{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.146028139473914,
  "eval_steps": 3000,
  "global_step": 36000,
  "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
    },
    {
      "entropy": 5.8786396980285645,
      "epoch": 1.835620029712488,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004663359003803475,
      "loss": 5.2625,
      "mean_token_accuracy": 0.20917077958583832,
      "num_tokens": 42542502.0,
      "step": 21005
    },
    {
      "entropy": 5.929272699356079,
      "epoch": 1.8360569780651927,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004663195026301258,
      "loss": 5.2877,
      "mean_token_accuracy": 0.195219586789608,
      "num_tokens": 42551837.0,
      "step": 21010
    },
    {
      "entropy": 5.992280721664429,
      "epoch": 1.8364939264178974,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046630310121029573,
      "loss": 5.3307,
      "mean_token_accuracy": 0.19740208387374877,
      "num_tokens": 42560537.0,
      "step": 21015
    },
    {
      "entropy": 5.987287807464599,
      "epoch": 1.8369308747706021,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046628669612117194,
      "loss": 5.3559,
      "mean_token_accuracy": 0.1935727670788765,
      "num_tokens": 42570386.0,
      "step": 21020
    },
    {
      "entropy": 5.9564755916595455,
      "epoch": 1.8373678231233068,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046627028736306907,
      "loss": 5.3971,
      "mean_token_accuracy": 0.18894872218370437,
      "num_tokens": 42579860.0,
      "step": 21025
    },
    {
      "entropy": 6.004134607315064,
      "epoch": 1.8378047714760115,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046625387493630187,
      "loss": 5.3783,
      "mean_token_accuracy": 0.19392509907484054,
      "num_tokens": 42590320.0,
      "step": 21030
    },
    {
      "entropy": 6.009503746032715,
      "epoch": 1.8382417198287162,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004662374588411852,
      "loss": 5.3221,
      "mean_token_accuracy": 0.19767068326473236,
      "num_tokens": 42600331.0,
      "step": 21035
    },
    {
      "entropy": 5.927414178848267,
      "epoch": 1.8386786681814211,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046622103907803384,
      "loss": 5.3527,
      "mean_token_accuracy": 0.1995800569653511,
      "num_tokens": 42610860.0,
      "step": 21040
    },
    {
      "entropy": 5.972218608856201,
      "epoch": 1.8391156165341256,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046620461564716265,
      "loss": 5.3941,
      "mean_token_accuracy": 0.18869002312421798,
      "num_tokens": 42620305.0,
      "step": 21045
    },
    {
      "entropy": 5.9703638553619385,
      "epoch": 1.8395525648868305,
      "grad_norm": 1.125,
      "learning_rate": 0.0004661881885488869,
      "loss": 5.2037,
      "mean_token_accuracy": 0.2125360459089279,
      "num_tokens": 42629667.0,
      "step": 21050
    },
    {
      "entropy": 5.935352993011475,
      "epoch": 1.839989513239535,
      "grad_norm": 1.125,
      "learning_rate": 0.0004661717577835214,
      "loss": 5.4024,
      "mean_token_accuracy": 0.18830619901418685,
      "num_tokens": 42640116.0,
      "step": 21055
    },
    {
      "entropy": 5.900828552246094,
      "epoch": 1.8404264615922399,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004661553233513815,
      "loss": 5.3116,
      "mean_token_accuracy": 0.19940631687641144,
      "num_tokens": 42650014.0,
      "step": 21060
    },
    {
      "entropy": 6.00567045211792,
      "epoch": 1.8408634099449444,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004661388852527822,
      "loss": 5.4028,
      "mean_token_accuracy": 0.19217278957366943,
      "num_tokens": 42661280.0,
      "step": 21065
    },
    {
      "entropy": 5.995460510253906,
      "epoch": 1.8413003582976493,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046612244348803906,
      "loss": 5.3971,
      "mean_token_accuracy": 0.19706904590129853,
      "num_tokens": 42671530.0,
      "step": 21070
    },
    {
      "entropy": 5.971308803558349,
      "epoch": 1.841737306650354,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004661059980574672,
      "loss": 5.3713,
      "mean_token_accuracy": 0.19689275920391083,
      "num_tokens": 42682738.0,
      "step": 21075
    },
    {
      "entropy": 5.953145170211792,
      "epoch": 1.8421742550030586,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004660895489613822,
      "loss": 5.374,
      "mean_token_accuracy": 0.1956367313861847,
      "num_tokens": 42694271.0,
      "step": 21080
    },
    {
      "entropy": 5.9715556621551515,
      "epoch": 1.8426112033557633,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004660730962000996,
      "loss": 5.4274,
      "mean_token_accuracy": 0.18757243305444718,
      "num_tokens": 42704896.0,
      "step": 21085
    },
    {
      "entropy": 5.995157861709595,
      "epoch": 1.843048151708468,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004660566397739348,
      "loss": 5.3107,
      "mean_token_accuracy": 0.19387827068567276,
      "num_tokens": 42715506.0,
      "step": 21090
    },
    {
      "entropy": 6.022855186462403,
      "epoch": 1.8434851000611727,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046604017968320357,
      "loss": 5.3387,
      "mean_token_accuracy": 0.19733892232179642,
      "num_tokens": 42725428.0,
      "step": 21095
    },
    {
      "entropy": 5.935022687911987,
      "epoch": 1.8439220484138774,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004660237159282216,
      "loss": 5.3288,
      "mean_token_accuracy": 0.20399669110774993,
      "num_tokens": 42735105.0,
      "step": 21100
    },
    {
      "entropy": 5.917203092575074,
      "epoch": 1.8443589967665823,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046600724850930466,
      "loss": 5.2902,
      "mean_token_accuracy": 0.20530568212270736,
      "num_tokens": 42744191.0,
      "step": 21105
    },
    {
      "entropy": 6.018963289260864,
      "epoch": 1.8447959451192868,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004659907774267686,
      "loss": 5.4492,
      "mean_token_accuracy": 0.19052611589431762,
      "num_tokens": 42755416.0,
      "step": 21110
    },
    {
      "entropy": 5.987232494354248,
      "epoch": 1.8452328934719917,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046597430268092933,
      "loss": 5.2647,
      "mean_token_accuracy": 0.20135606527328492,
      "num_tokens": 42764974.0,
      "step": 21115
    },
    {
      "entropy": 5.926096677780151,
      "epoch": 1.8456698418246962,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046595782427210285,
      "loss": 5.2518,
      "mean_token_accuracy": 0.1989995926618576,
      "num_tokens": 42775080.0,
      "step": 21120
    },
    {
      "entropy": 5.908840036392212,
      "epoch": 1.846106790177401,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004659413422006053,
      "loss": 5.3295,
      "mean_token_accuracy": 0.1948503941297531,
      "num_tokens": 42784333.0,
      "step": 21125
    },
    {
      "entropy": 6.020711374282837,
      "epoch": 1.8465437385301058,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004659248564667527,
      "loss": 5.3466,
      "mean_token_accuracy": 0.19612756073474885,
      "num_tokens": 42792912.0,
      "step": 21130
    },
    {
      "entropy": 5.982763671875,
      "epoch": 1.8469806868828105,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046590836707086135,
      "loss": 5.3835,
      "mean_token_accuracy": 0.19051459133625032,
      "num_tokens": 42803812.0,
      "step": 21135
    },
    {
      "entropy": 6.0001002788543705,
      "epoch": 1.8474176352355152,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046589187401324753,
      "loss": 5.4228,
      "mean_token_accuracy": 0.1915706604719162,
      "num_tokens": 42815172.0,
      "step": 21140
    },
    {
      "entropy": 6.010838174819947,
      "epoch": 1.8478545835882199,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004658753772942274,
      "loss": 5.3027,
      "mean_token_accuracy": 0.19950002878904344,
      "num_tokens": 42824958.0,
      "step": 21145
    },
    {
      "entropy": 5.950379705429077,
      "epoch": 1.8482915319409245,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046585887691411764,
      "loss": 5.3946,
      "mean_token_accuracy": 0.19144682139158248,
      "num_tokens": 42834940.0,
      "step": 21150
    },
    {
      "entropy": 5.9229084014892575,
      "epoch": 1.8487284802936292,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004658423728732346,
      "loss": 5.2668,
      "mean_token_accuracy": 0.20359189212322235,
      "num_tokens": 42845102.0,
      "step": 21155
    },
    {
      "entropy": 5.912109804153443,
      "epoch": 1.8491654286463342,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004658258651718948,
      "loss": 5.2919,
      "mean_token_accuracy": 0.20210865885019302,
      "num_tokens": 42853876.0,
      "step": 21160
    },
    {
      "entropy": 5.986409425735474,
      "epoch": 1.8496023769990386,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046580935381041496,
      "loss": 5.4465,
      "mean_token_accuracy": 0.18192440271377563,
      "num_tokens": 42863592.0,
      "step": 21165
    },
    {
      "entropy": 5.963451480865478,
      "epoch": 1.8500393253517435,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004657928387891117,
      "loss": 5.287,
      "mean_token_accuracy": 0.2027474343776703,
      "num_tokens": 42873478.0,
      "step": 21170
    },
    {
      "entropy": 6.022353792190552,
      "epoch": 1.850476273704448,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046577632010830184,
      "loss": 5.2823,
      "mean_token_accuracy": 0.20333863496780397,
      "num_tokens": 42882591.0,
      "step": 21175
    },
    {
      "entropy": 5.8997344970703125,
      "epoch": 1.850913222057153,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004657597977683022,
      "loss": 5.2449,
      "mean_token_accuracy": 0.20732143819332122,
      "num_tokens": 42892391.0,
      "step": 21180
    },
    {
      "entropy": 5.9400559425354,
      "epoch": 1.8513501704098576,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004657432717694297,
      "loss": 5.3385,
      "mean_token_accuracy": 0.19634882509708404,
      "num_tokens": 42902839.0,
      "step": 21185
    },
    {
      "entropy": 6.024962902069092,
      "epoch": 1.8517871187625623,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004657267421120012,
      "loss": 5.4238,
      "mean_token_accuracy": 0.18829901963472367,
      "num_tokens": 42912404.0,
      "step": 21190
    },
    {
      "entropy": 6.015098571777344,
      "epoch": 1.852224067115267,
      "grad_norm": 1.0,
      "learning_rate": 0.0004657102087963339,
      "loss": 5.326,
      "mean_token_accuracy": 0.19925577491521834,
      "num_tokens": 42922939.0,
      "step": 21195
    },
    {
      "entropy": 5.916782569885254,
      "epoch": 1.8526610154679717,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046569367182274494,
      "loss": 5.3209,
      "mean_token_accuracy": 0.2001183107495308,
      "num_tokens": 42932095.0,
      "step": 21200
    },
    {
      "entropy": 5.972196483612061,
      "epoch": 1.8530979638206764,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004656771311915513,
      "loss": 5.3697,
      "mean_token_accuracy": 0.19507252126932145,
      "num_tokens": 42941910.0,
      "step": 21205
    },
    {
      "entropy": 5.986444187164307,
      "epoch": 1.853534912173381,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004656605869030704,
      "loss": 5.3254,
      "mean_token_accuracy": 0.20411270260810851,
      "num_tokens": 42952869.0,
      "step": 21210
    },
    {
      "entropy": 5.98563323020935,
      "epoch": 1.853971860526086,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046564403895761966,
      "loss": 5.4855,
      "mean_token_accuracy": 0.18780758380889892,
      "num_tokens": 42963300.0,
      "step": 21215
    },
    {
      "entropy": 6.007974767684937,
      "epoch": 1.8544088088787904,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004656274873555162,
      "loss": 5.2861,
      "mean_token_accuracy": 0.19782135635614395,
      "num_tokens": 42973505.0,
      "step": 21220
    },
    {
      "entropy": 5.940877199172974,
      "epoch": 1.8548457572314954,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004656109320970778,
      "loss": 5.3333,
      "mean_token_accuracy": 0.19467673897743226,
      "num_tokens": 42982813.0,
      "step": 21225
    },
    {
      "entropy": 5.90571084022522,
      "epoch": 1.8552827055841998,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004655943731826217,
      "loss": 5.3678,
      "mean_token_accuracy": 0.19594168066978454,
      "num_tokens": 42991792.0,
      "step": 21230
    },
    {
      "entropy": 5.910856151580811,
      "epoch": 1.8557196539369047,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004655778106124658,
      "loss": 5.29,
      "mean_token_accuracy": 0.20972459614276887,
      "num_tokens": 43002191.0,
      "step": 21235
    },
    {
      "entropy": 5.96629695892334,
      "epoch": 1.8561566022896092,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004655612443869275,
      "loss": 5.306,
      "mean_token_accuracy": 0.20009464174509048,
      "num_tokens": 43013771.0,
      "step": 21240
    },
    {
      "entropy": 5.974319314956665,
      "epoch": 1.8565935506423141,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004655446745063247,
      "loss": 5.4076,
      "mean_token_accuracy": 0.18504276871681213,
      "num_tokens": 43023982.0,
      "step": 21245
    },
    {
      "entropy": 6.013309955596924,
      "epoch": 1.8570304989950188,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004655281009709753,
      "loss": 5.4082,
      "mean_token_accuracy": 0.18903540968894958,
      "num_tokens": 43034827.0,
      "step": 21250
    },
    {
      "entropy": 5.96611442565918,
      "epoch": 1.8574674473477235,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046551152378119696,
      "loss": 5.2802,
      "mean_token_accuracy": 0.20682353079319,
      "num_tokens": 43044811.0,
      "step": 21255
    },
    {
      "entropy": 5.9746068000793455,
      "epoch": 1.8579043957004282,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004654949429373078,
      "loss": 5.3578,
      "mean_token_accuracy": 0.19759218245744706,
      "num_tokens": 43054513.0,
      "step": 21260
    },
    {
      "entropy": 5.9367938995361325,
      "epoch": 1.858341344053133,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004654783584396258,
      "loss": 5.3492,
      "mean_token_accuracy": 0.19405039250850678,
      "num_tokens": 43064356.0,
      "step": 21265
    },
    {
      "entropy": 5.921123886108399,
      "epoch": 1.8587782924058376,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046546177028846915,
      "loss": 5.225,
      "mean_token_accuracy": 0.20465039908885957,
      "num_tokens": 43074663.0,
      "step": 21270
    },
    {
      "entropy": 5.981184720993042,
      "epoch": 1.8592152407585423,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004654451784841559,
      "loss": 5.3326,
      "mean_token_accuracy": 0.19548612087965012,
      "num_tokens": 43084542.0,
      "step": 21275
    },
    {
      "entropy": 5.8561724662780765,
      "epoch": 1.8596521891112472,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046542858302700443,
      "loss": 5.1728,
      "mean_token_accuracy": 0.2164452388882637,
      "num_tokens": 43093864.0,
      "step": 21280
    },
    {
      "entropy": 5.8967102527618405,
      "epoch": 1.8600891374639517,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046541198391733286,
      "loss": 5.2906,
      "mean_token_accuracy": 0.20328271239995957,
      "num_tokens": 43104541.0,
      "step": 21285
    },
    {
      "entropy": 5.9666530132293705,
      "epoch": 1.8605260858166566,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004653953811554596,
      "loss": 5.3289,
      "mean_token_accuracy": 0.198917955160141,
      "num_tokens": 43114378.0,
      "step": 21290
    },
    {
      "entropy": 6.038884401321411,
      "epoch": 1.860963034169361,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004653787747417033,
      "loss": 5.4961,
      "mean_token_accuracy": 0.18860555589199066,
      "num_tokens": 43125503.0,
      "step": 21295
    },
    {
      "entropy": 5.955950212478638,
      "epoch": 1.861399982522066,
      "grad_norm": 1.25,
      "learning_rate": 0.00046536216467638227,
      "loss": 5.2894,
      "mean_token_accuracy": 0.2072637140750885,
      "num_tokens": 43135034.0,
      "step": 21300
    },
    {
      "entropy": 5.93865647315979,
      "epoch": 1.8618369308747706,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004653455509598152,
      "loss": 5.3707,
      "mean_token_accuracy": 0.19625181555747986,
      "num_tokens": 43145359.0,
      "step": 21305
    },
    {
      "entropy": 5.894393825531006,
      "epoch": 1.8622738792274753,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004653289335923206,
      "loss": 5.2462,
      "mean_token_accuracy": 0.2069503426551819,
      "num_tokens": 43154995.0,
      "step": 21310
    },
    {
      "entropy": 5.9252012252807615,
      "epoch": 1.86271082758018,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004653123125742175,
      "loss": 5.3406,
      "mean_token_accuracy": 0.1978338837623596,
      "num_tokens": 43164703.0,
      "step": 21315
    },
    {
      "entropy": 5.978585815429687,
      "epoch": 1.8631477759328847,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004652956879058245,
      "loss": 5.3562,
      "mean_token_accuracy": 0.20009975731372834,
      "num_tokens": 43174193.0,
      "step": 21320
    },
    {
      "entropy": 5.967437267303467,
      "epoch": 1.8635847242855894,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046527905958746043,
      "loss": 5.3934,
      "mean_token_accuracy": 0.19694218337535857,
      "num_tokens": 43184236.0,
      "step": 21325
    },
    {
      "entropy": 5.963978672027588,
      "epoch": 1.864021672638294,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046526242761944434,
      "loss": 5.3468,
      "mean_token_accuracy": 0.19457553923130036,
      "num_tokens": 43194421.0,
      "step": 21330
    },
    {
      "entropy": 5.939884757995605,
      "epoch": 1.864458620990999,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046524579200209513,
      "loss": 5.3751,
      "mean_token_accuracy": 0.20116622298955916,
      "num_tokens": 43205200.0,
      "step": 21335
    },
    {
      "entropy": 5.919119834899902,
      "epoch": 1.8648955693437035,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046522915273573194,
      "loss": 5.3102,
      "mean_token_accuracy": 0.19919493943452835,
      "num_tokens": 43215415.0,
      "step": 21340
    },
    {
      "entropy": 5.977560663223267,
      "epoch": 1.8653325176964084,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004652125098206739,
      "loss": 5.34,
      "mean_token_accuracy": 0.1976194277405739,
      "num_tokens": 43225648.0,
      "step": 21345
    },
    {
      "entropy": 5.991477394104004,
      "epoch": 1.8657694660491129,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046519586325724033,
      "loss": 5.329,
      "mean_token_accuracy": 0.19047651886940004,
      "num_tokens": 43236358.0,
      "step": 21350
    },
    {
      "entropy": 6.009267282485962,
      "epoch": 1.8662064144018178,
      "grad_norm": 1.125,
      "learning_rate": 0.00046517921304575045,
      "loss": 5.3623,
      "mean_token_accuracy": 0.19506969153881074,
      "num_tokens": 43247022.0,
      "step": 21355
    },
    {
      "entropy": 5.950340747833252,
      "epoch": 1.8666433627545223,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046516255918652356,
      "loss": 5.4389,
      "mean_token_accuracy": 0.18702463060617447,
      "num_tokens": 43258249.0,
      "step": 21360
    },
    {
      "entropy": 5.976329612731933,
      "epoch": 1.8670803111072272,
      "grad_norm": 1.125,
      "learning_rate": 0.0004651459016798791,
      "loss": 5.3849,
      "mean_token_accuracy": 0.1914568617939949,
      "num_tokens": 43268997.0,
      "step": 21365
    },
    {
      "entropy": 5.966186666488648,
      "epoch": 1.8675172594599319,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004651292405261368,
      "loss": 5.3108,
      "mean_token_accuracy": 0.20096232146024703,
      "num_tokens": 43279165.0,
      "step": 21370
    },
    {
      "entropy": 5.95704493522644,
      "epoch": 1.8679542078126365,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046511257572561584,
      "loss": 5.358,
      "mean_token_accuracy": 0.2007625848054886,
      "num_tokens": 43289778.0,
      "step": 21375
    },
    {
      "entropy": 5.916252565383911,
      "epoch": 1.8683911561653412,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004650959072786362,
      "loss": 5.2711,
      "mean_token_accuracy": 0.20084256380796434,
      "num_tokens": 43299865.0,
      "step": 21380
    },
    {
      "entropy": 5.965196514129639,
      "epoch": 1.868828104518046,
      "grad_norm": 1.125,
      "learning_rate": 0.0004650792351855173,
      "loss": 5.312,
      "mean_token_accuracy": 0.19751300364732743,
      "num_tokens": 43309224.0,
      "step": 21385
    },
    {
      "entropy": 5.969532251358032,
      "epoch": 1.8692650528707508,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004650625594465792,
      "loss": 5.2441,
      "mean_token_accuracy": 0.19817610681056977,
      "num_tokens": 43320001.0,
      "step": 21390
    },
    {
      "entropy": 5.933201694488526,
      "epoch": 1.8697020012234553,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004650458800621415,
      "loss": 5.2876,
      "mean_token_accuracy": 0.19720535725355148,
      "num_tokens": 43329790.0,
      "step": 21395
    },
    {
      "entropy": 5.999037551879883,
      "epoch": 1.8701389495761602,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004650291970325244,
      "loss": 5.4024,
      "mean_token_accuracy": 0.19209298193454744,
      "num_tokens": 43340394.0,
      "step": 21400
    },
    {
      "entropy": 5.964137315750122,
      "epoch": 1.8705758979288647,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004650125103580476,
      "loss": 5.3901,
      "mean_token_accuracy": 0.19661134630441665,
      "num_tokens": 43351404.0,
      "step": 21405
    },
    {
      "entropy": 5.827948141098022,
      "epoch": 1.8710128462815696,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004649958200390313,
      "loss": 5.1964,
      "mean_token_accuracy": 0.21443381011486054,
      "num_tokens": 43361017.0,
      "step": 21410
    },
    {
      "entropy": 6.005615139007569,
      "epoch": 1.871449794634274,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004649791260757956,
      "loss": 5.4135,
      "mean_token_accuracy": 0.18955627977848052,
      "num_tokens": 43371680.0,
      "step": 21415
    },
    {
      "entropy": 5.998482942581177,
      "epoch": 1.871886742986979,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004649624284686608,
      "loss": 5.3968,
      "mean_token_accuracy": 0.19530851840972902,
      "num_tokens": 43381908.0,
      "step": 21420
    },
    {
      "entropy": 5.990721225738525,
      "epoch": 1.8723236913396837,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000464945727217947,
      "loss": 5.4406,
      "mean_token_accuracy": 0.1863081857562065,
      "num_tokens": 43393234.0,
      "step": 21425
    },
    {
      "entropy": 6.034473085403443,
      "epoch": 1.8727606396923884,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046492902232397464,
      "loss": 5.4335,
      "mean_token_accuracy": 0.19003948867321013,
      "num_tokens": 43404065.0,
      "step": 21430
    },
    {
      "entropy": 5.960765409469604,
      "epoch": 1.873197588045093,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046491231378706415,
      "loss": 5.359,
      "mean_token_accuracy": 0.1890498086810112,
      "num_tokens": 43414417.0,
      "step": 21435
    },
    {
      "entropy": 5.974086093902588,
      "epoch": 1.8736345363977978,
      "grad_norm": 1.125,
      "learning_rate": 0.000464895601607536,
      "loss": 5.3425,
      "mean_token_accuracy": 0.1925170138478279,
      "num_tokens": 43424265.0,
      "step": 21440
    },
    {
      "entropy": 5.948543500900269,
      "epoch": 1.8740714847505024,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004648788857857106,
      "loss": 5.2646,
      "mean_token_accuracy": 0.20223428905010224,
      "num_tokens": 43434699.0,
      "step": 21445
    },
    {
      "entropy": 5.991467714309692,
      "epoch": 1.8745084331032071,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046486216632190874,
      "loss": 5.3707,
      "mean_token_accuracy": 0.20030924826860427,
      "num_tokens": 43444113.0,
      "step": 21450
    },
    {
      "entropy": 5.940961503982544,
      "epoch": 1.874945381455912,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046484544321645103,
      "loss": 5.318,
      "mean_token_accuracy": 0.1934821218252182,
      "num_tokens": 43454256.0,
      "step": 21455
    },
    {
      "entropy": 5.975553226470947,
      "epoch": 1.8753823298086165,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004648287164696583,
      "loss": 5.3162,
      "mean_token_accuracy": 0.1988134801387787,
      "num_tokens": 43463182.0,
      "step": 21460
    },
    {
      "entropy": 5.961956548690796,
      "epoch": 1.8758192781613214,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004648119860818512,
      "loss": 5.3039,
      "mean_token_accuracy": 0.1998920813202858,
      "num_tokens": 43473326.0,
      "step": 21465
    },
    {
      "entropy": 5.983966922760009,
      "epoch": 1.876256226514026,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046479525205335084,
      "loss": 5.3668,
      "mean_token_accuracy": 0.19570816606283187,
      "num_tokens": 43483210.0,
      "step": 21470
    },
    {
      "entropy": 5.997408199310303,
      "epoch": 1.8766931748667308,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004647785143844781,
      "loss": 5.3124,
      "mean_token_accuracy": 0.1975392609834671,
      "num_tokens": 43493834.0,
      "step": 21475
    },
    {
      "entropy": 5.953333473205566,
      "epoch": 1.8771301232194355,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046476177307555394,
      "loss": 5.3243,
      "mean_token_accuracy": 0.1957929641008377,
      "num_tokens": 43504522.0,
      "step": 21480
    },
    {
      "entropy": 5.98978819847107,
      "epoch": 1.8775670715721402,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046474502812689954,
      "loss": 5.4171,
      "mean_token_accuracy": 0.19325313866138458,
      "num_tokens": 43514950.0,
      "step": 21485
    },
    {
      "entropy": 5.954551887512207,
      "epoch": 1.878004019924845,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004647282795388361,
      "loss": 5.3642,
      "mean_token_accuracy": 0.19693048298358917,
      "num_tokens": 43526463.0,
      "step": 21490
    },
    {
      "entropy": 5.946456527709961,
      "epoch": 1.8784409682775496,
      "grad_norm": 1.125,
      "learning_rate": 0.0004647115273116849,
      "loss": 5.3551,
      "mean_token_accuracy": 0.19512995183467866,
      "num_tokens": 43536104.0,
      "step": 21495
    },
    {
      "entropy": 6.090953254699707,
      "epoch": 1.8788779166302543,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046469477144576716,
      "loss": 5.4637,
      "mean_token_accuracy": 0.19071454256772996,
      "num_tokens": 43547863.0,
      "step": 21500
    },
    {
      "entropy": 6.046039485931397,
      "epoch": 1.879314864982959,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004646780119414043,
      "loss": 5.3238,
      "mean_token_accuracy": 0.19950437545776367,
      "num_tokens": 43558024.0,
      "step": 21505
    },
    {
      "entropy": 5.985542249679566,
      "epoch": 1.8797518133356639,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004646612487989178,
      "loss": 5.3611,
      "mean_token_accuracy": 0.19862153977155686,
      "num_tokens": 43567464.0,
      "step": 21510
    },
    {
      "entropy": 5.9040404796600345,
      "epoch": 1.8801887616883683,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004646444820186291,
      "loss": 5.2709,
      "mean_token_accuracy": 0.20635577142238617,
      "num_tokens": 43576826.0,
      "step": 21515
    },
    {
      "entropy": 5.955849838256836,
      "epoch": 1.8806257100410733,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046462771160085995,
      "loss": 5.2899,
      "mean_token_accuracy": 0.19987305402755737,
      "num_tokens": 43586106.0,
      "step": 21520
    },
    {
      "entropy": 5.981764459609986,
      "epoch": 1.8810626583937777,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046461093754593185,
      "loss": 5.4058,
      "mean_token_accuracy": 0.1865234225988388,
      "num_tokens": 43597077.0,
      "step": 21525
    },
    {
      "entropy": 6.03364748954773,
      "epoch": 1.8814996067464826,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004645941598541667,
      "loss": 5.4176,
      "mean_token_accuracy": 0.1891277477145195,
      "num_tokens": 43608773.0,
      "step": 21530
    },
    {
      "entropy": 6.084595108032227,
      "epoch": 1.8819365550991871,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004645773785258862,
      "loss": 5.3576,
      "mean_token_accuracy": 0.19895598590373992,
      "num_tokens": 43618017.0,
      "step": 21535
    },
    {
      "entropy": 5.985270738601685,
      "epoch": 1.882373503451892,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004645605935614122,
      "loss": 5.3584,
      "mean_token_accuracy": 0.1939827397465706,
      "num_tokens": 43627865.0,
      "step": 21540
    },
    {
      "entropy": 5.986138772964478,
      "epoch": 1.8828104518045967,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046454380496106665,
      "loss": 5.3656,
      "mean_token_accuracy": 0.18932264000177385,
      "num_tokens": 43637772.0,
      "step": 21545
    },
    {
      "entropy": 5.915783548355103,
      "epoch": 1.8832474001573014,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046452701272517167,
      "loss": 5.2968,
      "mean_token_accuracy": 0.20135699808597565,
      "num_tokens": 43648242.0,
      "step": 21550
    },
    {
      "entropy": 5.9579308986663815,
      "epoch": 1.883684348510006,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046451021685404927,
      "loss": 5.4005,
      "mean_token_accuracy": 0.19437652826309204,
      "num_tokens": 43658298.0,
      "step": 21555
    },
    {
      "entropy": 6.0063951969146725,
      "epoch": 1.8841212968627108,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046449341734802155,
      "loss": 5.3427,
      "mean_token_accuracy": 0.19973421692848206,
      "num_tokens": 43667792.0,
      "step": 21560
    },
    {
      "entropy": 5.951568984985352,
      "epoch": 1.8845582452154155,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004644766142074108,
      "loss": 5.2505,
      "mean_token_accuracy": 0.21105149686336516,
      "num_tokens": 43678571.0,
      "step": 21565
    },
    {
      "entropy": 5.973385334014893,
      "epoch": 1.8849951935681202,
      "grad_norm": 1.125,
      "learning_rate": 0.00046445980743253936,
      "loss": 5.3988,
      "mean_token_accuracy": 0.19107061326503755,
      "num_tokens": 43690851.0,
      "step": 21570
    },
    {
      "entropy": 5.946140623092651,
      "epoch": 1.885432141920825,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046444299702372955,
      "loss": 5.3206,
      "mean_token_accuracy": 0.1972106710076332,
      "num_tokens": 43700960.0,
      "step": 21575
    },
    {
      "entropy": 6.0195588111877445,
      "epoch": 1.8858690902735296,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004644261829813038,
      "loss": 5.308,
      "mean_token_accuracy": 0.20092339664697648,
      "num_tokens": 43710466.0,
      "step": 21580
    },
    {
      "entropy": 6.006682443618774,
      "epoch": 1.8863060386262345,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046440936530558455,
      "loss": 5.3735,
      "mean_token_accuracy": 0.19982051104307175,
      "num_tokens": 43720261.0,
      "step": 21585
    },
    {
      "entropy": 5.978329515457153,
      "epoch": 1.886742986978939,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004643925439968944,
      "loss": 5.3706,
      "mean_token_accuracy": 0.1959306165575981,
      "num_tokens": 43731045.0,
      "step": 21590
    },
    {
      "entropy": 6.036318683624268,
      "epoch": 1.8871799353316439,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046437571905555604,
      "loss": 5.3742,
      "mean_token_accuracy": 0.1944679483771324,
      "num_tokens": 43741873.0,
      "step": 21595
    },
    {
      "entropy": 5.974472332000732,
      "epoch": 1.8876168836843485,
      "grad_norm": 1.25,
      "learning_rate": 0.00046435889048189216,
      "loss": 5.3794,
      "mean_token_accuracy": 0.20000085830688477,
      "num_tokens": 43751473.0,
      "step": 21600
    },
    {
      "entropy": 5.973597240447998,
      "epoch": 1.8880538320370532,
      "grad_norm": 1.0,
      "learning_rate": 0.00046434205827622555,
      "loss": 5.3758,
      "mean_token_accuracy": 0.19781383574008943,
      "num_tokens": 43762517.0,
      "step": 21605
    },
    {
      "entropy": 6.039477682113647,
      "epoch": 1.888490780389758,
      "grad_norm": 1.125,
      "learning_rate": 0.0004643252224388791,
      "loss": 5.3899,
      "mean_token_accuracy": 0.1935605525970459,
      "num_tokens": 43772491.0,
      "step": 21610
    },
    {
      "entropy": 5.969558811187744,
      "epoch": 1.8889277287424626,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046430838297017556,
      "loss": 5.373,
      "mean_token_accuracy": 0.19177722036838532,
      "num_tokens": 43781646.0,
      "step": 21615
    },
    {
      "entropy": 5.944332408905029,
      "epoch": 1.8893646770951673,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046429153987043815,
      "loss": 5.2635,
      "mean_token_accuracy": 0.20484237223863602,
      "num_tokens": 43792020.0,
      "step": 21620
    },
    {
      "entropy": 5.9345863342285154,
      "epoch": 1.889801625447872,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004642746931399898,
      "loss": 5.2828,
      "mean_token_accuracy": 0.20632764101028442,
      "num_tokens": 43801638.0,
      "step": 21625
    },
    {
      "entropy": 5.9991905212402346,
      "epoch": 1.890238573800577,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046425784277915363,
      "loss": 5.4313,
      "mean_token_accuracy": 0.18519504070281984,
      "num_tokens": 43811475.0,
      "step": 21630
    },
    {
      "entropy": 6.002260541915893,
      "epoch": 1.8906755221532814,
      "grad_norm": 1.125,
      "learning_rate": 0.0004642409887882528,
      "loss": 5.3309,
      "mean_token_accuracy": 0.19440586864948273,
      "num_tokens": 43820811.0,
      "step": 21635
    },
    {
      "entropy": 6.044103956222534,
      "epoch": 1.8911124705059863,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004642241311676107,
      "loss": 5.3908,
      "mean_token_accuracy": 0.18596543222665787,
      "num_tokens": 43830804.0,
      "step": 21640
    },
    {
      "entropy": 5.869978904724121,
      "epoch": 1.8915494188586908,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004642072699175506,
      "loss": 5.214,
      "mean_token_accuracy": 0.21336660236120225,
      "num_tokens": 43840156.0,
      "step": 21645
    },
    {
      "entropy": 5.956230401992798,
      "epoch": 1.8919863672113957,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000464190405038396,
      "loss": 5.316,
      "mean_token_accuracy": 0.193955659866333,
      "num_tokens": 43849622.0,
      "step": 21650
    },
    {
      "entropy": 6.041838073730469,
      "epoch": 1.8924233155641004,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004641735365304703,
      "loss": 5.3959,
      "mean_token_accuracy": 0.18920937776565552,
      "num_tokens": 43860184.0,
      "step": 21655
    },
    {
      "entropy": 5.863009405136109,
      "epoch": 1.892860263916805,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046415666439409694,
      "loss": 5.2279,
      "mean_token_accuracy": 0.2157977506518364,
      "num_tokens": 43870157.0,
      "step": 21660
    },
    {
      "entropy": 5.94944314956665,
      "epoch": 1.8932972122695098,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046413978862959964,
      "loss": 5.3093,
      "mean_token_accuracy": 0.19482292085886002,
      "num_tokens": 43879867.0,
      "step": 21665
    },
    {
      "entropy": 5.917213106155396,
      "epoch": 1.8937341606222144,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004641229092373022,
      "loss": 5.3246,
      "mean_token_accuracy": 0.20086213648319245,
      "num_tokens": 43890314.0,
      "step": 21670
    },
    {
      "entropy": 6.05149359703064,
      "epoch": 1.8941711089749191,
      "grad_norm": 1.125,
      "learning_rate": 0.0004641060262175282,
      "loss": 5.3698,
      "mean_token_accuracy": 0.1908307820558548,
      "num_tokens": 43901075.0,
      "step": 21675
    },
    {
      "entropy": 6.005564165115357,
      "epoch": 1.8946080573276238,
      "grad_norm": 1.125,
      "learning_rate": 0.00046408913957060144,
      "loss": 5.3863,
      "mean_token_accuracy": 0.1920474052429199,
      "num_tokens": 43911777.0,
      "step": 21680
    },
    {
      "entropy": 5.9469747066497805,
      "epoch": 1.8950450056803287,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046407224929684597,
      "loss": 5.2957,
      "mean_token_accuracy": 0.19914434999227523,
      "num_tokens": 43921901.0,
      "step": 21685
    },
    {
      "entropy": 5.987225532531738,
      "epoch": 1.8954819540330332,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004640553553965857,
      "loss": 5.3839,
      "mean_token_accuracy": 0.19349030703306197,
      "num_tokens": 43931142.0,
      "step": 21690
    },
    {
      "entropy": 6.014645671844482,
      "epoch": 1.8959189023857381,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004640384578701447,
      "loss": 5.3079,
      "mean_token_accuracy": 0.19765571653842925,
      "num_tokens": 43940642.0,
      "step": 21695
    },
    {
      "entropy": 6.0021960735321045,
      "epoch": 1.8963558507384426,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046402155671784696,
      "loss": 5.3923,
      "mean_token_accuracy": 0.18912360221147537,
      "num_tokens": 43950808.0,
      "step": 21700
    },
    {
      "entropy": 6.063928937911987,
      "epoch": 1.8967927990911475,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046400465194001674,
      "loss": 5.4427,
      "mean_token_accuracy": 0.18787715584039688,
      "num_tokens": 43961349.0,
      "step": 21705
    },
    {
      "entropy": 6.022521066665649,
      "epoch": 1.897229747443852,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046398774353697826,
      "loss": 5.3127,
      "mean_token_accuracy": 0.19873567521572114,
      "num_tokens": 43970072.0,
      "step": 21710
    },
    {
      "entropy": 5.970760631561279,
      "epoch": 1.897666695796557,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004639708315090558,
      "loss": 5.3871,
      "mean_token_accuracy": 0.19710342288017274,
      "num_tokens": 43980076.0,
      "step": 21715
    },
    {
      "entropy": 5.912375974655151,
      "epoch": 1.8981036441492616,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004639539158565738,
      "loss": 5.1938,
      "mean_token_accuracy": 0.21539719998836518,
      "num_tokens": 43989876.0,
      "step": 21720
    },
    {
      "entropy": 6.038564395904541,
      "epoch": 1.8985405925019663,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004639369965798567,
      "loss": 5.385,
      "mean_token_accuracy": 0.18598264157772065,
      "num_tokens": 44000920.0,
      "step": 21725
    },
    {
      "entropy": 5.918316411972046,
      "epoch": 1.898977540854671,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046392007367922894,
      "loss": 5.2177,
      "mean_token_accuracy": 0.20443010926246644,
      "num_tokens": 44011539.0,
      "step": 21730
    },
    {
      "entropy": 5.965426874160767,
      "epoch": 1.8994144892073757,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004639031471550152,
      "loss": 5.345,
      "mean_token_accuracy": 0.19669246673583984,
      "num_tokens": 44022624.0,
      "step": 21735
    },
    {
      "entropy": 5.9511191844940186,
      "epoch": 1.8998514375600803,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046388621700754014,
      "loss": 5.3024,
      "mean_token_accuracy": 0.20036210268735885,
      "num_tokens": 44033105.0,
      "step": 21740
    },
    {
      "entropy": 5.962542867660522,
      "epoch": 1.900288385912785,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046386928323712847,
      "loss": 5.298,
      "mean_token_accuracy": 0.19554940462112427,
      "num_tokens": 44042805.0,
      "step": 21745
    },
    {
      "entropy": 5.944049644470215,
      "epoch": 1.90072533426549,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000463852345844105,
      "loss": 5.2824,
      "mean_token_accuracy": 0.2053304985165596,
      "num_tokens": 44052471.0,
      "step": 21750
    },
    {
      "entropy": 5.944161653518677,
      "epoch": 1.9011622826181944,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004638354048287945,
      "loss": 5.3913,
      "mean_token_accuracy": 0.1895996257662773,
      "num_tokens": 44063358.0,
      "step": 21755
    },
    {
      "entropy": 5.964716148376465,
      "epoch": 1.9015992309708993,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046381846019152194,
      "loss": 5.3569,
      "mean_token_accuracy": 0.19630459398031236,
      "num_tokens": 44073598.0,
      "step": 21760
    },
    {
      "entropy": 5.9959033012390135,
      "epoch": 1.9020361793236038,
      "grad_norm": 1.125,
      "learning_rate": 0.00046380151193261247,
      "loss": 5.3452,
      "mean_token_accuracy": 0.200860595703125,
      "num_tokens": 44084752.0,
      "step": 21765
    },
    {
      "entropy": 5.916839599609375,
      "epoch": 1.9024731276763087,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004637845600523911,
      "loss": 5.3085,
      "mean_token_accuracy": 0.20161595940589905,
      "num_tokens": 44094857.0,
      "step": 21770
    },
    {
      "entropy": 6.009101867675781,
      "epoch": 1.9029100760290134,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004637676045511829,
      "loss": 5.3814,
      "mean_token_accuracy": 0.19449977427721024,
      "num_tokens": 44105700.0,
      "step": 21775
    },
    {
      "entropy": 5.934243965148926,
      "epoch": 1.903347024381718,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004637506454293131,
      "loss": 5.3209,
      "mean_token_accuracy": 0.19799017757177353,
      "num_tokens": 44116588.0,
      "step": 21780
    },
    {
      "entropy": 5.971315956115722,
      "epoch": 1.9037839727344228,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004637336826871071,
      "loss": 5.3808,
      "mean_token_accuracy": 0.1952686369419098,
      "num_tokens": 44126948.0,
      "step": 21785
    },
    {
      "entropy": 6.037430620193481,
      "epoch": 1.9042209210871275,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046371671632489017,
      "loss": 5.4324,
      "mean_token_accuracy": 0.18884895890951156,
      "num_tokens": 44136239.0,
      "step": 21790
    },
    {
      "entropy": 5.95443320274353,
      "epoch": 1.9046578694398322,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046369974634298766,
      "loss": 5.2392,
      "mean_token_accuracy": 0.20692387521266936,
      "num_tokens": 44146314.0,
      "step": 21795
    },
    {
      "entropy": 5.969230031967163,
      "epoch": 1.9050948177925369,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004636827727417252,
      "loss": 5.3595,
      "mean_token_accuracy": 0.20070180892944336,
      "num_tokens": 44155767.0,
      "step": 21800
    },
    {
      "entropy": 5.970375633239746,
      "epoch": 1.9055317661452418,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046366579552142823,
      "loss": 5.3706,
      "mean_token_accuracy": 0.1992092490196228,
      "num_tokens": 44165830.0,
      "step": 21805
    },
    {
      "entropy": 5.948747539520264,
      "epoch": 1.9059687144979462,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004636488146824225,
      "loss": 5.2553,
      "mean_token_accuracy": 0.20749661028385163,
      "num_tokens": 44175526.0,
      "step": 21810
    },
    {
      "entropy": 6.0286256790161135,
      "epoch": 1.9064056628506512,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004636318302250336,
      "loss": 5.3808,
      "mean_token_accuracy": 0.19891500771045684,
      "num_tokens": 44185950.0,
      "step": 21815
    },
    {
      "entropy": 5.979408121109008,
      "epoch": 1.9068426112033556,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004636148421495874,
      "loss": 5.3106,
      "mean_token_accuracy": 0.19879081845283508,
      "num_tokens": 44196518.0,
      "step": 21820
    },
    {
      "entropy": 5.9688112258911135,
      "epoch": 1.9072795595560605,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046359785045640967,
      "loss": 5.4123,
      "mean_token_accuracy": 0.19879419058561326,
      "num_tokens": 44207709.0,
      "step": 21825
    },
    {
      "entropy": 6.0383905410766605,
      "epoch": 1.9077165079087652,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004635808551458264,
      "loss": 5.3892,
      "mean_token_accuracy": 0.1853034257888794,
      "num_tokens": 44218311.0,
      "step": 21830
    },
    {
      "entropy": 6.016307783126831,
      "epoch": 1.90815345626147,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046356385621816343,
      "loss": 5.3365,
      "mean_token_accuracy": 0.19130146205425264,
      "num_tokens": 44229059.0,
      "step": 21835
    },
    {
      "entropy": 5.956440877914429,
      "epoch": 1.9085904046141746,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004635468536737469,
      "loss": 5.3371,
      "mean_token_accuracy": 0.1939671203494072,
      "num_tokens": 44239254.0,
      "step": 21840
    },
    {
      "entropy": 6.002814388275146,
      "epoch": 1.9090273529668793,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046352984751290297,
      "loss": 5.2295,
      "mean_token_accuracy": 0.20384658724069596,
      "num_tokens": 44248465.0,
      "step": 21845
    },
    {
      "entropy": 5.982296657562256,
      "epoch": 1.909464301319584,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046351283773595777,
      "loss": 5.3357,
      "mean_token_accuracy": 0.19522083103656768,
      "num_tokens": 44259121.0,
      "step": 21850
    },
    {
      "entropy": 6.005945110321045,
      "epoch": 1.9099012496722887,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046349582434323757,
      "loss": 5.3276,
      "mean_token_accuracy": 0.20128542482852935,
      "num_tokens": 44269291.0,
      "step": 21855
    },
    {
      "entropy": 5.878613376617432,
      "epoch": 1.9103381980249936,
      "grad_norm": 1.125,
      "learning_rate": 0.00046347880733506867,
      "loss": 5.3215,
      "mean_token_accuracy": 0.1954540103673935,
      "num_tokens": 44280985.0,
      "step": 21860
    },
    {
      "entropy": 5.927117824554443,
      "epoch": 1.910775146377698,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046346178671177746,
      "loss": 5.3082,
      "mean_token_accuracy": 0.20088924765586852,
      "num_tokens": 44290370.0,
      "step": 21865
    },
    {
      "entropy": 5.942974948883057,
      "epoch": 1.911212094730403,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046344476247369044,
      "loss": 5.3217,
      "mean_token_accuracy": 0.19429877251386643,
      "num_tokens": 44300670.0,
      "step": 21870
    },
    {
      "entropy": 5.992985105514526,
      "epoch": 1.9116490430831075,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046342773462113414,
      "loss": 5.3564,
      "mean_token_accuracy": 0.19414859265089035,
      "num_tokens": 44311578.0,
      "step": 21875
    },
    {
      "entropy": 5.929243278503418,
      "epoch": 1.9120859914358124,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046341070315443516,
      "loss": 5.2712,
      "mean_token_accuracy": 0.20441189259290696,
      "num_tokens": 44321692.0,
      "step": 21880
    },
    {
      "entropy": 5.903761005401611,
      "epoch": 1.9125229397885168,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046339366807392017,
      "loss": 5.3118,
      "mean_token_accuracy": 0.20191961228847505,
      "num_tokens": 44331468.0,
      "step": 21885
    },
    {
      "entropy": 5.968597936630249,
      "epoch": 1.9129598881412218,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046337662937991585,
      "loss": 5.3404,
      "mean_token_accuracy": 0.19412423819303512,
      "num_tokens": 44341491.0,
      "step": 21890
    },
    {
      "entropy": 5.997011995315551,
      "epoch": 1.9133968364939264,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004633595870727492,
      "loss": 5.4168,
      "mean_token_accuracy": 0.18706524968147278,
      "num_tokens": 44351568.0,
      "step": 21895
    },
    {
      "entropy": 6.075109338760376,
      "epoch": 1.9138337848466311,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046334254115274685,
      "loss": 5.3961,
      "mean_token_accuracy": 0.19470628947019578,
      "num_tokens": 44362009.0,
      "step": 21900
    },
    {
      "entropy": 5.962094306945801,
      "epoch": 1.9142707331993358,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004633254916202359,
      "loss": 5.2905,
      "mean_token_accuracy": 0.20516868084669113,
      "num_tokens": 44372671.0,
      "step": 21905
    },
    {
      "entropy": 5.892187213897705,
      "epoch": 1.9147076815520405,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004633084384755434,
      "loss": 5.2912,
      "mean_token_accuracy": 0.20124102532863616,
      "num_tokens": 44382658.0,
      "step": 21910
    },
    {
      "entropy": 5.980854272842407,
      "epoch": 1.9151446299047452,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004632913817189963,
      "loss": 5.3087,
      "mean_token_accuracy": 0.19909146726131438,
      "num_tokens": 44392090.0,
      "step": 21915
    },
    {
      "entropy": 5.998489189147949,
      "epoch": 1.91558157825745,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004632743213509219,
      "loss": 5.393,
      "mean_token_accuracy": 0.18818808645009993,
      "num_tokens": 44402896.0,
      "step": 21920
    },
    {
      "entropy": 5.995146036148071,
      "epoch": 1.9160185266101548,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046325725737164734,
      "loss": 5.3445,
      "mean_token_accuracy": 0.19438591748476028,
      "num_tokens": 44413700.0,
      "step": 21925
    },
    {
      "entropy": 6.003905344009399,
      "epoch": 1.9164554749628593,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004632401897814999,
      "loss": 5.2512,
      "mean_token_accuracy": 0.20949110835790635,
      "num_tokens": 44423442.0,
      "step": 21930
    },
    {
      "entropy": 5.91346173286438,
      "epoch": 1.9168924233155642,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004632231185808071,
      "loss": 5.3054,
      "mean_token_accuracy": 0.1997228369116783,
      "num_tokens": 44434197.0,
      "step": 21935
    },
    {
      "entropy": 5.9272966384887695,
      "epoch": 1.9173293716682687,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046320604376989604,
      "loss": 5.3114,
      "mean_token_accuracy": 0.2026067465543747,
      "num_tokens": 44444483.0,
      "step": 21940
    },
    {
      "entropy": 5.941142606735229,
