5.4687,
      "mean_token_accuracy": 0.18370471149683,
      "num_tokens": 26741453.0,
      "step": 13200
    },
    {
      "entropy": 5.902590274810791,
      "epoch": 1.1539805994931398,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048726678900086327,
      "loss": 5.4483,
      "mean_token_accuracy": 0.18379174172878265,
      "num_tokens": 26750627.0,
      "step": 13205
    },
    {
      "entropy": 5.96935510635376,
      "epoch": 1.1544175478458447,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048725645287022644,
      "loss": 5.4307,
      "mean_token_accuracy": 0.18417043387889862,
      "num_tokens": 26760820.0,
      "step": 13210
    },
    {
      "entropy": 5.912472295761108,
      "epoch": 1.1548544961985494,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004872461126683996,
      "loss": 5.3527,
      "mean_token_accuracy": 0.18626405596733092,
      "num_tokens": 26771535.0,
      "step": 13215
    },
    {
      "entropy": 5.9118694305419925,
      "epoch": 1.155291444551254,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004872357683955812,
      "loss": 5.4705,
      "mean_token_accuracy": 0.17806640565395354,
      "num_tokens": 26782755.0,
      "step": 13220
    },
    {
      "entropy": 5.979738759994507,
      "epoch": 1.1557283929039588,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048722542005196954,
      "loss": 5.412,
      "mean_token_accuracy": 0.1880972608923912,
      "num_tokens": 26792834.0,
      "step": 13225
    },
    {
      "entropy": 5.87319540977478,
      "epoch": 1.1561653412566635,
      "grad_norm": 1.0,
      "learning_rate": 0.00048721506763776306,
      "loss": 5.4387,
      "mean_token_accuracy": 0.18915383964776994,
      "num_tokens": 26802865.0,
      "step": 13230
    },
    {
      "entropy": 5.895325994491577,
      "epoch": 1.1566022896093682,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004872047111531604,
      "loss": 5.4187,
      "mean_token_accuracy": 0.1856602504849434,
      "num_tokens": 26813566.0,
      "step": 13235
    },
    {
      "entropy": 5.939857244491577,
      "epoch": 1.1570392379620729,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004871943505983602,
      "loss": 5.3867,
      "mean_token_accuracy": 0.19569242149591445,
      "num_tokens": 26823885.0,
      "step": 13240
    },
    {
      "entropy": 5.8192572593688965,
      "epoch": 1.1574761863147776,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004871839859735612,
      "loss": 5.3825,
      "mean_token_accuracy": 0.1912005826830864,
      "num_tokens": 26833236.0,
      "step": 13245
    },
    {
      "entropy": 5.891085815429688,
      "epoch": 1.1579131346674822,
      "grad_norm": 1.0,
      "learning_rate": 0.00048717361727896203,
      "loss": 5.4812,
      "mean_token_accuracy": 0.18334946185350418,
      "num_tokens": 26842651.0,
      "step": 13250
    },
    {
      "entropy": 5.916226053237915,
      "epoch": 1.158350083020187,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048716324451476183,
      "loss": 5.3832,
      "mean_token_accuracy": 0.18329836428165436,
      "num_tokens": 26852198.0,
      "step": 13255
    },
    {
      "entropy": 5.931845712661743,
      "epoch": 1.1587870313728916,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004871528676811594,
      "loss": 5.5016,
      "mean_token_accuracy": 0.18253403306007385,
      "num_tokens": 26862807.0,
      "step": 13260
    },
    {
      "entropy": 5.970778274536133,
      "epoch": 1.1592239797255965,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004871424867783538,
      "loss": 5.4739,
      "mean_token_accuracy": 0.1757626548409462,
      "num_tokens": 26871992.0,
      "step": 13265
    },
    {
      "entropy": 5.899312162399292,
      "epoch": 1.1596609280783012,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048713210180654414,
      "loss": 5.4483,
      "mean_token_accuracy": 0.1861603006720543,
      "num_tokens": 26881417.0,
      "step": 13270
    },
    {
      "entropy": 5.863110828399658,
      "epoch": 1.160097876431006,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048712171276592964,
      "loss": 5.3673,
      "mean_token_accuracy": 0.19361750036478043,
      "num_tokens": 26891879.0,
      "step": 13275
    },
    {
      "entropy": 5.903482723236084,
      "epoch": 1.1605348247837106,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004871113196567095,
      "loss": 5.529,
      "mean_token_accuracy": 0.18085136860609055,
      "num_tokens": 26902729.0,
      "step": 13280
    },
    {
      "entropy": 5.877122116088867,
      "epoch": 1.1609717731364153,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048710092247908316,
      "loss": 5.4204,
      "mean_token_accuracy": 0.18764560371637345,
      "num_tokens": 26914415.0,
      "step": 13285
    },
    {
      "entropy": 5.97261323928833,
      "epoch": 1.16140872148912,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048709052123324995,
      "loss": 5.4605,
      "mean_token_accuracy": 0.1869677037000656,
      "num_tokens": 26924513.0,
      "step": 13290
    },
    {
      "entropy": 5.837568998336792,
      "epoch": 1.1618456698418247,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048708011591940936,
      "loss": 5.355,
      "mean_token_accuracy": 0.19860585927963256,
      "num_tokens": 26935039.0,
      "step": 13295
    },
    {
      "entropy": 5.847805690765381,
      "epoch": 1.1622826181945294,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000487069706537761,
      "loss": 5.3439,
      "mean_token_accuracy": 0.19541891664266586,
      "num_tokens": 26945148.0,
      "step": 13300
    },
    {
      "entropy": 5.893123531341553,
      "epoch": 1.162719566547234,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004870592930885046,
      "loss": 5.3863,
      "mean_token_accuracy": 0.18971287310123444,
      "num_tokens": 26955227.0,
      "step": 13305
    },
    {
      "entropy": 5.889494466781616,
      "epoch": 1.1631565148999388,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048704887557183993,
      "loss": 5.3735,
      "mean_token_accuracy": 0.18409813195466995,
      "num_tokens": 26964389.0,
      "step": 13310
    },
    {
      "entropy": 5.912398481369019,
      "epoch": 1.1635934632526435,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048703845398796665,
      "loss": 5.3723,
      "mean_token_accuracy": 0.1910355031490326,
      "num_tokens": 26974695.0,
      "step": 13315
    },
    {
      "entropy": 5.931927299499511,
      "epoch": 1.1640304116053484,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048702802833708466,
      "loss": 5.3691,
      "mean_token_accuracy": 0.19444576501846314,
      "num_tokens": 26984851.0,
      "step": 13320
    },
    {
      "entropy": 5.9451830863952635,
      "epoch": 1.1644673599580528,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048701759861939403,
      "loss": 5.4339,
      "mean_token_accuracy": 0.19086842685937883,
      "num_tokens": 26995967.0,
      "step": 13325
    },
    {
      "entropy": 5.8642237186431885,
      "epoch": 1.1649043083107578,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048700716483509476,
      "loss": 5.3709,
      "mean_token_accuracy": 0.1980260968208313,
      "num_tokens": 27006469.0,
      "step": 13330
    },
    {
      "entropy": 5.939092969894409,
      "epoch": 1.1653412566634624,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000486996726984387,
      "loss": 5.4872,
      "mean_token_accuracy": 0.18335025608539582,
      "num_tokens": 27016419.0,
      "step": 13335
    },
    {
      "entropy": 5.934697532653809,
      "epoch": 1.1657782050161671,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004869862850674709,
      "loss": 5.4746,
      "mean_token_accuracy": 0.18127274811267852,
      "num_tokens": 27026322.0,
      "step": 13340
    },
    {
      "entropy": 5.861299753189087,
      "epoch": 1.1662151533688718,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048697583908454675,
      "loss": 5.4071,
      "mean_token_accuracy": 0.1835213616490364,
      "num_tokens": 27036518.0,
      "step": 13345
    },
    {
      "entropy": 5.928552484512329,
      "epoch": 1.1666521017215765,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000486965389035815,
      "loss": 5.4608,
      "mean_token_accuracy": 0.1823159247636795,
      "num_tokens": 27048035.0,
      "step": 13350
    },
    {
      "entropy": 5.922189426422119,
      "epoch": 1.1670890500742812,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048695493492147596,
      "loss": 5.4298,
      "mean_token_accuracy": 0.18342245668172835,
      "num_tokens": 27058507.0,
      "step": 13355
    },
    {
      "entropy": 5.87492094039917,
      "epoch": 1.167525998426986,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048694447674173025,
      "loss": 5.4282,
      "mean_token_accuracy": 0.18995942771434784,
      "num_tokens": 27068350.0,
      "step": 13360
    },
    {
      "entropy": 5.835802030563355,
      "epoch": 1.1679629467796906,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004869340144967784,
      "loss": 5.415,
      "mean_token_accuracy": 0.18773578256368637,
      "num_tokens": 27077840.0,
      "step": 13365
    },
    {
      "entropy": 5.870659685134887,
      "epoch": 1.1683998951323953,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048692354818682113,
      "loss": 5.3788,
      "mean_token_accuracy": 0.1861558347940445,
      "num_tokens": 27087687.0,
      "step": 13370
    },
    {
      "entropy": 5.969425201416016,
      "epoch": 1.1688368434851,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004869130778120591,
      "loss": 5.5377,
      "mean_token_accuracy": 0.18092977851629258,
      "num_tokens": 27099057.0,
      "step": 13375
    },
    {
      "entropy": 5.910357904434204,
      "epoch": 1.1692737918378047,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004869026033726933,
      "loss": 5.4976,
      "mean_token_accuracy": 0.18495589047670363,
      "num_tokens": 27109617.0,
      "step": 13380
    },
    {
      "entropy": 5.926921653747558,
      "epoch": 1.1697107401905096,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004868921248689244,
      "loss": 5.4608,
      "mean_token_accuracy": 0.1765725150704384,
      "num_tokens": 27119456.0,
      "step": 13385
    },
    {
      "entropy": 5.881301307678223,
      "epoch": 1.1701476885432143,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004868816423009535,
      "loss": 5.3661,
      "mean_token_accuracy": 0.18941093981266022,
      "num_tokens": 27129333.0,
      "step": 13390
    },
    {
      "entropy": 5.897729063034058,
      "epoch": 1.170584636895919,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004868711556689818,
      "loss": 5.4021,
      "mean_token_accuracy": 0.1831916019320488,
      "num_tokens": 27139205.0,
      "step": 13395
    },
    {
      "entropy": 5.922893524169922,
      "epoch": 1.1710215852486237,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048686066497321027,
      "loss": 5.4497,
      "mean_token_accuracy": 0.19185693860054015,
      "num_tokens": 27148530.0,
      "step": 13400
    },
    {
      "entropy": 5.9587808609008786,
      "epoch": 1.1714585336013283,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048685017021384014,
      "loss": 5.5191,
      "mean_token_accuracy": 0.1831662252545357,
      "num_tokens": 27159000.0,
      "step": 13405
    },
    {
      "entropy": 5.972366571426392,
      "epoch": 1.171895481954033,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048683967139107276,
      "loss": 5.4456,
      "mean_token_accuracy": 0.18697132915258408,
      "num_tokens": 27169035.0,
      "step": 13410
    },
    {
      "entropy": 5.911486959457397,
      "epoch": 1.1723324303067377,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004868291685051095,
      "loss": 5.4601,
      "mean_token_accuracy": 0.1824149027466774,
      "num_tokens": 27179809.0,
      "step": 13415
    },
    {
      "entropy": 5.902546787261963,
      "epoch": 1.1727693786594424,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004868186615561518,
      "loss": 5.4278,
      "mean_token_accuracy": 0.1891435220837593,
      "num_tokens": 27190445.0,
      "step": 13420
    },
    {
      "entropy": 5.934472942352295,
      "epoch": 1.173206327012147,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048680815054440114,
      "loss": 5.3884,
      "mean_token_accuracy": 0.18509516417980193,
      "num_tokens": 27201097.0,
      "step": 13425
    },
    {
      "entropy": 5.908805179595947,
      "epoch": 1.1736432753648518,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048679763547005925,
      "loss": 5.3873,
      "mean_token_accuracy": 0.19070715308189393,
      "num_tokens": 27210181.0,
      "step": 13430
    },
    {
      "entropy": 5.90688419342041,
      "epoch": 1.1740802237175565,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004867871163333276,
      "loss": 5.3807,
      "mean_token_accuracy": 0.18854737877845765,
      "num_tokens": 27219869.0,
      "step": 13435
    },
    {
      "entropy": 5.926894760131836,
      "epoch": 1.1745171720702614,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004867765931344082,
      "loss": 5.4253,
      "mean_token_accuracy": 0.18806055933237076,
      "num_tokens": 27229730.0,
      "step": 13440
    },
    {
      "entropy": 5.896069622039795,
      "epoch": 1.174954120422966,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004867660658735027,
      "loss": 5.432,
      "mean_token_accuracy": 0.18719345182180405,
      "num_tokens": 27239534.0,
      "step": 13445
    },
    {
      "entropy": 5.895239400863647,
      "epoch": 1.1753910687756708,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004867555345508131,
      "loss": 5.4984,
      "mean_token_accuracy": 0.1786297157406807,
      "num_tokens": 27249017.0,
      "step": 13450
    },
    {
      "entropy": 5.9561872482299805,
      "epoch": 1.1758280171283755,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004867449991665414,
      "loss": 5.5262,
      "mean_token_accuracy": 0.18092494755983352,
      "num_tokens": 27261135.0,
      "step": 13455
    },
    {
      "entropy": 5.941755580902099,
      "epoch": 1.1762649654810802,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004867344597208897,
      "loss": 5.4002,
      "mean_token_accuracy": 0.1828320249915123,
      "num_tokens": 27270763.0,
      "step": 13460
    },
    {
      "entropy": 5.92633090019226,
      "epoch": 1.1767019138337849,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004867239162140601,
      "loss": 5.46,
      "mean_token_accuracy": 0.18304837346076966,
      "num_tokens": 27280115.0,
      "step": 13465
    },
    {
      "entropy": 5.902360200881958,
      "epoch": 1.1771388621864896,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004867133686462548,
      "loss": 5.4052,
      "mean_token_accuracy": 0.18600647449493407,
      "num_tokens": 27290162.0,
      "step": 13470
    },
    {
      "entropy": 5.892708873748779,
      "epoch": 1.1775758105391942,
      "grad_norm": 1.0,
      "learning_rate": 0.00048670281701767616,
      "loss": 5.3514,
      "mean_token_accuracy": 0.199081489443779,
      "num_tokens": 27300467.0,
      "step": 13475
    },
    {
      "entropy": 5.89604868888855,
      "epoch": 1.178012758891899,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004866922613285266,
      "loss": 5.4112,
      "mean_token_accuracy": 0.19364064931869507,
      "num_tokens": 27310524.0,
      "step": 13480
    },
    {
      "entropy": 5.906346416473388,
      "epoch": 1.1784497072446036,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004866817015790085,
      "loss": 5.4434,
      "mean_token_accuracy": 0.18836638033390046,
      "num_tokens": 27320861.0,
      "step": 13485
    },
    {
      "entropy": 5.8906749248504635,
      "epoch": 1.1788866555973083,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048667113776932453,
      "loss": 5.4613,
      "mean_token_accuracy": 0.1855959713459015,
      "num_tokens": 27331771.0,
      "step": 13490
    },
    {
      "entropy": 5.941611909866333,
      "epoch": 1.1793236039500132,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004866605698996772,
      "loss": 5.4318,
      "mean_token_accuracy": 0.18423112332820893,
      "num_tokens": 27341449.0,
      "step": 13495
    },
    {
      "entropy": 5.936126279830932,
      "epoch": 1.1797605523027177,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004866499979702691,
      "loss": 5.5081,
      "mean_token_accuracy": 0.18165443688631058,
      "num_tokens": 27353127.0,
      "step": 13500
    },
    {
      "entropy": 5.929058408737182,
      "epoch": 1.1801975006554226,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004866394219813032,
      "loss": 5.3956,
      "mean_token_accuracy": 0.18730926662683486,
      "num_tokens": 27362116.0,
      "step": 13505
    },
    {
      "entropy": 5.879386854171753,
      "epoch": 1.1806344490081273,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048662884193298233,
      "loss": 5.4162,
      "mean_token_accuracy": 0.18988312482833863,
      "num_tokens": 27372204.0,
      "step": 13510
    },
    {
      "entropy": 5.940651226043701,
      "epoch": 1.181071397360832,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004866182578255094,
      "loss": 5.4418,
      "mean_token_accuracy": 0.18776460438966752,
      "num_tokens": 27382963.0,
      "step": 13515
    },
    {
      "entropy": 5.929487848281861,
      "epoch": 1.1815083457135367,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048660766965908736,
      "loss": 5.427,
      "mean_token_accuracy": 0.18676802963018418,
      "num_tokens": 27392426.0,
      "step": 13520
    },
    {
      "entropy": 5.861693906784057,
      "epoch": 1.1819452940662414,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004865970774339194,
      "loss": 5.3855,
      "mean_token_accuracy": 0.19154687821865082,
      "num_tokens": 27402679.0,
      "step": 13525
    },
    {
      "entropy": 5.922818851470947,
      "epoch": 1.182382242418946,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048658648115020855,
      "loss": 5.463,
      "mean_token_accuracy": 0.1868848666548729,
      "num_tokens": 27412911.0,
      "step": 13530
    },
    {
      "entropy": 5.9512992858886715,
      "epoch": 1.1828191907716508,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004865758808081581,
      "loss": 5.3933,
      "mean_token_accuracy": 0.18820076435804367,
      "num_tokens": 27423394.0,
      "step": 13535
    },
    {
      "entropy": 5.882066106796264,
      "epoch": 1.1832561391243555,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048656527640797144,
      "loss": 5.3816,
      "mean_token_accuracy": 0.19377702474594116,
      "num_tokens": 27433016.0,
      "step": 13540
    },
    {
      "entropy": 5.899059629440307,
      "epoch": 1.1836930874770601,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048655466794985194,
      "loss": 5.4097,
      "mean_token_accuracy": 0.1836561754345894,
      "num_tokens": 27442414.0,
      "step": 13545
    },
    {
      "entropy": 5.862533760070801,
      "epoch": 1.1841300358297648,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000486544055434003,
      "loss": 5.4803,
      "mean_token_accuracy": 0.18564198166131973,
      "num_tokens": 27452410.0,
      "step": 13550
    },
    {
      "entropy": 5.936806726455688,
      "epoch": 1.1845669841824695,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004865334388606282,
      "loss": 5.3876,
      "mean_token_accuracy": 0.19439693689346313,
      "num_tokens": 27462196.0,
      "step": 13555
    },
    {
      "entropy": 5.935538673400879,
      "epoch": 1.1850039325351744,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048652281822993127,
      "loss": 5.4936,
      "mean_token_accuracy": 0.17901546210050584,
      "num_tokens": 27473934.0,
      "step": 13560
    },
    {
      "entropy": 5.948485279083252,
      "epoch": 1.1854408808878791,
      "grad_norm": 1.0,
      "learning_rate": 0.00048651219354211585,
      "loss": 5.5122,
      "mean_token_accuracy": 0.1835333913564682,
      "num_tokens": 27484218.0,
      "step": 13565
    },
    {
      "entropy": 5.8962428092956545,
      "epoch": 1.1858778292405838,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048650156479738575,
      "loss": 5.4854,
      "mean_token_accuracy": 0.18982964307069777,
      "num_tokens": 27494663.0,
      "step": 13570
    },
    {
      "entropy": 5.941255569458008,
      "epoch": 1.1863147775932885,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048649093199594473,
      "loss": 5.3812,
      "mean_token_accuracy": 0.1890908494591713,
      "num_tokens": 27504233.0,
      "step": 13575
    },
    {
      "entropy": 5.912873315811157,
      "epoch": 1.1867517259459932,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004864802951379969,
      "loss": 5.4133,
      "mean_token_accuracy": 0.1848856672644615,
      "num_tokens": 27514032.0,
      "step": 13580
    },
    {
      "entropy": 5.906877183914185,
      "epoch": 1.187188674298698,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048646965422374615,
      "loss": 5.4945,
      "mean_token_accuracy": 0.187015537917614,
      "num_tokens": 27524659.0,
      "step": 13585
    },
    {
      "entropy": 5.900476408004761,
      "epoch": 1.1876256226514026,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004864590092533966,
      "loss": 5.3897,
      "mean_token_accuracy": 0.18745032995939254,
      "num_tokens": 27534003.0,
      "step": 13590
    },
    {
      "entropy": 5.8948493003845215,
      "epoch": 1.1880625710041073,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004864483602271524,
      "loss": 5.3801,
      "mean_token_accuracy": 0.1897505447268486,
      "num_tokens": 27543586.0,
      "step": 13595
    },
    {
      "entropy": 5.83880844116211,
      "epoch": 1.188499519356812,
      "grad_norm": 1.0,
      "learning_rate": 0.000486437707145218,
      "loss": 5.3755,
      "mean_token_accuracy": 0.19126184582710265,
      "num_tokens": 27554210.0,
      "step": 13600
    },
    {
      "entropy": 5.961216688156128,
      "epoch": 1.1889364677095167,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048642705000779754,
      "loss": 5.4879,
      "mean_token_accuracy": 0.18774239718914032,
      "num_tokens": 27565324.0,
      "step": 13605
    },
    {
      "entropy": 5.959632873535156,
      "epoch": 1.1893734160622214,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004864163888150955,
      "loss": 5.4078,
      "mean_token_accuracy": 0.18437501341104506,
      "num_tokens": 27575192.0,
      "step": 13610
    },
    {
      "entropy": 5.8900378227233885,
      "epoch": 1.1898103644149263,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048640572356731623,
      "loss": 5.4052,
      "mean_token_accuracy": 0.1854822114109993,
      "num_tokens": 27585294.0,
      "step": 13615
    },
    {
      "entropy": 5.900397109985351,
      "epoch": 1.190247312767631,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048639505426466447,
      "loss": 5.4214,
      "mean_token_accuracy": 0.18857598155736924,
      "num_tokens": 27595667.0,
      "step": 13620
    },
    {
      "entropy": 5.930958986282349,
      "epoch": 1.1906842611203357,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004863843809073448,
      "loss": 5.4026,
      "mean_token_accuracy": 0.19348912984132766,
      "num_tokens": 27605816.0,
      "step": 13625
    },
    {
      "entropy": 5.918566513061523,
      "epoch": 1.1911212094730403,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000486373703495562,
      "loss": 5.4232,
      "mean_token_accuracy": 0.19261527359485625,
      "num_tokens": 27614874.0,
      "step": 13630
    },
    {
      "entropy": 5.901441287994385,
      "epoch": 1.191558157825745,
      "grad_norm": 1.25,
      "learning_rate": 0.0004863630220295207,
      "loss": 5.3831,
      "mean_token_accuracy": 0.18751581460237504,
      "num_tokens": 27625610.0,
      "step": 13635
    },
    {
      "entropy": 5.884318685531616,
      "epoch": 1.1919951061784497,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048635233650942596,
      "loss": 5.4459,
      "mean_token_accuracy": 0.18078674972057343,
      "num_tokens": 27636646.0,
      "step": 13640
    },
    {
      "entropy": 5.939125442504883,
      "epoch": 1.1924320545311544,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048634164693548267,
      "loss": 5.4321,
      "mean_token_accuracy": 0.18753560483455659,
      "num_tokens": 27647311.0,
      "step": 13645
    },
    {
      "entropy": 5.918542003631591,
      "epoch": 1.192869002883859,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004863309533078957,
      "loss": 5.459,
      "mean_token_accuracy": 0.18094117492437362,
      "num_tokens": 27658347.0,
      "step": 13650
    },
    {
      "entropy": 5.846793556213379,
      "epoch": 1.1933059512365638,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048632025562687047,
      "loss": 5.4061,
      "mean_token_accuracy": 0.19455469846725465,
      "num_tokens": 27669211.0,
      "step": 13655
    },
    {
      "entropy": 5.889489555358887,
      "epoch": 1.1937428995892685,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048630955389261193,
      "loss": 5.4029,
      "mean_token_accuracy": 0.18614855110645295,
      "num_tokens": 27678904.0,
      "step": 13660
    },
    {
      "entropy": 5.913536930084229,
      "epoch": 1.1941798479419732,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048629884810532547,
      "loss": 5.3597,
      "mean_token_accuracy": 0.19368251860141755,
      "num_tokens": 27688928.0,
      "step": 13665
    },
    {
      "entropy": 5.8937318325042725,
      "epoch": 1.1946167962946779,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004862881382652164,
      "loss": 5.4529,
      "mean_token_accuracy": 0.1843964383006096,
      "num_tokens": 27699855.0,
      "step": 13670
    },
    {
      "entropy": 5.931383991241455,
      "epoch": 1.1950537446473826,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004862774243724901,
      "loss": 5.4276,
      "mean_token_accuracy": 0.19214920848608016,
      "num_tokens": 27709857.0,
      "step": 13675
    },
    {
      "entropy": 5.9646812915802006,
      "epoch": 1.1954906930000875,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004862667064273521,
      "loss": 5.4524,
      "mean_token_accuracy": 0.18666862100362777,
      "num_tokens": 27719612.0,
      "step": 13680
    },
    {
      "entropy": 5.863973379135132,
      "epoch": 1.1959276413527922,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004862559844300079,
      "loss": 5.4297,
      "mean_token_accuracy": 0.18873930871486663,
      "num_tokens": 27730052.0,
      "step": 13685
    },
    {
      "entropy": 5.844094944000244,
      "epoch": 1.1963645897054969,
      "grad_norm": 1.0,
      "learning_rate": 0.0004862452583806633,
      "loss": 5.3274,
      "mean_token_accuracy": 0.20251837968826295,
      "num_tokens": 27739980.0,
      "step": 13690
    },
    {
      "entropy": 5.900234794616699,
      "epoch": 1.1968015380582016,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048623452827952385,
      "loss": 5.4387,
      "mean_token_accuracy": 0.18855753242969514,
      "num_tokens": 27748863.0,
      "step": 13695
    },
    {
      "entropy": 5.863984251022339,
      "epoch": 1.1972384864109062,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004862237941267956,
      "loss": 5.404,
      "mean_token_accuracy": 0.1916792333126068,
      "num_tokens": 27758672.0,
      "step": 13700
    },
    {
      "entropy": 5.932933187484741,
      "epoch": 1.197675434763611,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004862130559226842,
      "loss": 5.4353,
      "mean_token_accuracy": 0.18824968189001084,
      "num_tokens": 27768195.0,
      "step": 13705
    },
    {
      "entropy": 5.83626389503479,
      "epoch": 1.1981123831163156,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004862023136673958,
      "loss": 5.3657,
      "mean_token_accuracy": 0.19778419882059098,
      "num_tokens": 27778135.0,
      "step": 13710
    },
    {
      "entropy": 5.947547149658203,
      "epoch": 1.1985493314690203,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048619156736113634,
      "loss": 5.4614,
      "mean_token_accuracy": 0.18169374465942384,
      "num_tokens": 27788723.0,
      "step": 13715
    },
    {
      "entropy": 5.89508466720581,
      "epoch": 1.198986279821725,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004861808170041119,
      "loss": 5.4445,
      "mean_token_accuracy": 0.1878601238131523,
      "num_tokens": 27799060.0,
      "step": 13720
    },
    {
      "entropy": 5.914816856384277,
      "epoch": 1.1994232281744297,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004861700625965287,
      "loss": 5.4684,
      "mean_token_accuracy": 0.18419447988271714,
      "num_tokens": 27809542.0,
      "step": 13725
    },
    {
      "entropy": 5.905847835540771,
      "epoch": 1.1998601765271344,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048615930413859317,
      "loss": 5.4376,
      "mean_token_accuracy": 0.18501867353916168,
      "num_tokens": 27819738.0,
      "step": 13730
    },
    {
      "entropy": 5.942208576202392,
      "epoch": 1.2002971248798393,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048614854163051143,
      "loss": 5.4067,
      "mean_token_accuracy": 0.18465443551540375,
      "num_tokens": 27829052.0,
      "step": 13735
    },
    {
      "entropy": 5.9071800231933596,
      "epoch": 1.200734073232544,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004861377750724901,
      "loss": 5.3842,
      "mean_token_accuracy": 0.19011755138635636,
      "num_tokens": 27839277.0,
      "step": 13740
    },
    {
      "entropy": 5.93560004234314,
      "epoch": 1.2011710215852487,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004861270044647356,
      "loss": 5.4239,
      "mean_token_accuracy": 0.18515996932983397,
      "num_tokens": 27848550.0,
      "step": 13745
    },
    {
      "entropy": 5.953679084777832,
      "epoch": 1.2016079699379534,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048611622980745446,
      "loss": 5.4855,
      "mean_token_accuracy": 0.17401159256696702,
      "num_tokens": 27858546.0,
      "step": 13750
    },
    {
      "entropy": 5.922816562652588,
      "epoch": 1.202044918290658,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048610545110085347,
      "loss": 5.5033,
      "mean_token_accuracy": 0.17675969749689102,
      "num_tokens": 27869192.0,
      "step": 13755
    },
    {
      "entropy": 5.960075187683105,
      "epoch": 1.2024818666433628,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004860946683451393,
      "loss": 5.4659,
      "mean_token_accuracy": 0.19015710055828094,
      "num_tokens": 27878534.0,
      "step": 13760
    },
    {
      "entropy": 5.88929967880249,
      "epoch": 1.2029188149960675,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048608388154051875,
      "loss": 5.3713,
      "mean_token_accuracy": 0.1933937653899193,
      "num_tokens": 27888074.0,
      "step": 13765
    },
    {
      "entropy": 5.844871139526367,
      "epoch": 1.2033557633487721,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048607309068719875,
      "loss": 5.4592,
      "mean_token_accuracy": 0.17977949231863022,
      "num_tokens": 27898210.0,
      "step": 13770
    },
    {
      "entropy": 5.890784168243409,
      "epoch": 1.2037927117014768,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004860622957853863,
      "loss": 5.3721,
      "mean_token_accuracy": 0.1929750382900238,
      "num_tokens": 27907472.0,
      "step": 13775
    },
    {
      "entropy": 5.886312341690063,
      "epoch": 1.2042296600541815,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004860514968352885,
      "loss": 5.437,
      "mean_token_accuracy": 0.18656102567911148,
      "num_tokens": 27917265.0,
      "step": 13780
    },
    {
      "entropy": 5.875347232818603,
      "epoch": 1.2046666084068862,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048604069383711225,
      "loss": 5.3519,
      "mean_token_accuracy": 0.1993434876203537,
      "num_tokens": 27926883.0,
      "step": 13785
    },
    {
      "entropy": 5.961084842681885,
      "epoch": 1.2051035567595911,
      "grad_norm": 1.0,
      "learning_rate": 0.00048602988679106497,
      "loss": 5.3989,
      "mean_token_accuracy": 0.18489478081464766,
      "num_tokens": 27937167.0,
      "step": 13790
    },
    {
      "entropy": 5.878257322311401,
      "epoch": 1.2055405051122958,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004860190756973539,
      "loss": 5.3745,
      "mean_token_accuracy": 0.19751642942428588,
      "num_tokens": 27947422.0,
      "step": 13795
    },
    {
      "entropy": 5.88252272605896,
      "epoch": 1.2059774534650005,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048600826055618634,
      "loss": 5.4479,
      "mean_token_accuracy": 0.1842876046895981,
      "num_tokens": 27958407.0,
      "step": 13800
    },
    {
      "entropy": 5.91728835105896,
      "epoch": 1.2064144018177052,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048599744136776986,
      "loss": 5.3,
      "mean_token_accuracy": 0.1949502110481262,
      "num_tokens": 27968277.0,
      "step": 13805
    },
    {
      "entropy": 5.880324411392212,
      "epoch": 1.20685135017041,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048598661813231177,
      "loss": 5.342,
      "mean_token_accuracy": 0.19631233662366868,
      "num_tokens": 27978116.0,
      "step": 13810
    },
    {
      "entropy": 5.892468118667603,
      "epoch": 1.2072882985231146,
      "grad_norm": 1.125,
      "learning_rate": 0.00048597579085001984,
      "loss": 5.3544,
      "mean_token_accuracy": 0.1891017660498619,
      "num_tokens": 27988383.0,
      "step": 13815
    },
    {
      "entropy": 5.858812761306763,
      "epoch": 1.2077252468758193,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048596495952110176,
      "loss": 5.3647,
      "mean_token_accuracy": 0.19857774525880814,
      "num_tokens": 27999118.0,
      "step": 13820
    },
    {
      "entropy": 5.868727016448974,
      "epoch": 1.208162195228524,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004859541241457651,
      "loss": 5.3704,
      "mean_token_accuracy": 0.19236791729927064,
      "num_tokens": 28008849.0,
      "step": 13825
    },
    {
      "entropy": 5.904741477966309,
      "epoch": 1.2085991435812287,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004859432847242179,
      "loss": 5.4072,
      "mean_token_accuracy": 0.19068050533533096,
      "num_tokens": 28019063.0,
      "step": 13830
    },
    {
      "entropy": 5.872762823104859,
      "epoch": 1.2090360919339334,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048593244125666793,
      "loss": 5.3206,
      "mean_token_accuracy": 0.1917506128549576,
      "num_tokens": 28028333.0,
      "step": 13835
    },
    {
      "entropy": 5.906183338165283,
      "epoch": 1.209473040286638,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004859215937433232,
      "loss": 5.3949,
      "mean_token_accuracy": 0.19371555596590043,
      "num_tokens": 28037704.0,
      "step": 13840
    },
    {
      "entropy": 5.893036603927612,
      "epoch": 1.2099099886393427,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004859107421843918,
      "loss": 5.4178,
      "mean_token_accuracy": 0.19110537469387054,
      "num_tokens": 28047157.0,
      "step": 13845
    },
    {
      "entropy": 5.9495766162872314,
      "epoch": 1.2103469369920474,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048589988658008184,
      "loss": 5.3618,
      "mean_token_accuracy": 0.18787849992513656,
      "num_tokens": 28057582.0,
      "step": 13850
    },
    {
      "entropy": 5.879031324386597,
      "epoch": 1.2107838853447523,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048588902693060156,
      "loss": 5.4552,
      "mean_token_accuracy": 0.17862686961889268,
      "num_tokens": 28067458.0,
      "step": 13855
    },
    {
      "entropy": 5.926748895645142,
      "epoch": 1.211220833697457,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004858781632361592,
      "loss": 5.412,
      "mean_token_accuracy": 0.18901762515306472,
      "num_tokens": 28078967.0,
      "step": 13860
    },
    {
      "entropy": 5.880393171310425,
      "epoch": 1.2116577820501617,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004858672954969632,
      "loss": 5.3732,
      "mean_token_accuracy": 0.18931224048137665,
      "num_tokens": 28088498.0,
      "step": 13865
    },
    {
      "entropy": 5.850058507919312,
      "epoch": 1.2120947304028664,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048585642371322196,
      "loss": 5.4669,
      "mean_token_accuracy": 0.18889996707439421,
      "num_tokens": 28098884.0,
      "step": 13870
    },
    {
      "entropy": 5.8794450759887695,
      "epoch": 1.212531678755571,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004858455478851441,
      "loss": 5.3797,
      "mean_token_accuracy": 0.19863649755716323,
      "num_tokens": 28109629.0,
      "step": 13875
    },
    {
      "entropy": 5.931327295303345,
      "epoch": 1.2129686271082758,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000485834668012938,
      "loss": 5.3266,
      "mean_token_accuracy": 0.18935124278068544,
      "num_tokens": 28118555.0,
      "step": 13880
    },
    {
      "entropy": 5.966088724136353,
      "epoch": 1.2134055754609805,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004858237840968127,
      "loss": 5.4838,
      "mean_token_accuracy": 0.17752571105957032,
      "num_tokens": 28129028.0,
      "step": 13885
    },
    {
      "entropy": 5.851723909378052,
      "epoch": 1.2138425238136852,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048581289613697654,
      "loss": 5.4087,
      "mean_token_accuracy": 0.1870487689971924,
      "num_tokens": 28139188.0,
      "step": 13890
    },
    {
      "entropy": 5.882381391525269,
      "epoch": 1.2142794721663899,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048580200413363866,
      "loss": 5.3512,
      "mean_token_accuracy": 0.19721489399671555,
      "num_tokens": 28148199.0,
      "step": 13895
    },
    {
      "entropy": 5.8507630825042725,
      "epoch": 1.2147164205190946,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004857911080870079,
      "loss": 5.3312,
      "mean_token_accuracy": 0.19044904559850692,
      "num_tokens": 28157811.0,
      "step": 13900
    },
    {
      "entropy": 5.80710768699646,
      "epoch": 1.2151533688717993,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048578020799729327,
      "loss": 5.3563,
      "mean_token_accuracy": 0.19959521144628525,
      "num_tokens": 28168644.0,
      "step": 13905
    },
    {
      "entropy": 5.851896715164185,
      "epoch": 1.2155903172245042,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004857693038647037,
      "loss": 5.4431,
      "mean_token_accuracy": 0.18612070083618165,
      "num_tokens": 28179846.0,
      "step": 13910
    },
    {
      "entropy": 5.908466529846192,
      "epoch": 1.2160272655772089,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004857583956894485,
      "loss": 5.3755,
      "mean_token_accuracy": 0.19613765627145768,
      "num_tokens": 28189400.0,
      "step": 13915
    },
    {
      "entropy": 5.9199409008026125,
      "epoch": 1.2164642139299136,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004857474834717368,
      "loss": 5.459,
      "mean_token_accuracy": 0.18525611907243728,
      "num_tokens": 28199695.0,
      "step": 13920
    },
    {
      "entropy": 5.846132278442383,
      "epoch": 1.2169011622826182,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048573656721177805,
      "loss": 5.3578,
      "mean_token_accuracy": 0.19568415284156798,
      "num_tokens": 28209977.0,
      "step": 13925
    },
    {
      "entropy": 5.949631261825561,
      "epoch": 1.217338110635323,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004857256469097814,
      "loss": 5.4276,
      "mean_token_accuracy": 0.1857695609331131,
      "num_tokens": 28220619.0,
      "step": 13930
    },
    {
      "entropy": 5.910973167419433,
      "epoch": 1.2177750589880276,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004857147225659564,
      "loss": 5.3959,
      "mean_token_accuracy": 0.19005665928125381,
      "num_tokens": 28230065.0,
      "step": 13935
    },
    {
      "entropy": 5.883505964279175,
      "epoch": 1.2182120073407323,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048570379418051273,
      "loss": 5.3739,
      "mean_token_accuracy": 0.19453093260526658,
      "num_tokens": 28239307.0,
      "step": 13940
    },
    {
      "entropy": 5.927099180221558,
      "epoch": 1.218648955693437,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004856928617536598,
      "loss": 5.3646,
      "mean_token_accuracy": 0.1852310746908188,
      "num_tokens": 28249421.0,
      "step": 13945
    },
    {
      "entropy": 5.86376543045044,
      "epoch": 1.2190859040461417,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004856819252856073,
      "loss": 5.363,
      "mean_token_accuracy": 0.1815093994140625,
      "num_tokens": 28258125.0,
      "step": 13950
    },
    {
      "entropy": 5.862662649154663,
      "epoch": 1.2195228523988464,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048567098477656526,
      "loss": 5.3488,
      "mean_token_accuracy": 0.19125979542732238,
      "num_tokens": 28268897.0,
      "step": 13955
    },
    {
      "entropy": 5.883247327804566,
      "epoch": 1.219959800751551,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004856600402267432,
      "loss": 5.3844,
      "mean_token_accuracy": 0.19681299179792405,
      "num_tokens": 28278506.0,
      "step": 13960
    },
    {
      "entropy": 5.897805786132812,
      "epoch": 1.220396749104256,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004856490916363513,
      "loss": 5.3393,
      "mean_token_accuracy": 0.19090089648962022,
      "num_tokens": 28287905.0,
      "step": 13965
    },
    {
      "entropy": 5.936423015594483,
      "epoch": 1.2208336974569605,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004856381390055994,
      "loss": 5.4639,
      "mean_token_accuracy": 0.18472230881452562,
      "num_tokens": 28297123.0,
      "step": 13970
    },
    {
      "entropy": 5.935361099243164,
      "epoch": 1.2212706458096654,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048562718233469756,
      "loss": 5.4872,
      "mean_token_accuracy": 0.17922871708869934,
      "num_tokens": 28306621.0,
      "step": 13975
    },
    {
      "entropy": 5.943904781341553,
      "epoch": 1.22170759416237,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048561622162385606,
      "loss": 5.5233,
      "mean_token_accuracy": 0.17819749414920807,
      "num_tokens": 28318369.0,
      "step": 13980
    },
    {
      "entropy": 5.933570384979248,
      "epoch": 1.2221445425150748,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048560525687328497,
      "loss": 5.46,
      "mean_token_accuracy": 0.179440501332283,
      "num_tokens": 28328817.0,
      "step": 13985
    },
    {
      "entropy": 5.961049222946167,
      "epoch": 1.2225814908677795,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048559428808319483,
      "loss": 5.4983,
      "mean_token_accuracy": 0.17671767622232437,
      "num_tokens": 28339448.0,
      "step": 13990
    },
    {
      "entropy": 5.861384534835816,
      "epoch": 1.2230184392204841,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004855833152537958,
      "loss": 5.3607,
      "mean_token_accuracy": 0.19180475324392318,
      "num_tokens": 28349243.0,
      "step": 13995
    },
    {
      "entropy": 5.882990026473999,
      "epoch": 1.2234553875731888,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004855723383852985,
      "loss": 5.4124,
      "mean_token_accuracy": 0.19370826929807664,
      "num_tokens": 28359392.0,
      "step": 14000
    },
    {
      "entropy": 5.956579256057739,
      "epoch": 1.2238923359258935,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004855613574779134,
      "loss": 5.3779,
      "mean_token_accuracy": 0.18940700590610504,
      "num_tokens": 28369160.0,
      "step": 14005
    },
    {
      "entropy": 5.894655704498291,
      "epoch": 1.2243292842785982,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048555037253185114,
      "loss": 5.4467,
      "mean_token_accuracy": 0.1892506405711174,
      "num_tokens": 28379107.0,
      "step": 14010
    },
    {
      "entropy": 5.927813625335693,
      "epoch": 1.224766232631303,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048553938354732243,
      "loss": 5.4123,
      "mean_token_accuracy": 0.1894097864627838,
      "num_tokens": 28388956.0,
      "step": 14015
    },
    {
      "entropy": 5.878145933151245,
      "epoch": 1.2252031809840076,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048552839052453794,
      "loss": 5.3889,
      "mean_token_accuracy": 0.18979301899671555,
      "num_tokens": 28398925.0,
      "step": 14020
    },
    {
      "entropy": 5.972030544281006,
      "epoch": 1.2256401293367123,
      "grad_norm": 1.0,
      "learning_rate": 0.00048551739346370864,
      "loss": 5.4943,
      "mean_token_accuracy": 0.1777285009622574,
      "num_tokens": 28408714.0,
      "step": 14025
    },
    {
      "entropy": 6.012389326095581,
      "epoch": 1.2260770776894172,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004855063923650455,
      "loss": 5.4667,
      "mean_token_accuracy": 0.18359135687351227,
      "num_tokens": 28419408.0,
      "step": 14030
    },
    {
      "entropy": 5.8626504898071286,
      "epoch": 1.226514026042122,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004854953872287593,
      "loss": 5.4037,
      "mean_token_accuracy": 0.18574604541063308,
      "num_tokens": 28430044.0,
      "step": 14035
    },
    {
      "entropy": 5.896631145477295,
      "epoch": 1.2269509743948266,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004854843780550614,
      "loss": 5.3532,
      "mean_token_accuracy": 0.19011349231004715,
      "num_tokens": 28440100.0,
      "step": 14040
    },
    {
      "entropy": 5.904385137557983,
      "epoch": 1.2273879227475313,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004854733648441628,
      "loss": 5.3496,
      "mean_token_accuracy": 0.19333548843860626,
      "num_tokens": 28449497.0,
      "step": 14045
    },
    {
      "entropy": 5.897369861602783,
      "epoch": 1.227824871100236,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048546234759627477,
      "loss": 5.4249,
      "mean_token_accuracy": 0.1942518472671509,
      "num_tokens": 28459789.0,
      "step": 14050
    },
    {
      "entropy": 5.905897188186645,
      "epoch": 1.2282618194529407,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048545132631160856,
      "loss": 5.4805,
      "mean_token_accuracy": 0.1796879529953003,
      "num_tokens": 28471728.0,
      "step": 14055
    },
    {
      "entropy": 5.89410810470581,
      "epoch": 1.2286987678056454,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048544030099037576,
      "loss": 5.4424,
      "mean_token_accuracy": 0.1886451229453087,
      "num_tokens": 28481876.0,
      "step": 14060
    },
    {
      "entropy": 5.920392847061157,
      "epoch": 1.22913571615835,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048542927163278766,
      "loss": 5.4039,
      "mean_token_accuracy": 0.19030335545539856,
      "num_tokens": 28491800.0,
      "step": 14065
    },
    {
      "entropy": 5.874304580688476,
      "epoch": 1.2295726645110547,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004854182382390559,
      "loss": 5.373,
      "mean_token_accuracy": 0.19378072172403335,
      "num_tokens": 28502033.0,
      "step": 14070
    },
    {
      "entropy": 5.9291047096252445,
      "epoch": 1.2300096128637594,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000485407200809392,
      "loss": 5.4369,
      "mean_token_accuracy": 0.18223156332969664,
      "num_tokens": 28511921.0,
      "step": 14075
    },
    {
      "entropy": 5.961735343933105,
      "epoch": 1.2304465612164641,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004853961593440078,
      "loss": 5.3728,
      "mean_token_accuracy": 0.19469387382268905,
      "num_tokens": 28521449.0,
      "step": 14080
    },
    {
      "entropy": 5.85452709197998,
      "epoch": 1.230883509569169,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000485385113843115,
      "loss": 5.365,
      "mean_token_accuracy": 0.19389043003320694,
      "num_tokens": 28532088.0,
      "step": 14085
    },
    {
      "entropy": 5.912873601913452,
      "epoch": 1.2313204579218737,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004853740643069255,
      "loss": 5.4334,
      "mean_token_accuracy": 0.19485167413949966,
      "num_tokens": 28541835.0,
      "step": 14090
    },
    {
      "entropy": 5.938327169418335,
      "epoch": 1.2317574062745784,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048536301073565113,
      "loss": 5.4116,
      "mean_token_accuracy": 0.18792740553617476,
      "num_tokens": 28551601.0,
      "step": 14095
    },
    {
      "entropy": 5.889505863189697,
      "epoch": 1.232194354627283,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004853519531295041,
      "loss": 5.3715,
      "mean_token_accuracy": 0.19640056192874908,
      "num_tokens": 28561012.0,
      "step": 14100
    },
    {
      "entropy": 5.900825643539429,
      "epoch": 1.2326313029799878,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004853408914886963,
      "loss": 5.4675,
      "mean_token_accuracy": 0.18700266927480697,
      "num_tokens": 28571431.0,
      "step": 14105
    },
    {
      "entropy": 5.8342852115631105,
      "epoch": 1.2330682513326925,
      "grad_norm": 1.0,
      "learning_rate": 0.00048532982581344004,
      "loss": 5.3516,
      "mean_token_accuracy": 0.19761367589235307,
      "num_tokens": 28581185.0,
      "step": 14110
    },
    {
      "entropy": 5.873185634613037,
      "epoch": 1.2335051996853972,
      "grad_norm": 1.125,
      "learning_rate": 0.0004853187561039475,
      "loss": 5.3877,
      "mean_token_accuracy": 0.19380698204040528,
      "num_tokens": 28591164.0,
      "step": 14115
    },
    {
      "entropy": 5.844310665130616,
      "epoch": 1.2339421480381019,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048530768236043094,
      "loss": 5.319,
      "mean_token_accuracy": 0.2011661559343338,
      "num_tokens": 28600931.0,
      "step": 14120
    },
    {
      "entropy": 5.881939363479614,
      "epoch": 1.2343790963908066,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004852966045831029,
      "loss": 5.4408,
      "mean_token_accuracy": 0.19059082716703415,
      "num_tokens": 28611908.0,
      "step": 14125
    },
    {
      "entropy": 5.932625961303711,
      "epoch": 1.2348160447435113,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004852855227721758,
      "loss": 5.4668,
      "mean_token_accuracy": 0.18799239993095399,
      "num_tokens": 28622106.0,
      "step": 14130
    },
    {
      "entropy": 5.956688594818115,
      "epoch": 1.235252993096216,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048527443692786217,
      "loss": 5.4753,
      "mean_token_accuracy": 0.1807318851351738,
      "num_tokens": 28632634.0,
      "step": 14135
    },
    {
      "entropy": 5.921898317337036,
      "epoch": 1.2356899414489209,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004852633470503746,
      "loss": 5.4229,
      "mean_token_accuracy": 0.18444126546382905,
      "num_tokens": 28642936.0,
      "step": 14140
    },
    {
      "entropy": 5.931942129135132,
      "epoch": 1.2361268898016253,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048525225313992594,
      "loss": 5.3937,
      "mean_token_accuracy": 0.1861627370119095,
      "num_tokens": 28653539.0,
      "step": 14145
    },
    {
      "entropy": 5.8976305484771725,
      "epoch": 1.2365638381543302,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004852411551967289,
      "loss": 5.4399,
      "mean_token_accuracy": 0.18275326043367385,
      "num_tokens": 28664212.0,
      "step": 14150
    },
    {
      "entropy": 5.966551637649536,
      "epoch": 1.237000786507035,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048523005322099626,
      "loss": 5.4032,
      "mean_token_accuracy": 0.1928525149822235,
      "num_tokens": 28676028.0,
      "step": 14155
    },
    {
      "entropy": 5.972525596618652,
      "epoch": 1.2374377348597396,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048521894721294115,
      "loss": 5.368,
      "mean_token_accuracy": 0.1926775485277176,
      "num_tokens": 28685370.0,
      "step": 14160
    },
    {
      "entropy": 5.8508946895599365,
      "epoch": 1.2378746832124443,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048520783717277645,
      "loss": 5.3554,
      "mean_token_accuracy": 0.19577284306287765,
      "num_tokens": 28695300.0,
      "step": 14165
    },
    {
      "entropy": 5.8799950122833256,
      "epoch": 1.238311631565149,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048519672310071523,
      "loss": 5.3901,
      "mean_token_accuracy": 0.19527616947889329,
      "num_tokens": 28705485.0,
      "step": 14170
    },
    {
      "entropy": 5.9531818389892575,
      "epoch": 1.2387485799178537,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048518560499697066,
      "loss": 5.4253,
      "mean_token_accuracy": 0.18646642565727234,
      "num_tokens": 28715407.0,
      "step": 14175
    },
    {
      "entropy": 5.9337883472442625,