      "epoch": 1.9177663200209736,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004631889653490946,
      "loss": 5.3263,
      "mean_token_accuracy": 0.20008074641227722,
      "num_tokens": 44456303.0,
      "step": 21945
    },
    {
      "entropy": 5.952383518218994,
      "epoch": 1.9182032683736783,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046317188331873024,
      "loss": 5.3663,
      "mean_token_accuracy": 0.20039328187704086,
      "num_tokens": 44467314.0,
      "step": 21950
    },
    {
      "entropy": 5.987460613250732,
      "epoch": 1.918640216726383,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004631547976791305,
      "loss": 5.3433,
      "mean_token_accuracy": 0.19674230068922044,
      "num_tokens": 44476466.0,
      "step": 21955
    },
    {
      "entropy": 5.978013706207276,
      "epoch": 1.9190771650790877,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004631377084306231,
      "loss": 5.3507,
      "mean_token_accuracy": 0.19546225070953369,
      "num_tokens": 44486247.0,
      "step": 21960
    },
    {
      "entropy": 5.971577596664429,
      "epoch": 1.9195141134317923,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046312061557353585,
      "loss": 5.3247,
      "mean_token_accuracy": 0.19959349781274796,
      "num_tokens": 44497166.0,
      "step": 21965
    },
    {
      "entropy": 6.027429389953613,
      "epoch": 1.919951061784497,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004631035191081967,
      "loss": 5.3568,
      "mean_token_accuracy": 0.19265873432159425,
      "num_tokens": 44507677.0,
      "step": 21970
    },
    {
      "entropy": 5.957675218582153,
      "epoch": 1.9203880101372017,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004630864190349334,
      "loss": 5.2949,
      "mean_token_accuracy": 0.2083557978272438,
      "num_tokens": 44517392.0,
      "step": 21975
    },
    {
      "entropy": 5.986942148208618,
      "epoch": 1.9208249584899066,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046306931535407416,
      "loss": 5.3324,
      "mean_token_accuracy": 0.19862686097621918,
      "num_tokens": 44528914.0,
      "step": 21980
    },
    {
      "entropy": 6.024699306488037,
      "epoch": 1.9212619068426111,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046305220806594683,
      "loss": 5.4682,
      "mean_token_accuracy": 0.1860966607928276,
      "num_tokens": 44539419.0,
      "step": 21985
    },
    {
      "entropy": 5.9925620555877686,
      "epoch": 1.921698855195316,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004630350971708797,
      "loss": 5.2977,
      "mean_token_accuracy": 0.20379093587398528,
      "num_tokens": 44548704.0,
      "step": 21990
    },
    {
      "entropy": 5.941236352920532,
      "epoch": 1.9221358035480205,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046301798266920076,
      "loss": 5.3076,
      "mean_token_accuracy": 0.20235534459352494,
      "num_tokens": 44559110.0,
      "step": 21995
    },
    {
      "entropy": 6.0077269077301025,
      "epoch": 1.9225727519007254,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004630008645612384,
      "loss": 5.3778,
      "mean_token_accuracy": 0.19249722212553025,
      "num_tokens": 44570414.0,
      "step": 22000
    },
    {
      "entropy": 6.050540208816528,
      "epoch": 1.92300970025343,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046298374284732096,
      "loss": 5.4803,
      "mean_token_accuracy": 0.18668873161077498,
      "num_tokens": 44580194.0,
      "step": 22005
    },
    {
      "entropy": 5.959224081039428,
      "epoch": 1.9234466486061348,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004629666175277768,
      "loss": 5.3249,
      "mean_token_accuracy": 0.20353677719831467,
      "num_tokens": 44590174.0,
      "step": 22010
    },
    {
      "entropy": 6.005481004714966,
      "epoch": 1.9238835969588395,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004629494886029345,
      "loss": 5.4136,
      "mean_token_accuracy": 0.19222104251384736,
      "num_tokens": 44600940.0,
      "step": 22015
    },
    {
      "entropy": 6.008962249755859,
      "epoch": 1.9243205453115442,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004629323560731225,
      "loss": 5.3936,
      "mean_token_accuracy": 0.19404432475566863,
      "num_tokens": 44610852.0,
      "step": 22020
    },
    {
      "entropy": 6.013247442245484,
      "epoch": 1.9247574936642489,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046291521993866943,
      "loss": 5.4183,
      "mean_token_accuracy": 0.19684708267450332,
      "num_tokens": 44621479.0,
      "step": 22025
    },
    {
      "entropy": 5.99975438117981,
      "epoch": 1.9251944420169536,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000462898080199904,
      "loss": 5.356,
      "mean_token_accuracy": 0.1953789547085762,
      "num_tokens": 44631838.0,
      "step": 22030
    },
    {
      "entropy": 5.960122013092041,
      "epoch": 1.9256313903696585,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004628809368571549,
      "loss": 5.3677,
      "mean_token_accuracy": 0.19015876650810243,
      "num_tokens": 44643103.0,
      "step": 22035
    },
    {
      "entropy": 6.058140611648559,
      "epoch": 1.926068338722363,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046286378991075095,
      "loss": 5.4892,
      "mean_token_accuracy": 0.18383235931396485,
      "num_tokens": 44653659.0,
      "step": 22040
    },
    {
      "entropy": 6.021159362792969,
      "epoch": 1.9265052870750679,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004628466393610211,
      "loss": 5.3422,
      "mean_token_accuracy": 0.19734666496515274,
      "num_tokens": 44665064.0,
      "step": 22045
    },
    {
      "entropy": 5.9649523258209225,
      "epoch": 1.9269422354277723,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046282948520829433,
      "loss": 5.3058,
      "mean_token_accuracy": 0.20294871628284455,
      "num_tokens": 44674256.0,
      "step": 22050
    },
    {
      "entropy": 5.935221099853516,
      "epoch": 1.9273791837804772,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004628123274528996,
      "loss": 5.2864,
      "mean_token_accuracy": 0.2001873090863228,
      "num_tokens": 44683715.0,
      "step": 22055
    },
    {
      "entropy": 6.0111610889434814,
      "epoch": 1.9278161321331817,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046279516609516594,
      "loss": 5.477,
      "mean_token_accuracy": 0.18932333439588547,
      "num_tokens": 44694637.0,
      "step": 22060
    },
    {
      "entropy": 6.011595916748047,
      "epoch": 1.9282530804858866,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004627780011354226,
      "loss": 5.3266,
      "mean_token_accuracy": 0.20312012881040573,
      "num_tokens": 44705307.0,
      "step": 22065
    },
    {
      "entropy": 5.99512414932251,
      "epoch": 1.9286900288385913,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004627608325739989,
      "loss": 5.378,
      "mean_token_accuracy": 0.19776866137981414,
      "num_tokens": 44715293.0,
      "step": 22070
    },
    {
      "entropy": 6.033068132400513,
      "epoch": 1.929126977191296,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000462743660411224,
      "loss": 5.443,
      "mean_token_accuracy": 0.1864391893148422,
      "num_tokens": 44726406.0,
      "step": 22075
    },
    {
      "entropy": 5.9982085704803465,
      "epoch": 1.9295639255440007,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046272648464742733,
      "loss": 5.3752,
      "mean_token_accuracy": 0.19542793482542037,
      "num_tokens": 44737166.0,
      "step": 22080
    },
    {
      "entropy": 5.9512584686279295,
      "epoch": 1.9300008738967054,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046270930528293826,
      "loss": 5.2659,
      "mean_token_accuracy": 0.20087360143661498,
      "num_tokens": 44746693.0,
      "step": 22085
    },
    {
      "entropy": 5.952630472183228,
      "epoch": 1.93043782224941,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004626921223180864,
      "loss": 5.2962,
      "mean_token_accuracy": 0.2017673060297966,
      "num_tokens": 44757499.0,
      "step": 22090
    },
    {
      "entropy": 5.913655710220337,
      "epoch": 1.9308747706021148,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046267493575320127,
      "loss": 5.3472,
      "mean_token_accuracy": 0.19926172941923143,
      "num_tokens": 44768112.0,
      "step": 22095
    },
    {
      "entropy": 5.945527124404907,
      "epoch": 1.9313117189548197,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004626577455886125,
      "loss": 5.2863,
      "mean_token_accuracy": 0.19811212867498398,
      "num_tokens": 44778317.0,
      "step": 22100
    },
    {
      "entropy": 6.01431975364685,
      "epoch": 1.9317486673075241,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004626405518246499,
      "loss": 5.4285,
      "mean_token_accuracy": 0.19000308513641356,
      "num_tokens": 44790483.0,
      "step": 22105
    },
    {
      "entropy": 5.971853399276734,
      "epoch": 1.932185615660229,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046262335446164315,
      "loss": 5.2842,
      "mean_token_accuracy": 0.20164819657802582,
      "num_tokens": 44799891.0,
      "step": 22110
    },
    {
      "entropy": 5.984445571899414,
      "epoch": 1.9326225640129335,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004626061534999222,
      "loss": 5.3879,
      "mean_token_accuracy": 0.19286690801382064,
      "num_tokens": 44810728.0,
      "step": 22115
    },
    {
      "entropy": 6.012849903106689,
      "epoch": 1.9330595123656384,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046258894893981686,
      "loss": 5.3984,
      "mean_token_accuracy": 0.20070509165525435,
      "num_tokens": 44820597.0,
      "step": 22120
    },
    {
      "entropy": 5.955709505081177,
      "epoch": 1.9334964607183431,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004625717407816572,
      "loss": 5.3514,
      "mean_token_accuracy": 0.19705803096294403,
      "num_tokens": 44832570.0,
      "step": 22125
    },
    {
      "entropy": 5.994314765930175,
      "epoch": 1.9339334090710478,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004625545290257733,
      "loss": 5.3982,
      "mean_token_accuracy": 0.19569030553102493,
      "num_tokens": 44841936.0,
      "step": 22130
    },
    {
      "entropy": 5.9172608852386475,
      "epoch": 1.9343703574237525,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046253731367249527,
      "loss": 5.3382,
      "mean_token_accuracy": 0.20409472435712814,
      "num_tokens": 44852329.0,
      "step": 22135
    },
    {
      "entropy": 6.002935552597046,
      "epoch": 1.9348073057764572,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004625200947221532,
      "loss": 5.3567,
      "mean_token_accuracy": 0.19056169390678407,
      "num_tokens": 44862183.0,
      "step": 22140
    },
    {
      "entropy": 5.980534315109253,
      "epoch": 1.935244254129162,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004625028721750776,
      "loss": 5.3715,
      "mean_token_accuracy": 0.19294256269931792,
      "num_tokens": 44872712.0,
      "step": 22145
    },
    {
      "entropy": 5.962975978851318,
      "epoch": 1.9356812024818666,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046248564603159855,
      "loss": 5.3683,
      "mean_token_accuracy": 0.1976391166448593,
      "num_tokens": 44884423.0,
      "step": 22150
    },
    {
      "entropy": 6.023768091201783,
      "epoch": 1.9361181508345715,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004624684162920466,
      "loss": 5.3533,
      "mean_token_accuracy": 0.18964893221855164,
      "num_tokens": 44894556.0,
      "step": 22155
    },
    {
      "entropy": 5.951013469696045,
      "epoch": 1.936555099187276,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046245118295675213,
      "loss": 5.2807,
      "mean_token_accuracy": 0.19842213094234468,
      "num_tokens": 44905208.0,
      "step": 22160
    },
    {
      "entropy": 5.94128565788269,
      "epoch": 1.9369920475399809,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046243394602604585,
      "loss": 5.296,
      "mean_token_accuracy": 0.20040609538555146,
      "num_tokens": 44915674.0,
      "step": 22165
    },
    {
      "entropy": 5.962712717056275,
      "epoch": 1.9374289958926854,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004624167055002582,
      "loss": 5.3485,
      "mean_token_accuracy": 0.19720822870731353,
      "num_tokens": 44926382.0,
      "step": 22170
    },
    {
      "entropy": 5.986813306808472,
      "epoch": 1.9378659442453903,
      "grad_norm": 1.125,
      "learning_rate": 0.00046239946137971997,
      "loss": 5.2854,
      "mean_token_accuracy": 0.20724829584360122,
      "num_tokens": 44935963.0,
      "step": 22175
    },
    {
      "entropy": 6.001313304901123,
      "epoch": 1.9383028925980947,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046238221366476183,
      "loss": 5.3956,
      "mean_token_accuracy": 0.19136762171983718,
      "num_tokens": 44946312.0,
      "step": 22180
    },
    {
      "entropy": 5.90595817565918,
      "epoch": 1.9387398409507997,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046236496235571475,
      "loss": 5.3122,
      "mean_token_accuracy": 0.19563453644514084,
      "num_tokens": 44956213.0,
      "step": 22185
    },
    {
      "entropy": 6.011289262771607,
      "epoch": 1.9391767893035043,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004623477074529094,
      "loss": 5.4671,
      "mean_token_accuracy": 0.18076505959033967,
      "num_tokens": 44966575.0,
      "step": 22190
    },
    {
      "entropy": 6.031469058990479,
      "epoch": 1.939613737656209,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004623304489566769,
      "loss": 5.3773,
      "mean_token_accuracy": 0.19320639818906785,
      "num_tokens": 44976353.0,
      "step": 22195
    },
    {
      "entropy": 5.954381465911865,
      "epoch": 1.9400506860089137,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046231318686734826,
      "loss": 5.3073,
      "mean_token_accuracy": 0.1928958475589752,
      "num_tokens": 44987046.0,
      "step": 22200
    },
    {
      "entropy": 6.010937166213989,
      "epoch": 1.9404876343616184,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004622959211852545,
      "loss": 5.2865,
      "mean_token_accuracy": 0.19815263748168946,
      "num_tokens": 44996606.0,
      "step": 22205
    },
    {
      "entropy": 6.0451610565185545,
      "epoch": 1.9409245827143233,
      "grad_norm": 1.125,
      "learning_rate": 0.00046227865191072684,
      "loss": 5.407,
      "mean_token_accuracy": 0.19318957775831222,
      "num_tokens": 45006895.0,
      "step": 22210
    },
    {
      "entropy": 5.9608980178833,
      "epoch": 1.9413615310670278,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004622613790440965,
      "loss": 5.2832,
      "mean_token_accuracy": 0.19868330359458924,
      "num_tokens": 45017075.0,
      "step": 22215
    },
    {
      "entropy": 5.994738817214966,
      "epoch": 1.9417984794197327,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004622441025856948,
      "loss": 5.3247,
      "mean_token_accuracy": 0.195194211602211,
      "num_tokens": 45027331.0,
      "step": 22220
    },
    {
      "entropy": 5.976983594894409,
      "epoch": 1.9422354277724372,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004622268225358531,
      "loss": 5.3246,
      "mean_token_accuracy": 0.19028964787721633,
      "num_tokens": 45037623.0,
      "step": 22225
    },
    {
      "entropy": 6.001293420791626,
      "epoch": 1.942672376125142,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046220953889490287,
      "loss": 5.3967,
      "mean_token_accuracy": 0.18830335289239883,
      "num_tokens": 45048191.0,
      "step": 22230
    },
    {
      "entropy": 5.996423578262329,
      "epoch": 1.9431093244778466,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046219225166317554,
      "loss": 5.3813,
      "mean_token_accuracy": 0.19468885958194732,
      "num_tokens": 45057906.0,
      "step": 22235
    },
    {
      "entropy": 5.979166460037232,
      "epoch": 1.9435462728305515,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046217496084100267,
      "loss": 5.3764,
      "mean_token_accuracy": 0.20665381997823715,
      "num_tokens": 45067842.0,
      "step": 22240
    },
    {
      "entropy": 5.987281608581543,
      "epoch": 1.9439832211832562,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046215766642871607,
      "loss": 5.3075,
      "mean_token_accuracy": 0.19508747458457948,
      "num_tokens": 45077388.0,
      "step": 22245
    },
    {
      "entropy": 5.888488292694092,
      "epoch": 1.9444201695359609,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046214036842664736,
      "loss": 5.2476,
      "mean_token_accuracy": 0.20221011936664582,
      "num_tokens": 45086001.0,
      "step": 22250
    },
    {
      "entropy": 5.951739931106568,
      "epoch": 1.9448571178886656,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004621230668351283,
      "loss": 5.3013,
      "mean_token_accuracy": 0.20360518842935563,
      "num_tokens": 45096203.0,
      "step": 22255
    },
    {
      "entropy": 5.959019184112549,
      "epoch": 1.9452940662413702,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046210576165449076,
      "loss": 5.2793,
      "mean_token_accuracy": 0.20589797496795653,
      "num_tokens": 45104997.0,
      "step": 22260
    },
    {
      "entropy": 5.987274026870727,
      "epoch": 1.945731014594075,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046208845288506667,
      "loss": 5.3377,
      "mean_token_accuracy": 0.20092995911836625,
      "num_tokens": 45114150.0,
      "step": 22265
    },
    {
      "entropy": 5.920574522018432,
      "epoch": 1.9461679629467796,
      "grad_norm": 1.109375,
      "learning_rate": 0.000462071140527188,
      "loss": 5.3515,
      "mean_token_accuracy": 0.20754557996988296,
      "num_tokens": 45124668.0,
      "step": 22270
    },
    {
      "entropy": 5.967814111709595,
      "epoch": 1.9466049112994845,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004620538245811869,
      "loss": 5.3433,
      "mean_token_accuracy": 0.19587946087121963,
      "num_tokens": 45135164.0,
      "step": 22275
    },
    {
      "entropy": 5.949343967437744,
      "epoch": 1.947041859652189,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046203650504739543,
      "loss": 5.2029,
      "mean_token_accuracy": 0.2107423350214958,
      "num_tokens": 45144453.0,
      "step": 22280
    },
    {
      "entropy": 5.984271240234375,
      "epoch": 1.947478808004894,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046201918192614573,
      "loss": 5.3759,
      "mean_token_accuracy": 0.19176603406667708,
      "num_tokens": 45154557.0,
      "step": 22285
    },
    {
      "entropy": 5.972058820724487,
      "epoch": 1.9479157563575984,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004620018552177701,
      "loss": 5.3347,
      "mean_token_accuracy": 0.20043175518512726,
      "num_tokens": 45164938.0,
      "step": 22290
    },
    {
      "entropy": 6.075237655639649,
      "epoch": 1.9483527047103033,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046198452492260096,
      "loss": 5.4638,
      "mean_token_accuracy": 0.17996990829706191,
      "num_tokens": 45175823.0,
      "step": 22295
    },
    {
      "entropy": 5.911475038528442,
      "epoch": 1.948789653063008,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004619671910409706,
      "loss": 5.2539,
      "mean_token_accuracy": 0.2133048266172409,
      "num_tokens": 45185589.0,
      "step": 22300
    },
    {
      "entropy": 5.946296882629395,
      "epoch": 1.9492266014157127,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004619498535732116,
      "loss": 5.3986,
      "mean_token_accuracy": 0.19284196346998214,
      "num_tokens": 45195654.0,
      "step": 22305
    },
    {
      "entropy": 6.042048645019531,
      "epoch": 1.9496635497684174,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046193251251965646,
      "loss": 5.4175,
      "mean_token_accuracy": 0.19091781228780746,
      "num_tokens": 45205824.0,
      "step": 22310
    },
    {
      "entropy": 6.036351728439331,
      "epoch": 1.950100498121122,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004619151678806377,
      "loss": 5.3666,
      "mean_token_accuracy": 0.1968584954738617,
      "num_tokens": 45215228.0,
      "step": 22315
    },
    {
      "entropy": 5.945999574661255,
      "epoch": 1.9505374464738268,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004618978196564882,
      "loss": 5.2657,
      "mean_token_accuracy": 0.19797850102186204,
      "num_tokens": 45225043.0,
      "step": 22320
    },
    {
      "entropy": 5.977666187286377,
      "epoch": 1.9509743948265315,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004618804678475405,
      "loss": 5.3846,
      "mean_token_accuracy": 0.18967183083295822,
      "num_tokens": 45234837.0,
      "step": 22325
    },
    {
      "entropy": 5.956893634796143,
      "epoch": 1.9514113431792364,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046186311245412746,
      "loss": 5.2593,
      "mean_token_accuracy": 0.20254109054803848,
      "num_tokens": 45243540.0,
      "step": 22330
    },
    {
      "entropy": 5.931342029571534,
      "epoch": 1.9518482915319408,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004618457534765821,
      "loss": 5.3199,
      "mean_token_accuracy": 0.19375266134738922,
      "num_tokens": 45253053.0,
      "step": 22335
    },
    {
      "entropy": 5.914459657669068,
      "epoch": 1.9522852398846458,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004618283909152372,
      "loss": 5.2699,
      "mean_token_accuracy": 0.20102924555540086,
      "num_tokens": 45262973.0,
      "step": 22340
    },
    {
      "entropy": 5.948253202438354,
      "epoch": 1.9527221882373502,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004618110247704259,
      "loss": 5.2724,
      "mean_token_accuracy": 0.2085190460085869,
      "num_tokens": 45272951.0,
      "step": 22345
    },
    {
      "entropy": 6.012061738967896,
      "epoch": 1.9531591365900551,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046179365504248125,
      "loss": 5.3797,
      "mean_token_accuracy": 0.19831132590770723,
      "num_tokens": 45283596.0,
      "step": 22350
    },
    {
      "entropy": 5.9497106075286865,
      "epoch": 1.9535960849427596,
      "grad_norm": 1.125,
      "learning_rate": 0.00046177628173173647,
      "loss": 5.2121,
      "mean_token_accuracy": 0.21149141639471053,
      "num_tokens": 45294712.0,
      "step": 22355
    },
    {
      "entropy": 5.9305109024047855,
      "epoch": 1.9540330332954645,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004617589048385247,
      "loss": 5.3278,
      "mean_token_accuracy": 0.19764360338449477,
      "num_tokens": 45305681.0,
      "step": 22360
    },
    {
      "entropy": 5.966127061843872,
      "epoch": 1.9544699816481692,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004617415243631793,
      "loss": 5.3191,
      "mean_token_accuracy": 0.20282640755176545,
      "num_tokens": 45315332.0,
      "step": 22365
    },
    {
      "entropy": 5.902161598205566,
      "epoch": 1.954906930000874,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046172414030603356,
      "loss": 5.275,
      "mean_token_accuracy": 0.19772151559591294,
      "num_tokens": 45324526.0,
      "step": 22370
    },
    {
      "entropy": 5.953669166564941,
      "epoch": 1.9553438783535786,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000461706752667421,
      "loss": 5.3278,
      "mean_token_accuracy": 0.19918453097343444,
      "num_tokens": 45333989.0,
      "step": 22375
    },
    {
      "entropy": 5.984906435012817,
      "epoch": 1.9557808267062833,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004616893614476751,
      "loss": 5.2862,
      "mean_token_accuracy": 0.19828191697597503,
      "num_tokens": 45344545.0,
      "step": 22380
    },
    {
      "entropy": 5.9427565574646,
      "epoch": 1.956217775058988,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046167196664712945,
      "loss": 5.3296,
      "mean_token_accuracy": 0.19135435074567794,
      "num_tokens": 45354125.0,
      "step": 22385
    },
    {
      "entropy": 6.018604040145874,
      "epoch": 1.9566547234116927,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046165456826611763,
      "loss": 5.3818,
      "mean_token_accuracy": 0.19122352451086044,
      "num_tokens": 45364577.0,
      "step": 22390
    },
    {
      "entropy": 6.004033470153809,
      "epoch": 1.9570916717643976,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046163716630497336,
      "loss": 5.3529,
      "mean_token_accuracy": 0.20121804773807525,
      "num_tokens": 45374655.0,
      "step": 22395
    },
    {
      "entropy": 5.971172523498535,
      "epoch": 1.957528620117102,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046161976076403047,
      "loss": 5.3453,
      "mean_token_accuracy": 0.19005915075540541,
      "num_tokens": 45384383.0,
      "step": 22400
    },
    {
      "entropy": 5.9863605976104735,
      "epoch": 1.957965568469807,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004616023516436228,
      "loss": 5.345,
      "mean_token_accuracy": 0.19596186131238938,
      "num_tokens": 45394645.0,
      "step": 22405
    },
    {
      "entropy": 5.951040887832642,
      "epoch": 1.9584025168225114,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046158493894408423,
      "loss": 5.2567,
      "mean_token_accuracy": 0.2091382160782814,
      "num_tokens": 45404661.0,
      "step": 22410
    },
    {
      "entropy": 5.983701515197754,
      "epoch": 1.9588394651752163,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004615675226657487,
      "loss": 5.4191,
      "mean_token_accuracy": 0.19480680227279662,
      "num_tokens": 45413915.0,
      "step": 22415
    },
    {
      "entropy": 6.0368828773498535,
      "epoch": 1.959276413527921,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004615501028089504,
      "loss": 5.3114,
      "mean_token_accuracy": 0.19448467195034028,
      "num_tokens": 45424019.0,
      "step": 22420
    },
    {
      "entropy": 5.981932163238525,
      "epoch": 1.9597133618806257,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004615326793740233,
      "loss": 5.3056,
      "mean_token_accuracy": 0.19413352757692337,
      "num_tokens": 45435087.0,
      "step": 22425
    },
    {
      "entropy": 6.0161192417144775,
      "epoch": 1.9601503102333304,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046151525236130174,
      "loss": 5.3956,
      "mean_token_accuracy": 0.1911984995007515,
      "num_tokens": 45446363.0,
      "step": 22430
    },
    {
      "entropy": 6.01465482711792,
      "epoch": 1.960587258586035,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004614978217711199,
      "loss": 5.3418,
      "mean_token_accuracy": 0.19941654801368713,
      "num_tokens": 45456268.0,
      "step": 22435
    },
    {
      "entropy": 5.945563936233521,
      "epoch": 1.9610242069387398,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046148038760381217,
      "loss": 5.3011,
      "mean_token_accuracy": 0.20126101225614548,
      "num_tokens": 45466335.0,
      "step": 22440
    },
    {
      "entropy": 5.9413598537445065,
      "epoch": 1.9614611552914445,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004614629498597128,
      "loss": 5.3392,
      "mean_token_accuracy": 0.1895208552479744,
      "num_tokens": 45476128.0,
      "step": 22445
    },
    {
      "entropy": 5.947769355773926,
      "epoch": 1.9618981036441494,
      "grad_norm": 1.125,
      "learning_rate": 0.0004614455085391564,
      "loss": 5.2776,
      "mean_token_accuracy": 0.19739657491445542,
      "num_tokens": 45485986.0,
      "step": 22450
    },
    {
      "entropy": 5.974871110916138,
      "epoch": 1.9623350519968539,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004614280636424774,
      "loss": 5.3183,
      "mean_token_accuracy": 0.19653785973787308,
      "num_tokens": 45497270.0,
      "step": 22455
    },
    {
      "entropy": 5.94278883934021,
      "epoch": 1.9627720003495588,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004614106151700105,
      "loss": 5.3396,
      "mean_token_accuracy": 0.20182654857635499,
      "num_tokens": 45507415.0,
      "step": 22460
    },
    {
      "entropy": 5.954515552520752,
      "epoch": 1.9632089487022633,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004613931631220903,
      "loss": 5.3609,
      "mean_token_accuracy": 0.19426361322402955,
      "num_tokens": 45518025.0,
      "step": 22465
    },
    {
      "entropy": 5.990345621109009,
      "epoch": 1.9636458970549682,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046137570749905157,
      "loss": 5.3776,
      "mean_token_accuracy": 0.19676245152950286,
      "num_tokens": 45528834.0,
      "step": 22470
    },
    {
      "entropy": 6.0056641578674315,
      "epoch": 1.9640828454076729,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004613582483012291,
      "loss": 5.3003,
      "mean_token_accuracy": 0.19819145202636718,
      "num_tokens": 45538543.0,
      "step": 22475
    },
    {
      "entropy": 6.004667472839356,
      "epoch": 1.9645197937603776,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004613407855289578,
      "loss": 5.3947,
      "mean_token_accuracy": 0.19497627317905425,
      "num_tokens": 45549298.0,
      "step": 22480
    },
    {
      "entropy": 6.002863788604737,
      "epoch": 1.9649567421130822,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004613233191825726,
      "loss": 5.3688,
      "mean_token_accuracy": 0.19904514849185945,
      "num_tokens": 45560064.0,
      "step": 22485
    },
    {
      "entropy": 6.0056017398834225,
      "epoch": 1.965393690465787,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004613058492624084,
      "loss": 5.2454,
      "mean_token_accuracy": 0.20324240773916244,
      "num_tokens": 45569780.0,
      "step": 22490
    },
    {
      "entropy": 5.888987445831299,
      "epoch": 1.9658306388184916,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004612883757688005,
      "loss": 5.2516,
      "mean_token_accuracy": 0.20491398274898528,
      "num_tokens": 45580256.0,
      "step": 22495
    },
    {
      "entropy": 5.963739633560181,
      "epoch": 1.9662675871711963,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046127089870208395,
      "loss": 5.3567,
      "mean_token_accuracy": 0.19158184677362441,
      "num_tokens": 45590615.0,
      "step": 22500
    },
    {
      "entropy": 5.959015417098999,
      "epoch": 1.9667045355239012,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004612534180625939,
      "loss": 5.287,
      "mean_token_accuracy": 0.19879740178585054,
      "num_tokens": 45600215.0,
      "step": 22505
    },
    {
      "entropy": 5.909812974929809,
      "epoch": 1.9671414838766057,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046123593385066574,
      "loss": 5.3216,
      "mean_token_accuracy": 0.20621325224637985,
      "num_tokens": 45610484.0,
      "step": 22510
    },
    {
      "entropy": 5.960053396224976,
      "epoch": 1.9675784322293106,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046121844606663475,
      "loss": 5.3642,
      "mean_token_accuracy": 0.19556097388267518,
      "num_tokens": 45620528.0,
      "step": 22515
    },
    {
      "entropy": 6.0141009330749515,
      "epoch": 1.968015380582015,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004612009547108364,
      "loss": 5.2974,
      "mean_token_accuracy": 0.19666298925876619,
      "num_tokens": 45629985.0,
      "step": 22520
    },
    {
      "entropy": 6.032186841964721,
      "epoch": 1.96845232893472,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046118345978360615,
      "loss": 5.3614,
      "mean_token_accuracy": 0.19304227232933044,
      "num_tokens": 45639750.0,
      "step": 22525
    },
    {
      "entropy": 5.917926359176636,
      "epoch": 1.9688892772874245,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004611659612852796,
      "loss": 5.3077,
      "mean_token_accuracy": 0.1993032529950142,
      "num_tokens": 45650779.0,
      "step": 22530
    },
    {
      "entropy": 5.969728279113769,
      "epoch": 1.9693262256401294,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046114845921619234,
      "loss": 5.3046,
      "mean_token_accuracy": 0.19499922394752503,
      "num_tokens": 45660363.0,
      "step": 22535
    },
    {
      "entropy": 5.967263841629029,
      "epoch": 1.969763173992834,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004611309535766801,
      "loss": 5.3616,
      "mean_token_accuracy": 0.20016192495822907,
      "num_tokens": 45670474.0,
      "step": 22540
    },
    {
      "entropy": 5.880279064178467,
      "epoch": 1.9702001223455388,
      "grad_norm": 1.125,
      "learning_rate": 0.0004611134443670787,
      "loss": 5.2436,
      "mean_token_accuracy": 0.20852898359298705,
      "num_tokens": 45680157.0,
      "step": 22545
    },
    {
      "entropy": 5.972974729537964,
      "epoch": 1.9706370706982435,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004610959315877239,
      "loss": 5.3313,
      "mean_token_accuracy": 0.2018501028418541,
      "num_tokens": 45689340.0,
      "step": 22550
    },
    {
      "entropy": 6.01200122833252,
      "epoch": 1.9710740190509481,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046107841523895153,
      "loss": 5.3423,
      "mean_token_accuracy": 0.19452092796564102,
      "num_tokens": 45699453.0,
      "step": 22555
    },
    {
      "entropy": 5.929769229888916,
      "epoch": 1.9715109674036528,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046106089532109776,
      "loss": 5.3085,
      "mean_token_accuracy": 0.20042839646339417,
      "num_tokens": 45711513.0,
      "step": 22560
    },
    {
      "entropy": 5.9612384796142575,
      "epoch": 1.9719479157563575,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004610433718344985,
      "loss": 5.3918,
      "mean_token_accuracy": 0.19315457046031953,
      "num_tokens": 45722939.0,
      "step": 22565
    },
    {
      "entropy": 5.9739477157592775,
      "epoch": 1.9723848641090624,
      "grad_norm": 2.859375,
      "learning_rate": 0.00046102584477948985,
      "loss": 5.3324,
      "mean_token_accuracy": 0.20178834944963456,
      "num_tokens": 45733099.0,
      "step": 22570
    },
    {
      "entropy": 5.9389043807983395,
      "epoch": 1.972821812461767,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004610083141564079,
      "loss": 5.3718,
      "mean_token_accuracy": 0.20048662573099135,
      "num_tokens": 45743161.0,
      "step": 22575
    },
    {
      "entropy": 5.992891693115235,
      "epoch": 1.9732587608144718,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046099077996558923,
      "loss": 5.4263,
      "mean_token_accuracy": 0.19105946719646455,
      "num_tokens": 45752903.0,
      "step": 22580
    },
    {
      "entropy": 5.954986238479615,
      "epoch": 1.9736957091671763,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046097324220736993,
      "loss": 5.2589,
      "mean_token_accuracy": 0.19870272278785706,
      "num_tokens": 45763360.0,
      "step": 22585
    },
    {
      "entropy": 5.916384649276734,
      "epoch": 1.9741326575198812,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004609557008820863,
      "loss": 5.2104,
      "mean_token_accuracy": 0.21421562433242797,
      "num_tokens": 45773025.0,
      "step": 22590
    },
    {
      "entropy": 5.912256717681885,
      "epoch": 1.974569605872586,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004609381559900749,
      "loss": 5.2407,
      "mean_token_accuracy": 0.19684404134750366,
      "num_tokens": 45782844.0,
      "step": 22595
    },
    {
      "entropy": 5.943379592895508,
      "epoch": 1.9750065542252906,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046092060753167224,
      "loss": 5.3527,
      "mean_token_accuracy": 0.19516193270683288,
      "num_tokens": 45793058.0,
      "step": 22600
    },
    {
      "entropy": 6.010842990875244,
      "epoch": 1.9754435025779953,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004609030555072149,
      "loss": 5.3263,
      "mean_token_accuracy": 0.1995937556028366,
      "num_tokens": 45802865.0,
      "step": 22605
    },
    {
      "entropy": 5.952635622024536,
      "epoch": 1.9758804509307,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046088549991703953,
      "loss": 5.2817,
      "mean_token_accuracy": 0.20552800744771957,
      "num_tokens": 45812981.0,
      "step": 22610
    },
    {
      "entropy": 5.939983987808228,
      "epoch": 1.9763173992834047,
      "grad_norm": 1.125,
      "learning_rate": 0.000460867940761483,
      "loss": 5.2436,
      "mean_token_accuracy": 0.2000185191631317,
      "num_tokens": 45822477.0,
      "step": 22615
    },
    {
      "entropy": 5.938186502456665,
      "epoch": 1.9767543476361094,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004608503780408818,
      "loss": 5.4158,
      "mean_token_accuracy": 0.18821091055870057,
      "num_tokens": 45833642.0,
      "step": 22620
    },
    {
      "entropy": 6.014611530303955,
      "epoch": 1.9771912959888143,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004608328117555731,
      "loss": 5.3311,
      "mean_token_accuracy": 0.2001780942082405,
      "num_tokens": 45843079.0,
      "step": 22625
    },
    {
      "entropy": 6.0136415481567385,
      "epoch": 1.9776282443415187,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004608152419058936,
      "loss": 5.3556,
      "mean_token_accuracy": 0.19679764211177825,
      "num_tokens": 45852987.0,
      "step": 22630
    },
    {
      "entropy": 5.970058345794678,
      "epoch": 1.9780651926942237,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046079766849218043,
      "loss": 5.3182,
      "mean_token_accuracy": 0.19756889194250107,
      "num_tokens": 45863148.0,
      "step": 22635
    },
    {
      "entropy": 5.900472497940063,
      "epoch": 1.9785021410469281,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046078009151477063,
      "loss": 5.315,
      "mean_token_accuracy": 0.19952883273363115,
      "num_tokens": 45873430.0,
      "step": 22640
    },
    {
      "entropy": 6.03760027885437,
      "epoch": 1.978939089399633,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046076251097400126,
      "loss": 5.3321,
      "mean_token_accuracy": 0.1956624984741211,
      "num_tokens": 45883169.0,
      "step": 22645
    },
    {
      "entropy": 6.060557842254639,
      "epoch": 1.9793760377523377,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046074492687020967,
      "loss": 5.3765,
      "mean_token_accuracy": 0.19401061832904815,
      "num_tokens": 45893948.0,
      "step": 22650
    },
    {
      "entropy": 5.907715940475464,
      "epoch": 1.9798129861050424,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000460727339203733,
      "loss": 5.3435,
      "mean_token_accuracy": 0.2025020256638527,
      "num_tokens": 45904887.0,
      "step": 22655
    },
    {
      "entropy": 5.996358156204224,
      "epoch": 1.980249934457747,
      "grad_norm": 1.125,
      "learning_rate": 0.00046070974797490875,
      "loss": 5.3508,
      "mean_token_accuracy": 0.19170649498701095,
      "num_tokens": 45914845.0,
      "step": 22660
    },
    {
      "entropy": 5.973856353759766,
      "epoch": 1.9806868828104518,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004606921531840741,
      "loss": 5.3372,
      "mean_token_accuracy": 0.20307995229959488,
      "num_tokens": 45925104.0,
      "step": 22665
    },
    {
      "entropy": 5.953500699996948,
      "epoch": 1.9811238311631565,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004606745548315667,
      "loss": 5.2912,
      "mean_token_accuracy": 0.19308414906263352,
      "num_tokens": 45934934.0,
      "step": 22670
    },
    {
      "entropy": 5.952045679092407,
      "epoch": 1.9815607795158612,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004606569529177241,
      "loss": 5.3201,
      "mean_token_accuracy": 0.20539211183786393,
      "num_tokens": 45945449.0,
      "step": 22675
    },
    {
      "entropy": 6.014133024215698,
      "epoch": 1.981997727868566,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004606393474428838,
      "loss": 5.371,
      "mean_token_accuracy": 0.1894163027405739,
      "num_tokens": 45954897.0,
      "step": 22680
    },
    {
      "entropy": 5.965905714035034,
      "epoch": 1.9824346762212706,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046062173840738356,
      "loss": 5.2965,
      "mean_token_accuracy": 0.20758665204048157,
      "num_tokens": 45965085.0,
      "step": 22685
    },
    {
      "entropy": 5.97070722579956,
      "epoch": 1.9828716245739755,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004606041258115611,
      "loss": 5.3428,
      "mean_token_accuracy": 0.2057381048798561,
      "num_tokens": 45974610.0,
      "step": 22690
    },
    {
      "entropy": 6.004184579849243,
      "epoch": 1.98330857292668,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004605865096557542,
      "loss": 5.298,
      "mean_token_accuracy": 0.1948241412639618,
      "num_tokens": 45984922.0,
      "step": 22695
    },
    {
      "entropy": 5.955970811843872,
      "epoch": 1.9837455212793849,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004605688899403009,
      "loss": 5.303,
      "mean_token_accuracy": 0.20215994864702225,
      "num_tokens": 45994591.0,
      "step": 22700
    },
    {
      "entropy": 5.9890546798706055,
      "epoch": 1.9841824696320893,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046055126666553906,
      "loss": 5.4142,
      "mean_token_accuracy": 0.19223107844591142,
      "num_tokens": 46005291.0,
      "step": 22705
    },
    {
      "entropy": 5.9434051513671875,
      "epoch": 1.9846194179847942,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046053363983180657,
      "loss": 5.2964,
      "mean_token_accuracy": 0.1960642784833908,
      "num_tokens": 46014970.0,
      "step": 22710
    },
    {
      "entropy": 6.019408655166626,
      "epoch": 1.985056366337499,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046051600943944175,
      "loss": 5.3902,
      "mean_token_accuracy": 0.193535216152668,
      "num_tokens": 46025042.0,
      "step": 22715
    },
    {
      "entropy": 6.052403783798217,
      "epoch": 1.9854933146902036,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046049837548878264,
      "loss": 5.4081,
      "mean_token_accuracy": 0.18673135787248613,
      "num_tokens": 46035440.0,
      "step": 22720
    },
    {
      "entropy": 5.940548038482666,
      "epoch": 1.9859302630429083,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046048073798016746,
      "loss": 5.3268,
      "mean_token_accuracy": 0.20195142924785614,
      "num_tokens": 46045344.0,
      "step": 22725
    },
    {
      "entropy": 6.002640914916992,
      "epoch": 1.986367211395613,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046046309691393455,
      "loss": 5.3982,
      "mean_token_accuracy": 0.19363316744565964,
      "num_tokens": 46055054.0,
      "step": 22730
    },
    {
      "entropy": 6.045628404617309,
      "epoch": 1.9868041597483177,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046044545229042226,
      "loss": 5.3609,
      "mean_token_accuracy": 0.19384893923997878,
      "num_tokens": 46063835.0,
      "step": 22735
    },
    {
      "entropy": 5.990070819854736,
      "epoch": 1.9872411081010224,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046042780410996906,
      "loss": 5.3752,
      "mean_token_accuracy": 0.1930243566632271,
      "num_tokens": 46074256.0,
      "step": 22740
    },
    {
      "entropy": 6.0628180503845215,
      "epoch": 1.9876780564537273,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046041015237291346,
      "loss": 5.4256,
      "mean_token_accuracy": 0.19554405957460402,
      "num_tokens": 46085563.0,
      "step": 22745
    },
    {
      "entropy": 6.032908391952515,
      "epoch": 1.9881150048064318,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046039249707959386,
      "loss": 5.3382,
      "mean_token_accuracy": 0.20578670203685762,
      "num_tokens": 46094956.0,
      "step": 22750
    },
    {
      "entropy": 5.95951247215271,
      "epoch": 1.9885519531591367,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046037483823034906,
      "loss": 5.3443,
      "mean_token_accuracy": 0.19877746254205703,
      "num_tokens": 46105292.0,
      "step": 22755
    },
    {
      "entropy": 6.037571811676026,
      "epoch": 1.9889889015118412,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004603571758255177,
      "loss": 5.5041,
      "mean_token_accuracy": 0.18195998817682266,
      "num_tokens": 46117254.0,
      "step": 22760
    },
    {
      "entropy": 6.008747100830078,
      "epoch": 1.989425849864546,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046033950986543865,
      "loss": 5.3904,
      "mean_token_accuracy": 0.19268152862787247,
      "num_tokens": 46127846.0,
      "step": 22765
    },
    {
      "entropy": 5.988293075561524,
      "epoch": 1.9898627982172508,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046032184035045066,
      "loss": 5.4055,
      "mean_token_accuracy": 0.19476241022348403,
      "num_tokens": 46138476.0,
      "step": 22770
    },
    {
      "entropy": 5.973821020126342,
      "epoch": 1.9902997465699555,
      "grad_norm": 1.125,
      "learning_rate": 0.00046030416728089267,
      "loss": 5.2843,
      "mean_token_accuracy": 0.20004335939884185,
      "num_tokens": 46147789.0,
      "step": 22775
    },
    {
      "entropy": 5.982377624511718,
      "epoch": 1.9907366949226601,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046028649065710367,
      "loss": 5.4096,
      "mean_token_accuracy": 0.19571513235569,
      "num_tokens": 46158484.0,
      "step": 22780
    },
    {
      "entropy": 6.039363622665405,
      "epoch": 1.9911736432753648,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046026881047942273,
      "loss": 5.3721,
      "mean_token_accuracy": 0.19029768109321593,
      "num_tokens": 46168062.0,
      "step": 22785
    },
    {
      "entropy": 6.035207986831665,
      "epoch": 1.9916105916280695,
      "grad_norm": 1.125,
      "learning_rate": 0.00046025112674818883,
      "loss": 5.3223,
      "mean_token_accuracy": 0.19832725822925568,
      "num_tokens": 46178710.0,
      "step": 22790
    },
    {
      "entropy": 5.9457337856292725,
      "epoch": 1.9920475399807742,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046023343946374127,
      "loss": 5.3767,
      "mean_token_accuracy": 0.18763104379177092,
      "num_tokens": 46188728.0,
      "step": 22795
    },
    {
      "entropy": 5.927577114105224,
      "epoch": 1.9924844883334791,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046021574862641925,
      "loss": 5.3341,
      "mean_token_accuracy": 0.2028247222304344,
      "num_tokens": 46199716.0,
      "step": 22800
    },
    {
      "entropy": 6.049821949005127,
      "epoch": 1.9929214366861836,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046019805423656217,
      "loss": 5.4014,
      "mean_token_accuracy": 0.19125034064054489,
      "num_tokens": 46210885.0,
      "step": 22805
    },
    {
      "entropy": 5.929156303405762,
      "epoch": 1.9933583850388885,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004601803562945094,
      "loss": 5.3393,
      "mean_token_accuracy": 0.1993194043636322,
      "num_tokens": 46220879.0,
      "step": 22810
    },
    {
      "entropy": 5.9476096630096436,
      "epoch": 1.993795333391593,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004601626548006003,
      "loss": 5.384,
      "mean_token_accuracy": 0.19260996729135513,
      "num_tokens": 46230420.0,
      "step": 22815
    },
    {
      "entropy": 5.963873291015625,
      "epoch": 1.994232281744298,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004601449497551745,
      "loss": 5.3107,
      "mean_token_accuracy": 0.20054134577512742,
      "num_tokens": 46240548.0,
      "step": 22820
    },
    {
      "entropy": 5.991069173812866,
      "epoch": 1.9946692300970026,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046012724115857153,
      "loss": 5.3202,
      "mean_token_accuracy": 0.20280917286872863,
      "num_tokens": 46250484.0,
      "step": 22825
    },
    {
      "entropy": 6.030648326873779,
      "epoch": 1.9951061784497073,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046010952901113107,
      "loss": 5.3894,
      "mean_token_accuracy": 0.19339952766895294,
      "num_tokens": 46261184.0,
      "step": 22830
    },
    {
      "entropy": 5.933788061141968,
      "epoch": 1.995543126802412,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004600918133131928,
      "loss": 5.3215,
      "mean_token_accuracy": 0.20271418392658233,
      "num_tokens": 46271301.0,
      "step": 22835
    },
    {
      "entropy": 5.982035112380982,
      "epoch": 1.9959800751551167,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046007409406509663,
      "loss": 5.3706,
      "mean_token_accuracy": 0.18944068998098373,
      "num_tokens": 46281201.0,
      "step": 22840
    },
    {
      "entropy": 6.022264051437378,
      "epoch": 1.9964170235078214,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004600563712671823,
      "loss": 5.3865,
      "mean_token_accuracy": 0.1937905177474022,
      "num_tokens": 46291969.0,
      "step": 22845
    },
    {
      "entropy": 5.886409330368042,
      "epoch": 1.996853971860526,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046003864491978984,
      "loss": 5.1769,
      "mean_token_accuracy": 0.21325488835573198,
      "num_tokens": 46301308.0,
      "step": 22850
    },
    {
      "entropy": 5.976172828674317,
      "epoch": 1.997290920213231,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046002091502325926,
      "loss": 5.3367,
      "mean_token_accuracy": 0.19929159432649612,
      "num_tokens": 46311564.0,
      "step": 22855
    },
    {
      "entropy": 5.972661638259888,
      "epoch": 1.9977278685659354,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004600031815779306,
      "loss": 5.3038,
      "mean_token_accuracy": 0.204234579205513,
      "num_tokens": 46321843.0,
      "step": 22860
    },
    {
      "entropy": 5.950051259994507,
      "epoch": 1.9981648169186403,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045998544458414385,
      "loss": 5.2816,
      "mean_token_accuracy": 0.20022589713335037,
      "num_tokens": 46331274.0,
      "step": 22865
    },
    {
      "entropy": 6.048787260055542,
      "epoch": 1.9986017652713448,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045996770404223937,
      "loss": 5.4177,
      "mean_token_accuracy": 0.1896192654967308,
      "num_tokens": 46341385.0,
      "step": 22870
    },
    {
      "entropy": 5.966534852981567,
      "epoch": 1.9990387136240497,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045994995995255746,
      "loss": 5.3509,
      "mean_token_accuracy": 0.20243673175573348,
      "num_tokens": 46350385.0,
      "step": 22875
    },
    {
      "entropy": 5.900974893569947,
      "epoch": 1.9994756619767542,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045993221231543837,
      "loss": 5.231,
      "mean_token_accuracy": 0.2020118787884712,
      "num_tokens": 46361111.0,
      "step": 22880
    },
    {
      "entropy": 5.987285089492798,
      "epoch": 1.999912610329459,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004599144611312226,
      "loss": 5.3829,
      "mean_token_accuracy": 0.19374679923057556,
      "num_tokens": 46371316.0,
      "step": 22885
    },
    {
      "entropy": 5.994890022277832,
      "epoch": 2.0003495586821636,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045989670640025047,
      "loss": 5.303,
      "mean_token_accuracy": 0.20071399658918382,
      "num_tokens": 46381151.0,
      "step": 22890
    },
    {
      "entropy": 6.003724908828735,
      "epoch": 2.0007865070348685,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045987894812286276,
      "loss": 5.2705,
      "mean_token_accuracy": 0.20056966543197632,
      "num_tokens": 46391420.0,
      "step": 22895
    },
    {
      "entropy": 5.946291542053222,
      "epoch": 2.0012234553875734,
      "grad_norm": 1.125,
      "learning_rate": 0.00045986118629939984,
      "loss": 5.2443,
      "mean_token_accuracy": 0.20043755769729615,
      "num_tokens": 46401575.0,
      "step": 22900
    },
    {
      "entropy": 5.95262508392334,
      "epoch": 2.001660403740278,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004598434209302025,
      "loss": 5.2285,
      "mean_token_accuracy": 0.21127583533525468,
      "num_tokens": 46411275.0,
      "step": 22905
    },
    {
      "entropy": 5.931780004501343,
      "epoch": 2.002097352092983,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004598256520156116,
      "loss": 5.2882,
      "mean_token_accuracy": 0.20510634779930115,
      "num_tokens": 46421179.0,
      "step": 22910
    },
    {
      "entropy": 5.972609043121338,
      "epoch": 2.0025343004456873,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004598078795559678,
      "loss": 5.3371,
      "mean_token_accuracy": 0.19837545305490495,
      "num_tokens": 46431865.0,
      "step": 22915
    },
    {
      "entropy": 6.007204532623291,
      "epoch": 2.002971248798392,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045979010355161196,
      "loss": 5.3834,
      "mean_token_accuracy": 0.190933658182621,
      "num_tokens": 46443398.0,
      "step": 22920
    },
    {
      "entropy": 6.062713670730591,
      "epoch": 2.0034081971510966,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004597723240028852,
      "loss": 5.342,
      "mean_token_accuracy": 0.20048344433307647,
      "num_tokens": 46454818.0,
      "step": 22925
    },
    {
      "entropy": 6.037443828582764,
      "epoch": 2.0038451455038016,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004597545409101283,
      "loss": 5.3399,
      "mean_token_accuracy": 0.19832110553979873,
      "num_tokens": 46465231.0,
      "step": 22930
    },
    {
      "entropy": 5.994971084594726,
      "epoch": 2.004282093856506,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004597367542736826,
      "loss": 5.3126,
      "mean_token_accuracy": 0.20500663220882415,
      "num_tokens": 46476323.0,
      "step": 22935
    },
    {
      "entropy": 5.953034162521362,
      "epoch": 2.004719042209211,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004597189640938892,
      "loss": 5.2657,
      "mean_token_accuracy": 0.1992411121726036,
      "num_tokens": 46486573.0,
      "step": 22940
    },
    {
      "entropy": 5.926540374755859,
      "epoch": 2.0051559905619154,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004597011703710892,
      "loss": 5.2574,
      "mean_token_accuracy": 0.20334928929805757,
      "num_tokens": 46497779.0,
      "step": 22945
    },
    {
      "entropy": 5.961461114883423,
      "epoch": 2.0055929389146203,
      "grad_norm": 1.046875,
      "learning_rate": 0.000459683373105624,
      "loss": 5.3194,
      "mean_token_accuracy": 0.19812862277030946,
      "num_tokens": 46509074.0,
      "step": 22950
    },
    {
      "entropy": 5.924068832397461,
      "epoch": 2.006029887267325,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045966557229783493,
      "loss": 5.1583,
      "mean_token_accuracy": 0.21278628408908845,
      "num_tokens": 46519292.0,
      "step": 22955
    },
    {
      "entropy": 5.901964521408081,
      "epoch": 2.0064668356200297,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004596477679480634,
      "loss": 5.2383,
      "mean_token_accuracy": 0.21018197238445283,
      "num_tokens": 46529689.0,
      "step": 22960
    },
    {
      "entropy": 6.006374502182007,
      "epoch": 2.0069037839727346,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045962996005665096,
      "loss": 5.302,
      "mean_token_accuracy": 0.19701653569936753,
      "num_tokens": 46538616.0,
      "step": 22965
    },
    {
      "entropy": 5.974369859695434,
      "epoch": 2.007340732325439,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045961214862393914,
      "loss": 5.1723,
      "mean_token_accuracy": 0.208843794465065,
      "num_tokens": 46548372.0,
      "step": 22970
    },
    {
      "entropy": 5.947838830947876,
      "epoch": 2.007777680678144,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004595943336502695,
      "loss": 5.2948,
      "mean_token_accuracy": 0.19371893256902695,
      "num_tokens": 46558991.0,
      "step": 22975
    },
    {
      "entropy": 5.9144574165344235,
      "epoch": 2.0082146290308485,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004595765151359839,
      "loss": 5.2681,
      "mean_token_accuracy": 0.2063501298427582,
      "num_tokens": 46569531.0,
      "step": 22980
    },
    {
      "entropy": 6.029057836532592,
      "epoch": 2.0086515773835534,
      "grad_norm": 1.1875,
      "learning_rate": 0.000459558693081424,
      "loss": 5.3056,
      "mean_token_accuracy": 0.20244320034980773,
      "num_tokens": 46579774.0,
      "step": 22985
    },
    {
      "entropy": 6.022578144073487,
      "epoch": 2.009088525736258,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004595408674869317,
      "loss": 5.2584,
      "mean_token_accuracy": 0.20083824396133423,
      "num_tokens": 46589810.0,
      "step": 22990
    },
    {
      "entropy": 5.946121644973755,
      "epoch": 2.0095254740889628,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004595230383528488,
      "loss": 5.2383,
      "mean_token_accuracy": 0.20509810894727706,
      "num_tokens": 46599347.0,
      "step": 22995
    },
    {
      "entropy": 6.021002006530762,
      "epoch": 2.0099624224416672,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045950520567951734,
      "loss": 5.326,
      "mean_token_accuracy": 0.20064252763986587,
      "num_tokens": 46609926.0,
      "step": 23000
    },
    {
      "entropy": 5.9716143131256105,
      "epoch": 2.010399370794372,
      "grad_norm": 1.125,
      "learning_rate": 0.0004594873694672794,
      "loss": 5.2546,
      "mean_token_accuracy": 0.20461439043283464,
      "num_tokens": 46620736.0,
      "step": 23005
    },
    {
      "entropy": 5.963765239715576,
      "epoch": 2.0108363191470766,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045946952971647696,
      "loss": 5.2384,
      "mean_token_accuracy": 0.20040605068206788,
      "num_tokens": 46630713.0,
      "step": 23010
    },
    {
      "entropy": 5.948656415939331,
      "epoch": 2.0112732674997815,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004594516864274524,
      "loss": 5.2685,
      "mean_token_accuracy": 0.2053532660007477,
      "num_tokens": 46641197.0,
      "step": 23015
    },
    {
      "entropy": 5.887697505950928,
      "epoch": 2.0117102158524864,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045943383960054775,
      "loss": 5.1232,
      "mean_token_accuracy": 0.22843295335769653,
      "num_tokens": 46651325.0,
      "step": 23020
    },
    {
      "entropy": 5.93201322555542,
      "epoch": 2.012147164205191,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045941598923610543,
      "loss": 5.2773,
      "mean_token_accuracy": 0.20390161573886872,
      "num_tokens": 46662100.0,
      "step": 23025
    },
    {
      "entropy": 5.980863094329834,
      "epoch": 2.012584112557896,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045939813533446783,
      "loss": 5.2849,
      "mean_token_accuracy": 0.20124830901622773,
      "num_tokens": 46671460.0,
      "step": 23030
    },
    {
      "entropy": 6.021858263015747,
      "epoch": 2.0130210609106003,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004593802778959773,
      "loss": 5.3268,
      "mean_token_accuracy": 0.1975738525390625,
      "num_tokens": 46681835.0,
      "step": 23035
    },
    {
      "entropy": 5.894870281219482,
      "epoch": 2.013458009263305,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004593624169209765,
      "loss": 5.2454,
      "mean_token_accuracy": 0.20720392167568208,
      "num_tokens": 46691976.0,
      "step": 23040
    },
    {
      "entropy": 5.930480813980102,
      "epoch": 2.0138949576160097,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045934455240980793,
      "loss": 5.2409,
      "mean_token_accuracy": 0.20573797971010208,
      "num_tokens": 46702589.0,
      "step": 23045
    },
    {
      "entropy": 6.041665029525757,
      "epoch": 2.0143319059687146,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004593266843628143,
      "loss": 5.3606,
      "mean_token_accuracy": 0.19139478206634522,
      "num_tokens": 46712167.0,
      "step": 23050
    },
    {
      "entropy": 5.966901683807373,
      "epoch": 2.014768854321419,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004593088127803382,
      "loss": 5.2701,
      "mean_token_accuracy": 0.20076039135456086,
      "num_tokens": 46722684.0,
      "step": 23055
    },
    {
      "entropy": 5.942961645126343,
      "epoch": 2.015205802674124,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045929093766272253,
      "loss": 5.2761,
      "mean_token_accuracy": 0.20035096406936645,
      "num_tokens": 46733856.0,
      "step": 23060
    },
    {
      "entropy": 6.040079545974732,
      "epoch": 2.0156427510268284,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004592730590103101,
      "loss": 5.3536,
      "mean_token_accuracy": 0.19467889666557311,
      "num_tokens": 46744210.0,
      "step": 23065
    },
    {
      "entropy": 6.0176109790802,
      "epoch": 2.0160796993795334,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004592551768234438,
      "loss": 5.2306,
      "mean_token_accuracy": 0.20563261657953263,
      "num_tokens": 46754611.0,
      "step": 23070
    },
    {
      "entropy": 6.019075536727906,
      "epoch": 2.0165166477322383,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004592372911024667,
      "loss": 5.3487,
      "mean_token_accuracy": 0.19727785587310792,
      "num_tokens": 46765647.0,
      "step": 23075
    },
    {
      "entropy": 5.920847415924072,
      "epoch": 2.0169535960849427,
      "grad_norm": 1.125,
      "learning_rate": 0.00045921940184772184,
      "loss": 5.3234,
      "mean_token_accuracy": 0.19733468890190126,
      "num_tokens": 46776052.0,
      "step": 23080
    },
    {
      "entropy": 5.990440082550049,
      "epoch": 2.0173905444376476,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004592015090595523,
      "loss": 5.2637,
      "mean_token_accuracy": 0.2040497899055481,
      "num_tokens": 46786255.0,
      "step": 23085
    },
    {
      "entropy": 5.965693759918213,
      "epoch": 2.017827492790352,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004591836127383013,
      "loss": 5.2776,
      "mean_token_accuracy": 0.20574386566877365,
      "num_tokens": 46795958.0,
      "step": 23090
    },
    {
      "entropy": 5.955945730209351,
      "epoch": 2.018264441143057,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045916571288431206,
      "loss": 5.2947,
      "mean_token_accuracy": 0.19509933292865753,
      "num_tokens": 46806508.0,
      "step": 23095
    },
    {
      "entropy": 5.996252059936523,
      "epoch": 2.0187013894957615,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045914780949792797,
      "loss": 5.3498,
      "mean_token_accuracy": 0.19067890346050262,
      "num_tokens": 46816868.0,
      "step": 23100
    },
    {
      "entropy": 5.970154809951782,
      "epoch": 2.0191383378484664,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004591299025794924,
      "loss": 5.2832,
      "mean_token_accuracy": 0.20413204580545424,
      "num_tokens": 46827591.0,
      "step": 23105
    },
    {
      "entropy": 5.9498663425445555,
      "epoch": 2.019575286201171,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004591119921293488,
      "loss": 5.235,
      "mean_token_accuracy": 0.20680221766233445,
      "num_tokens": 46837394.0,
      "step": 23110
    },
    {
      "entropy": 5.935481977462769,
      "epoch": 2.020012234553876,
      "grad_norm": 1.0,
      "learning_rate": 0.00045909407814784074,
      "loss": 5.2415,
      "mean_token_accuracy": 0.20633308291435243,
      "num_tokens": 46847881.0,
      "step": 23115
    },
    {
      "entropy": 5.9885826587677,
      "epoch": 2.0204491829065803,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045907616063531185,
      "loss": 5.2784,
      "mean_token_accuracy": 0.19889021962881087,
      "num_tokens": 46857728.0,
      "step": 23120
    },
    {
      "entropy": 6.035140132904052,
      "epoch": 2.020886131259285,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004590582395921057,
      "loss": 5.3369,
      "mean_token_accuracy": 0.19391704946756363,
      "num_tokens": 46866991.0,
      "step": 23125
    },
    {
      "entropy": 6.048573112487793,
      "epoch": 2.0213230796119896,
      "grad_norm": 1.203125,
      "learning_rate": 0.000459040315018566,
      "loss": 5.3581,
      "mean_token_accuracy": 0.1964619994163513,
      "num_tokens": 46876253.0,
      "step": 23130
    },
    {
      "entropy": 5.979772663116455,
      "epoch": 2.0217600279646946,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045902238691503675,
      "loss": 5.2685,
      "mean_token_accuracy": 0.19772214144468309,
      "num_tokens": 46887309.0,
      "step": 23135
    },
    {
      "entropy": 6.005744123458863,
      "epoch": 2.0221969763173995,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045900445528186163,
      "loss": 5.3423,
      "mean_token_accuracy": 0.19848988801240922,
      "num_tokens": 46896722.0,
      "step": 23140
    },
    {
      "entropy": 5.913441133499146,
      "epoch": 2.022633924670104,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045898652011938456,
      "loss": 5.3005,
      "mean_token_accuracy": 0.200257308781147,
      "num_tokens": 46907519.0,
      "step": 23145
    },
    {
      "entropy": 5.943037366867065,
      "epoch": 2.023070873022809,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045896858142794976,
      "loss": 5.2679,
      "mean_token_accuracy": 0.203258715569973,
      "num_tokens": 46917387.0,
      "step": 23150
    },
    {
      "entropy": 6.036884927749634,
      "epoch": 2.0235078213755133,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004589506392079011,
      "loss": 5.2801,
      "mean_token_accuracy": 0.20892366766929626,
      "num_tokens": 46926477.0,
      "step": 23155
    },
    {
      "entropy": 5.99068808555603,
      "epoch": 2.0239447697282182,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045893269345958284,
      "loss": 5.2724,
      "mean_token_accuracy": 0.2054175928235054,
      "num_tokens": 46935807.0,
      "step": 23160
    },
    {
      "entropy": 5.927919721603393,
      "epoch": 2.0243817180809227,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004589147441833391,
      "loss": 5.288,
      "mean_token_accuracy": 0.20221598744392394,
      "num_tokens": 46946843.0,
      "step": 23165
    },
    {
      "entropy": 5.934528923034668,
      "epoch": 2.0248186664336276,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004588967913795142,
      "loss": 5.2448,
      "mean_token_accuracy": 0.20843347907066345,
      "num_tokens": 46956289.0,
      "step": 23170
    },
    {
      "entropy": 6.019388484954834,
      "epoch": 2.025255614786332,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045887883504845244,
      "loss": 5.2687,
      "mean_token_accuracy": 0.204288487136364,
      "num_tokens": 46965606.0,
      "step": 23175
    },
    {
      "entropy": 5.958960819244385,
      "epoch": 2.025692563139037,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045886087519049837,
      "loss": 5.2069,
      "mean_token_accuracy": 0.2084169641137123,
      "num_tokens": 46974769.0,
      "step": 23180
    },
    {
      "entropy": 6.019436025619507,
      "epoch": 2.0261295114917415,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045884291180599626,
      "loss": 5.3147,
      "mean_token_accuracy": 0.1921847090125084,
      "num_tokens": 46984753.0,
      "step": 23185
    },
    {
      "entropy": 5.912743282318115,
      "epoch": 2.0265664598444464,
      "grad_norm": 1.25,
      "learning_rate": 0.00045882494489529085,
      "loss": 5.2475,
      "mean_token_accuracy": 0.20293320566415787,
      "num_tokens": 46995403.0,
      "step": 23190
    },
    {
      "entropy": 6.003033065795899,
      "epoch": 2.0270034081971513,
      "grad_norm": 1.125,
      "learning_rate": 0.0004588069744587267,
      "loss": 5.276,
      "mean_token_accuracy": 0.2003016322851181,
      "num_tokens": 47005054.0,
      "step": 23195
    },
    {
      "entropy": 5.965152931213379,
      "epoch": 2.0274403565498558,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004587890004966484,
      "loss": 5.2825,
      "mean_token_accuracy": 0.19975292086601257,
      "num_tokens": 47016315.0,
      "step": 23200
    },
    {
      "entropy": 6.006221199035645,
      "epoch": 2.0278773049025607,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045877102300940077,
      "loss": 5.2261,
      "mean_token_accuracy": 0.19747364073991774,
      "num_tokens": 47025981.0,
      "step": 23205
    },
    {
      "entropy": 5.966948461532593,
      "epoch": 2.028314253255265,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045875304199732863,
      "loss": 5.2598,
      "mean_token_accuracy": 0.2041460931301117,
      "num_tokens": 47035906.0,
      "step": 23210
    },
    {
      "entropy": 5.968922233581543,
      "epoch": 2.02875120160797,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045873505746077685,
      "loss": 5.2865,
      "mean_token_accuracy": 0.19521610736846923,
      "num_tokens": 47045884.0,
      "step": 23215
    },
    {
      "entropy": 5.987559127807617,
      "epoch": 2.0291881499606745,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004587170694000904,
      "loss": 5.3267,
      "mean_token_accuracy": 0.20016647428274154,
      "num_tokens": 47055575.0,
      "step": 23220
    },
    {
      "entropy": 5.912248039245606,
      "epoch": 2.0296250983133794,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004586990778156144,
      "loss": 5.3035,
      "mean_token_accuracy": 0.20076699256896974,
      "num_tokens": 47065484.0,
      "step": 23225
    },
    {
      "entropy": 5.989146757125854,
      "epoch": 2.030062046666084,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045868108270769373,
      "loss": 5.2365,
      "mean_token_accuracy": 0.20653180927038192,
      "num_tokens": 47076604.0,
      "step": 23230
    },
    {
      "entropy": 5.939270925521851,
      "epoch": 2.030498995018789,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004586630840766736,
      "loss": 5.2466,
      "mean_token_accuracy": 0.2092230424284935,
      "num_tokens": 47086452.0,
      "step": 23235
    },
    {
      "entropy": 5.933835124969482,
      "epoch": 2.0309359433714933,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045864508192289933,
      "loss": 5.3377,
      "mean_token_accuracy": 0.19230954796075822,
      "num_tokens": 47096425.0,
      "step": 23240
    },
    {
      "entropy": 5.957029676437378,
      "epoch": 2.031372891724198,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045862707624671616,
      "loss": 5.2785,
      "mean_token_accuracy": 0.2017475962638855,
      "num_tokens": 47107214.0,
      "step": 23245
    },
    {
      "entropy": 5.981607341766358,
      "epoch": 2.031809840076903,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045860906704846944,
      "loss": 5.2949,
      "mean_token_accuracy": 0.20124715864658355,
      "num_tokens": 47117498.0,
      "step": 23250
    },
    {
      "entropy": 5.948982620239258,
      "epoch": 2.0322467884296076,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045859105432850464,
      "loss": 5.235,
      "mean_token_accuracy": 0.2013936907052994,
      "num_tokens": 47128031.0,
      "step": 23255
    },
    {
      "entropy": 6.037002468109131,
      "epoch": 2.0326837367823125,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004585730380871672,
      "loss": 5.2856,
      "mean_token_accuracy": 0.20617474466562272,
      "num_tokens": 47138347.0,
      "step": 23260
    },
    {
      "entropy": 5.884390735626221,
      "epoch": 2.033120685135017,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004585550183248027,
      "loss": 5.1828,
      "mean_token_accuracy": 0.20321082174777985,
      "num_tokens": 47147642.0,
      "step": 23265
    },
    {
      "entropy": 5.927065134048462,
      "epoch": 2.033557633487722,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045853699504175673,
      "loss": 5.2906,
      "mean_token_accuracy": 0.20556440204381943,
      "num_tokens": 47156855.0,
      "step": 23270
    },
    {
      "entropy": 5.99628586769104,
      "epoch": 2.0339945818404264,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045851896823837505,
      "loss": 5.2719,
      "mean_token_accuracy": 0.20318920761346818,
      "num_tokens": 47167030.0,
      "step": 23275
    },
    {
      "entropy": 5.974725723266602,
      "epoch": 2.0344315301931313,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045850093791500334,
      "loss": 5.2839,
      "mean_token_accuracy": 0.2034066677093506,
      "num_tokens": 47177471.0,
      "step": 23280
    },
    {
      "entropy": 6.010871934890747,
      "epoch": 2.0348684785458357,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045848290407198754,
      "loss": 5.3375,
      "mean_token_accuracy": 0.19647724479436873,
      "num_tokens": 47187911.0,
      "step": 23285
    },
    {
      "entropy": 5.981044101715088,
      "epoch": 2.0353054268985407,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004584648667096736,
      "loss": 5.2276,
      "mean_token_accuracy": 0.20514477640390397,
      "num_tokens": 47198622.0,
      "step": 23290
    },
    {
      "entropy": 5.947878646850586,
      "epoch": 2.035742375251245,
      "grad_norm": 1.125,
      "learning_rate": 0.0004584468258284072,
      "loss": 5.3492,
      "mean_token_accuracy": 0.19826371669769288,
      "num_tokens": 47207593.0,
      "step": 23295
    },
    {
      "entropy": 6.006932878494263,
      "epoch": 2.03617932360395,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004584287814285347,
      "loss": 5.2942,
      "mean_token_accuracy": 0.19581635147333146,
      "num_tokens": 47218734.0,
      "step": 23300
    },
    {
      "entropy": 6.0089562892913815,
      "epoch": 2.0366162719566545,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045841073351040196,
      "loss": 5.3172,
      "mean_token_accuracy": 0.19651291966438295,
      "num_tokens": 47228662.0,
      "step": 23305
    },
    {
      "entropy": 6.04684419631958,
      "epoch": 2.0370532203093594,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004583926820743553,
      "loss": 5.4038,
      "mean_token_accuracy": 0.1997642546892166,
      "num_tokens": 47238460.0,
      "step": 23310
    },
    {
      "entropy": 6.022532796859741,
      "epoch": 2.0374901686620643,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004583746271207409,
      "loss": 5.309,
      "mean_token_accuracy": 0.19307074546813965,
      "num_tokens": 47247789.0,
      "step": 23315
    },
    {
      "entropy": 5.955890226364136,
      "epoch": 2.037927117014769,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045835656864990504,
      "loss": 5.3205,
      "mean_token_accuracy": 0.1980265870690346,
      "num_tokens": 47257901.0,
      "step": 23320
    },
    {
      "entropy": 6.016879749298096,
      "epoch": 2.0383640653674737,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004583385066621941,
      "loss": 5.2912,
      "mean_token_accuracy": 0.20296597480773926,
      "num_tokens": 47268362.0,
      "step": 23325
    },
    {
      "entropy": 5.9696752548217775,
      "epoch": 2.038801013720178,
      "grad_norm": 1.25,
      "learning_rate": 0.00045832044115795464,
      "loss": 5.2874,
      "mean_token_accuracy": 0.20396648645401,
      "num_tokens": 47278297.0,
      "step": 23330
    },
    {
      "entropy": 5.997499179840088,
      "epoch": 2.039237962072883,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045830237213753287,
      "loss": 5.2895,
      "mean_token_accuracy": 0.1932594895362854,
      "num_tokens": 47288795.0,
      "step": 23335
    },
    {
      "entropy": 5.966319131851196,
      "epoch": 2.0396749104255876,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004582842996012757,
      "loss": 5.249,
      "mean_token_accuracy": 0.20559716373682022,
      "num_tokens": 47298373.0,
      "step": 23340
    },
    {
      "entropy": 5.958977365493775,
      "epoch": 2.0401118587782925,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045826622354952956,
      "loss": 5.3346,
      "mean_token_accuracy": 0.19926563203334807,
      "num_tokens": 47308160.0,
      "step": 23345
    },
    {
      "entropy": 6.021490001678467,
      "epoch": 2.040548807130997,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045824814398264126,
      "loss": 5.2887,
      "mean_token_accuracy": 0.20548501759767532,
      "num_tokens": 47317626.0,
      "step": 23350
    },
    {
      "entropy": 6.017267894744873,
      "epoch": 2.040985755483702,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045823006090095757,
      "loss": 5.3676,
      "mean_token_accuracy": 0.1934916615486145,
      "num_tokens": 47327221.0,
      "step": 23355
    },
    {
      "entropy": 6.005812788009644,
      "epoch": 2.0414227038364063,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004582119743048253,
      "loss": 5.3192,
      "mean_token_accuracy": 0.20307274162769318,
      "num_tokens": 47337385.0,
      "step": 23360
    },
    {
      "entropy": 5.979656839370728,
      "epoch": 2.0418596521891113,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004581938841945912,
      "loss": 5.2718,
      "mean_token_accuracy": 0.2008368194103241,
      "num_tokens": 47346894.0,
      "step": 23365
    },
    {
      "entropy": 5.976390409469604,
      "epoch": 2.042296600541816,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045817579057060254,
      "loss": 5.3178,
      "mean_token_accuracy": 0.201950141787529,
      "num_tokens": 47356192.0,
      "step": 23370
    },
    {
      "entropy": 5.957108020782471,
      "epoch": 2.0427335488945206,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004581576934332063,
      "loss": 5.2346,
      "mean_token_accuracy": 0.20256718695163728,
      "num_tokens": 47366704.0,
      "step": 23375
    },
    {
      "entropy": 5.976712989807129,
      "epoch": 2.0431704972472255,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004581395927827494,
      "loss": 5.2459,
      "mean_token_accuracy": 0.205047345161438,
      "num_tokens": 47377068.0,
      "step": 23380
    },
    {
      "entropy": 5.938499164581299,
      "epoch": 2.04360744559993,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045812148861957923,
      "loss": 5.2827,
      "mean_token_accuracy": 0.20463809072971345,
      "num_tokens": 47387268.0,
      "step": 23385
    },
    {
      "entropy": 5.97277946472168,
      "epoch": 2.044044393952635,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004581033809440429,
      "loss": 5.2346,
      "mean_token_accuracy": 0.2053939715027809,
      "num_tokens": 47397508.0,
      "step": 23390
    },
    {
      "entropy": 5.9094995021820065,
      "epoch": 2.0444813423053394,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004580852697564878,
      "loss": 5.2938,
      "mean_token_accuracy": 0.19867694079875947,
      "num_tokens": 47407871.0,
      "step": 23395
    },
    {
      "entropy": 5.991778564453125,
      "epoch": 2.0449182906580443,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045806715505726135,
      "loss": 5.3918,
      "mean_token_accuracy": 0.18727514743804932,
      "num_tokens": 47417089.0,
      "step": 23400
    },
    {
      "entropy": 6.019281625747681,
      "epoch": 2.045355239010749,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045804903684671083,
      "loss": 5.269,
      "mean_token_accuracy": 0.20527396053075792,
      "num_tokens": 47426746.0,
      "step": 23405
    },
    {
      "entropy": 6.004535675048828,
      "epoch": 2.0457921873634537,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045803091512518397,
      "loss": 5.3109,
      "mean_token_accuracy": 0.20276014059782027,
      "num_tokens": 47437873.0,
      "step": 23410
    },
    {
      "entropy": 5.916994237899781,
      "epoch": 2.046229135716158,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045801278989302815,
      "loss": 5.2446,
      "mean_token_accuracy": 0.2033937469124794,
      "num_tokens": 47448520.0,
      "step": 23415
    },
    {
      "entropy": 5.999811506271362,
      "epoch": 2.046666084068863,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004579946611505912,
      "loss": 5.302,
      "mean_token_accuracy": 0.19262860268354415,
      "num_tokens": 47458323.0,
      "step": 23420
    },
    {
      "entropy": 6.012288188934326,
      "epoch": 2.0471030324215675,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004579765288982207,
      "loss": 5.3783,
      "mean_token_accuracy": 0.19708235263824464,
      "num_tokens": 47468061.0,
      "step": 23425
    },
    {
      "entropy": 5.985861206054688,
      "epoch": 2.0475399807742725,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045795839313626455,
      "loss": 5.3256,
      "mean_token_accuracy": 0.19986306875944138,
      "num_tokens": 47477541.0,
      "step": 23430
    },
    {
      "entropy": 5.951329565048217,
      "epoch": 2.0479769291269774,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004579402538650705,
      "loss": 5.2497,
      "mean_token_accuracy": 0.20517164021730422,
      "num_tokens": 47488570.0,
      "step": 23435
    },
    {
      "entropy": 6.002024984359741,
      "epoch": 2.048413877479682,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004579221110849866,
      "loss": 5.2859,
      "mean_token_accuracy": 0.19863407015800477,
      "num_tokens": 47498386.0,
      "step": 23440
    },
    {
      "entropy": 5.953047370910644,
      "epoch": 2.0488508258323868,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004579039647963607,
      "loss": 5.1884,
      "mean_token_accuracy": 0.20728788375854493,
      "num_tokens": 47507292.0,
      "step": 23445
    },
    {
      "entropy": 6.011994218826294,
      "epoch": 2.0492877741850912,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045788581499954094,
      "loss": 5.2951,
      "mean_token_accuracy": 0.1959016278386116,
      "num_tokens": 47516971.0,
      "step": 23450
    },
    {
      "entropy": 5.919968700408935,
      "epoch": 2.049724722537796,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045786766169487536,
      "loss": 5.17,
      "mean_token_accuracy": 0.21738371551036834,
      "num_tokens": 47527233.0,
      "step": 23455
    },
    {
      "entropy": 5.951792907714844,
      "epoch": 2.0501616708905006,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045784950488271227,
      "loss": 5.2359,
      "mean_token_accuracy": 0.20075401812791824,
      "num_tokens": 47537542.0,
      "step": 23460
    },
    {
      "entropy": 5.978523826599121,
      "epoch": 2.0505986192432055,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045783134456339985,
      "loss": 5.2267,
      "mean_token_accuracy": 0.20551769882440568,
      "num_tokens": 47546613.0,
      "step": 23465
    },
    {
      "entropy": 5.983701086044311,
      "epoch": 2.05103556759591,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045781318073728647,
      "loss": 5.2516,
      "mean_token_accuracy": 0.2048887848854065,
      "num_tokens": 47555462.0,
      "step": 23470
    },
    {
      "entropy": 5.993268871307373,
      "epoch": 2.051472515948615,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004577950134047204,
      "loss": 5.2302,
      "mean_token_accuracy": 0.20169572085142135,
      "num_tokens": 47565429.0,
      "step": 23475
    },
    {
      "entropy": 6.027115154266357,
      "epoch": 2.0519094643013194,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004577768425660503,
      "loss": 5.3803,
      "mean_token_accuracy": 0.19613909870386123,
      "num_tokens": 47576021.0,
      "step": 23480
    },
    {
      "entropy": 6.0205694198608395,
      "epoch": 2.0523464126540243,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004577586682216245,
      "loss": 5.3147,
      "mean_token_accuracy": 0.20120210349559783,
      "num_tokens": 47586016.0,
      "step": 23485
    },
    {
      "entropy": 5.9906919479370115,
      "epoch": 2.052783361006729,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004577404903717917,
      "loss": 5.2959,
      "mean_token_accuracy": 0.2004617214202881,
      "num_tokens": 47595673.0,
      "step": 23490
    },
    {
      "entropy": 5.957867765426636,
      "epoch": 2.0532203093594337,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004577223090169006,
      "loss": 5.2514,
      "mean_token_accuracy": 0.20546159893274307,
      "num_tokens": 47605252.0,
      "step": 23495
    },
    {
      "entropy": 5.948510217666626,
      "epoch": 2.0536572577121386,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004577041241572998,
      "loss": 5.267,
      "mean_token_accuracy": 0.19856192916631699,
      "num_tokens": 47614726.0,
      "step": 23500
    },
    {
      "entropy": 5.927587699890137,
      "epoch": 2.054094206064843,
      "grad_norm": 1.125,
      "learning_rate": 0.00045768593579333816,
      "loss": 5.2746,
      "mean_token_accuracy": 0.20949845612049103,
      "num_tokens": 47625618.0,
      "step": 23505
    },
    {
      "entropy": 6.022929000854492,
      "epoch": 2.054531154417548,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004576677439253646,
      "loss": 5.332,
      "mean_token_accuracy": 0.19899977296590804,
      "num_tokens": 47635702.0,
      "step": 23510
    },
    {
      "entropy": 5.9340517044067385,
      "epoch": 2.0549681027702524,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045764954855372796,
      "loss": 5.2967,
      "mean_token_accuracy": 0.20427961647510529,
      "num_tokens": 47647301.0,
      "step": 23515
    },
    {
      "entropy": 6.004848432540894,
      "epoch": 2.0554050511229573,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004576313496787772,
      "loss": 5.4169,
      "mean_token_accuracy": 0.18829720467329025,
      "num_tokens": 47656844.0,
      "step": 23520
    },
    {
      "entropy": 6.062983465194702,
      "epoch": 2.055841999475662,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004576131473008616,
      "loss": 5.2553,
      "mean_token_accuracy": 0.20448684692382812,
      "num_tokens": 47666196.0,
      "step": 23525
    },
    {
      "entropy": 5.95715651512146,
      "epoch": 2.0562789478283667,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004575949414203301,
      "loss": 5.2603,
      "mean_token_accuracy": 0.20045100003480912,
      "num_tokens": 47675329.0,
      "step": 23530
    },
    {
      "entropy": 5.88839898109436,
      "epoch": 2.056715896181071,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045757673203753195,
      "loss": 5.2572,
      "mean_token_accuracy": 0.20455398559570312,
      "num_tokens": 47686806.0,
      "step": 23535
    },
    {
      "entropy": 6.003586006164551,
      "epoch": 2.057152844533776,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004575585191528164,
      "loss": 5.2465,
      "mean_token_accuracy": 0.20412567257881165,
      "num_tokens": 47696664.0,
      "step": 23540
    },
    {
      "entropy": 5.992003679275513,
      "epoch": 2.057589792886481,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045754030276653275,
      "loss": 5.3169,
      "mean_token_accuracy": 0.19945477694272995,
      "num_tokens": 47707257.0,
      "step": 23545
    },
    {
      "entropy": 5.9581766605377195,
      "epoch": 2.0580267412391855,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004575220828790305,
      "loss": 5.2292,
      "mean_token_accuracy": 0.20634962022304534,
      "num_tokens": 47716345.0,
      "step": 23550
    },
    {
      "entropy": 5.92804479598999,
      "epoch": 2.0584636895918904,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045750385949065904,
      "loss": 5.2269,
      "mean_token_accuracy": 0.2083663135766983,
      "num_tokens": 47727779.0,
      "step": 23555
    },
    {
      "entropy": 5.981906890869141,
      "epoch": 2.058900637944595,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004574856326017679,
      "loss": 5.3249,
      "mean_token_accuracy": 0.20108399242162706,
      "num_tokens": 47737591.0,
      "step": 23560
    },
    {
      "entropy": 6.0139703273773195,
      "epoch": 2.0593375862973,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045746740221270664,
      "loss": 5.2734,
      "mean_token_accuracy": 0.20077636390924453,
      "num_tokens": 47746871.0,
      "step": 23565
    },
    {
      "entropy": 6.027799987792969,
      "epoch": 2.0597745346500043,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004574491683238251,
      "loss": 5.3054,
      "mean_token_accuracy": 0.20017154216766359,
      "num_tokens": 47757089.0,
      "step": 23570
    },
    {
      "entropy": 5.968857097625732,
      "epoch": 2.060211483002709,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004574309309354729,
      "loss": 5.3113,
      "mean_token_accuracy": 0.19811677634716035,
      "num_tokens": 47766853.0,
      "step": 23575
    },
    {
      "entropy": 5.974587249755859,
      "epoch": 2.0606484313554136,
      "grad_norm": 1.28125,
      "learning_rate": 0.00045741269004799984,
      "loss": 5.2662,
      "mean_token_accuracy": 0.20334978103637696,
      "num_tokens": 47776549.0,
      "step": 23580
    },
    {
      "entropy": 6.0003743171691895,
      "epoch": 2.0610853797081186,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045739444566175584,
      "loss": 5.3277,
      "mean_token_accuracy": 0.1940478652715683,
      "num_tokens": 47786641.0,
      "step": 23585
    },
    {
      "entropy": 6.015611982345581,
      "epoch": 2.061522328060823,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045737619777709077,
      "loss": 5.3451,
      "mean_token_accuracy": 0.19363989681005478,
      "num_tokens": 47796678.0,
      "step": 23590
    },
    {
      "entropy": 5.999029207229614,
      "epoch": 2.061959276413528,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004573579463943546,
      "loss": 5.2828,
      "mean_token_accuracy": 0.20389488488435745,
      "num_tokens": 47806678.0,
      "step": 23595
    },
    {
      "entropy": 5.9816371440887455,
      "epoch": 2.062396224766233,
      "grad_norm": 1.125,
      "learning_rate": 0.00045733969151389754,
      "loss": 5.3016,
      "mean_token_accuracy": 0.19748468995094298,
      "num_tokens": 47816619.0,
      "step": 23600
    },
    {
      "entropy": 5.92728590965271,
      "epoch": 2.0628331731189373,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004573214331360696,
      "loss": 5.2726,
      "mean_token_accuracy": 0.20210895985364913,
      "num_tokens": 47825437.0,
      "step": 23605
    },
    {
      "entropy": 5.992086553573609,
      "epoch": 2.0632701214716422,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004573031712612211,
      "loss": 5.2533,
      "mean_token_accuracy": 0.20195819437503815,
      "num_tokens": 47835604.0,
      "step": 23610
    },
    {
      "entropy": 5.9613371849060055,
      "epoch": 2.0637070698243467,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004572849058897022,
      "loss": 5.2384,
      "mean_token_accuracy": 0.20954297333955765,
      "num_tokens": 47845299.0,
      "step": 23615
    },
    {
      "entropy": 5.911047744750976,
      "epoch": 2.0641440181770516,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004572666370218633,
      "loss": 5.2673,
      "mean_token_accuracy": 0.2068694517016411,
      "num_tokens": 47855293.0,
      "step": 23620
    },
    {
      "entropy": 5.990138339996338,
      "epoch": 2.064580966529756,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045724836465805483,
      "loss": 5.2978,
      "mean_token_accuracy": 0.19968814104795457,
      "num_tokens": 47864809.0,
      "step": 23625
    },
    {
      "entropy": 5.9842897891998295,
      "epoch": 2.065017914882461,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004572300887986272,
      "loss": 5.317,
      "mean_token_accuracy": 0.19274878054857253,
      "num_tokens": 47874032.0,
      "step": 23630
    },
    {
      "entropy": 5.976044178009033,
      "epoch": 2.0654548632351655,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045721180944393095,
      "loss": 5.2704,
      "mean_token_accuracy": 0.20396225154399872,
      "num_tokens": 47884465.0,
      "step": 23635
    },
    {
      "entropy": 5.946150493621826,
      "epoch": 2.0658918115878704,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004571935265943168,
      "loss": 5.2428,
      "mean_token_accuracy": 0.21182102411985398,
      "num_tokens": 47894030.0,
      "step": 23640
    },
    {
      "entropy": 6.0196695804595945,
      "epoch": 2.066328759940575,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045717524025013526,
      "loss": 5.2999,
      "mean_token_accuracy": 0.19555599987506866,
      "num_tokens": 47904338.0,
      "step": 23645
    },
    {
      "entropy": 5.986866855621338,
      "epoch": 2.0667657082932798,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004571569504117372,
      "loss": 5.3193,
      "mean_token_accuracy": 0.1922439754009247,
      "num_tokens": 47914384.0,
      "step": 23650
    },
    {
      "entropy": 5.942481231689453,
      "epoch": 2.0672026566459842,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004571386570794734,
      "loss": 5.2163,
      "mean_token_accuracy": 0.20585158169269563,
      "num_tokens": 47924386.0,
      "step": 23655
    },
    {
      "entropy": 6.0230378150939945,
      "epoch": 2.067639604998689,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045712036025369465,
      "loss": 5.3009,
      "mean_token_accuracy": 0.20295849740505217,
      "num_tokens": 47935379.0,
      "step": 23660
    },
    {
      "entropy": 5.958348512649536,
      "epoch": 2.068076553351394,
      "grad_norm": 1.078125,
      "learning_rate": 0.000457102059934752,
      "loss": 5.2729,
      "mean_token_accuracy": 0.19996372908353804,
      "num_tokens": 47945657.0,
      "step": 23665
    },
    {
      "entropy": 5.968524026870727,
      "epoch": 2.0685135017040985,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045708375612299646,
      "loss": 5.2528,
      "mean_token_accuracy": 0.19964286386966706,
      "num_tokens": 47956294.0,
      "step": 23670
    },
    {
      "entropy": 5.9689921855926515,
      "epoch": 2.0689504500568034,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045706544881877895,
      "loss": 5.3849,
      "mean_token_accuracy": 0.1901520609855652,
      "num_tokens": 47967362.0,
      "step": 23675
    },
    {
      "entropy": 6.001163578033447,
      "epoch": 2.069387398409508,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004570471380224508,
      "loss": 5.2892,
      "mean_token_accuracy": 0.2072314515709877,
      "num_tokens": 47977521.0,
      "step": 23680
    },
    {
      "entropy": 5.972126436233521,
      "epoch": 2.069824346762213,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045702882373436317,
      "loss": 5.291,
      "mean_token_accuracy": 0.2062172695994377,
      "num_tokens": 47988485.0,
      "step": 23685
    },
    {
      "entropy": 5.942620897293091,
      "epoch": 2.0702612951149173,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004570105059548673,
      "loss": 5.3069,
      "mean_token_accuracy": 0.1943725526332855,
      "num_tokens": 47998696.0,
      "step": 23690
    },
    {
      "entropy": 5.936844301223755,
      "epoch": 2.070698243467622,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004569921846843146,
      "loss": 5.1285,
      "mean_token_accuracy": 0.217216295003891,
      "num_tokens": 48007843.0,
      "step": 23695
    },
    {
      "entropy": 5.997303676605225,
      "epoch": 2.0711351918203267,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004569738599230563,
      "loss": 5.2551,
      "mean_token_accuracy": 0.20477324724197388,
      "num_tokens": 48017579.0,
      "step": 23700
    },
    {
      "entropy": 5.93058500289917,
      "epoch": 2.0715721401730316,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045695553167144417,
      "loss": 5.2495,
      "mean_token_accuracy": 0.20687267780303956,
      "num_tokens": 48027267.0,
      "step": 23705
    },
    {
      "entropy": 5.986223459243774,
      "epoch": 2.072009088525736,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045693719992982956,
      "loss": 5.2873,
      "mean_token_accuracy": 0.2016694501042366,
      "num_tokens": 48038064.0,
      "step": 23710
    },
    {
      "entropy": 6.079101991653443,
      "epoch": 2.072446036878441,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045691886469856406,
      "loss": 5.3826,
      "mean_token_accuracy": 0.19004255086183547,
      "num_tokens": 48048316.0,
      "step": 23715
    },
    {
      "entropy": 5.953252935409546,
      "epoch": 2.072882985231146,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004569005259779995,
      "loss": 5.3431,
      "mean_token_accuracy": 0.19733480215072632,
      "num_tokens": 48059083.0,
      "step": 23720
    },
    {
      "entropy": 5.965694379806519,
      "epoch": 2.0733199335838504,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045688218376848744,
      "loss": 5.2694,
      "mean_token_accuracy": 0.1957753121852875,
      "num_tokens": 48069184.0,
      "step": 23725
    },
    {
      "entropy": 5.993287229537964,
      "epoch": 2.0737568819365553,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004568638380703799,
      "loss": 5.3194,
      "mean_token_accuracy": 0.19417759776115417,
      "num_tokens": 48078737.0,
      "step": 23730
    },
    {
      "entropy": 5.976222324371338,
      "epoch": 2.0741938302892597,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004568454888840285,
      "loss": 5.3032,
      "mean_token_accuracy": 0.1958940237760544,
      "num_tokens": 48088581.0,
      "step": 23735
    },
    {
      "entropy": 5.857799291610718,
      "epoch": 2.0746307786419647,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004568271362097855,
      "loss": 5.1534,
      "mean_token_accuracy": 0.22632022351026534,
      "num_tokens": 48098941.0,
      "step": 23740
    },
    {
      "entropy": 5.956092882156372,
      "epoch": 2.075067726994669,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045680878004800265,
      "loss": 5.2684,
      "mean_token_accuracy": 0.1985607072710991,
      "num_tokens": 48108773.0,
      "step": 23745
    },
    {
      "entropy": 6.012767314910889,
      "epoch": 2.075504675347374,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004567904203990322,
      "loss": 5.3124,
      "mean_token_accuracy": 0.20393790006637574,
      "num_tokens": 48117699.0,
      "step": 23750
    },
    {
      "entropy": 5.981554841995239,
      "epoch": 2.0759416237000785,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004567720572632261,
      "loss": 5.2123,
      "mean_token_accuracy": 0.2135670781135559,
      "num_tokens": 48127693.0,
      "step": 23755
    },
    {
      "entropy": 6.0019303321838375,
      "epoch": 2.0763785720527834,
      "grad_norm": 1.125,
      "learning_rate": 0.00045675369064093686,
      "loss": 5.3395,
      "mean_token_accuracy": 0.19261172711849212,
      "num_tokens": 48138073.0,
      "step": 23760
    },
    {
      "entropy": 5.958858919143677,
      "epoch": 2.076815520405488,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004567353205325165,
      "loss": 5.1963,
      "mean_token_accuracy": 0.21042153835296631,
      "num_tokens": 48147504.0,
      "step": 23765
    },
    {
      "entropy": 6.012579727172851,
      "epoch": 2.077252468758193,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004567169469383175,
      "loss": 5.2816,
      "mean_token_accuracy": 0.1972183033823967,
      "num_tokens": 48156739.0,
      "step": 23770
    },
    {
      "entropy": 5.967541265487671,
      "epoch": 2.0776894171108973,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045669856985869225,
      "loss": 5.2762,
      "mean_token_accuracy": 0.19794319421052933,
      "num_tokens": 48166883.0,
      "step": 23775
    },
    {
      "entropy": 5.932852697372437,
      "epoch": 2.078126365463602,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004566801892939932,
      "loss": 5.218,
      "mean_token_accuracy": 0.21063703149557114,
      "num_tokens": 48177738.0,
      "step": 23780
    },
    {
      "entropy": 5.948004722595215,
      "epoch": 2.078563313816307,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004566618052445729,
      "loss": 5.2248,
      "mean_token_accuracy": 0.20694657564163207,
      "num_tokens": 48187979.0,
      "step": 23785
    },
    {
      "entropy": 5.964308547973633,
      "epoch": 2.0790002621690116,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000456643417710784,
      "loss": 5.3123,
      "mean_token_accuracy": 0.1970108762383461,
      "num_tokens": 48198579.0,
      "step": 23790
    },
    {
      "entropy": 6.027055978775024,
      "epoch": 2.0794372105217165,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045662502669297915,
      "loss": 5.414,
      "mean_token_accuracy": 0.19000465720891951,
      "num_tokens": 48208860.0,
      "step": 23795
    },
    {
      "entropy": 6.013506317138672,
      "epoch": 2.079874158874421,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004566066321915112,
      "loss": 5.2965,
      "mean_token_accuracy": 0.20841639190912248,
      "num_tokens": 48218842.0,
      "step": 23800
    },
    {
      "entropy": 5.969681310653686,
      "epoch": 2.080311107227126,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004565882342067328,
      "loss": 5.2388,
      "mean_token_accuracy": 0.20356912016868592,
      "num_tokens": 48229082.0,
      "step": 23805
    },
    {
      "entropy": 6.018150377273559,
      "epoch": 2.0807480555798303,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045656983273899694,
      "loss": 5.3358,
      "mean_token_accuracy": 0.19954806715250015,
      "num_tokens": 48239158.0,
      "step": 23810
    },
    {
      "entropy": 5.978731155395508,
      "epoch": 2.0811850039325352,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004565514277886565,
      "loss": 5.2788,
      "mean_token_accuracy": 0.19992275536060333,
      "num_tokens": 48249663.0,
      "step": 23815
    },
    {
      "entropy": 5.991799163818359,
      "epoch": 2.0816219522852397,
      "grad_norm": 1.125,
      "learning_rate": 0.00045653301935606456,
      "loss": 5.3493,
      "mean_token_accuracy": 0.19129498153924943,
      "num_tokens": 48260319.0,
      "step": 23820
    },
    {
      "entropy": 5.988545036315918,
      "epoch": 2.0820589006379446,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045651460744157424,
      "loss": 5.3181,
      "mean_token_accuracy": 0.19308842420578004,
      "num_tokens": 48270107.0,
      "step": 23825
    },
    {
      "entropy": 5.942237710952758,
      "epoch": 2.082495848990649,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004564961920455387,
      "loss": 5.2918,
      "mean_token_accuracy": 0.20339775681495667,
      "num_tokens": 48280529.0,
      "step": 23830
    },