      "epoch": 1.2391855282705584,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048517448286175617,
      "loss": 5.4538,
      "mean_token_accuracy": 0.185136778652668,
      "num_tokens": 28725283.0,
      "step": 14180
    },
    {
      "entropy": 5.863058471679688,
      "epoch": 1.239622476623263,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048516335669528495,
      "loss": 5.3518,
      "mean_token_accuracy": 0.19374218881130217,
      "num_tokens": 28736131.0,
      "step": 14185
    },
    {
      "entropy": 5.892153930664063,
      "epoch": 1.2400594249759678,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004851522264977704,
      "loss": 5.3672,
      "mean_token_accuracy": 0.19060200601816177,
      "num_tokens": 28745446.0,
      "step": 14190
    },
    {
      "entropy": 5.87888994216919,
      "epoch": 1.2404963733286725,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000485141092269426,
      "loss": 5.326,
      "mean_token_accuracy": 0.19347751140594482,
      "num_tokens": 28754639.0,
      "step": 14195
    },
    {
      "entropy": 5.9117982387542725,
      "epoch": 1.2409333216813772,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004851299540104654,
      "loss": 5.4039,
      "mean_token_accuracy": 0.18946437090635299,
      "num_tokens": 28764397.0,
      "step": 14200
    },
    {
      "entropy": 5.886824083328247,
      "epoch": 1.241370270034082,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048511881172110216,
      "loss": 5.3776,
      "mean_token_accuracy": 0.1943126231431961,
      "num_tokens": 28773680.0,
      "step": 14205
    },
    {
      "entropy": 5.937914180755615,
      "epoch": 1.2418072183867868,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048510766540154997,
      "loss": 5.4537,
      "mean_token_accuracy": 0.19156129360198976,
      "num_tokens": 28784034.0,
      "step": 14210
    },
    {
      "entropy": 5.889170217514038,
      "epoch": 1.2422441667394915,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048509651505202275,
      "loss": 5.4477,
      "mean_token_accuracy": 0.18654171973466874,
      "num_tokens": 28794487.0,
      "step": 14215
    },
    {
      "entropy": 5.967399740219117,
      "epoch": 1.2426811150921961,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004850853606727342,
      "loss": 5.3497,
      "mean_token_accuracy": 0.1961347535252571,
      "num_tokens": 28803942.0,
      "step": 14220
    },
    {
      "entropy": 5.930499029159546,
      "epoch": 1.2431180634449008,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004850742022638984,
      "loss": 5.3787,
      "mean_token_accuracy": 0.19075165390968324,
      "num_tokens": 28814498.0,
      "step": 14225
    },
    {
      "entropy": 5.85469160079956,
      "epoch": 1.2435550117976055,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048506303982572924,
      "loss": 5.3706,
      "mean_token_accuracy": 0.1892199009656906,
      "num_tokens": 28824513.0,
      "step": 14230
    },
    {
      "entropy": 5.915068292617798,
      "epoch": 1.2439919601503102,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000485051873358441,
      "loss": 5.5006,
      "mean_token_accuracy": 0.17539184242486955,
      "num_tokens": 28835643.0,
      "step": 14235
    },
    {
      "entropy": 6.015800857543946,
      "epoch": 1.244428908503015,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004850407028622477,
      "loss": 5.5034,
      "mean_token_accuracy": 0.18126879185438155,
      "num_tokens": 28845615.0,
      "step": 14240
    },
    {
      "entropy": 5.901418972015381,
      "epoch": 1.2448658568557196,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048502952833736365,
      "loss": 5.3788,
      "mean_token_accuracy": 0.19435674399137498,
      "num_tokens": 28856152.0,
      "step": 14245
    },
    {
      "entropy": 5.842335557937622,
      "epoch": 1.2453028052084243,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048501834978400323,
      "loss": 5.3313,
      "mean_token_accuracy": 0.1958809182047844,
      "num_tokens": 28866926.0,
      "step": 14250
    },
    {
      "entropy": 5.9133342742919925,
      "epoch": 1.245739753561129,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004850071672023809,
      "loss": 5.4179,
      "mean_token_accuracy": 0.18564320653676986,
      "num_tokens": 28877188.0,
      "step": 14255
    },
    {
      "entropy": 5.911687660217285,
      "epoch": 1.246176701913834,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048499598059271087,
      "loss": 5.4017,
      "mean_token_accuracy": 0.18861016035079955,
      "num_tokens": 28887710.0,
      "step": 14260
    },
    {
      "entropy": 5.945541429519653,
      "epoch": 1.2466136502665386,
      "grad_norm": 1.015625,
      "learning_rate": 0.000484984789955208,
      "loss": 5.4318,
      "mean_token_accuracy": 0.18691859841346742,
      "num_tokens": 28897064.0,
      "step": 14265
    },
    {
      "entropy": 5.914587354660034,
      "epoch": 1.2470505986192433,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048497359529008684,
      "loss": 5.4248,
      "mean_token_accuracy": 0.18552234172821044,
      "num_tokens": 28907103.0,
      "step": 14270
    },
    {
      "entropy": 5.93072509765625,
      "epoch": 1.247487546971948,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004849623965975621,
      "loss": 5.4178,
      "mean_token_accuracy": 0.18577226549386977,
      "num_tokens": 28916851.0,
      "step": 14275
    },
    {
      "entropy": 5.946148538589478,
      "epoch": 1.2479244953246527,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004849511938778485,
      "loss": 5.4263,
      "mean_token_accuracy": 0.1852695882320404,
      "num_tokens": 28926638.0,
      "step": 14280
    },
    {
      "entropy": 5.920434665679932,
      "epoch": 1.2483614436773574,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048493998713116103,
      "loss": 5.3994,
      "mean_token_accuracy": 0.1885548636317253,
      "num_tokens": 28936764.0,
      "step": 14285
    },
    {
      "entropy": 5.867129373550415,
      "epoch": 1.248798392030062,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048492877635771464,
      "loss": 5.4036,
      "mean_token_accuracy": 0.18681362569332122,
      "num_tokens": 28946494.0,
      "step": 14290
    },
    {
      "entropy": 5.883354663848877,
      "epoch": 1.2492353403827667,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048491756155772433,
      "loss": 5.3763,
      "mean_token_accuracy": 0.19864395260810852,
      "num_tokens": 28955626.0,
      "step": 14295
    },
    {
      "entropy": 5.947961235046387,
      "epoch": 1.2496722887354714,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048490634273140517,
      "loss": 5.3947,
      "mean_token_accuracy": 0.18782537430524826,
      "num_tokens": 28966317.0,
      "step": 14300
    },
    {
      "entropy": 5.877606344223023,
      "epoch": 1.2501092370881761,
      "grad_norm": 1.125,
      "learning_rate": 0.00048489511987897236,
      "loss": 5.4161,
      "mean_token_accuracy": 0.19131755232810974,
      "num_tokens": 28976063.0,
      "step": 14305
    },
    {
      "entropy": 5.915600061416626,
      "epoch": 1.2505461854408808,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004848838930006411,
      "loss": 5.3873,
      "mean_token_accuracy": 0.19216327071189881,
      "num_tokens": 28985789.0,
      "step": 14310
    },
    {
      "entropy": 5.935839414596558,
      "epoch": 1.2509831337935857,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004848726620966269,
      "loss": 5.475,
      "mean_token_accuracy": 0.1871756449341774,
      "num_tokens": 28996458.0,
      "step": 14315
    },
    {
      "entropy": 5.934056520462036,
      "epoch": 1.2514200821462902,
      "grad_norm": 0.96875,
      "learning_rate": 0.000484861427167145,
      "loss": 5.3913,
      "mean_token_accuracy": 0.1895337864756584,
      "num_tokens": 29007641.0,
      "step": 14320
    },
    {
      "entropy": 5.996636962890625,
      "epoch": 1.251857030498995,
      "grad_norm": 1.0,
      "learning_rate": 0.00048485018821241104,
      "loss": 5.4139,
      "mean_token_accuracy": 0.184527288377285,
      "num_tokens": 29017351.0,
      "step": 14325
    },
    {
      "entropy": 5.934348440170288,
      "epoch": 1.2522939788516998,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048483894523264045,
      "loss": 5.4403,
      "mean_token_accuracy": 0.18092443197965621,
      "num_tokens": 29026951.0,
      "step": 14330
    },
    {
      "entropy": 5.858847141265869,
      "epoch": 1.2527309272044045,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004848276982280489,
      "loss": 5.3948,
      "mean_token_accuracy": 0.19367213398218155,
      "num_tokens": 29037596.0,
      "step": 14335
    },
    {
      "entropy": 5.881970024108886,
      "epoch": 1.2531678755571092,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004848164471988522,
      "loss": 5.4034,
      "mean_token_accuracy": 0.18898246735334395,
      "num_tokens": 29049010.0,
      "step": 14340
    },
    {
      "entropy": 5.917917060852051,
      "epoch": 1.2536048239098139,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048480519214526616,
      "loss": 5.44,
      "mean_token_accuracy": 0.1865866556763649,
      "num_tokens": 29059028.0,
      "step": 14345
    },
    {
      "entropy": 5.910173034667968,
      "epoch": 1.2540417722625186,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048479393306750647,
      "loss": 5.3978,
      "mean_token_accuracy": 0.18645863980054855,
      "num_tokens": 29069661.0,
      "step": 14350
    },
    {
      "entropy": 5.921068811416626,
      "epoch": 1.2544787206152233,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004847826699657893,
      "loss": 5.3803,
      "mean_token_accuracy": 0.1934783563017845,
      "num_tokens": 29079971.0,
      "step": 14355
    },
    {
      "entropy": 5.904225540161133,
      "epoch": 1.254915668967928,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048477140284033054,
      "loss": 5.4501,
      "mean_token_accuracy": 0.18418473452329637,
      "num_tokens": 29091378.0,
      "step": 14360
    },
    {
      "entropy": 5.881290626525879,
      "epoch": 1.2553526173206326,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004847601316913464,
      "loss": 5.3869,
      "mean_token_accuracy": 0.19446357041597367,
      "num_tokens": 29101516.0,
      "step": 14365
    },
    {
      "entropy": 5.9012552261352536,
      "epoch": 1.2557895656733375,
      "grad_norm": 0.953125,
      "learning_rate": 0.000484748856519053,
      "loss": 5.4624,
      "mean_token_accuracy": 0.18427134603261947,
      "num_tokens": 29112147.0,
      "step": 14370
    },
    {
      "entropy": 5.915145826339722,
      "epoch": 1.256226514026042,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004847375773236667,
      "loss": 5.3606,
      "mean_token_accuracy": 0.19371346831321717,
      "num_tokens": 29122084.0,
      "step": 14375
    },
    {
      "entropy": 5.948915910720825,
      "epoch": 1.256663462378747,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004847262941054037,
      "loss": 5.4038,
      "mean_token_accuracy": 0.18872934430837632,
      "num_tokens": 29131988.0,
      "step": 14380
    },
    {
      "entropy": 5.920507764816284,
      "epoch": 1.2571004107314516,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004847150068644805,
      "loss": 5.4767,
      "mean_token_accuracy": 0.18241386860609055,
      "num_tokens": 29141731.0,
      "step": 14385
    },
    {
      "entropy": 5.93372278213501,
      "epoch": 1.2575373590841563,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048470371560111353,
      "loss": 5.4778,
      "mean_token_accuracy": 0.18316574543714523,
      "num_tokens": 29152185.0,
      "step": 14390
    },
    {
      "entropy": 5.928674650192261,
      "epoch": 1.257974307436861,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004846924203155194,
      "loss": 5.3216,
      "mean_token_accuracy": 0.195791994035244,
      "num_tokens": 29162337.0,
      "step": 14395
    },
    {
      "entropy": 5.899833297729492,
      "epoch": 1.2584112557895657,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048468112100791483,
      "loss": 5.46,
      "mean_token_accuracy": 0.18569039702415466,
      "num_tokens": 29173246.0,
      "step": 14400
    },
    {
      "entropy": 5.936389970779419,
      "epoch": 1.2588482041422704,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004846698176785165,
      "loss": 5.4018,
      "mean_token_accuracy": 0.19019022583961487,
      "num_tokens": 29183700.0,
      "step": 14405
    },
    {
      "entropy": 5.951324081420898,
      "epoch": 1.259285152494975,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004846585103275412,
      "loss": 5.4271,
      "mean_token_accuracy": 0.182702699303627,
      "num_tokens": 29193686.0,
      "step": 14410
    },
    {
      "entropy": 5.861006832122802,
      "epoch": 1.2597221008476798,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048464719895520577,
      "loss": 5.4382,
      "mean_token_accuracy": 0.1845213770866394,
      "num_tokens": 29203575.0,
      "step": 14415
    },
    {
      "entropy": 5.910084009170532,
      "epoch": 1.2601590492003845,
      "grad_norm": 1.0,
      "learning_rate": 0.0004846358835617273,
      "loss": 5.403,
      "mean_token_accuracy": 0.18603634536266328,
      "num_tokens": 29213591.0,
      "step": 14420
    },
    {
      "entropy": 5.91565089225769,
      "epoch": 1.2605959975530892,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048462456414732257,
      "loss": 5.3628,
      "mean_token_accuracy": 0.19475086629390717,
      "num_tokens": 29222799.0,
      "step": 14425
    },
    {
      "entropy": 5.913440942764282,
      "epoch": 1.2610329459057938,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048461324071220896,
      "loss": 5.4057,
      "mean_token_accuracy": 0.18695778101682664,
      "num_tokens": 29232477.0,
      "step": 14430
    },
    {
      "entropy": 5.920862627029419,
      "epoch": 1.2614698942584988,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048460191325660356,
      "loss": 5.3927,
      "mean_token_accuracy": 0.1855441451072693,
      "num_tokens": 29243477.0,
      "step": 14435
    },
    {
      "entropy": 5.920644044876099,
      "epoch": 1.2619068426112032,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048459058178072366,
      "loss": 5.4731,
      "mean_token_accuracy": 0.178337961435318,
      "num_tokens": 29253426.0,
      "step": 14440
    },
    {
      "entropy": 5.908987283706665,
      "epoch": 1.2623437909639081,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004845792462847866,
      "loss": 5.4242,
      "mean_token_accuracy": 0.1899403840303421,
      "num_tokens": 29263639.0,
      "step": 14445
    },
    {
      "entropy": 5.923262977600098,
      "epoch": 1.2627807393166128,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048456790676900963,
      "loss": 5.446,
      "mean_token_accuracy": 0.18739333599805832,
      "num_tokens": 29273980.0,
      "step": 14450
    },
    {
      "entropy": 5.888284635543823,
      "epoch": 1.2632176876693175,
      "grad_norm": 1.0,
      "learning_rate": 0.00048455656323361055,
      "loss": 5.3524,
      "mean_token_accuracy": 0.19866549372673034,
      "num_tokens": 29283983.0,
      "step": 14455
    },
    {
      "entropy": 5.9734782695770265,
      "epoch": 1.2636546360220222,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004845452156788067,
      "loss": 5.4681,
      "mean_token_accuracy": 0.18402894586324692,
      "num_tokens": 29294460.0,
      "step": 14460
    },
    {
      "entropy": 5.8873453617095945,
      "epoch": 1.264091584374727,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004845338641048158,
      "loss": 5.4596,
      "mean_token_accuracy": 0.18593526631593704,
      "num_tokens": 29305788.0,
      "step": 14465
    },
    {
      "entropy": 5.9676384925842285,
      "epoch": 1.2645285327274316,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048452250851185565,
      "loss": 5.4272,
      "mean_token_accuracy": 0.19221847653388976,
      "num_tokens": 29315457.0,
      "step": 14470
    },
    {
      "entropy": 5.941789674758911,
      "epoch": 1.2649654810801363,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000484511148900144,
      "loss": 5.371,
      "mean_token_accuracy": 0.18988537937402725,
      "num_tokens": 29325713.0,
      "step": 14475
    },
    {
      "entropy": 5.854087924957275,
      "epoch": 1.265402429432841,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004844997852698987,
      "loss": 5.3929,
      "mean_token_accuracy": 0.18959095627069472,
      "num_tokens": 29336489.0,
      "step": 14480
    },
    {
      "entropy": 5.890285539627075,
      "epoch": 1.2658393777855457,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004844884176213377,
      "loss": 5.3912,
      "mean_token_accuracy": 0.1929282456636429,
      "num_tokens": 29345567.0,
      "step": 14485
    },
    {
      "entropy": 5.951479816436768,
      "epoch": 1.2662763261382506,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048447704595467917,
      "loss": 5.3996,
      "mean_token_accuracy": 0.18794978708028792,
      "num_tokens": 29355890.0,
      "step": 14490
    },
    {
      "entropy": 5.945492076873779,
      "epoch": 1.266713274490955,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048446567027014104,
      "loss": 5.4219,
      "mean_token_accuracy": 0.19309927821159362,
      "num_tokens": 29365948.0,
      "step": 14495
    },
    {
      "entropy": 5.843845701217651,
      "epoch": 1.26715022284366,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004844542905679417,
      "loss": 5.378,
      "mean_token_accuracy": 0.19008505046367646,
      "num_tokens": 29375673.0,
      "step": 14500
    },
    {
      "entropy": 5.916299772262573,
      "epoch": 1.2675871711963647,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004844429068482993,
      "loss": 5.3881,
      "mean_token_accuracy": 0.1841655269265175,
      "num_tokens": 29384567.0,
      "step": 14505
    },
    {
      "entropy": 5.903764438629151,
      "epoch": 1.2680241195490694,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048443151911143213,
      "loss": 5.2998,
      "mean_token_accuracy": 0.19877238869667052,
      "num_tokens": 29393747.0,
      "step": 14510
    },
    {
      "entropy": 5.923214340209961,
      "epoch": 1.268461067901774,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004844201273575587,
      "loss": 5.3931,
      "mean_token_accuracy": 0.1868215724825859,
      "num_tokens": 29404095.0,
      "step": 14515
    },
    {
      "entropy": 5.849318122863769,
      "epoch": 1.2688980162544787,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048440873158689754,
      "loss": 5.3925,
      "mean_token_accuracy": 0.18995471745729448,
      "num_tokens": 29414450.0,
      "step": 14520
    },
    {
      "entropy": 5.87549147605896,
      "epoch": 1.2693349646071834,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004843973317996672,
      "loss": 5.3698,
      "mean_token_accuracy": 0.19018177688121796,
      "num_tokens": 29424054.0,
      "step": 14525
    },
    {
      "entropy": 5.93612847328186,
      "epoch": 1.2697719129598881,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004843859279960862,
      "loss": 5.4528,
      "mean_token_accuracy": 0.19034417867660522,
      "num_tokens": 29434585.0,
      "step": 14530
    },
    {
      "entropy": 5.91095380783081,
      "epoch": 1.2702088613125928,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004843745201763735,
      "loss": 5.4022,
      "mean_token_accuracy": 0.19487695693969725,
      "num_tokens": 29443863.0,
      "step": 14535
    },
    {
      "entropy": 5.903210783004761,
      "epoch": 1.2706458096652975,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048436310834074764,
      "loss": 5.3745,
      "mean_token_accuracy": 0.1903134986758232,
      "num_tokens": 29453851.0,
      "step": 14540
    },
    {
      "entropy": 5.868008232116699,
      "epoch": 1.2710827580180022,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004843516924894278,
      "loss": 5.3291,
      "mean_token_accuracy": 0.19449379593133925,
      "num_tokens": 29465152.0,
      "step": 14545
    },
    {
      "entropy": 5.892236852645874,
      "epoch": 1.2715197063707069,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048434027262263274,
      "loss": 5.3305,
      "mean_token_accuracy": 0.19877527505159379,
      "num_tokens": 29474749.0,
      "step": 14550
    },
    {
      "entropy": 5.905594158172607,
      "epoch": 1.2719566547234118,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004843288487405816,
      "loss": 5.3904,
      "mean_token_accuracy": 0.1906062662601471,
      "num_tokens": 29483554.0,
      "step": 14555
    },
    {
      "entropy": 5.882736921310425,
      "epoch": 1.2723936030761165,
      "grad_norm": 1.0,
      "learning_rate": 0.00048431742084349347,
      "loss": 5.3662,
      "mean_token_accuracy": 0.19256820231676103,
      "num_tokens": 29493718.0,
      "step": 14560
    },
    {
      "entropy": 5.908563327789307,
      "epoch": 1.2728305514288212,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004843059889315874,
      "loss": 5.4306,
      "mean_token_accuracy": 0.18146968334913255,
      "num_tokens": 29504407.0,
      "step": 14565
    },
    {
      "entropy": 5.960625457763672,
      "epoch": 1.2732674997815259,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004842945530050829,
      "loss": 5.4389,
      "mean_token_accuracy": 0.18569350987672806,
      "num_tokens": 29514992.0,
      "step": 14570
    },
    {
      "entropy": 5.964413738250732,
      "epoch": 1.2737044481342306,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048428311306419913,
      "loss": 5.4652,
      "mean_token_accuracy": 0.18309289067983628,
      "num_tokens": 29525109.0,
      "step": 14575
    },
    {
      "entropy": 5.900123167037964,
      "epoch": 1.2741413964869353,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048427166910915565,
      "loss": 5.4528,
      "mean_token_accuracy": 0.18809206038713455,
      "num_tokens": 29536131.0,
      "step": 14580
    },
    {
      "entropy": 5.834575843811035,
      "epoch": 1.27457834483964,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004842602211401719,
      "loss": 5.3519,
      "mean_token_accuracy": 0.1990059047937393,
      "num_tokens": 29547614.0,
      "step": 14585
    },
    {
      "entropy": 5.899036931991577,
      "epoch": 1.2750152931923446,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004842487691574675,
      "loss": 5.4119,
      "mean_token_accuracy": 0.19102126955986024,
      "num_tokens": 29557499.0,
      "step": 14590
    },
    {
      "entropy": 5.963096380233765,
      "epoch": 1.2754522415450493,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048423731316126194,
      "loss": 5.4621,
      "mean_token_accuracy": 0.18404424786567689,
      "num_tokens": 29569909.0,
      "step": 14595
    },
    {
      "entropy": 5.997011470794678,
      "epoch": 1.275889189897754,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004842258531517752,
      "loss": 5.5566,
      "mean_token_accuracy": 0.17832042127847672,
      "num_tokens": 29581261.0,
      "step": 14600
    },
    {
      "entropy": 5.928091669082642,
      "epoch": 1.2763261382504587,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004842143891292269,
      "loss": 5.3576,
      "mean_token_accuracy": 0.18998321890830994,
      "num_tokens": 29590604.0,
      "step": 14605
    },
    {
      "entropy": 5.908301496505738,
      "epoch": 1.2767630866031636,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048420292109383704,
      "loss": 5.4446,
      "mean_token_accuracy": 0.18559692949056625,
      "num_tokens": 29600129.0,
      "step": 14610
    },
    {
      "entropy": 5.894982528686524,
      "epoch": 1.277200034955868,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048419144904582545,
      "loss": 5.3209,
      "mean_token_accuracy": 0.19524644315242767,
      "num_tokens": 29609608.0,
      "step": 14615
    },
    {
      "entropy": 5.873043012619019,
      "epoch": 1.277636983308573,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048417997298541235,
      "loss": 5.3164,
      "mean_token_accuracy": 0.19929681867361068,
      "num_tokens": 29618764.0,
      "step": 14620
    },
    {
      "entropy": 5.907800722122192,
      "epoch": 1.2780739316612777,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048416849291281765,
      "loss": 5.5015,
      "mean_token_accuracy": 0.18639667332172394,
      "num_tokens": 29628703.0,
      "step": 14625
    },
    {
      "entropy": 5.902586650848389,
      "epoch": 1.2785108800139824,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048415700882826167,
      "loss": 5.3881,
      "mean_token_accuracy": 0.19106873273849487,
      "num_tokens": 29639291.0,
      "step": 14630
    },
    {
      "entropy": 5.921809720993042,
      "epoch": 1.278947828366687,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048414552073196477,
      "loss": 5.391,
      "mean_token_accuracy": 0.187796850502491,
      "num_tokens": 29649058.0,
      "step": 14635
    },
    {
      "entropy": 5.927960777282715,
      "epoch": 1.2793847767193918,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004841340286241471,
      "loss": 5.4098,
      "mean_token_accuracy": 0.18895344585180282,
      "num_tokens": 29658629.0,
      "step": 14640
    },
    {
      "entropy": 5.9130716800689695,
      "epoch": 1.2798217250720965,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004841225325050292,
      "loss": 5.4809,
      "mean_token_accuracy": 0.18124763965606688,
      "num_tokens": 29668875.0,
      "step": 14645
    },
    {
      "entropy": 5.930905342102051,
      "epoch": 1.2802586734248012,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048411103237483156,
      "loss": 5.4636,
      "mean_token_accuracy": 0.18665360361337663,
      "num_tokens": 29678427.0,
      "step": 14650
    },
    {
      "entropy": 5.87141261100769,
      "epoch": 1.2806956217775058,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048409952823377474,
      "loss": 5.3585,
      "mean_token_accuracy": 0.19337610453367232,
      "num_tokens": 29688975.0,
      "step": 14655
    },
    {
      "entropy": 5.931174659729004,
      "epoch": 1.2811325701302105,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048408802008207937,
      "loss": 5.4408,
      "mean_token_accuracy": 0.1845469132065773,
      "num_tokens": 29699550.0,
      "step": 14660
    },
    {
      "entropy": 5.930015277862549,
      "epoch": 1.2815695184829154,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048407650791996624,
      "loss": 5.3824,
      "mean_token_accuracy": 0.19021725952625274,
      "num_tokens": 29709597.0,
      "step": 14665
    },
    {
      "entropy": 5.924608993530273,
      "epoch": 1.28200646683562,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004840649917476561,
      "loss": 5.4279,
      "mean_token_accuracy": 0.18564238399267197,
      "num_tokens": 29719250.0,
      "step": 14670
    },
    {
      "entropy": 5.94475359916687,
      "epoch": 1.2824434151883248,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048405347156536995,
      "loss": 5.4541,
      "mean_token_accuracy": 0.1824955552816391,
      "num_tokens": 29728809.0,
      "step": 14675
    },
    {
      "entropy": 5.935526943206787,
      "epoch": 1.2828803635410295,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004840419473733287,
      "loss": 5.361,
      "mean_token_accuracy": 0.19212491512298585,
      "num_tokens": 29739240.0,
      "step": 14680
    },
    {
      "entropy": 5.903755521774292,
      "epoch": 1.2833173118937342,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048403041917175323,
      "loss": 5.4166,
      "mean_token_accuracy": 0.1896335855126381,
      "num_tokens": 29749284.0,
      "step": 14685
    },
    {
      "entropy": 5.926655578613281,
      "epoch": 1.283754260246439,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048401888696086485,
      "loss": 5.4525,
      "mean_token_accuracy": 0.18348005264997483,
      "num_tokens": 29758770.0,
      "step": 14690
    },
    {
      "entropy": 5.913109588623047,
      "epoch": 1.2841912085991436,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004840073507408847,
      "loss": 5.392,
      "mean_token_accuracy": 0.1908774584531784,
      "num_tokens": 29770111.0,
      "step": 14695
    },
    {
      "entropy": 5.920510149002075,
      "epoch": 1.2846281569518483,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000483995810512034,
      "loss": 5.3664,
      "mean_token_accuracy": 0.19222841560840606,
      "num_tokens": 29780210.0,
      "step": 14700
    },
    {
      "entropy": 5.925009250640869,
      "epoch": 1.285065105304553,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048398426627453423,
      "loss": 5.3844,
      "mean_token_accuracy": 0.19015737175941466,
      "num_tokens": 29789903.0,
      "step": 14705
    },
    {
      "entropy": 5.953707695007324,
      "epoch": 1.2855020536572577,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048397271802860667,
      "loss": 5.4748,
      "mean_token_accuracy": 0.18402190655469894,
      "num_tokens": 29799377.0,
      "step": 14710
    },
    {
      "entropy": 5.921931791305542,
      "epoch": 1.2859390020099624,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004839611657744729,
      "loss": 5.3569,
      "mean_token_accuracy": 0.1932017520070076,
      "num_tokens": 29809173.0,
      "step": 14715
    },
    {
      "entropy": 5.863224697113037,
      "epoch": 1.286375950362667,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048394960951235444,
      "loss": 5.4846,
      "mean_token_accuracy": 0.1858229249715805,
      "num_tokens": 29819475.0,
      "step": 14720
    },
    {
      "entropy": 5.8966460704803465,
      "epoch": 1.2868128987153717,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000483938049242473,
      "loss": 5.3389,
      "mean_token_accuracy": 0.18775721341371537,
      "num_tokens": 29829109.0,
      "step": 14725
    },
    {
      "entropy": 5.955277776718139,
      "epoch": 1.2872498470680767,
      "grad_norm": 1.0,
      "learning_rate": 0.00048392648496505024,
      "loss": 5.3829,
      "mean_token_accuracy": 0.1888130784034729,
      "num_tokens": 29839735.0,
      "step": 14730
    },
    {
      "entropy": 5.8912272453308105,
      "epoch": 1.2876867954207811,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004839149166803081,
      "loss": 5.3588,
      "mean_token_accuracy": 0.18991594016551971,
      "num_tokens": 29850039.0,
      "step": 14735
    },
    {
      "entropy": 5.931095170974731,
      "epoch": 1.288123743773486,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004839033443884684,
      "loss": 5.3754,
      "mean_token_accuracy": 0.1888427659869194,
      "num_tokens": 29859411.0,
      "step": 14740
    },
    {
      "entropy": 5.901830863952637,
      "epoch": 1.2885606921261907,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004838917680897531,
      "loss": 5.4293,
      "mean_token_accuracy": 0.18556695729494094,
      "num_tokens": 29870426.0,
      "step": 14745
    },
    {
      "entropy": 5.939640235900879,
      "epoch": 1.2889976404788954,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048388018778438413,
      "loss": 5.3844,
      "mean_token_accuracy": 0.19092077314853667,
      "num_tokens": 29881992.0,
      "step": 14750
    },
    {
      "entropy": 5.925490760803223,
      "epoch": 1.2894345888316001,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004838686034725838,
      "loss": 5.3956,
      "mean_token_accuracy": 0.19109199345111846,
      "num_tokens": 29891496.0,
      "step": 14755
    },
    {
      "entropy": 5.957470369338989,
      "epoch": 1.2898715371843048,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048385701515457413,
      "loss": 5.3912,
      "mean_token_accuracy": 0.18377087116241456,
      "num_tokens": 29901529.0,
      "step": 14760
    },
    {
      "entropy": 5.862794637680054,
      "epoch": 1.2903084855370095,
      "grad_norm": 1.125,
      "learning_rate": 0.0004838454228305775,
      "loss": 5.4505,
      "mean_token_accuracy": 0.18958282917737962,
      "num_tokens": 29910773.0,
      "step": 14765
    },
    {
      "entropy": 5.90316424369812,
      "epoch": 1.2907454338897142,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048383382650081633,
      "loss": 5.4312,
      "mean_token_accuracy": 0.18617956340312958,
      "num_tokens": 29919381.0,
      "step": 14770
    },
    {
      "entropy": 5.979472398757935,
      "epoch": 1.2911823822424189,
      "grad_norm": 1.0,
      "learning_rate": 0.0004838222261655129,
      "loss": 5.4375,
      "mean_token_accuracy": 0.18564701974391937,
      "num_tokens": 29929179.0,
      "step": 14775
    },
    {
      "entropy": 5.972487449645996,
      "epoch": 1.2916193305951236,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048381062182488964,
      "loss": 5.3667,
      "mean_token_accuracy": 0.1945200264453888,
      "num_tokens": 29938252.0,
      "step": 14780
    },
    {
      "entropy": 5.89096040725708,
      "epoch": 1.2920562789478285,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048379901347916926,
      "loss": 5.4187,
      "mean_token_accuracy": 0.18726425915956496,
      "num_tokens": 29947489.0,
      "step": 14785
    },
    {
      "entropy": 5.87551531791687,
      "epoch": 1.292493227300533,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048378740112857443,
      "loss": 5.3577,
      "mean_token_accuracy": 0.19502917528152466,
      "num_tokens": 29956788.0,
      "step": 14790
    },
    {
      "entropy": 5.9463177680969235,
      "epoch": 1.2929301756532379,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048377578477332785,
      "loss": 5.4526,
      "mean_token_accuracy": 0.18452025949954987,
      "num_tokens": 29967161.0,
      "step": 14795
    },
    {
      "entropy": 5.880333423614502,
      "epoch": 1.2933671240059426,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004837641644136523,
      "loss": 5.2794,
      "mean_token_accuracy": 0.20454811155796052,
      "num_tokens": 29976739.0,
      "step": 14800
    },
    {
      "entropy": 5.879928159713745,
      "epoch": 1.2938040723586473,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048375254004977074,
      "loss": 5.396,
      "mean_token_accuracy": 0.19256442338228225,
      "num_tokens": 29986739.0,
      "step": 14805
    },
    {
      "entropy": 5.8898272037506105,
      "epoch": 1.294241020711352,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004837409116819061,
      "loss": 5.371,
      "mean_token_accuracy": 0.19188270717859268,
      "num_tokens": 29996687.0,
      "step": 14810
    },
    {
      "entropy": 5.88683066368103,
      "epoch": 1.2946779690640566,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004837292793102813,
      "loss": 5.4103,
      "mean_token_accuracy": 0.18896131068468094,
      "num_tokens": 30007757.0,
      "step": 14815
    },
    {
      "entropy": 5.830907583236694,
      "epoch": 1.2951149174167613,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048371764293511963,
      "loss": 5.3604,
      "mean_token_accuracy": 0.19820277094841005,
      "num_tokens": 30018129.0,
      "step": 14820
    },
    {
      "entropy": 5.932209205627442,
      "epoch": 1.295551865769466,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004837060025566441,
      "loss": 5.3814,
      "mean_token_accuracy": 0.1881450593471527,
      "num_tokens": 30028194.0,
      "step": 14825
    },
    {
      "entropy": 5.936330509185791,
      "epoch": 1.2959888141221707,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004836943581750782,
      "loss": 5.5161,
      "mean_token_accuracy": 0.17465783059597015,
      "num_tokens": 30038182.0,
      "step": 14830
    },
    {
      "entropy": 5.979481601715088,
      "epoch": 1.2964257624748754,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004836827097906451,
      "loss": 5.4108,
      "mean_token_accuracy": 0.19148010164499282,
      "num_tokens": 30047848.0,
      "step": 14835
    },
    {
      "entropy": 5.932941436767578,
      "epoch": 1.2968627108275803,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048367105740356825,
      "loss": 5.4752,
      "mean_token_accuracy": 0.19045577943325043,
      "num_tokens": 30058697.0,
      "step": 14840
    },
    {
      "entropy": 5.9263989448547365,
      "epoch": 1.2972996591802848,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048365940101407123,
      "loss": 5.3125,
      "mean_token_accuracy": 0.19342112988233567,
      "num_tokens": 30068611.0,
      "step": 14845
    },
    {
      "entropy": 5.881247615814209,
      "epoch": 1.2977366075329897,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004836477406223775,
      "loss": 5.3719,
      "mean_token_accuracy": 0.1940523236989975,
      "num_tokens": 30077992.0,
      "step": 14850
    },
    {
      "entropy": 5.996506452560425,
      "epoch": 1.2981735558856944,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048363607622871086,
      "loss": 5.5431,
      "mean_token_accuracy": 0.1776387706398964,
      "num_tokens": 30088940.0,
      "step": 14855
    },
    {
      "entropy": 5.981442260742187,
      "epoch": 1.298610504238399,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048362440783329485,
      "loss": 5.4183,
      "mean_token_accuracy": 0.1882338672876358,
      "num_tokens": 30099249.0,
      "step": 14860
    },
    {
      "entropy": 5.87814507484436,
      "epoch": 1.2990474525911038,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004836127354363535,
      "loss": 5.4495,
      "mean_token_accuracy": 0.19085088074207307,
      "num_tokens": 30110548.0,
      "step": 14865
    },
    {
      "entropy": 5.904420232772827,
      "epoch": 1.2994844009438085,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048360105903811045,
      "loss": 5.3952,
      "mean_token_accuracy": 0.1964625224471092,
      "num_tokens": 30120842.0,
      "step": 14870
    },
    {
      "entropy": 5.948130130767822,
      "epoch": 1.2999213492965132,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048358937863878984,
      "loss": 5.4364,
      "mean_token_accuracy": 0.18821181803941728,
      "num_tokens": 30130439.0,
      "step": 14875
    },
    {
      "entropy": 5.948386526107788,
      "epoch": 1.3003582976492178,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004835776942386157,
      "loss": 5.4691,
      "mean_token_accuracy": 0.18767090290784835,
      "num_tokens": 30140976.0,
      "step": 14880
    },
    {
      "entropy": 5.896849155426025,
      "epoch": 1.3007952460019225,
      "grad_norm": 0.9375,
      "learning_rate": 0.000483566005837812,
      "loss": 5.4155,
      "mean_token_accuracy": 0.18939503729343415,
      "num_tokens": 30150989.0,
      "step": 14885
    },
    {
      "entropy": 5.903592252731324,
      "epoch": 1.3012321943546272,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048355431343660303,
      "loss": 5.3735,
      "mean_token_accuracy": 0.1896089255809784,
      "num_tokens": 30161605.0,
      "step": 14890
    },
    {
      "entropy": 5.913569068908691,
      "epoch": 1.301669142707332,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004835426170352131,
      "loss": 5.3828,
      "mean_token_accuracy": 0.19199938625097274,
      "num_tokens": 30171361.0,
      "step": 14895
    },
    {
      "entropy": 5.923840665817261,
      "epoch": 1.3021060910600366,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048353091663386655,
      "loss": 5.3286,
      "mean_token_accuracy": 0.19344531744718552,
      "num_tokens": 30180352.0,
      "step": 14900
    },
    {
      "entropy": 5.88422384262085,
      "epoch": 1.3025430394127415,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004835192122327877,
      "loss": 5.3425,
      "mean_token_accuracy": 0.19159792959690095,
      "num_tokens": 30190624.0,
      "step": 14905
    },
    {
      "entropy": 5.887854385375976,
      "epoch": 1.302979987765446,
      "grad_norm": 1.125,
      "learning_rate": 0.00048350750383220105,
      "loss": 5.3709,
      "mean_token_accuracy": 0.19495064169168472,
      "num_tokens": 30200203.0,
      "step": 14910
    },
    {
      "entropy": 5.868716430664063,
      "epoch": 1.303416936118151,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048349579143233126,
      "loss": 5.3572,
      "mean_token_accuracy": 0.20026013106107712,
      "num_tokens": 30210546.0,
      "step": 14915
    },
    {
      "entropy": 5.982860660552978,
      "epoch": 1.3038538844708556,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004834840750334029,
      "loss": 5.5209,
      "mean_token_accuracy": 0.1813124969601631,
      "num_tokens": 30221316.0,
      "step": 14920
    },
    {
      "entropy": 5.961442232131958,
      "epoch": 1.3042908328235603,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004834723546356408,
      "loss": 5.4507,
      "mean_token_accuracy": 0.18683552443981172,
      "num_tokens": 30232580.0,
      "step": 14925
    },
    {
      "entropy": 5.946694183349609,
      "epoch": 1.304727781176265,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048346063023926964,
      "loss": 5.3807,
      "mean_token_accuracy": 0.19309913367033005,
      "num_tokens": 30241976.0,
      "step": 14930
    },
    {
      "entropy": 5.910878229141235,
      "epoch": 1.3051647295289697,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048344890184451436,
      "loss": 5.3834,
      "mean_token_accuracy": 0.186969393491745,
      "num_tokens": 30252723.0,
      "step": 14935
    },
    {
      "entropy": 5.9180943965911865,
      "epoch": 1.3056016778816744,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048343716945159997,
      "loss": 5.4295,
      "mean_token_accuracy": 0.1871197462081909,
      "num_tokens": 30263125.0,
      "step": 14940
    },
    {
      "entropy": 5.849191570281983,
      "epoch": 1.306038626234379,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004834254330607514,
      "loss": 5.333,
      "mean_token_accuracy": 0.20228434205055237,
      "num_tokens": 30273113.0,
      "step": 14945
    },
    {
      "entropy": 5.8994098663330075,
      "epoch": 1.3064755745870837,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004834136926721939,
      "loss": 5.3253,
      "mean_token_accuracy": 0.19868248403072358,
      "num_tokens": 30282550.0,
      "step": 14950
    },
    {
      "entropy": 5.864881277084351,
      "epoch": 1.3069125229397884,
      "grad_norm": 1.0,
      "learning_rate": 0.00048340194828615246,
      "loss": 5.3454,
      "mean_token_accuracy": 0.18766660541296004,
      "num_tokens": 30293377.0,
      "step": 14955
    },
    {
      "entropy": 5.910787057876587,
      "epoch": 1.3073494712924933,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004833901999028525,
      "loss": 5.3807,
      "mean_token_accuracy": 0.1926659718155861,
      "num_tokens": 30303147.0,
      "step": 14960
    },
    {
      "entropy": 5.917950105667114,
      "epoch": 1.3077864196451978,
      "grad_norm": 1.0,
      "learning_rate": 0.0004833784475225192,
      "loss": 5.4997,
      "mean_token_accuracy": 0.18232237547636032,
      "num_tokens": 30313426.0,
      "step": 14965
    },
    {
      "entropy": 5.927287149429321,
      "epoch": 1.3082233679979027,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004833666911453783,
      "loss": 5.4229,
      "mean_token_accuracy": 0.18010252863168716,
      "num_tokens": 30323896.0,
      "step": 14970
    },
    {
      "entropy": 5.909463262557983,
      "epoch": 1.3086603163506074,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000483354930771655,
      "loss": 5.414,
      "mean_token_accuracy": 0.1837685227394104,
      "num_tokens": 30333523.0,
      "step": 14975
    },
    {
      "entropy": 5.900996303558349,
      "epoch": 1.3090972647033121,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048334316640157493,
      "loss": 5.3764,
      "mean_token_accuracy": 0.19616176337003707,
      "num_tokens": 30344035.0,
      "step": 14980
    },
    {
      "entropy": 5.925342464447022,
      "epoch": 1.3095342130560168,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004833313980353637,
      "loss": 5.4723,
      "mean_token_accuracy": 0.19031932353973388,
      "num_tokens": 30354385.0,
      "step": 14985
    },
    {
      "entropy": 5.943136215209961,
      "epoch": 1.3099711614087215,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048331962567324717,
      "loss": 5.5089,
      "mean_token_accuracy": 0.18158982694149017,
      "num_tokens": 30364941.0,
      "step": 14990
    },
    {
      "entropy": 5.8967584609985355,
      "epoch": 1.3104081097614262,
      "grad_norm": 0.984375,
      "learning_rate": 0.000483307849315451,
      "loss": 5.3528,
      "mean_token_accuracy": 0.1892830029129982,
      "num_tokens": 30376160.0,
      "step": 14995
    },
    {
      "entropy": 5.957175970077515,
      "epoch": 1.3108450581141309,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048329606896220116,
      "loss": 5.4647,
      "mean_token_accuracy": 0.1824781909584999,
      "num_tokens": 30385018.0,
      "step": 15000
    },
    {
      "epoch": 1.3108450581141309,
      "eval_entropy": 5.6811567013429105,
      "eval_loss": 5.449954509735107,
      "eval_mean_token_accuracy": 0.1964771876682054,
      "eval_num_tokens": 30385018.0,
      "eval_runtime": 21.5708,
      "eval_samples_per_second": 1715.931,
      "eval_steps_per_second": 214.503,
      "step": 15000
    },
    {
      "entropy": 5.907427549362183,
      "epoch": 1.3112820064668356,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004832842846137236,
      "loss": 5.3885,
      "mean_token_accuracy": 0.1931648001074791,
      "num_tokens": 30394618.0,
      "step": 15005
    },
    {
      "entropy": 5.907952070236206,
      "epoch": 1.3117189548195403,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048327249627024433,
      "loss": 5.403,
      "mean_token_accuracy": 0.19143993258476258,
      "num_tokens": 30404081.0,
      "step": 15010
    },
    {
      "entropy": 5.921749019622803,
      "epoch": 1.3121559031722452,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004832607039319894,
      "loss": 5.3724,
      "mean_token_accuracy": 0.18711319714784622,
      "num_tokens": 30413892.0,
      "step": 15015
    },
    {
      "entropy": 5.967968463897705,
      "epoch": 1.3125928515249496,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048324890759918513,
      "loss": 5.4603,
      "mean_token_accuracy": 0.1837619051337242,
      "num_tokens": 30424523.0,
      "step": 15020
    },
    {
      "entropy": 5.903027868270874,
      "epoch": 1.3130297998776546,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004832371072720576,
      "loss": 5.4014,
      "mean_token_accuracy": 0.1950627952814102,
      "num_tokens": 30435092.0,
      "step": 15025
    },
    {
      "entropy": 5.81185622215271,
      "epoch": 1.3134667482303592,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004832253029508332,
      "loss": 5.3413,
      "mean_token_accuracy": 0.1893490508198738,
      "num_tokens": 30445628.0,
      "step": 15030
    },
    {
      "entropy": 5.864007091522216,
      "epoch": 1.313903696583064,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004832134946357384,
      "loss": 5.4135,
      "mean_token_accuracy": 0.18919628858566284,
      "num_tokens": 30457433.0,
      "step": 15035
    },
    {
      "entropy": 5.9238440036773685,
      "epoch": 1.3143406449357686,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004832016823269998,
      "loss": 5.3926,
      "mean_token_accuracy": 0.18801021277904512,
      "num_tokens": 30467987.0,
      "step": 15040
    },
    {
      "entropy": 5.967406558990478,
      "epoch": 1.3147775932884733,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004831898660248437,
      "loss": 5.4112,
      "mean_token_accuracy": 0.18333270847797395,
      "num_tokens": 30476577.0,
      "step": 15045
    },
    {
      "entropy": 5.866452217102051,
      "epoch": 1.315214541641178,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00048317804572949697,
      "loss": 5.3494,
      "mean_token_accuracy": 0.19156383574008942,
      "num_tokens": 30487410.0,
      "step": 15050
    },
    {
      "entropy": 5.946942090988159,
      "epoch": 1.3156514899938827,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004831662214411862,
      "loss": 5.4669,
      "mean_token_accuracy": 0.1835633859038353,
      "num_tokens": 30497778.0,
      "step": 15055
    },
    {
      "entropy": 5.923778629302978,
      "epoch": 1.3160884383465874,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004831543931601382,
      "loss": 5.39,
      "mean_token_accuracy": 0.19037847518920897,
      "num_tokens": 30507369.0,
      "step": 15060
    },
    {
      "entropy": 5.922081422805786,
      "epoch": 1.316525386699292,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048314256088657985,
      "loss": 5.4256,
      "mean_token_accuracy": 0.18685904145240784,
      "num_tokens": 30518126.0,
      "step": 15065
    },
    {
      "entropy": 5.889039373397827,
      "epoch": 1.3169623350519968,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004831307246207382,
      "loss": 5.4007,
      "mean_token_accuracy": 0.19415348023176193,
      "num_tokens": 30528167.0,
      "step": 15070
    },
    {
      "entropy": 5.889727830886841,
      "epoch": 1.3173992834047015,
      "grad_norm": 1.0,
      "learning_rate": 0.00048311888436284017,
      "loss": 5.3369,
      "mean_token_accuracy": 0.1999618425965309,
      "num_tokens": 30538111.0,
      "step": 15075
    },
    {
      "entropy": 5.905206918716431,
      "epoch": 1.3178362317574064,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004831070401131129,
      "loss": 5.4359,
      "mean_token_accuracy": 0.19576275050640107,
      "num_tokens": 30547773.0,
      "step": 15080
    },
    {
      "entropy": 5.904333400726318,
      "epoch": 1.3182731801101109,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004830951918717834,
      "loss": 5.3706,
      "mean_token_accuracy": 0.19121160358190536,
      "num_tokens": 30558058.0,
      "step": 15085
    },
    {
      "entropy": 5.899756717681885,
      "epoch": 1.3187101284628158,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048308333963907927,
      "loss": 5.3357,
      "mean_token_accuracy": 0.1947065144777298,
      "num_tokens": 30567297.0,
      "step": 15090
    },
    {
      "entropy": 5.936825561523437,
      "epoch": 1.3191470768155205,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048307148341522756,
      "loss": 5.3994,
      "mean_token_accuracy": 0.1913631707429886,
      "num_tokens": 30578159.0,
      "step": 15095
    },
    {
      "entropy": 5.9228386878967285,
      "epoch": 1.3195840251682251,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048305962320045574,
      "loss": 5.4138,
      "mean_token_accuracy": 0.1887746974825859,
      "num_tokens": 30588734.0,
      "step": 15100
    },
    {
      "entropy": 5.8877880573272705,
      "epoch": 1.3200209735209298,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048304775899499143,
      "loss": 5.3214,
      "mean_token_accuracy": 0.19779499620199203,
      "num_tokens": 30598496.0,
      "step": 15105
    },
    {
      "entropy": 5.933681344985962,
      "epoch": 1.3204579218736345,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000483035890799062,
      "loss": 5.3953,
      "mean_token_accuracy": 0.1913222074508667,
      "num_tokens": 30608330.0,
      "step": 15110
    },
    {
      "entropy": 5.942589950561524,
      "epoch": 1.3208948702263392,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004830240186128952,
      "loss": 5.3963,
      "mean_token_accuracy": 0.19469435065984725,
      "num_tokens": 30618887.0,
      "step": 15115
    },
    {
      "entropy": 5.912452650070191,
      "epoch": 1.321331818579044,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048301214243671884,
      "loss": 5.3457,
      "mean_token_accuracy": 0.1928478166460991,
      "num_tokens": 30629479.0,
      "step": 15120
    },
    {
      "entropy": 5.857966327667237,
      "epoch": 1.3217687669317486,
      "grad_norm": 1.125,
      "learning_rate": 0.0004830002622707604,
      "loss": 5.3437,
      "mean_token_accuracy": 0.19111318588256837,
      "num_tokens": 30639270.0,
      "step": 15125
    },
    {
      "entropy": 5.944682359695435,
      "epoch": 1.3222057152844533,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004829883781152481,
      "loss": 5.4578,
      "mean_token_accuracy": 0.19346543699502944,
      "num_tokens": 30649156.0,
      "step": 15130
    },
    {
      "entropy": 5.907151794433593,
      "epoch": 1.3226426636371582,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048297648997040966,
      "loss": 5.4714,
      "mean_token_accuracy": 0.1834176406264305,
      "num_tokens": 30659919.0,
      "step": 15135
    },
    {
      "entropy": 5.911435556411743,
      "epoch": 1.3230796119898627,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048296459783647304,
      "loss": 5.437,
      "mean_token_accuracy": 0.18487379103899002,
      "num_tokens": 30670523.0,
      "step": 15140
    },
    {
      "entropy": 5.9640114307403564,
      "epoch": 1.3235165603425676,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048295270171366665,
      "loss": 5.4431,
      "mean_token_accuracy": 0.18818518668413162,
      "num_tokens": 30680803.0,
      "step": 15145
    },
    {
      "entropy": 5.927008724212646,
      "epoch": 1.3239535086952723,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048294080160221837,
      "loss": 5.2964,