    {
      "entropy": 6.035485506057739,
      "epoch": 2.082932797343354,
      "grad_norm": 1.125,
      "learning_rate": 0.0004564777731683109,
      "loss": 5.3501,
      "mean_token_accuracy": 0.19646989107131957,
      "num_tokens": 48290672.0,
      "step": 23835
    },
    {
      "entropy": 6.074013900756836,
      "epoch": 2.083369745696059,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004564593508102445,
      "loss": 5.3616,
      "mean_token_accuracy": 0.19723010212182998,
      "num_tokens": 48301327.0,
      "step": 23840
    },
    {
      "entropy": 6.013739109039307,
      "epoch": 2.0838066940487634,
      "grad_norm": 1.125,
      "learning_rate": 0.0004564409249716925,
      "loss": 5.3306,
      "mean_token_accuracy": 0.19409627169370652,
      "num_tokens": 48310861.0,
      "step": 23845
    },
    {
      "entropy": 5.986939334869385,
      "epoch": 2.0842436424014683,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045642249565300866,
      "loss": 5.2667,
      "mean_token_accuracy": 0.1999209627509117,
      "num_tokens": 48320381.0,
      "step": 23850
    },
    {
      "entropy": 6.006286001205444,
      "epoch": 2.084680590754173,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045640406285454616,
      "loss": 5.2231,
      "mean_token_accuracy": 0.20856190770864486,
      "num_tokens": 48330566.0,
      "step": 23855
    },
    {
      "entropy": 5.956300449371338,
      "epoch": 2.0851175391068777,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045638562657665885,
      "loss": 5.2697,
      "mean_token_accuracy": 0.19988733381032944,
      "num_tokens": 48340773.0,
      "step": 23860
    },
    {
      "entropy": 5.9660279750823975,
      "epoch": 2.085554487459582,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045636718681969996,
      "loss": 5.2563,
      "mean_token_accuracy": 0.20337688475847243,
      "num_tokens": 48350972.0,
      "step": 23865
    },
    {
      "entropy": 6.036467027664185,
      "epoch": 2.085991435812287,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004563487435840235,
      "loss": 5.2618,
      "mean_token_accuracy": 0.20075617581605912,
      "num_tokens": 48360606.0,
      "step": 23870
    },
    {
      "entropy": 5.9941614151000975,
      "epoch": 2.0864283841649915,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045633029686998304,
      "loss": 5.3594,
      "mean_token_accuracy": 0.201019811630249,
      "num_tokens": 48371970.0,
      "step": 23875
    },
    {
      "entropy": 5.914083480834961,
      "epoch": 2.0868653325176965,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004563118466779326,
      "loss": 5.2396,
      "mean_token_accuracy": 0.2049369215965271,
      "num_tokens": 48382917.0,
      "step": 23880
    },
    {
      "entropy": 5.959013271331787,
      "epoch": 2.087302280870401,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045629339300822593,
      "loss": 5.2935,
      "mean_token_accuracy": 0.20374876856803895,
      "num_tokens": 48393428.0,
      "step": 23885
    },
    {
      "entropy": 6.1024669170379635,
      "epoch": 2.087739229223106,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004562749358612168,
      "loss": 5.3622,
      "mean_token_accuracy": 0.1867930054664612,
      "num_tokens": 48404616.0,
      "step": 23890
    },
    {
      "entropy": 5.982398891448975,
      "epoch": 2.0881761775758108,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004562564752372596,
      "loss": 5.1977,
      "mean_token_accuracy": 0.2111396998167038,
      "num_tokens": 48414753.0,
      "step": 23895
    },
    {
      "entropy": 5.9884387969970705,
      "epoch": 2.0886131259285152,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004562380111367082,
      "loss": 5.3143,
      "mean_token_accuracy": 0.19504892826080322,
      "num_tokens": 48424759.0,
      "step": 23900
    },
    {
      "entropy": 5.976785278320312,
      "epoch": 2.08905007428122,
      "grad_norm": 1.125,
      "learning_rate": 0.00045621954355991675,
      "loss": 5.2802,
      "mean_token_accuracy": 0.20730550289154054,
      "num_tokens": 48435158.0,
      "step": 23905
    },
    {
      "entropy": 5.9468724727630615,
      "epoch": 2.0894870226339246,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045620107250723946,
      "loss": 5.2539,
      "mean_token_accuracy": 0.20153072476387024,
      "num_tokens": 48445234.0,
      "step": 23910
    },
    {
      "entropy": 5.972158765792846,
      "epoch": 2.0899239709866295,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045618259797903073,
      "loss": 5.2523,
      "mean_token_accuracy": 0.20746831893920897,
      "num_tokens": 48456053.0,
      "step": 23915
    },
    {
      "entropy": 5.993629455566406,
      "epoch": 2.090360919339334,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004561641199756447,
      "loss": 5.2695,
      "mean_token_accuracy": 0.20250408500432968,
      "num_tokens": 48466293.0,
      "step": 23920
    },
    {
      "entropy": 5.926103687286377,
      "epoch": 2.090797867692039,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045614563849743605,
      "loss": 5.2349,
      "mean_token_accuracy": 0.20896509736776353,
      "num_tokens": 48476170.0,
      "step": 23925
    },
    {
      "entropy": 6.048404312133789,
      "epoch": 2.0912348160447434,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045612715354475907,
      "loss": 5.2736,
      "mean_token_accuracy": 0.1966005191206932,
      "num_tokens": 48485728.0,
      "step": 23930
    },
    {
      "entropy": 5.962942409515381,
      "epoch": 2.0916717643974483,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045610866511796845,
      "loss": 5.2652,
      "mean_token_accuracy": 0.2022349417209625,
      "num_tokens": 48496531.0,
      "step": 23935
    },
    {
      "entropy": 5.9755817413330075,
      "epoch": 2.0921087127501528,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045609017321741863,
      "loss": 5.2402,
      "mean_token_accuracy": 0.20457208454608916,
      "num_tokens": 48506006.0,
      "step": 23940
    },
    {
      "entropy": 5.975103378295898,
      "epoch": 2.0925456611028577,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004560716778434644,
      "loss": 5.3102,
      "mean_token_accuracy": 0.20323970168828964,
      "num_tokens": 48515902.0,
      "step": 23945
    },
    {
      "entropy": 5.973331785202026,
      "epoch": 2.0929826094555626,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045605317899646066,
      "loss": 5.3364,
      "mean_token_accuracy": 0.20117953568696975,
      "num_tokens": 48526263.0,
      "step": 23950
    },
    {
      "entropy": 5.969903945922852,
      "epoch": 2.093419557808267,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045603467667676195,
      "loss": 5.2593,
      "mean_token_accuracy": 0.2055455282330513,
      "num_tokens": 48537415.0,
      "step": 23955
    },
    {
      "entropy": 5.9920093536376955,
      "epoch": 2.093856506160972,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004560161708847233,
      "loss": 5.2859,
      "mean_token_accuracy": 0.19954429864883422,
      "num_tokens": 48548911.0,
      "step": 23960
    },
    {
      "entropy": 5.99702033996582,
      "epoch": 2.0942934545136764,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004559976616206996,
      "loss": 5.2873,
      "mean_token_accuracy": 0.20600546896457672,
      "num_tokens": 48560266.0,
      "step": 23965
    },
    {
      "entropy": 6.039608764648437,
      "epoch": 2.0947304028663813,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045597914888504597,
      "loss": 5.3682,
      "mean_token_accuracy": 0.19138525277376175,
      "num_tokens": 48570696.0,
      "step": 23970
    },
    {
      "entropy": 6.005309820175171,
      "epoch": 2.095167351219086,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045596063267811746,
      "loss": 5.2686,
      "mean_token_accuracy": 0.20150866508483886,
      "num_tokens": 48580725.0,
      "step": 23975
    },
    {
      "entropy": 5.956534481048584,
      "epoch": 2.0956042995717907,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004559421130002691,
      "loss": 5.3401,
      "mean_token_accuracy": 0.19627320021390915,
      "num_tokens": 48591968.0,
      "step": 23980
    },
    {
      "entropy": 5.948630237579346,
      "epoch": 2.096041247924495,
      "grad_norm": 1.125,
      "learning_rate": 0.0004559235898518563,
      "loss": 5.2631,
      "mean_token_accuracy": 0.20522929579019547,
      "num_tokens": 48602074.0,
      "step": 23985
    },
    {
      "entropy": 5.985196447372436,
      "epoch": 2.0964781962772,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004559050632332342,
      "loss": 5.2986,
      "mean_token_accuracy": 0.2014070436358452,
      "num_tokens": 48613323.0,
      "step": 23990
    },
    {
      "entropy": 6.050700998306274,
      "epoch": 2.0969151446299046,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004558865331447581,
      "loss": 5.3343,
      "mean_token_accuracy": 0.19677257239818574,
      "num_tokens": 48623386.0,
      "step": 23995
    },
    {
      "entropy": 5.916219472885132,
      "epoch": 2.0973520929826095,
      "grad_norm": 1.25,
      "learning_rate": 0.0004558679995867837,
      "loss": 5.2158,
      "mean_token_accuracy": 0.21357180029153824,
      "num_tokens": 48633379.0,
      "step": 24000
    },
    {
      "epoch": 2.0973520929826095,
      "eval_entropy": 5.70637079096263,
      "eval_loss": 5.342724323272705,
      "eval_mean_token_accuracy": 0.20755516394879409,
      "eval_num_tokens": 48633379.0,
      "eval_runtime": 21.9936,
      "eval_samples_per_second": 1682.946,
      "eval_steps_per_second": 210.38,
      "step": 24000
    },
    {
      "entropy": 5.948476886749267,
      "epoch": 2.097789041335314,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004558494625596662,
      "loss": 5.2677,
      "mean_token_accuracy": 0.20135288089513778,
      "num_tokens": 48643918.0,
      "step": 24005
    },
    {
      "entropy": 5.999200820922852,
      "epoch": 2.098225989688019,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004558309220637612,
      "loss": 5.3607,
      "mean_token_accuracy": 0.19238023310899735,
      "num_tokens": 48654534.0,
      "step": 24010
    },
    {
      "entropy": 5.987173366546631,
      "epoch": 2.098662938040724,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004558123780994244,
      "loss": 5.2649,
      "mean_token_accuracy": 0.20801436752080918,
      "num_tokens": 48664928.0,
      "step": 24015
    },
    {
      "entropy": 5.951588869094849,
      "epoch": 2.0990998863934283,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004557938306670115,
      "loss": 5.2338,
      "mean_token_accuracy": 0.20749987065792083,
      "num_tokens": 48675441.0,
      "step": 24020
    },
    {
      "entropy": 5.931415748596192,
      "epoch": 2.099536834746133,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045577527976687813,
      "loss": 5.2168,
      "mean_token_accuracy": 0.20139683336019515,
      "num_tokens": 48685173.0,
      "step": 24025
    },
    {
      "entropy": 5.9447705268859865,
      "epoch": 2.0999737830988376,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045575672539938014,
      "loss": 5.258,
      "mean_token_accuracy": 0.20311667025089264,
      "num_tokens": 48695539.0,
      "step": 24030
    },
    {
      "entropy": 5.964308834075927,
      "epoch": 2.1004107314515426,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045573816756487345,
      "loss": 5.3026,
      "mean_token_accuracy": 0.20116329491138457,
      "num_tokens": 48705977.0,
      "step": 24035
    },
    {
      "entropy": 5.9740485668182375,
      "epoch": 2.100847679804247,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004557196062637141,
      "loss": 5.2478,
      "mean_token_accuracy": 0.19610935896635057,
      "num_tokens": 48715924.0,
      "step": 24040
    },
    {
      "entropy": 6.008978080749512,
      "epoch": 2.101284628156952,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004557010414962579,
      "loss": 5.2916,
      "mean_token_accuracy": 0.1933056965470314,
      "num_tokens": 48727071.0,
      "step": 24045
    },
    {
      "entropy": 5.966390800476074,
      "epoch": 2.1017215765096564,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045568247326286103,
      "loss": 5.287,
      "mean_token_accuracy": 0.19596810191869735,
      "num_tokens": 48736959.0,
      "step": 24050
    },
    {
      "entropy": 5.938596391677857,
      "epoch": 2.1021585248623613,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045566390156387974,
      "loss": 5.3156,
      "mean_token_accuracy": 0.19665297567844392,
      "num_tokens": 48748419.0,
      "step": 24055
    },
    {
      "entropy": 5.997144508361816,
      "epoch": 2.102595473215066,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004556453263996701,
      "loss": 5.2715,
      "mean_token_accuracy": 0.20210712701082229,
      "num_tokens": 48758480.0,
      "step": 24060
    },
    {
      "entropy": 6.078323936462402,
      "epoch": 2.1030324215677707,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045562674777058844,
      "loss": 5.3795,
      "mean_token_accuracy": 0.19399535953998565,
      "num_tokens": 48768182.0,
      "step": 24065
    },
    {
      "entropy": 5.991031885147095,
      "epoch": 2.103469369920475,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004556081656769911,
      "loss": 5.2405,
      "mean_token_accuracy": 0.20374003946781158,
      "num_tokens": 48777729.0,
      "step": 24070
    },
    {
      "entropy": 6.0071799755096436,
      "epoch": 2.10390631827318,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045558958011923456,
      "loss": 5.261,
      "mean_token_accuracy": 0.20351017862558365,
      "num_tokens": 48787109.0,
      "step": 24075
    },
    {
      "entropy": 6.017028331756592,
      "epoch": 2.104343266625885,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004555709910976752,
      "loss": 5.3144,
      "mean_token_accuracy": 0.19395648390054704,
      "num_tokens": 48797152.0,
      "step": 24080
    },
    {
      "entropy": 5.979705333709717,
      "epoch": 2.1047802149785895,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00045555239861266964,
      "loss": 5.24,
      "mean_token_accuracy": 0.20510568469762802,
      "num_tokens": 48806288.0,
      "step": 24085
    },
    {
      "entropy": 5.881852436065674,
      "epoch": 2.1052171633312944,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004555338026645744,
      "loss": 5.2011,
      "mean_token_accuracy": 0.21550232470035552,
      "num_tokens": 48816025.0,
      "step": 24090
    },
    {
      "entropy": 5.947009801864624,
      "epoch": 2.105654111683999,
      "grad_norm": 1.125,
      "learning_rate": 0.0004555152032537463,
      "loss": 5.3227,
      "mean_token_accuracy": 0.19922163784503938,
      "num_tokens": 48827150.0,
      "step": 24095
    },
    {
      "entropy": 5.996705150604248,
      "epoch": 2.1060910600367038,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045549660038054205,
      "loss": 5.2555,
      "mean_token_accuracy": 0.19747764021158218,
      "num_tokens": 48837190.0,
      "step": 24100
    },
    {
      "entropy": 6.045696878433228,
      "epoch": 2.1065280083894082,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045547799404531826,
      "loss": 5.3208,
      "mean_token_accuracy": 0.19347089380025864,
      "num_tokens": 48847573.0,
      "step": 24105
    },
    {
      "entropy": 5.8955323696136475,
      "epoch": 2.106964956742113,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045545938424843214,
      "loss": 5.2203,
      "mean_token_accuracy": 0.201906980574131,
      "num_tokens": 48858966.0,
      "step": 24110
    },
    {
      "entropy": 5.976616764068604,
      "epoch": 2.1074019050948176,
      "grad_norm": 1.125,
      "learning_rate": 0.0004554407709902404,
      "loss": 5.299,
      "mean_token_accuracy": 0.19440461695194244,
      "num_tokens": 48869027.0,
      "step": 24115
    },
    {
      "entropy": 5.975147771835327,
      "epoch": 2.1078388534475225,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004554221542711001,
      "loss": 5.2841,
      "mean_token_accuracy": 0.2084512159228325,
      "num_tokens": 48879325.0,
      "step": 24120
    },
    {
      "entropy": 5.960361003875732,
      "epoch": 2.108275801800227,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045540353409136836,
      "loss": 5.1847,
      "mean_token_accuracy": 0.2127418264746666,
      "num_tokens": 48888939.0,
      "step": 24125
    },
    {
      "entropy": 5.952237844467163,
      "epoch": 2.108712750152932,
      "grad_norm": 1.125,
      "learning_rate": 0.00045538491045140234,
      "loss": 5.2785,
      "mean_token_accuracy": 0.19790051281452178,
      "num_tokens": 48898997.0,
      "step": 24130
    },
    {
      "entropy": 5.97725920677185,
      "epoch": 2.109149698505637,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045536628335155914,
      "loss": 5.2736,
      "mean_token_accuracy": 0.2051587775349617,
      "num_tokens": 48909063.0,
      "step": 24135
    },
    {
      "entropy": 5.902054214477539,
      "epoch": 2.1095866468583413,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045534765279219626,
      "loss": 5.1728,
      "mean_token_accuracy": 0.21292358934879302,
      "num_tokens": 48918694.0,
      "step": 24140
    },
    {
      "entropy": 6.000258111953736,
      "epoch": 2.110023595211046,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004553290187736708,
      "loss": 5.3574,
      "mean_token_accuracy": 0.19658932089805603,
      "num_tokens": 48928223.0,
      "step": 24145
    },
    {
      "entropy": 5.985530281066895,
      "epoch": 2.1104605435637507,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004553103812963402,
      "loss": 5.2758,
      "mean_token_accuracy": 0.20125031918287278,
      "num_tokens": 48938193.0,
      "step": 24150
    },
    {
      "entropy": 6.013857078552246,
      "epoch": 2.1108974919164556,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045529174036056217,
      "loss": 5.3623,
      "mean_token_accuracy": 0.18812162429094315,
      "num_tokens": 48949693.0,
      "step": 24155
    },
    {
      "entropy": 5.899487829208374,
      "epoch": 2.11133444026916,
      "grad_norm": 1.28125,
      "learning_rate": 0.00045527309596669397,
      "loss": 5.245,
      "mean_token_accuracy": 0.20490187853574754,
      "num_tokens": 48960199.0,
      "step": 24160
    },
    {
      "entropy": 5.99813060760498,
      "epoch": 2.111771388621865,
      "grad_norm": 1.125,
      "learning_rate": 0.00045525444811509336,
      "loss": 5.3243,
      "mean_token_accuracy": 0.19890498369932175,
      "num_tokens": 48969380.0,
      "step": 24165
    },
    {
      "entropy": 5.973342609405518,
      "epoch": 2.1122083369745694,
      "grad_norm": 1.15625,
      "learning_rate": 0.000455235796806118,
      "loss": 5.2196,
      "mean_token_accuracy": 0.20543456822633743,
      "num_tokens": 48978963.0,
      "step": 24170
    },
    {
      "entropy": 5.972637605667114,
      "epoch": 2.1126452853272744,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004552171420401256,
      "loss": 5.3739,
      "mean_token_accuracy": 0.19974524676799774,
      "num_tokens": 48989568.0,
      "step": 24175
    },
    {
      "entropy": 6.005236339569092,
      "epoch": 2.113082233679979,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045519848381747396,
      "loss": 5.3171,
      "mean_token_accuracy": 0.20017559230327606,
      "num_tokens": 49000684.0,
      "step": 24180
    },
    {
      "entropy": 6.045384740829467,
      "epoch": 2.1135191820326837,
      "grad_norm": 1.125,
      "learning_rate": 0.000455179822138521,
      "loss": 5.3627,
      "mean_token_accuracy": 0.19144562184810637,
      "num_tokens": 49010194.0,
      "step": 24185
    },
    {
      "entropy": 6.007035875320435,
      "epoch": 2.1139561303853887,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045516115700362466,
      "loss": 5.2721,
      "mean_token_accuracy": 0.2037833720445633,
      "num_tokens": 49020180.0,
      "step": 24190
    },
    {
      "entropy": 5.988349151611328,
      "epoch": 2.114393078738093,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045514248841314295,
      "loss": 5.3004,
      "mean_token_accuracy": 0.2047158882021904,
      "num_tokens": 49030665.0,
      "step": 24195
    },
    {
      "entropy": 6.033586120605468,
      "epoch": 2.114830027090798,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004551238163674338,
      "loss": 5.269,
      "mean_token_accuracy": 0.20375679880380632,
      "num_tokens": 49040197.0,
      "step": 24200
    },
    {
      "entropy": 6.004416370391846,
      "epoch": 2.1152669754435025,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045510514086685553,
      "loss": 5.2965,
      "mean_token_accuracy": 0.20025161057710647,
      "num_tokens": 49049996.0,
      "step": 24205
    },
    {
      "entropy": 5.994514846801758,
      "epoch": 2.1157039237962074,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004550864619117664,
      "loss": 5.3694,
      "mean_token_accuracy": 0.19138968735933304,
      "num_tokens": 49061691.0,
      "step": 24210
    },
    {
      "entropy": 6.0420920848846436,
      "epoch": 2.116140872148912,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004550677795025244,
      "loss": 5.4069,
      "mean_token_accuracy": 0.19028496444225312,
      "num_tokens": 49071605.0,
      "step": 24215
    },
    {
      "entropy": 5.984131908416748,
      "epoch": 2.116577820501617,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004550490936394881,
      "loss": 5.2307,
      "mean_token_accuracy": 0.20902906209230424,
      "num_tokens": 49081317.0,
      "step": 24220
    },
    {
      "entropy": 5.908083057403564,
      "epoch": 2.1170147688543213,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045503040432301586,
      "loss": 5.2447,
      "mean_token_accuracy": 0.20008406192064285,
      "num_tokens": 49090662.0,
      "step": 24225
    },
    {
      "entropy": 6.052128648757934,
      "epoch": 2.117451717207026,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004550117115534661,
      "loss": 5.3856,
      "mean_token_accuracy": 0.1914703443646431,
      "num_tokens": 49102046.0,
      "step": 24230
    },
    {
      "entropy": 6.073166751861573,
      "epoch": 2.1178886655597307,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045499301533119736,
      "loss": 5.2702,
      "mean_token_accuracy": 0.20426298826932907,
      "num_tokens": 49111047.0,
      "step": 24235
    },
    {
      "entropy": 5.976275110244751,
      "epoch": 2.1183256139124356,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004549743156565684,
      "loss": 5.2409,
      "mean_token_accuracy": 0.2048878252506256,
      "num_tokens": 49120606.0,
      "step": 24240
    },
    {
      "entropy": 5.960862064361573,
      "epoch": 2.1187625622651405,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004549556125299377,
      "loss": 5.3293,
      "mean_token_accuracy": 0.1985153943300247,
      "num_tokens": 49130981.0,
      "step": 24245
    },
    {
      "entropy": 6.008634042739868,
      "epoch": 2.119199510617845,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004549369059516639,
      "loss": 5.3639,
      "mean_token_accuracy": 0.19475304037332536,
      "num_tokens": 49140843.0,
      "step": 24250
    },
    {
      "entropy": 6.062795543670655,
      "epoch": 2.11963645897055,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004549181959221061,
      "loss": 5.3391,
      "mean_token_accuracy": 0.18970508873462677,
      "num_tokens": 49150492.0,
      "step": 24255
    },
    {
      "entropy": 5.972259569168091,
      "epoch": 2.1200734073232543,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045489948244162304,
      "loss": 5.2784,
      "mean_token_accuracy": 0.1997545838356018,
      "num_tokens": 49159973.0,
      "step": 24260
    },
    {
      "entropy": 6.047921752929687,
      "epoch": 2.1205103556759592,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004548807655105736,
      "loss": 5.3503,
      "mean_token_accuracy": 0.1986815944314003,
      "num_tokens": 49170711.0,
      "step": 24265
    },
    {
      "entropy": 6.028760290145874,
      "epoch": 2.1209473040286637,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045486204512931684,
      "loss": 5.3498,
      "mean_token_accuracy": 0.19593408703804016,
      "num_tokens": 49181121.0,
      "step": 24270
    },
    {
      "entropy": 5.861601829528809,
      "epoch": 2.1213842523813686,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004548433212982118,
      "loss": 5.1806,
      "mean_token_accuracy": 0.21345833986997603,
      "num_tokens": 49191528.0,
      "step": 24275
    },
    {
      "entropy": 5.973892641067505,
      "epoch": 2.121821200734073,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004548245940176176,
      "loss": 5.3133,
      "mean_token_accuracy": 0.1893392652273178,
      "num_tokens": 49202244.0,
      "step": 24280
    },
    {
      "entropy": 6.003243780136108,
      "epoch": 2.122258149086778,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004548058632878935,
      "loss": 5.3001,
      "mean_token_accuracy": 0.2023636668920517,
      "num_tokens": 49212964.0,
      "step": 24285
    },
    {
      "entropy": 6.022773599624633,
      "epoch": 2.1226950974394825,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045478712910939874,
      "loss": 5.2928,
      "mean_token_accuracy": 0.20139617323875428,
      "num_tokens": 49223478.0,
      "step": 24290
    },
    {
      "entropy": 5.982235383987427,
      "epoch": 2.1231320457921874,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045476839148249256,
      "loss": 5.3378,
      "mean_token_accuracy": 0.1967831388115883,
      "num_tokens": 49233933.0,
      "step": 24295
    },
    {
      "entropy": 6.00782527923584,
      "epoch": 2.123568994144892,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045474965040753453,
      "loss": 5.2333,
      "mean_token_accuracy": 0.20958125740289688,
      "num_tokens": 49243147.0,
      "step": 24300
    },
    {
      "entropy": 5.960189247131348,
      "epoch": 2.1240059424975968,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004547309058848839,
      "loss": 5.2314,
      "mean_token_accuracy": 0.20718719512224198,
      "num_tokens": 49252524.0,
      "step": 24305
    },
    {
      "entropy": 5.975556325912476,
      "epoch": 2.1244428908503017,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045471215791490047,
      "loss": 5.3227,
      "mean_token_accuracy": 0.20295278877019882,
      "num_tokens": 49261845.0,
      "step": 24310
    },
    {
      "entropy": 5.974141263961792,
      "epoch": 2.124879839203006,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045469340649794353,
      "loss": 5.2475,
      "mean_token_accuracy": 0.20254990309476853,
      "num_tokens": 49272437.0,
      "step": 24315
    },
    {
      "entropy": 5.987542724609375,
      "epoch": 2.125316787555711,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000454674651634373,
      "loss": 5.2215,
      "mean_token_accuracy": 0.2147807776927948,
      "num_tokens": 49281687.0,
      "step": 24320
    },
    {
      "entropy": 5.970121669769287,
      "epoch": 2.1257537359084155,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045465589332454844,
      "loss": 5.2029,
      "mean_token_accuracy": 0.2023642659187317,
      "num_tokens": 49291524.0,
      "step": 24325
    },
    {
      "entropy": 5.97864179611206,
      "epoch": 2.1261906842611205,
      "grad_norm": 1.125,
      "learning_rate": 0.00045463713156882976,
      "loss": 5.2926,
      "mean_token_accuracy": 0.20039637237787247,
      "num_tokens": 49303124.0,
      "step": 24330
    },
    {
      "entropy": 6.01218204498291,
      "epoch": 2.126627632613825,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045461836636757674,
      "loss": 5.2354,
      "mean_token_accuracy": 0.2069285675883293,
      "num_tokens": 49313050.0,
      "step": 24335
    },
    {
      "entropy": 5.934906959533691,
      "epoch": 2.12706458096653,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045459959772114935,
      "loss": 5.2328,
      "mean_token_accuracy": 0.20723114609718324,
      "num_tokens": 49321652.0,
      "step": 24340
    },
    {
      "entropy": 5.890747308731079,
      "epoch": 2.1275015293192343,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045458082562990754,
      "loss": 5.2304,
      "mean_token_accuracy": 0.20994477868080139,
      "num_tokens": 49331713.0,
      "step": 24345
    },
    {
      "entropy": 5.927346897125244,
      "epoch": 2.127938477671939,
      "grad_norm": 1.125,
      "learning_rate": 0.0004545620500942114,
      "loss": 5.2785,
      "mean_token_accuracy": 0.20163070559501647,
      "num_tokens": 49340370.0,
      "step": 24350
    },
    {
      "entropy": 6.011865520477295,
      "epoch": 2.1283754260246437,
      "grad_norm": 1.25,
      "learning_rate": 0.0004545432711144211,
      "loss": 5.3,
      "mean_token_accuracy": 0.1996089458465576,
      "num_tokens": 49349894.0,
      "step": 24355
    },
    {
      "entropy": 5.970243978500366,
      "epoch": 2.1288123743773486,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004545244886908967,
      "loss": 5.1896,
      "mean_token_accuracy": 0.21071835607290268,
      "num_tokens": 49359738.0,
      "step": 24360
    },
    {
      "entropy": 5.9314733982086185,
      "epoch": 2.129249322730053,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045450570282399856,
      "loss": 5.2843,
      "mean_token_accuracy": 0.2026189848780632,
      "num_tokens": 49370495.0,
      "step": 24365
    },
    {
      "entropy": 6.005051946640014,
      "epoch": 2.129686271082758,
      "grad_norm": 1.171875,
      "learning_rate": 0.000454486913514087,
      "loss": 5.2891,
      "mean_token_accuracy": 0.20458750128746034,
      "num_tokens": 49382191.0,
      "step": 24370
    },
    {
      "entropy": 6.028448009490967,
      "epoch": 2.130123219435463,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004544681207615224,
      "loss": 5.3321,
      "mean_token_accuracy": 0.1970113307237625,
      "num_tokens": 49391724.0,
      "step": 24375
    },
    {
      "entropy": 6.00970687866211,
      "epoch": 2.1305601677881674,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004544493245666652,
      "loss": 5.2649,
      "mean_token_accuracy": 0.20098417550325393,
      "num_tokens": 49401117.0,
      "step": 24380
    },
    {
      "entropy": 5.94898567199707,
      "epoch": 2.1309971161408723,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004544305249298759,
      "loss": 5.2524,
      "mean_token_accuracy": 0.20278632193803786,
      "num_tokens": 49410731.0,
      "step": 24385
    },
    {
      "entropy": 5.976503658294678,
      "epoch": 2.1314340644935768,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004544117218515152,
      "loss": 5.323,
      "mean_token_accuracy": 0.20137679278850557,
      "num_tokens": 49420373.0,
      "step": 24390
    },
    {
      "entropy": 5.98778886795044,
      "epoch": 2.1318710128462817,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045439291533194356,
      "loss": 5.2051,
      "mean_token_accuracy": 0.20642664134502411,
      "num_tokens": 49430021.0,
      "step": 24395
    },
    {
      "entropy": 5.955783462524414,
      "epoch": 2.132307961198986,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004543741053715218,
      "loss": 5.2138,
      "mean_token_accuracy": 0.20640814304351807,
      "num_tokens": 49439687.0,
      "step": 24400
    },
    {
      "entropy": 5.972831773757934,
      "epoch": 2.132744909551691,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004543552919706107,
      "loss": 5.3933,
      "mean_token_accuracy": 0.1924498125910759,
      "num_tokens": 49450320.0,
      "step": 24405
    },
    {
      "entropy": 5.967776536941528,
      "epoch": 2.1331818579043955,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004543364751295712,
      "loss": 5.2114,
      "mean_token_accuracy": 0.20429467856884004,
      "num_tokens": 49461272.0,
      "step": 24410
    },
    {
      "entropy": 6.016030550003052,
      "epoch": 2.1336188062571004,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045431765484876406,
      "loss": 5.3345,
      "mean_token_accuracy": 0.189921635389328,
      "num_tokens": 49473087.0,
      "step": 24415
    },
    {
      "entropy": 5.9144762516021725,
      "epoch": 2.134055754609805,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004542988311285505,
      "loss": 5.2261,
      "mean_token_accuracy": 0.2090281456708908,
      "num_tokens": 49483174.0,
      "step": 24420
    },
    {
      "entropy": 5.987238883972168,
      "epoch": 2.13449270296251,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045428000396929125,
      "loss": 5.2697,
      "mean_token_accuracy": 0.19994195252656938,
      "num_tokens": 49492267.0,
      "step": 24425
    },
    {
      "entropy": 5.960227918624878,
      "epoch": 2.1349296513152147,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045426117337134757,
      "loss": 5.2206,
      "mean_token_accuracy": 0.2042364865541458,
      "num_tokens": 49502338.0,
      "step": 24430
    },
    {
      "entropy": 5.966682243347168,
      "epoch": 2.135366599667919,
      "grad_norm": 1.40625,
      "learning_rate": 0.00045424233933508075,
      "loss": 5.2411,
      "mean_token_accuracy": 0.2119140699505806,
      "num_tokens": 49512991.0,
      "step": 24435
    },
    {
      "entropy": 5.960899782180786,
      "epoch": 2.135803548020624,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004542235018608519,
      "loss": 5.175,
      "mean_token_accuracy": 0.2150188758969307,
      "num_tokens": 49522224.0,
      "step": 24440
    },
    {
      "entropy": 5.950961685180664,
      "epoch": 2.1362404963733286,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045420466094902234,
      "loss": 5.3153,
      "mean_token_accuracy": 0.19898143112659455,
      "num_tokens": 49531982.0,
      "step": 24445
    },
    {
      "entropy": 6.005917501449585,
      "epoch": 2.1366774447260335,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004541858165999535,
      "loss": 5.2102,
      "mean_token_accuracy": 0.20760295689105987,
      "num_tokens": 49541487.0,
      "step": 24450
    },
    {
      "entropy": 5.990510892868042,
      "epoch": 2.137114393078738,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004541669688140068,
      "loss": 5.358,
      "mean_token_accuracy": 0.1909318447113037,
      "num_tokens": 49552052.0,
      "step": 24455
    },
    {
      "entropy": 5.960030174255371,
      "epoch": 2.137551341431443,
      "grad_norm": 1.125,
      "learning_rate": 0.0004541481175915438,
      "loss": 5.3039,
      "mean_token_accuracy": 0.19525336027145385,
      "num_tokens": 49561807.0,
      "step": 24460
    },
    {
      "entropy": 6.038075637817383,
      "epoch": 2.1379882897841473,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000454129262932926,
      "loss": 5.3256,
      "mean_token_accuracy": 0.1994215279817581,
      "num_tokens": 49570687.0,
      "step": 24465
    },
    {
      "entropy": 6.000555515289307,
      "epoch": 2.1384252381368523,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045411040483851506,
      "loss": 5.2924,
      "mean_token_accuracy": 0.20063157975673676,
      "num_tokens": 49581000.0,
      "step": 24470
    },
    {
      "entropy": 5.968602085113526,
      "epoch": 2.1388621864895567,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045409154330867274,
      "loss": 5.3345,
      "mean_token_accuracy": 0.19663477391004563,
      "num_tokens": 49590442.0,
      "step": 24475
    },
    {
      "entropy": 6.000287675857544,
      "epoch": 2.1392991348422616,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045407267834376075,
      "loss": 5.28,
      "mean_token_accuracy": 0.20048447251319884,
      "num_tokens": 49600068.0,
      "step": 24480
    },
    {
      "entropy": 6.041955900192261,
      "epoch": 2.1397360831949666,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000454053809944141,
      "loss": 5.306,
      "mean_token_accuracy": 0.2003210410475731,
      "num_tokens": 49610696.0,
      "step": 24485
    },
    {
      "entropy": 6.03258695602417,
      "epoch": 2.140173031547671,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004540349381101753,
      "loss": 5.3014,
      "mean_token_accuracy": 0.20439715832471847,
      "num_tokens": 49619489.0,
      "step": 24490
    },
    {
      "entropy": 6.004000616073609,
      "epoch": 2.140609979900376,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045401606284222563,
      "loss": 5.2968,
      "mean_token_accuracy": 0.19551292359828948,
      "num_tokens": 49629572.0,
      "step": 24495
    },
    {
      "entropy": 6.027883720397949,
      "epoch": 2.1410469282530804,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045399718414065415,
      "loss": 5.2474,
      "mean_token_accuracy": 0.200345915555954,
      "num_tokens": 49640035.0,
      "step": 24500
    },
    {
      "entropy": 5.9540962219238285,
      "epoch": 2.1414838766057853,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004539783020058229,
      "loss": 5.2576,
      "mean_token_accuracy": 0.20641531348228453,
      "num_tokens": 49649671.0,
      "step": 24505
    },
    {
      "entropy": 5.943430805206299,
      "epoch": 2.14192082495849,
      "grad_norm": 1.25,
      "learning_rate": 0.000453959416438094,
      "loss": 5.2926,
      "mean_token_accuracy": 0.20425975173711777,
      "num_tokens": 49659122.0,
      "step": 24510
    },
    {
      "entropy": 5.968747329711914,
      "epoch": 2.1423577733111947,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004539405274378297,
      "loss": 5.2405,
      "mean_token_accuracy": 0.20249828249216079,
      "num_tokens": 49669938.0,
      "step": 24515
    },
    {
      "entropy": 6.012790489196777,
      "epoch": 2.142794721663899,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004539216350053924,
      "loss": 5.2583,
      "mean_token_accuracy": 0.20460319221019746,
      "num_tokens": 49680783.0,
      "step": 24520
    },
    {
      "entropy": 5.992510890960693,
      "epoch": 2.143231670016604,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004539027391411443,
      "loss": 5.3255,
      "mean_token_accuracy": 0.19709013551473617,
      "num_tokens": 49691271.0,
      "step": 24525
    },
    {
      "entropy": 6.001753902435302,
      "epoch": 2.1436686183693086,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045388383984544794,
      "loss": 5.2988,
      "mean_token_accuracy": 0.19887901246547698,
      "num_tokens": 49702496.0,
      "step": 24530
    },
    {
      "entropy": 5.987268733978271,
      "epoch": 2.1441055667220135,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045386493711866577,
      "loss": 5.3088,
      "mean_token_accuracy": 0.19808293730020524,
      "num_tokens": 49712145.0,
      "step": 24535
    },
    {
      "entropy": 5.9787458896636965,
      "epoch": 2.1445425150747184,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045384603096116044,
      "loss": 5.2892,
      "mean_token_accuracy": 0.20414726436138153,
      "num_tokens": 49722243.0,
      "step": 24540
    },
    {
      "entropy": 6.0496257781982425,
      "epoch": 2.144979463427423,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004538271213732945,
      "loss": 5.3453,
      "mean_token_accuracy": 0.19548353403806687,
      "num_tokens": 49731804.0,
      "step": 24545
    },
    {
      "entropy": 6.018684530258179,
      "epoch": 2.1454164117801278,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045380820835543066,
      "loss": 5.2468,
      "mean_token_accuracy": 0.19833640307188033,
      "num_tokens": 49741275.0,
      "step": 24550
    },
    {
      "entropy": 5.941843557357788,
      "epoch": 2.1458533601328322,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004537892919079317,
      "loss": 5.233,
      "mean_token_accuracy": 0.20132200419902802,
      "num_tokens": 49751140.0,
      "step": 24555
    },
    {
      "entropy": 5.974742603302002,
      "epoch": 2.146290308485537,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045377037203116053,
      "loss": 5.2995,
      "mean_token_accuracy": 0.20266111493110656,
      "num_tokens": 49760747.0,
      "step": 24560
    },
    {
      "entropy": 6.032704162597656,
      "epoch": 2.1467272568382416,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045375144872547975,
      "loss": 5.315,
      "mean_token_accuracy": 0.19498913139104843,
      "num_tokens": 49771491.0,
      "step": 24565
    },
    {
      "entropy": 6.027551555633545,
      "epoch": 2.1471642051909465,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004537325219912527,
      "loss": 5.2852,
      "mean_token_accuracy": 0.205619415640831,
      "num_tokens": 49781523.0,
      "step": 24570
    },
    {
      "entropy": 6.015030002593994,
      "epoch": 2.147601153543651,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004537135918288421,
      "loss": 5.3626,
      "mean_token_accuracy": 0.19302117973566055,
      "num_tokens": 49791560.0,
      "step": 24575
    },
    {
      "entropy": 5.9680499076843265,
      "epoch": 2.148038101896356,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004536946582386112,
      "loss": 5.238,
      "mean_token_accuracy": 0.19804274141788483,
      "num_tokens": 49800588.0,
      "step": 24580
    },
    {
      "entropy": 5.958343648910523,
      "epoch": 2.1484750502490604,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045367572122092316,
      "loss": 5.2873,
      "mean_token_accuracy": 0.20094806253910064,
      "num_tokens": 49810381.0,
      "step": 24585
    },
    {
      "entropy": 5.999561405181884,
      "epoch": 2.1489119986017653,
      "grad_norm": 1.125,
      "learning_rate": 0.0004536567807761411,
      "loss": 5.3118,
      "mean_token_accuracy": 0.19956807643175126,
      "num_tokens": 49820931.0,
      "step": 24590
    },
    {
      "entropy": 5.979767608642578,
      "epoch": 2.14934894695447,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045363783690462845,
      "loss": 5.3339,
      "mean_token_accuracy": 0.19946508556604386,
      "num_tokens": 49831264.0,
      "step": 24595
    },
    {
      "entropy": 5.988063383102417,
      "epoch": 2.1497858953071747,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004536188896067483,
      "loss": 5.3222,
      "mean_token_accuracy": 0.20080913454294205,
      "num_tokens": 49841656.0,
      "step": 24600
    },
    {
      "entropy": 5.973040676116943,
      "epoch": 2.1502228436598796,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004535999388828645,
      "loss": 5.1786,
      "mean_token_accuracy": 0.21574768126010896,
      "num_tokens": 49851127.0,
      "step": 24605
    },
    {
      "entropy": 5.992838191986084,
      "epoch": 2.150659792012584,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004535809847333402,
      "loss": 5.2452,
      "mean_token_accuracy": 0.205888794362545,
      "num_tokens": 49860563.0,
      "step": 24610
    },
    {
      "entropy": 5.92738561630249,
      "epoch": 2.151096740365289,
      "grad_norm": 1.15625,
      "learning_rate": 0.000453562027158539,
      "loss": 5.2081,
      "mean_token_accuracy": 0.20450471490621566,
      "num_tokens": 49869741.0,
      "step": 24615
    },
    {
      "entropy": 6.010134649276734,
      "epoch": 2.1515336887179934,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004535430661588245,
      "loss": 5.3017,
      "mean_token_accuracy": 0.2007271856069565,
      "num_tokens": 49879664.0,
      "step": 24620
    },
    {
      "entropy": 5.988369655609131,
      "epoch": 2.1519706370706984,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045352410173456057,
      "loss": 5.2485,
      "mean_token_accuracy": 0.20682136714458466,
      "num_tokens": 49889322.0,
      "step": 24625
    },
    {
      "entropy": 5.9994570255279545,
      "epoch": 2.152407585423403,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045350513388611074,
      "loss": 5.3198,
      "mean_token_accuracy": 0.19514103531837462,
      "num_tokens": 49899172.0,
      "step": 24630
    },
    {
      "entropy": 5.9998682022094725,
      "epoch": 2.1528445337761077,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045348616261383897,
      "loss": 5.2792,
      "mean_token_accuracy": 0.20267634242773055,
      "num_tokens": 49909211.0,
      "step": 24635
    },
    {
      "entropy": 6.00224609375,
      "epoch": 2.153281482128812,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045346718791810896,
      "loss": 5.2841,
      "mean_token_accuracy": 0.2110156610608101,
      "num_tokens": 49918884.0,
      "step": 24640
    },
    {
      "entropy": 5.960064697265625,
      "epoch": 2.153718430481517,
      "grad_norm": 2.140625,
      "learning_rate": 0.00045344820979928484,
      "loss": 5.2496,
      "mean_token_accuracy": 0.21263094246387482,
      "num_tokens": 49929558.0,
      "step": 24645
    },
    {
      "entropy": 5.970487308502197,
      "epoch": 2.1541553788342216,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045342922825773057,
      "loss": 5.3075,
      "mean_token_accuracy": 0.20321640372276306,
      "num_tokens": 49939831.0,
      "step": 24650
    },
    {
      "entropy": 6.02623987197876,
      "epoch": 2.1545923271869265,
      "grad_norm": 1.28125,
      "learning_rate": 0.00045341024329381024,
      "loss": 5.2561,
      "mean_token_accuracy": 0.2026459664106369,
      "num_tokens": 49949690.0,
      "step": 24655
    },
    {
      "entropy": 6.0329869270324705,
      "epoch": 2.1550292755396314,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004533912549078879,
      "loss": 5.2955,
      "mean_token_accuracy": 0.19196694791316987,
      "num_tokens": 49959154.0,
      "step": 24660
    },
    {
      "entropy": 5.956251955032348,
      "epoch": 2.155466223892336,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045337226310032783,
      "loss": 5.3373,
      "mean_token_accuracy": 0.19950353354215622,
      "num_tokens": 49970185.0,
      "step": 24665
    },
    {
      "entropy": 6.024576091766358,
      "epoch": 2.155903172245041,
      "grad_norm": 1.125,
      "learning_rate": 0.0004533532678714943,
      "loss": 5.3629,
      "mean_token_accuracy": 0.19860281199216842,
      "num_tokens": 49980489.0,
      "step": 24670
    },
    {
      "entropy": 6.001733922958374,
      "epoch": 2.1563401205977453,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045333426922175156,
      "loss": 5.312,
      "mean_token_accuracy": 0.20223289132118225,
      "num_tokens": 49990038.0,
      "step": 24675
    },
    {
      "entropy": 6.0133130073547365,
      "epoch": 2.15677706895045,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004533152671514641,
      "loss": 5.2946,
      "mean_token_accuracy": 0.20060777366161348,
      "num_tokens": 49998888.0,
      "step": 24680
    },
    {
      "entropy": 5.907683038711548,
      "epoch": 2.1572140173031547,
      "grad_norm": 1.125,
      "learning_rate": 0.00045329626166099653,
      "loss": 5.2248,
      "mean_token_accuracy": 0.20888834744691848,
      "num_tokens": 50009349.0,
      "step": 24685
    },
    {
      "entropy": 6.031356143951416,
      "epoch": 2.1576509656558596,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045327725275071313,
      "loss": 5.3678,
      "mean_token_accuracy": 0.1948721557855606,
      "num_tokens": 50019626.0,
      "step": 24690
    },
    {
      "entropy": 5.975992918014526,
      "epoch": 2.158087914008564,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004532582404209786,
      "loss": 5.2472,
      "mean_token_accuracy": 0.2002884179353714,
      "num_tokens": 50029281.0,
      "step": 24695
    },
    {
      "entropy": 5.972818756103516,
      "epoch": 2.158524862361269,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004532392246721576,
      "loss": 5.2662,
      "mean_token_accuracy": 0.20598067045211793,
      "num_tokens": 50038700.0,
      "step": 24700
    },
    {
      "entropy": 5.98949499130249,
      "epoch": 2.1589618107139734,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004532202055046149,
      "loss": 5.245,
      "mean_token_accuracy": 0.19961014091968537,
      "num_tokens": 50048904.0,
      "step": 24705
    },
    {
      "entropy": 5.96722993850708,
      "epoch": 2.1593987590666783,
      "grad_norm": 1.25,
      "learning_rate": 0.0004532011829187153,
      "loss": 5.2842,
      "mean_token_accuracy": 0.19866975545883178,
      "num_tokens": 50059188.0,
      "step": 24710
    },
    {
      "entropy": 5.9413182735443115,
      "epoch": 2.159835707419383,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004531821569148236,
      "loss": 5.2458,
      "mean_token_accuracy": 0.20765288770198823,
      "num_tokens": 50069069.0,
      "step": 24715
    },
    {
      "entropy": 5.976128435134887,
      "epoch": 2.1602726557720877,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045316312749330473,
      "loss": 5.2811,
      "mean_token_accuracy": 0.1988314211368561,
      "num_tokens": 50078904.0,
      "step": 24720
    },
    {
      "entropy": 5.929215717315674,
      "epoch": 2.1607096041247926,
      "grad_norm": 1.34375,
      "learning_rate": 0.00045314409465452385,
      "loss": 5.2151,
      "mean_token_accuracy": 0.20767030268907546,
      "num_tokens": 50089897.0,
      "step": 24725
    },
    {
      "entropy": 5.991294002532959,
      "epoch": 2.161146552477497,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045312505839884573,
      "loss": 5.3218,
      "mean_token_accuracy": 0.19744386821985244,
      "num_tokens": 50100030.0,
      "step": 24730
    },
    {
      "entropy": 6.014613819122315,
      "epoch": 2.161583500830202,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045310601872663574,
      "loss": 5.2782,
      "mean_token_accuracy": 0.19920214116573334,
      "num_tokens": 50110043.0,
      "step": 24735
    },
    {
      "entropy": 5.977018690109253,
      "epoch": 2.1620204491829065,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045308697563825894,
      "loss": 5.2598,
      "mean_token_accuracy": 0.20718759596347808,
      "num_tokens": 50120081.0,
      "step": 24740
    },
    {
      "entropy": 5.9413896083831785,
      "epoch": 2.1624573975356114,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004530679291340807,
      "loss": 5.2739,
      "mean_token_accuracy": 0.20787776112556458,
      "num_tokens": 50129834.0,
      "step": 24745
    },
    {
      "entropy": 5.9613062858581545,
      "epoch": 2.162894345888316,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004530488792144662,
      "loss": 5.3152,
      "mean_token_accuracy": 0.19530644565820693,
      "num_tokens": 50140853.0,
      "step": 24750
    },
    {
      "entropy": 6.004955291748047,
      "epoch": 2.1633312942410208,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000453029825879781,
      "loss": 5.3808,
      "mean_token_accuracy": 0.20032420754432678,
      "num_tokens": 50151341.0,
      "step": 24755
    },
    {
      "entropy": 5.9946857452392575,
      "epoch": 2.1637682425937252,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045301076913039035,
      "loss": 5.2154,
      "mean_token_accuracy": 0.207000832259655,
      "num_tokens": 50160826.0,
      "step": 24760
    },
    {
      "entropy": 5.981246423721314,
      "epoch": 2.16420519094643,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004529917089666599,
      "loss": 5.254,
      "mean_token_accuracy": 0.20070670396089554,
      "num_tokens": 50170720.0,
      "step": 24765
    },
    {
      "entropy": 5.959883117675782,
      "epoch": 2.1646421392991346,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004529726453889552,
      "loss": 5.2886,
      "mean_token_accuracy": 0.20104080587625503,
      "num_tokens": 50180707.0,
      "step": 24770
    },
    {
      "entropy": 5.873113965988159,
      "epoch": 2.1650790876518395,
      "grad_norm": 1.125,
      "learning_rate": 0.00045295357839764193,
      "loss": 5.2223,
      "mean_token_accuracy": 0.21604481488466262,
      "num_tokens": 50190714.0,
      "step": 24775
    },
    {
      "entropy": 6.029397630691529,
      "epoch": 2.1655160360045445,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004529345079930858,
      "loss": 5.3268,
      "mean_token_accuracy": 0.20171083360910416,
      "num_tokens": 50201023.0,
      "step": 24780
    },
    {
      "entropy": 6.01704421043396,
      "epoch": 2.165952984357249,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004529154341756525,
      "loss": 5.2954,
      "mean_token_accuracy": 0.20035841912031174,
      "num_tokens": 50211064.0,
      "step": 24785
    },
    {
      "entropy": 6.043945837020874,
      "epoch": 2.166389932709954,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045289635694570796,
      "loss": 5.3387,
      "mean_token_accuracy": 0.1971260905265808,
      "num_tokens": 50220696.0,
      "step": 24790
    },
    {
      "entropy": 5.930728721618652,
      "epoch": 2.1668268810626583,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045287727630361814,
      "loss": 5.2854,
      "mean_token_accuracy": 0.20559623539447786,
      "num_tokens": 50230629.0,
      "step": 24795
    },
    {
      "entropy": 5.928195810317993,
      "epoch": 2.167263829415363,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045285819224974894,
      "loss": 5.2494,
      "mean_token_accuracy": 0.20787600576877593,
      "num_tokens": 50241042.0,
      "step": 24800
    },
    {
      "entropy": 5.977256917953492,
      "epoch": 2.1677007777680677,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004528391047844664,
      "loss": 5.2084,
      "mean_token_accuracy": 0.20777381658554078,
      "num_tokens": 50251731.0,
      "step": 24805
    },
    {
      "entropy": 5.986799001693726,
      "epoch": 2.1681377261207726,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004528200139081366,
      "loss": 5.3494,
      "mean_token_accuracy": 0.19539202749729156,
      "num_tokens": 50262893.0,
      "step": 24810
    },
    {
      "entropy": 6.01557354927063,
      "epoch": 2.168574674473477,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004528009196211258,
      "loss": 5.3934,
      "mean_token_accuracy": 0.19798990488052368,
      "num_tokens": 50273666.0,
      "step": 24815
    },
    {
      "entropy": 6.00052695274353,
      "epoch": 2.169011622826182,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004527818219238002,
      "loss": 5.251,
      "mean_token_accuracy": 0.20280583649873735,
      "num_tokens": 50283165.0,
      "step": 24820
    },
    {
      "entropy": 5.990179586410522,
      "epoch": 2.1694485711788865,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004527627208165261,
      "loss": 5.2354,
      "mean_token_accuracy": 0.20299988538026809,
      "num_tokens": 50291958.0,
      "step": 24825
    },
    {
      "entropy": 6.0079107761383055,
      "epoch": 2.1698855195315914,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045274361629966985,
      "loss": 5.3466,
      "mean_token_accuracy": 0.19664695858955383,
      "num_tokens": 50302944.0,
      "step": 24830
    },
    {
      "entropy": 5.981020927429199,
      "epoch": 2.1703224678842963,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045272450837359783,
      "loss": 5.2705,
      "mean_token_accuracy": 0.20685114860534667,
      "num_tokens": 50313088.0,
      "step": 24835
    },
    {
      "entropy": 5.991263341903687,
      "epoch": 2.1707594162370007,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045270539703867675,
      "loss": 5.2964,
      "mean_token_accuracy": 0.19730859398841857,
      "num_tokens": 50323121.0,
      "step": 24840
    },
    {
      "entropy": 5.994693756103516,
      "epoch": 2.1711963645897057,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000452686282295273,
      "loss": 5.2883,
      "mean_token_accuracy": 0.2052124783396721,
      "num_tokens": 50333091.0,
      "step": 24845
    },
    {
      "entropy": 5.965065622329712,
      "epoch": 2.17163331294241,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045266716414375324,
      "loss": 5.2183,
      "mean_token_accuracy": 0.20806433409452438,
      "num_tokens": 50342752.0,
      "step": 24850
    },
    {
      "entropy": 5.92313404083252,
      "epoch": 2.172070261295115,
      "grad_norm": 1.125,
      "learning_rate": 0.00045264804258448416,
      "loss": 5.2423,
      "mean_token_accuracy": 0.20364870131015778,
      "num_tokens": 50352213.0,
      "step": 24855
    },
    {
      "entropy": 5.9317467212677,
      "epoch": 2.1725072096478195,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004526289176178325,
      "loss": 5.2994,
      "mean_token_accuracy": 0.20067029893398286,
      "num_tokens": 50362586.0,
      "step": 24860
    },
    {
      "entropy": 5.945126008987427,
      "epoch": 2.1729441580005244,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004526097892441652,
      "loss": 5.191,
      "mean_token_accuracy": 0.20695283114910126,
      "num_tokens": 50372742.0,
      "step": 24865
    },
    {
      "entropy": 5.90234603881836,
      "epoch": 2.173381106353229,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045259065746384904,
      "loss": 5.1996,
      "mean_token_accuracy": 0.20818976014852525,
      "num_tokens": 50382857.0,
      "step": 24870
    },
    {
      "entropy": 6.0087216854095455,
      "epoch": 2.173818054705934,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000452571522277251,
      "loss": 5.2965,
      "mean_token_accuracy": 0.20335424244403838,
      "num_tokens": 50393638.0,
      "step": 24875
    },
    {
      "entropy": 5.967329216003418,
      "epoch": 2.1742550030586383,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004525523836847382,
      "loss": 5.2993,
      "mean_token_accuracy": 0.19960007816553116,
      "num_tokens": 50404928.0,
      "step": 24880
    },
    {
      "entropy": 5.952121305465698,
      "epoch": 2.174691951411343,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004525332416866776,
      "loss": 5.2077,
      "mean_token_accuracy": 0.20648495405912398,
      "num_tokens": 50414382.0,
      "step": 24885
    },
    {
      "entropy": 5.930358362197876,
      "epoch": 2.175128899764048,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004525140962834363,
      "loss": 5.2593,
      "mean_token_accuracy": 0.20583373606204985,
      "num_tokens": 50425203.0,
      "step": 24890
    },
    {
      "entropy": 6.072073984146118,
      "epoch": 2.1755658481167526,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004524949474753817,
      "loss": 5.4207,
      "mean_token_accuracy": 0.18636932522058486,
      "num_tokens": 50436395.0,
      "step": 24895
    },
    {
      "entropy": 5.951963281631469,
      "epoch": 2.1760027964694575,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004524757952628811,
      "loss": 5.272,
      "mean_token_accuracy": 0.20682541280984879,
      "num_tokens": 50446990.0,
      "step": 24900
    },
    {
      "entropy": 6.003990554809571,
      "epoch": 2.176439744822162,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045245663964630164,
      "loss": 5.3452,
      "mean_token_accuracy": 0.1900661900639534,
      "num_tokens": 50457964.0,
      "step": 24905
    },
    {
      "entropy": 6.051761722564697,
      "epoch": 2.176876693174867,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004524374806260108,
      "loss": 5.2691,
      "mean_token_accuracy": 0.20283871442079543,
      "num_tokens": 50467329.0,
      "step": 24910
    },
    {
      "entropy": 6.0116044044494625,
      "epoch": 2.1773136415275713,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004524183182023762,
      "loss": 5.2845,
      "mean_token_accuracy": 0.20525500327348709,
      "num_tokens": 50476852.0,
      "step": 24915
    },
    {
      "entropy": 5.947391223907471,
      "epoch": 2.1777505898802763,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045239915237576525,
      "loss": 5.2862,
      "mean_token_accuracy": 0.20616008937358857,
      "num_tokens": 50486207.0,
      "step": 24920
    },
    {
      "entropy": 5.9663480758667,
      "epoch": 2.1781875382329807,
      "grad_norm": 1.125,
      "learning_rate": 0.00045237998314654555,
      "loss": 5.3327,
      "mean_token_accuracy": 0.20211797058582306,
      "num_tokens": 50496457.0,
      "step": 24925
    },
    {
      "entropy": 6.052412605285644,
      "epoch": 2.1786244865856856,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004523608105150849,
      "loss": 5.3177,
      "mean_token_accuracy": 0.1909082144498825,
      "num_tokens": 50506430.0,
      "step": 24930
    },
    {
      "entropy": 5.996081399917602,
      "epoch": 2.17906143493839,
      "grad_norm": 1.078125,
      "learning_rate": 0.000452341634481751,
      "loss": 5.2359,
      "mean_token_accuracy": 0.19893730580806732,
      "num_tokens": 50516204.0,
      "step": 24935
    },
    {
      "entropy": 5.9653754234313965,
      "epoch": 2.179498383291095,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004523224550469115,
      "loss": 5.2926,
      "mean_token_accuracy": 0.2012923002243042,
      "num_tokens": 50525978.0,
      "step": 24940
    },
    {
      "entropy": 5.943626260757446,
      "epoch": 2.1799353316438,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004523032722109345,
      "loss": 5.2384,
      "mean_token_accuracy": 0.2041111782193184,
      "num_tokens": 50535579.0,
      "step": 24945
    },
    {
      "entropy": 5.985019254684448,
      "epoch": 2.1803722799965044,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004522840859741878,
      "loss": 5.2535,
      "mean_token_accuracy": 0.2093922421336174,
      "num_tokens": 50545765.0,
      "step": 24950
    },
    {
      "entropy": 6.0700188159942625,
      "epoch": 2.1808092283492093,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045226489633703944,
      "loss": 5.3834,
      "mean_token_accuracy": 0.19823744148015976,
      "num_tokens": 50556237.0,
      "step": 24955
    },
    {
      "entropy": 6.022035598754883,
      "epoch": 2.181246176701914,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045224570329985747,
      "loss": 5.3233,
      "mean_token_accuracy": 0.19936123937368394,
      "num_tokens": 50567720.0,
      "step": 24960
    },
    {
      "entropy": 6.001087427139282,
      "epoch": 2.1816831250546187,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045222650686301,
      "loss": 5.2474,
      "mean_token_accuracy": 0.20453023612499238,
      "num_tokens": 50577234.0,
      "step": 24965
    },
    {
      "entropy": 5.961590576171875,
      "epoch": 2.182120073407323,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004522073070268654,
      "loss": 5.318,
      "mean_token_accuracy": 0.20344094485044478,
      "num_tokens": 50587416.0,
      "step": 24970
    },
    {
      "entropy": 5.978789472579956,
      "epoch": 2.182557021760028,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004521881037917917,
      "loss": 5.1932,
      "mean_token_accuracy": 0.21895358264446257,
      "num_tokens": 50598317.0,
      "step": 24975
    },
    {
      "entropy": 6.02926344871521,
      "epoch": 2.1829939701127326,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004521688971581573,
      "loss": 5.3268,
      "mean_token_accuracy": 0.19442544579505922,
      "num_tokens": 50608507.0,
      "step": 24980
    },
    {
      "entropy": 5.953371572494507,
      "epoch": 2.1834309184654375,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045214968712633066,
      "loss": 5.267,
      "mean_token_accuracy": 0.20755741149187087,
      "num_tokens": 50618625.0,
      "step": 24985
    },
    {
      "entropy": 6.017662286758423,
      "epoch": 2.183867866818142,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004521304736966802,
      "loss": 5.2714,
      "mean_token_accuracy": 0.20593022108078002,
      "num_tokens": 50628696.0,
      "step": 24990
    },
    {
      "entropy": 6.0272540092468265,
      "epoch": 2.184304815170847,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045211125686957443,
      "loss": 5.3304,
      "mean_token_accuracy": 0.19830611497163772,
      "num_tokens": 50639313.0,
      "step": 24995
    },
    {
      "entropy": 5.988324403762817,
      "epoch": 2.1847417635235513,
      "grad_norm": 1.125,
      "learning_rate": 0.00045209203664538195,
      "loss": 5.2379,
      "mean_token_accuracy": 0.20508057922124862,
      "num_tokens": 50650715.0,
      "step": 25000
    },
    {
      "entropy": 5.979454708099365,
      "epoch": 2.1851787118762562,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004520728130244714,
      "loss": 5.2535,
      "mean_token_accuracy": 0.20115463584661483,
      "num_tokens": 50660224.0,
      "step": 25005
    },
    {
      "entropy": 6.000765132904053,
      "epoch": 2.185615660228961,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004520535860072115,
      "loss": 5.2629,
      "mean_token_accuracy": 0.20440563559532166,
      "num_tokens": 50670346.0,
      "step": 25010
    },
    {
      "entropy": 5.947476243972778,
      "epoch": 2.1860526085816656,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045203435559397113,
      "loss": 5.2852,
      "mean_token_accuracy": 0.1979052484035492,
      "num_tokens": 50680274.0,
      "step": 25015
    },
    {
      "entropy": 5.894933795928955,
      "epoch": 2.1864895569343705,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004520151217851189,
      "loss": 5.225,
      "mean_token_accuracy": 0.20967770665884017,
      "num_tokens": 50690360.0,
      "step": 25020
    },
    {
      "entropy": 5.967007875442505,
      "epoch": 2.186926505287075,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000451995884581024,
      "loss": 5.3046,
      "mean_token_accuracy": 0.20117821395397187,
      "num_tokens": 50700798.0,
      "step": 25025
    },
    {
      "entropy": 6.010558462142944,
      "epoch": 2.18736345363978,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004519766439820552,
      "loss": 5.3162,
      "mean_token_accuracy": 0.1973426267504692,
      "num_tokens": 50710120.0,
      "step": 25030
    },
    {
      "entropy": 6.031559753417969,
      "epoch": 2.1878004019924844,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004519573999885817,
      "loss": 5.3779,
      "mean_token_accuracy": 0.198570616543293,
      "num_tokens": 50719903.0,
      "step": 25035
    },
    {
      "entropy": 5.984703540802002,
      "epoch": 2.1882373503451893,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045193815260097253,
      "loss": 5.3182,
      "mean_token_accuracy": 0.20209044814109803,
      "num_tokens": 50730632.0,
      "step": 25040
    },
    {
      "entropy": 6.056163263320923,
      "epoch": 2.1886742986978938,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045191890181959696,
      "loss": 5.3212,
      "mean_token_accuracy": 0.1937124624848366,
      "num_tokens": 50741263.0,
      "step": 25045
    },
    {
      "entropy": 5.972694253921508,
      "epoch": 2.1891112470505987,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045189964764482407,
      "loss": 5.2596,
      "mean_token_accuracy": 0.2011943519115448,
      "num_tokens": 50751394.0,
      "step": 25050
    },
    {
      "entropy": 5.9762438297271725,
      "epoch": 2.189548195403303,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004518803900770232,
      "loss": 5.2817,
      "mean_token_accuracy": 0.20256065130233764,
      "num_tokens": 50762038.0,
      "step": 25055
    },
    {
      "entropy": 6.062092065811157,
      "epoch": 2.189985143756008,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045186112911656387,
      "loss": 5.3301,
      "mean_token_accuracy": 0.1977305829524994,
      "num_tokens": 50774063.0,
      "step": 25060
    },
    {
      "entropy": 6.051634645462036,
      "epoch": 2.1904220921087125,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004518418647638153,
      "loss": 5.339,
      "mean_token_accuracy": 0.19859573543071746,
      "num_tokens": 50784606.0,
      "step": 25065
    },
    {
      "entropy": 5.999018287658691,
      "epoch": 2.1908590404614174,
      "grad_norm": 1.25,
      "learning_rate": 0.0004518225970191471,
      "loss": 5.289,
      "mean_token_accuracy": 0.20028154850006102,
      "num_tokens": 50794892.0,
      "step": 25070
    },
    {
      "entropy": 5.950182771682739,
      "epoch": 2.1912959888141224,
      "grad_norm": 1.296875,
      "learning_rate": 0.00045180332588292896,
      "loss": 5.3158,
      "mean_token_accuracy": 0.20342053472995758,
      "num_tokens": 50805903.0,
      "step": 25075
    },
    {
      "entropy": 5.974352836608887,
      "epoch": 2.191732937166827,
      "grad_norm": 1.25,
      "learning_rate": 0.0004517840513555303,
      "loss": 5.1937,
      "mean_token_accuracy": 0.21271340996026994,
      "num_tokens": 50815339.0,
      "step": 25080
    },
    {
      "entropy": 6.0051473617553714,
      "epoch": 2.1921698855195317,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004517647734373209,
      "loss": 5.3184,
      "mean_token_accuracy": 0.19903482496738434,
      "num_tokens": 50825575.0,
      "step": 25085
    },
    {
      "entropy": 6.038287782669068,
      "epoch": 2.192606833872236,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045174549212867053,
      "loss": 5.3701,
      "mean_token_accuracy": 0.19301172941923142,
      "num_tokens": 50836106.0,
      "step": 25090
    },
    {
      "entropy": 6.021002721786499,
      "epoch": 2.193043782224941,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00045172620742994904,
      "loss": 5.3477,
      "mean_token_accuracy": 0.1996574729681015,
      "num_tokens": 50847350.0,
      "step": 25095
    },
    {
      "entropy": 5.997040224075318,
      "epoch": 2.1934807305776456,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004517069193415263,
      "loss": 5.219,
      "mean_token_accuracy": 0.2023195967078209,
      "num_tokens": 50857543.0,
      "step": 25100
    },
    {
      "entropy": 6.010534763336182,
      "epoch": 2.1939176789303505,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045168762786377214,
      "loss": 5.3228,
      "mean_token_accuracy": 0.19866396784782409,
      "num_tokens": 50868791.0,
      "step": 25105
    },
    {
      "entropy": 6.006191778182983,
      "epoch": 2.194354627283055,
      "grad_norm": 1.125,
      "learning_rate": 0.00045166833299705686,
      "loss": 5.2762,
      "mean_token_accuracy": 0.2025529697537422,
      "num_tokens": 50879244.0,
      "step": 25110
    },
    {
      "entropy": 6.010280704498291,
      "epoch": 2.19479157563576,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045164903474175026,
      "loss": 5.2916,
      "mean_token_accuracy": 0.1960947707295418,
      "num_tokens": 50888935.0,
      "step": 25115
    },
    {
      "entropy": 5.992211580276489,
      "epoch": 2.1952285239884644,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045162973309822264,
      "loss": 5.2891,
      "mean_token_accuracy": 0.20282559394836425,
      "num_tokens": 50898152.0,
      "step": 25120
    },
    {
      "entropy": 5.970980834960938,
      "epoch": 2.1956654723411693,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004516104280668442,
      "loss": 5.2913,
      "mean_token_accuracy": 0.19259297847747803,
      "num_tokens": 50908555.0,
      "step": 25125
    },
    {
      "entropy": 6.033415603637695,
      "epoch": 2.196102420693874,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045159111964798514,
      "loss": 5.4134,
      "mean_token_accuracy": 0.19205807000398636,
      "num_tokens": 50919078.0,
      "step": 25130
    },
    {
      "entropy": 6.03166127204895,
      "epoch": 2.1965393690465786,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045157180784201594,
      "loss": 5.2874,
      "mean_token_accuracy": 0.20091313123703003,
      "num_tokens": 50928879.0,
      "step": 25135
    },
    {
      "entropy": 5.999827909469604,
      "epoch": 2.1969763173992836,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000451552492649307,
      "loss": 5.303,
      "mean_token_accuracy": 0.2043769806623459,
      "num_tokens": 50939456.0,
      "step": 25140
    },
    {
      "entropy": 5.953745794296265,
      "epoch": 2.197413265751988,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045153317407022866,
      "loss": 5.2044,
      "mean_token_accuracy": 0.2134213700890541,
      "num_tokens": 50948789.0,
      "step": 25145
    },
    {
      "entropy": 5.95056037902832,
      "epoch": 2.197850214104693,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045151385210515153,
      "loss": 5.2101,
      "mean_token_accuracy": 0.20353306978940963,
      "num_tokens": 50958492.0,
      "step": 25150
    },
    {
      "entropy": 5.96875901222229,
      "epoch": 2.1982871624573974,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045149452675444625,
      "loss": 5.2871,
      "mean_token_accuracy": 0.20676224827766418,
      "num_tokens": 50968814.0,
      "step": 25155
    },
    {
      "entropy": 6.02197265625,
      "epoch": 2.1987241108101023,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045147519801848346,
      "loss": 5.2739,
      "mean_token_accuracy": 0.19699014872312545,
      "num_tokens": 50979680.0,
      "step": 25160
    },
    {
      "entropy": 5.972609519958496,
      "epoch": 2.199161059162807,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045145586589763394,
      "loss": 5.3415,
      "mean_token_accuracy": 0.19267342686653138,
      "num_tokens": 50990827.0,
      "step": 25165
    },
    {
      "entropy": 5.969339847564697,
      "epoch": 2.1995980075155117,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004514365303922684,
      "loss": 5.2634,
      "mean_token_accuracy": 0.20149824768304825,
      "num_tokens": 51001065.0,
      "step": 25170
    },
    {
      "entropy": 5.982540512084961,
      "epoch": 2.200034955868216,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045141719150275773,
      "loss": 5.2644,
      "mean_token_accuracy": 0.19874315112829208,
      "num_tokens": 51011348.0,
      "step": 25175
    },
    {
      "entropy": 6.051080799102783,
      "epoch": 2.200471904220921,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045139784922947296,
      "loss": 5.2564,
      "mean_token_accuracy": 0.20074936598539353,
      "num_tokens": 51022033.0,
      "step": 25180
    },
    {
      "entropy": 5.9793445587158205,
      "epoch": 2.200908852573626,
      "grad_norm": 1.140625,
      "learning_rate": 0.000451378503572785,
      "loss": 5.3178,
      "mean_token_accuracy": 0.19772969037294388,
      "num_tokens": 51032001.0,
      "step": 25185
    },
    {
      "entropy": 5.965353202819824,
      "epoch": 2.2013458009263305,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045135915453306487,
      "loss": 5.346,
      "mean_token_accuracy": 0.20046821981668472,
      "num_tokens": 51041227.0,
      "step": 25190
    },
    {
      "entropy": 5.912197160720825,
      "epoch": 2.2017827492790354,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045133980211068374,
      "loss": 5.1516,
      "mean_token_accuracy": 0.21722314804792403,
      "num_tokens": 51050642.0,
      "step": 25195
    },
    {
      "entropy": 5.996251630783081,
      "epoch": 2.20221969763174,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004513204463060129,
      "loss": 5.2342,
      "mean_token_accuracy": 0.2069326564669609,
      "num_tokens": 51060646.0,
      "step": 25200
    },
    {
      "entropy": 6.0047870635986325,
      "epoch": 2.2026566459844448,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004513010871194234,
      "loss": 5.3431,
      "mean_token_accuracy": 0.20534094870090486,
      "num_tokens": 51071762.0,
      "step": 25205
    },
    {
      "entropy": 5.976371383666992,
      "epoch": 2.2030935943371492,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045128172455128674,
      "loss": 5.3035,
      "mean_token_accuracy": 0.20422996580600739,
      "num_tokens": 51080802.0,
      "step": 25210
    },
    {
      "entropy": 5.96469259262085,
      "epoch": 2.203530542689854,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004512623586019742,
      "loss": 5.2965,
      "mean_token_accuracy": 0.20304841101169585,
      "num_tokens": 51091802.0,
      "step": 25215
    },
    {
      "entropy": 6.018932294845581,
      "epoch": 2.2039674910425586,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004512429892718573,
      "loss": 5.2729,
      "mean_token_accuracy": 0.20943634063005448,
      "num_tokens": 51100939.0,
      "step": 25220
    },
    {
      "entropy": 5.986187553405761,
      "epoch": 2.2044044393952635,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004512236165613074,
      "loss": 5.2414,
      "mean_token_accuracy": 0.2029520481824875,
      "num_tokens": 51110928.0,
      "step": 25225
    },
    {
      "entropy": 5.93722767829895,
      "epoch": 2.204841387747968,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045120424047069633,
      "loss": 5.2824,
      "mean_token_accuracy": 0.20102397203445435,
      "num_tokens": 51121126.0,
      "step": 25230
    },
    {
      "entropy": 6.031739330291748,
      "epoch": 2.205278336100673,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045118486100039553,
      "loss": 5.2167,
      "mean_token_accuracy": 0.20787395983934404,
      "num_tokens": 51130425.0,
      "step": 25235
    },
    {
      "entropy": 6.009005212783814,
      "epoch": 2.205715284453378,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004511654781507768,
      "loss": 5.252,
      "mean_token_accuracy": 0.20489756613969803,
      "num_tokens": 51139994.0,
      "step": 25240
    },
    {
      "entropy": 5.964145660400391,
      "epoch": 2.2061522328060823,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045114609192221196,
      "loss": 5.3008,
      "mean_token_accuracy": 0.19633836895227433,
      "num_tokens": 51149762.0,
      "step": 25245
    },
    {
      "entropy": 5.872948455810547,
      "epoch": 2.206589181158787,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004511267023150727,
      "loss": 5.284,
      "mean_token_accuracy": 0.21058533489704132,
      "num_tokens": 51161091.0,
      "step": 25250
    },
    {
      "entropy": 6.030750703811646,
      "epoch": 2.2070261295114917,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000451107309329731,
      "loss": 5.2822,
      "mean_token_accuracy": 0.1964680641889572,
      "num_tokens": 51170749.0,
      "step": 25255
    },
    {
      "entropy": 6.018183612823487,
      "epoch": 2.2074630778641966,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045108791296655877,
      "loss": 5.2722,
      "mean_token_accuracy": 0.19996788650751113,
      "num_tokens": 51179515.0,
      "step": 25260
    },
    {
      "entropy": 5.9684101104736325,
      "epoch": 2.207900026216901,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045106851322592817,
      "loss": 5.3203,
      "mean_token_accuracy": 0.20337124317884445,
      "num_tokens": 51189036.0,
      "step": 25265
    },
    {
      "entropy": 6.02715425491333,
      "epoch": 2.208336974569606,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004510491101082112,
      "loss": 5.3187,
      "mean_token_accuracy": 0.200125253200531,
      "num_tokens": 51199967.0,
      "step": 25270
    },
    {
      "entropy": 5.975202465057373,
      "epoch": 2.2087739229223105,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004510297036137801,
      "loss": 5.2819,
      "mean_token_accuracy": 0.20551377087831496,
      "num_tokens": 51210281.0,
      "step": 25275
    },
    {
      "entropy": 5.991356706619262,
      "epoch": 2.2092108712750154,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045101029374300703,
      "loss": 5.3306,
      "mean_token_accuracy": 0.20230619609355927,
      "num_tokens": 51220209.0,
      "step": 25280
    },
    {
      "entropy": 6.010218620300293,
      "epoch": 2.20964781962772,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004509908804962643,
      "loss": 5.302,
      "mean_token_accuracy": 0.20277660340070724,
      "num_tokens": 51231108.0,
      "step": 25285
    },
    {
      "entropy": 5.9315252780914305,
      "epoch": 2.2100847679804247,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045097146387392423,
      "loss": 5.2603,
      "mean_token_accuracy": 0.20345693975687026,
      "num_tokens": 51241316.0,
      "step": 25290
    },
    {
      "entropy": 5.957946014404297,
      "epoch": 2.210521716333129,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004509520438763593,
      "loss": 5.2863,
      "mean_token_accuracy": 0.20318453907966613,
      "num_tokens": 51250854.0,
      "step": 25295
    },
    {
      "entropy": 5.973989200592041,
      "epoch": 2.210958664685834,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045093262050394193,
      "loss": 5.2863,
      "mean_token_accuracy": 0.19954634010791777,
      "num_tokens": 51261001.0,
      "step": 25300
    },
    {
      "entropy": 5.9872862815856935,
      "epoch": 2.211395613038539,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045091319375704475,
      "loss": 5.3302,
      "mean_token_accuracy": 0.2040487840771675,
      "num_tokens": 51270815.0,
      "step": 25305
    },
    {
      "entropy": 6.029658699035645,
      "epoch": 2.2118325613912435,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045089376363604037,
      "loss": 5.2851,
      "mean_token_accuracy": 0.19787714779376983,
      "num_tokens": 51281370.0,
      "step": 25310
    },
    {
      "entropy": 5.994758653640747,
      "epoch": 2.2122695097439484,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045087433014130143,
      "loss": 5.2796,
      "mean_token_accuracy": 0.20998667031526566,
      "num_tokens": 51290147.0,
      "step": 25315
    },
    {
      "entropy": 6.019383907318115,
      "epoch": 2.212706458096653,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004508548932732006,
      "loss": 5.2161,
      "mean_token_accuracy": 0.21161465495824813,
      "num_tokens": 51299487.0,
      "step": 25320
    },
    {
      "entropy": 6.006859064102173,
      "epoch": 2.213143406449358,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004508354530321108,
      "loss": 5.2703,
      "mean_token_accuracy": 0.20177474319934846,
      "num_tokens": 51309425.0,
      "step": 25325
    },
    {
      "entropy": 5.974910879135132,
      "epoch": 2.2135803548020623,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004508160094184049,
      "loss": 5.2556,
      "mean_token_accuracy": 0.204846653342247,
      "num_tokens": 51319962.0,
      "step": 25330
    },
    {
      "entropy": 5.958051586151123,
      "epoch": 2.214017303154767,
      "grad_norm": 1.125,
      "learning_rate": 0.0004507965624324558,
      "loss": 5.2689,
      "mean_token_accuracy": 0.20965221971273423,
      "num_tokens": 51330892.0,
      "step": 25335
    },
    {
      "entropy": 6.011083126068115,
      "epoch": 2.2144542515074717,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045077711207463654,
      "loss": 5.2939,
      "mean_token_accuracy": 0.20400772243738174,
      "num_tokens": 51342783.0,
      "step": 25340
    },
    {
      "entropy": 6.080199241638184,
      "epoch": 2.2148911998601766,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045075765834532024,
      "loss": 5.3303,
      "mean_token_accuracy": 0.1953320324420929,
      "num_tokens": 51352890.0,
      "step": 25345
    },
    {
      "entropy": 6.027039575576782,
      "epoch": 2.215328148212881,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004507382012448798,
      "loss": 5.3475,
      "mean_token_accuracy": 0.19352855235338212,
      "num_tokens": 51363479.0,
      "step": 25350
    },
    {
      "entropy": 6.014998960494995,
      "epoch": 2.215765096565586,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004507187407736886,
      "loss": 5.2881,
      "mean_token_accuracy": 0.2030559003353119,
      "num_tokens": 51374304.0,
      "step": 25355
    },
    {
      "entropy": 5.993007612228394,
      "epoch": 2.2162020449182904,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004506992769321199,
      "loss": 5.3071,
      "mean_token_accuracy": 0.20199086964130403,
      "num_tokens": 51384649.0,
      "step": 25360
    },
    {
      "entropy": 5.994898462295533,
      "epoch": 2.2166389932709953,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045067980972054707,
      "loss": 5.2684,
      "mean_token_accuracy": 0.20396547466516496,
      "num_tokens": 51394582.0,
      "step": 25365
    },
    {
      "entropy": 5.981873655319214,
      "epoch": 2.2170759416237003,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004506603391393434,
      "loss": 5.2602,
      "mean_token_accuracy": 0.2050446331501007,
      "num_tokens": 51404618.0,
      "step": 25370
    },
    {
      "entropy": 5.985753345489502,
      "epoch": 2.2175128899764047,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004506408651888824,
      "loss": 5.2968,
      "mean_token_accuracy": 0.201531320810318,
      "num_tokens": 51414810.0,
      "step": 25375
    },
    {
      "entropy": 5.99204568862915,
      "epoch": 2.2179498383291096,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004506213878695375,
      "loss": 5.3633,
      "mean_token_accuracy": 0.1980898216366768,
      "num_tokens": 51426118.0,
      "step": 25380
    },
    {
      "entropy": 6.005024480819702,
      "epoch": 2.218386786681814,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045060190718168244,
      "loss": 5.2632,
      "mean_token_accuracy": 0.20656161308288573,
      "num_tokens": 51435260.0,
      "step": 25385
    },
    {
      "entropy": 5.97500433921814,
      "epoch": 2.218823735034519,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004505824231256907,
      "loss": 5.2596,
      "mean_token_accuracy": 0.2074566289782524,
      "num_tokens": 51444778.0,
      "step": 25390
    },
    {
      "entropy": 5.946860599517822,
      "epoch": 2.2192606833872235,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004505629357019361,
      "loss": 5.2506,
      "mean_token_accuracy": 0.20769355893135072,
      "num_tokens": 51456030.0,
      "step": 25395
    },
    {
      "entropy": 6.035029315948487,
      "epoch": 2.2196976317399284,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004505434449107925,
      "loss": 5.284,
      "mean_token_accuracy": 0.20077827870845794,
      "num_tokens": 51465836.0,
      "step": 25400
    },
    {
      "entropy": 5.936590576171875,
      "epoch": 2.220134580092633,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004505239507526335,
      "loss": 5.2238,
      "mean_token_accuracy": 0.2082556426525116,
      "num_tokens": 51475330.0,
      "step": 25405
    },
    {
      "entropy": 5.938610982894898,
      "epoch": 2.220571528445338,
      "grad_norm": 1.125,
      "learning_rate": 0.00045050445322783326,
      "loss": 5.2371,
      "mean_token_accuracy": 0.20780044049024582,
      "num_tokens": 51486646.0,
      "step": 25410
    },
    {
      "entropy": 5.957927799224853,
      "epoch": 2.2210084767980423,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004504849523367656,
      "loss": 5.2563,
      "mean_token_accuracy": 0.20362223088741302,
      "num_tokens": 51496878.0,
      "step": 25415
    },
    {
      "entropy": 6.0101292610168455,
      "epoch": 2.221445425150747,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004504654480798046,
      "loss": 5.3516,
      "mean_token_accuracy": 0.19800446629524232,
      "num_tokens": 51507092.0,
      "step": 25420
    },
    {
      "entropy": 5.988156032562256,
      "epoch": 2.221882373503452,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004504459404573243,
      "loss": 5.2742,
      "mean_token_accuracy": 0.20291563868522644,
      "num_tokens": 51517521.0,
      "step": 25425
    },
    {
      "entropy": 6.017628908157349,
      "epoch": 2.2223193218561565,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004504264294696991,
      "loss": 5.3483,
      "mean_token_accuracy": 0.1981750637292862,
      "num_tokens": 51527902.0,
      "step": 25430
    },
    {
      "entropy": 5.977133417129517,
      "epoch": 2.2227562702088615,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004504069151173029,
      "loss": 5.302,
      "mean_token_accuracy": 0.20484108179807664,
      "num_tokens": 51539062.0,
      "step": 25435
    },
    {
      "entropy": 6.001537036895752,
      "epoch": 2.223193218561566,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004503873974005102,
      "loss": 5.1992,
      "mean_token_accuracy": 0.2193001389503479,
      "num_tokens": 51548938.0,
      "step": 25440
    },
    {
      "entropy": 5.947586107254028,
      "epoch": 2.223630166914271,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004503678763196953,
      "loss": 5.2967,
      "mean_token_accuracy": 0.20248588770627976,
      "num_tokens": 51559330.0,
      "step": 25445
    },
    {
      "entropy": 5.954498720169068,
      "epoch": 2.2240671152669753,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004503483518752327,
      "loss": 5.3047,
      "mean_token_accuracy": 0.20180090367794037,
      "num_tokens": 51569617.0,
      "step": 25450
    },
    {
      "entropy": 5.950854349136352,
      "epoch": 2.2245040636196802,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004503288240674968,
      "loss": 5.2741,
      "mean_token_accuracy": 0.19765046685934068,
      "num_tokens": 51579545.0,
      "step": 25455
    },
    {
      "entropy": 5.997538280487061,
      "epoch": 2.2249410119723847,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004503092928968622,
      "loss": 5.2749,
      "mean_token_accuracy": 0.20021436661481856,
      "num_tokens": 51589375.0,
      "step": 25460
    },
    {
      "entropy": 5.976735925674438,
      "epoch": 2.2253779603250896,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004502897583637035,
      "loss": 5.26,
      "mean_token_accuracy": 0.20506629347801208,
      "num_tokens": 51600586.0,
      "step": 25465
    },
    {
      "entropy": 5.987039947509766,
      "epoch": 2.225814908677794,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045027022046839534,
      "loss": 5.2187,
      "mean_token_accuracy": 0.20402833372354506,
      "num_tokens": 51609568.0,
      "step": 25470
    },
    {
      "entropy": 5.990793514251709,
      "epoch": 2.226251857030499,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004502506792113126,
      "loss": 5.2309,
      "mean_token_accuracy": 0.21199878454208373,
      "num_tokens": 51619450.0,
      "step": 25475
    },
    {
      "entropy": 5.923298501968384,
      "epoch": 2.226688805383204,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004502311345928299,
      "loss": 5.2404,
      "mean_token_accuracy": 0.205440653860569,
      "num_tokens": 51630528.0,
      "step": 25480
    },
    {
      "entropy": 5.947795629501343,
      "epoch": 2.2271257537359084,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045021158661332227,
      "loss": 5.2643,
      "mean_token_accuracy": 0.2041957661509514,
      "num_tokens": 51640204.0,
      "step": 25485
    },
    {
      "entropy": 6.052309513092041,
      "epoch": 2.2275627020886133,
      "grad_norm": 1.125,
      "learning_rate": 0.0004501920352731646,
      "loss": 5.4123,
      "mean_token_accuracy": 0.19460852146148683,
      "num_tokens": 51649531.0,
      "step": 25490
    },
    {
      "entropy": 6.06924729347229,
      "epoch": 2.2279996504413178,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004501724805727319,
      "loss": 5.3456,
      "mean_token_accuracy": 0.19539033323526384,
      "num_tokens": 51660943.0,
      "step": 25495
    },
    {
      "entropy": 5.978160762786866,
      "epoch": 2.2284365987940227,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004501529225123992,
      "loss": 5.2101,
      "mean_token_accuracy": 0.2139250785112381,
      "num_tokens": 51670606.0,
      "step": 25500
    },
    {
      "entropy": 5.964802169799805,
      "epoch": 2.228873547146727,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045013336109254174,
      "loss": 5.2716,
      "mean_token_accuracy": 0.20278995484113693,
      "num_tokens": 51681634.0,
      "step": 25505
    },
    {
      "entropy": 6.023851442337036,
      "epoch": 2.229310495499432,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045011379631353457,
      "loss": 5.3092,
      "mean_token_accuracy": 0.1998673066496849,
      "num_tokens": 51691579.0,
      "step": 25510
    },
    {
      "entropy": 5.930956268310547,
      "epoch": 2.2297474438521365,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045009422817575304,
      "loss": 5.1949,
      "mean_token_accuracy": 0.21773786842823029,
      "num_tokens": 51700739.0,
      "step": 25515
    },
    {
      "entropy": 6.001932191848755,
      "epoch": 2.2301843922048414,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004500746566795725,
      "loss": 5.2995,
      "mean_token_accuracy": 0.20415406227111815,
      "num_tokens": 51711905.0,
      "step": 25520
    },
    {
      "entropy": 6.005658340454102,
      "epoch": 2.230621340557546,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045005508182536824,
      "loss": 5.2197,
      "mean_token_accuracy": 0.203158962726593,
      "num_tokens": 51722020.0,
      "step": 25525
    },
    {
      "entropy": 5.996915435791015,
      "epoch": 2.231058288910251,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004500355036135158,
      "loss": 5.3224,
      "mean_token_accuracy": 0.19707472771406173,
      "num_tokens": 51731763.0,
      "step": 25530
    },
    {
      "entropy": 6.0075780868530275,
      "epoch": 2.2314952372629557,
      "grad_norm": 1.8203125,
      "learning_rate": 0.00045001592204439077,
      "loss": 5.3097,
      "mean_token_accuracy": 0.2004112184047699,
      "num_tokens": 51741425.0,
      "step": 25535
    },
    {
      "entropy": 6.006615018844604,
      "epoch": 2.23193218561566,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004499963371183686,
      "loss": 5.3268,
      "mean_token_accuracy": 0.19427379816770554,
      "num_tokens": 51751788.0,
      "step": 25540
    },
    {
      "entropy": 6.008364343643189,
      "epoch": 2.232369133968365,
      "grad_norm": 1.125,
      "learning_rate": 0.00044997674883582494,
      "loss": 5.3037,
      "mean_token_accuracy": 0.19809815883636475,
      "num_tokens": 51763205.0,
      "step": 25545
    },
    {
      "entropy": 6.1081784725189205,
      "epoch": 2.2328060823210696,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004499571571971356,
      "loss": 5.2776,
      "mean_token_accuracy": 0.20329545587301254,
      "num_tokens": 51772817.0,
      "step": 25550
    },
    {
      "entropy": 6.041207504272461,
      "epoch": 2.2332430306737745,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004499375622026763,
      "loss": 5.3249,
      "mean_token_accuracy": 0.1993855655193329,
      "num_tokens": 51781596.0,
      "step": 25555
    },
    {
      "entropy": 5.928060483932495,
      "epoch": 2.233679979026479,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004499179638528229,
      "loss": 5.2363,
      "mean_token_accuracy": 0.20944161862134933,
      "num_tokens": 51792452.0,
      "step": 25560
    },
    {
      "entropy": 6.014696741104126,
      "epoch": 2.234116927379184,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004498983621479513,
      "loss": 5.314,
      "mean_token_accuracy": 0.2001222476363182,
      "num_tokens": 51802554.0,
      "step": 25565
    },
    {
      "entropy": 6.04367299079895,
      "epoch": 2.2345538757318884,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004498787570884374,
      "loss": 5.2952,
      "mean_token_accuracy": 0.19833528399467468,
      "num_tokens": 51812855.0,
      "step": 25570
    },
    {
      "entropy": 5.980140686035156,
      "epoch": 2.2349908240845933,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004498591486746573,
      "loss": 5.3061,
      "mean_token_accuracy": 0.2002284824848175,
      "num_tokens": 51822306.0,
      "step": 25575
    },
    {
      "entropy": 6.017872619628906,
      "epoch": 2.2354277724372977,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044983953690698724,
      "loss": 5.3239,
      "mean_token_accuracy": 0.2023647829890251,
      "num_tokens": 51832209.0,
      "step": 25580
    },
    {
      "entropy": 5.989811182022095,
      "epoch": 2.2358647207900026,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004498199217858032,
      "loss": 5.2997,
      "mean_token_accuracy": 0.20347013771533967,
      "num_tokens": 51841794.0,
      "step": 25585
    },
    {
      "entropy": 5.9941479682922365,
      "epoch": 2.2363016691427076,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044980030331148144,
      "loss": 5.3085,
      "mean_token_accuracy": 0.19830113649368286,
      "num_tokens": 51851973.0,
      "step": 25590
    },
    {
      "entropy": 6.030292749404907,
      "epoch": 2.236738617495412,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044978068148439826,
      "loss": 5.3274,
      "mean_token_accuracy": 0.19767968207597733,
      "num_tokens": 51862173.0,
      "step": 25595
    },
    {
      "entropy": 6.02427830696106,
      "epoch": 2.237175565848117,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044976105630493,
      "loss": 5.3098,
      "mean_token_accuracy": 0.20223543643951417,
      "num_tokens": 51872960.0,
      "step": 25600
    },
    {
      "entropy": 5.968071556091308,
      "epoch": 2.2376125142008214,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004497414277734532,
      "loss": 5.2582,
      "mean_token_accuracy": 0.2004844442009926,
      "num_tokens": 51883911.0,
      "step": 25605
    },
    {
      "entropy": 6.008838129043579,
      "epoch": 2.2380494625535263,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044972179589034407,
      "loss": 5.2799,
      "mean_token_accuracy": 0.199517884850502,
      "num_tokens": 51893982.0,
      "step": 25610
    },
    {
      "entropy": 5.993571996688843,
      "epoch": 2.238486410906231,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00044970216065597947,
      "loss": 5.1882,
      "mean_token_accuracy": 0.21238960921764374,
      "num_tokens": 51903808.0,
      "step": 25615
    },
    {
      "entropy": 5.929039525985718,
      "epoch": 2.2389233592589357,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044968252207073584,
      "loss": 5.2218,
      "mean_token_accuracy": 0.21100116819143294,
      "num_tokens": 51912918.0,
      "step": 25620
    },
    {
      "entropy": 5.983486175537109,
      "epoch": 2.23936030761164,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044966288013499,
      "loss": 5.3176,
      "mean_token_accuracy": 0.20117924213409424,
      "num_tokens": 51923894.0,
      "step": 25625
    },
    {
      "entropy": 6.006154632568359,
      "epoch": 2.239797255964345,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004496432348491185,
      "loss": 5.3186,
      "mean_token_accuracy": 0.19669201821088791,
      "num_tokens": 51934569.0,
      "step": 25630
    },
    {
      "entropy": 6.00641655921936,
      "epoch": 2.2402342043170496,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004496235862134982,
      "loss": 5.299,
      "mean_token_accuracy": 0.1915728196501732,
      "num_tokens": 51944678.0,
      "step": 25635
    },
    {
      "entropy": 6.0540080070495605,
      "epoch": 2.2406711526697545,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044960393422850616,
      "loss": 5.3835,
      "mean_token_accuracy": 0.18910786956548692,
      "num_tokens": 51954785.0,
      "step": 25640
    },
    {
      "entropy": 5.981863689422608,
      "epoch": 2.241108101022459,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004495842788945191,
      "loss": 5.2159,
      "mean_token_accuracy": 0.2051376387476921,
      "num_tokens": 51964427.0,
      "step": 25645
    },
    {
      "entropy": 6.026783800125122,
      "epoch": 2.241545049375164,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044956462021191406,
      "loss": 5.4338,
      "mean_token_accuracy": 0.1912957489490509,
      "num_tokens": 51974367.0,
      "step": 25650
    },
    {
      "entropy": 6.001294898986816,
      "epoch": 2.2419819977278688,
      "grad_norm": 1.125,
      "learning_rate": 0.00044954495818106814,
      "loss": 5.2496,
      "mean_token_accuracy": 0.20351927876472473,
      "num_tokens": 51984935.0,
      "step": 25655
    },
    {
      "entropy": 5.995336580276489,
      "epoch": 2.2424189460805732,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004495252928023585,
      "loss": 5.2809,
      "mean_token_accuracy": 0.20406294465065003,
      "num_tokens": 51995132.0,
      "step": 25660
    },
    {
      "entropy": 5.9943763256073,
      "epoch": 2.242855894433278,
      "grad_norm": 1.25,
      "learning_rate": 0.0004495056240761623,
      "loss": 5.3328,
      "mean_token_accuracy": 0.19841758757829667,
      "num_tokens": 52005828.0,
      "step": 25665
    },
    {
      "entropy": 5.925562381744385,
      "epoch": 2.2432928427859826,
      "grad_norm": 1.28125,
      "learning_rate": 0.00044948595200285677,
      "loss": 5.1745,
      "mean_token_accuracy": 0.2117704838514328,
      "num_tokens": 52015381.0,
      "step": 25670
    },
    {
      "entropy": 5.969022035598755,
      "epoch": 2.2437297911386875,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044946627658281924,
      "loss": 5.3126,
      "mean_token_accuracy": 0.19968122988939285,
      "num_tokens": 52025699.0,
      "step": 25675
    },
    {
      "entropy": 6.01761302947998,
      "epoch": 2.244166739491392,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004494465978164271,
      "loss": 5.2686,
      "mean_token_accuracy": 0.20824065655469895,
      "num_tokens": 52035903.0,
      "step": 25680
    },
    {
      "entropy": 5.9868498802185055,
      "epoch": 2.244603687844097,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004494269157040579,
      "loss": 5.2474,
      "mean_token_accuracy": 0.20900060385465621,
      "num_tokens": 52046039.0,
      "step": 25685
    },
    {
      "entropy": 6.037782287597656,
      "epoch": 2.2450406361968014,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044940723024608893,
      "loss": 5.3281,
      "mean_token_accuracy": 0.19393403977155685,
      "num_tokens": 52057146.0,
      "step": 25690
    },
    {
      "entropy": 5.984255409240722,
      "epoch": 2.2454775845495063,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044938754144289793,
      "loss": 5.2854,
      "mean_token_accuracy": 0.20470805913209916,
      "num_tokens": 52067355.0,
      "step": 25695
    },
    {
      "entropy": 6.007954549789429,
      "epoch": 2.2459145329022108,
      "grad_norm": 1.125,
      "learning_rate": 0.00044936784929486257,
      "loss": 5.2906,
      "mean_token_accuracy": 0.2007986932992935,
      "num_tokens": 52077062.0,
      "step": 25700
    },
    {
      "entropy": 5.962127733230591,
      "epoch": 2.2463514812549157,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004493481538023604,
      "loss": 5.2426,
      "mean_token_accuracy": 0.20367797315120698,
      "num_tokens": 52087918.0,
      "step": 25705
    },
    {
      "entropy": 5.956518077850342,
      "epoch": 2.24678842960762,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044932845496576936,
      "loss": 5.2922,
      "mean_token_accuracy": 0.19732857644557952,
      "num_tokens": 52097283.0,
      "step": 25710
    },
    {
      "entropy": 5.993821477890014,
      "epoch": 2.247225377960325,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004493087527854672,
      "loss": 5.3256,
      "mean_token_accuracy": 0.20067543983459474,