      "mean_token_accuracy": 0.19815680384635925,
      "num_tokens": 30690401.0,
      "step": 15150
    },
    {
      "entropy": 5.966713094711304,
      "epoch": 1.324390457047977,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048292889750235665,
      "loss": 5.4516,
      "mean_token_accuracy": 0.18730060309171676,
      "num_tokens": 30700779.0,
      "step": 15155
    },
    {
      "entropy": 5.952288675308227,
      "epoch": 1.3248274054006817,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048291698941430955,
      "loss": 5.4201,
      "mean_token_accuracy": 0.18509871363639832,
      "num_tokens": 30711556.0,
      "step": 15160
    },
    {
      "entropy": 5.838730144500732,
      "epoch": 1.3252643537533864,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004829050773383058,
      "loss": 5.3089,
      "mean_token_accuracy": 0.19998255223035813,
      "num_tokens": 30721929.0,
      "step": 15165
    },
    {
      "entropy": 5.927463388442993,
      "epoch": 1.325701302106091,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004828931612745737,
      "loss": 5.4411,
      "mean_token_accuracy": 0.18473042100667952,
      "num_tokens": 30732848.0,
      "step": 15170
    },
    {
      "entropy": 5.959762859344482,
      "epoch": 1.3261382504587957,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048288124122334176,
      "loss": 5.3973,
      "mean_token_accuracy": 0.186854487657547,
      "num_tokens": 30742628.0,
      "step": 15175
    },
    {
      "entropy": 5.933885669708252,
      "epoch": 1.3265751988115004,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004828693171848388,
      "loss": 5.4406,
      "mean_token_accuracy": 0.18824684023857116,
      "num_tokens": 30752866.0,
      "step": 15180
    },
    {
      "entropy": 5.838097953796387,
      "epoch": 1.3270121471642051,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004828573891592933,
      "loss": 5.3666,
      "mean_token_accuracy": 0.19719793498516083,
      "num_tokens": 30762861.0,
      "step": 15185
    },
    {
      "entropy": 5.933246803283692,
      "epoch": 1.32744909551691,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004828454571469342,
      "loss": 5.4183,
      "mean_token_accuracy": 0.18982189297676086,
      "num_tokens": 30772555.0,
      "step": 15190
    },
    {
      "entropy": 5.980365562438965,
      "epoch": 1.3278860438696145,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004828335211479903,
      "loss": 5.4106,
      "mean_token_accuracy": 0.18651817739009857,
      "num_tokens": 30782739.0,
      "step": 15195
    },
    {
      "entropy": 5.936884212493896,
      "epoch": 1.3283229922223194,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048282158116269065,
      "loss": 5.3824,
      "mean_token_accuracy": 0.19274972826242448,
      "num_tokens": 30792600.0,
      "step": 15200
    },
    {
      "entropy": 5.943372058868408,
      "epoch": 1.3287599405750241,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004828096371912641,
      "loss": 5.3908,
      "mean_token_accuracy": 0.18674497455358505,
      "num_tokens": 30801433.0,
      "step": 15205
    },
    {
      "entropy": 5.92948546409607,
      "epoch": 1.3291968889277288,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048279768923393987,
      "loss": 5.4693,
      "mean_token_accuracy": 0.1827809453010559,
      "num_tokens": 30811277.0,
      "step": 15210
    },
    {
      "entropy": 5.941763210296631,
      "epoch": 1.3296338372804335,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048278573729094705,
      "loss": 5.4385,
      "mean_token_accuracy": 0.18668775856494904,
      "num_tokens": 30821120.0,
      "step": 15215
    },
    {
      "entropy": 5.904952621459961,
      "epoch": 1.3300707856331382,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048277378136251487,
      "loss": 5.3535,
      "mean_token_accuracy": 0.19506777673959733,
      "num_tokens": 30831660.0,
      "step": 15220
    },
    {
      "entropy": 5.971986770629883,
      "epoch": 1.3305077339858429,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048276182144887274,
      "loss": 5.4261,
      "mean_token_accuracy": 0.18057018369436265,
      "num_tokens": 30842530.0,
      "step": 15225
    },
    {
      "entropy": 5.916555404663086,
      "epoch": 1.3309446823385476,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048274985755025003,
      "loss": 5.4164,
      "mean_token_accuracy": 0.1839665174484253,
      "num_tokens": 30852954.0,
      "step": 15230
    },
    {
      "entropy": 5.860643815994263,
      "epoch": 1.3313816306912523,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048273788966687615,
      "loss": 5.3123,
      "mean_token_accuracy": 0.19132208675146103,
      "num_tokens": 30862864.0,
      "step": 15235
    },
    {
      "entropy": 5.918424844741821,
      "epoch": 1.331818579043957,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004827259177989807,
      "loss": 5.3819,
      "mean_token_accuracy": 0.18915876746177673,
      "num_tokens": 30873533.0,
      "step": 15240
    },
    {
      "entropy": 5.965174055099487,
      "epoch": 1.3322555273966616,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048271394194679336,
      "loss": 5.3287,
      "mean_token_accuracy": 0.1960371643304825,
      "num_tokens": 30883564.0,
      "step": 15245
    },
    {
      "entropy": 5.928645467758178,
      "epoch": 1.3326924757493663,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004827019621105436,
      "loss": 5.3581,
      "mean_token_accuracy": 0.19878839552402497,
      "num_tokens": 30893516.0,
      "step": 15250
    },
    {
      "entropy": 5.840205526351928,
      "epoch": 1.3331294241020712,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048268997829046156,
      "loss": 5.3733,
      "mean_token_accuracy": 0.18698578029870988,
      "num_tokens": 30903104.0,
      "step": 15255
    },
    {
      "entropy": 5.891952705383301,
      "epoch": 1.3335663724547757,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048267799048677684,
      "loss": 5.3569,
      "mean_token_accuracy": 0.19129989445209503,
      "num_tokens": 30912759.0,
      "step": 15260
    },
    {
      "entropy": 5.929360866546631,
      "epoch": 1.3340033208074806,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004826659986997194,
      "loss": 5.3957,
      "mean_token_accuracy": 0.19075891375541687,
      "num_tokens": 30921361.0,
      "step": 15265
    },
    {
      "entropy": 5.921074438095093,
      "epoch": 1.3344402691601853,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004826540029295193,
      "loss": 5.3788,
      "mean_token_accuracy": 0.18813875913619996,
      "num_tokens": 30931006.0,
      "step": 15270
    },
    {
      "entropy": 5.898694276809692,
      "epoch": 1.33487721751289,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048264200317640665,
      "loss": 5.3767,
      "mean_token_accuracy": 0.19070641100406646,
      "num_tokens": 30940647.0,
      "step": 15275
    },
    {
      "entropy": 5.934998321533203,
      "epoch": 1.3353141658655947,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004826299994406116,
      "loss": 5.4349,
      "mean_token_accuracy": 0.1851475328207016,
      "num_tokens": 30951028.0,
      "step": 15280
    },
    {
      "entropy": 5.915450096130371,
      "epoch": 1.3357511142182994,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048261799172236435,
      "loss": 5.3867,
      "mean_token_accuracy": 0.18982872664928435,
      "num_tokens": 30960726.0,
      "step": 15285
    },
    {
      "entropy": 5.937544631958008,
      "epoch": 1.336188062571004,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004826059800218952,
      "loss": 5.4659,
      "mean_token_accuracy": 0.1879633381962776,
      "num_tokens": 30971061.0,
      "step": 15290
    },
    {
      "entropy": 5.854364919662475,
      "epoch": 1.3366250109237088,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004825939643394346,
      "loss": 5.258,
      "mean_token_accuracy": 0.20702720284461976,
      "num_tokens": 30980518.0,
      "step": 15295
    },
    {
      "entropy": 5.898902511596679,
      "epoch": 1.3370619592764135,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048258194467521294,
      "loss": 5.4737,
      "mean_token_accuracy": 0.18529762625694274,
      "num_tokens": 30990759.0,
      "step": 15300
    },
    {
      "entropy": 6.000565910339356,
      "epoch": 1.3374989076291182,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048256992102946084,
      "loss": 5.4262,
      "mean_token_accuracy": 0.18242829740047456,
      "num_tokens": 31002218.0,
      "step": 15305
    },
    {
      "entropy": 5.88933482170105,
      "epoch": 1.337935855981823,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004825578934024089,
      "loss": 5.3828,
      "mean_token_accuracy": 0.19623757004737855,
      "num_tokens": 31012434.0,
      "step": 15310
    },
    {
      "entropy": 5.963655614852906,
      "epoch": 1.3383728043345275,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048254586179428776,
      "loss": 5.4859,
      "mean_token_accuracy": 0.1833385244011879,
      "num_tokens": 31022805.0,
      "step": 15315
    },
    {
      "entropy": 5.933218383789063,
      "epoch": 1.3388097526872325,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004825338262053283,
      "loss": 5.3694,
      "mean_token_accuracy": 0.19402150064706802,
      "num_tokens": 31032563.0,
      "step": 15320
    },
    {
      "entropy": 5.88496618270874,
      "epoch": 1.3392467010399371,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004825217866357613,
      "loss": 5.3599,
      "mean_token_accuracy": 0.19509214013814927,
      "num_tokens": 31041445.0,
      "step": 15325
    },
    {
      "entropy": 5.879874610900879,
      "epoch": 1.3396836493926418,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048250974308581765,
      "loss": 5.3829,
      "mean_token_accuracy": 0.19183432459831237,
      "num_tokens": 31051413.0,
      "step": 15330
    },
    {
      "entropy": 5.877159357070923,
      "epoch": 1.3401205977453465,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048249769555572844,
      "loss": 5.3892,
      "mean_token_accuracy": 0.1919635221362114,
      "num_tokens": 31061822.0,
      "step": 15335
    },
    {
      "entropy": 5.8798950672149655,
      "epoch": 1.3405575460980512,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048248564404572463,
      "loss": 5.3554,
      "mean_token_accuracy": 0.1902351051568985,
      "num_tokens": 31071889.0,
      "step": 15340
    },
    {
      "entropy": 5.8901854991912845,
      "epoch": 1.340994494450756,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048247358855603754,
      "loss": 5.3234,
      "mean_token_accuracy": 0.19608905166387558,
      "num_tokens": 31080933.0,
      "step": 15345
    },
    {
      "entropy": 5.938668584823608,
      "epoch": 1.3414314428034606,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004824615290868982,
      "loss": 5.4272,
      "mean_token_accuracy": 0.18879844546318053,
      "num_tokens": 31091349.0,
      "step": 15350
    },
    {
      "entropy": 6.0049355030059814,
      "epoch": 1.3418683911561653,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004824494656385381,
      "loss": 5.4404,
      "mean_token_accuracy": 0.18345837146043778,
      "num_tokens": 31101485.0,
      "step": 15355
    },
    {
      "entropy": 5.9244713306427,
      "epoch": 1.34230533950887,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004824373982111886,
      "loss": 5.3768,
      "mean_token_accuracy": 0.18395518660545349,
      "num_tokens": 31111965.0,
      "step": 15360
    },
    {
      "entropy": 5.910174894332886,
      "epoch": 1.3427422878615747,
      "grad_norm": 1.03125,
      "learning_rate": 0.000482425326805081,
      "loss": 5.375,
      "mean_token_accuracy": 0.19336099028587342,
      "num_tokens": 31122284.0,
      "step": 15365
    },
    {
      "entropy": 5.930029964447021,
      "epoch": 1.3431792362142794,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048241325142044697,
      "loss": 5.3923,
      "mean_token_accuracy": 0.18796140253543853,
      "num_tokens": 31132156.0,
      "step": 15370
    },
    {
      "entropy": 5.928755521774292,
      "epoch": 1.3436161845669843,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048240117205751817,
      "loss": 5.3196,
      "mean_token_accuracy": 0.1949525237083435,
      "num_tokens": 31141609.0,
      "step": 15375
    },
    {
      "entropy": 5.835082626342773,
      "epoch": 1.344053132919689,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004823890887165261,
      "loss": 5.3841,
      "mean_token_accuracy": 0.1877817764878273,
      "num_tokens": 31153159.0,
      "step": 15380
    },
    {
      "entropy": 5.8788244247436525,
      "epoch": 1.3444900812723937,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048237700139770273,
      "loss": 5.4261,
      "mean_token_accuracy": 0.19060512632131577,
      "num_tokens": 31163330.0,
      "step": 15385
    },
    {
      "entropy": 5.998034572601318,
      "epoch": 1.3449270296250984,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004823649101012797,
      "loss": 5.4597,
      "mean_token_accuracy": 0.18410485088825226,
      "num_tokens": 31173874.0,
      "step": 15390
    },
    {
      "entropy": 5.89088659286499,
      "epoch": 1.345363977977803,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004823528148274891,
      "loss": 5.3227,
      "mean_token_accuracy": 0.19295923709869384,
      "num_tokens": 31183933.0,
      "step": 15395
    },
    {
      "entropy": 5.871858882904053,
      "epoch": 1.3458009263305077,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048234071557656295,
      "loss": 5.3294,
      "mean_token_accuracy": 0.20088240653276443,
      "num_tokens": 31194308.0,
      "step": 15400
    },
    {
      "entropy": 5.907775259017944,
      "epoch": 1.3462378746832124,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004823286123487331,
      "loss": 5.3696,
      "mean_token_accuracy": 0.1927136540412903,
      "num_tokens": 31203757.0,
      "step": 15405
    },
    {
      "entropy": 5.958554077148437,
      "epoch": 1.3466748230359171,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048231650514423194,
      "loss": 5.4502,
      "mean_token_accuracy": 0.1855989933013916,
      "num_tokens": 31213980.0,
      "step": 15410
    },
    {
      "entropy": 5.982662916183472,
      "epoch": 1.3471117713886218,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048230439396329143,
      "loss": 5.4029,
      "mean_token_accuracy": 0.18790283054113388,
      "num_tokens": 31225034.0,
      "step": 15415
    },
    {
      "entropy": 5.894769716262817,
      "epoch": 1.3475487197413265,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048229227880614413,
      "loss": 5.3911,
      "mean_token_accuracy": 0.18703666776418687,
      "num_tokens": 31236393.0,
      "step": 15420
    },
    {
      "entropy": 5.905417156219483,
      "epoch": 1.3479856680940312,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004822801596730222,
      "loss": 5.3885,
      "mean_token_accuracy": 0.19105320125818254,
      "num_tokens": 31246577.0,
      "step": 15425
    },
    {
      "entropy": 5.921939420700073,
      "epoch": 1.348422616446736,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048226803656415833,
      "loss": 5.3651,
      "mean_token_accuracy": 0.18947611451148988,
      "num_tokens": 31256004.0,
      "step": 15430
    },
    {
      "entropy": 5.973615407943726,
      "epoch": 1.3488595647994406,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004822559094797849,
      "loss": 5.4362,
      "mean_token_accuracy": 0.18820781260728836,
      "num_tokens": 31267294.0,
      "step": 15435
    },
    {
      "entropy": 5.89126877784729,
      "epoch": 1.3492965131521455,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048224377842013443,
      "loss": 5.372,
      "mean_token_accuracy": 0.19635307490825654,
      "num_tokens": 31276896.0,
      "step": 15440
    },
    {
      "entropy": 5.839965391159057,
      "epoch": 1.3497334615048502,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004822316433854397,
      "loss": 5.4185,
      "mean_token_accuracy": 0.18946751207113266,
      "num_tokens": 31287248.0,
      "step": 15445
    },
    {
      "entropy": 5.901244354248047,
      "epoch": 1.3501704098575549,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048221950437593353,
      "loss": 5.3643,
      "mean_token_accuracy": 0.19348460137844087,
      "num_tokens": 31297422.0,
      "step": 15450
    },
    {
      "entropy": 5.9371990203857425,
      "epoch": 1.3506073582102596,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048220736139184864,
      "loss": 5.4353,
      "mean_token_accuracy": 0.18870948106050492,
      "num_tokens": 31307182.0,
      "step": 15455
    },
    {
      "entropy": 5.937413358688355,
      "epoch": 1.3510443065629643,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000482195214433418,
      "loss": 5.5006,
      "mean_token_accuracy": 0.18698268979787827,
      "num_tokens": 31317234.0,
      "step": 15460
    },
    {
      "entropy": 5.954547882080078,
      "epoch": 1.351481254915669,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004821830635008746,
      "loss": 5.4825,
      "mean_token_accuracy": 0.18243261575698852,
      "num_tokens": 31328294.0,
      "step": 15465
    },
    {
      "entropy": 5.974853277206421,
      "epoch": 1.3519182032683736,
      "grad_norm": 1.125,
      "learning_rate": 0.0004821709085944515,
      "loss": 5.3523,
      "mean_token_accuracy": 0.19491346627473832,
      "num_tokens": 31337250.0,
      "step": 15470
    },
    {
      "entropy": 5.900555801391602,
      "epoch": 1.3523551516210783,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004821587497143817,
      "loss": 5.4543,
      "mean_token_accuracy": 0.1824741244316101,
      "num_tokens": 31346900.0,
      "step": 15475
    },
    {
      "entropy": 5.891288566589355,
      "epoch": 1.352792099973783,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004821465868608986,
      "loss": 5.4306,
      "mean_token_accuracy": 0.1812546133995056,
      "num_tokens": 31357233.0,
      "step": 15480
    },
    {
      "entropy": 5.946888542175293,
      "epoch": 1.353229048326488,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004821344200342355,
      "loss": 5.3302,
      "mean_token_accuracy": 0.19245522767305373,
      "num_tokens": 31367043.0,
      "step": 15485
    },
    {
      "entropy": 5.93652310371399,
      "epoch": 1.3536659966791924,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004821222492346256,
      "loss": 5.3545,
      "mean_token_accuracy": 0.18722770512104034,
      "num_tokens": 31376770.0,
      "step": 15490
    },
    {
      "entropy": 5.9118452072143555,
      "epoch": 1.3541029450318973,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048211007446230243,
      "loss": 5.4123,
      "mean_token_accuracy": 0.19272248595952987,
      "num_tokens": 31386056.0,
      "step": 15495
    },
    {
      "entropy": 5.942300176620483,
      "epoch": 1.354539893384602,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048209789571749954,
      "loss": 5.4545,
      "mean_token_accuracy": 0.18530717641115188,
      "num_tokens": 31395129.0,
      "step": 15500
    },
    {
      "entropy": 6.025916337966919,
      "epoch": 1.3549768417373067,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048208571300045045,
      "loss": 5.4688,
      "mean_token_accuracy": 0.1825476571917534,
      "num_tokens": 31404720.0,
      "step": 15505
    },
    {
      "entropy": 5.955374240875244,
      "epoch": 1.3554137900900114,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048207352631138897,
      "loss": 5.4117,
      "mean_token_accuracy": 0.18924205601215363,
      "num_tokens": 31415588.0,
      "step": 15510
    },
    {
      "entropy": 5.919925546646118,
      "epoch": 1.355850738442716,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048206133565054863,
      "loss": 5.4865,
      "mean_token_accuracy": 0.18127070963382722,
      "num_tokens": 31425904.0,
      "step": 15515
    },
    {
      "entropy": 5.943467473983764,
      "epoch": 1.3562876867954208,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004820491410181634,
      "loss": 5.4722,
      "mean_token_accuracy": 0.18373198807239532,
      "num_tokens": 31435282.0,
      "step": 15520
    },
    {
      "entropy": 5.995621538162231,
      "epoch": 1.3567246351481255,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048203694241446724,
      "loss": 5.5115,
      "mean_token_accuracy": 0.18461267948150634,
      "num_tokens": 31445681.0,
      "step": 15525
    },
    {
      "entropy": 5.8674421310424805,
      "epoch": 1.3571615835008302,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048202473983969396,
      "loss": 5.2188,
      "mean_token_accuracy": 0.20154461413621902,
      "num_tokens": 31456934.0,
      "step": 15530
    },
    {
      "entropy": 5.916478252410888,
      "epoch": 1.3575985318535349,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048201253329407766,
      "loss": 5.3764,
      "mean_token_accuracy": 0.19577232599258423,
      "num_tokens": 31466692.0,
      "step": 15535
    },
    {
      "entropy": 5.911339044570923,
      "epoch": 1.3580354802062395,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004820003227778526,
      "loss": 5.4041,
      "mean_token_accuracy": 0.19368426352739335,
      "num_tokens": 31477166.0,
      "step": 15540
    },
    {
      "entropy": 5.97406759262085,
      "epoch": 1.3584724285589442,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004819881082912528,
      "loss": 5.4058,
      "mean_token_accuracy": 0.18580401241779326,
      "num_tokens": 31486998.0,
      "step": 15545
    },
    {
      "entropy": 5.949998188018799,
      "epoch": 1.3589093769116491,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048197588983451265,
      "loss": 5.4457,
      "mean_token_accuracy": 0.18588638454675674,
      "num_tokens": 31496674.0,
      "step": 15550
    },
    {
      "entropy": 5.909130716323853,
      "epoch": 1.3593463252643536,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048196366740786646,
      "loss": 5.4172,
      "mean_token_accuracy": 0.18799530118703842,
      "num_tokens": 31507412.0,
      "step": 15555
    },
    {
      "entropy": 5.934013605117798,
      "epoch": 1.3597832736170585,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048195144101154867,
      "loss": 5.3977,
      "mean_token_accuracy": 0.19013418704271318,
      "num_tokens": 31517048.0,
      "step": 15560
    },
    {
      "entropy": 5.938172149658203,
      "epoch": 1.3602202219697632,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048193921064579385,
      "loss": 5.4698,
      "mean_token_accuracy": 0.1879833683371544,
      "num_tokens": 31526353.0,
      "step": 15565
    },
    {
      "entropy": 5.942385196685791,
      "epoch": 1.360657170322468,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004819269763108365,
      "loss": 5.4599,
      "mean_token_accuracy": 0.1853959709405899,
      "num_tokens": 31537194.0,
      "step": 15570
    },
    {
      "entropy": 5.9838580131530765,
      "epoch": 1.3610941186751726,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048191473800691145,
      "loss": 5.3917,
      "mean_token_accuracy": 0.18582662045955659,
      "num_tokens": 31547896.0,
      "step": 15575
    },
    {
      "entropy": 5.852573108673096,
      "epoch": 1.3615310670278773,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004819024957342531,
      "loss": 5.4113,
      "mean_token_accuracy": 0.18565790206193925,
      "num_tokens": 31558750.0,
      "step": 15580
    },
    {
      "entropy": 5.9180803298950195,
      "epoch": 1.361968015380582,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048189024949309655,
      "loss": 5.3832,
      "mean_token_accuracy": 0.18662008494138718,
      "num_tokens": 31568721.0,
      "step": 15585
    },
    {
      "entropy": 5.865088224411011,
      "epoch": 1.3624049637332867,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004818779992836766,
      "loss": 5.3576,
      "mean_token_accuracy": 0.19057433009147645,
      "num_tokens": 31579907.0,
      "step": 15590
    },
    {
      "entropy": 5.914302062988281,
      "epoch": 1.3628419120859914,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048186574510622823,
      "loss": 5.4469,
      "mean_token_accuracy": 0.18582863807678224,
      "num_tokens": 31590137.0,
      "step": 15595
    },
    {
      "entropy": 5.936892175674439,
      "epoch": 1.363278860438696,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048185348696098653,
      "loss": 5.4529,
      "mean_token_accuracy": 0.18439133167266847,
      "num_tokens": 31599969.0,
      "step": 15600
    },
    {
      "entropy": 5.942535543441773,
      "epoch": 1.363715808791401,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004818412248481865,
      "loss": 5.4307,
      "mean_token_accuracy": 0.18277610689401627,
      "num_tokens": 31609951.0,
      "step": 15605
    },
    {
      "entropy": 5.967480039596557,
      "epoch": 1.3641527571441054,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048182895876806343,
      "loss": 5.351,
      "mean_token_accuracy": 0.19561637043952942,
      "num_tokens": 31619214.0,
      "step": 15610
    },
    {
      "entropy": 5.9452317237854,
      "epoch": 1.3645897054968104,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004818166887208524,
      "loss": 5.3771,
      "mean_token_accuracy": 0.19294212460517884,
      "num_tokens": 31629562.0,
      "step": 15615
    },
    {
      "entropy": 5.93432650566101,
      "epoch": 1.365026653849515,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004818044147067891,
      "loss": 5.3574,
      "mean_token_accuracy": 0.19308866411447526,
      "num_tokens": 31639801.0,
      "step": 15620
    },
    {
      "entropy": 5.965869188308716,
      "epoch": 1.3654636022022197,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048179213672610866,
      "loss": 5.4388,
      "mean_token_accuracy": 0.18222691863775253,
      "num_tokens": 31650593.0,
      "step": 15625
    },
    {
      "entropy": 5.922257709503174,
      "epoch": 1.3659005505549244,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048177985477904675,
      "loss": 5.3852,
      "mean_token_accuracy": 0.1885321021080017,
      "num_tokens": 31660942.0,
      "step": 15630
    },
    {
      "entropy": 5.950902938842773,
      "epoch": 1.3663374989076291,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048176756886583884,
      "loss": 5.4391,
      "mean_token_accuracy": 0.18292322009801865,
      "num_tokens": 31671105.0,
      "step": 15635
    },
    {
      "entropy": 5.9649045944213865,
      "epoch": 1.3667744472603338,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004817552789867206,
      "loss": 5.3942,
      "mean_token_accuracy": 0.19148000329732895,
      "num_tokens": 31680696.0,
      "step": 15640
    },
    {
      "entropy": 5.829177761077881,
      "epoch": 1.3672113956130385,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048174298514192784,
      "loss": 5.3337,
      "mean_token_accuracy": 0.19585084915161133,
      "num_tokens": 31689755.0,
      "step": 15645
    },
    {
      "entropy": 5.86336088180542,
      "epoch": 1.3676483439657432,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048173068733169615,
      "loss": 5.3657,
      "mean_token_accuracy": 0.19218877404928209,
      "num_tokens": 31699614.0,
      "step": 15650
    },
    {
      "entropy": 6.001069450378418,
      "epoch": 1.3680852923184479,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004817183855562617,
      "loss": 5.3658,
      "mean_token_accuracy": 0.1948654055595398,
      "num_tokens": 31709106.0,
      "step": 15655
    },
    {
      "entropy": 5.967438697814941,
      "epoch": 1.3685222406711528,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048170607981586023,
      "loss": 5.4905,
      "mean_token_accuracy": 0.18355343043804168,
      "num_tokens": 31719542.0,
      "step": 15660
    },
    {
      "entropy": 5.953834342956543,
      "epoch": 1.3689591890238573,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048169377011072784,
      "loss": 5.386,
      "mean_token_accuracy": 0.18963207602500914,
      "num_tokens": 31730477.0,
      "step": 15665
    },
    {
      "entropy": 5.910719871520996,
      "epoch": 1.3693961373765622,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004816814564411006,
      "loss": 5.3769,
      "mean_token_accuracy": 0.19145529270172118,
      "num_tokens": 31740269.0,
      "step": 15670
    },
    {
      "entropy": 5.919774007797241,
      "epoch": 1.3698330857292669,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004816691388072148,
      "loss": 5.3668,
      "mean_token_accuracy": 0.18916304856538774,
      "num_tokens": 31750153.0,
      "step": 15675
    },
    {
      "entropy": 5.982410192489624,
      "epoch": 1.3702700340819716,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048165681720930653,
      "loss": 5.4412,
      "mean_token_accuracy": 0.1835963785648346,
      "num_tokens": 31760171.0,
      "step": 15680
    },
    {
      "entropy": 5.943642854690552,
      "epoch": 1.3707069824346763,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004816444916476123,
      "loss": 5.4107,
      "mean_token_accuracy": 0.18434692770242692,
      "num_tokens": 31771601.0,
      "step": 15685
    },
    {
      "entropy": 5.965985584259033,
      "epoch": 1.371143930787381,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048163216212236844,
      "loss": 5.3293,
      "mean_token_accuracy": 0.19703148454427719,
      "num_tokens": 31781423.0,
      "step": 15690
    },
    {
      "entropy": 5.883529615402222,
      "epoch": 1.3715808791400856,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004816198286338115,
      "loss": 5.2951,
      "mean_token_accuracy": 0.19994248151779176,
      "num_tokens": 31791025.0,
      "step": 15695
    },
    {
      "entropy": 5.813385772705078,
      "epoch": 1.3720178274927903,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048160749118217783,
      "loss": 5.4173,
      "mean_token_accuracy": 0.18607256859540938,
      "num_tokens": 31801125.0,
      "step": 15700
    },
    {
      "entropy": 5.977494573593139,
      "epoch": 1.372454775845495,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004815951497677043,
      "loss": 5.4217,
      "mean_token_accuracy": 0.1817416787147522,
      "num_tokens": 31811100.0,
      "step": 15705
    },
    {
      "entropy": 5.983398580551148,
      "epoch": 1.3728917241981997,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004815828043906275,
      "loss": 5.4126,
      "mean_token_accuracy": 0.18428441733121873,
      "num_tokens": 31821667.0,
      "step": 15710
    },
    {
      "entropy": 5.922186231613159,
      "epoch": 1.3733286725509044,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004815704550511843,
      "loss": 5.3684,
      "mean_token_accuracy": 0.1934458002448082,
      "num_tokens": 31831859.0,
      "step": 15715
    },
    {
      "entropy": 5.942836284637451,
      "epoch": 1.373765620903609,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004815581017496115,
      "loss": 5.3522,
      "mean_token_accuracy": 0.18955797106027603,
      "num_tokens": 31842952.0,
      "step": 15720
    },
    {
      "entropy": 5.860512733459473,
      "epoch": 1.374202569256314,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048154574448614607,
      "loss": 5.4607,
      "mean_token_accuracy": 0.1875806674361229,
      "num_tokens": 31852869.0,
      "step": 15725
    },
    {
      "entropy": 5.926542711257935,
      "epoch": 1.3746395176090185,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004815333832610251,
      "loss": 5.3613,
      "mean_token_accuracy": 0.1953507259488106,
      "num_tokens": 31862590.0,
      "step": 15730
    },
    {
      "entropy": 5.901823329925537,
      "epoch": 1.3750764659617234,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048152101807448554,
      "loss": 5.4103,
      "mean_token_accuracy": 0.19443651139736176,
      "num_tokens": 31873071.0,
      "step": 15735
    },
    {
      "entropy": 5.906974124908447,
      "epoch": 1.375513414314428,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004815086489267647,
      "loss": 5.4175,
      "mean_token_accuracy": 0.19121112078428268,
      "num_tokens": 31883325.0,
      "step": 15740
    },
    {
      "entropy": 5.90954065322876,
      "epoch": 1.3759503626671328,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048149627581809974,
      "loss": 5.4369,
      "mean_token_accuracy": 0.18698863536119462,
      "num_tokens": 31893071.0,
      "step": 15745
    },
    {
      "entropy": 5.896205186843872,
      "epoch": 1.3763873110198375,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000481483898748728,
      "loss": 5.3631,
      "mean_token_accuracy": 0.1976223722100258,
      "num_tokens": 31902948.0,
      "step": 15750
    },
    {
      "entropy": 5.933066701889038,
      "epoch": 1.3768242593725422,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004814715177188869,
      "loss": 5.3995,
      "mean_token_accuracy": 0.18631459772586823,
      "num_tokens": 31913425.0,
      "step": 15755
    },
    {
      "entropy": 5.876261425018311,
      "epoch": 1.3772612077252468,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004814591327288139,
      "loss": 5.4133,
      "mean_token_accuracy": 0.1883334755897522,
      "num_tokens": 31923608.0,
      "step": 15760
    },
    {
      "entropy": 5.89783935546875,
      "epoch": 1.3776981560779515,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048144674377874656,
      "loss": 5.336,
      "mean_token_accuracy": 0.19112644344568253,
      "num_tokens": 31932863.0,
      "step": 15765
    },
    {
      "entropy": 5.939556407928467,
      "epoch": 1.3781351044306562,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048143435086892244,
      "loss": 5.4061,
      "mean_token_accuracy": 0.18697402775287628,
      "num_tokens": 31942425.0,
      "step": 15770
    },
    {
      "entropy": 5.909504508972168,
      "epoch": 1.378572052783361,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004814219539995793,
      "loss": 5.4019,
      "mean_token_accuracy": 0.18929953128099442,
      "num_tokens": 31952152.0,
      "step": 15775
    },
    {
      "entropy": 5.930842876434326,
      "epoch": 1.3790090011360658,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048140955317095503,
      "loss": 5.4386,
      "mean_token_accuracy": 0.18445633798837663,
      "num_tokens": 31962224.0,
      "step": 15780
    },
    {
      "entropy": 5.918753242492675,
      "epoch": 1.3794459494887703,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004813971483832873,
      "loss": 5.3697,
      "mean_token_accuracy": 0.19660432636737823,
      "num_tokens": 31971517.0,
      "step": 15785
    },
    {
      "entropy": 5.936580848693848,
      "epoch": 1.3798828978414752,
      "grad_norm": 1.0,
      "learning_rate": 0.0004813847396368141,
      "loss": 5.4189,
      "mean_token_accuracy": 0.1913294717669487,
      "num_tokens": 31982461.0,
      "step": 15790
    },
    {
      "entropy": 5.902597332000733,
      "epoch": 1.38031984619418,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004813723269317734,
      "loss": 5.3278,
      "mean_token_accuracy": 0.2018505647778511,
      "num_tokens": 31992202.0,
      "step": 15795
    },
    {
      "entropy": 5.963249921798706,
      "epoch": 1.3807567945468846,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004813599102684034,
      "loss": 5.4,
      "mean_token_accuracy": 0.1937546655535698,
      "num_tokens": 32002283.0,
      "step": 15800
    },
    {
      "entropy": 5.934924077987671,
      "epoch": 1.3811937428995893,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048134748964694227,
      "loss": 5.3767,
      "mean_token_accuracy": 0.1859770745038986,
      "num_tokens": 32013135.0,
      "step": 15805
    },
    {
      "entropy": 5.911010456085205,
      "epoch": 1.381630691252294,
      "grad_norm": 1.125,
      "learning_rate": 0.0004813350650676281,
      "loss": 5.414,
      "mean_token_accuracy": 0.18636096119880677,
      "num_tokens": 32022631.0,
      "step": 15810
    },
    {
      "entropy": 5.9712158203125,
      "epoch": 1.3820676396049987,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004813226365306992,
      "loss": 5.4183,
      "mean_token_accuracy": 0.1774699032306671,
      "num_tokens": 32032216.0,
      "step": 15815
    },
    {
      "entropy": 5.915241050720215,
      "epoch": 1.3825045879577034,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048131020403639405,
      "loss": 5.408,
      "mean_token_accuracy": 0.19716142416000365,
      "num_tokens": 32042382.0,
      "step": 15820
    },
    {
      "entropy": 5.921321630477905,
      "epoch": 1.382941536310408,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048129776758495116,
      "loss": 5.3657,
      "mean_token_accuracy": 0.19254121482372283,
      "num_tokens": 32052592.0,
      "step": 15825
    },
    {
      "entropy": 5.90276870727539,
      "epoch": 1.3833784846631128,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004812853271766089,
      "loss": 5.3106,
      "mean_token_accuracy": 0.19231097996234894,
      "num_tokens": 32062260.0,
      "step": 15830
    },
    {
      "entropy": 5.92298526763916,
      "epoch": 1.3838154330158177,
      "grad_norm": 1.09375,
      "learning_rate": 0.000481272882811606,
      "loss": 5.4055,
      "mean_token_accuracy": 0.1911344513297081,
      "num_tokens": 32072079.0,
      "step": 15835
    },
    {
      "entropy": 5.898052978515625,
      "epoch": 1.3842523813685221,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048126043449018115,
      "loss": 5.4122,
      "mean_token_accuracy": 0.19184153378009797,
      "num_tokens": 32082151.0,
      "step": 15840
    },
    {
      "entropy": 5.997023582458496,
      "epoch": 1.384689329721227,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000481247982212573,
      "loss": 5.3964,
      "mean_token_accuracy": 0.19457949101924896,
      "num_tokens": 32091893.0,
      "step": 15845
    },
    {
      "entropy": 5.940587091445923,
      "epoch": 1.3851262780739317,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048123552597902053,
      "loss": 5.3575,
      "mean_token_accuracy": 0.19238243848085404,
      "num_tokens": 32101550.0,
      "step": 15850
    },
    {
      "entropy": 5.793041515350342,
      "epoch": 1.3855632264266364,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048122306578976266,
      "loss": 5.2603,
      "mean_token_accuracy": 0.20557148307561873,
      "num_tokens": 32112278.0,
      "step": 15855
    },
    {
      "entropy": 5.93659348487854,
      "epoch": 1.3860001747793411,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048121060164503835,
      "loss": 5.4407,
      "mean_token_accuracy": 0.18138837665319443,
      "num_tokens": 32122284.0,
      "step": 15860
    },
    {
      "entropy": 5.98124680519104,
      "epoch": 1.3864371231320458,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004811981335450865,
      "loss": 5.4075,
      "mean_token_accuracy": 0.18542490452528,
      "num_tokens": 32131552.0,
      "step": 15865
    },
    {
      "entropy": 6.013390731811524,
      "epoch": 1.3868740714847505,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048118566149014653,
      "loss": 5.4777,
      "mean_token_accuracy": 0.1803198605775833,
      "num_tokens": 32141758.0,
      "step": 15870
    },
    {
      "entropy": 5.89038724899292,
      "epoch": 1.3873110198374552,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004811731854804575,
      "loss": 5.3693,
      "mean_token_accuracy": 0.191001957654953,
      "num_tokens": 32150679.0,
      "step": 15875
    },
    {
      "entropy": 5.822209787368775,
      "epoch": 1.3877479681901599,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004811607055162587,
      "loss": 5.3431,
      "mean_token_accuracy": 0.19070906937122345,
      "num_tokens": 32160967.0,
      "step": 15880
    },
    {
      "entropy": 5.999260091781617,
      "epoch": 1.3881849165428646,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004811482215977895,
      "loss": 5.4347,
      "mean_token_accuracy": 0.18152179718017578,
      "num_tokens": 32170516.0,
      "step": 15885
    },
    {
      "entropy": 5.948375463485718,
      "epoch": 1.3886218648955693,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048113573372528956,
      "loss": 5.4256,
      "mean_token_accuracy": 0.18740050941705705,
      "num_tokens": 32180741.0,
      "step": 15890
    },
    {
      "entropy": 5.9992334842681885,
      "epoch": 1.389058813248274,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004811232418989981,
      "loss": 5.4343,
      "mean_token_accuracy": 0.18890583366155625,
      "num_tokens": 32190646.0,
      "step": 15895
    },
    {
      "entropy": 5.900910425186157,
      "epoch": 1.3894957616009789,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048111074611915483,
      "loss": 5.3947,
      "mean_token_accuracy": 0.1879427030682564,
      "num_tokens": 32201361.0,
      "step": 15900
    },
    {
      "entropy": 5.945524597167969,
      "epoch": 1.3899327099536833,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004810982463859996,
      "loss": 5.4869,
      "mean_token_accuracy": 0.1790340393781662,
      "num_tokens": 32211759.0,
      "step": 15905
    },
    {
      "entropy": 5.959325838088989,
      "epoch": 1.3903696583063883,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004810857426997719,
      "loss": 5.4469,
      "mean_token_accuracy": 0.18835808336734772,
      "num_tokens": 32222582.0,
      "step": 15910
    },
    {
      "entropy": 5.907512140274048,
      "epoch": 1.390806606659093,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048107323506071167,
      "loss": 5.4189,
      "mean_token_accuracy": 0.19367744624614716,
      "num_tokens": 32232936.0,
      "step": 15915
    },
    {
      "entropy": 5.937647771835327,
      "epoch": 1.3912435550117976,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048106072346905875,
      "loss": 5.4342,
      "mean_token_accuracy": 0.183608378469944,
      "num_tokens": 32242392.0,
      "step": 15920
    },
    {
      "entropy": 5.996935510635376,
      "epoch": 1.3916805033645023,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048104820792505335,
      "loss": 5.4856,
      "mean_token_accuracy": 0.18290978223085402,
      "num_tokens": 32252385.0,
      "step": 15925
    },
    {
      "entropy": 5.9627251625061035,
      "epoch": 1.392117451717207,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048103568842893527,
      "loss": 5.4511,
      "mean_token_accuracy": 0.1800887554883957,
      "num_tokens": 32263537.0,
      "step": 15930
    },
    {
      "entropy": 5.929212045669556,
      "epoch": 1.3925544000699117,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048102316498094467,
      "loss": 5.4249,
      "mean_token_accuracy": 0.18811047524213792,
      "num_tokens": 32273450.0,
      "step": 15935
    },
    {
      "entropy": 5.949815893173218,
      "epoch": 1.3929913484226164,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048101063758132186,
      "loss": 5.4143,
      "mean_token_accuracy": 0.18962176591157914,
      "num_tokens": 32283243.0,
      "step": 15940
    },
    {
      "entropy": 5.87922945022583,
      "epoch": 1.393428296775321,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004809981062303071,
      "loss": 5.3502,
      "mean_token_accuracy": 0.19418465942144394,
      "num_tokens": 32292458.0,
      "step": 15945
    },
    {
      "entropy": 5.947406530380249,
      "epoch": 1.3938652451280258,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048098557092814066,
      "loss": 5.3555,
      "mean_token_accuracy": 0.1953027606010437,
      "num_tokens": 32302574.0,
      "step": 15950
    },
    {
      "entropy": 5.9246203899383545,
      "epoch": 1.3943021934807307,
      "grad_norm": 1.046875,
      "learning_rate": 0.000480973031675063,
      "loss": 5.3735,
      "mean_token_accuracy": 0.19516233950853348,
      "num_tokens": 32312658.0,
      "step": 15955
    },
    {
      "entropy": 5.936765718460083,
      "epoch": 1.3947391418334352,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004809604884713148,
      "loss": 5.4559,
      "mean_token_accuracy": 0.18382129222154617,
      "num_tokens": 32322717.0,
      "step": 15960
    },
    {
      "entropy": 5.835982465744019,
      "epoch": 1.39517609018614,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048094794131713645,
      "loss": 5.3656,
      "mean_token_accuracy": 0.19576722085475923,
      "num_tokens": 32333316.0,
      "step": 15965
    },
    {
      "entropy": 5.945925235748291,
      "epoch": 1.3956130385388448,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004809353902127686,
      "loss": 5.3967,
      "mean_token_accuracy": 0.19223746210336684,
      "num_tokens": 32343857.0,
      "step": 15970
    },
    {
      "entropy": 5.934004831314087,
      "epoch": 1.3960499868915495,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048092283515845216,
      "loss": 5.3037,
      "mean_token_accuracy": 0.197795769572258,
      "num_tokens": 32353883.0,
      "step": 15975
    },
    {
      "entropy": 5.82573618888855,
      "epoch": 1.3964869352442542,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004809102761544278,
      "loss": 5.3142,
      "mean_token_accuracy": 0.19158945232629776,
      "num_tokens": 32363355.0,
      "step": 15980
    },
    {
      "entropy": 5.909144449234009,
      "epoch": 1.3969238835969588,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004808977132009364,
      "loss": 5.3916,
      "mean_token_accuracy": 0.18794154077768327,
      "num_tokens": 32373066.0,
      "step": 15985
    },
    {
      "entropy": 5.928152465820313,
      "epoch": 1.3973608319496635,
      "grad_norm": 1.015625,
      "learning_rate": 0.000480885146298219,
      "loss": 5.377,
      "mean_token_accuracy": 0.1937905728816986,
      "num_tokens": 32383720.0,
      "step": 15990
    },
    {
      "entropy": 5.959723091125488,
      "epoch": 1.3977977803023682,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004808725754465167,
      "loss": 5.3833,
      "mean_token_accuracy": 0.18545129597187043,
      "num_tokens": 32393823.0,
      "step": 15995
    },
    {
      "entropy": 5.803153562545776,
      "epoch": 1.398234728655073,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048086000064607044,
      "loss": 5.3601,
      "mean_token_accuracy": 0.19230881184339524,
      "num_tokens": 32404745.0,
      "step": 16000
    },
    {
      "entropy": 5.88792667388916,
      "epoch": 1.3986716770077776,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048084742189712147,
      "loss": 5.3603,
      "mean_token_accuracy": 0.19908595979213714,
      "num_tokens": 32415325.0,
      "step": 16005
    },
    {
      "entropy": 5.9637792110443115,
      "epoch": 1.3991086253604825,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004808348391999112,
      "loss": 5.4461,
      "mean_token_accuracy": 0.19039007276296616,
      "num_tokens": 32425810.0,
      "step": 16010
    },
    {
      "entropy": 5.9417750358581545,
      "epoch": 1.399545573713187,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004808222525546808,
      "loss": 5.4351,
      "mean_token_accuracy": 0.18690823763608932,
      "num_tokens": 32436799.0,
      "step": 16015
    },
    {
      "entropy": 5.915511178970337,
      "epoch": 1.399982522065892,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048080966196167163,
      "loss": 5.3527,
      "mean_token_accuracy": 0.18955630213022232,
      "num_tokens": 32446409.0,
      "step": 16020
    },
    {
      "entropy": 5.948288345336914,
      "epoch": 1.4004194704185966,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004807970674211255,
      "loss": 5.4063,
      "mean_token_accuracy": 0.18342116624116897,
      "num_tokens": 32456268.0,
      "step": 16025
    },
    {
      "entropy": 5.872245931625367,
      "epoch": 1.4008564187713013,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048078446893328364,
      "loss": 5.4907,
      "mean_token_accuracy": 0.18287250250577927,
      "num_tokens": 32467622.0,
      "step": 16030
    },
    {
      "entropy": 5.9805042266845705,
      "epoch": 1.401293367124006,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004807718664983879,
      "loss": 5.4265,
      "mean_token_accuracy": 0.18713460713624955,
      "num_tokens": 32477753.0,
      "step": 16035
    },
    {
      "entropy": 5.934474468231201,
      "epoch": 1.4017303154767107,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048075926011667997,
      "loss": 5.3105,
      "mean_token_accuracy": 0.19666449576616288,
      "num_tokens": 32487758.0,
      "step": 16040
    },
    {
      "entropy": 5.902699136734009,
      "epoch": 1.4021672638294154,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048074664978840147,
      "loss": 5.3525,
      "mean_token_accuracy": 0.19005557745695115,
      "num_tokens": 32497725.0,
      "step": 16045
    },
    {
      "entropy": 5.888387966156006,
      "epoch": 1.40260421218212,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004807340355137945,
      "loss": 5.2774,
      "mean_token_accuracy": 0.20323949605226516,
      "num_tokens": 32507515.0,
      "step": 16050
    },
    {
      "entropy": 5.895556783676147,
      "epoch": 1.4030411605348247,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000480721417293101,
      "loss": 5.3292,
      "mean_token_accuracy": 0.19752669781446458,
      "num_tokens": 32516943.0,
      "step": 16055
    },
    {
      "entropy": 5.83926682472229,
      "epoch": 1.4034781088875294,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004807087951265628,
      "loss": 5.2661,
      "mean_token_accuracy": 0.20339374244213104,
      "num_tokens": 32527223.0,
      "step": 16060
    },
    {
      "entropy": 5.878695964813232,
      "epoch": 1.4039150572402341,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004806961690144222,
      "loss": 5.354,
      "mean_token_accuracy": 0.19479555636644363,
      "num_tokens": 32536668.0,
      "step": 16065
    },
    {
      "entropy": 5.950730705261231,
      "epoch": 1.4043520055929388,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004806835389569213,
      "loss": 5.3919,
      "mean_token_accuracy": 0.19276262521743776,
      "num_tokens": 32546566.0,
      "step": 16070
    },
    {
      "entropy": 5.8865399837493895,
      "epoch": 1.4047889539456437,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048067090495430235,
      "loss": 5.3052,
      "mean_token_accuracy": 0.19126076102256775,
      "num_tokens": 32555927.0,
      "step": 16075
    },
    {
      "entropy": 5.92861704826355,
      "epoch": 1.4052259022983482,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048065826700680763,
      "loss": 5.3552,
      "mean_token_accuracy": 0.19382271766662598,
      "num_tokens": 32566569.0,
      "step": 16080
    },
    {
      "entropy": 5.875193881988525,
      "epoch": 1.4056628506510531,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048064562511467956,
      "loss": 5.3327,
      "mean_token_accuracy": 0.19615849256515502,
      "num_tokens": 32577076.0,
      "step": 16085
    },
    {
      "entropy": 5.861883592605591,
      "epoch": 1.4060997990037578,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048063297927816067,
      "loss": 5.4059,
      "mean_token_accuracy": 0.18204522281885147,
      "num_tokens": 32587926.0,
      "step": 16090
    },
    {
      "entropy": 5.921440172195434,
      "epoch": 1.4065367473564625,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004806203294974935,
      "loss": 5.4534,
      "mean_token_accuracy": 0.19232698082923888,
      "num_tokens": 32599012.0,
      "step": 16095
    },
    {
      "entropy": 6.022868299484253,
      "epoch": 1.4069736957091672,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048060767577292067,
      "loss": 5.3525,
      "mean_token_accuracy": 0.19329773634672165,
      "num_tokens": 32609094.0,
      "step": 16100
    },
    {
      "entropy": 5.897890377044678,
      "epoch": 1.4074106440618719,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004805950181046849,
      "loss": 5.4243,
      "mean_token_accuracy": 0.1879214257001877,
      "num_tokens": 32620356.0,
      "step": 16105
    },
    {
      "entropy": 5.93077712059021,
      "epoch": 1.4078475924145766,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000480582356493029,
      "loss": 5.4134,
      "mean_token_accuracy": 0.18897202759981155,
      "num_tokens": 32629841.0,
      "step": 16110
    },
    {
      "entropy": 5.950577306747436,
      "epoch": 1.4082845407672813,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004805696909381957,
      "loss": 5.4423,
      "mean_token_accuracy": 0.18740929961204528,
      "num_tokens": 32639006.0,
      "step": 16115
    },
    {
      "entropy": 5.956612825393677,
      "epoch": 1.408721489119986,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048055702144042794,
      "loss": 5.4203,
      "mean_token_accuracy": 0.18856735676527023,
      "num_tokens": 32648419.0,
      "step": 16120
    },
    {
      "entropy": 5.911210775375366,
      "epoch": 1.4091584374726907,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004805443479999689,
      "loss": 5.4119,
      "mean_token_accuracy": 0.1930314615368843,
      "num_tokens": 32658050.0,
      "step": 16125
    },
    {
      "entropy": 5.906906175613403,