      "num_tokens": 52109037.0,
      "step": 25715
    },
    {
      "entropy": 6.035325622558593,
      "epoch": 2.24766232631303,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004492890472618318,
      "loss": 5.2856,
      "mean_token_accuracy": 0.2036233589053154,
      "num_tokens": 52119809.0,
      "step": 25720
    },
    {
      "entropy": 5.960057353973388,
      "epoch": 2.2480992746657344,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044926933839524116,
      "loss": 5.2209,
      "mean_token_accuracy": 0.20789601951837539,
      "num_tokens": 52129641.0,
      "step": 25725
    },
    {
      "entropy": 5.951611518859863,
      "epoch": 2.2485362230184394,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044924962618607324,
      "loss": 5.2378,
      "mean_token_accuracy": 0.20806556046009064,
      "num_tokens": 52139701.0,
      "step": 25730
    },
    {
      "entropy": 5.969723272323608,
      "epoch": 2.248973171371144,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004492299106347062,
      "loss": 5.2173,
      "mean_token_accuracy": 0.20986126661300658,
      "num_tokens": 52148402.0,
      "step": 25735
    },
    {
      "entropy": 5.914966058731079,
      "epoch": 2.2494101197238487,
      "grad_norm": 1.25,
      "learning_rate": 0.0004492101917415183,
      "loss": 5.2724,
      "mean_token_accuracy": 0.1973956510424614,
      "num_tokens": 52159238.0,
      "step": 25740
    },
    {
      "entropy": 5.973027038574219,
      "epoch": 2.249847068076553,
      "grad_norm": 1.3125,
      "learning_rate": 0.00044919046950688744,
      "loss": 5.2119,
      "mean_token_accuracy": 0.2090591847896576,
      "num_tokens": 52169172.0,
      "step": 25745
    },
    {
      "entropy": 5.951249885559082,
      "epoch": 2.250284016429258,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044917074393119224,
      "loss": 5.2314,
      "mean_token_accuracy": 0.2049234539270401,
      "num_tokens": 52179432.0,
      "step": 25750
    },
    {
      "entropy": 5.950696086883545,
      "epoch": 2.2507209647819626,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044915101501481086,
      "loss": 5.2768,
      "mean_token_accuracy": 0.20506273061037064,
      "num_tokens": 52190939.0,
      "step": 25755
    },
    {
      "entropy": 5.921615123748779,
      "epoch": 2.2511579131346675,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044913128275812184,
      "loss": 5.2412,
      "mean_token_accuracy": 0.20697939544916152,
      "num_tokens": 52201644.0,
      "step": 25760
    },
    {
      "entropy": 6.02143874168396,
      "epoch": 2.251594861487372,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004491115471615035,
      "loss": 5.2425,
      "mean_token_accuracy": 0.2021028593182564,
      "num_tokens": 52211353.0,
      "step": 25765
    },
    {
      "entropy": 6.04316782951355,
      "epoch": 2.252031809840077,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004490918082253344,
      "loss": 5.3338,
      "mean_token_accuracy": 0.19333953112363816,
      "num_tokens": 52221717.0,
      "step": 25770
    },
    {
      "entropy": 5.958343982696533,
      "epoch": 2.252468758192782,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044907206594999325,
      "loss": 5.3263,
      "mean_token_accuracy": 0.20213766992092133,
      "num_tokens": 52232419.0,
      "step": 25775
    },
    {
      "entropy": 6.028144598007202,
      "epoch": 2.2529057065454863,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044905232033585867,
      "loss": 5.3037,
      "mean_token_accuracy": 0.1980423301458359,
      "num_tokens": 52242386.0,
      "step": 25780
    },
    {
      "entropy": 6.019436407089233,
      "epoch": 2.253342654898191,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004490325713833094,
      "loss": 5.2779,
      "mean_token_accuracy": 0.2029683455824852,
      "num_tokens": 52252569.0,
      "step": 25785
    },
    {
      "entropy": 5.908410692214966,
      "epoch": 2.2537796032508957,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004490128190927242,
      "loss": 5.1771,
      "mean_token_accuracy": 0.2145699977874756,
      "num_tokens": 52263101.0,
      "step": 25790
    },
    {
      "entropy": 5.988770437240601,
      "epoch": 2.2542165516036006,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044899306346448195,
      "loss": 5.2747,
      "mean_token_accuracy": 0.2016826704144478,
      "num_tokens": 52273250.0,
      "step": 25795
    },
    {
      "entropy": 6.035788631439209,
      "epoch": 2.254653499956305,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004489733044989616,
      "loss": 5.3757,
      "mean_token_accuracy": 0.19502123594284057,
      "num_tokens": 52284351.0,
      "step": 25800
    },
    {
      "entropy": 5.945617389678955,
      "epoch": 2.25509044830901,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044895354219654204,
      "loss": 5.2552,
      "mean_token_accuracy": 0.2070562556385994,
      "num_tokens": 52294642.0,
      "step": 25805
    },
    {
      "entropy": 5.939660501480103,
      "epoch": 2.2555273966617144,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044893377655760245,
      "loss": 5.1989,
      "mean_token_accuracy": 0.20447924882173538,
      "num_tokens": 52303793.0,
      "step": 25810
    },
    {
      "entropy": 5.975811195373535,
      "epoch": 2.2559643450144193,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044891400758252187,
      "loss": 5.3085,
      "mean_token_accuracy": 0.1926594227552414,
      "num_tokens": 52313541.0,
      "step": 25815
    },
    {
      "entropy": 6.021809387207031,
      "epoch": 2.256401293367124,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004488942352716795,
      "loss": 5.2804,
      "mean_token_accuracy": 0.20189407020807265,
      "num_tokens": 52325428.0,
      "step": 25820
    },
    {
      "entropy": 5.980365753173828,
      "epoch": 2.2568382417198287,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004488744596254545,
      "loss": 5.3077,
      "mean_token_accuracy": 0.20167040973901748,
      "num_tokens": 52335837.0,
      "step": 25825
    },
    {
      "entropy": 5.934809589385987,
      "epoch": 2.2572751900725336,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044885468064422636,
      "loss": 5.2499,
      "mean_token_accuracy": 0.203309965133667,
      "num_tokens": 52346925.0,
      "step": 25830
    },
    {
      "entropy": 5.985050630569458,
      "epoch": 2.257712138425238,
      "grad_norm": 3.625,
      "learning_rate": 0.0004488348983283743,
      "loss": 5.3803,
      "mean_token_accuracy": 0.194162617623806,
      "num_tokens": 52357244.0,
      "step": 25835
    },
    {
      "entropy": 5.972400522232055,
      "epoch": 2.258149086777943,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004488151126782779,
      "loss": 5.2615,
      "mean_token_accuracy": 0.20487895607948303,
      "num_tokens": 52367657.0,
      "step": 25840
    },
    {
      "entropy": 6.019529724121094,
      "epoch": 2.2585860351306475,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044879532369431635,
      "loss": 5.3064,
      "mean_token_accuracy": 0.19760869592428207,
      "num_tokens": 52377298.0,
      "step": 25845
    },
    {
      "entropy": 6.010892915725708,
      "epoch": 2.2590229834833524,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044877553137686953,
      "loss": 5.3424,
      "mean_token_accuracy": 0.20083278715610503,
      "num_tokens": 52387829.0,
      "step": 25850
    },
    {
      "entropy": 5.963274669647217,
      "epoch": 2.259459931836057,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044875573572631706,
      "loss": 5.2814,
      "mean_token_accuracy": 0.2069653421640396,
      "num_tokens": 52398214.0,
      "step": 25855
    },
    {
      "entropy": 5.974958705902099,
      "epoch": 2.259896880188762,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004487359367430384,
      "loss": 5.2496,
      "mean_token_accuracy": 0.20205751359462737,
      "num_tokens": 52407619.0,
      "step": 25860
    },
    {
      "entropy": 5.975138139724732,
      "epoch": 2.2603338285414663,
      "grad_norm": 1.125,
      "learning_rate": 0.0004487161344274135,
      "loss": 5.2967,
      "mean_token_accuracy": 0.19795904755592347,
      "num_tokens": 52417465.0,
      "step": 25865
    },
    {
      "entropy": 5.924899673461914,
      "epoch": 2.260770776894171,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004486963287798221,
      "loss": 5.2632,
      "mean_token_accuracy": 0.20705086886882781,
      "num_tokens": 52427000.0,
      "step": 25870
    },
    {
      "entropy": 6.019606447219848,
      "epoch": 2.2612077252468756,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004486765198006441,
      "loss": 5.3007,
      "mean_token_accuracy": 0.19729628711938857,
      "num_tokens": 52437449.0,
      "step": 25875
    },
    {
      "entropy": 6.061234951019287,
      "epoch": 2.2616446735995805,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044865670749025943,
      "loss": 5.3485,
      "mean_token_accuracy": 0.1940264418721199,
      "num_tokens": 52446946.0,
      "step": 25880
    },
    {
      "entropy": 5.915046977996826,
      "epoch": 2.2620816219522855,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044863689184904815,
      "loss": 5.2511,
      "mean_token_accuracy": 0.20301511287689208,
      "num_tokens": 52456993.0,
      "step": 25885
    },
    {
      "entropy": 5.993336820602417,
      "epoch": 2.26251857030499,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004486170728773903,
      "loss": 5.3082,
      "mean_token_accuracy": 0.1954186499118805,
      "num_tokens": 52467999.0,
      "step": 25890
    },
    {
      "entropy": 6.041971111297608,
      "epoch": 2.262955518657695,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000448597250575666,
      "loss": 5.3606,
      "mean_token_accuracy": 0.1918350115418434,
      "num_tokens": 52479556.0,
      "step": 25895
    },
    {
      "entropy": 6.03965015411377,
      "epoch": 2.2633924670103993,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004485774249442555,
      "loss": 5.2949,
      "mean_token_accuracy": 0.20411693304777145,
      "num_tokens": 52489160.0,
      "step": 25900
    },
    {
      "entropy": 5.960551452636719,
      "epoch": 2.2638294153631042,
      "grad_norm": 1.125,
      "learning_rate": 0.000448557595983539,
      "loss": 5.2448,
      "mean_token_accuracy": 0.19842283427715302,
      "num_tokens": 52500102.0,
      "step": 25905
    },
    {
      "entropy": 5.954697799682617,
      "epoch": 2.2642663637158087,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044853776369389685,
      "loss": 5.2755,
      "mean_token_accuracy": 0.2018350601196289,
      "num_tokens": 52509615.0,
      "step": 25910
    },
    {
      "entropy": 5.94775881767273,
      "epoch": 2.2647033120685136,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004485179280757095,
      "loss": 5.2532,
      "mean_token_accuracy": 0.212725430727005,
      "num_tokens": 52520796.0,
      "step": 25915
    },
    {
      "entropy": 5.976756429672241,
      "epoch": 2.265140260421218,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044849808912935735,
      "loss": 5.234,
      "mean_token_accuracy": 0.20306461155414582,
      "num_tokens": 52530641.0,
      "step": 25920
    },
    {
      "entropy": 5.950080251693725,
      "epoch": 2.265577208773923,
      "grad_norm": 1.078125,
      "learning_rate": 0.000448478246855221,
      "loss": 5.2849,
      "mean_token_accuracy": 0.20711189210414888,
      "num_tokens": 52541626.0,
      "step": 25925
    },
    {
      "entropy": 5.962403583526611,
      "epoch": 2.2660141571266275,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004484584012536809,
      "loss": 5.3285,
      "mean_token_accuracy": 0.20161781311035157,
      "num_tokens": 52552301.0,
      "step": 25930
    },
    {
      "entropy": 5.927394962310791,
      "epoch": 2.2664511054793324,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004484385523251177,
      "loss": 5.2031,
      "mean_token_accuracy": 0.21619621217250823,
      "num_tokens": 52562463.0,
      "step": 25935
    },
    {
      "entropy": 6.033977746963501,
      "epoch": 2.2668880538320373,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004484187000699123,
      "loss": 5.3256,
      "mean_token_accuracy": 0.19327559769153596,
      "num_tokens": 52572782.0,
      "step": 25940
    },
    {
      "entropy": 5.978731441497803,
      "epoch": 2.2673250021847418,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044839884448844534,
      "loss": 5.2493,
      "mean_token_accuracy": 0.20490159392356871,
      "num_tokens": 52582293.0,
      "step": 25945
    },
    {
      "entropy": 5.981101179122925,
      "epoch": 2.2677619505374462,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004483789855810977,
      "loss": 5.2851,
      "mean_token_accuracy": 0.19982358664274216,
      "num_tokens": 52591950.0,
      "step": 25950
    },
    {
      "entropy": 5.947430038452149,
      "epoch": 2.268198898890151,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044835912334825024,
      "loss": 5.2751,
      "mean_token_accuracy": 0.20461659282445907,
      "num_tokens": 52603998.0,
      "step": 25955
    },
    {
      "entropy": 5.997095155715942,
      "epoch": 2.268635847242856,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044833925779028393,
      "loss": 5.3167,
      "mean_token_accuracy": 0.1961570143699646,
      "num_tokens": 52614955.0,
      "step": 25960
    },
    {
      "entropy": 6.046394109725952,
      "epoch": 2.2690727955955605,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004483193889075799,
      "loss": 5.3057,
      "mean_token_accuracy": 0.2004019260406494,
      "num_tokens": 52625305.0,
      "step": 25965
    },
    {
      "entropy": 5.975561618804932,
      "epoch": 2.2695097439482654,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004482995167005191,
      "loss": 5.2164,
      "mean_token_accuracy": 0.20655145347118378,
      "num_tokens": 52635310.0,
      "step": 25970
    },
    {
      "entropy": 5.972252893447876,
      "epoch": 2.26994669230097,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044827964116948284,
      "loss": 5.3029,
      "mean_token_accuracy": 0.2004230797290802,
      "num_tokens": 52644946.0,
      "step": 25975
    },
    {
      "entropy": 5.981488370895386,
      "epoch": 2.270383640653675,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004482597623148522,
      "loss": 5.2729,
      "mean_token_accuracy": 0.19713380932807922,
      "num_tokens": 52655056.0,
      "step": 25980
    },
    {
      "entropy": 5.991111707687378,
      "epoch": 2.2708205890063793,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004482398801370085,
      "loss": 5.2979,
      "mean_token_accuracy": 0.20120468884706497,
      "num_tokens": 52665423.0,
      "step": 25985
    },
    {
      "entropy": 6.045563077926635,
      "epoch": 2.271257537359084,
      "grad_norm": 1.125,
      "learning_rate": 0.00044821999463633323,
      "loss": 5.3313,
      "mean_token_accuracy": 0.19937362819910048,
      "num_tokens": 52675300.0,
      "step": 25990
    },
    {
      "entropy": 6.03339638710022,
      "epoch": 2.271694485711789,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044820010581320764,
      "loss": 5.2863,
      "mean_token_accuracy": 0.20210985839366913,
      "num_tokens": 52685431.0,
      "step": 25995
    },
    {
      "entropy": 6.005045223236084,
      "epoch": 2.2721314340644936,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044818021366801325,
      "loss": 5.3797,
      "mean_token_accuracy": 0.19421911239624023,
      "num_tokens": 52695756.0,
      "step": 26000
    },
    {
      "entropy": 5.984430313110352,
      "epoch": 2.272568382417198,
      "grad_norm": 1.25,
      "learning_rate": 0.0004481603182011317,
      "loss": 5.2383,
      "mean_token_accuracy": 0.20345399528741837,
      "num_tokens": 52704889.0,
      "step": 26005
    },
    {
      "entropy": 5.986238718032837,
      "epoch": 2.273005330769903,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044814041941294435,
      "loss": 5.2188,
      "mean_token_accuracy": 0.20425159484148026,
      "num_tokens": 52714227.0,
      "step": 26010
    },
    {
      "entropy": 5.928758955001831,
      "epoch": 2.273442279122608,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044812051730383316,
      "loss": 5.2113,
      "mean_token_accuracy": 0.21331766992807388,
      "num_tokens": 52725150.0,
      "step": 26015
    },
    {
      "entropy": 5.976894426345825,
      "epoch": 2.2738792274753123,
      "grad_norm": 1.125,
      "learning_rate": 0.0004481006118741798,
      "loss": 5.2455,
      "mean_token_accuracy": 0.19970356673002243,
      "num_tokens": 52734825.0,
      "step": 26020
    },
    {
      "entropy": 5.98288106918335,
      "epoch": 2.2743161758280173,
      "grad_norm": 1.125,
      "learning_rate": 0.0004480807031243658,
      "loss": 5.2552,
      "mean_token_accuracy": 0.21018600314855576,
      "num_tokens": 52745185.0,
      "step": 26025
    },
    {
      "entropy": 5.976874303817749,
      "epoch": 2.2747531241807217,
      "grad_norm": 1.25,
      "learning_rate": 0.00044806079105477335,
      "loss": 5.3268,
      "mean_token_accuracy": 0.20187413990497588,
      "num_tokens": 52755122.0,
      "step": 26030
    },
    {
      "entropy": 5.954514741897583,
      "epoch": 2.2751900725334266,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004480408756657842,
      "loss": 5.2864,
      "mean_token_accuracy": 0.19901591986417771,
      "num_tokens": 52765338.0,
      "step": 26035
    },
    {
      "entropy": 6.039728116989136,
      "epoch": 2.275627020886131,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044802095695778047,
      "loss": 5.3654,
      "mean_token_accuracy": 0.19008792489767073,
      "num_tokens": 52775045.0,
      "step": 26040
    },
    {
      "entropy": 5.997090768814087,
      "epoch": 2.276063969238836,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044800103493114403,
      "loss": 5.2987,
      "mean_token_accuracy": 0.20002927035093307,
      "num_tokens": 52785942.0,
      "step": 26045
    },
    {
      "entropy": 5.9279797077178955,
      "epoch": 2.2765009175915405,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004479811095862571,
      "loss": 5.2379,
      "mean_token_accuracy": 0.21205351501703262,
      "num_tokens": 52794905.0,
      "step": 26050
    },
    {
      "entropy": 5.982124757766724,
      "epoch": 2.2769378659442454,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004479611809235019,
      "loss": 5.1804,
      "mean_token_accuracy": 0.21022075563669204,
      "num_tokens": 52804091.0,
      "step": 26055
    },
    {
      "entropy": 5.998239946365357,
      "epoch": 2.27737481429695,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004479412489432605,
      "loss": 5.3046,
      "mean_token_accuracy": 0.1973986431956291,
      "num_tokens": 52814400.0,
      "step": 26060
    },
    {
      "entropy": 5.999188375473023,
      "epoch": 2.277811762649655,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044792131364591544,
      "loss": 5.3016,
      "mean_token_accuracy": 0.19932857304811477,
      "num_tokens": 52824423.0,
      "step": 26065
    },
    {
      "entropy": 5.981936502456665,
      "epoch": 2.2782487110023597,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044790137503184895,
      "loss": 5.3021,
      "mean_token_accuracy": 0.20564573854207993,
      "num_tokens": 52833852.0,
      "step": 26070
    },
    {
      "entropy": 5.976498031616211,
      "epoch": 2.278685659355064,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004478814331014435,
      "loss": 5.3013,
      "mean_token_accuracy": 0.20225505977869035,
      "num_tokens": 52844759.0,
      "step": 26075
    },
    {
      "entropy": 5.961839056015014,
      "epoch": 2.279122607707769,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004478614878550815,
      "loss": 5.2465,
      "mean_token_accuracy": 0.20384530872106552,
      "num_tokens": 52854302.0,
      "step": 26080
    },
    {
      "entropy": 5.981349992752075,
      "epoch": 2.2795595560604736,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004478415392931457,
      "loss": 5.2755,
      "mean_token_accuracy": 0.2013041228055954,
      "num_tokens": 52864086.0,
      "step": 26085
    },
    {
      "entropy": 5.954524040222168,
      "epoch": 2.2799965044131785,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044782158741601854,
      "loss": 5.2074,
      "mean_token_accuracy": 0.20539432466030122,
      "num_tokens": 52874894.0,
      "step": 26090
    },
    {
      "entropy": 5.876962280273437,
      "epoch": 2.280433452765883,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044780163222408277,
      "loss": 5.2569,
      "mean_token_accuracy": 0.20539739578962327,
      "num_tokens": 52885418.0,
      "step": 26095
    },
    {
      "entropy": 6.005073118209839,
      "epoch": 2.280870401118588,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044778167371772117,
      "loss": 5.3267,
      "mean_token_accuracy": 0.20011919885873794,
      "num_tokens": 52895515.0,
      "step": 26100
    },
    {
      "entropy": 5.980842590332031,
      "epoch": 2.2813073494712923,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004477617118973165,
      "loss": 5.2702,
      "mean_token_accuracy": 0.2099935531616211,
      "num_tokens": 52905599.0,
      "step": 26105
    },
    {
      "entropy": 6.034142398834229,
      "epoch": 2.2817442978239972,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004477417467632517,
      "loss": 5.3772,
      "mean_token_accuracy": 0.19549345076084138,
      "num_tokens": 52916599.0,
      "step": 26110
    },
    {
      "entropy": 5.974320554733277,
      "epoch": 2.2821812461767017,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044772177831590965,
      "loss": 5.2632,
      "mean_token_accuracy": 0.21193960756063462,
      "num_tokens": 52928131.0,
      "step": 26115
    },
    {
      "entropy": 5.992448186874389,
      "epoch": 2.2826181945294066,
      "grad_norm": 1.125,
      "learning_rate": 0.00044770180655567343,
      "loss": 5.3221,
      "mean_token_accuracy": 0.19565028548240662,
      "num_tokens": 52939305.0,
      "step": 26120
    },
    {
      "entropy": 5.944539594650268,
      "epoch": 2.2830551428821115,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004476818314829261,
      "loss": 5.1758,
      "mean_token_accuracy": 0.2073556065559387,
      "num_tokens": 52948814.0,
      "step": 26125
    },
    {
      "entropy": 5.98690619468689,
      "epoch": 2.283492091234816,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004476618530980506,
      "loss": 5.326,
      "mean_token_accuracy": 0.1951696440577507,
      "num_tokens": 52958964.0,
      "step": 26130
    },
    {
      "entropy": 5.976001358032226,
      "epoch": 2.283929039587521,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004476418714014305,
      "loss": 5.3387,
      "mean_token_accuracy": 0.19576144069433213,
      "num_tokens": 52969837.0,
      "step": 26135
    },
    {
      "entropy": 5.994408845901489,
      "epoch": 2.2843659879402254,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044762188639344875,
      "loss": 5.2447,
      "mean_token_accuracy": 0.20784379839897155,
      "num_tokens": 52979184.0,
      "step": 26140
    },
    {
      "entropy": 6.025073051452637,
      "epoch": 2.2848029362929303,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044760189807448874,
      "loss": 5.4019,
      "mean_token_accuracy": 0.1948528990149498,
      "num_tokens": 52989867.0,
      "step": 26145
    },
    {
      "entropy": 5.9744054794311525,
      "epoch": 2.2852398846456348,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004475819064449338,
      "loss": 5.3329,
      "mean_token_accuracy": 0.1984528422355652,
      "num_tokens": 53000177.0,
      "step": 26150
    },
    {
      "entropy": 5.941235589981079,
      "epoch": 2.2856768329983397,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044756191150516764,
      "loss": 5.2081,
      "mean_token_accuracy": 0.20006634145975113,
      "num_tokens": 53009372.0,
      "step": 26155
    },
    {
      "entropy": 5.980286550521851,
      "epoch": 2.286113781351044,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004475419132555735,
      "loss": 5.2617,
      "mean_token_accuracy": 0.1987534984946251,
      "num_tokens": 53020405.0,
      "step": 26160
    },
    {
      "entropy": 5.983349895477295,
      "epoch": 2.286550729703749,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044752191169653504,
      "loss": 5.318,
      "mean_token_accuracy": 0.19906822293996812,
      "num_tokens": 53030713.0,
      "step": 26165
    },
    {
      "entropy": 5.977909851074219,
      "epoch": 2.2869876780564535,
      "grad_norm": 1.125,
      "learning_rate": 0.00044750190682843596,
      "loss": 5.2665,
      "mean_token_accuracy": 0.20394909530878066,
      "num_tokens": 53040868.0,
      "step": 26170
    },
    {
      "entropy": 5.965624475479126,
      "epoch": 2.2874246264091584,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004474818986516599,
      "loss": 5.1807,
      "mean_token_accuracy": 0.20517090260982512,
      "num_tokens": 53050349.0,
      "step": 26175
    },
    {
      "entropy": 5.965198040008545,
      "epoch": 2.2878615747618634,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044746188716659064,
      "loss": 5.2668,
      "mean_token_accuracy": 0.2023068130016327,
      "num_tokens": 53060446.0,
      "step": 26180
    },
    {
      "entropy": 5.9778642654418945,
      "epoch": 2.288298523114568,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044744187237361207,
      "loss": 5.3054,
      "mean_token_accuracy": 0.19863708913326264,
      "num_tokens": 53071099.0,
      "step": 26185
    },
    {
      "entropy": 5.962195920944214,
      "epoch": 2.2887354714672727,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044742185427310797,
      "loss": 5.227,
      "mean_token_accuracy": 0.20450480729341508,
      "num_tokens": 53080712.0,
      "step": 26190
    },
    {
      "entropy": 6.0206627368927,
      "epoch": 2.289172419819977,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004474018328654623,
      "loss": 5.3799,
      "mean_token_accuracy": 0.19296462833881378,
      "num_tokens": 53091160.0,
      "step": 26195
    },
    {
      "entropy": 5.920805835723877,
      "epoch": 2.289609368172682,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004473818081510592,
      "loss": 5.3137,
      "mean_token_accuracy": 0.2031278133392334,
      "num_tokens": 53100767.0,
      "step": 26200
    },
    {
      "entropy": 6.006697654724121,
      "epoch": 2.2900463165253866,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044736178013028276,
      "loss": 5.2914,
      "mean_token_accuracy": 0.20342222452163697,
      "num_tokens": 53111641.0,
      "step": 26205
    },
    {
      "entropy": 5.99641432762146,
      "epoch": 2.2904832648780915,
      "grad_norm": 1.078125,
      "learning_rate": 0.000447341748803517,
      "loss": 5.2572,
      "mean_token_accuracy": 0.20721617937088013,
      "num_tokens": 53121982.0,
      "step": 26210
    },
    {
      "entropy": 6.023488283157349,
      "epoch": 2.290920213230796,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044732171417114614,
      "loss": 5.3219,
      "mean_token_accuracy": 0.19784843772649766,
      "num_tokens": 53132649.0,
      "step": 26215
    },
    {
      "entropy": 5.968764925003052,
      "epoch": 2.291357161583501,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004473016762335546,
      "loss": 5.3388,
      "mean_token_accuracy": 0.2000483900308609,
      "num_tokens": 53142249.0,
      "step": 26220
    },
    {
      "entropy": 6.013354969024658,
      "epoch": 2.2917941099362054,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004472816349911266,
      "loss": 5.2629,
      "mean_token_accuracy": 0.20667946189641953,
      "num_tokens": 53152375.0,
      "step": 26225
    },
    {
      "entropy": 6.015341091156006,
      "epoch": 2.2922310582889103,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004472615904442465,
      "loss": 5.3416,
      "mean_token_accuracy": 0.2020195886492729,
      "num_tokens": 53162699.0,
      "step": 26230
    },
    {
      "entropy": 6.033641767501831,
      "epoch": 2.292668006641615,
      "grad_norm": 1.328125,
      "learning_rate": 0.00044724154259329893,
      "loss": 5.3268,
      "mean_token_accuracy": 0.19925981909036636,
      "num_tokens": 53172890.0,
      "step": 26235
    },
    {
      "entropy": 6.005310344696045,
      "epoch": 2.2931049549943197,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004472214914386682,
      "loss": 5.3046,
      "mean_token_accuracy": 0.203839772939682,
      "num_tokens": 53183735.0,
      "step": 26240
    },
    {
      "entropy": 6.047776412963867,
      "epoch": 2.2935419033470246,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004472014369807391,
      "loss": 5.3054,
      "mean_token_accuracy": 0.1979999750852585,
      "num_tokens": 53193046.0,
      "step": 26245
    },
    {
      "entropy": 6.004939317703247,
      "epoch": 2.293978851699729,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004471813792198962,
      "loss": 5.2487,
      "mean_token_accuracy": 0.2107205033302307,
      "num_tokens": 53203367.0,
      "step": 26250
    },
    {
      "entropy": 6.000207042694091,
      "epoch": 2.294415800052434,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044716131815652425,
      "loss": 5.4151,
      "mean_token_accuracy": 0.19387611746788025,
      "num_tokens": 53214207.0,
      "step": 26255
    },
    {
      "entropy": 6.027893209457398,
      "epoch": 2.2948527484051384,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000447141253791008,
      "loss": 5.291,
      "mean_token_accuracy": 0.19945420920848847,
      "num_tokens": 53225061.0,
      "step": 26260
    },
    {
      "entropy": 6.0364399433135985,
      "epoch": 2.2952896967578433,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044712118612373223,
      "loss": 5.3116,
      "mean_token_accuracy": 0.19774167537689208,
      "num_tokens": 53235463.0,
      "step": 26265
    },
    {
      "entropy": 6.025462055206299,
      "epoch": 2.295726645110548,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000447101115155082,
      "loss": 5.3365,
      "mean_token_accuracy": 0.1958286166191101,
      "num_tokens": 53245923.0,
      "step": 26270
    },
    {
      "entropy": 5.974324464797974,
      "epoch": 2.2961635934632527,
      "grad_norm": 1.125,
      "learning_rate": 0.00044708104088544216,
      "loss": 5.2792,
      "mean_token_accuracy": 0.20022417604923248,
      "num_tokens": 53255756.0,
      "step": 26275
    },
    {
      "entropy": 5.971109867095947,
      "epoch": 2.296600541815957,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004470609633151978,
      "loss": 5.28,
      "mean_token_accuracy": 0.20030320286750794,
      "num_tokens": 53266176.0,
      "step": 26280
    },
    {
      "entropy": 5.9809760570526125,
      "epoch": 2.297037490168662,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044704088244473404,
      "loss": 5.2941,
      "mean_token_accuracy": 0.20084454566240312,
      "num_tokens": 53276703.0,
      "step": 26285
    },
    {
      "entropy": 6.046857500076294,
      "epoch": 2.297474438521367,
      "grad_norm": 1.28125,
      "learning_rate": 0.000447020798274436,
      "loss": 5.3803,
      "mean_token_accuracy": 0.19681245535612107,
      "num_tokens": 53286511.0,
      "step": 26290
    },
    {
      "entropy": 5.987572431564331,
      "epoch": 2.2979113868740715,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044700071080468884,
      "loss": 5.2826,
      "mean_token_accuracy": 0.20324795842170715,
      "num_tokens": 53297261.0,
      "step": 26295
    },
    {
      "entropy": 6.035807943344116,
      "epoch": 2.298348335226776,
      "grad_norm": 1.140625,
      "learning_rate": 0.000446980620035878,
      "loss": 5.3026,
      "mean_token_accuracy": 0.196852970123291,
      "num_tokens": 53307054.0,
      "step": 26300
    },
    {
      "entropy": 5.990601634979248,
      "epoch": 2.298785283579481,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044696052596838873,
      "loss": 5.2821,
      "mean_token_accuracy": 0.20361760258674622,
      "num_tokens": 53318022.0,
      "step": 26305
    },
    {
      "entropy": 5.9850739479064945,
      "epoch": 2.299222231932186,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004469404286026064,
      "loss": 5.2854,
      "mean_token_accuracy": 0.20306855589151382,
      "num_tokens": 53328462.0,
      "step": 26310
    },
    {
      "entropy": 5.960739231109619,
      "epoch": 2.2996591802848902,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044692032793891665,
      "loss": 5.2195,
      "mean_token_accuracy": 0.21259954273700715,
      "num_tokens": 53337837.0,
      "step": 26315
    },
    {
      "entropy": 5.957828998565674,
      "epoch": 2.300096128637595,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044690022397770484,
      "loss": 5.2594,
      "mean_token_accuracy": 0.204704087972641,
      "num_tokens": 53348026.0,
      "step": 26320
    },
    {
      "entropy": 6.01363582611084,
      "epoch": 2.3005330769902996,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004468801167193567,
      "loss": 5.2946,
      "mean_token_accuracy": 0.2019259139895439,
      "num_tokens": 53358881.0,
      "step": 26325
    },
    {
      "entropy": 5.961754846572876,
      "epoch": 2.3009700253430045,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00044686000616425795,
      "loss": 5.268,
      "mean_token_accuracy": 0.20571938753128052,
      "num_tokens": 53369732.0,
      "step": 26330
    },
    {
      "entropy": 5.960969495773315,
      "epoch": 2.301406973695709,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044683989231279406,
      "loss": 5.2924,
      "mean_token_accuracy": 0.20252117216587068,
      "num_tokens": 53379607.0,
      "step": 26335
    },
    {
      "entropy": 6.057096290588379,
      "epoch": 2.301843922048414,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044681977516535106,
      "loss": 5.3246,
      "mean_token_accuracy": 0.19147243052721025,
      "num_tokens": 53390630.0,
      "step": 26340
    },
    {
      "entropy": 5.958010578155518,
      "epoch": 2.3022808704011184,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004467996547223148,
      "loss": 5.1906,
      "mean_token_accuracy": 0.2090715378522873,
      "num_tokens": 53400830.0,
      "step": 26345
    },
    {
      "entropy": 5.962417078018189,
      "epoch": 2.3027178187538233,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044677953098407107,
      "loss": 5.2948,
      "mean_token_accuracy": 0.20014173835515975,
      "num_tokens": 53411150.0,
      "step": 26350
    },
    {
      "entropy": 5.955305194854736,
      "epoch": 2.303154767106528,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004467594039510059,
      "loss": 5.268,
      "mean_token_accuracy": 0.2047977030277252,
      "num_tokens": 53421072.0,
      "step": 26355
    },
    {
      "entropy": 6.023828411102295,
      "epoch": 2.3035917154592327,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044673927362350545,
      "loss": 5.2435,
      "mean_token_accuracy": 0.20247279554605485,
      "num_tokens": 53430659.0,
      "step": 26360
    },
    {
      "entropy": 5.94343934059143,
      "epoch": 2.3040286638119376,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044671914000195566,
      "loss": 5.1732,
      "mean_token_accuracy": 0.20643617808818818,
      "num_tokens": 53440883.0,
      "step": 26365
    },
    {
      "entropy": 5.951247262954712,
      "epoch": 2.304465612164642,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044669900308674283,
      "loss": 5.2887,
      "mean_token_accuracy": 0.20575280338525773,
      "num_tokens": 53450491.0,
      "step": 26370
    },
    {
      "entropy": 5.936660623550415,
      "epoch": 2.304902560517347,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044667886287825316,
      "loss": 5.2829,
      "mean_token_accuracy": 0.20236023962497712,
      "num_tokens": 53460864.0,
      "step": 26375
    },
    {
      "entropy": 5.979042482376099,
      "epoch": 2.3053395088700515,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044665871937687296,
      "loss": 5.2791,
      "mean_token_accuracy": 0.20138145387172698,
      "num_tokens": 53470965.0,
      "step": 26380
    },
    {
      "entropy": 5.985240840911866,
      "epoch": 2.3057764572227564,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004466385725829885,
      "loss": 5.1959,
      "mean_token_accuracy": 0.2118961438536644,
      "num_tokens": 53480272.0,
      "step": 26385
    },
    {
      "entropy": 5.937760686874389,
      "epoch": 2.306213405575461,
      "grad_norm": 1.125,
      "learning_rate": 0.0004466184224969863,
      "loss": 5.3145,
      "mean_token_accuracy": 0.20038765221834182,
      "num_tokens": 53490285.0,
      "step": 26390
    },
    {
      "entropy": 5.971506023406983,
      "epoch": 2.3066503539281658,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004465982691192529,
      "loss": 5.1995,
      "mean_token_accuracy": 0.2026042878627777,
      "num_tokens": 53499712.0,
      "step": 26395
    },
    {
      "entropy": 5.966630935668945,
      "epoch": 2.3070873022808702,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004465781124501747,
      "loss": 5.2704,
      "mean_token_accuracy": 0.20495169162750243,
      "num_tokens": 53510144.0,
      "step": 26400
    },
    {
      "entropy": 6.00187201499939,
      "epoch": 2.307524250633575,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044655795249013845,
      "loss": 5.2783,
      "mean_token_accuracy": 0.21021487265825273,
      "num_tokens": 53519936.0,
      "step": 26405
    },
    {
      "entropy": 5.989297103881836,
      "epoch": 2.3079611989862796,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004465377892395308,
      "loss": 5.312,
      "mean_token_accuracy": 0.20199796706438064,
      "num_tokens": 53530447.0,
      "step": 26410
    },
    {
      "entropy": 6.0300336360931395,
      "epoch": 2.3083981473389845,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004465176226987384,
      "loss": 5.3436,
      "mean_token_accuracy": 0.20175949931144715,
      "num_tokens": 53540929.0,
      "step": 26415
    },
    {
      "entropy": 5.956923484802246,
      "epoch": 2.3088350956916894,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004464974528681481,
      "loss": 5.2076,
      "mean_token_accuracy": 0.21550011783838272,
      "num_tokens": 53551575.0,
      "step": 26420
    },
    {
      "entropy": 5.993300199508667,
      "epoch": 2.309272044044394,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004464772797481469,
      "loss": 5.3472,
      "mean_token_accuracy": 0.19375950396060942,
      "num_tokens": 53563081.0,
      "step": 26425
    },
    {
      "entropy": 5.968145561218262,
      "epoch": 2.309708992397099,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044645710333912153,
      "loss": 5.2929,
      "mean_token_accuracy": 0.2048858806490898,
      "num_tokens": 53572303.0,
      "step": 26430
    },
    {
      "entropy": 5.987769317626953,
      "epoch": 2.3101459407498033,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044643692364145914,
      "loss": 5.2483,
      "mean_token_accuracy": 0.20565159916877745,
      "num_tokens": 53582033.0,
      "step": 26435
    },
    {
      "entropy": 5.8483068466186525,
      "epoch": 2.310582889102508,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044641674065554665,
      "loss": 5.088,
      "mean_token_accuracy": 0.23245813548564911,
      "num_tokens": 53592140.0,
      "step": 26440
    },
    {
      "entropy": 5.998413562774658,
      "epoch": 2.3110198374552127,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044639655438177134,
      "loss": 5.273,
      "mean_token_accuracy": 0.20060949474573136,
      "num_tokens": 53602601.0,
      "step": 26445
    },
    {
      "entropy": 5.989663600921631,
      "epoch": 2.3114567858079176,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044637636482052027,
      "loss": 5.2787,
      "mean_token_accuracy": 0.20266793817281722,
      "num_tokens": 53613292.0,
      "step": 26450
    },
    {
      "entropy": 5.99614520072937,
      "epoch": 2.311893734160622,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004463561719721807,
      "loss": 5.2674,
      "mean_token_accuracy": 0.20227111428976058,
      "num_tokens": 53624493.0,
      "step": 26455
    },
    {
      "entropy": 6.003597784042358,
      "epoch": 2.312330682513327,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044633597583713995,
      "loss": 5.2691,
      "mean_token_accuracy": 0.20297378599643706,
      "num_tokens": 53633739.0,
      "step": 26460
    },
    {
      "entropy": 6.019600772857666,
      "epoch": 2.3127676308660314,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004463157764157855,
      "loss": 5.2428,
      "mean_token_accuracy": 0.20576695799827577,
      "num_tokens": 53643025.0,
      "step": 26465
    },
    {
      "entropy": 5.993663644790649,
      "epoch": 2.3132045792187363,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004462955737085046,
      "loss": 5.2389,
      "mean_token_accuracy": 0.20662827789783478,
      "num_tokens": 53652905.0,
      "step": 26470
    },
    {
      "entropy": 5.903075885772705,
      "epoch": 2.3136415275714413,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004462753677156849,
      "loss": 5.2382,
      "mean_token_accuracy": 0.20683277696371077,
      "num_tokens": 53663189.0,
      "step": 26475
    },
    {
      "entropy": 5.953018999099731,
      "epoch": 2.3140784759241457,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004462551584377138,
      "loss": 5.304,
      "mean_token_accuracy": 0.20128540098667144,
      "num_tokens": 53673068.0,
      "step": 26480
    },
    {
      "entropy": 6.026959466934204,
      "epoch": 2.3145154242768506,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004462349458749791,
      "loss": 5.3258,
      "mean_token_accuracy": 0.19912416934967042,
      "num_tokens": 53683929.0,
      "step": 26485
    },
    {
      "entropy": 6.013142156600952,
      "epoch": 2.314952372629555,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044621473002786846,
      "loss": 5.3246,
      "mean_token_accuracy": 0.19885779470205306,
      "num_tokens": 53694170.0,
      "step": 26490
    },
    {
      "entropy": 5.979232215881348,
      "epoch": 2.31538932098226,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044619451089676946,
      "loss": 5.257,
      "mean_token_accuracy": 0.20716357678174974,
      "num_tokens": 53704491.0,
      "step": 26495
    },
    {
      "entropy": 6.01428189277649,
      "epoch": 2.3158262693349645,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004461742884820701,
      "loss": 5.2947,
      "mean_token_accuracy": 0.19946308434009552,
      "num_tokens": 53714971.0,
      "step": 26500
    },
    {
      "entropy": 5.779849910736084,
      "epoch": 2.3162632176876694,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004461540627841582,
      "loss": 5.0703,
      "mean_token_accuracy": 0.2358221098780632,
      "num_tokens": 53723965.0,
      "step": 26505
    },
    {
      "entropy": 5.982090473175049,
      "epoch": 2.316700166040374,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044613383380342163,
      "loss": 5.2644,
      "mean_token_accuracy": 0.20165972411632538,
      "num_tokens": 53734561.0,
      "step": 26510
    },
    {
      "entropy": 6.00323748588562,
      "epoch": 2.317137114393079,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004461136015402485,
      "loss": 5.2576,
      "mean_token_accuracy": 0.20267353951931,
      "num_tokens": 53745115.0,
      "step": 26515
    },
    {
      "entropy": 5.929891443252563,
      "epoch": 2.3175740627457833,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004460933659950269,
      "loss": 5.1867,
      "mean_token_accuracy": 0.20972787141799926,
      "num_tokens": 53754884.0,
      "step": 26520
    },
    {
      "entropy": 5.944492530822754,
      "epoch": 2.318011011098488,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004460731271681448,
      "loss": 5.2893,
      "mean_token_accuracy": 0.1979304239153862,
      "num_tokens": 53764772.0,
      "step": 26525
    },
    {
      "entropy": 6.0488067150115965,
      "epoch": 2.318447959451193,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044605288505999047,
      "loss": 5.3413,
      "mean_token_accuracy": 0.19361234009265899,
      "num_tokens": 53775391.0,
      "step": 26530
    },
    {
      "entropy": 6.009644174575806,
      "epoch": 2.3188849078038976,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044603263967095224,
      "loss": 5.3005,
      "mean_token_accuracy": 0.2000970348715782,
      "num_tokens": 53785698.0,
      "step": 26535
    },
    {
      "entropy": 6.020865058898925,
      "epoch": 2.3193218561566025,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044601239100141833,
      "loss": 5.2715,
      "mean_token_accuracy": 0.20225870460271836,
      "num_tokens": 53795432.0,
      "step": 26540
    },
    {
      "entropy": 6.023139619827271,
      "epoch": 2.319758804509307,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044599213905177724,
      "loss": 5.3367,
      "mean_token_accuracy": 0.20013624876737596,
      "num_tokens": 53805295.0,
      "step": 26545
    },
    {
      "entropy": 5.97481255531311,
      "epoch": 2.320195752862012,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004459718838224171,
      "loss": 5.2965,
      "mean_token_accuracy": 0.20735813528299332,
      "num_tokens": 53814765.0,
      "step": 26550
    },
    {
      "entropy": 5.993768215179443,
      "epoch": 2.3206327012147163,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044595162531372684,
      "loss": 5.2436,
      "mean_token_accuracy": 0.2019777461886406,
      "num_tokens": 53824154.0,
      "step": 26555
    },
    {
      "entropy": 6.001479625701904,
      "epoch": 2.3210696495674212,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044593136352609475,
      "loss": 5.2954,
      "mean_token_accuracy": 0.19926044791936875,
      "num_tokens": 53834867.0,
      "step": 26560
    },
    {
      "entropy": 6.059103202819824,
      "epoch": 2.3215065979201257,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004459110984599096,
      "loss": 5.3428,
      "mean_token_accuracy": 0.19337670058012008,
      "num_tokens": 53844901.0,
      "step": 26565
    },
    {
      "entropy": 5.988993787765503,
      "epoch": 2.3219435462728306,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044589083011555995,
      "loss": 5.3731,
      "mean_token_accuracy": 0.20253414362668992,
      "num_tokens": 53854897.0,
      "step": 26570
    },
    {
      "entropy": 6.005941152572632,
      "epoch": 2.322380494625535,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004458705584934347,
      "loss": 5.3242,
      "mean_token_accuracy": 0.19291084557771682,
      "num_tokens": 53865152.0,
      "step": 26575
    },
    {
      "entropy": 6.058319425582885,
      "epoch": 2.32281744297824,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044585028359392256,
      "loss": 5.3059,
      "mean_token_accuracy": 0.19708448499441147,
      "num_tokens": 53875261.0,
      "step": 26580
    },
    {
      "entropy": 6.015552091598511,
      "epoch": 2.323254391330945,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004458300054174125,
      "loss": 5.251,
      "mean_token_accuracy": 0.2007886663079262,
      "num_tokens": 53885281.0,
      "step": 26585
    },
    {
      "entropy": 5.905110263824463,
      "epoch": 2.3236913396836494,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004458097239642933,
      "loss": 5.2734,
      "mean_token_accuracy": 0.2046502411365509,
      "num_tokens": 53895282.0,
      "step": 26590
    },
    {
      "entropy": 6.002722883224488,
      "epoch": 2.3241282880363543,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004457894392349542,
      "loss": 5.3292,
      "mean_token_accuracy": 0.1995193362236023,
      "num_tokens": 53905843.0,
      "step": 26595
    },
    {
      "entropy": 5.9950127601623535,
      "epoch": 2.3245652363890588,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004457691512297841,
      "loss": 5.3606,
      "mean_token_accuracy": 0.19995751082897187,
      "num_tokens": 53915270.0,
      "step": 26600
    },
    {
      "entropy": 6.029417324066162,
      "epoch": 2.3250021847417637,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004457488599491723,
      "loss": 5.4008,
      "mean_token_accuracy": 0.19153389185667039,
      "num_tokens": 53924548.0,
      "step": 26605
    },
    {
      "entropy": 6.017051744461059,
      "epoch": 2.325439133094468,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044572856539350784,
      "loss": 5.3391,
      "mean_token_accuracy": 0.20058583766222,
      "num_tokens": 53935696.0,
      "step": 26610
    },
    {
      "entropy": 6.05939679145813,
      "epoch": 2.325876081447173,
      "grad_norm": 1.28125,
      "learning_rate": 0.00044570826756318006,
      "loss": 5.2954,
      "mean_token_accuracy": 0.20305695682764052,
      "num_tokens": 53945499.0,
      "step": 26615
    },
    {
      "entropy": 5.984999084472657,
      "epoch": 2.3263130297998775,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004456879664585782,
      "loss": 5.2167,
      "mean_token_accuracy": 0.20989808142185212,
      "num_tokens": 53955478.0,
      "step": 26620
    },
    {
      "entropy": 6.021804618835449,
      "epoch": 2.3267499781525824,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004456676620800917,
      "loss": 5.243,
      "mean_token_accuracy": 0.20199108719825745,
      "num_tokens": 53964908.0,
      "step": 26625
    },
    {
      "entropy": 5.984299421310425,
      "epoch": 2.327186926505287,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004456473544281101,
      "loss": 5.2366,
      "mean_token_accuracy": 0.2061626747250557,
      "num_tokens": 53974850.0,
      "step": 26630
    },
    {
      "entropy": 5.929650831222534,
      "epoch": 2.327623874857992,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004456270435030228,
      "loss": 5.2074,
      "mean_token_accuracy": 0.20826065689325332,
      "num_tokens": 53984458.0,
      "step": 26635
    },
    {
      "entropy": 5.952507400512696,
      "epoch": 2.3280608232106967,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004456067293052193,
      "loss": 5.2755,
      "mean_token_accuracy": 0.19922115355730058,
      "num_tokens": 53994560.0,
      "step": 26640
    },
    {
      "entropy": 6.052654218673706,
      "epoch": 2.328497771563401,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044558641183508945,
      "loss": 5.2669,
      "mean_token_accuracy": 0.20442079454660417,
      "num_tokens": 54003698.0,
      "step": 26645
    },
    {
      "entropy": 5.960959482192993,
      "epoch": 2.3289347199161057,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044556609109302277,
      "loss": 5.1966,
      "mean_token_accuracy": 0.21855515986680984,
      "num_tokens": 54013405.0,
      "step": 26650
    },
    {
      "entropy": 5.981247520446777,
      "epoch": 2.3293716682688106,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004455457670794092,
      "loss": 5.2913,
      "mean_token_accuracy": 0.2044383853673935,
      "num_tokens": 54023140.0,
      "step": 26655
    },
    {
      "entropy": 5.987320518493652,
      "epoch": 2.3298086166215155,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00044552543979463826,
      "loss": 5.2526,
      "mean_token_accuracy": 0.19957018047571182,
      "num_tokens": 54032506.0,