      "epoch": 1.4095953858253956,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004805316706170615,
      "loss": 5.3934,
      "mean_token_accuracy": 0.18748139441013337,
      "num_tokens": 32668176.0,
      "step": 16130
    },
    {
      "entropy": 5.874647760391236,
      "epoch": 1.4100323341781,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048051898929194907,
      "loss": 5.3234,
      "mean_token_accuracy": 0.19439055919647216,
      "num_tokens": 32678517.0,
      "step": 16135
    },
    {
      "entropy": 5.971245861053466,
      "epoch": 1.410469282530805,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004805063040248747,
      "loss": 5.3733,
      "mean_token_accuracy": 0.19361704885959624,
      "num_tokens": 32688459.0,
      "step": 16140
    },
    {
      "entropy": 5.914208316802979,
      "epoch": 1.4109062308835096,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004804936148160817,
      "loss": 5.3714,
      "mean_token_accuracy": 0.19330816417932511,
      "num_tokens": 32698481.0,
      "step": 16145
    },
    {
      "entropy": 5.865154981613159,
      "epoch": 1.4113431792362143,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004804809216658135,
      "loss": 5.416,
      "mean_token_accuracy": 0.18777146190404892,
      "num_tokens": 32709031.0,
      "step": 16150
    },
    {
      "entropy": 5.877122163772583,
      "epoch": 1.411780127588919,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048046822457431347,
      "loss": 5.3564,
      "mean_token_accuracy": 0.20140321105718612,
      "num_tokens": 32720293.0,
      "step": 16155
    },
    {
      "entropy": 5.932410383224488,
      "epoch": 1.4122170759416237,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048045552354182536,
      "loss": 5.408,
      "mean_token_accuracy": 0.19444107711315156,
      "num_tokens": 32730108.0,
      "step": 16160
    },
    {
      "entropy": 5.94854884147644,
      "epoch": 1.4126540242943284,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048044281856859255,
      "loss": 5.4145,
      "mean_token_accuracy": 0.19040635228157043,
      "num_tokens": 32740614.0,
      "step": 16165
    },
    {
      "entropy": 5.959013414382935,
      "epoch": 1.413090972647033,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004804301096548589,
      "loss": 5.3715,
      "mean_token_accuracy": 0.19229773432016373,
      "num_tokens": 32751155.0,
      "step": 16170
    },
    {
      "entropy": 5.88377275466919,
      "epoch": 1.4135279209997378,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004804173968008681,
      "loss": 5.2643,
      "mean_token_accuracy": 0.20098385959863663,
      "num_tokens": 32761067.0,
      "step": 16175
    },
    {
      "entropy": 5.802181196212769,
      "epoch": 1.4139648693524425,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048040468000686395,
      "loss": 5.2651,
      "mean_token_accuracy": 0.2066579833626747,
      "num_tokens": 32771357.0,
      "step": 16180
    },
    {
      "entropy": 5.92797818183899,
      "epoch": 1.4144018177051472,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004803919592730904,
      "loss": 5.4596,
      "mean_token_accuracy": 0.18854528665542603,
      "num_tokens": 32781116.0,
      "step": 16185
    },
    {
      "entropy": 5.94848599433899,
      "epoch": 1.4148387660578519,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004803792345997915,
      "loss": 5.4142,
      "mean_token_accuracy": 0.1864178240299225,
      "num_tokens": 32791660.0,
      "step": 16190
    },
    {
      "entropy": 5.895408868789673,
      "epoch": 1.4152757144105568,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048036650598721125,
      "loss": 5.3673,
      "mean_token_accuracy": 0.1915145680308342,
      "num_tokens": 32801413.0,
      "step": 16195
    },
    {
      "entropy": 5.920891141891479,
      "epoch": 1.4157126627632615,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048035377343559377,
      "loss": 5.4455,
      "mean_token_accuracy": 0.18711694926023484,
      "num_tokens": 32811293.0,
      "step": 16200
    },
    {
      "entropy": 5.940282917022705,
      "epoch": 1.4161496111159662,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048034103694518336,
      "loss": 5.3986,
      "mean_token_accuracy": 0.1932341441512108,
      "num_tokens": 32821609.0,
      "step": 16205
    },
    {
      "entropy": 5.954688930511475,
      "epoch": 1.4165865594686708,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048032829651622425,
      "loss": 5.3693,
      "mean_token_accuracy": 0.19395513236522674,
      "num_tokens": 32831219.0,
      "step": 16210
    },
    {
      "entropy": 5.945771932601929,
      "epoch": 1.4170235078213755,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048031555214896076,
      "loss": 5.4175,
      "mean_token_accuracy": 0.18480684608221054,
      "num_tokens": 32841085.0,
      "step": 16215
    },
    {
      "entropy": 5.9158872127532955,
      "epoch": 1.4174604561740802,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004803028038436375,
      "loss": 5.3602,
      "mean_token_accuracy": 0.19622723162174224,
      "num_tokens": 32851592.0,
      "step": 16220
    },
    {
      "entropy": 5.922158813476562,
      "epoch": 1.417897404526785,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048029005160049886,
      "loss": 5.3573,
      "mean_token_accuracy": 0.1916589096188545,
      "num_tokens": 32862124.0,
      "step": 16225
    },
    {
      "entropy": 5.854958724975586,
      "epoch": 1.4183343528794896,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004802772954197894,
      "loss": 5.3706,
      "mean_token_accuracy": 0.19481732696294785,
      "num_tokens": 32872760.0,
      "step": 16230
    },
    {
      "entropy": 5.936741209030151,
      "epoch": 1.4187713012321943,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048026453530175393,
      "loss": 5.4373,
      "mean_token_accuracy": 0.18855455219745637,
      "num_tokens": 32883542.0,
      "step": 16235
    },
    {
      "entropy": 5.945894432067871,
      "epoch": 1.419208249584899,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048025177124663704,
      "loss": 5.3589,
      "mean_token_accuracy": 0.18813807964324952,
      "num_tokens": 32893922.0,
      "step": 16240
    },
    {
      "entropy": 5.944129180908203,
      "epoch": 1.4196451979376037,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048023900325468367,
      "loss": 5.4086,
      "mean_token_accuracy": 0.19249464869499205,
      "num_tokens": 32903907.0,
      "step": 16245
    },
    {
      "entropy": 5.95349087715149,
      "epoch": 1.4200821462903086,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048022623132613863,
      "loss": 5.3966,
      "mean_token_accuracy": 0.19005063325166702,
      "num_tokens": 32914656.0,
      "step": 16250
    },
    {
      "entropy": 5.964418506622314,
      "epoch": 1.420519094643013,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000480213455461247,
      "loss": 5.4127,
      "mean_token_accuracy": 0.18650437891483307,
      "num_tokens": 32925250.0,
      "step": 16255
    },
    {
      "entropy": 5.950198602676392,
      "epoch": 1.420956042995718,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048020067566025373,
      "loss": 5.4017,
      "mean_token_accuracy": 0.18971362560987473,
      "num_tokens": 32935435.0,
      "step": 16260
    },
    {
      "entropy": 5.8930731296539305,
      "epoch": 1.4213929913484227,
      "grad_norm": 1.125,
      "learning_rate": 0.000480187891923404,
      "loss": 5.3466,
      "mean_token_accuracy": 0.19211373329162598,
      "num_tokens": 32946166.0,
      "step": 16265
    },
    {
      "entropy": 5.879304265975952,
      "epoch": 1.4218299397011274,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004801751042509429,
      "loss": 5.3315,
      "mean_token_accuracy": 0.19240437150001527,
      "num_tokens": 32955751.0,
      "step": 16270
    },
    {
      "entropy": 5.9611396312713625,
      "epoch": 1.422266888053832,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004801623126431159,
      "loss": 5.4221,
      "mean_token_accuracy": 0.19137579798698426,
      "num_tokens": 32966121.0,
      "step": 16275
    },
    {
      "entropy": 5.869462251663208,
      "epoch": 1.4227038364065367,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004801495171001681,
      "loss": 5.3732,
      "mean_token_accuracy": 0.19691570699214936,
      "num_tokens": 32976950.0,
      "step": 16280
    },
    {
      "entropy": 5.93516058921814,
      "epoch": 1.4231407847592414,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004801367176223452,
      "loss": 5.3733,
      "mean_token_accuracy": 0.1975690945982933,
      "num_tokens": 32987278.0,
      "step": 16285
    },
    {
      "entropy": 5.962365961074829,
      "epoch": 1.4235777331119461,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004801239142098926,
      "loss": 5.4307,
      "mean_token_accuracy": 0.1881343975663185,
      "num_tokens": 32997050.0,
      "step": 16290
    },
    {
      "entropy": 5.957546329498291,
      "epoch": 1.4240146814646508,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004801111068630557,
      "loss": 5.3817,
      "mean_token_accuracy": 0.19174536913633347,
      "num_tokens": 33007799.0,
      "step": 16295
    },
    {
      "entropy": 6.0058403491973875,
      "epoch": 1.4244516298173555,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048009829558208044,
      "loss": 5.379,
      "mean_token_accuracy": 0.19253240376710892,
      "num_tokens": 33017476.0,
      "step": 16300
    },
    {
      "entropy": 5.884482097625733,
      "epoch": 1.4248885781700604,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048008548036721226,
      "loss": 5.3601,
      "mean_token_accuracy": 0.19435580521821977,
      "num_tokens": 33027617.0,
      "step": 16305
    },
    {
      "entropy": 5.8657091617584225,
      "epoch": 1.425325526522765,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004800726612186972,
      "loss": 5.3827,
      "mean_token_accuracy": 0.188617305457592,
      "num_tokens": 33037257.0,
      "step": 16310
    },
    {
      "entropy": 5.898414754867554,
      "epoch": 1.4257624748754698,
      "grad_norm": 1.125,
      "learning_rate": 0.0004800598381367811,
      "loss": 5.3354,
      "mean_token_accuracy": 0.1972358763217926,
      "num_tokens": 33047406.0,
      "step": 16315
    },
    {
      "entropy": 5.929562616348266,
      "epoch": 1.4261994232281745,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004800470111217096,
      "loss": 5.3858,
      "mean_token_accuracy": 0.1960243284702301,
      "num_tokens": 33057975.0,
      "step": 16320
    },
    {
      "entropy": 5.983206272125244,
      "epoch": 1.4266363715808792,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048003418017372924,
      "loss": 5.4203,
      "mean_token_accuracy": 0.1886355310678482,
      "num_tokens": 33068645.0,
      "step": 16325
    },
    {
      "entropy": 5.963585376739502,
      "epoch": 1.4270733199335839,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004800213452930857,
      "loss": 5.3963,
      "mean_token_accuracy": 0.18947566747665406,
      "num_tokens": 33079147.0,
      "step": 16330
    },
    {
      "entropy": 5.897152996063232,
      "epoch": 1.4275102682862886,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004800085064800255,
      "loss": 5.3399,
      "mean_token_accuracy": 0.1878763496875763,
      "num_tokens": 33089229.0,
      "step": 16335
    },
    {
      "entropy": 5.935441303253174,
      "epoch": 1.4279472166389933,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004799956637347946,
      "loss": 5.3657,
      "mean_token_accuracy": 0.19219464659690857,
      "num_tokens": 33099688.0,
      "step": 16340
    },
    {
      "entropy": 5.9365355491638185,
      "epoch": 1.428384164991698,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004799828170576395,
      "loss": 5.3949,
      "mean_token_accuracy": 0.19047276079654693,
      "num_tokens": 33110019.0,
      "step": 16345
    },
    {
      "entropy": 5.816471338272095,
      "epoch": 1.4288211133444026,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004799699664488065,
      "loss": 5.2757,
      "mean_token_accuracy": 0.20463430285453796,
      "num_tokens": 33120141.0,
      "step": 16350
    },
    {
      "entropy": 5.919538974761963,
      "epoch": 1.4292580616971073,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004799571119085422,
      "loss": 5.3734,
      "mean_token_accuracy": 0.19327963441610335,
      "num_tokens": 33130047.0,
      "step": 16355
    },
    {
      "entropy": 5.941774320602417,
      "epoch": 1.429695010049812,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004799442534370931,
      "loss": 5.406,
      "mean_token_accuracy": 0.18735291957855224,
      "num_tokens": 33141234.0,
      "step": 16360
    },
    {
      "entropy": 5.896886110305786,
      "epoch": 1.4301319584025167,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004799313910347057,
      "loss": 5.3995,
      "mean_token_accuracy": 0.19915665090084075,
      "num_tokens": 33150788.0,
      "step": 16365
    },
    {
      "entropy": 5.890523290634155,
      "epoch": 1.4305689067552216,
      "grad_norm": 0.984375,
      "learning_rate": 0.000479918524701627,
      "loss": 5.371,
      "mean_token_accuracy": 0.19702991247177123,
      "num_tokens": 33160928.0,
      "step": 16370
    },
    {
      "entropy": 6.013475036621093,
      "epoch": 1.431005855107926,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047990565443810357,
      "loss": 5.4685,
      "mean_token_accuracy": 0.18067269921302795,
      "num_tokens": 33171333.0,
      "step": 16375
    },
    {
      "entropy": 5.8375959396362305,
      "epoch": 1.431442803460631,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047989278024438233,
      "loss": 5.264,
      "mean_token_accuracy": 0.20181096196174622,
      "num_tokens": 33180396.0,
      "step": 16380
    },
    {
      "entropy": 5.901287412643432,
      "epoch": 1.4318797518133357,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004798799021207102,
      "loss": 5.3633,
      "mean_token_accuracy": 0.19351983517408372,
      "num_tokens": 33189571.0,
      "step": 16385
    },
    {
      "entropy": 5.914506292343139,
      "epoch": 1.4323167001660404,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047986702006733417,
      "loss": 5.3278,
      "mean_token_accuracy": 0.19347851276397704,
      "num_tokens": 33198979.0,
      "step": 16390
    },
    {
      "entropy": 5.903127813339234,
      "epoch": 1.432753648518745,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047985413408450136,
      "loss": 5.3866,
      "mean_token_accuracy": 0.19590976983308792,
      "num_tokens": 33208862.0,
      "step": 16395
    },
    {
      "entropy": 5.888301944732666,
      "epoch": 1.4331905968714498,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004798412441724589,
      "loss": 5.3984,
      "mean_token_accuracy": 0.18671912997961043,
      "num_tokens": 33218804.0,
      "step": 16400
    },
    {
      "entropy": 5.948255968093872,
      "epoch": 1.4336275452241545,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047982835033145413,
      "loss": 5.4061,
      "mean_token_accuracy": 0.19083763509988785,
      "num_tokens": 33228783.0,
      "step": 16405
    },
    {
      "entropy": 6.010579347610474,
      "epoch": 1.4340644935768592,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047981545256173415,
      "loss": 5.441,
      "mean_token_accuracy": 0.18903158307075502,
      "num_tokens": 33241029.0,
      "step": 16410
    },
    {
      "entropy": 5.966364240646362,
      "epoch": 1.4345014419295639,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004798025508635465,
      "loss": 5.446,
      "mean_token_accuracy": 0.1886860564351082,
      "num_tokens": 33250383.0,
      "step": 16415
    },
    {
      "entropy": 5.838119411468506,
      "epoch": 1.4349383902822685,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004797896452371386,
      "loss": 5.2635,
      "mean_token_accuracy": 0.20365190356969834,
      "num_tokens": 33260424.0,
      "step": 16420
    },
    {
      "entropy": 5.950293636322021,
      "epoch": 1.4353753386349735,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047977673568275805,
      "loss": 5.3339,
      "mean_token_accuracy": 0.18798238933086395,
      "num_tokens": 33270200.0,
      "step": 16425
    },
    {
      "entropy": 5.974686479568481,
      "epoch": 1.435812286987678,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004797638222006524,
      "loss": 5.5067,
      "mean_token_accuracy": 0.18515574038028718,
      "num_tokens": 33279850.0,
      "step": 16430
    },
    {
      "entropy": 5.911031484603882,
      "epoch": 1.4362492353403828,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004797509047910693,
      "loss": 5.3389,
      "mean_token_accuracy": 0.19940691888332368,
      "num_tokens": 33289053.0,
      "step": 16435
    },
    {
      "entropy": 5.927047348022461,
      "epoch": 1.4366861836930875,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047973798345425655,
      "loss": 5.3789,
      "mean_token_accuracy": 0.19348515272140504,
      "num_tokens": 33298988.0,
      "step": 16440
    },
    {
      "entropy": 5.9274373054504395,
      "epoch": 1.4371231320457922,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047972505819046207,
      "loss": 5.3964,
      "mean_token_accuracy": 0.19789183884859085,
      "num_tokens": 33308683.0,
      "step": 16445
    },
    {
      "entropy": 5.890329122543335,
      "epoch": 1.437560080398497,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004797121289999336,
      "loss": 5.3621,
      "mean_token_accuracy": 0.18902698308229446,
      "num_tokens": 33319230.0,
      "step": 16450
    },
    {
      "entropy": 5.962348365783692,
      "epoch": 1.4379970287512016,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004796991958829193,
      "loss": 5.4017,
      "mean_token_accuracy": 0.18928273916244506,
      "num_tokens": 33329277.0,
      "step": 16455
    },
    {
      "entropy": 5.962646436691284,
      "epoch": 1.4384339771039063,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004796862588396672,
      "loss": 5.3734,
      "mean_token_accuracy": 0.19195094406604768,
      "num_tokens": 33338644.0,
      "step": 16460
    },
    {
      "entropy": 5.87282886505127,
      "epoch": 1.438870925456611,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047967331787042536,
      "loss": 5.2954,
      "mean_token_accuracy": 0.20512934923171997,
      "num_tokens": 33348996.0,
      "step": 16465
    },
    {
      "entropy": 5.918225812911987,
      "epoch": 1.4393078738093157,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047966037297544195,
      "loss": 5.4043,
      "mean_token_accuracy": 0.18746281266212464,
      "num_tokens": 33358581.0,
      "step": 16470
    },
    {
      "entropy": 5.956439971923828,
      "epoch": 1.4397448221620204,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047964742415496553,
      "loss": 5.4147,
      "mean_token_accuracy": 0.19281542748212815,
      "num_tokens": 33368925.0,
      "step": 16475
    },
    {
      "entropy": 5.931121873855591,
      "epoch": 1.4401817705147253,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004796344714092441,
      "loss": 5.4095,
      "mean_token_accuracy": 0.19207444936037063,
      "num_tokens": 33378918.0,
      "step": 16480
    },
    {
      "entropy": 5.908322143554687,
      "epoch": 1.4406187188674298,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004796215147385264,
      "loss": 5.3815,
      "mean_token_accuracy": 0.19624494165182113,
      "num_tokens": 33389039.0,
      "step": 16485
    },
    {
      "entropy": 5.934713459014892,
      "epoch": 1.4410556672201347,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004796085541430608,
      "loss": 5.3381,
      "mean_token_accuracy": 0.19729611426591873,
      "num_tokens": 33398893.0,
      "step": 16490
    },
    {
      "entropy": 5.978250455856323,
      "epoch": 1.4414926155728394,
      "grad_norm": 1.125,
      "learning_rate": 0.0004795955896230959,
      "loss": 5.4314,
      "mean_token_accuracy": 0.18476485311985016,
      "num_tokens": 33408188.0,
      "step": 16495
    },
    {
      "entropy": 5.9172686576843265,
      "epoch": 1.441929563925544,
      "grad_norm": 1.0,
      "learning_rate": 0.00047958262117888035,
      "loss": 5.3492,
      "mean_token_accuracy": 0.1991400822997093,
      "num_tokens": 33418926.0,
      "step": 16500
    },
    {
      "entropy": 5.949209403991699,
      "epoch": 1.4423665122782487,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047956964881066297,
      "loss": 5.4035,
      "mean_token_accuracy": 0.19406010955572128,
      "num_tokens": 33429790.0,
      "step": 16505
    },
    {
      "entropy": 5.874335908889771,
      "epoch": 1.4428034606309534,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004795566725186925,
      "loss": 5.3962,
      "mean_token_accuracy": 0.19424377083778382,
      "num_tokens": 33440470.0,
      "step": 16510
    },
    {
      "entropy": 5.951537942886352,
      "epoch": 1.4432404089836581,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004795436923032179,
      "loss": 5.3854,
      "mean_token_accuracy": 0.19475654661655425,
      "num_tokens": 33449875.0,
      "step": 16515
    },
    {
      "entropy": 5.991564035415649,
      "epoch": 1.4436773573363628,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004795307081644881,
      "loss": 5.3666,
      "mean_token_accuracy": 0.18781429529190063,
      "num_tokens": 33458841.0,
      "step": 16520
    },
    {
      "entropy": 5.924037981033325,
      "epoch": 1.4441143056890675,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004795177201027521,
      "loss": 5.3239,
      "mean_token_accuracy": 0.19647502452135085,
      "num_tokens": 33468656.0,
      "step": 16525
    },
    {
      "entropy": 5.903465366363525,
      "epoch": 1.4445512540417722,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047950472811825916,
      "loss": 5.3766,
      "mean_token_accuracy": 0.1932348906993866,
      "num_tokens": 33478977.0,
      "step": 16530
    },
    {
      "entropy": 5.953530120849609,
      "epoch": 1.444988202394477,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004794917322112583,
      "loss": 5.3407,
      "mean_token_accuracy": 0.19245980232954024,
      "num_tokens": 33488418.0,
      "step": 16535
    },
    {
      "entropy": 5.868187713623047,
      "epoch": 1.4454251507471816,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004794787323819989,
      "loss": 5.3206,
      "mean_token_accuracy": 0.19788368195295333,
      "num_tokens": 33499125.0,
      "step": 16540
    },
    {
      "entropy": 5.948830938339233,
      "epoch": 1.4458620990998865,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047946572863073023,
      "loss": 5.3797,
      "mean_token_accuracy": 0.1890175998210907,
      "num_tokens": 33508826.0,
      "step": 16545
    },
    {
      "entropy": 5.879712295532227,
      "epoch": 1.446299047452591,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047945272095770174,
      "loss": 5.2881,
      "mean_token_accuracy": 0.2040787935256958,
      "num_tokens": 33518398.0,
      "step": 16550
    },
    {
      "entropy": 5.893000841140747,
      "epoch": 1.4467359958052959,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000479439709363163,
      "loss": 5.3607,
      "mean_token_accuracy": 0.19072179943323136,
      "num_tokens": 33528289.0,
      "step": 16555
    },
    {
      "entropy": 5.956651258468628,
      "epoch": 1.4471729441580006,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004794266938473635,
      "loss": 5.4617,
      "mean_token_accuracy": 0.18273449391126634,
      "num_tokens": 33539228.0,
      "step": 16560
    },
    {
      "entropy": 5.967517471313476,
      "epoch": 1.4476098925107053,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047941367441055283,
      "loss": 5.4456,
      "mean_token_accuracy": 0.1923265352845192,
      "num_tokens": 33551015.0,
      "step": 16565
    },
    {
      "entropy": 5.926549768447876,
      "epoch": 1.44804684086341,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047940065105298073,
      "loss": 5.3758,
      "mean_token_accuracy": 0.1925606459379196,
      "num_tokens": 33560374.0,
      "step": 16570
    },
    {
      "entropy": 5.828077077865601,
      "epoch": 1.4484837892161146,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004793876237748972,
      "loss": 5.2664,
      "mean_token_accuracy": 0.2103278174996376,
      "num_tokens": 33569916.0,
      "step": 16575
    },
    {
      "entropy": 5.835581922531128,
      "epoch": 1.4489207375688193,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004793745925765518,
      "loss": 5.2823,
      "mean_token_accuracy": 0.2031272828578949,
      "num_tokens": 33579930.0,
      "step": 16580
    },
    {
      "entropy": 5.864585304260254,
      "epoch": 1.449357685921524,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004793615574581947,
      "loss": 5.3944,
      "mean_token_accuracy": 0.19269582033157348,
      "num_tokens": 33588612.0,
      "step": 16585
    },
    {
      "entropy": 5.98237247467041,
      "epoch": 1.4497946342742287,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047934851842007585,
      "loss": 5.4306,
      "mean_token_accuracy": 0.1815986305475235,
      "num_tokens": 33599253.0,
      "step": 16590
    },
    {
      "entropy": 5.961969947814941,
      "epoch": 1.4502315826269334,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004793354754624453,
      "loss": 5.304,
      "mean_token_accuracy": 0.19963324666023255,
      "num_tokens": 33609484.0,
      "step": 16595
    },
    {
      "entropy": 5.9463170051574705,
      "epoch": 1.4506685309796383,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004793224285855533,
      "loss": 5.3211,
      "mean_token_accuracy": 0.1961975187063217,
      "num_tokens": 33619660.0,
      "step": 16600
    },
    {
      "entropy": 5.861484956741333,
      "epoch": 1.4511054793323428,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047930937778965,
      "loss": 5.4197,
      "mean_token_accuracy": 0.188763627409935,
      "num_tokens": 33629505.0,
      "step": 16605
    },
    {
      "entropy": 5.90167555809021,
      "epoch": 1.4515424276850477,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004792963230749858,
      "loss": 5.4347,
      "mean_token_accuracy": 0.18968316465616225,
      "num_tokens": 33639680.0,
      "step": 16610
    },
    {
      "entropy": 5.968433094024658,
      "epoch": 1.4519793760377524,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047928326444181117,
      "loss": 5.4338,
      "mean_token_accuracy": 0.1863484725356102,
      "num_tokens": 33649200.0,
      "step": 16615
    },
    {
      "entropy": 5.939980411529541,
      "epoch": 1.452416324390457,
      "grad_norm": 1.125,
      "learning_rate": 0.00047927020189037634,
      "loss": 5.3703,
      "mean_token_accuracy": 0.18805228024721146,
      "num_tokens": 33658386.0,
      "step": 16620
    },
    {
      "entropy": 5.913007545471191,
      "epoch": 1.4528532727431618,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004792571354209321,
      "loss": 5.3381,
      "mean_token_accuracy": 0.19636524468660355,
      "num_tokens": 33669139.0,
      "step": 16625
    },
    {
      "entropy": 5.901395702362061,
      "epoch": 1.4532902210958665,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004792440650337289,
      "loss": 5.4122,
      "mean_token_accuracy": 0.1945533946156502,
      "num_tokens": 33679629.0,
      "step": 16630
    },
    {
      "entropy": 5.9825526714324955,
      "epoch": 1.4537271694485712,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004792309907290175,
      "loss": 5.4167,
      "mean_token_accuracy": 0.187764872610569,
      "num_tokens": 33689633.0,
      "step": 16635
    },
    {
      "entropy": 5.979506492614746,
      "epoch": 1.4541641178012759,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047921791250704875,
      "loss": 5.3762,
      "mean_token_accuracy": 0.18703658133745193,
      "num_tokens": 33699719.0,
      "step": 16640
    },
    {
      "entropy": 5.967195510864258,
      "epoch": 1.4546010661539805,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047920483036807346,
      "loss": 5.43,
      "mean_token_accuracy": 0.19106595814228058,
      "num_tokens": 33709389.0,
      "step": 16645
    },
    {
      "entropy": 5.941090822219849,
      "epoch": 1.4550380145066852,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047919174431234246,
      "loss": 5.4123,
      "mean_token_accuracy": 0.18370047807693482,
      "num_tokens": 33719063.0,
      "step": 16650
    },
    {
      "entropy": 6.010422134399414,
      "epoch": 1.4554749628593902,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004791786543401068,
      "loss": 5.4958,
      "mean_token_accuracy": 0.17961474061012267,
      "num_tokens": 33729798.0,
      "step": 16655
    },
    {
      "entropy": 5.922529840469361,
      "epoch": 1.4559119112120946,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004791655604516176,
      "loss": 5.3872,
      "mean_token_accuracy": 0.18207787424325944,
      "num_tokens": 33740012.0,
      "step": 16660
    },
    {
      "entropy": 5.924761343002319,
      "epoch": 1.4563488595647995,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047915246264712595,
      "loss": 5.3576,
      "mean_token_accuracy": 0.19462848603725433,
      "num_tokens": 33749478.0,
      "step": 16665
    },
    {
      "entropy": 5.811961698532104,
      "epoch": 1.4567858079175042,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004791393609268831,
      "loss": 5.2439,
      "mean_token_accuracy": 0.20810047686100006,
      "num_tokens": 33758935.0,
      "step": 16670
    },
    {
      "entropy": 5.8739457607269285,
      "epoch": 1.457222756270209,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004791262552911404,
      "loss": 5.3221,
      "mean_token_accuracy": 0.19595580846071242,
      "num_tokens": 33769035.0,
      "step": 16675
    },
    {
      "entropy": 5.909000587463379,
      "epoch": 1.4576597046229136,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047911314574014904,
      "loss": 5.3612,
      "mean_token_accuracy": 0.19351799488067628,
      "num_tokens": 33779288.0,
      "step": 16680
    },
    {
      "entropy": 5.940864324569702,
      "epoch": 1.4580966529756183,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004791000322741606,
      "loss": 5.3735,
      "mean_token_accuracy": 0.1909196048974991,
      "num_tokens": 33789138.0,
      "step": 16685
    },
    {
      "entropy": 5.969956731796264,
      "epoch": 1.458533601328323,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004790869148934267,
      "loss": 5.4013,
      "mean_token_accuracy": 0.18877887725830078,
      "num_tokens": 33798649.0,
      "step": 16690
    },
    {
      "entropy": 5.8169389247894285,
      "epoch": 1.4589705496810277,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047907379359819877,
      "loss": 5.2768,
      "mean_token_accuracy": 0.20962828844785691,
      "num_tokens": 33808742.0,
      "step": 16695
    },
    {
      "entropy": 5.861245250701904,
      "epoch": 1.4594074980337324,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004790606683887286,
      "loss": 5.3186,
      "mean_token_accuracy": 0.1982771039009094,
      "num_tokens": 33819507.0,
      "step": 16700
    },
    {
      "entropy": 5.955374097824096,
      "epoch": 1.459844446386437,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047904753926526777,
      "loss": 5.3855,
      "mean_token_accuracy": 0.19107490479946138,
      "num_tokens": 33829722.0,
      "step": 16705
    },
    {
      "entropy": 5.936424779891968,
      "epoch": 1.4602813947391418,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004790344062280683,
      "loss": 5.4308,
      "mean_token_accuracy": 0.18706141263246537,
      "num_tokens": 33839350.0,
      "step": 16710
    },
    {
      "entropy": 5.90481743812561,
      "epoch": 1.4607183430918464,
      "grad_norm": 1.125,
      "learning_rate": 0.00047902126927738196,
      "loss": 5.3552,
      "mean_token_accuracy": 0.19351010471582414,
      "num_tokens": 33849233.0,
      "step": 16715
    },
    {
      "entropy": 5.952498292922973,
      "epoch": 1.4611552914445514,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047900812841346086,
      "loss": 5.4076,
      "mean_token_accuracy": 0.19245013892650603,
      "num_tokens": 33859996.0,
      "step": 16720
    },
    {
      "entropy": 5.960275793075562,
      "epoch": 1.4615922397972558,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004789949836365569,
      "loss": 5.4156,
      "mean_token_accuracy": 0.19620639383792876,
      "num_tokens": 33870780.0,
      "step": 16725
    },
    {
      "entropy": 5.940114688873291,
      "epoch": 1.4620291881499607,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004789818349469222,
      "loss": 5.4092,
      "mean_token_accuracy": 0.19113557636737824,
      "num_tokens": 33880182.0,
      "step": 16730
    },
    {
      "entropy": 5.934760475158692,
      "epoch": 1.4624661365026654,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004789686823448091,
      "loss": 5.4136,
      "mean_token_accuracy": 0.18441389352083207,
      "num_tokens": 33889464.0,
      "step": 16735
    },
    {
      "entropy": 5.908947849273682,
      "epoch": 1.4629030848553701,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004789555258304698,
      "loss": 5.3676,
      "mean_token_accuracy": 0.1884254604578018,
      "num_tokens": 33899530.0,
      "step": 16740
    },
    {
      "entropy": 5.940669679641724,
      "epoch": 1.4633400332080748,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004789423654041566,
      "loss": 5.4114,
      "mean_token_accuracy": 0.1921267956495285,
      "num_tokens": 33910094.0,
      "step": 16745
    },
    {
      "entropy": 5.963872051239013,
      "epoch": 1.4637769815607795,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047892920106612204,
      "loss": 5.3339,
      "mean_token_accuracy": 0.19431649148464203,
      "num_tokens": 33921181.0,
      "step": 16750
    },
    {
      "entropy": 5.914497756958008,
      "epoch": 1.4642139299134842,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004789160328166184,
      "loss": 5.4811,
      "mean_token_accuracy": 0.18305710852146148,
      "num_tokens": 33931567.0,
      "step": 16755
    },
    {
      "entropy": 5.937298345565796,
      "epoch": 1.464650878266189,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047890286065589856,
      "loss": 5.4182,
      "mean_token_accuracy": 0.1914092034101486,
      "num_tokens": 33941513.0,
      "step": 16760
    },
    {
      "entropy": 5.9022399425506595,
      "epoch": 1.4650878266188936,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047888968458421487,
      "loss": 5.3457,
      "mean_token_accuracy": 0.19582749307155609,
      "num_tokens": 33953400.0,
      "step": 16765
    },
    {
      "entropy": 5.95798807144165,
      "epoch": 1.4655247749715983,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004788765046018204,
      "loss": 5.4294,
      "mean_token_accuracy": 0.18544412404298782,
      "num_tokens": 33964394.0,
      "step": 16770
    },
    {
      "entropy": 5.988860082626343,
      "epoch": 1.4659617233243032,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047886332070896747,
      "loss": 5.5193,
      "mean_token_accuracy": 0.18202537149190903,
      "num_tokens": 33975266.0,
      "step": 16775
    },
    {
      "entropy": 5.981400442123413,
      "epoch": 1.4663986716770077,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004788501329059094,
      "loss": 5.4674,
      "mean_token_accuracy": 0.18373313546180725,
      "num_tokens": 33986159.0,
      "step": 16780
    },
    {
      "entropy": 5.948021745681762,
      "epoch": 1.4668356200297126,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047883694119289896,
      "loss": 5.406,
      "mean_token_accuracy": 0.18902237713336945,
      "num_tokens": 33997792.0,
      "step": 16785
    },
    {
      "entropy": 5.928107738494873,
      "epoch": 1.4672725683824173,
      "grad_norm": 1.0,
      "learning_rate": 0.00047882374557018916,
      "loss": 5.3973,
      "mean_token_accuracy": 0.1929691269993782,
      "num_tokens": 34007945.0,
      "step": 16790
    },
    {
      "entropy": 5.976398038864136,
      "epoch": 1.467709516735122,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047881054603803307,
      "loss": 5.4597,
      "mean_token_accuracy": 0.19202382117509842,
      "num_tokens": 34018399.0,
      "step": 16795
    },
    {
      "entropy": 5.970090532302857,
      "epoch": 1.4681464650878266,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004787973425966839,
      "loss": 5.4887,
      "mean_token_accuracy": 0.1841201364994049,
      "num_tokens": 34028267.0,
      "step": 16800
    },
    {
      "entropy": 5.9967827796936035,
      "epoch": 1.4685834134405313,
      "grad_norm": 1.09375,
      "learning_rate": 0.000478784135246395,
      "loss": 5.5066,
      "mean_token_accuracy": 0.1798946425318718,
      "num_tokens": 34037904.0,
      "step": 16805
    },
    {
      "entropy": 5.873777103424072,
      "epoch": 1.469020361793236,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004787709239874196,
      "loss": 5.3425,
      "mean_token_accuracy": 0.19733866751194,
      "num_tokens": 34049203.0,
      "step": 16810
    },
    {
      "entropy": 5.935450601577759,
      "epoch": 1.4694573101459407,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000478757708820011,
      "loss": 5.4323,
      "mean_token_accuracy": 0.19147325158119202,
      "num_tokens": 34059670.0,
      "step": 16815
    },
    {
      "entropy": 5.931893634796142,
      "epoch": 1.4698942584986454,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004787444897444229,
      "loss": 5.3445,
      "mean_token_accuracy": 0.19922071397304536,
      "num_tokens": 34069730.0,
      "step": 16820
    },
    {
      "entropy": 5.97949538230896,
      "epoch": 1.47033120685135,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004787312667609086,
      "loss": 5.3675,
      "mean_token_accuracy": 0.19391518235206603,
      "num_tokens": 34079315.0,
      "step": 16825
    },
    {
      "entropy": 5.938223886489868,
      "epoch": 1.470768155204055,
      "grad_norm": 1.015625,
      "learning_rate": 0.000478718039869722,
      "loss": 5.3759,
      "mean_token_accuracy": 0.19052087366580964,
      "num_tokens": 34089431.0,
      "step": 16830
    },
    {
      "entropy": 5.967315816879273,
      "epoch": 1.4712051035567595,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047870480907111657,
      "loss": 5.337,
      "mean_token_accuracy": 0.19613843262195588,
      "num_tokens": 34100120.0,
      "step": 16835
    },
    {
      "entropy": 5.942030763626098,
      "epoch": 1.4716420519094644,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004786915743653461,
      "loss": 5.3775,
      "mean_token_accuracy": 0.19505628049373627,
      "num_tokens": 34110271.0,
      "step": 16840
    },
    {
      "entropy": 5.95726752281189,
      "epoch": 1.472079000262169,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047867833575266456,
      "loss": 5.4467,
      "mean_token_accuracy": 0.1906438335776329,
      "num_tokens": 34120240.0,
      "step": 16845
    },
    {
      "entropy": 5.953580570220947,
      "epoch": 1.4725159486148738,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004786650932333259,
      "loss": 5.3921,
      "mean_token_accuracy": 0.19524959623813629,
      "num_tokens": 34130813.0,
      "step": 16850
    },
    {
      "entropy": 5.930655813217163,
      "epoch": 1.4729528969675785,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004786518468075839,
      "loss": 5.4406,
      "mean_token_accuracy": 0.19372042417526245,
      "num_tokens": 34141169.0,
      "step": 16855
    },
    {
      "entropy": 5.90782356262207,
      "epoch": 1.4733898453202832,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004786385964756928,
      "loss": 5.3903,
      "mean_token_accuracy": 0.19269654601812364,
      "num_tokens": 34151869.0,
      "step": 16860
    },
    {
      "entropy": 5.992818403244018,
      "epoch": 1.4738267936729879,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004786253422379068,
      "loss": 5.493,
      "mean_token_accuracy": 0.1814376801252365,
      "num_tokens": 34162608.0,
      "step": 16865
    },
    {
      "entropy": 5.926175832748413,
      "epoch": 1.4742637420256925,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047861208409448,
      "loss": 5.4101,
      "mean_token_accuracy": 0.18960846364498138,
      "num_tokens": 34172648.0,
      "step": 16870
    },
    {
      "entropy": 5.957403659820557,
      "epoch": 1.4747006903783972,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004785988220456666,
      "loss": 5.3975,
      "mean_token_accuracy": 0.19460989385843278,
      "num_tokens": 34183603.0,
      "step": 16875
    },
    {
      "entropy": 5.976054573059082,
      "epoch": 1.475137638731102,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004785855560917213,
      "loss": 5.4374,
      "mean_token_accuracy": 0.1843220204114914,
      "num_tokens": 34193830.0,
      "step": 16880
    },
    {
      "entropy": 5.886058044433594,
      "epoch": 1.4755745870838066,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047857228623289825,
      "loss": 5.2633,
      "mean_token_accuracy": 0.2090531513094902,
      "num_tokens": 34203074.0,
      "step": 16885
    },
    {
      "entropy": 5.876571083068848,
      "epoch": 1.4760115354365113,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047855901246945205,
      "loss": 5.4738,
      "mean_token_accuracy": 0.1826240226626396,
      "num_tokens": 34212958.0,
      "step": 16890
    },
    {
      "entropy": 5.912948131561279,
      "epoch": 1.4764484837892162,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004785457348016374,
      "loss": 5.4224,
      "mean_token_accuracy": 0.18457893580198287,
      "num_tokens": 34223077.0,
      "step": 16895
    },
    {
      "entropy": 5.969263935089112,
      "epoch": 1.4768854321419207,
      "grad_norm": 1.0,
      "learning_rate": 0.00047853245322970885,
      "loss": 5.3872,
      "mean_token_accuracy": 0.20474308580160142,
      "num_tokens": 34233386.0,
      "step": 16900
    },
    {
      "entropy": 5.967006826400757,
      "epoch": 1.4773223804946256,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004785191677539212,
      "loss": 5.4052,
      "mean_token_accuracy": 0.19328354001045228,
      "num_tokens": 34244488.0,
      "step": 16905
    },
    {
      "entropy": 5.932929944992066,
      "epoch": 1.4777593288473303,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047850587837452936,
      "loss": 5.4097,
      "mean_token_accuracy": 0.18604549914598464,
      "num_tokens": 34254047.0,
      "step": 16910
    },
    {
      "entropy": 5.9183652877807615,
      "epoch": 1.478196277200035,
      "grad_norm": 1.0,
      "learning_rate": 0.0004784925850917881,
      "loss": 5.3916,
      "mean_token_accuracy": 0.1836699590086937,
      "num_tokens": 34264297.0,
      "step": 16915
    },
    {
      "entropy": 6.000126123428345,
      "epoch": 1.4786332255527397,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004784792879059524,
      "loss": 5.4533,
      "mean_token_accuracy": 0.18663615137338638,
      "num_tokens": 34274945.0,
      "step": 16920
    },
    {
      "entropy": 5.9227471351623535,
      "epoch": 1.4790701739054444,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047846598681727733,
      "loss": 5.311,
      "mean_token_accuracy": 0.189407517015934,
      "num_tokens": 34284914.0,
      "step": 16925
    },
    {
      "entropy": 5.914976930618286,
      "epoch": 1.479507122258149,
      "grad_norm": 1.125,
      "learning_rate": 0.000478452681826018,
      "loss": 5.3815,
      "mean_token_accuracy": 0.19138140231370926,
      "num_tokens": 34294277.0,
      "step": 16930
    },
    {
      "entropy": 5.958480501174927,
      "epoch": 1.4799440706108538,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047843937293242967,
      "loss": 5.419,
      "mean_token_accuracy": 0.18898237347602845,
      "num_tokens": 34304865.0,
      "step": 16935
    },
    {
      "entropy": 5.793494462966919,
      "epoch": 1.4803810189635584,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047842606013676753,
      "loss": 5.265,
      "mean_token_accuracy": 0.20624737739562987,
      "num_tokens": 34315404.0,
      "step": 16940
    },
    {
      "entropy": 6.015575408935547,
      "epoch": 1.4808179673162631,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047841274343928697,
      "loss": 5.3685,
      "mean_token_accuracy": 0.18702270984649658,
      "num_tokens": 34325448.0,
      "step": 16945
    },
    {
      "entropy": 5.910754203796387,
      "epoch": 1.481254915668968,
      "grad_norm": 1.125,
      "learning_rate": 0.0004783994228402434,
      "loss": 5.3823,
      "mean_token_accuracy": 0.20188578367233276,
      "num_tokens": 34335266.0,
      "step": 16950
    },
    {
      "entropy": 5.933721303939819,
      "epoch": 1.4816918640216725,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004783860983398924,
      "loss": 5.375,
      "mean_token_accuracy": 0.19482334107160568,
      "num_tokens": 34345847.0,
      "step": 16955
    },
    {
      "entropy": 5.926606130599976,
      "epoch": 1.4821288123743774,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004783727699384894,
      "loss": 5.3164,
      "mean_token_accuracy": 0.19397610276937485,
      "num_tokens": 34355718.0,
      "step": 16960
    },
    {
      "entropy": 5.888956546783447,
      "epoch": 1.4825657607270821,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004783594376362901,
      "loss": 5.4049,
      "mean_token_accuracy": 0.1935230702161789,
      "num_tokens": 34366100.0,
      "step": 16965
    },
    {
      "entropy": 5.970944404602051,
      "epoch": 1.4830027090797868,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004783461014335502,
      "loss": 5.4607,
      "mean_token_accuracy": 0.1856511726975441,
      "num_tokens": 34376977.0,
      "step": 16970
    },
    {
      "entropy": 6.001266860961914,
      "epoch": 1.4834396574324915,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004783327613305256,
      "loss": 5.4167,
      "mean_token_accuracy": 0.1878763496875763,
      "num_tokens": 34387782.0,
      "step": 16975
    },
    {
      "entropy": 5.92032470703125,
      "epoch": 1.4838766057851962,
      "grad_norm": 1.015625,
      "learning_rate": 0.000478319417327472,
      "loss": 5.3581,
      "mean_token_accuracy": 0.1938147008419037,
      "num_tokens": 34397321.0,
      "step": 16980
    },
    {
      "entropy": 5.925790596008301,
      "epoch": 1.484313554137901,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004783060694246455,
      "loss": 5.4223,
      "mean_token_accuracy": 0.1853831246495247,
      "num_tokens": 34407594.0,
      "step": 16985
    },
    {
      "entropy": 5.989916706085205,
      "epoch": 1.4847505024906056,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004782927176223021,
      "loss": 5.4069,
      "mean_token_accuracy": 0.18808092772960663,
      "num_tokens": 34417337.0,
      "step": 16990
    },
    {
      "entropy": 5.899311542510986,
      "epoch": 1.4851874508433103,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004782793619206977,
      "loss": 5.2737,
      "mean_token_accuracy": 0.1998074993491173,
      "num_tokens": 34427643.0,
      "step": 16995
    },
    {
      "entropy": 5.909807872772217,
      "epoch": 1.485624399196015,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004782660023200887,
      "loss": 5.4133,
      "mean_token_accuracy": 0.19217527955770491,
      "num_tokens": 34437224.0,
      "step": 17000
    },
    {
      "entropy": 5.99349455833435,
      "epoch": 1.4860613475487199,
      "grad_norm": 1.0,
      "learning_rate": 0.0004782526388207313,
      "loss": 5.408,
      "mean_token_accuracy": 0.1883452594280243,
      "num_tokens": 34447376.0,
      "step": 17005
    },
    {
      "entropy": 5.918891334533692,
      "epoch": 1.4864982959014243,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004782392714228818,
      "loss": 5.4352,
      "mean_token_accuracy": 0.19151339530944825,
      "num_tokens": 34457649.0,
      "step": 17010
    },
    {
      "entropy": 5.952408218383789,
      "epoch": 1.4869352442541293,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047822590012679654,
      "loss": 5.4169,
      "mean_token_accuracy": 0.19594855159521102,
      "num_tokens": 34467437.0,
      "step": 17015
    },
    {
      "entropy": 5.965356302261353,
      "epoch": 1.487372192606834,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047821252493273204,
      "loss": 5.3436,
      "mean_token_accuracy": 0.19102644622325898,
      "num_tokens": 34477615.0,
      "step": 17020
    },
    {
      "entropy": 5.923635721206665,
      "epoch": 1.4878091409595386,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004781991458409448,
      "loss": 5.3623,
      "mean_token_accuracy": 0.19544933140277862,
      "num_tokens": 34487619.0,
      "step": 17025
    },
    {
      "entropy": 5.920355558395386,
      "epoch": 1.4882460893122433,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047818576285169154,
      "loss": 5.3749,
      "mean_token_accuracy": 0.19765003770589828,
      "num_tokens": 34498837.0,
      "step": 17030
    },
    {
      "entropy": 5.906600427627564,
      "epoch": 1.488683037664948,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047817237596522885,
      "loss": 5.3877,
      "mean_token_accuracy": 0.19382711201906205,
      "num_tokens": 34508972.0,
      "step": 17035
    },
    {
      "entropy": 5.968051195144653,
      "epoch": 1.4891199860176527,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004781589851818135,
      "loss": 5.3859,
      "mean_token_accuracy": 0.18821389675140382,
      "num_tokens": 34519289.0,
      "step": 17040
    },
    {
      "entropy": 5.92430567741394,
      "epoch": 1.4895569343703574,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047814559050170236,
      "loss": 5.3306,
      "mean_token_accuracy": 0.19727715849876404,
      "num_tokens": 34528384.0,
      "step": 17045
    },
    {
      "entropy": 5.883686065673828,
      "epoch": 1.489993882723062,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047813219192515235,
      "loss": 5.3693,
      "mean_token_accuracy": 0.18729361295700073,
      "num_tokens": 34537442.0,
      "step": 17050
    },
    {
      "entropy": 5.929759359359741,
      "epoch": 1.4904308310757668,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004781187894524205,
      "loss": 5.3844,
      "mean_token_accuracy": 0.19306399375200273,
      "num_tokens": 34547960.0,
      "step": 17055
    },
    {
      "entropy": 6.009260463714599,
      "epoch": 1.4908677794284715,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047810538308376376,
      "loss": 5.3354,
      "mean_token_accuracy": 0.18930962085723876,
      "num_tokens": 34556730.0,
      "step": 17060
    },
    {
      "entropy": 5.929559898376465,
      "epoch": 1.4913047277811762,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004780919728194394,
      "loss": 5.3762,
      "mean_token_accuracy": 0.19069135934114456,
      "num_tokens": 34566751.0,
      "step": 17065
    },
    {
      "entropy": 5.968905019760132,
      "epoch": 1.491741676133881,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047807855865970454,
      "loss": 5.3604,
      "mean_token_accuracy": 0.19766390472650527,
      "num_tokens": 34576417.0,
      "step": 17070
    },
    {
      "entropy": 5.980312871932983,
      "epoch": 1.4921786244865856,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004780651406048166,
      "loss": 5.4631,
      "mean_token_accuracy": 0.1901880532503128,
      "num_tokens": 34586976.0,
      "step": 17075
    },
    {
      "entropy": 5.949950218200684,
      "epoch": 1.4926155728392905,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004780517186550328,
      "loss": 5.3771,
      "mean_token_accuracy": 0.19293468743562697,
      "num_tokens": 34597374.0,
      "step": 17080
    },
    {
      "entropy": 5.8726012229919435,
      "epoch": 1.4930525211919952,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004780382928106107,
      "loss": 5.3245,
      "mean_token_accuracy": 0.20052800476551055,
      "num_tokens": 34607586.0,
      "step": 17085
    },
    {
      "entropy": 5.967091846466064,
      "epoch": 1.4934894695446999,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047802486307180767,
      "loss": 5.4222,
      "mean_token_accuracy": 0.19459164440631865,
      "num_tokens": 34617152.0,
      "step": 17090
    },
    {
      "entropy": 5.922168159484864,
      "epoch": 1.4939264178974045,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047801142943888135,
      "loss": 5.3386,
      "mean_token_accuracy": 0.20066145807504654,
      "num_tokens": 34627421.0,
      "step": 17095
    },
    {
      "entropy": 5.933954811096191,
      "epoch": 1.4943633662501092,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047799799191208955,
      "loss": 5.3515,
      "mean_token_accuracy": 0.19438849538564681,
      "num_tokens": 34637687.0,
      "step": 17100
    },
    {
      "entropy": 5.9456541538238525,
      "epoch": 1.494800314602814,
      "grad_norm": 1.0,
      "learning_rate": 0.0004779845504916898,
      "loss": 5.4292,
      "mean_token_accuracy": 0.1898817464709282,
      "num_tokens": 34647884.0,
      "step": 17105
    },
    {
      "entropy": 5.9336553573608395,
      "epoch": 1.4952372629555186,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004779711051779401,
      "loss": 5.3851,
      "mean_token_accuracy": 0.19225801527500153,
      "num_tokens": 34658658.0,
      "step": 17110
    },
    {
      "entropy": 5.931031036376953,
      "epoch": 1.4956742113082233,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004779576559710982,
      "loss": 5.3839,
      "mean_token_accuracy": 0.1937255084514618,
      "num_tokens": 34668728.0,
      "step": 17115
    },
    {
      "entropy": 5.987097358703613,
      "epoch": 1.496111159660928,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047794420287142203,
      "loss": 5.4121,
      "mean_token_accuracy": 0.1867624282836914,
      "num_tokens": 34678576.0,
      "step": 17120
    },
    {
      "entropy": 6.047802591323853,
      "epoch": 1.496548108013633,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047793074587916984,
      "loss": 5.4658,
      "mean_token_accuracy": 0.18730274289846421,
      "num_tokens": 34688920.0,
      "step": 17125
    },
    {
      "entropy": 5.9621009349823,
      "epoch": 1.4969850563663374,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047791728499459955,
      "loss": 5.3768,
      "mean_token_accuracy": 0.19595859050750733,
      "num_tokens": 34699348.0,
      "step": 17130
    },
    {
      "entropy": 5.855591249465943,
      "epoch": 1.4974220047190423,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004779038202179694,
      "loss": 5.3101,
      "mean_token_accuracy": 0.1965723916888237,
      "num_tokens": 34708991.0,
      "step": 17135
    },
    {
      "entropy": 5.874743127822876,
      "epoch": 1.497858953071747,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047789035154953776,
      "loss": 5.3871,
      "mean_token_accuracy": 0.1878236100077629,
      "num_tokens": 34720869.0,
      "step": 17140
    },
    {
      "entropy": 5.9690995693206785,
      "epoch": 1.4982959014244517,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004778768789895627,
      "loss": 5.3462,
      "mean_token_accuracy": 0.19245240986347198,
      "num_tokens": 34731025.0,
      "step": 17145
    },
    {
      "entropy": 5.972511577606201,
      "epoch": 1.4987328497771564,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004778634025383029,
      "loss": 5.4105,
      "mean_token_accuracy": 0.1889700099825859,
      "num_tokens": 34741846.0,
      "step": 17150
    },
    {
      "entropy": 5.847237014770508,
      "epoch": 1.499169798129861,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047784992219601673,
      "loss": 5.2913,
      "mean_token_accuracy": 0.20646344274282455,
      "num_tokens": 34753294.0,
      "step": 17155
    },
    {
      "entropy": 5.973375225067139,
      "epoch": 1.4996067464825658,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004778364379629627,
      "loss": 5.3813,
      "mean_token_accuracy": 0.19462862759828567,
      "num_tokens": 34763997.0,
      "step": 17160
    },
    {
      "entropy": 5.906386423110962,
      "epoch": 1.5000436948352704,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047782294983939955,
      "loss": 5.2849,
      "mean_token_accuracy": 0.19262845069169998,
      "num_tokens": 34772925.0,
      "step": 17165
    },
    {
      "entropy": 5.961965608596802,
      "epoch": 1.5004806431879751,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004778094578255859,
      "loss": 5.3948,
      "mean_token_accuracy": 0.19218206107616426,
      "num_tokens": 34782608.0,
      "step": 17170
    },
    {
      "entropy": 5.906649446487426,
      "epoch": 1.5009175915406798,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004777959619217807,
      "loss": 5.3755,
      "mean_token_accuracy": 0.19553013890981674,
      "num_tokens": 34792533.0,
      "step": 17175
    },
    {
      "entropy": 5.987194013595581,
      "epoch": 1.5013545398933847,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047778246212824263,
      "loss": 5.4598,
      "mean_token_accuracy": 0.18557877391576766,
      "num_tokens": 34803069.0,
      "step": 17180
    },
    {
      "entropy": 5.95816330909729,
      "epoch": 1.5017914882460892,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004777689584452306,
      "loss": 5.4224,
      "mean_token_accuracy": 0.19091023802757262,
      "num_tokens": 34811925.0,
      "step": 17185
    },
    {
      "entropy": 5.91795859336853,
      "epoch": 1.5022284365987941,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004777554508730038,
      "loss": 5.339,
      "mean_token_accuracy": 0.1948755145072937,
      "num_tokens": 34821511.0,
      "step": 17190
    },
    {
      "entropy": 5.950110054016113,
      "epoch": 1.5026653849514986,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004777419394118212,
      "loss": 5.4458,
      "mean_token_accuracy": 0.1863289698958397,
      "num_tokens": 34831683.0,
      "step": 17195
    },
    {
      "entropy": 5.94433331489563,
      "epoch": 1.5031023333042035,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004777284240619419,
      "loss": 5.3991,
      "mean_token_accuracy": 0.19183634966611862,
      "num_tokens": 34841578.0,
      "step": 17200
    },
    {
      "entropy": 5.951744031906128,
      "epoch": 1.5035392816569082,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004777149048236252,
      "loss": 5.3947,
      "mean_token_accuracy": 0.19033548086881638,
      "num_tokens": 34851959.0,
      "step": 17205
    },
    {
      "entropy": 5.904899263381958,
      "epoch": 1.503976230009613,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047770138169713047,
      "loss": 5.3798,
      "mean_token_accuracy": 0.19468510746955872,
      "num_tokens": 34862557.0,
      "step": 17210
    },
    {
      "entropy": 5.967948770523071,
      "epoch": 1.5044131783623176,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047768785468271704,
      "loss": 5.4253,
      "mean_token_accuracy": 0.18892432600259781,
      "num_tokens": 34872394.0,
      "step": 17215
    },
    {
      "entropy": 5.98168249130249,
      "epoch": 1.5048501267150223,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047767432378064426,
      "loss": 5.4112,
      "mean_token_accuracy": 0.19354839026927947,
      "num_tokens": 34881938.0,
      "step": 17220
    },
    {
      "entropy": 5.945557451248169,
      "epoch": 1.505287075067727,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047766078899117174,
      "loss": 5.3343,
      "mean_token_accuracy": 0.19963189512491225,
      "num_tokens": 34892769.0,
      "step": 17225
    },
    {
      "entropy": 6.013900089263916,
      "epoch": 1.5057240234204317,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047764725031455927,
      "loss": 5.4927,
      "mean_token_accuracy": 0.1814967766404152,
      "num_tokens": 34903643.0,
      "step": 17230
    },
    {
      "entropy": 5.988615226745606,
      "epoch": 1.5061609717731366,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004776337077510662,
      "loss": 5.3884,
      "mean_token_accuracy": 0.18400030434131623,
      "num_tokens": 34913558.0,
      "step": 17235
    },
    {
      "entropy": 5.953996276855468,
      "epoch": 1.506597920125841,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047762016130095246,
      "loss": 5.3643,
      "mean_token_accuracy": 0.19351110905408858,
      "num_tokens": 34923653.0,
      "step": 17240
    },
    {
      "entropy": 5.899397945404052,
      "epoch": 1.507034868478546,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00047760661096447784,
      "loss": 5.3972,
      "mean_token_accuracy": 0.19157352298498154,
      "num_tokens": 34933381.0,
      "step": 17245
    },
    {
      "entropy": 5.9515159130096436,
      "epoch": 1.5074718168312504,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004775930567419023,
      "loss": 5.3165,
      "mean_token_accuracy": 0.19489477276802064,
      "num_tokens": 34942486.0,
      "step": 17250
    },
    {
      "entropy": 5.970848226547242,
      "epoch": 1.5079087651839553,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004775794986334857,
      "loss": 5.4483,
      "mean_token_accuracy": 0.18931927233934404,
      "num_tokens": 34952291.0,
      "step": 17255
    },
    {
      "entropy": 5.966502714157104,
      "epoch": 1.5083457135366598,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004775659366394883,
      "loss": 5.4199,
      "mean_token_accuracy": 0.19071502685546876,
      "num_tokens": 34963842.0,
      "step": 17260
    },
    {
      "entropy": 5.927987289428711,
      "epoch": 1.5087826618893647,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047755237076016997,
      "loss": 5.3337,
      "mean_token_accuracy": 0.1942000836133957,
      "num_tokens": 34973057.0,
      "step": 17265
    },
    {
      "entropy": 6.073458051681518,
      "epoch": 1.5092196102420694,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004775388009957911,
      "loss": 5.5228,
      "mean_token_accuracy": 0.17763173580169678,
      "num_tokens": 34983257.0,
      "step": 17270
    },
    {
      "entropy": 6.023300457000732,
      "epoch": 1.509656558594774,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047752522734661193,
      "loss": 5.3423,
      "mean_token_accuracy": 0.19471496641635894,
      "num_tokens": 34993066.0,
      "step": 17275
    },
    {
      "entropy": 5.953414678573608,
      "epoch": 1.5100935069474788,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047751164981289264,
      "loss": 5.313,
      "mean_token_accuracy": 0.19284505844116212,
      "num_tokens": 35002830.0,
      "step": 17280
    },
    {
      "entropy": 5.886883640289307,
      "epoch": 1.5105304553001835,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004774980683948939,
      "loss": 5.3337,
      "mean_token_accuracy": 0.193736432492733,
      "num_tokens": 35011837.0,
      "step": 17285
    },
    {
      "entropy": 5.971287488937378,
      "epoch": 1.5109674036528882,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047748448309287606,
      "loss": 5.3616,
      "mean_token_accuracy": 0.19102526605129241,
      "num_tokens": 35021589.0,
      "step": 17290
    },
    {
      "entropy": 5.9919655323028564,
      "epoch": 1.5114043520055929,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047747089390709973,
      "loss": 5.4466,
      "mean_token_accuracy": 0.19087186455726624,
      "num_tokens": 35032207.0,
      "step": 17295
    },
    {
      "entropy": 5.976630115509034,
      "epoch": 1.5118413003582978,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004774573008378256,
      "loss": 5.4468,
      "mean_token_accuracy": 0.18153705149888993,
      "num_tokens": 35042661.0,
      "step": 17300
    },
    {
      "entropy": 5.958442687988281,
      "epoch": 1.5122782487110022,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004774437038853143,
      "loss": 5.3807,
      "mean_token_accuracy": 0.19345025420188905,
      "num_tokens": 35053050.0,
      "step": 17305
    },
    {
      "entropy": 5.988296556472778,
      "epoch": 1.5127151970637072,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047743010304982674,
      "loss": 5.4126,
      "mean_token_accuracy": 0.18651075661182404,
      "num_tokens": 35064381.0,
      "step": 17310
    },
    {
      "entropy": 5.944477319717407,
      "epoch": 1.5131521454164116,
      "grad_norm": 1.0,
      "learning_rate": 0.0004774164983316237,
      "loss": 5.3384,
      "mean_token_accuracy": 0.18800942599773407,
      "num_tokens": 35074787.0,
      "step": 17315
    },
    {
      "entropy": 5.826354742050171,
      "epoch": 1.5135890937691165,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004774028897309662,
      "loss": 5.3351,
      "mean_token_accuracy": 0.19468536078929902,
      "num_tokens": 35085801.0,
      "step": 17320
    },
    {
      "entropy": 5.955341243743897,
      "epoch": 1.5140260421218212,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047738927724811514,
      "loss": 5.3993,
      "mean_token_accuracy": 0.19367965161800385,
      "num_tokens": 35096800.0,
      "step": 17325
    },
    {
      "entropy": 5.895774984359742,
      "epoch": 1.514462990474526,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004773756608833318,
      "loss": 5.2755,
      "mean_token_accuracy": 0.20646664202213288,
      "num_tokens": 35106735.0,
      "step": 17330
    },
    {
      "entropy": 5.846197032928467,
      "epoch": 1.5148999388272306,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047736204063687717,
      "loss": 5.3406,
      "mean_token_accuracy": 0.19663771688938142,
      "num_tokens": 35116221.0,
      "step": 17335
    },
    {
      "entropy": 5.906387186050415,
      "epoch": 1.5153368871799353,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047734841650901255,
      "loss": 5.3107,
      "mean_token_accuracy": 0.200717556476593,
      "num_tokens": 35124896.0,
      "step": 17340
    },
    {
      "entropy": 5.950955963134765,
      "epoch": 1.51577383553264,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004773347884999993,
      "loss": 5.3899,
      "mean_token_accuracy": 0.18890520930290222,
      "num_tokens": 35134576.0,
      "step": 17345
    },
    {
      "entropy": 5.923600530624389,
      "epoch": 1.5162107838853447,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004773211566100988,
      "loss": 5.2865,
      "mean_token_accuracy": 0.20567765980958938,
      "num_tokens": 35144210.0,
      "step": 17350
    },
    {
      "entropy": 5.906909561157226,
      "epoch": 1.5166477322380496,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004773075208395725,
      "loss": 5.2972,
      "mean_token_accuracy": 0.19822865575551987,
      "num_tokens": 35153669.0,
      "step": 17355
    },
    {
      "entropy": 5.921533298492432,
      "epoch": 1.517084680590754,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004772938811886818,
      "loss": 5.4226,
      "mean_token_accuracy": 0.1898806571960449,
      "num_tokens": 35163469.0,
      "step": 17360
    },
    {
      "entropy": 5.967339086532593,
      "epoch": 1.517521628943459,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004772802376576887,
      "loss": 5.4194,
      "mean_token_accuracy": 0.18661053776741027,
      "num_tokens": 35173479.0,
      "step": 17365
    },
    {
      "entropy": 5.9561514377594,
      "epoch": 1.5179585772961635,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004772665902468545,
      "loss": 5.3634,
      "mean_token_accuracy": 0.19533047676086426,
      "num_tokens": 35184570.0,
      "step": 17370
    },
    {
      "entropy": 5.963231897354126,
      "epoch": 1.5183955256488684,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004772529389564412,
      "loss": 5.3997,
      "mean_token_accuracy": 0.1845904141664505,
      "num_tokens": 35193714.0,
      "step": 17375
    },
    {
      "entropy": 5.934446477890015,
      "epoch": 1.518832474001573,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004772392837867104,
      "loss": 5.4141,
      "mean_token_accuracy": 0.19061145484447478,
      "num_tokens": 35203806.0,
      "step": 17380
    },
    {
      "entropy": 5.996020984649658,
      "epoch": 1.5192694223542778,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047722562473792433,
      "loss": 5.4295,
      "mean_token_accuracy": 0.19151860177516938,
      "num_tokens": 35213677.0,
      "step": 17385
    },
    {
      "entropy": 5.908690929412842,
      "epoch": 1.5197063707069824,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047721196181034465,
      "loss": 5.2758,
      "mean_token_accuracy": 0.1967557817697525,
      "num_tokens": 35223663.0,
      "step": 17390
    },
    {
      "entropy": 5.897973585128784,
      "epoch": 1.5201433190596871,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047719829500423374,
      "loss": 5.338,
      "mean_token_accuracy": 0.19006107300519942,
      "num_tokens": 35234551.0,
      "step": 17395
    },
    {
      "entropy": 5.936777973175049,
      "epoch": 1.5205802674123918,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047718462431985354,
      "loss": 5.3732,
      "mean_token_accuracy": 0.19428088068962096,
      "num_tokens": 35244350.0,
      "step": 17400
    },
    {
      "entropy": 5.942840051651001,
      "epoch": 1.5210172157650965,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004771709497574662,
      "loss": 5.3096,
      "mean_token_accuracy": 0.19966583251953124,
      "num_tokens": 35253970.0,
      "step": 17405
    },
    {
      "entropy": 5.895852136611938,
      "epoch": 1.5214541641178014,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047715727131733424,
      "loss": 5.3368,
      "mean_token_accuracy": 0.20081835389137268,
      "num_tokens": 35263316.0,
      "step": 17410
    },
    {
      "entropy": 5.890225505828857,
      "epoch": 1.521891112470506,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047714358899971984,
      "loss": 5.2843,
      "mean_token_accuracy": 0.198163141310215,
      "num_tokens": 35273407.0,
      "step": 17415
    },
    {
      "entropy": 5.917234134674072,
      "epoch": 1.5223280608232108,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004771299028048855,
      "loss": 5.4402,
      "mean_token_accuracy": 0.18566306978464125,
      "num_tokens": 35284225.0,
      "step": 17420
    },
    {
      "entropy": 5.942920589447022,
      "epoch": 1.5227650091759153,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047711621273309365,
      "loss": 5.2717,
      "mean_token_accuracy": 0.19619423300027847,
      "num_tokens": 35294747.0,
      "step": 17425
    },
    {
      "entropy": 5.9268474102020265,
      "epoch": 1.5232019575286202,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000477102518784607,
      "loss": 5.3812,
      "mean_token_accuracy": 0.1976624384522438,
      "num_tokens": 35304856.0,
      "step": 17430
    },
    {
      "entropy": 5.974918460845947,
      "epoch": 1.5236389058813247,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047708882095968817,
      "loss": 5.4313,
      "mean_token_accuracy": 0.19476866871118545,
      "num_tokens": 35315550.0,
      "step": 17435
    },
    {
      "entropy": 5.996383714675903,
      "epoch": 1.5240758542340296,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004770751192585997,
      "loss": 5.4568,
      "mean_token_accuracy": 0.18702697157859802,
      "num_tokens": 35325970.0,
      "step": 17440
    },
    {
      "entropy": 5.8929966449737545,
      "epoch": 1.5245128025867343,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004770614136816047,
      "loss": 5.3508,
      "mean_token_accuracy": 0.19216359853744508,
      "num_tokens": 35337277.0,
      "step": 17445
    },
    {
      "entropy": 5.9446330070495605,
      "epoch": 1.524949750939439,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047704770422896586,
      "loss": 5.4395,
      "mean_token_accuracy": 0.18613397926092148,
      "num_tokens": 35347686.0,
      "step": 17450
    },
    {
      "entropy": 6.0134594440460205,
      "epoch": 1.5253866992921437,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047703399090094623,
      "loss": 5.3406,
      "mean_token_accuracy": 0.1946569040417671,
      "num_tokens": 35358727.0,
      "step": 17455
    },
    {
      "entropy": 5.9185779094696045,
      "epoch": 1.5258236476448483,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004770202736978087,
      "loss": 5.4035,
      "mean_token_accuracy": 0.19185979813337325,
      "num_tokens": 35369741.0,
      "step": 17460
    },
    {
      "entropy": 5.94719557762146,
      "epoch": 1.526260595997553,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004770065526198166,
      "loss": 5.4366,
      "mean_token_accuracy": 0.19340642243623735,
      "num_tokens": 35380213.0,
      "step": 17465
    },
    {
      "entropy": 5.937335205078125,
      "epoch": 1.5266975443502577,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004769928276672328,
      "loss": 5.377,
      "mean_token_accuracy": 0.19517465978860854,
      "num_tokens": 35389937.0,
      "step": 17470
    },
    {
      "entropy": 5.934479761123657,
      "epoch": 1.5271344927029626,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004769790988403208,
      "loss": 5.3732,
      "mean_token_accuracy": 0.19371370673179628,
      "num_tokens": 35400534.0,
      "step": 17475
    },
    {
      "entropy": 5.937342357635498,
      "epoch": 1.5275714410556671,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004769653661393438,
      "loss": 5.3964,
      "mean_token_accuracy": 0.191179521381855,
      "num_tokens": 35410607.0,
      "step": 17480
    },
    {
      "entropy": 5.969351148605346,
      "epoch": 1.528008389408372,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004769516295645654,
      "loss": 5.3921,
      "mean_token_accuracy": 0.1909470096230507,
      "num_tokens": 35420242.0,
      "step": 17485
    },
    {
      "entropy": 5.879063940048217,
      "epoch": 1.5284453377610765,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047693788911624876,
      "loss": 5.3318,
      "mean_token_accuracy": 0.19992929548025132,
      "num_tokens": 35429469.0,
      "step": 17490
    },
    {
      "entropy": 5.899411630630493,
      "epoch": 1.5288822861137814,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004769241447946577,
      "loss": 5.397,
      "mean_token_accuracy": 0.19387564361095427,
      "num_tokens": 35439656.0,
      "step": 17495
    },
    {
      "entropy": 5.979111003875732,
      "epoch": 1.529319234466486,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004769103966000557,
      "loss": 5.2789,
      "mean_token_accuracy": 0.193449966609478,
      "num_tokens": 35448600.0,
      "step": 17500
    },
    {
      "entropy": 5.895920610427856,
      "epoch": 1.5297561828191908,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004768966445327065,
      "loss": 5.3677,
      "mean_token_accuracy": 0.19853124320507048,
      "num_tokens": 35458124.0,
      "step": 17505
    },
    {
      "entropy": 5.932988166809082,
      "epoch": 1.5301931311718955,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004768828885928738,
      "loss": 5.3416,
      "mean_token_accuracy": 0.19027460217475892,
      "num_tokens": 35467616.0,
      "step": 17510
    },
    {
      "entropy": 5.940256786346436,
      "epoch": 1.5306300795246002,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004768691287808216,
      "loss": 5.4104,
      "mean_token_accuracy": 0.1941087946295738,
      "num_tokens": 35478356.0,
      "step": 17515
    },
    {
      "entropy": 5.933048057556152,
      "epoch": 1.5310670278773049,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004768553650968137,
      "loss": 5.3952,
      "mean_token_accuracy": 0.19571640938520432,
      "num_tokens": 35490213.0,
      "step": 17520
    },
    {
      "entropy": 5.888135766983032,
      "epoch": 1.5315039762300096,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004768415975411142,
      "loss": 5.3032,
      "mean_token_accuracy": 0.20007455497980117,
      "num_tokens": 35499982.0,
      "step": 17525
    },
    {
      "entropy": 5.94724326133728,
      "epoch": 1.5319409245827145,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047682782611398693,
      "loss": 5.409,
      "mean_token_accuracy": 0.1909792885184288,
      "num_tokens": 35510572.0,
      "step": 17530
    },
    {
      "entropy": 6.034134769439698,
      "epoch": 1.532377872935419,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047681405081569637,
      "loss": 5.4737,
      "mean_token_accuracy": 0.1859863206744194,
      "num_tokens": 35521560.0,
      "step": 17535
    },
    {
      "entropy": 6.0142249584198,
      "epoch": 1.5328148212881239,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047680027164650644,
      "loss": 5.4497,
      "mean_token_accuracy": 0.18425901085138321,
      "num_tokens": 35530879.0,
      "step": 17540
    },
    {
      "entropy": 5.9494544506073,
      "epoch": 1.5332517696408283,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047678648860668163,
      "loss": 5.4063,
      "mean_token_accuracy": 0.19150852411985397,
      "num_tokens": 35541210.0,
      "step": 17545
    },
    {
      "entropy": 5.914788341522216,
      "epoch": 1.5336887179935332,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047677270169648616,
      "loss": 5.3675,
      "mean_token_accuracy": 0.1951254889369011,
      "num_tokens": 35550149.0,
      "step": 17550
    },
    {
      "entropy": 5.9883654594421385,
      "epoch": 1.534125666346238,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047675891091618464,
      "loss": 5.3437,
      "mean_token_accuracy": 0.19251646101474762,
      "num_tokens": 35561191.0,
      "step": 17555
    },
    {
      "entropy": 5.956315708160401,
      "epoch": 1.5345626146989426,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047674511626604144,
      "loss": 5.314,
      "mean_token_accuracy": 0.2014037311077118,
      "num_tokens": 35571636.0,
      "step": 17560
    },
    {
      "entropy": 5.94164719581604,
      "epoch": 1.5349995630516473,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004767313177463212,
      "loss": 5.3542,
      "mean_token_accuracy": 0.19030099213123322,
      "num_tokens": 35583191.0,
      "step": 17565
    },
    {
      "entropy": 5.869876670837402,
      "epoch": 1.535436511404352,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004767175153572885,
      "loss": 5.3228,
      "mean_token_accuracy": 0.1921442851424217,
      "num_tokens": 35592875.0,
      "step": 17570
    },
    {
      "entropy": 5.913808584213257,
      "epoch": 1.5358734597570567,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004767037090992083,
      "loss": 5.2984,
      "mean_token_accuracy": 0.20119868963956833,
      "num_tokens": 35602158.0,
      "step": 17575
    },
    {
      "entropy": 5.8755903244018555,
      "epoch": 1.5363104081097614,
      "grad_norm": 1.0,
      "learning_rate": 0.00047668989897234516,
      "loss": 5.3341,
      "mean_token_accuracy": 0.19669338911771775,
      "num_tokens": 35612077.0,
      "step": 17580
    },
    {
      "entropy": 5.911571502685547,
      "epoch": 1.5367473564624663,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047667608497696415,
      "loss": 5.3001,
      "mean_token_accuracy": 0.1925256848335266,
      "num_tokens": 35621856.0,
      "step": 17585
    },
    {
      "entropy": 5.9142168998718265,
      "epoch": 1.5371843048151708,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047666226711333014,
      "loss": 5.3951,
      "mean_token_accuracy": 0.1933210998773575,
      "num_tokens": 35633160.0,
      "step": 17590
    },
    {
      "entropy": 5.987384176254272,
      "epoch": 1.5376212531678757,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004766484453817081,
      "loss": 5.3999,
      "mean_token_accuracy": 0.18930098563432693,
      "num_tokens": 35642477.0,
      "step": 17595
    },
    {
      "entropy": 5.931742858886719,
      "epoch": 1.5380582015205801,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004766346197823633,
      "loss": 5.358,
      "mean_token_accuracy": 0.19075200855731964,
      "num_tokens": 35651938.0,
      "step": 17600
    },
    {
      "entropy": 5.937109136581421,
      "epoch": 1.538495149873285,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047662079031556085,
      "loss": 5.371,
      "mean_token_accuracy": 0.1920846074819565,
      "num_tokens": 35662881.0,
      "step": 17605
    },
    {
      "entropy": 5.936427068710327,
      "epoch": 1.5389320982259895,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004766069569815659,
      "loss": 5.3534,
      "mean_token_accuracy": 0.18855183869600295,
      "num_tokens": 35673436.0,
      "step": 17610
    },
    {
      "entropy": 5.932575273513794,
      "epoch": 1.5393690465786944,
      "grad_norm": 1.140625,
      "learning_rate": 0.000476593119780644,
      "loss": 5.3302,
      "mean_token_accuracy": 0.1930941179394722,
      "num_tokens": 35683411.0,
      "step": 17615
    },
    {
      "entropy": 5.931133890151978,
      "epoch": 1.5398059949313991,
      "grad_norm": 1.0,
      "learning_rate": 0.00047657927871306037,
      "loss": 5.3381,
      "mean_token_accuracy": 0.19494591355323793,
      "num_tokens": 35693806.0,
      "step": 17620
    },
    {
      "entropy": 5.90626163482666,
      "epoch": 1.5402429432841038,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004765654337790805,
      "loss": 5.3657,
      "mean_token_accuracy": 0.19733764976263046,
      "num_tokens": 35705015.0,
      "step": 17625
    },
    {
      "entropy": 5.991803503036499,
      "epoch": 1.5406798916368085,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004765515849789701,
      "loss": 5.4006,
      "mean_token_accuracy": 0.19187617301940918,
      "num_tokens": 35715668.0,
      "step": 17630
    },
    {
      "entropy": 5.986660385131836,
      "epoch": 1.5411168399895132,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047653773231299455,
      "loss": 5.3572,
      "mean_token_accuracy": 0.1966194838285446,
      "num_tokens": 35726381.0,
      "step": 17635
    },
    {
      "entropy": 5.9291564464569095,
      "epoch": 1.541553788342218,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047652387578141973,
      "loss": 5.3753,
      "mean_token_accuracy": 0.19292495548725128,
      "num_tokens": 35736223.0,
      "step": 17640
    },
    {
      "entropy": 5.945890426635742,
      "epoch": 1.5419907366949226,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047651001538451146,
      "loss": 5.3745,
      "mean_token_accuracy": 0.19145803153514862,
      "num_tokens": 35746959.0,
      "step": 17645
    },
    {
      "entropy": 5.9459137439727785,
      "epoch": 1.5424276850476275,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047649615112253547,
      "loss": 5.3506,
      "mean_token_accuracy": 0.20160731673240662,
      "num_tokens": 35756975.0,
      "step": 17650
    },
    {
      "entropy": 5.902245330810547,
      "epoch": 1.542864633400332,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047648228299575766,
      "loss": 5.3237,
      "mean_token_accuracy": 0.19980331361293793,
      "num_tokens": 35765890.0,
      "step": 17655
    },
    {
      "entropy": 5.906957960128784,
      "epoch": 1.5433015817530369,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047646841100444415,
      "loss": 5.2753,
      "mean_token_accuracy": 0.20187938511371611,
      "num_tokens": 35775630.0,
      "step": 17660
    },
    {
      "entropy": 5.8998230457305905,
      "epoch": 1.5437385301057414,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047645453514886095,
      "loss": 5.3931,
      "mean_token_accuracy": 0.1978357583284378,
      "num_tokens": 35786053.0,
      "step": 17665
    },
    {
      "entropy": 5.943258810043335,
      "epoch": 1.5441754784584463,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004764406554292742,
      "loss": 5.3963,
      "mean_token_accuracy": 0.19061651080846786,
      "num_tokens": 35795736.0,
      "step": 17670
    },
    {
      "entropy": 5.987040662765503,
      "epoch": 1.544612426811151,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004764267718459501,
      "loss": 5.3249,
      "mean_token_accuracy": 0.19682199954986573,
      "num_tokens": 35805887.0,
      "step": 17675
    },
    {
      "entropy": 5.876585817337036,
      "epoch": 1.5450493751638557,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000476412884399155,
      "loss": 5.3681,
      "mean_token_accuracy": 0.1935882106423378,
      "num_tokens": 35815625.0,
      "step": 17680
    },
    {
      "entropy": 5.956658506393433,
      "epoch": 1.5454863235165603,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004763989930891552,
      "loss": 5.3295,
      "mean_token_accuracy": 0.19058604389429093,
      "num_tokens": 35825617.0,
      "step": 17685
    },
    {
      "entropy": 5.928430843353271,
      "epoch": 1.545923271869265,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047638509791621714,
      "loss": 5.3751,
      "mean_token_accuracy": 0.18593814224004745,
      "num_tokens": 35835538.0,
      "step": 17690
    },
    {
      "entropy": 5.926581954956054,
      "epoch": 1.5463602202219697,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047637119888060747,
      "loss": 5.2649,
      "mean_token_accuracy": 0.2024490013718605,
      "num_tokens": 35846302.0,
      "step": 17695
    },
    {
      "entropy": 5.990643787384033,
      "epoch": 1.5467971685746744,
      "grad_norm": 1.125,
      "learning_rate": 0.00047635729598259266,
      "loss": 5.4519,
      "mean_token_accuracy": 0.18585081547498702,
      "num_tokens": 35857384.0,
      "step": 17700
    },
    {
      "entropy": 5.999239492416382,
      "epoch": 1.5472341169273793,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004763433892224394,
      "loss": 5.4011,
      "mean_token_accuracy": 0.18472544997930526,
      "num_tokens": 35867496.0,
      "step": 17705
    },
    {
      "entropy": 5.936768007278443,
      "epoch": 1.5476710652800838,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004763294786004144,
      "loss": 5.3911,
      "mean_token_accuracy": 0.1923845201730728,
      "num_tokens": 35878051.0,
      "step": 17710
    },
    {
      "entropy": 5.925836563110352,
      "epoch": 1.5481080136327887,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004763155641167846,
      "loss": 5.3318,
      "mean_token_accuracy": 0.1996160551905632,
      "num_tokens": 35887926.0,
      "step": 17715
    },
    {
      "entropy": 5.92617883682251,
      "epoch": 1.5485449619854932,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047630164577181666,
      "loss": 5.3404,
      "mean_token_accuracy": 0.1946843683719635,
      "num_tokens": 35897982.0,
      "step": 17720
    },
    {
      "entropy": 5.923866128921508,
      "epoch": 1.548981910338198,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004762877235657777,
      "loss": 5.3662,
      "mean_token_accuracy": 0.20035843998193742,
      "num_tokens": 35908805.0,
      "step": 17725
    },
    {
      "entropy": 5.942095184326172,
      "epoch": 1.5494188586909028,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004762737974989348,
      "loss": 5.3374,
      "mean_token_accuracy": 0.19958553463220596,
      "num_tokens": 35918739.0,
      "step": 17730
    },
    {
      "entropy": 5.939583110809326,
      "epoch": 1.5498558070436075,
      "grad_norm": 1.125,
      "learning_rate": 0.000476259867571555,
      "loss": 5.4375,
      "mean_token_accuracy": 0.1889088660478592,
      "num_tokens": 35928592.0,
      "step": 17735
    },
    {
      "entropy": 5.844542217254639,
      "epoch": 1.5502927553963122,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004762459337839055,
      "loss": 5.3475,
      "mean_token_accuracy": 0.19364595264196396,
      "num_tokens": 35938287.0,
      "step": 17740
    },
    {
      "entropy": 5.963957595825195,
      "epoch": 1.5507297037490169,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047623199613625343,
      "loss": 5.3552,
      "mean_token_accuracy": 0.19228036999702453,
      "num_tokens": 35947890.0,
      "step": 17745
    },
    {
      "entropy": 5.910624694824219,
      "epoch": 1.5511666521017216,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004762180546288663,
      "loss": 5.3543,
      "mean_token_accuracy": 0.19413044601678847,
      "num_tokens": 35956513.0,
      "step": 17750
    },
    {
      "entropy": 5.87700514793396,
      "epoch": 1.5516036004544262,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004762041092620114,
      "loss": 5.3469,
      "mean_token_accuracy": 0.2008675217628479,
      "num_tokens": 35966837.0,
      "step": 17755
    },
    {
      "entropy": 5.982762479782105,
      "epoch": 1.5520405488071312,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047619016003595625,
      "loss": 5.485,
      "mean_token_accuracy": 0.18736762851476668,
      "num_tokens": 35977076.0,
      "step": 17760
    },
    {
      "entropy": 5.984706211090088,
      "epoch": 1.5524774971598356,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047617620695096854,
      "loss": 5.309,
      "mean_token_accuracy": 0.1957852467894554,
      "num_tokens": 35986862.0,
      "step": 17765
    },
    {
      "entropy": 5.970073175430298,
      "epoch": 1.5529144455125405,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047616225000731566,
      "loss": 5.4073,
      "mean_token_accuracy": 0.18255013823509217,
      "num_tokens": 35997488.0,
      "step": 17770
    },
    {
      "entropy": 5.903728818893432,
      "epoch": 1.553351393865245,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004761482892052655,
      "loss": 5.3561,
      "mean_token_accuracy": 0.1927005648612976,
      "num_tokens": 36008045.0,
      "step": 17775
    },
    {
      "entropy": 5.958009052276611,
      "epoch": 1.55378834221795,
      "grad_norm": 1.125,
      "learning_rate": 0.00047613432454508565,
      "loss": 5.3558,
      "mean_token_accuracy": 0.18460554033517837,
      "num_tokens": 36017864.0,
      "step": 17780
    },
    {
      "entropy": 5.995775032043457,
      "epoch": 1.5542252905706544,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047612035602704407,
      "loss": 5.3393,
      "mean_token_accuracy": 0.19702621400356293,
      "num_tokens": 36027797.0,
      "step": 17785
    },
    {
      "entropy": 5.97640585899353,
      "epoch": 1.5546622389233593,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047610638365140867,
      "loss": 5.3624,
      "mean_token_accuracy": 0.18952467888593674,
      "num_tokens": 36037627.0,
      "step": 17790
    },
    {
      "entropy": 5.950860786437988,
      "epoch": 1.555099187276064,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047609240741844757,
      "loss": 5.357,
      "mean_token_accuracy": 0.19165082275867462,
      "num_tokens": 36046324.0,
      "step": 17795
    },
    {
      "entropy": 5.925674533843994,
      "epoch": 1.5555361356287687,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004760784273284286,
      "loss": 5.3211,
      "mean_token_accuracy": 0.19897984564304352,
      "num_tokens": 36056212.0,
      "step": 17800
    },
    {
      "entropy": 5.9572783470153805,
      "epoch": 1.5559730839814734,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004760644433816201,
      "loss": 5.4668,
      "mean_token_accuracy": 0.1824682965874672,
      "num_tokens": 36066745.0,
      "step": 17805
    },
    {
      "entropy": 6.005882120132446,
      "epoch": 1.556410032334178,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004760504555782902,
      "loss": 5.3373,
      "mean_token_accuracy": 0.190341018140316,
      "num_tokens": 36077415.0,
      "step": 17810
    },
    {
      "entropy": 5.913064861297608,
      "epoch": 1.5568469806868828,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047603646391870715,
      "loss": 5.3565,
      "mean_token_accuracy": 0.1969461366534233,
      "num_tokens": 36088381.0,
      "step": 17815
    },
    {
      "entropy": 5.852984428405762,
      "epoch": 1.5572839290395875,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047602246840313945,
      "loss": 5.3308,
      "mean_token_accuracy": 0.19753824472427367,
      "num_tokens": 36098333.0,
      "step": 17820
    },
    {
      "entropy": 5.930189990997315,
      "epoch": 1.5577208773922924,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047600846903185544,
      "loss": 5.3658,
      "mean_token_accuracy": 0.19420438408851623,
      "num_tokens": 36107843.0,
      "step": 17825
    },
    {
      "entropy": 5.921025085449219,
      "epoch": 1.5581578257449968,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004759944658051237,
      "loss": 5.3727,
      "mean_token_accuracy": 0.19476681798696518,
      "num_tokens": 36116948.0,
      "step": 17830
    },
    {
      "entropy": 5.961842107772827,
      "epoch": 1.5585947740977018,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004759804587232127,
      "loss": 5.374,
      "mean_token_accuracy": 0.19379333704710006,
      "num_tokens": 36126808.0,
      "step": 17835
    },
    {
      "entropy": 6.005739688873291,
      "epoch": 1.5590317224504062,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047596644778639133,
      "loss": 5.4753,
      "mean_token_accuracy": 0.18406805992126465,
      "num_tokens": 36136630.0,
      "step": 17840
    },
    {
      "entropy": 6.059730148315429,
      "epoch": 1.5594686708031111,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004759524329949281,
      "loss": 5.3903,
      "mean_token_accuracy": 0.18391201198101043,
      "num_tokens": 36146422.0,
      "step": 17845
    },
    {
      "entropy": 5.948061418533325,
      "epoch": 1.5599056191558158,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004759384143490919,
      "loss": 5.3753,
      "mean_token_accuracy": 0.19095561653375626,
      "num_tokens": 36156169.0,
      "step": 17850
    },
    {
      "entropy": 5.920959615707398,
      "epoch": 1.5603425675085205,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004759243918491517,
      "loss": 5.379,
      "mean_token_accuracy": 0.19626594036817552,
      "num_tokens": 36165173.0,
      "step": 17855
    },
    {
      "entropy": 5.987109041213989,
      "epoch": 1.5607795158612252,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047591036549537633,
      "loss": 5.4243,
      "mean_token_accuracy": 0.189918252825737,
      "num_tokens": 36175852.0,
      "step": 17860
    },
    {
      "entropy": 5.956524324417114,
      "epoch": 1.56121646421393,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004758963352880348,
      "loss": 5.3904,
      "mean_token_accuracy": 0.18796810060739516,
      "num_tokens": 36185288.0,
      "step": 17865
    },
    {
      "entropy": 5.962094736099243,
      "epoch": 1.5616534125666346,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004758823012273964,
      "loss": 5.3896,
      "mean_token_accuracy": 0.19087910354137422,
      "num_tokens": 36196374.0,
      "step": 17870
    },
    {
      "entropy": 5.939943933486939,
      "epoch": 1.5620903609193393,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047586826331373025,
      "loss": 5.394,
      "mean_token_accuracy": 0.19338358938694,
      "num_tokens": 36206714.0,
      "step": 17875
    },
    {
      "entropy": 5.9810943603515625,
      "epoch": 1.5625273092720442,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004758542215473054,
      "loss": 5.486,
      "mean_token_accuracy": 0.18169303983449936,
      "num_tokens": 36217891.0,
      "step": 17880
    },
    {
      "entropy": 5.966112899780273,
      "epoch": 1.5629642576247487,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004758401759283915,
      "loss": 5.3443,
      "mean_token_accuracy": 0.19357675462961196,
      "num_tokens": 36226776.0,
      "step": 17885
    },
    {
      "entropy": 5.957785654067993,
      "epoch": 1.5634012059774536,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047582612645725764,
      "loss": 5.397,
      "mean_token_accuracy": 0.1930985078215599,
      "num_tokens": 36237362.0,
      "step": 17890
    },
    {
      "entropy": 5.919914436340332,
      "epoch": 1.563838154330158,
      "grad_norm": 1.0,
      "learning_rate": 0.00047581207313417346,
      "loss": 5.3198,
      "mean_token_accuracy": 0.1961434319615364,
      "num_tokens": 36246264.0,
      "step": 17895
    },
    {
      "entropy": 5.9693090438842775,
      "epoch": 1.564275102682863,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047579801595940853,
      "loss": 5.4122,
      "mean_token_accuracy": 0.18978408277034758,
      "num_tokens": 36255676.0,
      "step": 17900
    },
    {
      "entropy": 5.96657133102417,
      "epoch": 1.5647120510355674,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047578395493323245,
      "loss": 5.2639,
      "mean_token_accuracy": 0.2070227161049843,
      "num_tokens": 36264346.0,
      "step": 17905
    },
    {
      "entropy": 5.930969858169556,
      "epoch": 1.5651489993882723,
      "grad_norm": 1.125,
      "learning_rate": 0.00047576989005591495,
      "loss": 5.3685,
      "mean_token_accuracy": 0.19207754284143447,
      "num_tokens": 36273569.0,
      "step": 17910
    },
    {
      "entropy": 5.882574653625488,
      "epoch": 1.565585947740977,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004757558213277257,
      "loss": 5.2776,
      "mean_token_accuracy": 0.19909806847572326,
      "num_tokens": 36283708.0,
      "step": 17915
    },
    {
      "entropy": 5.943996715545654,
      "epoch": 1.5660228960936817,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047574174874893456,
      "loss": 5.3566,
      "mean_token_accuracy": 0.19213392436504365,
      "num_tokens": 36293184.0,
      "step": 17920
    },
    {
      "entropy": 5.962588262557984,
      "epoch": 1.5664598444463864,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004757276723198115,
      "loss": 5.4451,
      "mean_token_accuracy": 0.18556798249483109,
      "num_tokens": 36303071.0,
      "step": 17925
    },
    {
      "entropy": 5.928398466110229,
      "epoch": 1.566896792799091,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004757135920406266,
      "loss": 5.3076,
      "mean_token_accuracy": 0.20150104016065598,
      "num_tokens": 36313245.0,
      "step": 17930
    },
    {
      "entropy": 5.9323567867279055,
      "epoch": 1.567333741151796,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004756995079116498,
      "loss": 5.3471,
      "mean_token_accuracy": 0.18859713524580002,
      "num_tokens": 36322711.0,
      "step": 17935
    },
    {
      "entropy": 5.910932350158691,
      "epoch": 1.5677706895045005,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047568541993315124,
      "loss": 5.3597,
      "mean_token_accuracy": 0.19202776551246642,
      "num_tokens": 36333667.0,
      "step": 17940
    },
    {
      "entropy": 6.017526245117187,
      "epoch": 1.5682076378572054,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047567132810540113,
      "loss": 5.4095,
      "mean_token_accuracy": 0.19304633736610413,
      "num_tokens": 36343660.0,
      "step": 17945
    },
    {
      "entropy": 5.960295867919922,
      "epoch": 1.5686445862099099,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047565723242866983,
      "loss": 5.4452,
      "mean_token_accuracy": 0.19347493648529052,
      "num_tokens": 36353581.0,
      "step": 17950
    },
    {
      "entropy": 5.962349987030029,
      "epoch": 1.5690815345626148,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047564313290322774,
      "loss": 5.3231,
      "mean_token_accuracy": 0.1901409685611725,
      "num_tokens": 36363818.0,
      "step": 17955
    },
    {
      "entropy": 5.915064334869385,
      "epoch": 1.5695184829153193,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004756290295293451,
      "loss": 5.362,
      "mean_token_accuracy": 0.19492893517017365,
      "num_tokens": 36373590.0,
      "step": 17960
    },
    {
      "entropy": 5.9700957298278805,
      "epoch": 1.5699554312680242,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004756149223072926,
      "loss": 5.4319,
      "mean_token_accuracy": 0.18828320503234863,
      "num_tokens": 36383556.0,
      "step": 17965
    },
    {
      "entropy": 5.944630908966064,
      "epoch": 1.5703923796207289,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004756008112373408,
      "loss": 5.3775,
      "mean_token_accuracy": 0.1888403832912445,
      "num_tokens": 36393194.0,
      "step": 17970
    },
    {
      "entropy": 5.90058445930481,
      "epoch": 1.5708293279734336,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004755866963197603,
      "loss": 5.3484,
      "mean_token_accuracy": 0.1893824055790901,
      "num_tokens": 36402962.0,
      "step": 17975
    },
    {
      "entropy": 5.97128324508667,
      "epoch": 1.5712662763261382,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004755725775548219,
      "loss": 5.3555,
      "mean_token_accuracy": 0.1917151227593422,
      "num_tokens": 36411932.0,
      "step": 17980
    },
    {
      "entropy": 5.934804821014405,
      "epoch": 1.571703224678843,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047555845494279633,
      "loss": 5.3467,
      "mean_token_accuracy": 0.19615339487791061,
      "num_tokens": 36421363.0,
      "step": 17985
    },
    {
      "entropy": 5.931304693222046,
      "epoch": 1.5721401730315476,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004755443284839545,
      "loss": 5.3529,
      "mean_token_accuracy": 0.19118622541427613,
      "num_tokens": 36431422.0,
      "step": 17990
    },
    {
      "entropy": 5.894608736038208,
      "epoch": 1.5725771213842523,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047553019817856736,
      "loss": 5.3091,
      "mean_token_accuracy": 0.20306738317012787,
      "num_tokens": 36441004.0,
      "step": 17995
    },
    {
      "entropy": 5.988725423812866,
      "epoch": 1.5730140697369572,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047551606402690595,
      "loss": 5.4409,
      "mean_token_accuracy": 0.18739915937185286,
      "num_tokens": 36451474.0,
      "step": 18000
    },
    {
      "epoch": 1.5730140697369572,
      "eval_entropy": 5.730322190744975,
      "eval_loss": 5.399023056030273,
      "eval_mean_token_accuracy": 0.20113133602061692,
      "eval_num_tokens": 36451474.0,
      "eval_runtime": 21.4786,
      "eval_samples_per_second": 1723.294,
      "eval_steps_per_second": 215.423,
      "step": 18000
    },
    {
      "entropy": 5.930827236175537,
      "epoch": 1.5734510180896617,
      "grad_norm": 1.125,
      "learning_rate": 0.0004755019260292414,
      "loss": 5.3485,
      "mean_token_accuracy": 0.1933707907795906,
      "num_tokens": 36461341.0,
      "step": 18005
    },
    {
      "entropy": 5.945961618423462,
      "epoch": 1.5738879664423666,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004754877841858448,
      "loss": 5.4627,
      "mean_token_accuracy": 0.18912752270698546,
      "num_tokens": 36471679.0,
      "step": 18010
    },
    {
      "entropy": 5.966172695159912,
      "epoch": 1.574324914795071,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004754736384969875,
      "loss": 5.385,
      "mean_token_accuracy": 0.19129979312419892,
      "num_tokens": 36481457.0,
      "step": 18015
    },
    {
      "entropy": 5.9558553218841555,
      "epoch": 1.574761863147776,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004754594889629407,
      "loss": 5.3491,
      "mean_token_accuracy": 0.19105580896139146,
      "num_tokens": 36490596.0,
      "step": 18020
    },
    {
      "entropy": 6.005193519592285,
      "epoch": 1.5751988115004807,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004754453355839759,
      "loss": 5.4078,
      "mean_token_accuracy": 0.19219863414764404,
      "num_tokens": 36501027.0,
      "step": 18025
    },
    {
      "entropy": 5.9588275909423825,
      "epoch": 1.5756357598531854,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047543117836036446,
      "loss": 5.3305,
      "mean_token_accuracy": 0.19428578913211822,
      "num_tokens": 36512089.0,
      "step": 18030
    },
    {
      "entropy": 5.91853895187378,
      "epoch": 1.57607270820589,
      "grad_norm": 1.0,
      "learning_rate": 0.00047541701729237815,
      "loss": 5.3349,
      "mean_token_accuracy": 0.19521148651838302,
      "num_tokens": 36522521.0,
      "step": 18035
    },
    {
      "entropy": 6.017958354949951,
      "epoch": 1.5765096565585948,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004754028523802883,
      "loss": 5.4294,
      "mean_token_accuracy": 0.1866988569498062,
      "num_tokens": 36533111.0,
      "step": 18040
    },
    {
      "entropy": 5.900738716125488,
      "epoch": 1.5769466049112995,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004753886836243668,
      "loss": 5.3389,
      "mean_token_accuracy": 0.20958115458488463,
      "num_tokens": 36542477.0,
      "step": 18045
    },
    {
      "entropy": 5.9259601593017575,
      "epoch": 1.5773835532640041,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047537451102488526,
      "loss": 5.3587,
      "mean_token_accuracy": 0.19652674198150635,
      "num_tokens": 36552137.0,
      "step": 18050
    },
    {
      "entropy": 5.950625228881836,
      "epoch": 1.577820501616709,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004753603345821157,
      "loss": 5.3829,
      "mean_token_accuracy": 0.1906690090894699,
      "num_tokens": 36562440.0,
      "step": 18055
    },
    {
      "entropy": 5.938940525054932,
      "epoch": 1.5782574499694135,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047534615429632986,
      "loss": 5.3575,
      "mean_token_accuracy": 0.19788594096899031,
      "num_tokens": 36571863.0,
      "step": 18060
    },
    {
      "entropy": 5.936715745925904,
      "epoch": 1.5786943983221184,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004753319701677999,
      "loss": 5.3537,
      "mean_token_accuracy": 0.19590529799461365,
      "num_tokens": 36581409.0,
      "step": 18065
    },
    {
      "entropy": 5.975103282928467,
      "epoch": 1.579131346674823,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047531778219679773,
      "loss": 5.3514,
      "mean_token_accuracy": 0.1873975560069084,
      "num_tokens": 36592779.0,
      "step": 18070
    },
    {
      "entropy": 5.956869220733642,
      "epoch": 1.5795682950275278,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047530359038359547,
      "loss": 5.3521,
      "mean_token_accuracy": 0.2036803647875786,
      "num_tokens": 36601173.0,
      "step": 18075
    },
    {
      "entropy": 5.9573814868927,