      "step": 26660
    },
    {
      "entropy": 5.94255256652832,
      "epoch": 2.33024556497422,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004455051092391002,
      "loss": 5.1702,
      "mean_token_accuracy": 0.21615176498889924,
      "num_tokens": 54042113.0,
      "step": 26665
    },
    {
      "entropy": 5.9424724102020265,
      "epoch": 2.330682513326925,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044548477541318473,
      "loss": 5.267,
      "mean_token_accuracy": 0.20628141313791276,
      "num_tokens": 54051139.0,
      "step": 26670
    },
    {
      "entropy": 5.943489217758179,
      "epoch": 2.3311194616796294,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004454644383172819,
      "loss": 5.2719,
      "mean_token_accuracy": 0.20310223251581191,
      "num_tokens": 54060929.0,
      "step": 26675
    },
    {
      "entropy": 6.050171899795532,
      "epoch": 2.3315564100323343,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004454440979517819,
      "loss": 5.3046,
      "mean_token_accuracy": 0.20131214559078217,
      "num_tokens": 54071654.0,
      "step": 26680
    },
    {
      "entropy": 5.996007442474365,
      "epoch": 2.3319933583850387,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004454237543170747,
      "loss": 5.2057,
      "mean_token_accuracy": 0.21012396216392518,
      "num_tokens": 54081497.0,
      "step": 26685
    },
    {
      "entropy": 5.989157676696777,
      "epoch": 2.3324303067377437,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00044540340741355063,
      "loss": 5.2764,
      "mean_token_accuracy": 0.2022074580192566,
      "num_tokens": 54091899.0,
      "step": 26690
    },
    {
      "entropy": 5.9222088813781735,
      "epoch": 2.332867255090448,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004453830572416,
      "loss": 5.2397,
      "mean_token_accuracy": 0.20325904488563537,
      "num_tokens": 54101734.0,
      "step": 26695
    },
    {
      "entropy": 6.012081861495972,
      "epoch": 2.333304203443153,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000445362703801613,
      "loss": 5.2807,
      "mean_token_accuracy": 0.20424388200044633,
      "num_tokens": 54112633.0,
      "step": 26700
    },
    {
      "entropy": 5.997182321548462,
      "epoch": 2.3337411517958575,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044534234709398,
      "loss": 5.261,
      "mean_token_accuracy": 0.19950223863124847,
      "num_tokens": 54122182.0,
      "step": 26705
    },
    {
      "entropy": 5.992593145370483,
      "epoch": 2.3341781001485624,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00044532198711909153,
      "loss": 5.3152,
      "mean_token_accuracy": 0.19070415645837785,
      "num_tokens": 54131410.0,
      "step": 26710
    },
    {
      "entropy": 5.925068140029907,
      "epoch": 2.3346150485012673,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004453016238773382,
      "loss": 5.208,
      "mean_token_accuracy": 0.20806040614843369,
      "num_tokens": 54141890.0,
      "step": 26715
    },
    {
      "entropy": 5.989047336578369,
      "epoch": 2.335051996853972,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004452812573691104,
      "loss": 5.2875,
      "mean_token_accuracy": 0.202058108150959,
      "num_tokens": 54151579.0,
      "step": 26720
    },
    {
      "entropy": 6.020118522644043,
      "epoch": 2.3354889452066767,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044526088759479893,
      "loss": 5.3187,
      "mean_token_accuracy": 0.20375843495130538,
      "num_tokens": 54161824.0,
      "step": 26725
    },
    {
      "entropy": 5.928245210647583,
      "epoch": 2.335925893559381,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044524051455479435,
      "loss": 5.2289,
      "mean_token_accuracy": 0.20583690106868743,
      "num_tokens": 54172110.0,
      "step": 26730
    },
    {
      "entropy": 5.949780607223511,
      "epoch": 2.336362841912086,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044522013824948756,
      "loss": 5.2273,
      "mean_token_accuracy": 0.2082776054739952,
      "num_tokens": 54181852.0,
      "step": 26735
    },
    {
      "entropy": 6.002387285232544,
      "epoch": 2.3367997902647906,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044519975867926933,
      "loss": 5.3112,
      "mean_token_accuracy": 0.20279179513454437,
      "num_tokens": 54192703.0,
      "step": 26740
    },
    {
      "entropy": 5.951768255233764,
      "epoch": 2.3372367386174955,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004451793758445305,
      "loss": 5.2184,
      "mean_token_accuracy": 0.21194638162851334,
      "num_tokens": 54201613.0,
      "step": 26745
    },
    {
      "entropy": 5.9921770095825195,
      "epoch": 2.3376736869702,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000445158989745662,
      "loss": 5.3598,
      "mean_token_accuracy": 0.1968343436717987,
      "num_tokens": 54211717.0,
      "step": 26750
    },
    {
      "entropy": 5.91988615989685,
      "epoch": 2.338110635322905,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044513860038305505,
      "loss": 5.1684,
      "mean_token_accuracy": 0.21248936653137207,
      "num_tokens": 54221481.0,
      "step": 26755
    },
    {
      "entropy": 6.04011869430542,
      "epoch": 2.3385475836756093,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004451182077571005,
      "loss": 5.266,
      "mean_token_accuracy": 0.2062675327062607,
      "num_tokens": 54231558.0,
      "step": 26760
    },
    {
      "entropy": 5.9967530250549315,
      "epoch": 2.3389845320283142,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044509781186818963,
      "loss": 5.2579,
      "mean_token_accuracy": 0.2027207598090172,
      "num_tokens": 54241150.0,
      "step": 26765
    },
    {
      "entropy": 5.984634637832642,
      "epoch": 2.339421480381019,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004450774127167136,
      "loss": 5.2764,
      "mean_token_accuracy": 0.20731580257415771,
      "num_tokens": 54251323.0,
      "step": 26770
    },
    {
      "entropy": 6.019742774963379,
      "epoch": 2.3398584287337236,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004450570103030637,
      "loss": 5.2796,
      "mean_token_accuracy": 0.19969063401222228,
      "num_tokens": 54260956.0,
      "step": 26775
    },
    {
      "entropy": 6.007708787918091,
      "epoch": 2.3402953770864285,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044503660462763116,
      "loss": 5.2732,
      "mean_token_accuracy": 0.204617378115654,
      "num_tokens": 54270382.0,
      "step": 26780
    },
    {
      "entropy": 6.025277614593506,
      "epoch": 2.340732325439133,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004450161956908075,
      "loss": 5.2948,
      "mean_token_accuracy": 0.2022526428103447,
      "num_tokens": 54280380.0,
      "step": 26785
    },
    {
      "entropy": 6.006008148193359,
      "epoch": 2.341169273791838,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004449957834929841,
      "loss": 5.2966,
      "mean_token_accuracy": 0.2015715643763542,
      "num_tokens": 54290536.0,
      "step": 26790
    },
    {
      "entropy": 5.916873264312744,
      "epoch": 2.3416062221445424,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044497536803455245,
      "loss": 5.2278,
      "mean_token_accuracy": 0.20941326022148132,
      "num_tokens": 54299835.0,
      "step": 26795
    },
    {
      "entropy": 5.982048845291137,
      "epoch": 2.3420431704972473,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004449549493159043,
      "loss": 5.2589,
      "mean_token_accuracy": 0.20691750943660736,
      "num_tokens": 54309496.0,
      "step": 26800
    },
    {
      "entropy": 5.9669544219970705,
      "epoch": 2.3424801188499518,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004449345273374311,
      "loss": 5.2412,
      "mean_token_accuracy": 0.20177914947271347,
      "num_tokens": 54318591.0,
      "step": 26805
    },
    {
      "entropy": 5.942092704772949,
      "epoch": 2.3429170672026567,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004449141020995246,
      "loss": 5.2589,
      "mean_token_accuracy": 0.20299112498760224,
      "num_tokens": 54327995.0,
      "step": 26810
    },
    {
      "entropy": 5.994596672058106,
      "epoch": 2.343354015555361,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004448936736025765,
      "loss": 5.3024,
      "mean_token_accuracy": 0.1960247576236725,
      "num_tokens": 54339122.0,
      "step": 26815
    },
    {
      "entropy": 6.034914541244507,
      "epoch": 2.343790963908066,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004448732418469789,
      "loss": 5.3281,
      "mean_token_accuracy": 0.2037929981946945,
      "num_tokens": 54349004.0,
      "step": 26820
    },
    {
      "entropy": 5.979753875732422,
      "epoch": 2.344227912260771,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044485280683312345,
      "loss": 5.2819,
      "mean_token_accuracy": 0.20838777720928192,
      "num_tokens": 54359859.0,
      "step": 26825
    },
    {
      "entropy": 5.979834175109863,
      "epoch": 2.3446648606134755,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004448323685614021,
      "loss": 5.2569,
      "mean_token_accuracy": 0.2044917479157448,
      "num_tokens": 54369386.0,
      "step": 26830
    },
    {
      "entropy": 5.946871662139893,
      "epoch": 2.3451018089661804,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044481192703220695,
      "loss": 5.221,
      "mean_token_accuracy": 0.20628986060619353,
      "num_tokens": 54379250.0,
      "step": 26835
    },
    {
      "entropy": 5.995414972305298,
      "epoch": 2.345538757318885,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044479148224593014,
      "loss": 5.3023,
      "mean_token_accuracy": 0.19839154481887816,
      "num_tokens": 54388980.0,
      "step": 26840
    },
    {
      "entropy": 5.880949831008911,
      "epoch": 2.3459757056715898,
      "grad_norm": 1.34375,
      "learning_rate": 0.00044477103420296365,
      "loss": 5.1472,
      "mean_token_accuracy": 0.21698672324419022,
      "num_tokens": 54398849.0,
      "step": 26845
    },
    {
      "entropy": 6.030484342575074,
      "epoch": 2.346412654024294,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004447505829036998,
      "loss": 5.315,
      "mean_token_accuracy": 0.2032279208302498,
      "num_tokens": 54409426.0,
      "step": 26850
    },
    {
      "entropy": 6.025615882873535,
      "epoch": 2.346849602376999,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004447301283485308,
      "loss": 5.3193,
      "mean_token_accuracy": 0.2040020301938057,
      "num_tokens": 54420212.0,
      "step": 26855
    },
    {
      "entropy": 5.985171890258789,
      "epoch": 2.3472865507297036,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044470967053784904,
      "loss": 5.3933,
      "mean_token_accuracy": 0.19423657208681105,
      "num_tokens": 54431179.0,
      "step": 26860
    },
    {
      "entropy": 5.984085941314698,
      "epoch": 2.3477234990824085,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004446892094720468,
      "loss": 5.2839,
      "mean_token_accuracy": 0.20956086814403535,
      "num_tokens": 54442247.0,
      "step": 26865
    },
    {
      "entropy": 6.0092212677001955,
      "epoch": 2.348160447435113,
      "grad_norm": 1.125,
      "learning_rate": 0.00044466874515151667,
      "loss": 5.241,
      "mean_token_accuracy": 0.20793827325105668,
      "num_tokens": 54451475.0,
      "step": 26870
    },
    {
      "entropy": 5.9848634719848635,
      "epoch": 2.348597395787818,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000444648277576651,
      "loss": 5.3628,
      "mean_token_accuracy": 0.1966217651963234,
      "num_tokens": 54463259.0,
      "step": 26875
    },
    {
      "entropy": 5.934948205947876,
      "epoch": 2.349034344140523,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044462780674784264,
      "loss": 5.2739,
      "mean_token_accuracy": 0.20516833066940307,
      "num_tokens": 54473573.0,
      "step": 26880
    },
    {
      "entropy": 6.1028179168701175,
      "epoch": 2.3494712924932273,
      "grad_norm": 1.25,
      "learning_rate": 0.000444607332665484,
      "loss": 5.3885,
      "mean_token_accuracy": 0.19680129140615463,
      "num_tokens": 54482907.0,
      "step": 26885
    },
    {
      "entropy": 6.012430477142334,
      "epoch": 2.349908240845932,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004445868553299678,
      "loss": 5.312,
      "mean_token_accuracy": 0.20593286752700807,
      "num_tokens": 54492993.0,
      "step": 26890
    },
    {
      "entropy": 5.964368581771851,
      "epoch": 2.3503451891986367,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004445663747416868,
      "loss": 5.3236,
      "mean_token_accuracy": 0.19334489554166795,
      "num_tokens": 54503826.0,
      "step": 26895
    },
    {
      "entropy": 5.964860343933106,
      "epoch": 2.3507821375513416,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00044454589090103406,
      "loss": 5.3138,
      "mean_token_accuracy": 0.2027054563164711,
      "num_tokens": 54513839.0,
      "step": 26900
    },
    {
      "entropy": 5.993250370025635,
      "epoch": 2.351219085904046,
      "grad_norm": 1.125,
      "learning_rate": 0.00044452540380840217,
      "loss": 5.2231,
      "mean_token_accuracy": 0.20615014135837556,
      "num_tokens": 54523742.0,
      "step": 26905
    },
    {
      "entropy": 6.091670846939087,
      "epoch": 2.351656034256751,
      "grad_norm": 1.359375,
      "learning_rate": 0.00044450491346418426,
      "loss": 5.3305,
      "mean_token_accuracy": 0.19553127139806747,
      "num_tokens": 54534017.0,
      "step": 26910
    },
    {
      "entropy": 5.9463303565979,
      "epoch": 2.3520929826094554,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004444844198687732,
      "loss": 5.2297,
      "mean_token_accuracy": 0.2066887930035591,
      "num_tokens": 54544445.0,
      "step": 26915
    },
    {
      "entropy": 5.916622924804687,
      "epoch": 2.3525299309621603,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004444639230225622,
      "loss": 5.2432,
      "mean_token_accuracy": 0.20301828533411026,
      "num_tokens": 54554966.0,
      "step": 26920
    },
    {
      "entropy": 5.998639249801636,
      "epoch": 2.352966879314865,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004444434229259444,
      "loss": 5.2434,
      "mean_token_accuracy": 0.20636873096227645,
      "num_tokens": 54565005.0,
      "step": 26925
    },
    {
      "entropy": 6.007830381393433,
      "epoch": 2.3534038276675697,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004444229195793129,
      "loss": 5.2852,
      "mean_token_accuracy": 0.20270363092422486,
      "num_tokens": 54575242.0,
      "step": 26930
    },
    {
      "entropy": 5.994860887527466,
      "epoch": 2.3538407760202746,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004444024129830611,
      "loss": 5.2916,
      "mean_token_accuracy": 0.19562040269374847,
      "num_tokens": 54585448.0,
      "step": 26935
    },
    {
      "entropy": 5.904260921478271,
      "epoch": 2.354277724372979,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004443819031375822,
      "loss": 5.2261,
      "mean_token_accuracy": 0.21159045100212098,
      "num_tokens": 54595891.0,
      "step": 26940
    },
    {
      "entropy": 6.01039309501648,
      "epoch": 2.3547146727256836,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004443613900432697,
      "loss": 5.3424,
      "mean_token_accuracy": 0.1951133653521538,
      "num_tokens": 54607407.0,
      "step": 26945
    },
    {
      "entropy": 6.049400758743286,
      "epoch": 2.3551516210783885,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044434087370051693,
      "loss": 5.2997,
      "mean_token_accuracy": 0.20127919614315032,
      "num_tokens": 54617337.0,
      "step": 26950
    },
    {
      "entropy": 6.006165456771851,
      "epoch": 2.3555885694310934,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004443203541097175,
      "loss": 5.2852,
      "mean_token_accuracy": 0.2023713305592537,
      "num_tokens": 54627914.0,
      "step": 26955
    },
    {
      "entropy": 5.967828273773193,
      "epoch": 2.356025517783798,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004442998312712649,
      "loss": 5.1915,
      "mean_token_accuracy": 0.20523437410593032,
      "num_tokens": 54636709.0,
      "step": 26960
    },
    {
      "entropy": 5.911697435379028,
      "epoch": 2.356462466136503,
      "grad_norm": 1.125,
      "learning_rate": 0.00044427930518555294,
      "loss": 5.2265,
      "mean_token_accuracy": 0.20489410310983658,
      "num_tokens": 54646167.0,
      "step": 26965
    },
    {
      "entropy": 6.003185224533081,
      "epoch": 2.3568994144892073,
      "grad_norm": 1.125,
      "learning_rate": 0.0004442587758529751,
      "loss": 5.3069,
      "mean_token_accuracy": 0.19936318099498748,
      "num_tokens": 54656663.0,
      "step": 26970
    },
    {
      "entropy": 6.008269119262695,
      "epoch": 2.357336362841912,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004442382432739253,
      "loss": 5.3028,
      "mean_token_accuracy": 0.19892603904008865,
      "num_tokens": 54667123.0,
      "step": 26975
    },
    {
      "entropy": 6.000832557678223,
      "epoch": 2.3577733111946166,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044421770744879725,
      "loss": 5.2824,
      "mean_token_accuracy": 0.20455645769834518,
      "num_tokens": 54677381.0,
      "step": 26980
    },
    {
      "entropy": 5.966068172454834,
      "epoch": 2.3582102595473216,
      "grad_norm": 1.1875,
      "learning_rate": 0.000444197168377985,
      "loss": 5.246,
      "mean_token_accuracy": 0.20683507174253463,
      "num_tokens": 54687190.0,
      "step": 26985
    },
    {
      "entropy": 6.03504114151001,
      "epoch": 2.3586472079000265,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00044417662606188236,
      "loss": 5.3258,
      "mean_token_accuracy": 0.20118242204189302,
      "num_tokens": 54697102.0,
      "step": 26990
    },
    {
      "entropy": 5.989965200424194,
      "epoch": 2.359084156252731,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004441560805008834,
      "loss": 5.2614,
      "mean_token_accuracy": 0.20483358055353165,
      "num_tokens": 54707315.0,
      "step": 26995
    },
    {
      "entropy": 6.006865787506103,
      "epoch": 2.3595211046054354,
      "grad_norm": 1.25,
      "learning_rate": 0.0004441355316953822,
      "loss": 5.3392,
      "mean_token_accuracy": 0.19471132308244704,
      "num_tokens": 54717873.0,
      "step": 27000
    },
    {
      "epoch": 2.3595211046054354,
      "eval_entropy": 5.72592621919324,
      "eval_loss": 5.330050945281982,
      "eval_mean_token_accuracy": 0.209662055484395,
      "eval_num_tokens": 54717873.0,
      "eval_runtime": 21.7127,
      "eval_samples_per_second": 1704.72,
      "eval_steps_per_second": 213.101,
      "step": 27000
    },
    {
      "entropy": 6.017616319656372,
      "epoch": 2.3599580529581403,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004441149796457728,
      "loss": 5.3482,
      "mean_token_accuracy": 0.19908980131149293,
      "num_tokens": 54727723.0,
      "step": 27005
    },
    {
      "entropy": 5.986052513122559,
      "epoch": 2.3603950013108452,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044409442435244947,
      "loss": 5.2505,
      "mean_token_accuracy": 0.20926570743322373,
      "num_tokens": 54738108.0,
      "step": 27010
    },
    {
      "entropy": 5.999160432815552,
      "epoch": 2.3608319496635497,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004440738658158065,
      "loss": 5.3506,
      "mean_token_accuracy": 0.19765647649765014,
      "num_tokens": 54749011.0,
      "step": 27015
    },
    {
      "entropy": 6.049213027954101,
      "epoch": 2.3612688980162546,
      "grad_norm": 1.125,
      "learning_rate": 0.0004440533040362382,
      "loss": 5.3023,
      "mean_token_accuracy": 0.2003503143787384,
      "num_tokens": 54759922.0,
      "step": 27020
    },
    {
      "entropy": 5.993977308273315,
      "epoch": 2.361705846368959,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044403273901413897,
      "loss": 5.3025,
      "mean_token_accuracy": 0.19645484983921052,
      "num_tokens": 54770134.0,
      "step": 27025
    },
    {
      "entropy": 6.0027570724487305,
      "epoch": 2.362142794721664,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044401217074990323,
      "loss": 5.2373,
      "mean_token_accuracy": 0.19907864779233933,
      "num_tokens": 54780131.0,
      "step": 27030
    },
    {
      "entropy": 6.005213785171509,
      "epoch": 2.3625797430743685,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044399159924392545,
      "loss": 5.3339,
      "mean_token_accuracy": 0.199359892308712,
      "num_tokens": 54790601.0,
      "step": 27035
    },
    {
      "entropy": 6.0193366527557375,
      "epoch": 2.3630166914270734,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004439710244966003,
      "loss": 5.2621,
      "mean_token_accuracy": 0.20804544240236283,
      "num_tokens": 54799995.0,
      "step": 27040
    },
    {
      "entropy": 6.022313070297241,
      "epoch": 2.363453639779778,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004439504465083223,
      "loss": 5.25,
      "mean_token_accuracy": 0.20103181302547454,
      "num_tokens": 54810299.0,
      "step": 27045
    },
    {
      "entropy": 6.0127684593200685,
      "epoch": 2.3638905881324828,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004439298652794863,
      "loss": 5.2429,
      "mean_token_accuracy": 0.2000145748257637,
      "num_tokens": 54820802.0,
      "step": 27050
    },
    {
      "entropy": 5.954245281219483,
      "epoch": 2.3643275364851872,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044390928081048693,
      "loss": 5.2584,
      "mean_token_accuracy": 0.20705511868000032,
      "num_tokens": 54831077.0,
      "step": 27055
    },
    {
      "entropy": 5.965565299987793,
      "epoch": 2.364764484837892,
      "grad_norm": 1.171875,
      "learning_rate": 0.000443888693101719,
      "loss": 5.2601,
      "mean_token_accuracy": 0.20559504628181458,
      "num_tokens": 54842574.0,
      "step": 27060
    },
    {
      "entropy": 5.945895481109619,
      "epoch": 2.365201433190597,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044386810215357747,
      "loss": 5.1867,
      "mean_token_accuracy": 0.21190289705991744,
      "num_tokens": 54852365.0,
      "step": 27065
    },
    {
      "entropy": 6.0058423519134525,
      "epoch": 2.3656383815433015,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044384750796645724,
      "loss": 5.273,
      "mean_token_accuracy": 0.20176127851009368,
      "num_tokens": 54863008.0,
      "step": 27070
    },
    {
      "entropy": 5.9978681087493895,
      "epoch": 2.3660753298960064,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004438269105407534,
      "loss": 5.2677,
      "mean_token_accuracy": 0.2003133177757263,
      "num_tokens": 54873134.0,
      "step": 27075
    },
    {
      "entropy": 6.024529886245728,
      "epoch": 2.366512278248711,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044380630987686086,
      "loss": 5.3194,
      "mean_token_accuracy": 0.2024843156337738,
      "num_tokens": 54883075.0,
      "step": 27080
    },
    {
      "entropy": 5.9680665016174315,
      "epoch": 2.366949226601416,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044378570597517495,
      "loss": 5.2673,
      "mean_token_accuracy": 0.20055705159902573,
      "num_tokens": 54892788.0,
      "step": 27085
    },
    {
      "entropy": 5.914882802963257,
      "epoch": 2.3673861749541203,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004437650988360906,
      "loss": 5.2317,
      "mean_token_accuracy": 0.20982354879379272,
      "num_tokens": 54902708.0,
      "step": 27090
    },
    {
      "entropy": 5.955168867111206,
      "epoch": 2.367823123306825,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004437444884600034,
      "loss": 5.2445,
      "mean_token_accuracy": 0.20770573019981384,
      "num_tokens": 54912906.0,
      "step": 27095
    },
    {
      "entropy": 5.956963682174683,
      "epoch": 2.3682600716595297,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044372387484730835,
      "loss": 5.294,
      "mean_token_accuracy": 0.1970170632004738,
      "num_tokens": 54922802.0,
      "step": 27100
    },
    {
      "entropy": 6.022427749633789,
      "epoch": 2.3686970200122346,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044370325799840106,
      "loss": 5.3129,
      "mean_token_accuracy": 0.19951269328594207,
      "num_tokens": 54933203.0,
      "step": 27105
    },
    {
      "entropy": 6.020863437652588,
      "epoch": 2.369133968364939,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044368263791367684,
      "loss": 5.2986,
      "mean_token_accuracy": 0.19809829890727998,
      "num_tokens": 54944081.0,
      "step": 27110
    },
    {
      "entropy": 6.0339234352111815,
      "epoch": 2.369570916717644,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044366201459353124,
      "loss": 5.2774,
      "mean_token_accuracy": 0.2010866165161133,
      "num_tokens": 54953421.0,
      "step": 27115
    },
    {
      "entropy": 5.915641403198242,
      "epoch": 2.370007865070349,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044364138803835984,
      "loss": 5.2633,
      "mean_token_accuracy": 0.20463054329156877,
      "num_tokens": 54964127.0,
      "step": 27120
    },
    {
      "entropy": 5.940655088424682,
      "epoch": 2.3704448134230534,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044362075824855816,
      "loss": 5.2358,
      "mean_token_accuracy": 0.204831101000309,
      "num_tokens": 54974636.0,
      "step": 27125
    },
    {
      "entropy": 6.089567613601685,
      "epoch": 2.3708817617757583,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044360012522452215,
      "loss": 5.3267,
      "mean_token_accuracy": 0.19494712203741074,
      "num_tokens": 54985542.0,
      "step": 27130
    },
    {
      "entropy": 6.031984090805054,
      "epoch": 2.3713187101284627,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004435794889666472,
      "loss": 5.2444,
      "mean_token_accuracy": 0.21003260761499404,
      "num_tokens": 54994973.0,
      "step": 27135
    },
    {
      "entropy": 5.955686044692993,
      "epoch": 2.3717556584811677,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044355884947532947,
      "loss": 5.356,
      "mean_token_accuracy": 0.1974317565560341,
      "num_tokens": 55006510.0,
      "step": 27140
    },
    {
      "entropy": 5.9748434066772464,
      "epoch": 2.372192606833872,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004435382067509647,
      "loss": 5.3489,
      "mean_token_accuracy": 0.2014079809188843,
      "num_tokens": 55016375.0,
      "step": 27145
    },
    {
      "entropy": 5.967666339874268,
      "epoch": 2.372629555186577,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044351756079394865,
      "loss": 5.2832,
      "mean_token_accuracy": 0.20414239764213563,
      "num_tokens": 55026213.0,
      "step": 27150
    },
    {
      "entropy": 6.054480457305909,
      "epoch": 2.3730665035392815,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004434969116046776,
      "loss": 5.3346,
      "mean_token_accuracy": 0.19741796106100082,
      "num_tokens": 55036836.0,
      "step": 27155
    },
    {
      "entropy": 6.003956079483032,
      "epoch": 2.3735034518919864,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004434762591835475,
      "loss": 5.2346,
      "mean_token_accuracy": 0.19876306504011154,
      "num_tokens": 55046243.0,
      "step": 27160
    },
    {
      "entropy": 5.961802577972412,
      "epoch": 2.373940400244691,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044345560353095437,
      "loss": 5.2722,
      "mean_token_accuracy": 0.20015566051006317,
      "num_tokens": 55055882.0,
      "step": 27165
    },
    {
      "entropy": 5.924121141433716,
      "epoch": 2.374377348597396,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004434349446472946,
      "loss": 5.1982,
      "mean_token_accuracy": 0.20726975053548813,
      "num_tokens": 55066398.0,
      "step": 27170
    },
    {
      "entropy": 6.026009654998779,
      "epoch": 2.3748142969501007,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044341428253296436,
      "loss": 5.3486,
      "mean_token_accuracy": 0.19728484600782395,
      "num_tokens": 55077053.0,
      "step": 27175
    },
    {
      "entropy": 6.015435457229614,
      "epoch": 2.375251245302805,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004433936171883598,
      "loss": 5.2608,
      "mean_token_accuracy": 0.2078452527523041,
      "num_tokens": 55086973.0,
      "step": 27180
    },
    {
      "entropy": 5.997517538070679,
      "epoch": 2.37568819365551,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004433729486138775,
      "loss": 5.2291,
      "mean_token_accuracy": 0.20740010738372802,
      "num_tokens": 55096162.0,
      "step": 27185
    },
    {
      "entropy": 5.94533405303955,
      "epoch": 2.3761251420082146,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044335227680991385,
      "loss": 5.2386,
      "mean_token_accuracy": 0.20711229294538497,
      "num_tokens": 55105852.0,
      "step": 27190
    },
    {
      "entropy": 5.924583578109742,
      "epoch": 2.3765620903609195,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004433316017768653,
      "loss": 5.1936,
      "mean_token_accuracy": 0.21455229222774505,
      "num_tokens": 55116092.0,
      "step": 27195
    },
    {
      "entropy": 6.053944396972656,
      "epoch": 2.376999038713624,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004433109235151284,
      "loss": 5.2444,
      "mean_token_accuracy": 0.21060768365859986,
      "num_tokens": 55125439.0,
      "step": 27200
    },
    {
      "entropy": 6.005931043624878,
      "epoch": 2.377435987066329,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044329024202509977,
      "loss": 5.3007,
      "mean_token_accuracy": 0.20088419020175935,
      "num_tokens": 55134684.0,
      "step": 27205
    },
    {
      "entropy": 5.929457664489746,
      "epoch": 2.3778729354190333,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044326955730717614,
      "loss": 5.2183,
      "mean_token_accuracy": 0.20725370049476624,
      "num_tokens": 55143853.0,
      "step": 27210
    },
    {
      "entropy": 6.075354146957397,
      "epoch": 2.3783098837717382,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004432488693617542,
      "loss": 5.4105,
      "mean_token_accuracy": 0.19055642038583756,
      "num_tokens": 55154190.0,
      "step": 27215
    },
    {
      "entropy": 6.042052698135376,
      "epoch": 2.3787468321244427,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044322817818923083,
      "loss": 5.3234,
      "mean_token_accuracy": 0.20635790526866912,
      "num_tokens": 55164158.0,
      "step": 27220
    },
    {
      "entropy": 6.006984376907349,
      "epoch": 2.3791837804771476,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00044320748379000284,
      "loss": 5.3692,
      "mean_token_accuracy": 0.19077155143022537,
      "num_tokens": 55173635.0,
      "step": 27225
    },
    {
      "entropy": 6.008430051803589,
      "epoch": 2.3796207288298525,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044318678616446717,
      "loss": 5.277,
      "mean_token_accuracy": 0.2101149633526802,
      "num_tokens": 55182972.0,
      "step": 27230
    },
    {
      "entropy": 6.039464521408081,
      "epoch": 2.380057677182557,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044316608531302085,
      "loss": 5.2889,
      "mean_token_accuracy": 0.20352879166603088,
      "num_tokens": 55192769.0,
      "step": 27235
    },
    {
      "entropy": 5.999217081069946,
      "epoch": 2.380494625535262,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044314538123606083,
      "loss": 5.3069,
      "mean_token_accuracy": 0.19983949214220048,
      "num_tokens": 55203241.0,
      "step": 27240
    },
    {
      "entropy": 5.893173456192017,
      "epoch": 2.3809315738879664,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044312467393398434,
      "loss": 5.2408,
      "mean_token_accuracy": 0.21061796844005584,
      "num_tokens": 55214272.0,
      "step": 27245
    },
    {
      "entropy": 5.947113656997681,
      "epoch": 2.3813685222406713,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044310396340718853,
      "loss": 5.2136,
      "mean_token_accuracy": 0.20764501690864562,
      "num_tokens": 55223947.0,
      "step": 27250
    },
    {
      "entropy": 6.000958871841431,
      "epoch": 2.3818054705933758,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044308324965607054,
      "loss": 5.337,
      "mean_token_accuracy": 0.19580454975366593,
      "num_tokens": 55234906.0,
      "step": 27255
    },
    {
      "entropy": 5.963118839263916,
      "epoch": 2.3822424189460807,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004430625326810278,
      "loss": 5.1609,
      "mean_token_accuracy": 0.21102575808763505,
      "num_tokens": 55244785.0,
      "step": 27260
    },
    {
      "entropy": 5.98688440322876,
      "epoch": 2.382679367298785,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044304181248245754,
      "loss": 5.3638,
      "mean_token_accuracy": 0.1987767145037651,
      "num_tokens": 55255053.0,
      "step": 27265
    },
    {
      "entropy": 5.972701501846314,
      "epoch": 2.38311631565149,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004430210890607574,
      "loss": 5.3168,
      "mean_token_accuracy": 0.20168333798646926,
      "num_tokens": 55264525.0,
      "step": 27270
    },
    {
      "entropy": 6.012744474411011,
      "epoch": 2.3835532640041945,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044300036241632463,
      "loss": 5.285,
      "mean_token_accuracy": 0.20157212316989898,
      "num_tokens": 55273486.0,
      "step": 27275
    },
    {
      "entropy": 6.003059673309326,
      "epoch": 2.3839902123568995,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044297963254955687,
      "loss": 5.2383,
      "mean_token_accuracy": 0.20783497542142867,
      "num_tokens": 55283164.0,
      "step": 27280
    },
    {
      "entropy": 6.020606565475464,
      "epoch": 2.3844271607096044,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044295889946085185,
      "loss": 5.3033,
      "mean_token_accuracy": 0.2035932496190071,
      "num_tokens": 55294657.0,
      "step": 27285
    },
    {
      "entropy": 6.021960258483887,
      "epoch": 2.384864109062309,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044293816315060704,
      "loss": 5.2891,
      "mean_token_accuracy": 0.19776412099599838,
      "num_tokens": 55304743.0,
      "step": 27290
    },
    {
      "entropy": 5.948904228210449,
      "epoch": 2.3853010574150133,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004429174236192203,
      "loss": 5.2513,
      "mean_token_accuracy": 0.21736903637647628,
      "num_tokens": 55314106.0,
      "step": 27295
    },
    {
      "entropy": 5.9962927341461185,
      "epoch": 2.385738005767718,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044289668086708935,
      "loss": 5.27,
      "mean_token_accuracy": 0.20801898390054702,
      "num_tokens": 55324407.0,
      "step": 27300
    },
    {
      "entropy": 5.9604206562042235,
      "epoch": 2.386174954120423,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044287593489461205,
      "loss": 5.2808,
      "mean_token_accuracy": 0.20465119183063507,
      "num_tokens": 55335853.0,
      "step": 27305
    },
    {
      "entropy": 6.006997728347779,
      "epoch": 2.3866119024731276,
      "grad_norm": 1.3125,
      "learning_rate": 0.00044285518570218635,
      "loss": 5.248,
      "mean_token_accuracy": 0.20608635544776915,
      "num_tokens": 55345947.0,
      "step": 27310
    },
    {
      "entropy": 6.010674238204956,
      "epoch": 2.3870488508258325,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004428344332902103,
      "loss": 5.2905,
      "mean_token_accuracy": 0.20702181607484818,
      "num_tokens": 55355709.0,
      "step": 27315
    },
    {
      "entropy": 5.998511409759521,
      "epoch": 2.387485799178537,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044281367765908186,
      "loss": 5.2176,
      "mean_token_accuracy": 0.20216851234436034,
      "num_tokens": 55365597.0,
      "step": 27320
    },
    {
      "entropy": 5.995793581008911,
      "epoch": 2.387922747531242,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00044279291880919915,
      "loss": 5.298,
      "mean_token_accuracy": 0.19898535460233688,
      "num_tokens": 55375508.0,
      "step": 27325
    },
    {
      "entropy": 5.969800996780395,
      "epoch": 2.3883596958839464,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044277215674096035,
      "loss": 5.2766,
      "mean_token_accuracy": 0.20457002967596055,
      "num_tokens": 55384928.0,
      "step": 27330
    },
    {
      "entropy": 6.022192478179932,
      "epoch": 2.3887966442366513,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004427513914547637,
      "loss": 5.2556,
      "mean_token_accuracy": 0.21045242995023727,
      "num_tokens": 55394805.0,
      "step": 27335
    },
    {
      "entropy": 6.01785364151001,
      "epoch": 2.3892335925893557,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004427306229510074,
      "loss": 5.321,
      "mean_token_accuracy": 0.20145614296197892,
      "num_tokens": 55405360.0,
      "step": 27340
    },
    {
      "entropy": 6.006945848464966,
      "epoch": 2.3896705409420607,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044270985123008984,
      "loss": 5.266,
      "mean_token_accuracy": 0.19296927899122238,
      "num_tokens": 55415124.0,
      "step": 27345
    },
    {
      "entropy": 5.995441198348999,
      "epoch": 2.390107489294765,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004426890762924096,
      "loss": 5.2605,
      "mean_token_accuracy": 0.20920797884464265,
      "num_tokens": 55425621.0,
      "step": 27350
    },
    {
      "entropy": 6.059401988983154,
      "epoch": 2.39054443764747,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004426682981383648,
      "loss": 5.3305,
      "mean_token_accuracy": 0.19722647070884705,
      "num_tokens": 55435812.0,
      "step": 27355
    },
    {
      "entropy": 5.995691251754761,
      "epoch": 2.390981386000175,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004426475167683543,
      "loss": 5.2957,
      "mean_token_accuracy": 0.20203914642333984,
      "num_tokens": 55445462.0,
      "step": 27360
    },
    {
      "entropy": 5.989670848846435,
      "epoch": 2.3914183343528794,
      "grad_norm": 1.125,
      "learning_rate": 0.00044262673218277654,
      "loss": 5.2877,
      "mean_token_accuracy": 0.19929719120264053,
      "num_tokens": 55455992.0,
      "step": 27365
    },
    {
      "entropy": 6.051963472366333,
      "epoch": 2.3918552827055843,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044260594438203025,
      "loss": 5.339,
      "mean_token_accuracy": 0.1956098884344101,
      "num_tokens": 55466715.0,
      "step": 27370
    },
    {
      "entropy": 6.019154644012451,
      "epoch": 2.392292231058289,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044258515336651416,
      "loss": 5.3091,
      "mean_token_accuracy": 0.2026679262518883,
      "num_tokens": 55476280.0,
      "step": 27375
    },
    {
      "entropy": 5.976699304580689,
      "epoch": 2.3927291794109937,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044256435913662696,
      "loss": 5.2625,
      "mean_token_accuracy": 0.1995115950703621,
      "num_tokens": 55486105.0,
      "step": 27380
    },
    {
      "entropy": 6.040324306488037,
      "epoch": 2.393166127763698,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004425435616927676,
      "loss": 5.266,
      "mean_token_accuracy": 0.20166035741567612,
      "num_tokens": 55496152.0,
      "step": 27385
    },
    {
      "entropy": 5.967291259765625,
      "epoch": 2.393603076116403,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004425227610353349,
      "loss": 5.2849,
      "mean_token_accuracy": 0.2069743975996971,
      "num_tokens": 55507104.0,
      "step": 27390
    },
    {
      "entropy": 5.954477548599243,
      "epoch": 2.3940400244691076,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004425019571647279,
      "loss": 5.2911,
      "mean_token_accuracy": 0.20369015783071517,
      "num_tokens": 55517114.0,
      "step": 27395
    },
    {
      "entropy": 5.923276042938232,
      "epoch": 2.3944769728218125,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044248115008134557,
      "loss": 5.1986,
      "mean_token_accuracy": 0.2130338117480278,
      "num_tokens": 55526719.0,
      "step": 27400
    },
    {
      "entropy": 5.983892583847046,
      "epoch": 2.394913921174517,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044246033978558697,
      "loss": 5.2415,
      "mean_token_accuracy": 0.20747818648815156,
      "num_tokens": 55536760.0,
      "step": 27405
    },
    {
      "entropy": 5.9937989711761475,
      "epoch": 2.395350869527222,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044243952627785136,
      "loss": 5.2701,
      "mean_token_accuracy": 0.20246149599552155,
      "num_tokens": 55547193.0,
      "step": 27410
    },
    {
      "entropy": 5.941078472137451,
      "epoch": 2.395787817879927,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044241870955853793,
      "loss": 5.2683,
      "mean_token_accuracy": 0.2026207685470581,
      "num_tokens": 55557257.0,
      "step": 27415
    },
    {
      "entropy": 5.966141986846924,
      "epoch": 2.3962247662326313,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044239788962804584,
      "loss": 5.2447,
      "mean_token_accuracy": 0.2081846132874489,
      "num_tokens": 55568191.0,
      "step": 27420
    },
    {
      "entropy": 5.984782314300537,
      "epoch": 2.396661714585336,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044237706648677456,
      "loss": 5.2618,
      "mean_token_accuracy": 0.20447543412446975,
      "num_tokens": 55578983.0,
      "step": 27425
    },
    {
      "entropy": 5.963319921493531,
      "epoch": 2.3970986629380406,
      "grad_norm": 1.25,
      "learning_rate": 0.0004423562401351234,
      "loss": 5.2472,
      "mean_token_accuracy": 0.20658429712057114,
      "num_tokens": 55588772.0,
      "step": 27430
    },
    {
      "entropy": 5.933617496490479,
      "epoch": 2.3975356112907455,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004423354105734919,
      "loss": 5.2679,
      "mean_token_accuracy": 0.2062872216105461,
      "num_tokens": 55599514.0,
      "step": 27435
    },
    {
      "entropy": 5.986964988708496,
      "epoch": 2.39797255964345,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004423145778022795,
      "loss": 5.2902,
      "mean_token_accuracy": 0.1965571627020836,
      "num_tokens": 55609183.0,
      "step": 27440
    },
    {
      "entropy": 6.042706775665283,
      "epoch": 2.398409507996155,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004422937418218859,
      "loss": 5.2927,
      "mean_token_accuracy": 0.20152009576559066,
      "num_tokens": 55618836.0,
      "step": 27445
    },
    {
      "entropy": 6.009373950958252,
      "epoch": 2.3988464563488594,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00044227290263271066,
      "loss": 5.318,
      "mean_token_accuracy": 0.19921126663684846,
      "num_tokens": 55629314.0,
      "step": 27450
    },
    {
      "entropy": 5.9197587966918945,
      "epoch": 2.3992834047015643,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004422520602351535,
      "loss": 5.1856,
      "mean_token_accuracy": 0.21892755925655366,
      "num_tokens": 55640542.0,
      "step": 27455
    },
    {
      "entropy": 6.021482038497925,
      "epoch": 2.399720353054269,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004422312146296141,
      "loss": 5.3579,
      "mean_token_accuracy": 0.19984910637140274,
      "num_tokens": 55650165.0,
      "step": 27460
    },
    {
      "entropy": 6.043282604217529,
      "epoch": 2.4001573014069737,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044221036581649245,
      "loss": 5.3031,
      "mean_token_accuracy": 0.19876542836427688,
      "num_tokens": 55658962.0,
      "step": 27465
    },
    {
      "entropy": 6.022459554672241,
      "epoch": 2.4005942497596786,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004421895137961884,
      "loss": 5.2771,
      "mean_token_accuracy": 0.1945583090186119,
      "num_tokens": 55668357.0,
      "step": 27470
    },
    {
      "entropy": 6.02305383682251,
      "epoch": 2.401031198112383,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044216865856910176,
      "loss": 5.3617,
      "mean_token_accuracy": 0.20274650752544404,
      "num_tokens": 55678313.0,
      "step": 27475
    },
    {
      "entropy": 6.007496690750122,
      "epoch": 2.401468146465088,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044214780013563277,
      "loss": 5.3081,
      "mean_token_accuracy": 0.19820630103349685,
      "num_tokens": 55689332.0,
      "step": 27480
    },
    {
      "entropy": 5.993176031112671,
      "epoch": 2.4019050948177925,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044212693849618136,
      "loss": 5.2628,
      "mean_token_accuracy": 0.20707063972949982,
      "num_tokens": 55699397.0,
      "step": 27485
    },
    {
      "entropy": 6.045131254196167,
      "epoch": 2.4023420431704974,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004421060736511476,
      "loss": 5.3461,
      "mean_token_accuracy": 0.19666240513324737,
      "num_tokens": 55708548.0,
      "step": 27490
    },
    {
      "entropy": 6.079915523529053,
      "epoch": 2.402778991523202,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044208520560093183,
      "loss": 5.325,
      "mean_token_accuracy": 0.1974760726094246,
      "num_tokens": 55718043.0,
      "step": 27495
    },
    {
      "entropy": 5.984104633331299,
      "epoch": 2.4032159398759068,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004420643343459343,
      "loss": 5.2944,
      "mean_token_accuracy": 0.19999710768461226,
      "num_tokens": 55727889.0,
      "step": 27500
    },
    {
      "entropy": 5.966103410720825,
      "epoch": 2.4036528882286112,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004420434598865553,
      "loss": 5.3233,
      "mean_token_accuracy": 0.20637630969285964,
      "num_tokens": 55738993.0,
      "step": 27505
    },
    {
      "entropy": 6.003294229507446,
      "epoch": 2.404089836581316,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004420225822231952,
      "loss": 5.3184,
      "mean_token_accuracy": 0.20091705918312072,
      "num_tokens": 55748858.0,
      "step": 27510
    },
    {
      "entropy": 5.9829692363739015,
      "epoch": 2.4045267849340206,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044200170135625446,
      "loss": 5.237,
      "mean_token_accuracy": 0.2057405725121498,
      "num_tokens": 55758639.0,
      "step": 27515
    },
    {
      "entropy": 6.046299648284912,
      "epoch": 2.4049637332867255,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004419808172861336,
      "loss": 5.3212,
      "mean_token_accuracy": 0.20568892806768418,
      "num_tokens": 55769632.0,
      "step": 27520
    },
    {
      "entropy": 6.084587526321411,
      "epoch": 2.4054006816394304,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004419599300132331,
      "loss": 5.329,
      "mean_token_accuracy": 0.1970178872346878,
      "num_tokens": 55780147.0,
      "step": 27525
    },
    {
      "entropy": 5.962748098373413,
      "epoch": 2.405837629992135,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044193903953795367,
      "loss": 5.2283,
      "mean_token_accuracy": 0.2096414402127266,
      "num_tokens": 55790933.0,
      "step": 27530
    },
    {
      "entropy": 5.927871417999268,
      "epoch": 2.40627457834484,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044191814586069605,
      "loss": 5.2305,
      "mean_token_accuracy": 0.2105153203010559,
      "num_tokens": 55800479.0,
      "step": 27535
    },
    {
      "entropy": 5.987896728515625,
      "epoch": 2.4067115266975443,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004418972489818609,
      "loss": 5.2548,
      "mean_token_accuracy": 0.2116370141506195,
      "num_tokens": 55810613.0,
      "step": 27540
    },
    {
      "entropy": 6.029828405380249,
      "epoch": 2.407148475050249,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044187634890184904,
      "loss": 5.3212,
      "mean_token_accuracy": 0.19973108619451524,
      "num_tokens": 55821507.0,
      "step": 27545
    },
    {
      "entropy": 5.991740417480469,
      "epoch": 2.4075854234029537,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004418554456210614,
      "loss": 5.3583,
      "mean_token_accuracy": 0.19498565346002578,
      "num_tokens": 55831878.0,
      "step": 27550
    },
    {
      "entropy": 5.9876025199890135,
      "epoch": 2.4080223717556586,
      "grad_norm": 1.125,
      "learning_rate": 0.00044183453913989885,
      "loss": 5.2301,
      "mean_token_accuracy": 0.20757532268762588,
      "num_tokens": 55841314.0,
      "step": 27555
    },
    {
      "entropy": 6.037330389022827,
      "epoch": 2.408459320108363,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004418136294587624,
      "loss": 5.2743,
      "mean_token_accuracy": 0.2049335151910782,
      "num_tokens": 55851056.0,
      "step": 27560
    },
    {
      "entropy": 5.958216142654419,
      "epoch": 2.408896268461068,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004417927165780531,
      "loss": 5.2394,
      "mean_token_accuracy": 0.20160573869943618,
      "num_tokens": 55860705.0,
      "step": 27565
    },
    {
      "entropy": 5.972125673294068,
      "epoch": 2.4093332168137724,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004417718004981723,
      "loss": 5.3022,
      "mean_token_accuracy": 0.20144228786230087,
      "num_tokens": 55870749.0,
      "step": 27570
    },
    {
      "entropy": 6.060548782348633,
      "epoch": 2.4097701651664774,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004417508812195207,
      "loss": 5.3623,
      "mean_token_accuracy": 0.19606729447841645,
      "num_tokens": 55880135.0,
      "step": 27575
    },
    {
      "entropy": 6.032803964614868,
      "epoch": 2.4102071135191823,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004417299587425001,
      "loss": 5.2296,
      "mean_token_accuracy": 0.20383723229169845,
      "num_tokens": 55890357.0,
      "step": 27580
    },
    {
      "entropy": 6.031777858734131,
      "epoch": 2.4106440618718867,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004417090330675114,
      "loss": 5.3057,
      "mean_token_accuracy": 0.20002656579017639,
      "num_tokens": 55900080.0,
      "step": 27585
    },
    {
      "entropy": 5.9240710735321045,
      "epoch": 2.4110810102245916,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044168810419495597,
      "loss": 5.2804,
      "mean_token_accuracy": 0.2010875463485718,
      "num_tokens": 55911326.0,
      "step": 27590
    },
    {
      "entropy": 6.010161924362182,
      "epoch": 2.411517958577296,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004416671721252355,
      "loss": 5.236,
      "mean_token_accuracy": 0.20693300515413285,
      "num_tokens": 55921872.0,
      "step": 27595
    },
    {
      "entropy": 5.944522142410278,
      "epoch": 2.411954906930001,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044164623685875133,
      "loss": 5.2381,