      "epoch": 1.5800052433802323,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004752893947284654,
      "loss": 5.393,
      "mean_token_accuracy": 0.19510993510484695,
      "num_tokens": 36610147.0,
      "step": 18080
    },
    {
      "entropy": 5.917958068847656,
      "epoch": 1.5804421917329372,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004752751952316799,
      "loss": 5.3733,
      "mean_token_accuracy": 0.1925841212272644,
      "num_tokens": 36620093.0,
      "step": 18085
    },
    {
      "entropy": 5.966060733795166,
      "epoch": 1.580879140085642,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047526099189351115,
      "loss": 5.373,
      "mean_token_accuracy": 0.19651580303907396,
      "num_tokens": 36631833.0,
      "step": 18090
    },
    {
      "entropy": 5.972617435455322,
      "epoch": 1.5813160884383466,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004752467847142316,
      "loss": 5.3327,
      "mean_token_accuracy": 0.196125827729702,
      "num_tokens": 36641123.0,
      "step": 18095
    },
    {
      "entropy": 5.932485198974609,
      "epoch": 1.5817530367910513,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047523257369411383,
      "loss": 5.3989,
      "mean_token_accuracy": 0.1942364528775215,
      "num_tokens": 36652310.0,
      "step": 18100
    },
    {
      "entropy": 6.0057779312133786,
      "epoch": 1.582189985143756,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004752183588334303,
      "loss": 5.4432,
      "mean_token_accuracy": 0.18548823297023773,
      "num_tokens": 36662783.0,
      "step": 18105
    },
    {
      "entropy": 5.90848331451416,
      "epoch": 1.5826269334964607,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047520414013245383,
      "loss": 5.348,
      "mean_token_accuracy": 0.19790913313627242,
      "num_tokens": 36672933.0,
      "step": 18110
    },
    {
      "entropy": 5.94440598487854,
      "epoch": 1.5830638818491654,
      "grad_norm": 1.125,
      "learning_rate": 0.0004751899175914571,
      "loss": 5.4319,
      "mean_token_accuracy": 0.18708040863275527,
      "num_tokens": 36682570.0,
      "step": 18115
    },
    {
      "entropy": 5.924571084976196,
      "epoch": 1.5835008302018703,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004751756912107127,
      "loss": 5.3335,
      "mean_token_accuracy": 0.20284097343683244,
      "num_tokens": 36693008.0,
      "step": 18120
    },
    {
      "entropy": 5.902623081207276,
      "epoch": 1.5839377785545747,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047516146099049376,
      "loss": 5.2318,
      "mean_token_accuracy": 0.20479210019111632,
      "num_tokens": 36702611.0,
      "step": 18125
    },
    {
      "entropy": 5.944227933883667,
      "epoch": 1.5843747269072797,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047514722693107304,
      "loss": 5.4123,
      "mean_token_accuracy": 0.1919320523738861,
      "num_tokens": 36712470.0,
      "step": 18130
    },
    {
      "entropy": 6.018637752532959,
      "epoch": 1.5848116752599841,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004751329890327237,
      "loss": 5.4221,
      "mean_token_accuracy": 0.19455501586198806,
      "num_tokens": 36722578.0,
      "step": 18135
    },
    {
      "entropy": 6.028703260421753,
      "epoch": 1.585248623612689,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004751187472957187,
      "loss": 5.3853,
      "mean_token_accuracy": 0.1970709264278412,
      "num_tokens": 36732334.0,
      "step": 18140
    },
    {
      "entropy": 5.906695079803467,
      "epoch": 1.5856855719653937,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004751045017203313,
      "loss": 5.4272,
      "mean_token_accuracy": 0.18399005234241486,
      "num_tokens": 36742328.0,
      "step": 18145
    },
    {
      "entropy": 5.932078886032104,
      "epoch": 1.5861225203180984,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004750902523068347,
      "loss": 5.3748,
      "mean_token_accuracy": 0.1980987474322319,
      "num_tokens": 36754363.0,
      "step": 18150
    },
    {
      "entropy": 6.023695755004883,
      "epoch": 1.586559468670803,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004750759990555022,
      "loss": 5.3827,
      "mean_token_accuracy": 0.18967308104038239,
      "num_tokens": 36765588.0,
      "step": 18155
    },
    {
      "entropy": 5.968165540695191,
      "epoch": 1.5869964170235078,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004750617419666072,
      "loss": 5.3361,
      "mean_token_accuracy": 0.192860709130764,
      "num_tokens": 36774851.0,
      "step": 18160
    },
    {
      "entropy": 5.864459085464477,
      "epoch": 1.5874333653762125,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047504748104042324,
      "loss": 5.3987,
      "mean_token_accuracy": 0.19324476569890975,
      "num_tokens": 36785836.0,
      "step": 18165
    },
    {
      "entropy": 5.911537218093872,
      "epoch": 1.5878703137289172,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047503321627722375,
      "loss": 5.3212,
      "mean_token_accuracy": 0.19682847410440446,
      "num_tokens": 36795367.0,
      "step": 18170
    },
    {
      "entropy": 6.02398452758789,
      "epoch": 1.588307262081622,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047501894767728225,
      "loss": 5.3255,
      "mean_token_accuracy": 0.19702719002962113,
      "num_tokens": 36804294.0,
      "step": 18175
    },
    {
      "entropy": 5.954411315917969,
      "epoch": 1.5887442104343266,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047500467524087266,
      "loss": 5.3845,
      "mean_token_accuracy": 0.19027767330408096,
      "num_tokens": 36813711.0,
      "step": 18180
    },
    {
      "entropy": 5.886231803894043,
      "epoch": 1.5891811587870315,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004749903989682685,
      "loss": 5.3602,
      "mean_token_accuracy": 0.194451966881752,
      "num_tokens": 36824498.0,
      "step": 18185
    },
    {
      "entropy": 5.922590017318726,
      "epoch": 1.589618107139736,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047497611885974366,
      "loss": 5.3292,
      "mean_token_accuracy": 0.19493328630924225,
      "num_tokens": 36833329.0,
      "step": 18190
    },
    {
      "entropy": 5.982392311096191,
      "epoch": 1.5900550554924409,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047496183491557215,
      "loss": 5.3843,
      "mean_token_accuracy": 0.19274914264678955,
      "num_tokens": 36842783.0,
      "step": 18195
    },
    {
      "entropy": 5.977981281280518,
      "epoch": 1.5904920038451456,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00047494754713602787,
      "loss": 5.3432,
      "mean_token_accuracy": 0.18667060285806655,
      "num_tokens": 36852785.0,
      "step": 18200
    },
    {
      "entropy": 5.972883987426758,
      "epoch": 1.5909289521978502,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047493325552138484,
      "loss": 5.3956,
      "mean_token_accuracy": 0.18446148037910462,
      "num_tokens": 36862337.0,
      "step": 18205
    },
    {
      "entropy": 6.003094291687011,
      "epoch": 1.591365900550555,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004749189600719172,
      "loss": 5.4096,
      "mean_token_accuracy": 0.18933023661375045,
      "num_tokens": 36872026.0,
      "step": 18210
    },
    {
      "entropy": 5.961195468902588,
      "epoch": 1.5918028489032596,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004749046607878992,
      "loss": 5.4037,
      "mean_token_accuracy": 0.19418194890022278,
      "num_tokens": 36882520.0,
      "step": 18215
    },
    {
      "entropy": 5.920127296447754,
      "epoch": 1.5922397972559643,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047489035766960504,
      "loss": 5.2998,
      "mean_token_accuracy": 0.2066112533211708,
      "num_tokens": 36892630.0,
      "step": 18220
    },
    {
      "entropy": 5.981991004943848,
      "epoch": 1.592676745608669,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047487605071730904,
      "loss": 5.3882,
      "mean_token_accuracy": 0.19742448180913924,
      "num_tokens": 36902692.0,
      "step": 18225
    },
    {
      "entropy": 6.017651081085205,
      "epoch": 1.593113693961374,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047486173993128566,
      "loss": 5.4865,
      "mean_token_accuracy": 0.183075250685215,
      "num_tokens": 36913581.0,
      "step": 18230
    },
    {
      "entropy": 5.967060708999634,
      "epoch": 1.5935506423140784,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004748474253118094,
      "loss": 5.3719,
      "mean_token_accuracy": 0.18754746913909912,
      "num_tokens": 36924031.0,
      "step": 18235
    },
    {
      "entropy": 5.95620288848877,
      "epoch": 1.5939875906667833,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004748331068591548,
      "loss": 5.3878,
      "mean_token_accuracy": 0.19211733490228652,
      "num_tokens": 36934622.0,
      "step": 18240
    },
    {
      "entropy": 5.958774137496948,
      "epoch": 1.5944245390194878,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004748187845735965,
      "loss": 5.395,
      "mean_token_accuracy": 0.1910262957215309,
      "num_tokens": 36944374.0,
      "step": 18245
    },
    {
      "entropy": 5.987827777862549,
      "epoch": 1.5948614873721927,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047480445845540927,
      "loss": 5.3385,
      "mean_token_accuracy": 0.1962900161743164,
      "num_tokens": 36954371.0,
      "step": 18250
    },
    {
      "entropy": 5.938724422454834,
      "epoch": 1.5952984357248972,
      "grad_norm": 1.0,
      "learning_rate": 0.00047479012850486777,
      "loss": 5.3676,
      "mean_token_accuracy": 0.19419587254524232,
      "num_tokens": 36964217.0,
      "step": 18255
    },
    {
      "entropy": 5.999371147155761,
      "epoch": 1.595735384077602,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047477579472224687,
      "loss": 5.4216,
      "mean_token_accuracy": 0.1880829080939293,
      "num_tokens": 36974580.0,
      "step": 18260
    },
    {
      "entropy": 5.898482084274292,
      "epoch": 1.5961723324303068,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004747614571078216,
      "loss": 5.2196,
      "mean_token_accuracy": 0.21170909702777863,
      "num_tokens": 36983434.0,
      "step": 18265
    },
    {
      "entropy": 5.932907009124756,
      "epoch": 1.5966092807830115,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004747471156618669,
      "loss": 5.4554,
      "mean_token_accuracy": 0.1889460176229477,
      "num_tokens": 36993813.0,
      "step": 18270
    },
    {
      "entropy": 5.97841968536377,
      "epoch": 1.5970462291357161,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004747327703846579,
      "loss": 5.3625,
      "mean_token_accuracy": 0.19432138949632644,
      "num_tokens": 37003855.0,
      "step": 18275
    },
    {
      "entropy": 5.950984764099121,
      "epoch": 1.5974831774884208,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047471842127646966,
      "loss": 5.3273,
      "mean_token_accuracy": 0.19477203339338303,
      "num_tokens": 37013660.0,
      "step": 18280
    },
    {
      "entropy": 5.87025556564331,
      "epoch": 1.5979201258411255,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047470406833757746,
      "loss": 5.2771,
      "mean_token_accuracy": 0.197653067111969,
      "num_tokens": 37023888.0,
      "step": 18285
    },
    {
      "entropy": 6.015871620178222,
      "epoch": 1.5983570741938302,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047468971156825656,
      "loss": 5.4721,
      "mean_token_accuracy": 0.17303868234157563,
      "num_tokens": 37034260.0,
      "step": 18290
    },
    {
      "entropy": 5.982002782821655,
      "epoch": 1.5987940225465351,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004746753509687824,
      "loss": 5.3885,
      "mean_token_accuracy": 0.19042999744415284,
      "num_tokens": 37045300.0,
      "step": 18295
    },
    {
      "entropy": 5.960189914703369,
      "epoch": 1.5992309708992396,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004746609865394303,
      "loss": 5.3369,
      "mean_token_accuracy": 0.19770914912223816,
      "num_tokens": 37054898.0,
      "step": 18300
    },
    {
      "entropy": 5.913305473327637,
      "epoch": 1.5996679192519445,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004746466182804759,
      "loss": 5.3052,
      "mean_token_accuracy": 0.1988921880722046,
      "num_tokens": 37064519.0,
      "step": 18305
    },
    {
      "entropy": 5.895228385925293,
      "epoch": 1.600104867604649,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004746322461921948,
      "loss": 5.3947,
      "mean_token_accuracy": 0.1941913291811943,
      "num_tokens": 37074967.0,
      "step": 18310
    },
    {
      "entropy": 5.9949079036712645,
      "epoch": 1.600541815957354,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004746178702748626,
      "loss": 5.3935,
      "mean_token_accuracy": 0.18901040703058242,
      "num_tokens": 37084736.0,
      "step": 18315
    },
    {
      "entropy": 5.986842155456543,
      "epoch": 1.6009787643100586,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000474603490528755,
      "loss": 5.3542,
      "mean_token_accuracy": 0.19494495093822478,
      "num_tokens": 37094327.0,
      "step": 18320
    },
    {
      "entropy": 5.89338788986206,
      "epoch": 1.6014157126627633,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004745891069541479,
      "loss": 5.3461,
      "mean_token_accuracy": 0.19611323475837708,
      "num_tokens": 37104099.0,
      "step": 18325
    },
    {
      "entropy": 5.95002932548523,
      "epoch": 1.601852661015468,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047457471955131714,
      "loss": 5.4127,
      "mean_token_accuracy": 0.1897410124540329,
      "num_tokens": 37115337.0,
      "step": 18330
    },
    {
      "entropy": 6.007960605621338,
      "epoch": 1.6022896093681727,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004745603283205387,
      "loss": 5.3543,
      "mean_token_accuracy": 0.19647108316421508,
      "num_tokens": 37125522.0,
      "step": 18335
    },
    {
      "entropy": 5.983588933944702,
      "epoch": 1.6027265577208774,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004745459332620885,
      "loss": 5.4356,
      "mean_token_accuracy": 0.18346741646528245,
      "num_tokens": 37135847.0,
      "step": 18340
    },
    {
      "entropy": 5.886105251312256,
      "epoch": 1.603163506073582,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004745315343762428,
      "loss": 5.3295,
      "mean_token_accuracy": 0.19727766513824463,
      "num_tokens": 37145714.0,
      "step": 18345
    },
    {
      "entropy": 5.944969511032104,
      "epoch": 1.603600454426287,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047451713166327776,
      "loss": 5.3677,
      "mean_token_accuracy": 0.19493056535720826,
      "num_tokens": 37154921.0,
      "step": 18350
    },
    {
      "entropy": 5.958466482162476,
      "epoch": 1.6040374027789914,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047450272512346953,
      "loss": 5.3493,
      "mean_token_accuracy": 0.19823681712150573,
      "num_tokens": 37165303.0,
      "step": 18355
    },
    {
      "entropy": 5.900684928894043,
      "epoch": 1.6044743511316963,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047448831475709454,
      "loss": 5.3335,
      "mean_token_accuracy": 0.19598163813352584,
      "num_tokens": 37175014.0,
      "step": 18360
    },
    {
      "entropy": 5.870084953308106,
      "epoch": 1.6049112994844008,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004744739005644291,
      "loss": 5.3337,
      "mean_token_accuracy": 0.20141167789697648,
      "num_tokens": 37185830.0,
      "step": 18365
    },
    {
      "entropy": 5.936510610580444,
      "epoch": 1.6053482478371057,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004744594825457498,
      "loss": 5.2604,
      "mean_token_accuracy": 0.19633207768201827,
      "num_tokens": 37196264.0,
      "step": 18370
    },
    {
      "entropy": 5.975880146026611,
      "epoch": 1.6057851961898104,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047444506070133305,
      "loss": 5.3969,
      "mean_token_accuracy": 0.19097890853881835,
      "num_tokens": 37206338.0,
      "step": 18375
    },
    {
      "entropy": 5.921074104309082,
      "epoch": 1.606222144542515,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004744306350314555,
      "loss": 5.3381,
      "mean_token_accuracy": 0.201911124587059,
      "num_tokens": 37216231.0,
      "step": 18380
    },
    {
      "entropy": 5.994329309463501,
      "epoch": 1.6066590928952198,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004744162055363939,
      "loss": 5.4585,
      "mean_token_accuracy": 0.18243957459926605,
      "num_tokens": 37227209.0,
      "step": 18385
    },
    {
      "entropy": 6.021249151229858,
      "epoch": 1.6070960412479245,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000474401772216425,
      "loss": 5.4529,
      "mean_token_accuracy": 0.19029203206300735,
      "num_tokens": 37238056.0,
      "step": 18390
    },
    {
      "entropy": 5.975427913665771,
      "epoch": 1.6075329896006292,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004743873350718257,
      "loss": 5.3555,
      "mean_token_accuracy": 0.19515093713998793,
      "num_tokens": 37247864.0,
      "step": 18395
    },
    {
      "entropy": 5.90982027053833,
      "epoch": 1.6079699379533339,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004743728941028727,
      "loss": 5.3066,
      "mean_token_accuracy": 0.1955193966627121,
      "num_tokens": 37256418.0,
      "step": 18400
    },
    {
      "entropy": 5.9319642066955565,
      "epoch": 1.6084068863060388,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004743584493098431,
      "loss": 5.3957,
      "mean_token_accuracy": 0.19182422757148743,
      "num_tokens": 37266540.0,
      "step": 18405
    },
    {
      "entropy": 5.97178692817688,
      "epoch": 1.6088438346587433,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047434400069301414,
      "loss": 5.3744,
      "mean_token_accuracy": 0.19100096374750136,
      "num_tokens": 37277011.0,
      "step": 18410
    },
    {
      "entropy": 5.928132057189941,
      "epoch": 1.6092807830114482,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004743295482526627,
      "loss": 5.3613,
      "mean_token_accuracy": 0.19248826652765275,
      "num_tokens": 37287203.0,
      "step": 18415
    },
    {
      "entropy": 5.946185874938965,
      "epoch": 1.6097177313641526,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004743150919890661,
      "loss": 5.3275,
      "mean_token_accuracy": 0.19885383993387223,
      "num_tokens": 37297293.0,
      "step": 18420
    },
    {
      "entropy": 5.99402642250061,
      "epoch": 1.6101546797168576,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047430063190250155,
      "loss": 5.4126,
      "mean_token_accuracy": 0.18812281787395477,
      "num_tokens": 37307882.0,
      "step": 18425
    },
    {
      "entropy": 5.943950271606445,
      "epoch": 1.610591628069562,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004742861679932464,
      "loss": 5.2849,
      "mean_token_accuracy": 0.20456938296556473,
      "num_tokens": 37317466.0,
      "step": 18430
    },
    {
      "entropy": 5.91976842880249,
      "epoch": 1.611028576422267,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047427170026157823,
      "loss": 5.4408,
      "mean_token_accuracy": 0.1833907037973404,
      "num_tokens": 37327544.0,
      "step": 18435
    },
    {
      "entropy": 5.949385261535644,
      "epoch": 1.6114655247749716,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004742572287077744,
      "loss": 5.4297,
      "mean_token_accuracy": 0.18350424170494078,
      "num_tokens": 37338082.0,
      "step": 18440
    },
    {
      "entropy": 5.986679315567017,
      "epoch": 1.6119024731276763,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047424275333211256,
      "loss": 5.361,
      "mean_token_accuracy": 0.19277454614639283,
      "num_tokens": 37348127.0,
      "step": 18445
    },
    {
      "entropy": 6.016973924636841,
      "epoch": 1.612339421480381,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004742282741348702,
      "loss": 5.4313,
      "mean_token_accuracy": 0.1839734822511673,
      "num_tokens": 37359547.0,
      "step": 18450
    },
    {
      "entropy": 6.0044639110565186,
      "epoch": 1.6127763698330857,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004742137911163253,
      "loss": 5.4731,
      "mean_token_accuracy": 0.1932380512356758,
      "num_tokens": 37369547.0,
      "step": 18455
    },
    {
      "entropy": 5.936719846725464,
      "epoch": 1.6132133181857904,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004741993042767554,
      "loss": 5.4669,
      "mean_token_accuracy": 0.1856607139110565,
      "num_tokens": 37380055.0,
      "step": 18460
    },
    {
      "entropy": 5.957195997238159,
      "epoch": 1.613650266538495,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047418481361643844,
      "loss": 5.4446,
      "mean_token_accuracy": 0.18539049625396728,
      "num_tokens": 37390179.0,
      "step": 18465
    },
    {
      "entropy": 6.004064416885376,
      "epoch": 1.6140872148912,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047417031913565247,
      "loss": 5.3845,
      "mean_token_accuracy": 0.19312862306833267,
      "num_tokens": 37400663.0,
      "step": 18470
    },
    {
      "entropy": 5.998154973983764,
      "epoch": 1.6145241632439045,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004741558208346753,
      "loss": 5.3742,
      "mean_token_accuracy": 0.19533357322216033,
      "num_tokens": 37410988.0,
      "step": 18475
    },
    {
      "entropy": 6.002672863006592,
      "epoch": 1.6149611115966094,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047414131871378526,
      "loss": 5.4421,
      "mean_token_accuracy": 0.18408506661653518,
      "num_tokens": 37422092.0,
      "step": 18480
    },
    {
      "entropy": 5.8937524318695065,
      "epoch": 1.6153980599493138,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004741268127732603,
      "loss": 5.3581,
      "mean_token_accuracy": 0.19370494037866592,
      "num_tokens": 37431637.0,
      "step": 18485
    },
    {
      "entropy": 5.991954708099366,
      "epoch": 1.6158350083020188,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047411230301337877,
      "loss": 5.4517,
      "mean_token_accuracy": 0.1810079663991928,
      "num_tokens": 37442678.0,
      "step": 18490
    },
    {
      "entropy": 6.037888669967652,
      "epoch": 1.6162719566547235,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004740977894344189,
      "loss": 5.3928,
      "mean_token_accuracy": 0.19230659455060958,
      "num_tokens": 37452324.0,
      "step": 18495
    },
    {
      "entropy": 6.0321227550506595,
      "epoch": 1.6167089050074281,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004740832720366591,
      "loss": 5.3265,
      "mean_token_accuracy": 0.18836831897497178,
      "num_tokens": 37462940.0,
      "step": 18500
    },
    {
      "entropy": 5.939798355102539,
      "epoch": 1.6171458533601328,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004740687508203779,
      "loss": 5.4417,
      "mean_token_accuracy": 0.18798425048589706,
      "num_tokens": 37474121.0,
      "step": 18505
    },
    {
      "entropy": 5.983417892456055,
      "epoch": 1.6175828017128375,
      "grad_norm": 1.125,
      "learning_rate": 0.0004740542257858536,
      "loss": 5.4082,
      "mean_token_accuracy": 0.19126715511083603,
      "num_tokens": 37483439.0,
      "step": 18510
    },
    {
      "entropy": 6.00217638015747,
      "epoch": 1.6180197500655422,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047403969693336507,
      "loss": 5.3799,
      "mean_token_accuracy": 0.19378245770931243,
      "num_tokens": 37493804.0,
      "step": 18515
    },
    {
      "entropy": 5.935322380065918,
      "epoch": 1.618456698418247,
      "grad_norm": 1.125,
      "learning_rate": 0.0004740251642631909,
      "loss": 5.2711,
      "mean_token_accuracy": 0.20048149675130844,
      "num_tokens": 37504158.0,
      "step": 18520
    },
    {
      "entropy": 5.92805585861206,
      "epoch": 1.6188936467709518,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004740106277756097,
      "loss": 5.3757,
      "mean_token_accuracy": 0.18845584541559218,
      "num_tokens": 37515069.0,
      "step": 18525
    },
    {
      "entropy": 5.9174963474273685,
      "epoch": 1.6193305951236563,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047399608747090033,
      "loss": 5.2892,
      "mean_token_accuracy": 0.19693935364484788,
      "num_tokens": 37525486.0,
      "step": 18530
    },
    {
      "entropy": 5.983233022689819,
      "epoch": 1.6197675434763612,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047398154334934187,
      "loss": 5.4183,
      "mean_token_accuracy": 0.19025291949510575,
      "num_tokens": 37535616.0,
      "step": 18535
    },
    {
      "entropy": 5.898658418655396,
      "epoch": 1.6202044918290657,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047396699541121296,
      "loss": 5.3117,
      "mean_token_accuracy": 0.2000321477651596,
      "num_tokens": 37546084.0,
      "step": 18540
    },
    {
      "entropy": 5.917351722717285,
      "epoch": 1.6206414401817706,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004739524436567929,
      "loss": 5.3737,
      "mean_token_accuracy": 0.19471491426229476,
      "num_tokens": 37555865.0,
      "step": 18545
    },
    {
      "entropy": 5.898860836029053,
      "epoch": 1.6210783885344753,
      "grad_norm": 1.125,
      "learning_rate": 0.0004739378880863608,
      "loss": 5.3606,
      "mean_token_accuracy": 0.19107866287231445,
      "num_tokens": 37565985.0,
      "step": 18550
    },
    {
      "entropy": 5.9978478908538815,
      "epoch": 1.62151533688718,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004739233287001957,
      "loss": 5.4408,
      "mean_token_accuracy": 0.18791046738624573,
      "num_tokens": 37575887.0,
      "step": 18555
    },
    {
      "entropy": 5.978249645233154,
      "epoch": 1.6219522852398847,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047390876549857695,
      "loss": 5.2923,
      "mean_token_accuracy": 0.19996708929538726,
      "num_tokens": 37585346.0,
      "step": 18560
    },
    {
      "entropy": 5.932308912277222,
      "epoch": 1.6223892335925894,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047389419848178384,
      "loss": 5.3556,
      "mean_token_accuracy": 0.19531055986881257,
      "num_tokens": 37595674.0,
      "step": 18565
    },
    {
      "entropy": 5.877446985244751,
      "epoch": 1.622826181945294,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047387962765009576,
      "loss": 5.3176,
      "mean_token_accuracy": 0.19735063016414642,
      "num_tokens": 37605537.0,
      "step": 18570
    },
    {
      "entropy": 6.018998432159424,
      "epoch": 1.6232631302979987,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047386505300379224,
      "loss": 5.4446,
      "mean_token_accuracy": 0.19094690382480622,
      "num_tokens": 37615801.0,
      "step": 18575
    },
    {
      "entropy": 5.982759618759156,
      "epoch": 1.6237000786507036,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004738504745431528,
      "loss": 5.3302,
      "mean_token_accuracy": 0.20413786470890044,
      "num_tokens": 37625582.0,
      "step": 18580
    },
    {
      "entropy": 5.9113953590393065,
      "epoch": 1.6241370270034081,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047383589226845707,
      "loss": 5.4411,
      "mean_token_accuracy": 0.19009063094854356,
      "num_tokens": 37636075.0,
      "step": 18585
    },
    {
      "entropy": 5.882681798934937,
      "epoch": 1.624573975356113,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00047382130617998477,
      "loss": 5.2978,
      "mean_token_accuracy": 0.1998710736632347,
      "num_tokens": 37645692.0,
      "step": 18590
    },
    {
      "entropy": 5.983788537979126,
      "epoch": 1.6250109237088175,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004738067162780156,
      "loss": 5.3918,
      "mean_token_accuracy": 0.1911251038312912,
      "num_tokens": 37655035.0,
      "step": 18595
    },
    {
      "entropy": 5.961643695831299,
      "epoch": 1.6254478720615224,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004737921225628294,
      "loss": 5.3222,
      "mean_token_accuracy": 0.19065113216638566,
      "num_tokens": 37665829.0,
      "step": 18600
    },
    {
      "entropy": 5.897986936569214,
      "epoch": 1.6258848204142269,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047377752503470613,
      "loss": 5.3487,
      "mean_token_accuracy": 0.2038196787238121,
      "num_tokens": 37676318.0,
      "step": 18605
    },
    {
      "entropy": 5.936885833740234,
      "epoch": 1.6263217687669318,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047376292369392584,
      "loss": 5.3798,
      "mean_token_accuracy": 0.1941177651286125,
      "num_tokens": 37686977.0,
      "step": 18610
    },
    {
      "entropy": 5.941699266433716,
      "epoch": 1.6267587171196365,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047374831854076843,
      "loss": 5.3396,
      "mean_token_accuracy": 0.19862248748540878,
      "num_tokens": 37697004.0,
      "step": 18615
    },
    {
      "entropy": 5.974614191055298,
      "epoch": 1.6271956654723412,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047373370957551423,
      "loss": 5.358,
      "mean_token_accuracy": 0.19609237760305404,
      "num_tokens": 37707613.0,
      "step": 18620
    },
    {
      "entropy": 5.977795505523682,
      "epoch": 1.6276326138250459,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004737190967984433,
      "loss": 5.3529,
      "mean_token_accuracy": 0.19070981740951537,
      "num_tokens": 37717167.0,
      "step": 18625
    },
    {
      "entropy": 5.983043766021728,
      "epoch": 1.6280695621777506,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004737044802098359,
      "loss": 5.46,
      "mean_token_accuracy": 0.18167573511600493,
      "num_tokens": 37727496.0,
      "step": 18630
    },
    {
      "entropy": 5.922756147384644,
      "epoch": 1.6285065105304553,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004736898598099725,
      "loss": 5.361,
      "mean_token_accuracy": 0.1944397881627083,
      "num_tokens": 37736844.0,
      "step": 18635
    },
    {
      "entropy": 5.973057651519776,
      "epoch": 1.62894345888316,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047367523559913347,
      "loss": 5.3832,
      "mean_token_accuracy": 0.18924760371446608,
      "num_tokens": 37746027.0,
      "step": 18640
    },
    {
      "entropy": 5.95055251121521,
      "epoch": 1.6293804072358649,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047366060757759933,
      "loss": 5.3172,
      "mean_token_accuracy": 0.1973396882414818,
      "num_tokens": 37756499.0,
      "step": 18645
    },
    {
      "entropy": 5.921862506866455,
      "epoch": 1.6298173555885693,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047364597574565064,
      "loss": 5.3311,
      "mean_token_accuracy": 0.19770945012569427,
      "num_tokens": 37766533.0,
      "step": 18650
    },
    {
      "entropy": 5.941161870956421,
      "epoch": 1.6302543039412742,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000473631340103568,
      "loss": 5.3584,
      "mean_token_accuracy": 0.2003634199500084,
      "num_tokens": 37775937.0,
      "step": 18655
    },
    {
      "entropy": 5.953216552734375,
      "epoch": 1.6306912522939787,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047361670065163225,
      "loss": 5.3388,
      "mean_token_accuracy": 0.19985182732343673,
      "num_tokens": 37786587.0,
      "step": 18660
    },
    {
      "entropy": 5.937740468978882,
      "epoch": 1.6311282006466836,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047360205739012406,
      "loss": 5.4047,
      "mean_token_accuracy": 0.1948538079857826,
      "num_tokens": 37797200.0,
      "step": 18665
    },
    {
      "entropy": 5.874578666687012,
      "epoch": 1.6315651489993883,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004735874103193244,
      "loss": 5.3507,
      "mean_token_accuracy": 0.20093490034341813,
      "num_tokens": 37806652.0,
      "step": 18670
    },
    {
      "entropy": 5.959755373001099,
      "epoch": 1.632002097352093,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004735727594395141,
      "loss": 5.4366,
      "mean_token_accuracy": 0.17869377881288528,
      "num_tokens": 37817057.0,
      "step": 18675
    },
    {
      "entropy": 5.9682197093963625,
      "epoch": 1.6324390457047977,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047355810475097423,
      "loss": 5.3564,
      "mean_token_accuracy": 0.18949784338474274,
      "num_tokens": 37826216.0,
      "step": 18680
    },
    {
      "entropy": 6.022679090499878,
      "epoch": 1.6328759940575024,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047354344625398596,
      "loss": 5.4332,
      "mean_token_accuracy": 0.18468377590179444,
      "num_tokens": 37835834.0,
      "step": 18685
    },
    {
      "entropy": 5.893345642089844,
      "epoch": 1.633312942410207,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004735287839488302,
      "loss": 5.3466,
      "mean_token_accuracy": 0.19803601056337355,
      "num_tokens": 37846252.0,
      "step": 18690
    },
    {
      "entropy": 5.8882077693939205,
      "epoch": 1.6337498907629118,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047351411783578846,
      "loss": 5.3189,
      "mean_token_accuracy": 0.1969886064529419,
      "num_tokens": 37856020.0,
      "step": 18695
    },
    {
      "entropy": 5.976502084732056,
      "epoch": 1.6341868391156167,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004734994479151418,
      "loss": 5.3485,
      "mean_token_accuracy": 0.1931903541088104,
      "num_tokens": 37866132.0,
      "step": 18700
    },
    {
      "entropy": 5.9602409362792965,
      "epoch": 1.6346237874683212,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004734847741871718,
      "loss": 5.4009,
      "mean_token_accuracy": 0.19420935660600663,
      "num_tokens": 37876779.0,
      "step": 18705
    },
    {
      "entropy": 5.9312550067901615,
      "epoch": 1.635060735821026,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047347009665215983,
      "loss": 5.3725,
      "mean_token_accuracy": 0.19233808815479278,
      "num_tokens": 37886981.0,
      "step": 18710
    },
    {
      "entropy": 5.902159595489502,
      "epoch": 1.6354976841737305,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004734554153103874,
      "loss": 5.3281,
      "mean_token_accuracy": 0.19896911084651947,
      "num_tokens": 37896974.0,
      "step": 18715
    },
    {
      "entropy": 5.999103116989136,
      "epoch": 1.6359346325264355,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047344073016213615,
      "loss": 5.4565,
      "mean_token_accuracy": 0.18812090158462524,
      "num_tokens": 37909020.0,
      "step": 18720
    },
    {
      "entropy": 5.994346237182617,
      "epoch": 1.63637158087914,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047342604120768764,
      "loss": 5.316,
      "mean_token_accuracy": 0.196539506316185,
      "num_tokens": 37919417.0,
      "step": 18725
    },
    {
      "entropy": 5.987838315963745,
      "epoch": 1.6368085292318448,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004734113484473237,
      "loss": 5.4471,
      "mean_token_accuracy": 0.18596353530883789,
      "num_tokens": 37929030.0,
      "step": 18730
    },
    {
      "entropy": 5.978737783432007,
      "epoch": 1.6372454775845495,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004733966518813262,
      "loss": 5.3163,
      "mean_token_accuracy": 0.1962161183357239,
      "num_tokens": 37939157.0,
      "step": 18735
    },
    {
      "entropy": 5.950700807571411,
      "epoch": 1.6376824259372542,
      "grad_norm": 1.125,
      "learning_rate": 0.00047338195150997686,
      "loss": 5.3341,
      "mean_token_accuracy": 0.19084887355566024,
      "num_tokens": 37950544.0,
      "step": 18740
    },
    {
      "entropy": 5.885601282119751,
      "epoch": 1.638119374289959,
      "grad_norm": 1.125,
      "learning_rate": 0.00047336724733355786,
      "loss": 5.3086,
      "mean_token_accuracy": 0.20149940699338914,
      "num_tokens": 37960059.0,
      "step": 18745
    },
    {
      "entropy": 5.938352108001709,
      "epoch": 1.6385563226426636,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047335253935235094,
      "loss": 5.3536,
      "mean_token_accuracy": 0.19937095642089844,
      "num_tokens": 37970669.0,
      "step": 18750
    },
    {
      "entropy": 5.9155913352966305,
      "epoch": 1.6389932709953685,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047333782756663853,
      "loss": 5.3589,
      "mean_token_accuracy": 0.2026721492409706,
      "num_tokens": 37980088.0,
      "step": 18755
    },
    {
      "entropy": 5.961941432952881,
      "epoch": 1.639430219348073,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047332311197670264,
      "loss": 5.4021,
      "mean_token_accuracy": 0.1957631841301918,
      "num_tokens": 37989433.0,
      "step": 18760
    },
    {
      "entropy": 6.004191970825195,
      "epoch": 1.639867167700778,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047330839258282544,
      "loss": 5.3726,
      "mean_token_accuracy": 0.1924380287528038,
      "num_tokens": 37999735.0,
      "step": 18765
    },
    {
      "entropy": 5.935877847671509,
      "epoch": 1.6403041160534824,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004732936693852895,
      "loss": 5.2621,
      "mean_token_accuracy": 0.20466379225254058,
      "num_tokens": 38008515.0,
      "step": 18770
    },
    {
      "entropy": 5.944166803359986,
      "epoch": 1.6407410644061873,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047327894238437696,
      "loss": 5.4152,
      "mean_token_accuracy": 0.18675880432128905,
      "num_tokens": 38018778.0,
      "step": 18775
    },
    {
      "entropy": 5.999041032791138,
      "epoch": 1.6411780127588917,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047326421158037045,
      "loss": 5.4172,
      "mean_token_accuracy": 0.19074871838092805,
      "num_tokens": 38029062.0,
      "step": 18780
    },
    {
      "entropy": 5.89801516532898,
      "epoch": 1.6416149611115967,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047324947697355253,
      "loss": 5.2761,
      "mean_token_accuracy": 0.20116819739341735,
      "num_tokens": 38039945.0,
      "step": 18785
    },
    {
      "entropy": 5.994245433807373,
      "epoch": 1.6420519094643014,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047323473856420573,
      "loss": 5.3915,
      "mean_token_accuracy": 0.1876007467508316,
      "num_tokens": 38050213.0,
      "step": 18790
    },
    {
      "entropy": 5.937754344940186,
      "epoch": 1.642488857817006,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047321999635261276,
      "loss": 5.3251,
      "mean_token_accuracy": 0.19378137588500977,
      "num_tokens": 38060147.0,
      "step": 18795
    },
    {
      "entropy": 5.975181961059571,
      "epoch": 1.6429258061697107,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004732052503390563,
      "loss": 5.3921,
      "mean_token_accuracy": 0.192252816259861,
      "num_tokens": 38070854.0,
      "step": 18800
    },
    {
      "entropy": 5.975453805923462,
      "epoch": 1.6433627545224154,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004731905005238194,
      "loss": 5.4069,
      "mean_token_accuracy": 0.1928907036781311,
      "num_tokens": 38080998.0,
      "step": 18805
    },
    {
      "entropy": 5.951515197753906,
      "epoch": 1.6437997028751201,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047317574690718486,
      "loss": 5.3752,
      "mean_token_accuracy": 0.19472001492977142,
      "num_tokens": 38091421.0,
      "step": 18810
    },
    {
      "entropy": 5.9296046733856205,
      "epoch": 1.6442366512278248,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047316098948943557,
      "loss": 5.3305,
      "mean_token_accuracy": 0.2016698583960533,
      "num_tokens": 38101106.0,
      "step": 18815
    },
    {
      "entropy": 5.9480362892150875,
      "epoch": 1.6446735995805297,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047314622827085466,
      "loss": 5.3456,
      "mean_token_accuracy": 0.18979759216308595,
      "num_tokens": 38111346.0,
      "step": 18820
    },
    {
      "entropy": 6.034021329879761,
      "epoch": 1.6451105479332342,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047313146325172534,
      "loss": 5.45,
      "mean_token_accuracy": 0.18599141538143157,
      "num_tokens": 38121798.0,
      "step": 18825
    },
    {
      "entropy": 5.971277141571045,
      "epoch": 1.645547496285939,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004731166944323307,
      "loss": 5.3958,
      "mean_token_accuracy": 0.18779970109462737,
      "num_tokens": 38134145.0,
      "step": 18830
    },
    {
      "entropy": 5.9830046653747555,
      "epoch": 1.6459844446386436,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047310192181295406,
      "loss": 5.3341,
      "mean_token_accuracy": 0.1939672499895096,
      "num_tokens": 38144213.0,
      "step": 18835
    },
    {
      "entropy": 5.909507322311401,
      "epoch": 1.6464213929913485,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004730871453938788,
      "loss": 5.3474,
      "mean_token_accuracy": 0.19774911403656006,
      "num_tokens": 38154245.0,
      "step": 18840
    },
    {
      "entropy": 5.94331693649292,
      "epoch": 1.6468583413440532,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004730723651753882,
      "loss": 5.375,
      "mean_token_accuracy": 0.1947465181350708,
      "num_tokens": 38163969.0,
      "step": 18845
    },
    {
      "entropy": 5.938192892074585,
      "epoch": 1.6472952896967579,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004730575811577659,
      "loss": 5.3069,
      "mean_token_accuracy": 0.1946923702955246,
      "num_tokens": 38173586.0,
      "step": 18850
    },
    {
      "entropy": 5.9168542385101315,
      "epoch": 1.6477322380494626,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004730427933412954,
      "loss": 5.3097,
      "mean_token_accuracy": 0.19491314142942429,
      "num_tokens": 38182702.0,
      "step": 18855
    },
    {
      "entropy": 5.917989206314087,
      "epoch": 1.6481691864021673,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004730280017262604,
      "loss": 5.3522,
      "mean_token_accuracy": 0.19614384472370147,
      "num_tokens": 38191937.0,
      "step": 18860
    },
    {
      "entropy": 5.894032573699951,
      "epoch": 1.648606134754872,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004730132063129445,
      "loss": 5.316,
      "mean_token_accuracy": 0.19804046899080277,
      "num_tokens": 38201234.0,
      "step": 18865
    },
    {
      "entropy": 5.925087404251099,
      "epoch": 1.6490430831075766,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047299840710163157,
      "loss": 5.3526,
      "mean_token_accuracy": 0.19888584762811662,
      "num_tokens": 38211790.0,
      "step": 18870
    },
    {
      "entropy": 5.979000473022461,
      "epoch": 1.6494800314602815,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047298360409260543,
      "loss": 5.3389,
      "mean_token_accuracy": 0.19509588479995726,
      "num_tokens": 38221946.0,
      "step": 18875
    },
    {
      "entropy": 5.960404586791992,
      "epoch": 1.649916979812986,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047296879728615,
      "loss": 5.3892,
      "mean_token_accuracy": 0.19035616815090178,
      "num_tokens": 38232366.0,
      "step": 18880
    },
    {
      "entropy": 5.979214000701904,
      "epoch": 1.650353928165691,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047295398668254933,
      "loss": 5.425,
      "mean_token_accuracy": 0.18807369768619536,
      "num_tokens": 38242504.0,
      "step": 18885
    },
    {
      "entropy": 6.040834093093872,
      "epoch": 1.6507908765183954,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004729391722820875,
      "loss": 5.4147,
      "mean_token_accuracy": 0.18911001682281495,
      "num_tokens": 38251916.0,
      "step": 18890
    },
    {
      "entropy": 5.9414527893066404,
      "epoch": 1.6512278248711003,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047292435408504855,
      "loss": 5.306,
      "mean_token_accuracy": 0.20286328196525574,
      "num_tokens": 38262438.0,
      "step": 18895
    },
    {
      "entropy": 5.842369461059571,
      "epoch": 1.6516647732238048,
      "grad_norm": 1.125,
      "learning_rate": 0.0004729095320917169,
      "loss": 5.3018,
      "mean_token_accuracy": 0.2098215863108635,
      "num_tokens": 38273095.0,
      "step": 18900
    },
    {
      "entropy": 5.960775899887085,
      "epoch": 1.6521017215765097,
      "grad_norm": 1.125,
      "learning_rate": 0.0004728947063023766,
      "loss": 5.3645,
      "mean_token_accuracy": 0.18719100207090378,
      "num_tokens": 38282827.0,
      "step": 18905
    },
    {
      "entropy": 6.036962080001831,
      "epoch": 1.6525386699292144,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004728798767173122,
      "loss": 5.3249,
      "mean_token_accuracy": 0.19393866062164306,
      "num_tokens": 38292827.0,
      "step": 18910
    },
    {
      "entropy": 5.959285640716553,
      "epoch": 1.652975618281919,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047286504333680804,
      "loss": 5.387,
      "mean_token_accuracy": 0.18669189363718033,
      "num_tokens": 38302658.0,
      "step": 18915
    },
    {
      "entropy": 5.926571321487427,
      "epoch": 1.6534125666346238,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004728502061611488,
      "loss": 5.35,
      "mean_token_accuracy": 0.188293194770813,
      "num_tokens": 38312154.0,
      "step": 18920
    },
    {
      "entropy": 5.981301355361938,
      "epoch": 1.6538495149873285,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047283536519061875,
      "loss": 5.3396,
      "mean_token_accuracy": 0.1971279874444008,
      "num_tokens": 38321317.0,
      "step": 18925
    },
    {
      "entropy": 5.965839147567749,
      "epoch": 1.6542864633400332,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047282052042550276,
      "loss": 5.3456,
      "mean_token_accuracy": 0.18779272437095643,
      "num_tokens": 38331632.0,
      "step": 18930
    },
    {
      "entropy": 5.947419548034668,
      "epoch": 1.6547234116927378,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047280567186608563,
      "loss": 5.4132,
      "mean_token_accuracy": 0.18169253915548325,
      "num_tokens": 38342072.0,
      "step": 18935
    },
    {
      "entropy": 6.012845993041992,
      "epoch": 1.6551603600454428,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000472790819512652,
      "loss": 5.3587,
      "mean_token_accuracy": 0.19356156289577484,
      "num_tokens": 38351105.0,
      "step": 18940
    },
    {
      "entropy": 6.005233860015869,
      "epoch": 1.6555973083981472,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004727759633654868,
      "loss": 5.4377,
      "mean_token_accuracy": 0.1839393451809883,
      "num_tokens": 38361521.0,
      "step": 18945
    },
    {
      "entropy": 5.929498386383057,
      "epoch": 1.6560342567508521,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004727611034248751,
      "loss": 5.301,
      "mean_token_accuracy": 0.19002585262060165,
      "num_tokens": 38370720.0,
      "step": 18950
    },
    {
      "entropy": 5.928011703491211,
      "epoch": 1.6564712051035566,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004727462396911017,
      "loss": 5.3916,
      "mean_token_accuracy": 0.19327943325042723,
      "num_tokens": 38380434.0,
      "step": 18955
    },
    {
      "entropy": 5.87253770828247,
      "epoch": 1.6569081534562615,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047273137216445176,
      "loss": 5.2785,
      "mean_token_accuracy": 0.21148537546396257,
      "num_tokens": 38390503.0,
      "step": 18960
    },
    {
      "entropy": 5.9877503395080565,
      "epoch": 1.6573451018089662,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004727165008452106,
      "loss": 5.4567,
      "mean_token_accuracy": 0.1855609118938446,
      "num_tokens": 38401757.0,
      "step": 18965
    },
    {
      "entropy": 5.958076620101929,
      "epoch": 1.657782050161671,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004727016257336633,
      "loss": 5.3879,
      "mean_token_accuracy": 0.19409504532814026,
      "num_tokens": 38412130.0,
      "step": 18970
    },
    {
      "entropy": 5.922611856460572,
      "epoch": 1.6582189985143756,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047268674683009526,
      "loss": 5.3399,
      "mean_token_accuracy": 0.1938109204173088,
      "num_tokens": 38422155.0,
      "step": 18975
    },
    {
      "entropy": 5.9239905834197994,
      "epoch": 1.6586559468670803,
      "grad_norm": 1.125,
      "learning_rate": 0.00047267186413479185,
      "loss": 5.2951,
      "mean_token_accuracy": 0.19954753965139388,
      "num_tokens": 38431621.0,
      "step": 18980
    },
    {
      "entropy": 6.021316480636597,
      "epoch": 1.659092895219785,
      "grad_norm": 1.0,
      "learning_rate": 0.00047265697764803843,
      "loss": 5.4323,
      "mean_token_accuracy": 0.1843567579984665,
      "num_tokens": 38442354.0,
      "step": 18985
    },
    {
      "entropy": 5.930760097503662,
      "epoch": 1.6595298435724897,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047264208737012064,
      "loss": 5.3753,
      "mean_token_accuracy": 0.1920338824391365,
      "num_tokens": 38452932.0,
      "step": 18990
    },
    {
      "entropy": 5.940264558792114,
      "epoch": 1.6599667919251946,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004726271933013241,
      "loss": 5.3665,
      "mean_token_accuracy": 0.1931345671415329,
      "num_tokens": 38464442.0,
      "step": 18995
    },
    {
      "entropy": 5.988611316680908,
      "epoch": 1.660403740277899,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047261229544193434,
      "loss": 5.3737,
      "mean_token_accuracy": 0.18998831063508986,
      "num_tokens": 38475718.0,
      "step": 19000
    },
    {
      "entropy": 5.956876564025879,
      "epoch": 1.660840688630604,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047259739379223726,
      "loss": 5.3445,
      "mean_token_accuracy": 0.1948675200343132,
      "num_tokens": 38485091.0,
      "step": 19005
    },
    {
      "entropy": 5.9416767120361325,
      "epoch": 1.6612776369833084,
      "grad_norm": 1.125,
      "learning_rate": 0.00047258248835251867,
      "loss": 5.3626,
      "mean_token_accuracy": 0.19392125010490419,
      "num_tokens": 38495748.0,
      "step": 19010
    },
    {
      "entropy": 6.014583730697632,
      "epoch": 1.6617145853360134,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047256757912306435,
      "loss": 5.4301,
      "mean_token_accuracy": 0.19182710498571395,
      "num_tokens": 38506152.0,
      "step": 19015
    },
    {
      "entropy": 5.986705017089844,
      "epoch": 1.662151533688718,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004725526661041604,
      "loss": 5.3114,
      "mean_token_accuracy": 0.19768983423709868,
      "num_tokens": 38515722.0,
      "step": 19020
    },
    {
      "entropy": 5.908015871047974,
      "epoch": 1.6625884820414227,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004725377492960927,
      "loss": 5.3391,
      "mean_token_accuracy": 0.1968207135796547,
      "num_tokens": 38526270.0,
      "step": 19025
    },
    {
      "entropy": 5.94363226890564,
      "epoch": 1.6630254303941274,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004725228286991475,
      "loss": 5.333,
      "mean_token_accuracy": 0.19104491919279099,
      "num_tokens": 38537436.0,
      "step": 19030
    },
    {
      "entropy": 5.986108636856079,
      "epoch": 1.6634623787468321,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004725079043136111,
      "loss": 5.3553,
      "mean_token_accuracy": 0.1987237364053726,
      "num_tokens": 38547298.0,
      "step": 19035
    },
    {
      "entropy": 5.964492130279541,
      "epoch": 1.6638993270995368,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004724929761397694,
      "loss": 5.3642,
      "mean_token_accuracy": 0.1899416595697403,
      "num_tokens": 38556465.0,
      "step": 19040
    },
    {
      "entropy": 5.9178753852844235,
      "epoch": 1.6643362754522415,
      "grad_norm": 1.03125,
      "learning_rate": 0.000472478044177909,
      "loss": 5.3796,
      "mean_token_accuracy": 0.19626868963241578,
      "num_tokens": 38568009.0,
      "step": 19045
    },
    {
      "entropy": 5.962144470214843,
      "epoch": 1.6647732238049464,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004724631084283163,
      "loss": 5.4168,
      "mean_token_accuracy": 0.19267766326665878,
      "num_tokens": 38578778.0,
      "step": 19050
    },
    {
      "entropy": 5.900454902648926,
      "epoch": 1.6652101721576509,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047244816889127763,
      "loss": 5.3232,
      "mean_token_accuracy": 0.19941655844449996,
      "num_tokens": 38588991.0,
      "step": 19055
    },
    {
      "entropy": 5.938519477844238,
      "epoch": 1.6656471205103558,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004724332255670797,
      "loss": 5.3187,
      "mean_token_accuracy": 0.20000976622104644,
      "num_tokens": 38598800.0,
      "step": 19060
    },
    {
      "entropy": 5.918172311782837,
      "epoch": 1.6660840688630603,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047241827845600896,
      "loss": 5.3625,
      "mean_token_accuracy": 0.19213091880083083,
      "num_tokens": 38608680.0,
      "step": 19065
    },
    {
      "entropy": 5.924151182174683,
      "epoch": 1.6665210172157652,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047240332755835217,
      "loss": 5.3683,
      "mean_token_accuracy": 0.19194802939891814,
      "num_tokens": 38618402.0,
      "step": 19070
    },
    {
      "entropy": 6.058193206787109,
      "epoch": 1.6669579655684696,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047238837287439625,
      "loss": 5.3698,
      "mean_token_accuracy": 0.19200772047042847,
      "num_tokens": 38628370.0,
      "step": 19075
    },
    {
      "entropy": 5.987363290786743,
      "epoch": 1.6673949139211746,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004723734144044278,
      "loss": 5.313,
      "mean_token_accuracy": 0.20074476897716523,
      "num_tokens": 38638138.0,
      "step": 19080
    },
    {
      "entropy": 5.979499673843383,
      "epoch": 1.6678318622738793,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047235845214873385,
      "loss": 5.4105,
      "mean_token_accuracy": 0.19160065650939942,
      "num_tokens": 38649412.0,
      "step": 19085
    },
    {
      "entropy": 5.960430383682251,
      "epoch": 1.668268810626584,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047234348610760145,
      "loss": 5.3228,
      "mean_token_accuracy": 0.18812565952539445,
      "num_tokens": 38660048.0,
      "step": 19090
    },
    {
      "entropy": 5.941126585006714,
      "epoch": 1.6687057589792886,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004723285162813175,
      "loss": 5.3162,
      "mean_token_accuracy": 0.20570667386054992,
      "num_tokens": 38671246.0,
      "step": 19095
    },
    {
      "entropy": 5.958766889572144,
      "epoch": 1.6691427073319933,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004723135426701693,
      "loss": 5.3713,
      "mean_token_accuracy": 0.19445665329694747,
      "num_tokens": 38680880.0,
      "step": 19100
    },
    {
      "entropy": 5.910894155502319,
      "epoch": 1.669579655684698,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004722985652744438,
      "loss": 5.4316,
      "mean_token_accuracy": 0.19288178086280822,
      "num_tokens": 38691440.0,
      "step": 19105
    },
    {
      "entropy": 5.997533702850342,
      "epoch": 1.6700166040374027,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047228358409442863,
      "loss": 5.3509,
      "mean_token_accuracy": 0.19898116290569307,
      "num_tokens": 38700846.0,
      "step": 19110
    },
    {
      "entropy": 6.0116127014160154,
      "epoch": 1.6704535523901076,
      "grad_norm": 1.125,
      "learning_rate": 0.00047226859913041084,
      "loss": 5.3786,
      "mean_token_accuracy": 0.19844187349081038,
      "num_tokens": 38712130.0,
      "step": 19115
    },
    {
      "entropy": 5.992485380172729,
      "epoch": 1.670890500742812,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047225361038267795,
      "loss": 5.4402,
      "mean_token_accuracy": 0.19342312961816788,
      "num_tokens": 38722482.0,
      "step": 19120
    },
    {
      "entropy": 5.893723964691162,
      "epoch": 1.671327449095517,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047223861785151744,
      "loss": 5.3344,
      "mean_token_accuracy": 0.19732031524181365,
      "num_tokens": 38731936.0,
      "step": 19125
    },
    {
      "entropy": 5.9072731018066404,
      "epoch": 1.6717643974482215,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004722236215372169,
      "loss": 5.2714,
      "mean_token_accuracy": 0.19629853814840317,
      "num_tokens": 38741444.0,
      "step": 19130
    },
    {
      "entropy": 6.073374557495117,
      "epoch": 1.6722013458009264,
      "grad_norm": 1.078125,
      "learning_rate": 0.000472208621440064,
      "loss": 5.4243,
      "mean_token_accuracy": 0.17860183715820313,
      "num_tokens": 38751454.0,
      "step": 19135
    },
    {
      "entropy": 6.01363434791565,
      "epoch": 1.672638294153631,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004721936175603463,
      "loss": 5.3819,
      "mean_token_accuracy": 0.19063351154327393,
      "num_tokens": 38761373.0,
      "step": 19140
    },
    {
      "entropy": 5.980135726928711,
      "epoch": 1.6730752425063358,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047217860989835184,
      "loss": 5.4341,
      "mean_token_accuracy": 0.19570788145065307,
      "num_tokens": 38772047.0,
      "step": 19145
    },
    {
      "entropy": 5.9762835025787355,
      "epoch": 1.6735121908590405,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004721635984543682,
      "loss": 5.3001,
      "mean_token_accuracy": 0.1967429041862488,
      "num_tokens": 38781393.0,
      "step": 19150
    },
    {
      "entropy": 5.982134866714477,
      "epoch": 1.6739491392117452,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004721485832286835,
      "loss": 5.3336,
      "mean_token_accuracy": 0.19541867524385453,
      "num_tokens": 38791104.0,
      "step": 19155
    },
    {
      "entropy": 5.907048273086548,
      "epoch": 1.6743860875644498,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004721335642215856,
      "loss": 5.4067,
      "mean_token_accuracy": 0.19334897696971892,
      "num_tokens": 38801855.0,
      "step": 19160
    },
    {
      "entropy": 5.933858966827392,
      "epoch": 1.6748230359171545,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004721185414333627,
      "loss": 5.2905,
      "mean_token_accuracy": 0.20036801248788833,
      "num_tokens": 38811983.0,
      "step": 19165
    },
    {
      "entropy": 5.968256092071533,
      "epoch": 1.6752599842698594,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004721035148643029,
      "loss": 5.346,
      "mean_token_accuracy": 0.18927782624959946,
      "num_tokens": 38821762.0,
      "step": 19170
    },
    {
      "entropy": 5.888788318634033,
      "epoch": 1.675696932622564,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004720884845146944,
      "loss": 5.3157,
      "mean_token_accuracy": 0.19734769612550734,
      "num_tokens": 38831840.0,
      "step": 19175
    },
    {
      "entropy": 5.8977960586547855,
      "epoch": 1.6761338809752688,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047207345038482545,
      "loss": 5.3,
      "mean_token_accuracy": 0.20271721929311753,
      "num_tokens": 38841506.0,
      "step": 19180
    },
    {
      "entropy": 5.969515085220337,
      "epoch": 1.6765708293279733,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004720584124749844,
      "loss": 5.3819,
      "mean_token_accuracy": 0.19514193683862685,
      "num_tokens": 38851793.0,
      "step": 19185
    },
    {
      "entropy": 5.884385251998902,
      "epoch": 1.6770077776806782,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047204337078545985,
      "loss": 5.3139,
      "mean_token_accuracy": 0.2012418732047081,
      "num_tokens": 38861958.0,
      "step": 19190
    },
    {
      "entropy": 5.943662691116333,
      "epoch": 1.677444726033383,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004720283253165402,
      "loss": 5.322,
      "mean_token_accuracy": 0.20048547983169557,
      "num_tokens": 38872181.0,
      "step": 19195
    },
    {
      "entropy": 5.973420143127441,
      "epoch": 1.6778816743860876,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047201327606851394,
      "loss": 5.3673,
      "mean_token_accuracy": 0.19793445616960526,
      "num_tokens": 38882423.0,
      "step": 19200
    },
    {
      "entropy": 5.982385683059692,
      "epoch": 1.6783186227387923,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004719982230416699,
      "loss": 5.3575,
      "mean_token_accuracy": 0.19366620630025863,
      "num_tokens": 38892833.0,
      "step": 19205
    },
    {
      "entropy": 5.994168043136597,
      "epoch": 1.678755571091497,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004719831662362967,
      "loss": 5.4235,
      "mean_token_accuracy": 0.18823584914207458,
      "num_tokens": 38904037.0,
      "step": 19210
    },
    {
      "entropy": 5.937631702423095,
      "epoch": 1.6791925194442017,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004719681056526831,
      "loss": 5.327,
      "mean_token_accuracy": 0.20106342285871506,
      "num_tokens": 38914194.0,
      "step": 19215
    },
    {
      "entropy": 5.918640995025635,
      "epoch": 1.6796294677969064,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047195304129111804,
      "loss": 5.3858,
      "mean_token_accuracy": 0.18994134962558745,
      "num_tokens": 38925354.0,
      "step": 19220
    },
    {
      "entropy": 6.023126125335693,
      "epoch": 1.6800664161496113,
      "grad_norm": 1.0,
      "learning_rate": 0.00047193797315189047,
      "loss": 5.4072,
      "mean_token_accuracy": 0.19175504446029662,
      "num_tokens": 38936170.0,
      "step": 19225
    },
    {
      "entropy": 6.011358737945557,
      "epoch": 1.6805033645023157,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004719229012352894,
      "loss": 5.3827,
      "mean_token_accuracy": 0.19631920605897904,
      "num_tokens": 38945316.0,
      "step": 19230
    },
    {
      "entropy": 5.983366870880127,
      "epoch": 1.6809403128550207,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047190782554160386,
      "loss": 5.3805,
      "mean_token_accuracy": 0.1906387373805046,
      "num_tokens": 38955444.0,
      "step": 19235
    },
    {
      "entropy": 5.949612760543824,
      "epoch": 1.6813772612077251,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004718927460711231,
      "loss": 5.3432,
      "mean_token_accuracy": 0.19840542525053023,
      "num_tokens": 38965202.0,
      "step": 19240
    },
    {
      "entropy": 5.925805902481079,
      "epoch": 1.68181420956043,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047187766282413623,
      "loss": 5.3688,
      "mean_token_accuracy": 0.19732916951179505,
      "num_tokens": 38975822.0,
      "step": 19245
    },
    {
      "entropy": 5.940431880950928,
      "epoch": 1.6822511579131345,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004718625758009327,
      "loss": 5.348,
      "mean_token_accuracy": 0.1928483471274376,
      "num_tokens": 38985890.0,
      "step": 19250
    },
    {
      "entropy": 5.93147497177124,
      "epoch": 1.6826881062658394,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047184748500180175,
      "loss": 5.3203,
      "mean_token_accuracy": 0.20132793337106705,
      "num_tokens": 38996374.0,
      "step": 19255
    },
    {
      "entropy": 5.949298620223999,
      "epoch": 1.6831250546185441,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004718323904270329,
      "loss": 5.3454,
      "mean_token_accuracy": 0.19260693341493607,
      "num_tokens": 39006371.0,
      "step": 19260
    },
    {
      "entropy": 5.942886686325073,
      "epoch": 1.6835620029712488,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004718172920769157,
      "loss": 5.3127,
      "mean_token_accuracy": 0.20009758174419404,
      "num_tokens": 39015514.0,
      "step": 19265
    },
    {
      "entropy": 5.981655502319336,
      "epoch": 1.6839989513239535,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004718021899517397,
      "loss": 5.3623,
      "mean_token_accuracy": 0.19803841263055802,
      "num_tokens": 39025700.0,
      "step": 19270
    },
    {
      "entropy": 5.920642995834351,
      "epoch": 1.6844358996766582,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047178708405179454,
      "loss": 5.3489,
      "mean_token_accuracy": 0.19966394156217576,
      "num_tokens": 39036157.0,
      "step": 19275
    },
    {
      "entropy": 5.992671775817871,
      "epoch": 1.6848728480293629,
      "grad_norm": 1.5,
      "learning_rate": 0.00047177197437737003,
      "loss": 5.4796,
      "mean_token_accuracy": 0.18412919193506241,
      "num_tokens": 39046044.0,
      "step": 19280
    },
    {
      "entropy": 5.9859947681427,
      "epoch": 1.6853097963820676,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004717568609287559,
      "loss": 5.4082,
      "mean_token_accuracy": 0.18998144268989564,
      "num_tokens": 39056034.0,
      "step": 19285
    },
    {
      "entropy": 5.929852676391602,
      "epoch": 1.6857467447347725,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004717417437062421,
      "loss": 5.3564,
      "mean_token_accuracy": 0.19433318972587585,
      "num_tokens": 39065687.0,
      "step": 19290
    },
    {
      "entropy": 5.994988012313843,
      "epoch": 1.686183693087477,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047172662271011855,
      "loss": 5.3604,
      "mean_token_accuracy": 0.1892445370554924,
      "num_tokens": 39077107.0,
      "step": 19295
    },
    {
      "entropy": 5.9737073421478275,
      "epoch": 1.6866206414401819,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047171149794067533,
      "loss": 5.357,
      "mean_token_accuracy": 0.19707521051168442,
      "num_tokens": 39087220.0,
      "step": 19300
    },
    {
      "entropy": 5.959347057342529,
      "epoch": 1.6870575897928863,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004716963693982025,
      "loss": 5.321,
      "mean_token_accuracy": 0.19903532266616822,
      "num_tokens": 39097706.0,
      "step": 19305
    },
    {
      "entropy": 5.97020697593689,
      "epoch": 1.6874945381455912,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004716812370829902,
      "loss": 5.3713,
      "mean_token_accuracy": 0.1886274591088295,
      "num_tokens": 39108044.0,
      "step": 19310
    },
    {
      "entropy": 5.94809136390686,
      "epoch": 1.687931486498296,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004716661009953287,
      "loss": 5.3816,
      "mean_token_accuracy": 0.18789347857236863,
      "num_tokens": 39116917.0,
      "step": 19315
    },
    {
      "entropy": 5.940630626678467,
      "epoch": 1.6883684348510006,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004716509611355084,
      "loss": 5.3479,
      "mean_token_accuracy": 0.20196391195058822,
      "num_tokens": 39126770.0,
      "step": 19320
    },
    {
      "entropy": 5.937542963027954,
      "epoch": 1.6888053832037053,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047163581750381956,
      "loss": 5.3649,
      "mean_token_accuracy": 0.19062499105930328,
      "num_tokens": 39137406.0,
      "step": 19325
    },
    {
      "entropy": 5.954419040679932,
      "epoch": 1.68924233155641,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047162067010055275,
      "loss": 5.3473,
      "mean_token_accuracy": 0.19816270768642424,
      "num_tokens": 39146950.0,
      "step": 19330
    },
    {
      "entropy": 5.978524589538575,
      "epoch": 1.6896792799091147,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004716055189259984,
      "loss": 5.2961,
      "mean_token_accuracy": 0.1967988446354866,
      "num_tokens": 39157382.0,
      "step": 19335
    },
    {
      "entropy": 5.948214292526245,
      "epoch": 1.6901162282618194,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047159036398044727,
      "loss": 5.3419,
      "mean_token_accuracy": 0.19455720484256744,
      "num_tokens": 39166641.0,
      "step": 19340
    },
    {
      "entropy": 5.948406505584717,
      "epoch": 1.6905531766145243,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004715752052641899,
      "loss": 5.4024,
      "mean_token_accuracy": 0.19691688120365142,
      "num_tokens": 39176731.0,
      "step": 19345
    },
    {
      "entropy": 6.0083435535430905,
      "epoch": 1.6909901249672288,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004715600427775171,
      "loss": 5.4188,
      "mean_token_accuracy": 0.18883825689554215,
      "num_tokens": 39187202.0,
      "step": 19350
    },
    {
      "entropy": 6.000610446929931,
      "epoch": 1.6914270733199337,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004715448765207196,
      "loss": 5.4036,
      "mean_token_accuracy": 0.19425127953290938,
      "num_tokens": 39197302.0,
      "step": 19355
    },
    {
      "entropy": 5.994425249099732,
      "epoch": 1.6918640216726382,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004715297064940885,
      "loss": 5.3803,
      "mean_token_accuracy": 0.19146058559417725,
      "num_tokens": 39208382.0,
      "step": 19360
    },
    {
      "entropy": 5.96797194480896,
      "epoch": 1.692300970025343,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004715145326979146,
      "loss": 5.3563,
      "mean_token_accuracy": 0.19052429050207137,
      "num_tokens": 39218529.0,
      "step": 19365
    },
    {
      "entropy": 5.926676416397095,
      "epoch": 1.6927379183780478,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047149935513248896,
      "loss": 5.3936,
      "mean_token_accuracy": 0.1927228406071663,
      "num_tokens": 39228976.0,
      "step": 19370
    },
    {
      "entropy": 5.924795436859131,
      "epoch": 1.6931748667307525,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004714841737981028,
      "loss": 5.2947,
      "mean_token_accuracy": 0.19764017760753633,
      "num_tokens": 39238921.0,
      "step": 19375
    },
    {
      "entropy": 5.983436059951782,
      "epoch": 1.6936118150834572,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047146898869504717,
      "loss": 5.3027,
      "mean_token_accuracy": 0.1901417061686516,
      "num_tokens": 39247978.0,
      "step": 19380
    },
    {
      "entropy": 5.92349944114685,
      "epoch": 1.6940487634361618,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047145379982361336,
      "loss": 5.358,
      "mean_token_accuracy": 0.19440321922302245,
      "num_tokens": 39257487.0,
      "step": 19385
    },
    {
      "entropy": 5.933723115921021,
      "epoch": 1.6944857117888665,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047143860718409285,
      "loss": 5.3252,
      "mean_token_accuracy": 0.1926820009946823,
      "num_tokens": 39267819.0,
      "step": 19390
    },
    {
      "entropy": 6.015988397598266,
      "epoch": 1.6949226601415712,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047142341077677685,
      "loss": 5.4464,
      "mean_token_accuracy": 0.18915971964597703,
      "num_tokens": 39277673.0,
      "step": 19395
    },
    {
      "entropy": 5.959174966812133,
      "epoch": 1.6953596084942761,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004714082106019569,
      "loss": 5.2816,
      "mean_token_accuracy": 0.19883922338485718,
      "num_tokens": 39287340.0,
      "step": 19400
    },
    {
      "entropy": 5.976446437835693,
      "epoch": 1.6957965568469806,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004713930066599246,
      "loss": 5.4084,
      "mean_token_accuracy": 0.19904834628105164,
      "num_tokens": 39298277.0,
      "step": 19405
    },
    {
      "entropy": 5.9894740104675295,
      "epoch": 1.6962335051996855,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047137779895097145,
      "loss": 5.3703,
      "mean_token_accuracy": 0.19155762642621993,
      "num_tokens": 39307996.0,
      "step": 19410
    },
    {
      "entropy": 5.949122571945191,
      "epoch": 1.69667045355239,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047136258747538926,
      "loss": 5.3617,
      "mean_token_accuracy": 0.19648578017950058,
      "num_tokens": 39317327.0,
      "step": 19415
    },
    {
      "entropy": 5.931555080413818,
      "epoch": 1.697107401905095,
      "grad_norm": 1.125,
      "learning_rate": 0.0004713473722334697,
      "loss": 5.3247,
      "mean_token_accuracy": 0.20154107362031937,
      "num_tokens": 39326728.0,
      "step": 19420
    },
    {
      "entropy": 5.963466596603394,
      "epoch": 1.6975443502577994,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047133215322550464,
      "loss": 5.4728,
      "mean_token_accuracy": 0.18820563852787017,
      "num_tokens": 39336606.0,
      "step": 19425
    },
    {
      "entropy": 6.001468849182129,
      "epoch": 1.6979812986105043,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004713169304517861,
      "loss": 5.3861,
      "mean_token_accuracy": 0.18377253860235215,
      "num_tokens": 39345622.0,
      "step": 19430
    },
    {
      "entropy": 5.9772333145141605,
      "epoch": 1.698418246963209,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047130170391260597,
      "loss": 5.3533,
      "mean_token_accuracy": 0.19372607171535491,
      "num_tokens": 39356868.0,
      "step": 19435
    },
    {
      "entropy": 5.920555305480957,
      "epoch": 1.6988551953159137,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004712864736082562,
      "loss": 5.2815,
      "mean_token_accuracy": 0.20022240281105042,
      "num_tokens": 39366866.0,
      "step": 19440
    },
    {
      "entropy": 5.932598114013672,
      "epoch": 1.6992921436686184,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047127123953902907,
      "loss": 5.3406,
      "mean_token_accuracy": 0.19484468102455138,
      "num_tokens": 39377303.0,
      "step": 19445
    },
    {
      "entropy": 5.933976554870606,
      "epoch": 1.699729092021323,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004712560017052167,
      "loss": 5.3025,
      "mean_token_accuracy": 0.2042732909321785,
      "num_tokens": 39387191.0,
      "step": 19450
    },
    {
      "entropy": 5.871959829330445,
      "epoch": 1.7001660403740277,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004712407601071114,
      "loss": 5.3198,
      "mean_token_accuracy": 0.20320574343204498,
      "num_tokens": 39397416.0,
      "step": 19455
    },
    {
      "entropy": 5.932036638259888,
      "epoch": 1.7006029887267324,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047122551474500555,
      "loss": 5.3558,
      "mean_token_accuracy": 0.19253724366426467,
      "num_tokens": 39407855.0,
      "step": 19460
    },
    {
      "entropy": 5.969138526916504,
      "epoch": 1.7010399370794373,
      "grad_norm": 1.25,
      "learning_rate": 0.00047121026561919137,
      "loss": 5.4189,
      "mean_token_accuracy": 0.19167513400316238,
      "num_tokens": 39417735.0,
      "step": 19465
    },
    {
      "entropy": 5.975672578811645,
      "epoch": 1.7014768854321418,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004711950127299616,
      "loss": 5.3293,
      "mean_token_accuracy": 0.19597136378288268,
      "num_tokens": 39427641.0,
      "step": 19470
    },
    {
      "entropy": 5.9086277961730955,
      "epoch": 1.7019138337848467,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004711797560776087,
      "loss": 5.2606,
      "mean_token_accuracy": 0.20626960694789886,
      "num_tokens": 39437372.0,
      "step": 19475
    },
    {
      "entropy": 6.004025268554687,
      "epoch": 1.7023507821375512,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047116449566242516,
      "loss": 5.3557,
      "mean_token_accuracy": 0.20205399692058562,
      "num_tokens": 39447800.0,
      "step": 19480
    },
    {
      "entropy": 5.987491416931152,
      "epoch": 1.7027877304902561,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004711492314847038,
      "loss": 5.3829,
      "mean_token_accuracy": 0.1952880084514618,
      "num_tokens": 39457310.0,
      "step": 19485
    },
    {
      "entropy": 5.919356727600098,
      "epoch": 1.7032246788429608,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047113396354473753,
      "loss": 5.3781,
      "mean_token_accuracy": 0.19549377113580704,
      "num_tokens": 39468579.0,
      "step": 19490
    },
    {
      "entropy": 5.98718671798706,
      "epoch": 1.7036616271956655,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000471118691842819,
      "loss": 5.3061,
      "mean_token_accuracy": 0.20133427232503892,
      "num_tokens": 39479027.0,
      "step": 19495
    },
    {
      "entropy": 5.922672367095947,
      "epoch": 1.7040985755483702,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047110341637924126,
      "loss": 5.3237,
      "mean_token_accuracy": 0.20764381140470506,
      "num_tokens": 39488651.0,
      "step": 19500
    },
    {
      "entropy": 5.95537052154541,
      "epoch": 1.7045355239010749,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004710881371542972,
      "loss": 5.3799,
      "mean_token_accuracy": 0.1921380266547203,
      "num_tokens": 39498633.0,
      "step": 19505
    },
    {
      "entropy": 5.891906452178955,
      "epoch": 1.7049724722537796,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047107285416827984,
      "loss": 5.2859,
      "mean_token_accuracy": 0.2038091629743576,
      "num_tokens": 39509351.0,
      "step": 19510
    },
    {
      "entropy": 5.884768724441528,
      "epoch": 1.7054094206064843,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004710575674214824,
      "loss": 5.334,
      "mean_token_accuracy": 0.20061837285757064,
      "num_tokens": 39519821.0,
      "step": 19515
    },
    {
      "entropy": 5.948180675506592,
      "epoch": 1.7058463689591892,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004710422769141982,
      "loss": 5.3713,
      "mean_token_accuracy": 0.19321709126234055,
      "num_tokens": 39530040.0,
      "step": 19520
    },
    {
      "entropy": 5.957509899139405,
      "epoch": 1.7062833173118936,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004710269826467203,
      "loss": 5.352,
      "mean_token_accuracy": 0.19331758916378022,
      "num_tokens": 39540669.0,
      "step": 19525
    },
    {
      "entropy": 5.996422052383423,
      "epoch": 1.7067202656645986,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047101168461934225,
      "loss": 5.3707,
      "mean_token_accuracy": 0.1921850323677063,
      "num_tokens": 39550651.0,
      "step": 19530
    },
    {
      "entropy": 5.9648847579956055,
      "epoch": 1.707157214017303,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004709963828323573,
      "loss": 5.3247,
      "mean_token_accuracy": 0.19209285974502563,
      "num_tokens": 39561748.0,
      "step": 19535
    },
    {
      "entropy": 5.960284042358398,
      "epoch": 1.707594162370008,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004709810772860591,
      "loss": 5.299,
      "mean_token_accuracy": 0.19776644855737685,
      "num_tokens": 39571328.0,
      "step": 19540
    },
    {
      "entropy": 5.873583030700684,
      "epoch": 1.7080311107227124,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000470965767980741,
      "loss": 5.3348,
      "mean_token_accuracy": 0.20177939981222154,
      "num_tokens": 39581931.0,
      "step": 19545
    },
    {
      "entropy": 5.971513795852661,
      "epoch": 1.7084680590754173,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047095045491669696,
      "loss": 5.4637,
      "mean_token_accuracy": 0.18473385125398636,
      "num_tokens": 39592818.0,
      "step": 19550
    },
    {
      "entropy": 5.962217903137207,
      "epoch": 1.708905007428122,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047093513809422043,
      "loss": 5.3503,
      "mean_token_accuracy": 0.19436137080192567,
      "num_tokens": 39602867.0,
      "step": 19555
    },
    {
      "entropy": 5.954106760025025,
      "epoch": 1.7093419557808267,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004709198175136053,
      "loss": 5.3637,
      "mean_token_accuracy": 0.19577980488538743,
      "num_tokens": 39613185.0,
      "step": 19560
    },
    {
      "entropy": 5.883975648880005,
      "epoch": 1.7097789041335314,
      "grad_norm": 1.125,
      "learning_rate": 0.0004709044931751454,
      "loss": 5.3112,
      "mean_token_accuracy": 0.19962934106588365,
      "num_tokens": 39623262.0,
      "step": 19565
    },
    {
      "entropy": 5.985095977783203,
      "epoch": 1.710215852486236,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004708891650791347,
      "loss": 5.3547,
      "mean_token_accuracy": 0.19547147005796434,
      "num_tokens": 39633184.0,
      "step": 19570
    },
    {
      "entropy": 5.986671447753906,
      "epoch": 1.710652800838941,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004708738332258671,
      "loss": 5.3869,
      "mean_token_accuracy": 0.197317673265934,
      "num_tokens": 39643778.0,
      "step": 19575
    },
    {
      "entropy": 5.939882421493531,
      "epoch": 1.7110897491916455,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004708584976156368,
      "loss": 5.2869,
      "mean_token_accuracy": 0.19924521297216416,
      "num_tokens": 39653519.0,
      "step": 19580
    },
    {
      "entropy": 5.943725347518921,
      "epoch": 1.7115266975443504,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047084315824873793,
      "loss": 5.3661,
      "mean_token_accuracy": 0.19805212318897247,
      "num_tokens": 39663824.0,
      "step": 19585
    },
    {
      "entropy": 5.965467691421509,
      "epoch": 1.7119636458970549,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004708278151254646,
      "loss": 5.3741,
      "mean_token_accuracy": 0.1934875637292862,
      "num_tokens": 39673631.0,
      "step": 19590
    },
    {
      "entropy": 6.050877857208252,
      "epoch": 1.7124005942497598,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004708124682461112,
      "loss": 5.4222,
      "mean_token_accuracy": 0.1872384652495384,
      "num_tokens": 39683118.0,
      "step": 19595
    },
    {
      "entropy": 5.9499608993530275,
      "epoch": 1.7128375426024642,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047079711761097197,
      "loss": 5.3443,
      "mean_token_accuracy": 0.1970664381980896,
      "num_tokens": 39693947.0,
      "step": 19600
    },
    {
      "entropy": 5.937475252151489,
      "epoch": 1.7132744909551691,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004707817632203415,
      "loss": 5.3684,
      "mean_token_accuracy": 0.18771580904722213,
      "num_tokens": 39703874.0,
      "step": 19605
    },
    {
      "entropy": 5.911107778549194,
      "epoch": 1.7137114393078738,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047076640507451425,
      "loss": 5.343,
      "mean_token_accuracy": 0.19749326407909393,
      "num_tokens": 39714997.0,
      "step": 19610
    },
    {
      "entropy": 6.021894454956055,
      "epoch": 1.7141483876605785,
      "grad_norm": 1.125,
      "learning_rate": 0.0004707510431737847,
      "loss": 5.3283,
      "mean_token_accuracy": 0.19457703381776809,
      "num_tokens": 39726042.0,
      "step": 19615
    },
    {
      "entropy": 5.967605829238892,
      "epoch": 1.7145853360132832,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004707356775184476,
      "loss": 5.3551,
      "mean_token_accuracy": 0.19531828463077544,
      "num_tokens": 39736733.0,
      "step": 19620
    },
    {
      "entropy": 5.99556679725647,
      "epoch": 1.715022284365988,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004707203081087977,
      "loss": 5.4645,
      "mean_token_accuracy": 0.18375043123960494,
      "num_tokens": 39746811.0,
      "step": 19625
    },
    {
      "entropy": 6.027449655532837,
      "epoch": 1.7154592327186926,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004707049349451297,
      "loss": 5.3866,
      "mean_token_accuracy": 0.1894560694694519,
      "num_tokens": 39757371.0,
      "step": 19630
    },
    {
      "entropy": 5.928915548324585,
      "epoch": 1.7158961810713973,
      "grad_norm": 1.125,
      "learning_rate": 0.0004706895580277384,
      "loss": 5.3461,
      "mean_token_accuracy": 0.19876503199338913,
      "num_tokens": 39767430.0,
      "step": 19635
    },
    {
      "entropy": 5.936293077468872,
      "epoch": 1.7163331294241022,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047067417735691884,
      "loss": 5.333,
      "mean_token_accuracy": 0.19789566993713378,
      "num_tokens": 39777270.0,
      "step": 19640
    },
    {
      "entropy": 5.9516379833221436,
      "epoch": 1.7167700777768067,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004706587929329662,
      "loss": 5.3406,
      "mean_token_accuracy": 0.19808469116687774,
      "num_tokens": 39787647.0,
      "step": 19645
    },
    {
      "entropy": 5.980694723129273,
      "epoch": 1.7172070261295116,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047064340475617527,
      "loss": 5.3984,
      "mean_token_accuracy": 0.19057171493768693,
      "num_tokens": 39797315.0,
      "step": 19650
    },
    {
      "entropy": 5.955020952224731,
      "epoch": 1.717643974482216,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004706280128268413,
      "loss": 5.2879,
      "mean_token_accuracy": 0.19807107746601105,
      "num_tokens": 39806625.0,
      "step": 19655
    },
    {
      "entropy": 5.933014154434204,
      "epoch": 1.718080922834921,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004706126171452596,
      "loss": 5.4101,
      "mean_token_accuracy": 0.19453221261501313,
      "num_tokens": 39818113.0,
      "step": 19660
    },
    {
      "entropy": 5.930294752120972,
      "epoch": 1.7185178711876257,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004705972177117254,
      "loss": 5.4002,
      "mean_token_accuracy": 0.19078071862459184,
      "num_tokens": 39827897.0,
      "step": 19665
    },
    {
      "entropy": 5.956164407730102,
      "epoch": 1.7189548195403304,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047058181452653403,
      "loss": 5.2364,
      "mean_token_accuracy": 0.20547852665185928,
      "num_tokens": 39837209.0,
      "step": 19670
    },
    {
      "entropy": 5.96647367477417,
      "epoch": 1.719391767893035,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000470566407589981,
      "loss": 5.3629,
      "mean_token_accuracy": 0.19827130883932115,
      "num_tokens": 39848023.0,
      "step": 19675
    },
    {
      "entropy": 5.972746753692627,
      "epoch": 1.7198287162457397,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004705509969023618,
      "loss": 5.4034,
      "mean_token_accuracy": 0.19092754870653153,
      "num_tokens": 39857992.0,
      "step": 19680
    },
    {
      "entropy": 6.003995132446289,
      "epoch": 1.7202656645984444,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000470535582463972,
      "loss": 5.4126,
      "mean_token_accuracy": 0.19368927627801896,
      "num_tokens": 39867927.0,
      "step": 19685
    },
    {
      "entropy": 5.952826976776123,
      "epoch": 1.7207026129511491,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004705201642751073,
      "loss": 5.3303,
      "mean_token_accuracy": 0.20198136270046235,
      "num_tokens": 39877327.0,
      "step": 19690
    },
    {
      "entropy": 5.902590131759643,
      "epoch": 1.721139561303854,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047050474233606334,
      "loss": 5.3521,
      "mean_token_accuracy": 0.1912250429391861,
      "num_tokens": 39887369.0,
      "step": 19695
    },
    {
      "entropy": 5.973301935195923,
      "epoch": 1.7215765096565585,
      "grad_norm": 1.03125,
      "learning_rate": 0.000470489316647136,
      "loss": 5.358,
      "mean_token_accuracy": 0.19962775856256484,
      "num_tokens": 39897530.0,
      "step": 19700
    },
    {
      "entropy": 5.950176668167114,
      "epoch": 1.7220134580092634,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004704738872086212,
      "loss": 5.3599,
      "mean_token_accuracy": 0.19185259491205214,
      "num_tokens": 39907234.0,
      "step": 19705
    },
    {
      "entropy": 5.990590000152588,
      "epoch": 1.722450406361968,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004704584540208148,
      "loss": 5.3954,
      "mean_token_accuracy": 0.18905961364507676,
      "num_tokens": 39917940.0,
      "step": 19710
    },
    {
      "entropy": 5.9447612285614015,
      "epoch": 1.7228873547146728,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004704430170840127,
      "loss": 5.3252,
      "mean_token_accuracy": 0.19938718974590303,
      "num_tokens": 39927569.0,
      "step": 19715
    },
    {
      "entropy": 5.864244794845581,
      "epoch": 1.7233243030673773,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004704275763985113,
      "loss": 5.3946,
      "mean_token_accuracy": 0.1947351276874542,
      "num_tokens": 39938302.0,
      "step": 19720
    },
    {
      "entropy": 5.976911354064941,
      "epoch": 1.7237612514200822,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004704121319646065,
      "loss": 5.3205,
      "mean_token_accuracy": 0.2018548369407654,
      "num_tokens": 39948133.0,
      "step": 19725
    },
    {
      "entropy": 5.945567989349366,
      "epoch": 1.7241981997727869,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047039668378259455,
      "loss": 5.3057,
      "mean_token_accuracy": 0.20177710950374603,
      "num_tokens": 39958924.0,
      "step": 19730
    },
    {
      "entropy": 5.9335045337677,
      "epoch": 1.7246351481254916,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047038123185277187,
      "loss": 5.2787,
      "mean_token_accuracy": 0.1971591204404831,
      "num_tokens": 39968578.0,
      "step": 19735
    },
    {
      "entropy": 5.889938926696777,
      "epoch": 1.7250720964781963,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047036577617543484,
      "loss": 5.263,
      "mean_token_accuracy": 0.21011730879545212,
      "num_tokens": 39978112.0,
      "step": 19740
    },
    {
      "entropy": 5.948252105712891,
      "epoch": 1.725509044830901,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047035031675087983,
      "loss": 5.3855,
      "mean_token_accuracy": 0.18325353264808655,
      "num_tokens": 39988071.0,
      "step": 19745
    },
    {
      "entropy": 6.006153345108032,
      "epoch": 1.7259459931836056,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004703348535794034,
      "loss": 5.392,
      "mean_token_accuracy": 0.18682860881090163,
      "num_tokens": 39998112.0,
      "step": 19750
    },
    {
      "entropy": 5.950923776626587,
      "epoch": 1.7263829415363103,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047031938666130206,
      "loss": 5.3433,
      "mean_token_accuracy": 0.19624405056238176,
      "num_tokens": 40008869.0,
      "step": 19755
    },
    {
      "entropy": 5.955561304092408,
      "epoch": 1.7268198898890152,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004703039159968726,
      "loss": 5.3608,
      "mean_token_accuracy": 0.19621913880109787,
      "num_tokens": 40018711.0,
      "step": 19760
    },
    {
      "entropy": 5.955186986923218,
      "epoch": 1.7272568382417197,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047028844158641166,
      "loss": 5.3685,
      "mean_token_accuracy": 0.1933722600340843,
      "num_tokens": 40029209.0,
      "step": 19765
    },
    {
      "entropy": 5.961617469787598,
      "epoch": 1.7276937865944246,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000470272963430216,
      "loss": 5.3332,
      "mean_token_accuracy": 0.1976212739944458,
      "num_tokens": 40039071.0,
      "step": 19770
    },
    {
      "entropy": 5.920843648910522,
      "epoch": 1.728130734947129,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004702574815285827,
      "loss": 5.3103,
      "mean_token_accuracy": 0.1958835169672966,
      "num_tokens": 40049765.0,
      "step": 19775
    },
    {
      "entropy": 5.906405067443847,
      "epoch": 1.728567683299834,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004702419958818085,
      "loss": 5.3063,
      "mean_token_accuracy": 0.1999536707997322,
      "num_tokens": 40060020.0,
      "step": 19780
    },
    {
      "entropy": 5.967951679229737,
      "epoch": 1.7290046316525387,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047022650649019063,
      "loss": 5.3313,
      "mean_token_accuracy": 0.19834507405757903,
      "num_tokens": 40070081.0,
      "step": 19785
    },
    {
      "entropy": 5.9688084602355955,
      "epoch": 1.7294415800052434,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000470211013354026,
      "loss": 5.3932,
      "mean_token_accuracy": 0.19364374727010727,
      "num_tokens": 40080532.0,
      "step": 19790
    },
    {
      "entropy": 5.998116159439087,
      "epoch": 1.729878528357948,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004701955164736118,
      "loss": 5.3604,
      "mean_token_accuracy": 0.19922006726264954,
      "num_tokens": 40091288.0,
      "step": 19795
    },
    {
      "entropy": 5.981817388534546,
      "epoch": 1.7303154767106528,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004701800158492454,
      "loss": 5.3262,
      "mean_token_accuracy": 0.19821074455976487,
      "num_tokens": 40100461.0,
      "step": 19800
    },
    {
      "entropy": 5.918224716186524,
      "epoch": 1.7307524250633575,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000470164511481224,
      "loss": 5.3105,
      "mean_token_accuracy": 0.1959414303302765,
      "num_tokens": 40109627.0,
      "step": 19805
    },
    {
      "entropy": 5.8767353057861325,
      "epoch": 1.7311893734160622,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000470149003369845,
      "loss": 5.3559,
      "mean_token_accuracy": 0.19707850217819214,
      "num_tokens": 40119183.0,
      "step": 19810
    },
    {
      "entropy": 5.9458245754241945,
      "epoch": 1.731626321768767,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004701334915154058,
      "loss": 5.3592,
      "mean_token_accuracy": 0.1941467195749283,
      "num_tokens": 40129569.0,
      "step": 19815
    },
    {
      "entropy": 5.993707847595215,
      "epoch": 1.7320632701214715,
      "grad_norm": 1.125,
      "learning_rate": 0.00047011797591820404,
      "loss": 5.3116,
      "mean_token_accuracy": 0.20003570020198821,
      "num_tokens": 40139618.0,
      "step": 19820
    },
    {
      "entropy": 5.891762828826904,
      "epoch": 1.7325002184741765,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047010245657853726,
      "loss": 5.2518,
      "mean_token_accuracy": 0.20437314063310624,
      "num_tokens": 40149349.0,
      "step": 19825
    },
    {
      "entropy": 5.967345714569092,
      "epoch": 1.732937166826881,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004700869334967031,
      "loss": 5.3734,
      "mean_token_accuracy": 0.19689346998929977,
      "num_tokens": 40159161.0,
      "step": 19830
    },
    {
      "entropy": 5.945078325271607,
      "epoch": 1.7333741151795858,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047007140667299944,
      "loss": 5.3672,
      "mean_token_accuracy": 0.19487204104661943,
      "num_tokens": 40169006.0,
      "step": 19835
    },
    {
      "entropy": 5.944540071487427,
      "epoch": 1.7338110635322905,
      "grad_norm": 1.0,
      "learning_rate": 0.00047005587610772384,
      "loss": 5.31,
      "mean_token_accuracy": 0.19780368208885193,
      "num_tokens": 40179964.0,
      "step": 19840
    },
    {
      "entropy": 5.961609983444214,
      "epoch": 1.7342480118849952,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047004034180117446,
      "loss": 5.3835,
      "mean_token_accuracy": 0.18961890637874604,
      "num_tokens": 40190598.0,
      "step": 19845
    },
    {
      "entropy": 5.965794706344605,
      "epoch": 1.7346849602377,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000470024803753649,
      "loss": 5.3977,
      "mean_token_accuracy": 0.19227588027715684,
      "num_tokens": 40199537.0,
      "step": 19850
    },
    {
      "entropy": 5.930172395706177,
      "epoch": 1.7351219085904046,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004700092619654456,
      "loss": 5.3282,
      "mean_token_accuracy": 0.20001808404922486,
      "num_tokens": 40209167.0,
      "step": 19855
    },
    {
      "entropy": 5.943537902832031,
      "epoch": 1.7355588569431093,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004699937164368625,
      "loss": 5.3355,
      "mean_token_accuracy": 0.19478426575660707,
      "num_tokens": 40218738.0,
      "step": 19860
    },
    {
      "entropy": 5.943896532058716,
      "epoch": 1.735995805295814,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004699781671681977,
      "loss": 5.3219,
      "mean_token_accuracy": 0.1962618798017502,
      "num_tokens": 40227897.0,
      "step": 19865
    },
    {
      "entropy": 5.961973524093628,
      "epoch": 1.736432753648519,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046996261415974946,
      "loss": 5.3672,
      "mean_token_accuracy": 0.19220719188451768,
      "num_tokens": 40238870.0,
      "step": 19870
    },
    {
      "entropy": 5.93774003982544,
      "epoch": 1.7368697020012234,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046994705741181615,
      "loss": 5.2931,
      "mean_token_accuracy": 0.19319459199905395,
      "num_tokens": 40247832.0,
      "step": 19875
    },
    {
      "entropy": 5.945313787460327,
      "epoch": 1.7373066503539283,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004699314969246961,
      "loss": 5.3271,
      "mean_token_accuracy": 0.19718337059020996,
      "num_tokens": 40258696.0,
      "step": 19880
    },
    {
      "entropy": 5.946334266662598,
      "epoch": 1.7377435987066328,
      "grad_norm": 1.125,
      "learning_rate": 0.00046991593269868776,
      "loss": 5.3058,
      "mean_token_accuracy": 0.20230603963136673,
      "num_tokens": 40267767.0,
      "step": 19885
    },
    {
      "entropy": 5.965389060974121,
      "epoch": 1.7381805470593377,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004699003647340898,
      "loss": 5.3875,
      "mean_token_accuracy": 0.1939070001244545,
      "num_tokens": 40277813.0,
      "step": 19890
    },
    {
      "entropy": 5.944714975357056,
      "epoch": 1.7386174954120421,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046988479303120066,
      "loss": 5.3266,
      "mean_token_accuracy": 0.20628851056098937,
      "num_tokens": 40289031.0,
      "step": 19895
    },
    {
      "entropy": 5.906374263763428,
      "epoch": 1.739054443764747,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046986921759031906,
      "loss": 5.3344,
      "mean_token_accuracy": 0.19758255630731583,
      "num_tokens": 40299428.0,
      "step": 19900
    },
    {
      "entropy": 5.925669240951538,
      "epoch": 1.7394913921174517,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004698536384117438,
      "loss": 5.2966,
      "mean_token_accuracy": 0.2008657619357109,
      "num_tokens": 40308996.0,
      "step": 19905
    },
    {
      "entropy": 5.973176527023315,
      "epoch": 1.7399283404701564,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046983805549577365,
      "loss": 5.391,
      "mean_token_accuracy": 0.18890541642904282,
      "num_tokens": 40319753.0,
      "step": 19910
    },
    {
      "entropy": 6.066070222854615,
      "epoch": 1.7403652888228611,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004698224688427074,
      "loss": 5.4981,
      "mean_token_accuracy": 0.1903664782643318,
      "num_tokens": 40330729.0,
      "step": 19915
    },
    {
      "entropy": 6.008314895629883,
      "epoch": 1.7408022371755658,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004698068784528442,
      "loss": 5.4116,
      "mean_token_accuracy": 0.1914690375328064,
      "num_tokens": 40340246.0,
      "step": 19920
    },
    {
      "entropy": 5.988184213638306,
      "epoch": 1.7412391855282705,
      "grad_norm": 1.078125,
      "learning_rate": 0.000469791284326483,
      "loss": 5.2916,
      "mean_token_accuracy": 0.19803860336542128,
      "num_tokens": 40350099.0,
      "step": 19925
    },
    {
      "entropy": 5.938072633743286,
      "epoch": 1.7416761338809752,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046977568646392284,
      "loss": 5.3362,
      "mean_token_accuracy": 0.2016901731491089,
      "num_tokens": 40360825.0,
      "step": 19930
    },
    {
      "entropy": 5.963312768936158,
      "epoch": 1.74211308223368,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000469760084865463,
      "loss": 5.4247,
      "mean_token_accuracy": 0.18622480183839799,
      "num_tokens": 40371025.0,
      "step": 19935
    },
    {
      "entropy": 5.984749984741211,
      "epoch": 1.7425500305863846,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004697444795314027,
      "loss": 5.3173,
      "mean_token_accuracy": 0.20238217413425447,
      "num_tokens": 40380338.0,
      "step": 19940
    },
    {
      "entropy": 5.962861680984497,
      "epoch": 1.7429869789390895,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00046972887046204114,
      "loss": 5.3598,
      "mean_token_accuracy": 0.20466173589229583,
      "num_tokens": 40391814.0,
      "step": 19945
    },
    {
      "entropy": 5.888778781890869,
      "epoch": 1.743423927291794,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046971325765767793,
      "loss": 5.3614,
      "mean_token_accuracy": 0.19153420329093934,
      "num_tokens": 40402013.0,
      "step": 19950
    },
    {
      "entropy": 5.939953708648682,
      "epoch": 1.7438608756444989,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046969764111861234,
      "loss": 5.3258,
      "mean_token_accuracy": 0.19152018576860427,
      "num_tokens": 40412205.0,
      "step": 19955
    },
    {
      "entropy": 5.869623327255249,
      "epoch": 1.7442978239972036,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046968202084514405,
      "loss": 5.2886,
      "mean_token_accuracy": 0.2071106642484665,
      "num_tokens": 40423212.0,
      "step": 19960
    },
    {
      "entropy": 5.944646692276001,
      "epoch": 1.7447347723499083,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004696663968375725,
      "loss": 5.2999,
      "mean_token_accuracy": 0.1982944890856743,
      "num_tokens": 40433272.0,
      "step": 19965
    },
    {
      "entropy": 5.980220699310303,
      "epoch": 1.745171720702613,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004696507690961975,
      "loss": 5.4079,
      "mean_token_accuracy": 0.19190803915262222,
      "num_tokens": 40443931.0,
      "step": 19970
    },
    {
      "entropy": 5.941680479049682,
      "epoch": 1.7456086690553176,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004696351376213188,
      "loss": 5.345,
      "mean_token_accuracy": 0.1941220834851265,
      "num_tokens": 40453008.0,
      "step": 19975
    },
    {
      "entropy": 5.931813097000122,
      "epoch": 1.7460456174080223,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000469619502413236,
      "loss": 5.4104,
      "mean_token_accuracy": 0.1965568706393242,
      "num_tokens": 40463081.0,
      "step": 19980
    },
    {
      "entropy": 5.954363441467285,
      "epoch": 1.746482565760727,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004696038634722493,
      "loss": 5.2644,
      "mean_token_accuracy": 0.197504460811615,
      "num_tokens": 40472390.0,
      "step": 19985
    },
    {
      "entropy": 5.975373935699463,
      "epoch": 1.746919514113432,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046958822079865855,
      "loss": 5.386,
      "mean_token_accuracy": 0.19939188212156295,
      "num_tokens": 40483068.0,
      "step": 19990
    },
    {
      "entropy": 6.006787872314453,
      "epoch": 1.7473564624661364,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004695725743927637,
      "loss": 5.3632,
      "mean_token_accuracy": 0.19816800504922866,
      "num_tokens": 40493125.0,
      "step": 19995
    },
    {
      "entropy": 5.973322439193725,
      "epoch": 1.7477934108188413,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046955692425486496,
      "loss": 5.3824,
      "mean_token_accuracy": 0.18999895602464675,
      "num_tokens": 40502618.0,
      "step": 20000
    },
    {
      "entropy": 5.959735631942749,
      "epoch": 1.7482303591715458,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004695412703852624,
      "loss": 5.3861,
      "mean_token_accuracy": 0.19456802755594255,
      "num_tokens": 40512878.0,
      "step": 20005
    },
    {
      "entropy": 5.916582202911377,
      "epoch": 1.7486673075242507,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004695256127842563,
      "loss": 5.3462,
      "mean_token_accuracy": 0.1970168188214302,
      "num_tokens": 40521983.0,
      "step": 20010
    },
    {
      "entropy": 5.944618654251099,
      "epoch": 1.7491042558769554,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046950995145214716,
      "loss": 5.345,
      "mean_token_accuracy": 0.19948479533195496,
      "num_tokens": 40533126.0,
      "step": 20015
    },
    {
      "entropy": 5.90579833984375,
      "epoch": 1.74954120422966,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004694942863892352,
      "loss": 5.3325,
      "mean_token_accuracy": 0.19609519690275193,
      "num_tokens": 40543321.0,
      "step": 20020
    },
    {
      "entropy": 5.939438152313232,
      "epoch": 1.7499781525823648,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004694786175958209,
      "loss": 5.364,
      "mean_token_accuracy": 0.19303788095712662,
      "num_tokens": 40554208.0,
      "step": 20025
    },
    {
      "entropy": 5.992564868927002,
      "epoch": 1.7504151009350695,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046946294507220477,
      "loss": 5.3121,
      "mean_token_accuracy": 0.19031067192554474,
      "num_tokens": 40563365.0,
      "step": 20030
    },
    {
      "entropy": 5.8383355140686035,
      "epoch": 1.7508520492877742,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004694472688186874,
      "loss": 5.2697,
      "mean_token_accuracy": 0.20086068511009217,
      "num_tokens": 40573765.0,
      "step": 20035
    },
    {
      "entropy": 5.926556205749511,
      "epoch": 1.7512889976404789,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004694315888355697,
      "loss": 5.3541,
      "mean_token_accuracy": 0.19645707458257675,
      "num_tokens": 40583786.0,
      "step": 20040
    },
    {
      "entropy": 6.012339448928833,
      "epoch": 1.7517259459931838,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046941590512315215,
      "loss": 5.3691,
      "mean_token_accuracy": 0.1901681676506996,
      "num_tokens": 40593686.0,
      "step": 20045
    },
    {
      "entropy": 5.949923706054688,
      "epoch": 1.7521628943458882,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004694002176817356,
      "loss": 5.3719,
      "mean_token_accuracy": 0.19580782651901246,
      "num_tokens": 40604326.0,
      "step": 20050
    },
    {
      "entropy": 6.009000682830811,
      "epoch": 1.7525998426985931,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004693845265116211,
      "loss": 5.3425,
      "mean_token_accuracy": 0.19292501658201217,
      "num_tokens": 40613683.0,
      "step": 20055
    },
    {
      "entropy": 5.978735446929932,
      "epoch": 1.7530367910512976,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046936883161310944,
      "loss": 5.3762,
      "mean_token_accuracy": 0.19804587662220002,
      "num_tokens": 40623675.0,
      "step": 20060
    },
    {
      "entropy": 5.966429281234741,
      "epoch": 1.7534737394040025,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046935313298650185,
      "loss": 5.3839,
      "mean_token_accuracy": 0.1928016200661659,
      "num_tokens": 40634006.0,
      "step": 20065
    },
    {
      "entropy": 5.91787805557251,
      "epoch": 1.753910687756707,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046933743063209927,
      "loss": 5.2806,
      "mean_token_accuracy": 0.20580178797245025,
      "num_tokens": 40643175.0,
      "step": 20070
    },
    {
      "entropy": 6.009709978103638,
      "epoch": 1.754347636109412,
      "grad_norm": 1.125,
      "learning_rate": 0.00046932172455020296,
      "loss": 5.3819,
      "mean_token_accuracy": 0.1888989269733429,
      "num_tokens": 40653447.0,
      "step": 20075
    },
    {
      "entropy": 5.923614740371704,
      "epoch": 1.7547845844621166,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004693060147411141,
      "loss": 5.2901,
      "mean_token_accuracy": 0.20090775340795516,
      "num_tokens": 40663499.0,
      "step": 20080
    },
    {
      "entropy": 5.950368642807007,
      "epoch": 1.7552215328148213,
      "grad_norm": 1.125,
      "learning_rate": 0.0004692903012051341,
      "loss": 5.3428,
      "mean_token_accuracy": 0.19515886902809143,
      "num_tokens": 40672943.0,
      "step": 20085
    },
    {
      "entropy": 5.944488573074341,
      "epoch": 1.755658481167526,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004692745839425643,
      "loss": 5.3438,
      "mean_token_accuracy": 0.20002549588680268,
      "num_tokens": 40683659.0,
      "step": 20090
    },
    {
      "entropy": 6.021074867248535,
      "epoch": 1.7560954295202307,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004692588629537062,
      "loss": 5.4141,
      "mean_token_accuracy": 0.19001730978488923,
      "num_tokens": 40693829.0,
      "step": 20095
    },
    {
      "entropy": 5.992730665206909,
      "epoch": 1.7565323778729354,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004692431382388613,
      "loss": 5.3799,
      "mean_token_accuracy": 0.19075458943843843,
      "num_tokens": 40704717.0,
      "step": 20100
    },
    {
      "entropy": 5.965110492706299,
      "epoch": 1.75696932622564,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046922740979833135,
      "loss": 5.3408,
      "mean_token_accuracy": 0.1943011164665222,
      "num_tokens": 40714970.0,
      "step": 20105
    },
    {
      "entropy": 5.976020860671997,
      "epoch": 1.757406274578345,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046921167763241777,
      "loss": 5.3254,
      "mean_token_accuracy": 0.19826027601957322,
      "num_tokens": 40723664.0,
      "step": 20110
    },
    {
      "entropy": 5.954362487792968,
      "epoch": 1.7578432229310494,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046919594174142247,
      "loss": 5.3331,
      "mean_token_accuracy": 0.19490178525447846,
      "num_tokens": 40733735.0,
      "step": 20115
    },
    {
      "entropy": 5.95610933303833,
      "epoch": 1.7582801712837544,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046918020212564725,
      "loss": 5.3798,
      "mean_token_accuracy": 0.19486182928085327,
      "num_tokens": 40743699.0,
      "step": 20120
    },
    {
      "entropy": 5.961549615859985,
      "epoch": 1.7587171196364588,
      "grad_norm": 1.046875,
      "learning_rate": 0.000469164458785394,
      "loss": 5.4028,
      "mean_token_accuracy": 0.19175568968057632,
      "num_tokens": 40754553.0,
      "step": 20125
    },
    {
      "entropy": 6.012901544570923,
      "epoch": 1.7591540679891637,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046914871172096473,
      "loss": 5.3843,
      "mean_token_accuracy": 0.18725629448890685,
      "num_tokens": 40764660.0,
      "step": 20130
    },
    {
      "entropy": 5.940416860580444,
      "epoch": 1.7595910163418684,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004691329609326614,
      "loss": 5.2372,
      "mean_token_accuracy": 0.20677722245454788,
      "num_tokens": 40774237.0,
      "step": 20135
    },
    {
      "entropy": 5.9530463218688965,
      "epoch": 1.7600279646945731,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004691172064207862,
      "loss": 5.3329,
      "mean_token_accuracy": 0.1988286018371582,
      "num_tokens": 40783608.0,
      "step": 20140
    },
    {
      "entropy": 6.003953647613526,
      "epoch": 1.7604649130472778,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004691014481856412,
      "loss": 5.3688,
      "mean_token_accuracy": 0.18875357508659363,
      "num_tokens": 40793987.0,
      "step": 20145
    },
    {
      "entropy": 6.00982403755188,
      "epoch": 1.7609018613999825,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046908568622752874,
      "loss": 5.397,
      "mean_token_accuracy": 0.19192990511655808,
      "num_tokens": 40805600.0,
      "step": 20150
    },
    {
      "entropy": 5.965030670166016,
      "epoch": 1.7613388097526872,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046906992054675115,
      "loss": 5.3659,
      "mean_token_accuracy": 0.19266253113746643,
      "num_tokens": 40815252.0,
      "step": 20155
    },
    {
      "entropy": 5.929058980941773,
      "epoch": 1.7617757581053919,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004690541511436107,
      "loss": 5.3444,
      "mean_token_accuracy": 0.20054872035980226,
      "num_tokens": 40825945.0,
      "step": 20160
    },
    {
      "entropy": 5.985772657394409,
      "epoch": 1.7622127064580968,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004690383780184101,
      "loss": 5.3965,
      "mean_token_accuracy": 0.1874737873673439,
      "num_tokens": 40835811.0,
      "step": 20165
    },
    {
      "entropy": 6.011438179016113,
      "epoch": 1.7626496548108013,
      "grad_norm": 1.25,
      "learning_rate": 0.00046902260117145163,
      "loss": 5.389,
      "mean_token_accuracy": 0.19082629978656768,
      "num_tokens": 40845088.0,
      "step": 20170
    },
    {
      "entropy": 5.957052373886109,
      "epoch": 1.7630866031635062,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000469006820603038,
      "loss": 5.4264,
      "mean_token_accuracy": 0.19309324473142625,
      "num_tokens": 40854703.0,
      "step": 20175
    },
    {
      "entropy": 5.967845630645752,
      "epoch": 1.7635235515162107,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046899103631347196,
      "loss": 5.3464,
      "mean_token_accuracy": 0.19681245386600493,
      "num_tokens": 40865056.0,
      "step": 20180
    },
    {
      "entropy": 5.924335050582886,
      "epoch": 1.7639604998689156,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004689752483030561,
      "loss": 5.2621,
      "mean_token_accuracy": 0.20219385772943496,
      "num_tokens": 40874355.0,
      "step": 20185
    },
    {
      "entropy": 5.979226970672608,
      "epoch": 1.7643974482216203,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004689594565720934,
      "loss": 5.3358,
      "mean_token_accuracy": 0.1995019093155861,
      "num_tokens": 40884329.0,
      "step": 20190
    },
    {
      "entropy": 5.960859727859497,
      "epoch": 1.764834396574325,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004689436611208867,
      "loss": 5.3482,
      "mean_token_accuracy": 0.19398341029882432,
      "num_tokens": 40894702.0,
      "step": 20195
    },
    {
      "entropy": 5.95725679397583,
      "epoch": 1.7652713449270296,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046892786194973897,
      "loss": 5.3075,
      "mean_token_accuracy": 0.19328475147485732,
      "num_tokens": 40905467.0,
      "step": 20200
    },
    {
      "entropy": 5.974929666519165,
      "epoch": 1.7657082932797343,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004689120590589532,
      "loss": 5.382,
      "mean_token_accuracy": 0.19120728671550752,
      "num_tokens": 40915915.0,
      "step": 20205
    },
    {
      "entropy": 5.931466341018677,
      "epoch": 1.766145241632439,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004688962524488325,
      "loss": 5.3043,
      "mean_token_accuracy": 0.19892434030771255,
      "num_tokens": 40925941.0,
      "step": 20210
    },
    {
      "entropy": 5.915340280532837,
      "epoch": 1.7665821899851437,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004688804421196801,
      "loss": 5.3673,
      "mean_token_accuracy": 0.2019046425819397,
      "num_tokens": 40935831.0,
      "step": 20215
    },
    {
      "entropy": 5.9245161533355715,
      "epoch": 1.7670191383378486,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046886462807179926,
      "loss": 5.3494,
      "mean_token_accuracy": 0.20024069547653198,
      "num_tokens": 40945341.0,
      "step": 20220
    },
    {
      "entropy": 5.960494136810302,
      "epoch": 1.767456086690553,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004688488103054931,
      "loss": 5.3697,
      "mean_token_accuracy": 0.1962823525071144,
      "num_tokens": 40955759.0,
      "step": 20225
    },
    {
      "entropy": 5.964919233322144,
      "epoch": 1.767893035043258,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004688329888210654,
      "loss": 5.3179,
      "mean_token_accuracy": 0.19228544235229492,
      "num_tokens": 40966184.0,
      "step": 20230
    },
    {
      "entropy": 5.975030040740966,
      "epoch": 1.7683299833959625,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004688171636188194,
      "loss": 5.3599,
      "mean_token_accuracy": 0.1957270935177803,
      "num_tokens": 40975797.0,
      "step": 20235
    },
    {
      "entropy": 6.000279188156128,
      "epoch": 1.7687669317486674,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004688013346990585,
      "loss": 5.3502,
      "mean_token_accuracy": 0.1955571785569191,
      "num_tokens": 40985830.0,
      "step": 20240
    },
    {
      "entropy": 5.893586444854736,
      "epoch": 1.7692038801013719,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046878550206208656,
      "loss": 5.2673,
      "mean_token_accuracy": 0.2043014794588089,
      "num_tokens": 40995692.0,
      "step": 20245
    },
    {
      "entropy": 5.957341909408569,
      "epoch": 1.7696408284540768,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004687696657082071,
      "loss": 5.396,
      "mean_token_accuracy": 0.19105567634105683,
      "num_tokens": 41005506.0,
      "step": 20250
    },
    {
      "entropy": 5.944362688064575,
      "epoch": 1.7700777768067815,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046875382563772396,
      "loss": 5.2823,
      "mean_token_accuracy": 0.1955369681119919,
      "num_tokens": 41014944.0,
      "step": 20255
    },
    {
      "entropy": 5.996970319747925,
      "epoch": 1.7705147251594862,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046873798185094093,
      "loss": 5.4303,
      "mean_token_accuracy": 0.18573116213083268,
      "num_tokens": 41025424.0,
      "step": 20260
    },
    {
      "entropy": 5.902832889556885,
      "epoch": 1.7709516735121908,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004687221343481618,
      "loss": 5.2691,
      "mean_token_accuracy": 0.20547903329133987,
      "num_tokens": 41035817.0,
      "step": 20265
    },
    {
      "entropy": 5.895861482620239,
      "epoch": 1.7713886218648955,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046870628312969064,
      "loss": 5.321,
      "mean_token_accuracy": 0.19946757704019547,
      "num_tokens": 41045730.0,
      "step": 20270
    },
    {
      "entropy": 6.013915348052978,
      "epoch": 1.7718255702176002,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004686904281958315,
      "loss": 5.4118,
      "mean_token_accuracy": 0.18563709408044815,
      "num_tokens": 41057921.0,
      "step": 20275
    },
    {
      "entropy": 6.010966014862061,
      "epoch": 1.772262518570305,
      "grad_norm": 1.0,
      "learning_rate": 0.00046867456954688847,
      "loss": 5.4093,
      "mean_token_accuracy": 0.19821885377168655,
      "num_tokens": 41068879.0,
      "step": 20280
    },
    {
      "entropy": 5.983388423919678,
      "epoch": 1.7726994669230098,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004686587071831657,
      "loss": 5.4137,
      "mean_token_accuracy": 0.19225437045097352,
      "num_tokens": 41079443.0,
      "step": 20285
    },
    {
      "entropy": 5.93249397277832,
      "epoch": 1.7731364152757143,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004686428411049675,
      "loss": 5.3247,
      "mean_token_accuracy": 0.19427616149187088,
      "num_tokens": 41089932.0,
      "step": 20290
    },
    {
      "entropy": 6.028458595275879,
      "epoch": 1.7735733636284192,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004686269713125981,
      "loss": 5.3865,
      "mean_token_accuracy": 0.19143144637346268,
      "num_tokens": 41100412.0,
      "step": 20295
    },
    {
      "entropy": 5.999599838256836,
      "epoch": 1.7740103119811237,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004686110978063619,
      "loss": 5.3496,
      "mean_token_accuracy": 0.19521638453006745,
      "num_tokens": 41110830.0,
      "step": 20300
    },
    {
      "entropy": 5.987091112136841,
      "epoch": 1.7744472603338286,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004685952205865634,
      "loss": 5.3517,
      "mean_token_accuracy": 0.19155488014221192,
      "num_tokens": 41120662.0,
      "step": 20305
    },
    {
      "entropy": 5.896356296539307,
      "epoch": 1.7748842086865333,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046857933965350714,
      "loss": 5.3029,
      "mean_token_accuracy": 0.20510941594839097,
      "num_tokens": 41130523.0,
      "step": 20310
    },
    {
      "entropy": 5.986127233505249,
      "epoch": 1.775321157039238,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004685634550074977,
      "loss": 5.4128,
      "mean_token_accuracy": 0.19181347340345384,
      "num_tokens": 41141659.0,
      "step": 20315
    },
    {
      "entropy": 6.023252725601196,
      "epoch": 1.7757581053919427,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004685475666488398,
      "loss": 5.4556,
      "mean_token_accuracy": 0.18151462972164153,
      "num_tokens": 41151038.0,
      "step": 20320
    },
    {
      "entropy": 5.962462997436523,
      "epoch": 1.7761950537446474,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046853167457783815,
      "loss": 5.3979,
      "mean_token_accuracy": 0.19599448889493942,
      "num_tokens": 41161888.0,
      "step": 20325
    },
    {
      "entropy": 6.07925066947937,
      "epoch": 1.776632002097352,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004685157787947975,
      "loss": 5.4476,
      "mean_token_accuracy": 0.19328658878803254,
      "num_tokens": 41172722.0,
      "step": 20330
    },
    {
      "entropy": 5.954113960266113,
      "epoch": 1.7770689504500568,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004684998793000229,
      "loss": 5.3058,
      "mean_token_accuracy": 0.19368577748537064,
      "num_tokens": 41181716.0,
      "step": 20335
    },
    {
      "entropy": 5.9371109962463375,
      "epoch": 1.7775058988027617,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004684839760938193,
      "loss": 5.3968,
      "mean_token_accuracy": 0.1860378697514534,
      "num_tokens": 41192903.0,
      "step": 20340
    },
    {
      "entropy": 5.963269853591919,
      "epoch": 1.7779428471554661,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046846806917649145,
      "loss": 5.3445,
      "mean_token_accuracy": 0.19525492042303086,
      "num_tokens": 41203283.0,
      "step": 20345
    },
    {
      "entropy": 6.000311851501465,
      "epoch": 1.778379795508171,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004684521585483447,
      "loss": 5.3466,
      "mean_token_accuracy": 0.19790537804365158,
      "num_tokens": 41213615.0,
      "step": 20350
    },
    {
      "entropy": 5.953909158706665,
      "epoch": 1.7788167438608755,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046843624420968435,
      "loss": 5.2944,
      "mean_token_accuracy": 0.20171753615140914,
      "num_tokens": 41222705.0,
      "step": 20355
    },
    {
      "entropy": 5.918466234207154,
      "epoch": 1.7792536922135804,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004684203261608153,
      "loss": 5.3107,
      "mean_token_accuracy": 0.2011592373251915,
      "num_tokens": 41233201.0,
      "step": 20360
    },
    {
      "entropy": 6.01002688407898,
      "epoch": 1.7796906405662851,
      "grad_norm": 1.125,
      "learning_rate": 0.0004684044044020431,
      "loss": 5.3809,
      "mean_token_accuracy": 0.18944962769746782,
      "num_tokens": 41243088.0,
      "step": 20365
    },
    {
      "entropy": 6.024571752548217,
      "epoch": 1.7801275889189898,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004683884789336731,
      "loss": 5.3974,
      "mean_token_accuracy": 0.1873878002166748,
      "num_tokens": 41253009.0,
      "step": 20370
    },
    {
      "entropy": 5.945598220825195,
      "epoch": 1.7805645372716945,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046837254975601073,
      "loss": 5.3939,
      "mean_token_accuracy": 0.19766019135713578,
      "num_tokens": 41262209.0,
      "step": 20375
    },
    {
      "entropy": 5.970803070068359,
      "epoch": 1.7810014856243992,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046835661686936154,
      "loss": 5.3561,
      "mean_token_accuracy": 0.19851401299238206,
      "num_tokens": 41272762.0,
      "step": 20380
    },
    {
      "entropy": 5.9873443126678465,
      "epoch": 1.7814384339771039,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046834068027403115,
      "loss": 5.4201,
      "mean_token_accuracy": 0.19131887704133987,
      "num_tokens": 41284306.0,
      "step": 20385
    },
    {
      "entropy": 5.994578123092651,
      "epoch": 1.7818753823298086,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046832473997032514,
      "loss": 5.3801,
      "mean_token_accuracy": 0.1905493840575218,
      "num_tokens": 41295211.0,
      "step": 20390
    },
    {
      "entropy": 5.9555176258087155,
      "epoch": 1.7823123306825135,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004683087959585493,
      "loss": 5.2316,
      "mean_token_accuracy": 0.2038902834057808,
      "num_tokens": 41303773.0,
      "step": 20395
    },
    {
      "entropy": 5.894587659835816,
      "epoch": 1.782749279035218,
      "grad_norm": 1.125,
      "learning_rate": 0.0004682928482390095,
      "loss": 5.3425,
      "mean_token_accuracy": 0.1977698400616646,
      "num_tokens": 41313536.0,
      "step": 20400
    },
    {
      "entropy": 5.992296743392944,
      "epoch": 1.7831862273879229,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046827689681201164,
      "loss": 5.3603,
      "mean_token_accuracy": 0.19356874674558638,
      "num_tokens": 41323831.0,
      "step": 20405
    },
    {
      "entropy": 6.009581708908081,
      "epoch": 1.7836231757406273,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004682609416778615,
      "loss": 5.3735,
      "mean_token_accuracy": 0.1928213581442833,
      "num_tokens": 41334295.0,
      "step": 20410
    },
    {
      "entropy": 5.96356348991394,
      "epoch": 1.7840601240933323,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004682449828368653,
      "loss": 5.364,
      "mean_token_accuracy": 0.19278717935085296,
      "num_tokens": 41344672.0,
      "step": 20415
    },
    {
      "entropy": 6.004062843322754,
      "epoch": 1.7844970724460367,
      "grad_norm": 1.125,
      "learning_rate": 0.0004682290202893291,
      "loss": 5.374,
      "mean_token_accuracy": 0.1902834579348564,
      "num_tokens": 41355108.0,
      "step": 20420
    },
    {
      "entropy": 6.0765709400177,
      "epoch": 1.7849340207987416,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046821305403555905,
      "loss": 5.4801,
      "mean_token_accuracy": 0.18523585945367813,
      "num_tokens": 41366277.0,
      "step": 20425
    },
    {
      "entropy": 6.019250440597534,
      "epoch": 1.7853709691514463,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046819708407586135,
      "loss": 5.3447,
      "mean_token_accuracy": 0.19644187688827514,
      "num_tokens": 41374757.0,
      "step": 20430
    },
    {
      "entropy": 5.951512861251831,
      "epoch": 1.785807917504151,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004681811104105423,
      "loss": 5.3306,
      "mean_token_accuracy": 0.20051602721214296,
      "num_tokens": 41384883.0,
      "step": 20435
    },
    {
      "entropy": 5.9537444591522215,
      "epoch": 1.7862448658568557,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004681651330399084,
      "loss": 5.3668,
      "mean_token_accuracy": 0.1960512161254883,
      "num_tokens": 41396492.0,
      "step": 20440
    },
    {
      "entropy": 5.95392484664917,
      "epoch": 1.7866818142095604,
      "grad_norm": 1.171875,
      "learning_rate": 0.000468149151964266,
      "loss": 5.3313,
      "mean_token_accuracy": 0.1954223781824112,
      "num_tokens": 41406217.0,
      "step": 20445
    },
    {
      "entropy": 5.957024002075196,
      "epoch": 1.787118762562265,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004681331671839217,
      "loss": 5.2851,
      "mean_token_accuracy": 0.19846437126398087,
      "num_tokens": 41415895.0,
      "step": 20450
    },
    {
      "entropy": 5.984970569610596,
      "epoch": 1.7875557109149698,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046811717869918196,
      "loss": 5.4185,
      "mean_token_accuracy": 0.18895849883556365,
      "num_tokens": 41426021.0,
      "step": 20455
    },
    {
      "entropy": 5.994244241714478,
      "epoch": 1.7879926592676747,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004681011865103536,
      "loss": 5.3236,
      "mean_token_accuracy": 0.19588702619075776,
      "num_tokens": 41434656.0,
      "step": 20460
    },
    {
      "entropy": 5.98708143234253,
      "epoch": 1.7884296076203792,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046808519061774324,
      "loss": 5.3639,
      "mean_token_accuracy": 0.18826863467693328,
      "num_tokens": 41443940.0,
      "step": 20465
    },
    {
      "entropy": 5.954663991928101,
      "epoch": 1.788866555973084,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046806919102165785,
      "loss": 5.4198,
      "mean_token_accuracy": 0.19702858626842498,
      "num_tokens": 41454907.0,
      "step": 20470
    },
    {
      "entropy": 5.962825965881348,
      "epoch": 1.7893035043257886,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004680531877224042,
      "loss": 5.2765,
      "mean_token_accuracy": 0.19965763688087462,
      "num_tokens": 41464248.0,
      "step": 20475
    },
    {
      "entropy": 5.94553427696228,
      "epoch": 1.7897404526784935,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004680371807202892,
      "loss": 5.317,
      "mean_token_accuracy": 0.2026899367570877,
      "num_tokens": 41475192.0,
      "step": 20480
    },
    {
      "entropy": 5.976460123062134,
      "epoch": 1.7901774010311982,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046802117001561996,
      "loss": 5.4218,
      "mean_token_accuracy": 0.18450224548578262,
      "num_tokens": 41484753.0,
      "step": 20485
    },
    {
      "entropy": 6.005007743835449,
      "epoch": 1.7906143493839028,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046800515560870346,
      "loss": 5.4089,
      "mean_token_accuracy": 0.1875214159488678,
      "num_tokens": 41495885.0,
      "step": 20490
    },
    {
      "entropy": 6.054569673538208,
      "epoch": 1.7910512977366075,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046798913749984694,
      "loss": 5.3718,
      "mean_token_accuracy": 0.19654789566993713,
      "num_tokens": 41506330.0,
      "step": 20495
    },
    {
      "entropy": 6.002534627914429,
      "epoch": 1.7914882460893122,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004679731156893577,
      "loss": 5.3662,
      "mean_token_accuracy": 0.19245149195194244,
      "num_tokens": 41515850.0,
      "step": 20500
    },
    {
      "entropy": 5.964077425003052,
      "epoch": 1.791925194442017,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000467957090177543,
      "loss": 5.3619,
      "mean_token_accuracy": 0.19491465240716935,
      "num_tokens": 41525910.0,
      "step": 20505
    },
    {
      "entropy": 5.964791345596313,
      "epoch": 1.7923621427947216,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046794106096471015,
      "loss": 5.2936,
      "mean_token_accuracy": 0.19656810462474822,
      "num_tokens": 41536417.0,
      "step": 20510
    },
    {
      "entropy": 5.953107929229736,
      "epoch": 1.7927990911474265,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004679250280511667,
      "loss": 5.3541,
      "mean_token_accuracy": 0.19228450506925582,
      "num_tokens": 41547215.0,
      "step": 20515
    },
    {
      "entropy": 6.005400609970093,
      "epoch": 1.793236039500131,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004679089914372201,
      "loss": 5.39,
      "mean_token_accuracy": 0.19130376428365709,
      "num_tokens": 41556036.0,
      "step": 20520
    },
    {
      "entropy": 5.959611177444458,
      "epoch": 1.793672987852836,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046789295112317796,
      "loss": 5.3819,
      "mean_token_accuracy": 0.19444650113582612,
      "num_tokens": 41565534.0,
      "step": 20525
    },
    {
      "entropy": 5.96167221069336,
      "epoch": 1.7941099362055404,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004678769071093479,
      "loss": 5.3313,
      "mean_token_accuracy": 0.19631841778755188,
      "num_tokens": 41576150.0,
      "step": 20530
    },
    {
      "entropy": 5.9254539012908936,
      "epoch": 1.7945468845582453,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046786085939603785,
      "loss": 5.3289,
      "mean_token_accuracy": 0.20324773490428924,
      "num_tokens": 41585248.0,
      "step": 20535
    },
    {
      "entropy": 5.952936697006225,
      "epoch": 1.7949838329109498,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004678448079835552,
      "loss": 5.3692,
      "mean_token_accuracy": 0.19381055384874343,
      "num_tokens": 41595215.0,
      "step": 20540
    },
    {
      "entropy": 6.029155302047729,
      "epoch": 1.7954207812636547,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004678287528722083,
      "loss": 5.4085,
      "mean_token_accuracy": 0.19139120429754258,
      "num_tokens": 41605268.0,
      "step": 20545
    },
    {
      "entropy": 5.9433846950531,
      "epoch": 1.7958577296163594,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004678126940623048,
      "loss": 5.2679,
      "mean_token_accuracy": 0.2019043266773224,
      "num_tokens": 41614554.0,
      "step": 20550
    },
    {
      "entropy": 5.94835467338562,
      "epoch": 1.796294677969064,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046779663155415287,
      "loss": 5.3784,
      "mean_token_accuracy": 0.19719138890504836,
      "num_tokens": 41624839.0,
      "step": 20555
    },
    {
      "entropy": 5.883698225021362,
      "epoch": 1.7967316263217687,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004677805653480604,
      "loss": 5.2555,
      "mean_token_accuracy": 0.21156887263059615,
      "num_tokens": 41634577.0,
      "step": 20560
    },
    {
      "entropy": 5.9624669551849365,
      "epoch": 1.7971685746744734,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046776449544433575,
      "loss": 5.4018,
      "mean_token_accuracy": 0.18531061708927155,
      "num_tokens": 41645213.0,
      "step": 20565
    },
    {
      "entropy": 5.932422208786011,
      "epoch": 1.7976055230271784,
      "grad_norm": 1.078125,
      "learning_rate": 0.000467748421843287,
      "loss": 5.2675,
      "mean_token_accuracy": 0.20247265994548796,
      "num_tokens": 41655160.0,
      "step": 20570
    },
    {
      "entropy": 5.962741184234619,
      "epoch": 1.7980424713798828,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046773234454522265,
      "loss": 5.3365,
      "mean_token_accuracy": 0.19085452258586882,
      "num_tokens": 41665861.0,
      "step": 20575
    },
    {
      "entropy": 5.978448438644409,
      "epoch": 1.7984794197325877,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046771626355045085,
      "loss": 5.3655,
      "mean_token_accuracy": 0.19811732023954393,
      "num_tokens": 41676699.0,
      "step": 20580
    },
    {
      "entropy": 5.991144037246704,
      "epoch": 1.7989163680852922,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004677001788592801,
      "loss": 5.3521,
      "mean_token_accuracy": 0.1998131662607193,
      "num_tokens": 41688162.0,
      "step": 20585
    },
    {
      "entropy": 5.988797330856324,
      "epoch": 1.7993533164379971,
      "grad_norm": 1.15625,
      "learning_rate": 0.000467684090472019,
      "loss": 5.3911,
      "mean_token_accuracy": 0.19503619968891145,
      "num_tokens": 41698088.0,
      "step": 20590
    },
    {
      "entropy": 5.964185953140259,
      "epoch": 1.7997902647907016,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004676679983889761,
      "loss": 5.3086,
      "mean_token_accuracy": 0.2052853062748909,
      "num_tokens": 41707446.0,
      "step": 20595
    },
    {
      "entropy": 5.987864255905151,
      "epoch": 1.8002272131434065,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046765190261046,
      "loss": 5.415,
      "mean_token_accuracy": 0.18706704974174498,
      "num_tokens": 41716388.0,
      "step": 20600
    },
    {
      "entropy": 5.984219408035278,
      "epoch": 1.8006641614961112,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004676358031367795,
      "loss": 5.4108,
      "mean_token_accuracy": 0.1964862674474716,
      "num_tokens": 41726328.0,
      "step": 20605
    },
    {
      "entropy": 6.010001993179321,
      "epoch": 1.8011011098488159,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046761969996824326,
      "loss": 5.3232,
      "mean_token_accuracy": 0.1936813101172447,
      "num_tokens": 41736128.0,
      "step": 20610
    },
    {
      "entropy": 5.9462262153625485,
      "epoch": 1.8015380582015206,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004676035931051604,
      "loss": 5.3056,
      "mean_token_accuracy": 0.1985646441578865,
      "num_tokens": 41746316.0,
      "step": 20615
    },
    {
      "entropy": 5.933775138854981,
      "epoch": 1.8019750065542253,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046758748254783957,
      "loss": 5.2742,
      "mean_token_accuracy": 0.2034897670149803,
      "num_tokens": 41756015.0,
      "step": 20620
    },
    {
      "entropy": 5.959750938415527,
      "epoch": 1.80241195490693,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046757136829658996,
      "loss": 5.3081,
      "mean_token_accuracy": 0.2024301528930664,
      "num_tokens": 41766360.0,
      "step": 20625
    },
    {
      "entropy": 6.0136803150177,
      "epoch": 1.8028489032596346,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046755525035172056,
      "loss": 5.4118,
      "mean_token_accuracy": 0.1937864601612091,
      "num_tokens": 41776640.0,
      "step": 20630
    },
    {
      "entropy": 5.964665412902832,
      "epoch": 1.8032858516123396,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004675391287135406,
      "loss": 5.4098,
      "mean_token_accuracy": 0.19876327216625214,
      "num_tokens": 41787329.0,
      "step": 20635
    },
    {
      "entropy": 6.004807424545288,
      "epoch": 1.803722799965044,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046752300338235924,
      "loss": 5.3825,
      "mean_token_accuracy": 0.1888030618429184,
      "num_tokens": 41797890.0,
      "step": 20640
    },
    {
      "entropy": 5.971687889099121,
      "epoch": 1.804159748317749,
      "grad_norm": 1.125,
      "learning_rate": 0.0004675068743584858,
      "loss": 5.3965,
      "mean_token_accuracy": 0.19000955820083618,
      "num_tokens": 41808845.0,
      "step": 20645
    },
    {
      "entropy": 5.962868642807007,
      "epoch": 1.8045966966704534,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004674907416422297,
      "loss": 5.3725,
      "mean_token_accuracy": 0.19740869104862213,
      "num_tokens": 41819499.0,
      "step": 20650
    },
    {
      "entropy": 6.0076569557189945,
      "epoch": 1.8050336450231583,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004674746052339002,
      "loss": 5.4038,
      "mean_token_accuracy": 0.1857535347342491,
      "num_tokens": 41827998.0,
      "step": 20655
    },
    {
      "entropy": 5.96401743888855,
      "epoch": 1.805470593375863,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046745846513380704,
      "loss": 5.394,
      "mean_token_accuracy": 0.19180378019809724,
      "num_tokens": 41839610.0,
      "step": 20660
    },
    {
      "entropy": 6.014907360076904,
      "epoch": 1.8059075417285677,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046744232134225954,
      "loss": 5.3122,
      "mean_token_accuracy": 0.19423591196537018,
      "num_tokens": 41850111.0,
      "step": 20665
    },
    {
      "entropy": 5.982804203033448,
      "epoch": 1.8063444900812724,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004674261738595676,
      "loss": 5.3452,
      "mean_token_accuracy": 0.19074386060237886,
      "num_tokens": 41859843.0,
      "step": 20670
    },
    {
      "entropy": 5.971756792068481,
      "epoch": 1.806781438433977,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046741002268604073,
      "loss": 5.3485,
      "mean_token_accuracy": 0.19779014140367507,
      "num_tokens": 41868864.0,
      "step": 20675
    },
    {
      "entropy": 5.9598226070404055,
      "epoch": 1.8072183867866818,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046739386782198883,
      "loss": 5.3711,
      "mean_token_accuracy": 0.19567952305078506,
      "num_tokens": 41878274.0,
      "step": 20680
    },
    {
      "entropy": 5.958711719512939,
      "epoch": 1.8076553351393865,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046737770926772175,
      "loss": 5.337,
      "mean_token_accuracy": 0.1938385933637619,
      "num_tokens": 41888005.0,
      "step": 20685
    },
    {
      "entropy": 6.052155208587647,
      "epoch": 1.8080922834920914,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004673615470235494,
      "loss": 5.4594,
      "mean_token_accuracy": 0.18697259873151778,
      "num_tokens": 41897415.0,
      "step": 20690
    },
    {
      "entropy": 5.986123657226562,
      "epoch": 1.8085292318447959,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004673453810897817,
      "loss": 5.3326,
      "mean_token_accuracy": 0.1972718968987465,
      "num_tokens": 41907021.0,
      "step": 20695
    },
    {
      "entropy": 5.9528210163116455,
      "epoch": 1.8089661801975008,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004673292114667289,
      "loss": 5.4095,
      "mean_token_accuracy": 0.1921335518360138,
      "num_tokens": 41918358.0,
      "step": 20700
    },
    {
      "entropy": 6.03670563697815,
      "epoch": 1.8094031285502052,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046731303815470104,
      "loss": 5.4123,
      "mean_token_accuracy": 0.18147315233945846,
      "num_tokens": 41929311.0,
      "step": 20705
    },
    {
      "entropy": 5.986992073059082,
      "epoch": 1.8098400769029102,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046729686115400826,
      "loss": 5.3696,
      "mean_token_accuracy": 0.19261527806520462,
      "num_tokens": 41939623.0,
      "step": 20710
    },
    {
      "entropy": 5.937793159484864,
      "epoch": 1.8102770252556146,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004672806804649611,
      "loss": 5.3323,
      "mean_token_accuracy": 0.19550692290067673,
      "num_tokens": 41949650.0,
      "step": 20715
    },
    {
      "entropy": 5.960168790817261,
      "epoch": 1.8107139736083195,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046726449608786957,
      "loss": 5.3172,
      "mean_token_accuracy": 0.19985006451606752,
      "num_tokens": 41960262.0,
      "step": 20720
    },
    {
      "entropy": 5.99585223197937,
      "epoch": 1.8111509219610242,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004672483080230443,
      "loss": 5.3681,
      "mean_token_accuracy": 0.1865314155817032,
      "num_tokens": 41969874.0,
      "step": 20725
    },
    {
      "entropy": 6.003401231765747,
      "epoch": 1.811587870313729,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046723211627079577,
      "loss": 5.3506,
      "mean_token_accuracy": 0.1937719151377678,
      "num_tokens": 41978810.0,
      "step": 20730
    },
    {
      "entropy": 5.949249649047852,
      "epoch": 1.8120248186664336,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004672159208314344,
      "loss": 5.3134,
      "mean_token_accuracy": 0.1966993048787117,
      "num_tokens": 41988785.0,
      "step": 20735
    },
    {
      "entropy": 5.926423120498657,
      "epoch": 1.8124617670191383,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004671997217052711,
      "loss": 5.3638,
      "mean_token_accuracy": 0.19981876611709595,
      "num_tokens": 41998698.0,
      "step": 20740
    },
    {
      "entropy": 5.939173364639283,
      "epoch": 1.812898715371843,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004671835188926164,
      "loss": 5.3283,
      "mean_token_accuracy": 0.19962072521448135,
      "num_tokens": 42009082.0,
      "step": 20745
    },
    {
      "entropy": 6.0124725818634035,
      "epoch": 1.8133356637245477,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004671673123937811,
      "loss": 5.3456,
      "mean_token_accuracy": 0.19037179797887802,
      "num_tokens": 42019960.0,
      "step": 20750
    },
    {
      "entropy": 5.918795013427735,
      "epoch": 1.8137726120772526,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000467151102209076,
      "loss": 5.2848,
      "mean_token_accuracy": 0.19726587533950807,
      "num_tokens": 42028799.0,
      "step": 20755
    },
    {
      "entropy": 5.928885221481323,
      "epoch": 1.814209560429957,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046713488833881214,
      "loss": 5.3408,
      "mean_token_accuracy": 0.18746046125888824,
      "num_tokens": 42038967.0,
      "step": 20760
    },
    {
      "entropy": 5.956796503067016,
      "epoch": 1.814646508782662,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046711867078330047,
      "loss": 5.3221,
      "mean_token_accuracy": 0.19678547382354736,
      "num_tokens": 42049982.0,
      "step": 20765
    },
    {
      "entropy": 5.990020990371704,
      "epoch": 1.8150834571353665,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046710244954285184,
      "loss": 5.4329,
      "mean_token_accuracy": 0.19020504355430604,
      "num_tokens": 42059835.0,
      "step": 20770
    },
    {
      "entropy": 5.952736473083496,
      "epoch": 1.8155204054880714,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046708622461777774,
      "loss": 5.2438,
      "mean_token_accuracy": 0.20219845026731492,
      "num_tokens": 42069339.0,
      "step": 20775
    },
    {
      "entropy": 5.914933490753174,
      "epoch": 1.815957353840776,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004670699960083892,
      "loss": 5.2788,
      "mean_token_accuracy": 0.19591066837310792,
      "num_tokens": 42080618.0,
      "step": 20780
    },
    {
      "entropy": 6.0219597816467285,
      "epoch": 1.8163943021934807,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004670537637149974,
      "loss": 5.438,
      "mean_token_accuracy": 0.1862032577395439,
      "num_tokens": 42091458.0,
      "step": 20785
    },
    {
      "entropy": 5.981135702133178,
      "epoch": 1.8168312505461854,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046703752773791387,
      "loss": 5.4132,
      "mean_token_accuracy": 0.19545188397169114,
      "num_tokens": 42102515.0,
      "step": 20790
    },
    {
      "entropy": 5.969221925735473,
      "epoch": 1.8172681988988901,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004670212880774499,
      "loss": 5.4061,
      "mean_token_accuracy": 0.18677465319633485,
      "num_tokens": 42113853.0,
      "step": 20795
    },
    {
      "entropy": 5.991488790512085,
      "epoch": 1.8177051472515948,
      "grad_norm": 1.078125,
      "learning_rate": 0.000467005044733917,
      "loss": 5.3928,
      "mean_token_accuracy": 0.18884593099355698,
      "num_tokens": 42123101.0,
      "step": 20800
    },
    {
      "entropy": 6.014171457290649,
      "epoch": 1.8181420956042995,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004669887977076267,
      "loss": 5.3575,
      "mean_token_accuracy": 0.19732308983802796,
      "num_tokens": 42133051.0,
      "step": 20805
    },
    {
      "entropy": 5.97155499458313,
      "epoch": 1.8185790439570044,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046697254699889064,
      "loss": 5.349,
      "mean_token_accuracy": 0.1996772199869156,
      "num_tokens": 42142705.0,
      "step": 20810
    },
    {
      "entropy": 5.916130876541137,
      "epoch": 1.819015992309709,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046695629260802064,
      "loss": 5.3345,
      "mean_token_accuracy": 0.19637890607118608,
      "num_tokens": 42152399.0,
      "step": 20815
    },
    {
      "entropy": 5.961959505081177,
      "epoch": 1.8194529406624138,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004669400345353283,
      "loss": 5.3558,
      "mean_token_accuracy": 0.19813742041587828,
      "num_tokens": 42162651.0,
      "step": 20820
    },
    {
      "entropy": 5.900429677963257,
      "epoch": 1.8198898890151183,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004669237727811255,
      "loss": 5.2533,
      "mean_token_accuracy": 0.20070793032646178,
      "num_tokens": 42172595.0,
      "step": 20825
    },
    {
      "entropy": 5.942414999008179,
      "epoch": 1.8203268373678232,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004669075073457242,
      "loss": 5.3616,
      "mean_token_accuracy": 0.1962771937251091,
      "num_tokens": 42182631.0,
      "step": 20830
    },
    {
      "entropy": 5.955499172210693,
      "epoch": 1.8207637857205279,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046689123822943643,
      "loss": 5.3467,
      "mean_token_accuracy": 0.19167760759592056,
      "num_tokens": 42192743.0,
      "step": 20835
    },
    {
      "entropy": 5.9779966354370115,
      "epoch": 1.8212007340732326,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004668749654325741,
      "loss": 5.3391,
      "mean_token_accuracy": 0.19472722113132476,
      "num_tokens": 42203090.0,
      "step": 20840
    },
    {
      "entropy": 5.978245687484741,
      "epoch": 1.8216376824259373,
      "grad_norm": 1.125,
      "learning_rate": 0.0004668586889554493,
      "loss": 5.4199,
      "mean_token_accuracy": 0.19089269936084746,
      "num_tokens": 42213341.0,
      "step": 20845
    },
    {
      "entropy": 5.970167398452759,
      "epoch": 1.822074630778642,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004668424087983744,
      "loss": 5.4197,
      "mean_token_accuracy": 0.19357079416513442,
      "num_tokens": 42222862.0,
      "step": 20850
    },
    {
      "entropy": 5.9638689994812015,
      "epoch": 1.8225115791313466,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046682612496166156,
      "loss": 5.4074,
      "mean_token_accuracy": 0.18884130716323852,
      "num_tokens": 42234045.0,
      "step": 20855
    },
    {
      "entropy": 6.026735877990722,
      "epoch": 1.8229485274840513,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046680983744562315,
      "loss": 5.3329,
      "mean_token_accuracy": 0.19571761339902877,
      "num_tokens": 42244093.0,
      "step": 20860
    },
    {
      "entropy": 5.933075094223023,
      "epoch": 1.8233854758367563,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004667935462505716,
      "loss": 5.3915,
      "mean_token_accuracy": 0.1935323193669319,
      "num_tokens": 42254865.0,
      "step": 20865
    },
    {
      "entropy": 5.937422227859497,
      "epoch": 1.8238224241894607,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046677725137681924,
      "loss": 5.303,
      "mean_token_accuracy": 0.2000357761979103,
      "num_tokens": 42264565.0,
      "step": 20870
    },
    {
      "entropy": 6.00654296875,
      "epoch": 1.8242593725421656,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046676095282467873,
      "loss": 5.4109,
      "mean_token_accuracy": 0.1904759079217911,
      "num_tokens": 42274289.0,
      "step": 20875
    },
    {
      "entropy": 6.020161771774292,
      "epoch": 1.82469632089487,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004667446505944627,
      "loss": 5.3583,
      "mean_token_accuracy": 0.19277170598506926,
      "num_tokens": 42283855.0,
      "step": 20880
    },
    {
      "entropy": 5.919912910461425,
      "epoch": 1.825133269247575,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046672834468648375,
      "loss": 5.3713,
      "mean_token_accuracy": 0.20074865520000457,
      "num_tokens": 42294871.0,
      "step": 20885
    },
    {
      "entropy": 5.975374698638916,
      "epoch": 1.8255702176002795,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004667120351010547,
      "loss": 5.3379,
      "mean_token_accuracy": 0.19493301808834076,
      "num_tokens": 42305315.0,
      "step": 20890
    },
    {
      "entropy": 5.980936050415039,
      "epoch": 1.8260071659529844,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046669572183848833,
      "loss": 5.3006,
      "mean_token_accuracy": 0.19916147589683533,
      "num_tokens": 42316893.0,
      "step": 20895
    },
    {
      "entropy": 5.924517011642456,
      "epoch": 1.826444114305689,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046667940489909766,
      "loss": 5.2989,
      "mean_token_accuracy": 0.19984648674726485,
      "num_tokens": 42326784.0,
      "step": 20900
    },
    {
      "entropy": 5.89840521812439,
      "epoch": 1.8268810626583938,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046666308428319547,
      "loss": 5.2893,
      "mean_token_accuracy": 0.2040782243013382,
      "num_tokens": 42336283.0,
      "step": 20905
    },
    {
      "entropy": 5.914764928817749,
      "epoch": 1.8273180110110985,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046664675999109494,
      "loss": 5.3139,
      "mean_token_accuracy": 0.203636434674263,
      "num_tokens": 42346933.0,
      "step": 20910
    },
    {
      "entropy": 5.9465855121612545,
      "epoch": 1.8277549593638032,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004666304320231091,
      "loss": 5.3808,
      "mean_token_accuracy": 0.1966951608657837,
      "num_tokens": 42357616.0,
      "step": 20915
    },
    {
      "entropy": 5.993144702911377,
      "epoch": 1.8281919077165079,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004666141003795512,
      "loss": 5.3652,
      "mean_token_accuracy": 0.1973184674978256,
      "num_tokens": 42367566.0,
      "step": 20920
    },
    {
      "entropy": 5.982747030258179,
      "epoch": 1.8286288560692125,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046659776506073446,
      "loss": 5.3204,
      "mean_token_accuracy": 0.20045436322689056,
      "num_tokens": 42377412.0,
      "step": 20925
    },
    {
      "entropy": 5.988457584381104,
      "epoch": 1.8290658044219175,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046658142606697216,
      "loss": 5.3377,
      "mean_token_accuracy": 0.19790948182344437,
      "num_tokens": 42387952.0,
      "step": 20930
    },
    {
      "entropy": 5.9854736328125,
      "epoch": 1.829502752774622,
      "grad_norm": 2.359375,
      "learning_rate": 0.0004665650833985777,
      "loss": 5.3029,
      "mean_token_accuracy": 0.19394005835056305,
      "num_tokens": 42398488.0,
      "step": 20935
    },
    {
      "entropy": 5.962787055969239,
      "epoch": 1.8299397011273268,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046654873705586464,
      "loss": 5.328,
      "mean_token_accuracy": 0.19906988739967346,
      "num_tokens": 42409137.0,
      "step": 20940
    },
    {
      "entropy": 6.024931526184082,
      "epoch": 1.8303766494800313,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004665323870391464,
      "loss": 5.4402,
      "mean_token_accuracy": 0.18443988561630248,
      "num_tokens": 42419867.0,
      "step": 20945
    },
    {
      "entropy": 5.975014305114746,
      "epoch": 1.8308135978327362,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004665160333487365,
      "loss": 5.3659,
      "mean_token_accuracy": 0.1961405798792839,
      "num_tokens": 42429590.0,
      "step": 20950
    },
    {
      "entropy": 5.9745433807373045,
      "epoch": 1.831250546185441,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004664996759849489,
      "loss": 5.3258,
      "mean_token_accuracy": 0.20013615787029265,
      "num_tokens": 42439777.0,
      "step": 20955
    },
    {
      "entropy": 5.918057537078857,
      "epoch": 1.8316874945381456,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004664833149480971,
      "loss": 5.3583,
      "mean_token_accuracy": 0.2003895491361618,
      "num_tokens": 42449843.0,
      "step": 20960
    },
    {
      "entropy": 5.884631538391114,
      "epoch": 1.8321244428908503,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046646695023849506,
      "loss": 5.2228,
      "mean_token_accuracy": 0.20322444587945937,
      "num_tokens": 42459056.0,
      "step": 20965
    },
    {
      "entropy": 5.956286859512329,
      "epoch": 1.832561391243555,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046645058185645644,
      "loss": 5.2719,
      "mean_token_accuracy": 0.1981494441628456,
      "num_tokens": 42468946.0,
      "step": 20970
    },
    {
      "entropy": 5.944397211074829,
      "epoch": 1.8329983395962597,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004664342098022954,
      "loss": 5.3381,
      "mean_token_accuracy": 0.19290238320827485,
      "num_tokens": 42478894.0,
      "step": 20975
    },
    {
      "entropy": 5.927860116958618,
      "epoch": 1.8334352879489644,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004664178340763259,
      "loss": 5.3321,
      "mean_token_accuracy": 0.20186780095100404,
      "num_tokens": 42490124.0,
      "step": 20980
    },
    {
      "entropy": 6.001922464370727,
      "epoch": 1.8338722363016693,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004664014546788621,
      "loss": 5.3771,
      "mean_token_accuracy": 0.19849551171064378,
      "num_tokens": 42499554.0,
      "step": 20985
    },
    {
      "entropy": 5.920125246047974,
      "epoch": 1.8343091846543738,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046638507161021807,
      "loss": 5.2334,
      "mean_token_accuracy": 0.20071903765201568,
      "num_tokens": 42510364.0,
      "step": 20990
    },
    {
      "entropy": 5.999049139022827,
      "epoch": 1.8347461330070787,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046636868487070804,
      "loss": 5.4119,
      "mean_token_accuracy": 0.18601226210594177,
      "num_tokens": 42521671.0,
      "step": 20995
    },
    {
      "entropy": 5.9901519298553465,
      "epoch": 1.8351830813597831,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046635229446064647,
      "loss": 5.3105,
      "mean_token_accuracy": 0.19882202595472337,
      "num_tokens": 42532784.0,
      "step": 21000
    },
    {
      "epoch": 1.8351830813597831,
      "eval_entropy": 5.732359239806206,
      "eval_loss": 5.370952606201172,
      "eval_mean_token_accuracy": 0.20386309489150503,
      "eval_num_tokens": 42532784.0,
      "eval_runtime": 21.6169,
      "eval_samples_per_second": 1712.271,
      "eval_steps_per_second": 214.045,
      "step": 21000
    }
  ],
  "logging_steps": 5,
  "max_steps": 114420,
  "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": 9441461251952640.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}