      "mean_token_accuracy": 0.21376649141311646,
      "num_tokens": 55932539.0,
      "step": 27600
    },
    {
      "entropy": 5.9885166645050045,
      "epoch": 2.4123918552827055,
      "grad_norm": 1.3125,
      "learning_rate": 0.000441625298395905,
      "loss": 5.318,
      "mean_token_accuracy": 0.20684414207935334,
      "num_tokens": 55942087.0,
      "step": 27605
    },
    {
      "entropy": 6.056195974349976,
      "epoch": 2.4128288036354104,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004416043567370981,
      "loss": 5.2664,
      "mean_token_accuracy": 0.19941044002771377,
      "num_tokens": 55951604.0,
      "step": 27610
    },
    {
      "entropy": 5.976415586471558,
      "epoch": 2.413265751988115,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004415834118827323,
      "loss": 5.2599,
      "mean_token_accuracy": 0.20743022710084916,
      "num_tokens": 55961687.0,
      "step": 27615
    },
    {
      "entropy": 5.9983306407928465,
      "epoch": 2.41370270034082,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004415624638332095,
      "loss": 5.312,
      "mean_token_accuracy": 0.20109530836343764,
      "num_tokens": 55972088.0,
      "step": 27620
    },
    {
      "entropy": 5.987062549591064,
      "epoch": 2.4141396486935243,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044154151258893125,
      "loss": 5.3187,
      "mean_token_accuracy": 0.19408823400735856,
      "num_tokens": 55981857.0,
      "step": 27625
    },
    {
      "entropy": 5.964003562927246,
      "epoch": 2.414576597046229,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004415205581502996,
      "loss": 5.2115,
      "mean_token_accuracy": 0.21475208997726442,
      "num_tokens": 55992196.0,
      "step": 27630
    },
    {
      "entropy": 5.99218339920044,
      "epoch": 2.415013545398934,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044149960051771643,
      "loss": 5.2408,
      "mean_token_accuracy": 0.2077663391828537,
      "num_tokens": 56002295.0,
      "step": 27635
    },
    {
      "entropy": 6.011464071273804,
      "epoch": 2.4154504937516386,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00044147863969158353,
      "loss": 5.2534,
      "mean_token_accuracy": 0.2010858252644539,
      "num_tokens": 56011847.0,
      "step": 27640
    },
    {
      "entropy": 5.944300556182862,
      "epoch": 2.415887442104343,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004414576756723032,
      "loss": 5.2173,
      "mean_token_accuracy": 0.21109281480312347,
      "num_tokens": 56021717.0,
      "step": 27645
    },
    {
      "entropy": 6.023686027526855,
      "epoch": 2.416324390457048,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004414367084602774,
      "loss": 5.2932,
      "mean_token_accuracy": 0.20196393877267838,
      "num_tokens": 56031845.0,
      "step": 27650
    },
    {
      "entropy": 5.988820886611938,
      "epoch": 2.416761338809753,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004414157380559083,
      "loss": 5.2866,
      "mean_token_accuracy": 0.20478367954492568,
      "num_tokens": 56042082.0,
      "step": 27655
    },
    {
      "entropy": 5.985181856155395,
      "epoch": 2.4171982871624573,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044139476445959817,
      "loss": 5.2795,
      "mean_token_accuracy": 0.20066423416137696,
      "num_tokens": 56052592.0,
      "step": 27660
    },
    {
      "entropy": 6.030622053146362,
      "epoch": 2.4176352355151622,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004413737876717492,
      "loss": 5.3605,
      "mean_token_accuracy": 0.19539632201194762,
      "num_tokens": 56062500.0,
      "step": 27665
    },
    {
      "entropy": 5.966926240921021,
      "epoch": 2.4180721838678667,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004413528076927639,
      "loss": 5.2028,
      "mean_token_accuracy": 0.20932901799678802,
      "num_tokens": 56072912.0,
      "step": 27670
    },
    {
      "entropy": 5.997545337677002,
      "epoch": 2.4185091322205716,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004413318245230446,
      "loss": 5.3333,
      "mean_token_accuracy": 0.19163413047790528,
      "num_tokens": 56082591.0,
      "step": 27675
    },
    {
      "entropy": 5.94212498664856,
      "epoch": 2.418946080573276,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004413108381629936,
      "loss": 5.1942,
      "mean_token_accuracy": 0.2119583562016487,
      "num_tokens": 56091738.0,
      "step": 27680
    },
    {
      "entropy": 5.979053688049317,
      "epoch": 2.419383028925981,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004412898486130137,
      "loss": 5.3061,
      "mean_token_accuracy": 0.19777825325727463,
      "num_tokens": 56102011.0,
      "step": 27685
    },
    {
      "entropy": 5.984834384918213,
      "epoch": 2.4198199772786855,
      "grad_norm": 1.125,
      "learning_rate": 0.00044126885587350735,
      "loss": 5.2716,
      "mean_token_accuracy": 0.20792069137096406,
      "num_tokens": 56112901.0,
      "step": 27690
    },
    {
      "entropy": 6.004150629043579,
      "epoch": 2.4202569256313904,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004412478599448772,
      "loss": 5.2791,
      "mean_token_accuracy": 0.2047337993979454,
      "num_tokens": 56123598.0,
      "step": 27695
    },
    {
      "entropy": 5.971405744552612,
      "epoch": 2.420693873984095,
      "grad_norm": 1.28125,
      "learning_rate": 0.000441226860827526,
      "loss": 5.3025,
      "mean_token_accuracy": 0.19638802856206894,
      "num_tokens": 56133686.0,
      "step": 27700
    },
    {
      "entropy": 6.013994121551514,
      "epoch": 2.4211308223367998,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044120585852185654,
      "loss": 5.2999,
      "mean_token_accuracy": 0.19695822447538375,
      "num_tokens": 56143722.0,
      "step": 27705
    },
    {
      "entropy": 6.017358303070068,
      "epoch": 2.4215677706895047,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044118485302827156,
      "loss": 5.2409,
      "mean_token_accuracy": 0.20762865692377092,
      "num_tokens": 56153781.0,
      "step": 27710
    },
    {
      "entropy": 5.900804281234741,
      "epoch": 2.422004719042209,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044116384434717405,
      "loss": 5.2412,
      "mean_token_accuracy": 0.21145824044942857,
      "num_tokens": 56163398.0,
      "step": 27715
    },
    {
      "entropy": 6.01810040473938,
      "epoch": 2.422441667394914,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004411428324789669,
      "loss": 5.3118,
      "mean_token_accuracy": 0.2066517174243927,
      "num_tokens": 56173418.0,
      "step": 27720
    },
    {
      "entropy": 6.0563544750213625,
      "epoch": 2.4228786157476185,
      "grad_norm": 1.125,
      "learning_rate": 0.00044112181742405324,
      "loss": 5.3652,
      "mean_token_accuracy": 0.19562885761260987,
      "num_tokens": 56184547.0,
      "step": 27725
    },
    {
      "entropy": 6.033981990814209,
      "epoch": 2.4233155641003234,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000441100799182836,
      "loss": 5.2589,
      "mean_token_accuracy": 0.20121364444494247,
      "num_tokens": 56195404.0,
      "step": 27730
    },
    {
      "entropy": 5.915733575820923,
      "epoch": 2.423752512453028,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004410797777557184,
      "loss": 5.2323,
      "mean_token_accuracy": 0.2122793823480606,
      "num_tokens": 56205552.0,
      "step": 27735
    },
    {
      "entropy": 5.997263193130493,
      "epoch": 2.424189460805733,
      "grad_norm": 1.125,
      "learning_rate": 0.00044105875314310374,
      "loss": 5.2365,
      "mean_token_accuracy": 0.20641426295042037,
      "num_tokens": 56215581.0,
      "step": 27740
    },
    {
      "entropy": 5.966891050338745,
      "epoch": 2.4246264091584373,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00044103772534539505,
      "loss": 5.1932,
      "mean_token_accuracy": 0.20871752202510835,
      "num_tokens": 56225428.0,
      "step": 27745
    },
    {
      "entropy": 5.979529476165771,
      "epoch": 2.425063357511142,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004410166943629959,
      "loss": 5.2863,
      "mean_token_accuracy": 0.20490991920232773,
      "num_tokens": 56235284.0,
      "step": 27750
    },
    {
      "entropy": 5.9713873863220215,
      "epoch": 2.4255003058638467,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004409956601963095,
      "loss": 5.2472,
      "mean_token_accuracy": 0.207872274518013,
      "num_tokens": 56244444.0,
      "step": 27755
    },
    {
      "entropy": 5.95171709060669,
      "epoch": 2.4259372542165516,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004409746228457393,
      "loss": 5.2645,
      "mean_token_accuracy": 0.20921646803617477,
      "num_tokens": 56254983.0,
      "step": 27760
    },
    {
      "entropy": 6.017758417129516,
      "epoch": 2.4263742025692565,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004409535823116889,
      "loss": 5.2789,
      "mean_token_accuracy": 0.20720958858728408,
      "num_tokens": 56264307.0,
      "step": 27765
    },
    {
      "entropy": 6.0052131652832035,
      "epoch": 2.426811150921961,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004409325385945619,
      "loss": 5.296,
      "mean_token_accuracy": 0.20183659046888353,
      "num_tokens": 56274418.0,
      "step": 27770
    },
    {
      "entropy": 6.076952171325684,
      "epoch": 2.427248099274666,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004409114916947618,
      "loss": 5.3837,
      "mean_token_accuracy": 0.19364988058805466,
      "num_tokens": 56285665.0,
      "step": 27775
    },
    {
      "entropy": 6.003330278396606,
      "epoch": 2.4276850476273704,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044089044161269237,
      "loss": 5.2704,
      "mean_token_accuracy": 0.20579207241535186,
      "num_tokens": 56296320.0,
      "step": 27780
    },
    {
      "entropy": 6.029161739349365,
      "epoch": 2.4281219959800753,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004408693883487574,
      "loss": 5.3794,
      "mean_token_accuracy": 0.19435993134975432,
      "num_tokens": 56308009.0,
      "step": 27785
    },
    {
      "entropy": 5.930944776535034,
      "epoch": 2.4285589443327797,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004408483319033606,
      "loss": 5.2799,
      "mean_token_accuracy": 0.20443665832281113,
      "num_tokens": 56318341.0,
      "step": 27790
    },
    {
      "entropy": 5.910695934295655,
      "epoch": 2.4289958926854847,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004408272722769059,
      "loss": 5.2032,
      "mean_token_accuracy": 0.21820038706064224,
      "num_tokens": 56328808.0,
      "step": 27795
    },
    {
      "entropy": 6.071496295928955,
      "epoch": 2.429432841038189,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004408062094697972,
      "loss": 5.288,
      "mean_token_accuracy": 0.20313290506601334,
      "num_tokens": 56339043.0,
      "step": 27800
    },
    {
      "entropy": 6.004401016235351,
      "epoch": 2.429869789390894,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044078514348243846,
      "loss": 5.2978,
      "mean_token_accuracy": 0.2003936469554901,
      "num_tokens": 56349518.0,
      "step": 27805
    },
    {
      "entropy": 5.94050030708313,
      "epoch": 2.4303067377435985,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044076407431523394,
      "loss": 5.2489,
      "mean_token_accuracy": 0.1974821373820305,
      "num_tokens": 56358739.0,
      "step": 27810
    },
    {
      "entropy": 5.989663934707641,
      "epoch": 2.4307436860963034,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004407430019685875,
      "loss": 5.2946,
      "mean_token_accuracy": 0.19897305965423584,
      "num_tokens": 56368049.0,
      "step": 27815
    },
    {
      "entropy": 6.03308424949646,
      "epoch": 2.4311806344490083,
      "grad_norm": 1.328125,
      "learning_rate": 0.00044072192644290355,
      "loss": 5.2572,
      "mean_token_accuracy": 0.20176074653863907,
      "num_tokens": 56377607.0,
      "step": 27820
    },
    {
      "entropy": 5.972359371185303,
      "epoch": 2.431617582801713,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004407008477385862,
      "loss": 5.2479,
      "mean_token_accuracy": 0.21049316972494125,
      "num_tokens": 56387188.0,
      "step": 27825
    },
    {
      "entropy": 5.996887254714966,
      "epoch": 2.4320545311544177,
      "grad_norm": 1.3125,
      "learning_rate": 0.00044067976585603964,
      "loss": 5.3721,
      "mean_token_accuracy": 0.19516438841819764,
      "num_tokens": 56396389.0,
      "step": 27830
    },
    {
      "entropy": 6.03840708732605,
      "epoch": 2.432491479507122,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004406586807956684,
      "loss": 5.3137,
      "mean_token_accuracy": 0.2036794826388359,
      "num_tokens": 56406909.0,
      "step": 27835
    },
    {
      "entropy": 6.025599002838135,
      "epoch": 2.432928427859827,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044063759255787686,
      "loss": 5.3138,
      "mean_token_accuracy": 0.1969681203365326,
      "num_tokens": 56416373.0,
      "step": 27840
    },
    {
      "entropy": 6.042899227142334,
      "epoch": 2.4333653762125316,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044061650114306954,
      "loss": 5.3482,
      "mean_token_accuracy": 0.19820954501628876,
      "num_tokens": 56426759.0,
      "step": 27845
    },
    {
      "entropy": 6.014265060424805,
      "epoch": 2.4338023245652365,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044059540655165097,
      "loss": 5.3068,
      "mean_token_accuracy": 0.20365864783525467,
      "num_tokens": 56436783.0,
      "step": 27850
    },
    {
      "entropy": 6.059355735778809,
      "epoch": 2.434239272917941,
      "grad_norm": 1.28125,
      "learning_rate": 0.00044057430878402564,
      "loss": 5.2983,
      "mean_token_accuracy": 0.19626158773899077,
      "num_tokens": 56447191.0,
      "step": 27855
    },
    {
      "entropy": 5.973565673828125,
      "epoch": 2.434676221270646,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004405532078405984,
      "loss": 5.3279,
      "mean_token_accuracy": 0.20042461901903152,
      "num_tokens": 56457119.0,
      "step": 27860
    },
    {
      "entropy": 5.993735647201538,
      "epoch": 2.4351131696233503,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004405321037217739,
      "loss": 5.2325,
      "mean_token_accuracy": 0.20937585532665254,
      "num_tokens": 56467139.0,
      "step": 27865
    },
    {
      "entropy": 6.021820688247681,
      "epoch": 2.4355501179760553,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004405109964279568,
      "loss": 5.2637,
      "mean_token_accuracy": 0.20351221263408661,
      "num_tokens": 56477188.0,
      "step": 27870
    },
    {
      "entropy": 6.027664709091186,
      "epoch": 2.43598706632876,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044048988595955216,
      "loss": 5.4221,
      "mean_token_accuracy": 0.19040421545505523,
      "num_tokens": 56487594.0,
      "step": 27875
    },
    {
      "entropy": 6.001716661453247,
      "epoch": 2.4364240146814646,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004404687723169647,
      "loss": 5.292,
      "mean_token_accuracy": 0.2093895584344864,
      "num_tokens": 56498008.0,
      "step": 27880
    },
    {
      "entropy": 5.974712657928467,
      "epoch": 2.4368609630341695,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044044765550059955,
      "loss": 5.1993,
      "mean_token_accuracy": 0.2083689883351326,
      "num_tokens": 56508409.0,
      "step": 27885
    },
    {
      "entropy": 5.982428979873657,
      "epoch": 2.437297911386874,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004404265355108616,
      "loss": 5.2812,
      "mean_token_accuracy": 0.20034931004047393,
      "num_tokens": 56519513.0,
      "step": 27890
    },
    {
      "entropy": 5.989485883712769,
      "epoch": 2.437734859739579,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004404054123481561,
      "loss": 5.255,
      "mean_token_accuracy": 0.2027032345533371,
      "num_tokens": 56528549.0,
      "step": 27895
    },
    {
      "entropy": 5.9587170600891115,
      "epoch": 2.4381718080922834,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044038428601288817,
      "loss": 5.2709,
      "mean_token_accuracy": 0.20314326882362366,
      "num_tokens": 56537859.0,
      "step": 27900
    },
    {
      "entropy": 5.95966649055481,
      "epoch": 2.4386087564449883,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044036315650546287,
      "loss": 5.328,
      "mean_token_accuracy": 0.1975555419921875,
      "num_tokens": 56547326.0,
      "step": 27905
    },
    {
      "entropy": 6.096992492675781,
      "epoch": 2.439045704797693,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044034202382628565,
      "loss": 5.3585,
      "mean_token_accuracy": 0.1972046375274658,
      "num_tokens": 56557460.0,
      "step": 27910
    },
    {
      "entropy": 5.986851501464844,
      "epoch": 2.4394826531503977,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004403208879757617,
      "loss": 5.2512,
      "mean_token_accuracy": 0.20841725319623947,
      "num_tokens": 56566928.0,
      "step": 27915
    },
    {
      "entropy": 5.8607649326324465,
      "epoch": 2.439919601503102,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004402997489542966,
      "loss": 5.2253,
      "mean_token_accuracy": 0.21006364673376082,
      "num_tokens": 56577576.0,
      "step": 27920
    },
    {
      "entropy": 6.058060550689698,
      "epoch": 2.440356549855807,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004402786067622955,
      "loss": 5.373,
      "mean_token_accuracy": 0.19836577624082566,
      "num_tokens": 56587744.0,
      "step": 27925
    },
    {
      "entropy": 6.030117511749268,
      "epoch": 2.440793498208512,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004402574614001643,
      "loss": 5.2832,
      "mean_token_accuracy": 0.20681770741939545,
      "num_tokens": 56598272.0,
      "step": 27930
    },
    {
      "entropy": 5.923111486434936,
      "epoch": 2.4412304465612165,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044023631286830835,
      "loss": 5.2414,
      "mean_token_accuracy": 0.20202074944972992,
      "num_tokens": 56609346.0,
      "step": 27935
    },
    {
      "entropy": 6.015246915817261,
      "epoch": 2.441667394913921,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00044021516116713334,
      "loss": 5.254,
      "mean_token_accuracy": 0.20097281485795976,
      "num_tokens": 56619210.0,
      "step": 27940
    },
    {
      "entropy": 6.0168966293334964,
      "epoch": 2.442104343266626,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000440194006297045,
      "loss": 5.2967,
      "mean_token_accuracy": 0.19960817247629165,
      "num_tokens": 56629135.0,
      "step": 27945
    },
    {
      "entropy": 6.020817756652832,
      "epoch": 2.4425412916193308,
      "grad_norm": 1.21875,
      "learning_rate": 0.000440172848258449,
      "loss": 5.2475,
      "mean_token_accuracy": 0.20342273116111756,
      "num_tokens": 56638900.0,
      "step": 27950
    },
    {
      "entropy": 6.0084583282470705,
      "epoch": 2.4429782399720352,
      "grad_norm": 1.328125,
      "learning_rate": 0.00044015168705175117,
      "loss": 5.3269,
      "mean_token_accuracy": 0.1968041479587555,
      "num_tokens": 56648605.0,
      "step": 27955
    },
    {
      "entropy": 5.950061988830567,
      "epoch": 2.44341518832474,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004401305226773576,
      "loss": 5.358,
      "mean_token_accuracy": 0.20388581454753876,
      "num_tokens": 56659556.0,
      "step": 27960
    },
    {
      "entropy": 6.090784883499145,
      "epoch": 2.4438521366774446,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004401093551356739,
      "loss": 5.3912,
      "mean_token_accuracy": 0.19347190856933594,
      "num_tokens": 56669727.0,
      "step": 27965
    },
    {
      "entropy": 6.064439058303833,
      "epoch": 2.4442890850301495,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004400881844271064,
      "loss": 5.3522,
      "mean_token_accuracy": 0.19656844288110734,
      "num_tokens": 56681308.0,
      "step": 27970
    },
    {
      "entropy": 6.01848578453064,
      "epoch": 2.444726033382854,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044006701055206095,
      "loss": 5.3024,
      "mean_token_accuracy": 0.20131366550922394,
      "num_tokens": 56691089.0,
      "step": 27975
    },
    {
      "entropy": 5.955787992477417,
      "epoch": 2.445162981735559,
      "grad_norm": 1.125,
      "learning_rate": 0.0004400458335109437,
      "loss": 5.2717,
      "mean_token_accuracy": 0.21171137690544128,
      "num_tokens": 56701599.0,
      "step": 27980
    },
    {
      "entropy": 5.949743223190308,
      "epoch": 2.4455999300882634,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044002465330416095,
      "loss": 5.19,
      "mean_token_accuracy": 0.21305831968784333,
      "num_tokens": 56711890.0,
      "step": 27985
    },
    {
      "entropy": 6.002938175201416,
      "epoch": 2.4460368784409683,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004400034699321189,
      "loss": 5.2544,
      "mean_token_accuracy": 0.20681423395872117,
      "num_tokens": 56722003.0,
      "step": 27990
    },
    {
      "entropy": 6.001338911056519,
      "epoch": 2.4464738267936728,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004399822833952238,
      "loss": 5.2204,
      "mean_token_accuracy": 0.20864820778369902,
      "num_tokens": 56731636.0,
      "step": 27995
    },
    {
      "entropy": 5.925555562973022,
      "epoch": 2.4469107751463777,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000439961093693882,
      "loss": 5.2051,
      "mean_token_accuracy": 0.20758059322834016,
      "num_tokens": 56741973.0,
      "step": 28000
    },
    {
      "entropy": 5.996609449386597,
      "epoch": 2.4473477234990826,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004399399008285001,
      "loss": 5.3657,
      "mean_token_accuracy": 0.1979817718267441,
      "num_tokens": 56751441.0,
      "step": 28005
    },
    {
      "entropy": 6.008587121963501,
      "epoch": 2.447784671851787,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004399187047994844,
      "loss": 5.2924,
      "mean_token_accuracy": 0.20876994282007216,
      "num_tokens": 56761156.0,
      "step": 28010
    },
    {
      "entropy": 6.0354249477386475,
      "epoch": 2.448221620204492,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043989750560724153,
      "loss": 5.2885,
      "mean_token_accuracy": 0.202756105363369,
      "num_tokens": 56771414.0,
      "step": 28015
    },
    {
      "entropy": 5.900220584869385,
      "epoch": 2.4486585685571964,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004398763032521781,
      "loss": 5.1936,
      "mean_token_accuracy": 0.21323088109493255,
      "num_tokens": 56781572.0,
      "step": 28020
    },
    {
      "entropy": 6.028608894348144,
      "epoch": 2.4490955169099013,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043985509773470077,
      "loss": 5.3378,
      "mean_token_accuracy": 0.19350051432847976,
      "num_tokens": 56791299.0,
      "step": 28025
    },
    {
      "entropy": 5.984430932998658,
      "epoch": 2.449532465262606,
      "grad_norm": 1.25,
      "learning_rate": 0.0004398338890552163,
      "loss": 5.2425,
      "mean_token_accuracy": 0.20214253962039946,
      "num_tokens": 56802485.0,
      "step": 28030
    },
    {
      "entropy": 5.98906569480896,
      "epoch": 2.4499694136153107,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004398126772141315,
      "loss": 5.3095,
      "mean_token_accuracy": 0.19430006444454193,
      "num_tokens": 56811694.0,
      "step": 28035
    },
    {
      "entropy": 5.956458568572998,
      "epoch": 2.450406361968015,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043979146221185313,
      "loss": 5.2203,
      "mean_token_accuracy": 0.2080234557390213,
      "num_tokens": 56821814.0,
      "step": 28040
    },
    {
      "entropy": 6.000380182266236,
      "epoch": 2.45084331032072,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004397702440487882,
      "loss": 5.2906,
      "mean_token_accuracy": 0.20272646248340606,
      "num_tokens": 56832257.0,
      "step": 28045
    },
    {
      "entropy": 6.039734792709351,
      "epoch": 2.4512802586734246,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004397490227253436,
      "loss": 5.2952,
      "mean_token_accuracy": 0.19864067882299424,
      "num_tokens": 56841789.0,
      "step": 28050
    },
    {
      "entropy": 5.9909224033355715,
      "epoch": 2.4517172070261295,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004397277982419265,
      "loss": 5.2699,
      "mean_token_accuracy": 0.19885912239551545,
      "num_tokens": 56851612.0,
      "step": 28055
    },
    {
      "entropy": 5.964059829711914,
      "epoch": 2.4521541553788344,
      "grad_norm": 1.125,
      "learning_rate": 0.00043970657059894385,
      "loss": 5.2346,
      "mean_token_accuracy": 0.20601497143507003,
      "num_tokens": 56861812.0,
      "step": 28060
    },
    {
      "entropy": 6.010293245315552,
      "epoch": 2.452591103731539,
      "grad_norm": 1.234375,
      "learning_rate": 0.000439685339796803,
      "loss": 5.2228,
      "mean_token_accuracy": 0.211328986287117,
      "num_tokens": 56871873.0,
      "step": 28065
    },
    {
      "entropy": 6.0562584400177,
      "epoch": 2.453028052084244,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004396641058359109,
      "loss": 5.3182,
      "mean_token_accuracy": 0.19674671888351442,
      "num_tokens": 56883579.0,
      "step": 28070
    },
    {
      "entropy": 5.93415470123291,
      "epoch": 2.4534650004369483,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043964286871667495,
      "loss": 5.2504,
      "mean_token_accuracy": 0.2053658574819565,
      "num_tokens": 56894305.0,
      "step": 28075
    },
    {
      "entropy": 6.010423946380615,
      "epoch": 2.453901948789653,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004396216284395026,
      "loss": 5.2489,
      "mean_token_accuracy": 0.201833114027977,
      "num_tokens": 56904393.0,
      "step": 28080
    },
    {
      "entropy": 5.9752356052398685,
      "epoch": 2.4543388971423576,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004396003850048011,
      "loss": 5.2799,
      "mean_token_accuracy": 0.20644329041242598,
      "num_tokens": 56914600.0,
      "step": 28085
    },
    {
      "entropy": 6.0111054420471195,
      "epoch": 2.4547758454950626,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004395791384129779,
      "loss": 5.3588,
      "mean_token_accuracy": 0.2020885854959488,
      "num_tokens": 56925948.0,
      "step": 28090
    },
    {
      "entropy": 5.949389314651489,
      "epoch": 2.455212793847767,
      "grad_norm": 1.25,
      "learning_rate": 0.00043955788866444073,
      "loss": 5.2063,
      "mean_token_accuracy": 0.20574226677417756,
      "num_tokens": 56935395.0,
      "step": 28095
    },
    {
      "entropy": 5.964217901229858,
      "epoch": 2.455649742200472,
      "grad_norm": 1.328125,
      "learning_rate": 0.00043953663575959696,
      "loss": 5.2063,
      "mean_token_accuracy": 0.21299299001693725,
      "num_tokens": 56944769.0,
      "step": 28100
    },
    {
      "entropy": 5.961055612564087,
      "epoch": 2.4560866905531764,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004395153796988543,
      "loss": 5.224,
      "mean_token_accuracy": 0.20844018310308457,
      "num_tokens": 56954352.0,
      "step": 28105
    },
    {
      "entropy": 5.957422876358033,
      "epoch": 2.4565236389058813,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004394941204826205,
      "loss": 5.2269,
      "mean_token_accuracy": 0.20683867037296294,
      "num_tokens": 56963759.0,
      "step": 28110
    },
    {
      "entropy": 5.952259588241577,
      "epoch": 2.4569605872585862,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043947285811130315,
      "loss": 5.2296,
      "mean_token_accuracy": 0.2054138168692589,
      "num_tokens": 56974962.0,
      "step": 28115
    },
    {
      "entropy": 6.008630275726318,
      "epoch": 2.4573975356112907,
      "grad_norm": 1.125,
      "learning_rate": 0.00043945159258531035,
      "loss": 5.2917,
      "mean_token_accuracy": 0.19653381705284118,
      "num_tokens": 56985637.0,
      "step": 28120
    },
    {
      "entropy": 6.010749483108521,
      "epoch": 2.4578344839639956,
      "grad_norm": 1.25,
      "learning_rate": 0.0004394303239050498,
      "loss": 5.2867,
      "mean_token_accuracy": 0.20718136429786682,
      "num_tokens": 56995779.0,
      "step": 28125
    },
    {
      "entropy": 5.979878616333008,
      "epoch": 2.4582714323167,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004394090520709294,
      "loss": 5.2396,
      "mean_token_accuracy": 0.20673224925994874,
      "num_tokens": 57005506.0,
      "step": 28130
    },
    {
      "entropy": 5.97655029296875,
      "epoch": 2.458708380669405,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043938777708335726,
      "loss": 5.3564,
      "mean_token_accuracy": 0.19155046194791794,
      "num_tokens": 57016001.0,
      "step": 28135
    },
    {
      "entropy": 6.015478801727295,
      "epoch": 2.4591453290221095,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043936649894274146,
      "loss": 5.3036,
      "mean_token_accuracy": 0.2018024280667305,
      "num_tokens": 57026418.0,
      "step": 28140
    },
    {
      "entropy": 6.033252239227295,
      "epoch": 2.4595822773748144,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043934521764949,
      "loss": 5.3114,
      "mean_token_accuracy": 0.20158193111419678,
      "num_tokens": 57037170.0,
      "step": 28145
    },
    {
      "entropy": 5.974526071548462,
      "epoch": 2.460019225727519,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004393239332040113,
      "loss": 5.2261,
      "mean_token_accuracy": 0.2081304058432579,
      "num_tokens": 57046775.0,
      "step": 28150
    },
    {
      "entropy": 5.942977237701416,
      "epoch": 2.4604561740802238,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043930264560671336,
      "loss": 5.2066,
      "mean_token_accuracy": 0.21097496896982193,
      "num_tokens": 57056360.0,
      "step": 28155
    },
    {
      "entropy": 5.969457817077637,
      "epoch": 2.4608931224329282,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004392813548580045,
      "loss": 5.2366,
      "mean_token_accuracy": 0.2037920042872429,
      "num_tokens": 57066355.0,
      "step": 28160
    },
    {
      "entropy": 5.9496739387512205,
      "epoch": 2.461330070785633,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004392600609582933,
      "loss": 5.2787,
      "mean_token_accuracy": 0.20797260850667953,
      "num_tokens": 57077158.0,
      "step": 28165
    },
    {
      "entropy": 6.021720027923584,
      "epoch": 2.461767019138338,
      "grad_norm": 1.140625,
      "learning_rate": 0.000439238763907988,
      "loss": 5.3384,
      "mean_token_accuracy": 0.202809938788414,
      "num_tokens": 57088005.0,
      "step": 28170
    },
    {
      "entropy": 5.969601058959961,
      "epoch": 2.4622039674910425,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004392174637074972,
      "loss": 5.2592,
      "mean_token_accuracy": 0.20450838953256606,
      "num_tokens": 57098399.0,
      "step": 28175
    },
    {
      "entropy": 5.981549835205078,
      "epoch": 2.4626409158437474,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004391961603572294,
      "loss": 5.2452,
      "mean_token_accuracy": 0.2047594666481018,
      "num_tokens": 57108949.0,
      "step": 28180
    },
    {
      "entropy": 6.00144395828247,
      "epoch": 2.463077864196452,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043917485385759316,
      "loss": 5.2974,
      "mean_token_accuracy": 0.203876231610775,
      "num_tokens": 57119441.0,
      "step": 28185
    },
    {
      "entropy": 5.989717864990235,
      "epoch": 2.463514812549157,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004391535442089971,
      "loss": 5.2946,
      "mean_token_accuracy": 0.20990421622991562,
      "num_tokens": 57129507.0,
      "step": 28190
    },
    {
      "entropy": 5.974894189834595,
      "epoch": 2.4639517609018613,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004391322314118502,
      "loss": 5.2635,
      "mean_token_accuracy": 0.20262532979249953,
      "num_tokens": 57140490.0,
      "step": 28195
    },
    {
      "entropy": 5.950553369522095,
      "epoch": 2.464388709254566,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043911091546656105,
      "loss": 5.2085,
      "mean_token_accuracy": 0.2119060382246971,
      "num_tokens": 57150706.0,
      "step": 28200
    },
    {
      "entropy": 5.964052009582519,
      "epoch": 2.4648256576072707,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043908959637353864,
      "loss": 5.2704,
      "mean_token_accuracy": 0.2050002858042717,
      "num_tokens": 57160259.0,
      "step": 28205
    },
    {
      "entropy": 5.944239711761474,
      "epoch": 2.4652626059599756,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043906827413319166,
      "loss": 5.2609,
      "mean_token_accuracy": 0.2054815486073494,
      "num_tokens": 57171831.0,
      "step": 28210
    },
    {
      "entropy": 6.031755113601685,
      "epoch": 2.46569955431268,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004390469487459292,
      "loss": 5.3625,
      "mean_token_accuracy": 0.198906871676445,
      "num_tokens": 57182172.0,
      "step": 28215
    },
    {
      "entropy": 5.954598712921142,
      "epoch": 2.466136502665385,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043902562021216033,
      "loss": 5.2306,
      "mean_token_accuracy": 0.21192415654659272,
      "num_tokens": 57191633.0,
      "step": 28220
    },
    {
      "entropy": 6.02432804107666,
      "epoch": 2.46657345101809,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004390042885322942,
      "loss": 5.2709,
      "mean_token_accuracy": 0.20245350003242493,
      "num_tokens": 57200953.0,
      "step": 28225
    },
    {
      "entropy": 6.036960315704346,
      "epoch": 2.4670103993707944,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043898295370673986,
      "loss": 5.3433,
      "mean_token_accuracy": 0.19941832423210143,
      "num_tokens": 57210893.0,
      "step": 28230
    },
    {
      "entropy": 5.9285973072052,
      "epoch": 2.4674473477234993,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043896161573590646,
      "loss": 5.2715,
      "mean_token_accuracy": 0.2088572010397911,
      "num_tokens": 57221458.0,
      "step": 28235
    },
    {
      "entropy": 5.970260238647461,
      "epoch": 2.4678842960762037,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004389402746202034,
      "loss": 5.3085,
      "mean_token_accuracy": 0.20003650486469268,
      "num_tokens": 57231450.0,
      "step": 28240
    },
    {
      "entropy": 6.042877817153931,
      "epoch": 2.4683212444289087,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043891893036004,
      "loss": 5.2956,
      "mean_token_accuracy": 0.204000024497509,
      "num_tokens": 57242820.0,
      "step": 28245
    },
    {
      "entropy": 6.083094263076783,
      "epoch": 2.468758192781613,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004388975829558256,
      "loss": 5.2998,
      "mean_token_accuracy": 0.19982342123985292,
      "num_tokens": 57253465.0,
      "step": 28250
    },
    {
      "entropy": 5.961714172363282,
      "epoch": 2.469195141134318,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043887623240796965,
      "loss": 5.2773,
      "mean_token_accuracy": 0.20379328429698945,
      "num_tokens": 57263521.0,
      "step": 28255
    },
    {
      "entropy": 5.993887329101563,
      "epoch": 2.4696320894870225,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004388548787168817,
      "loss": 5.2757,
      "mean_token_accuracy": 0.1957926020026207,
      "num_tokens": 57273568.0,
      "step": 28260
    },
    {
      "entropy": 6.076740646362305,
      "epoch": 2.4700690378397274,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004388335218829713,
      "loss": 5.3279,
      "mean_token_accuracy": 0.20292046070098876,
      "num_tokens": 57283960.0,
      "step": 28265
    },
    {
      "entropy": 5.972231340408325,
      "epoch": 2.470505986192432,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004388121619066482,
      "loss": 5.2798,
      "mean_token_accuracy": 0.20294946134090425,
      "num_tokens": 57294672.0,
      "step": 28270
    },
    {
      "entropy": 5.9782209396362305,
      "epoch": 2.470942934545137,
      "grad_norm": 1.25,
      "learning_rate": 0.0004387907987883219,
      "loss": 5.271,
      "mean_token_accuracy": 0.2072295680642128,
      "num_tokens": 57305594.0,
      "step": 28275
    },
    {
      "entropy": 6.001715087890625,
      "epoch": 2.4713798828978417,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004387694325284023,
      "loss": 5.2659,
      "mean_token_accuracy": 0.20878747552633287,
      "num_tokens": 57315250.0,
      "step": 28280
    },
    {
      "entropy": 6.013324785232544,
      "epoch": 2.471816831250546,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004387480631272991,
      "loss": 5.3013,
      "mean_token_accuracy": 0.20146892666816713,
      "num_tokens": 57326750.0,
      "step": 28285
    },
    {
      "entropy": 6.01450400352478,
      "epoch": 2.4722537796032507,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004387266905854223,
      "loss": 5.271,
      "mean_token_accuracy": 0.20026090294122695,
      "num_tokens": 57336151.0,
      "step": 28290
    },
    {
      "entropy": 5.995886421203613,
      "epoch": 2.4726907279559556,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004387053149031819,
      "loss": 5.2314,
      "mean_token_accuracy": 0.21070293486118316,
      "num_tokens": 57346761.0,
      "step": 28295
    },
    {
      "entropy": 6.0100915908813475,
      "epoch": 2.4731276763086605,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004386839360809876,
      "loss": 5.3271,
      "mean_token_accuracy": 0.20015607625246049,
      "num_tokens": 57356654.0,
      "step": 28300
    },
    {
      "entropy": 5.958705711364746,
      "epoch": 2.473564624661365,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004386625541192497,
      "loss": 5.2981,
      "mean_token_accuracy": 0.20085732489824296,
      "num_tokens": 57367888.0,
      "step": 28305
    },
    {
      "entropy": 5.9771040916442875,
      "epoch": 2.47400157301407,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043864116901837826,
      "loss": 5.2393,
      "mean_token_accuracy": 0.2108196958899498,
      "num_tokens": 57377167.0,
      "step": 28310
    },
    {
      "entropy": 5.947661447525024,
      "epoch": 2.4744385213667743,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043861978077878343,
      "loss": 5.2299,
      "mean_token_accuracy": 0.2058110311627388,
      "num_tokens": 57389076.0,
      "step": 28315
    },
    {
      "entropy": 5.974102735519409,
      "epoch": 2.4748754697194792,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004385983894008755,
      "loss": 5.2867,
      "mean_token_accuracy": 0.20293378680944443,
      "num_tokens": 57401136.0,
      "step": 28320
    },
    {
      "entropy": 5.991512632369995,
      "epoch": 2.4753124180721837,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043857699488506467,
      "loss": 5.2744,
      "mean_token_accuracy": 0.2026324987411499,
      "num_tokens": 57410548.0,
      "step": 28325
    },
    {
      "entropy": 5.959733486175537,
      "epoch": 2.4757493664248886,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004385555972317615,
      "loss": 5.322,
      "mean_token_accuracy": 0.19898713529109954,
      "num_tokens": 57420371.0,
      "step": 28330
    },
    {
      "entropy": 6.040572738647461,
      "epoch": 2.476186314777593,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043853419644137616,
      "loss": 5.3293,
      "mean_token_accuracy": 0.2015597030520439,
      "num_tokens": 57430079.0,
      "step": 28335
    },
    {
      "entropy": 6.031483602523804,
      "epoch": 2.476623263130298,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043851279251431937,
      "loss": 5.2744,
      "mean_token_accuracy": 0.20371359437704087,
      "num_tokens": 57440763.0,
      "step": 28340
    },
    {
      "entropy": 6.083889722824097,
      "epoch": 2.4770602114830025,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043849138545100136,
      "loss": 5.3441,
      "mean_token_accuracy": 0.20047691464424133,
      "num_tokens": 57451008.0,
      "step": 28345
    },
    {
      "entropy": 6.019879484176636,
      "epoch": 2.4774971598357074,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043846997525183305,
      "loss": 5.3673,
      "mean_token_accuracy": 0.19446358531713487,
      "num_tokens": 57460852.0,
      "step": 28350
    },
    {
      "entropy": 6.042668867111206,
      "epoch": 2.4779341081884123,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043844856191722486,
      "loss": 5.3007,
      "mean_token_accuracy": 0.2001379907131195,
      "num_tokens": 57470027.0,
      "step": 28355
    },
    {
      "entropy": 6.022334051132202,
      "epoch": 2.478371056541117,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043842714544758764,
      "loss": 5.2071,
      "mean_token_accuracy": 0.20228191018104552,
      "num_tokens": 57479846.0,
      "step": 28360
    },
    {
      "entropy": 6.052418661117554,
      "epoch": 2.4788080048938217,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043840572584333216,
      "loss": 5.3418,
      "mean_token_accuracy": 0.19497265964746474,
      "num_tokens": 57490569.0,
      "step": 28365
    },
    {
      "entropy": 5.980133056640625,
      "epoch": 2.479244953246526,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004383843031048692,
      "loss": 5.3153,
      "mean_token_accuracy": 0.19791275262832642,
      "num_tokens": 57501048.0,
      "step": 28370
    },
    {
      "entropy": 6.005629396438598,
      "epoch": 2.479681901599231,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043836287723260977,
      "loss": 5.3371,
      "mean_token_accuracy": 0.19643333405256272,
      "num_tokens": 57511510.0,
      "step": 28375
    },
    {
      "entropy": 5.986559534072876,
      "epoch": 2.4801188499519355,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043834144822696463,
      "loss": 5.2978,
      "mean_token_accuracy": 0.20255134403705596,
      "num_tokens": 57522229.0,
      "step": 28380
    },
    {
      "entropy": 6.0125524520874025,
      "epoch": 2.4805557983046405,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000438320016088345,
      "loss": 5.2399,
      "mean_token_accuracy": 0.20115377604961396,
      "num_tokens": 57532618.0,
      "step": 28385
    },
    {
      "entropy": 5.977360391616822,
      "epoch": 2.480992746657345,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004382985808171619,
      "loss": 5.2799,
      "mean_token_accuracy": 0.2069904625415802,
      "num_tokens": 57542698.0,
      "step": 28390
    },
    {
      "entropy": 5.915167570114136,
      "epoch": 2.48142969501005,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043827714241382644,
      "loss": 5.1578,
      "mean_token_accuracy": 0.21284923404455186,
      "num_tokens": 57552556.0,
      "step": 28395
    },
    {
      "entropy": 6.06952052116394,
      "epoch": 2.4818666433627543,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004382557008787498,
      "loss": 5.3179,
      "mean_token_accuracy": 0.20244660079479218,
      "num_tokens": 57562830.0,
      "step": 28400
    },
    {
      "entropy": 6.0083881378173825,
      "epoch": 2.4823035917154592,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004382342562123433,
      "loss": 5.2702,
      "mean_token_accuracy": 0.20655697137117385,
      "num_tokens": 57572584.0,
      "step": 28405
    },
    {
      "entropy": 5.911779308319092,
      "epoch": 2.482740540068164,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004382128084150182,
      "loss": 5.2591,
      "mean_token_accuracy": 0.20249946862459184,
      "num_tokens": 57582571.0,
      "step": 28410
    },
    {
      "entropy": 5.966986989974975,
      "epoch": 2.4831774884208686,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043819135748718595,
      "loss": 5.3159,
      "mean_token_accuracy": 0.20329163521528243,
      "num_tokens": 57593267.0,
      "step": 28415
    },
    {
      "entropy": 6.031068181991577,
      "epoch": 2.4836144367735735,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043816990342925797,
      "loss": 5.342,
      "mean_token_accuracy": 0.19337086379528046,
      "num_tokens": 57603708.0,
      "step": 28420
    },
    {
      "entropy": 6.048546266555786,
      "epoch": 2.484051385126278,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043814844624164576,
      "loss": 5.3676,
      "mean_token_accuracy": 0.19103354960680008,
      "num_tokens": 57613523.0,
      "step": 28425
    },
    {
      "entropy": 5.991291332244873,
      "epoch": 2.484488333478983,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004381269859247607,
      "loss": 5.2571,
      "mean_token_accuracy": 0.20960024744272232,
      "num_tokens": 57624181.0,
      "step": 28430
    },
    {
      "entropy": 6.0074241161346436,
      "epoch": 2.4849252818316874,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043810552247901475,
      "loss": 5.2417,
      "mean_token_accuracy": 0.20580746680498124,
      "num_tokens": 57633674.0,
      "step": 28435
    },
    {
      "entropy": 5.979611730575561,
      "epoch": 2.4853622301843923,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043808405590481934,
      "loss": 5.2759,
      "mean_token_accuracy": 0.2062771737575531,
      "num_tokens": 57644723.0,
      "step": 28440
    },
    {
      "entropy": 6.008944082260132,
      "epoch": 2.4857991785370968,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004380625862025863,
      "loss": 5.2432,
      "mean_token_accuracy": 0.20701296776533126,
      "num_tokens": 57654274.0,
      "step": 28445
    },
    {
      "entropy": 5.962491989135742,
      "epoch": 2.4862361268898017,
      "grad_norm": 1.125,
      "learning_rate": 0.0004380411133727274,
      "loss": 5.2126,
      "mean_token_accuracy": 0.20947698801755904,
      "num_tokens": 57664572.0,
      "step": 28450
    },
    {
      "entropy": 6.009991312026978,
      "epoch": 2.486673075242506,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043801963741565454,
      "loss": 5.3507,
      "mean_token_accuracy": 0.19242067486047745,
      "num_tokens": 57674048.0,
      "step": 28455
    },
    {
      "entropy": 6.030033874511719,
      "epoch": 2.487110023595211,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004379981583317796,
      "loss": 5.2983,
      "mean_token_accuracy": 0.2059917628765106,
      "num_tokens": 57685488.0,
      "step": 28460
    },
    {
      "entropy": 6.047135257720948,
      "epoch": 2.487546971947916,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043797667612151454,
      "loss": 5.3011,
      "mean_token_accuracy": 0.2033618077635765,
      "num_tokens": 57696044.0,
      "step": 28465
    },
    {
      "entropy": 6.007206821441651,
      "epoch": 2.4879839203006204,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043795519078527144,
      "loss": 5.272,
      "mean_token_accuracy": 0.20861012041568755,
      "num_tokens": 57705702.0,
      "step": 28470
    },
    {
      "entropy": 5.987147474288941,
      "epoch": 2.4884208686533253,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043793370232346246,
      "loss": 5.3233,
      "mean_token_accuracy": 0.19464080184698104,
      "num_tokens": 57715628.0,
      "step": 28475
    },
    {
      "entropy": 6.031042623519897,
      "epoch": 2.48885781700603,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043791221073649964,
      "loss": 5.3063,
      "mean_token_accuracy": 0.20013424903154373,
      "num_tokens": 57725319.0,
      "step": 28480
    },
    {
      "entropy": 5.995706367492676,
      "epoch": 2.4892947653587347,
      "grad_norm": 1.125,
      "learning_rate": 0.00043789071602479525,
      "loss": 5.3198,
      "mean_token_accuracy": 0.20007432252168655,
      "num_tokens": 57736483.0,
      "step": 28485
    },
    {
      "entropy": 6.016610956192016,
      "epoch": 2.489731713711439,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004378692181887616,
      "loss": 5.3056,
      "mean_token_accuracy": 0.20053367167711258,
      "num_tokens": 57746667.0,
      "step": 28490
    },
    {
      "entropy": 5.9702873706817625,
      "epoch": 2.490168662064144,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00043784771722881095,
      "loss": 5.188,
      "mean_token_accuracy": 0.21718259304761886,
      "num_tokens": 57755971.0,
      "step": 28495
    },
    {
      "entropy": 5.937647962570191,
      "epoch": 2.4906056104168486,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004378262131453558,
      "loss": 5.2291,
      "mean_token_accuracy": 0.21062637716531754,
      "num_tokens": 57765858.0,
      "step": 28500
    },
    {
      "entropy": 5.945595598220825,
      "epoch": 2.4910425587695535,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004378047059388086,
      "loss": 5.2807,
      "mean_token_accuracy": 0.20179225504398346,
      "num_tokens": 57775154.0,
      "step": 28505
    },
    {
      "entropy": 5.977473783493042,
      "epoch": 2.491479507122258,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004377831956095817,
      "loss": 5.2874,
      "mean_token_accuracy": 0.2027939036488533,
      "num_tokens": 57785416.0,
      "step": 28510
    },
    {
      "entropy": 5.977580165863037,
      "epoch": 2.491916455474963,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000437761682158088,
      "loss": 5.2829,
      "mean_token_accuracy": 0.20135335475206376,
      "num_tokens": 57795483.0,
      "step": 28515
    },
    {
      "entropy": 6.028756380081177,
      "epoch": 2.492353403827668,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043774016558473976,
      "loss": 5.2756,
      "mean_token_accuracy": 0.2085308313369751,
      "num_tokens": 57805772.0,
      "step": 28520
    },
    {
      "entropy": 6.001275682449341,
      "epoch": 2.4927903521803723,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043771864588994995,
      "loss": 5.3083,
      "mean_token_accuracy": 0.20058747828006745,
      "num_tokens": 57815514.0,
      "step": 28525
    },
    {
      "entropy": 5.934257888793946,
      "epoch": 2.493227300533077,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004376971230741313,
      "loss": 5.2293,
      "mean_token_accuracy": 0.2094150334596634,
      "num_tokens": 57826312.0,
      "step": 28530
    },
    {
      "entropy": 6.023898887634277,
      "epoch": 2.4936642488857816,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004376755971376965,
      "loss": 5.3442,
      "mean_token_accuracy": 0.194673353433609,
      "num_tokens": 57836718.0,
      "step": 28535
    },
    {
      "entropy": 6.034321355819702,
      "epoch": 2.4941011972384866,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004376540680810585,
      "loss": 5.2601,
      "mean_token_accuracy": 0.20366467982530595,
      "num_tokens": 57846833.0,
      "step": 28540
    },
    {
      "entropy": 6.0583618640899655,
      "epoch": 2.494538145591191,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004376325359046304,
      "loss": 5.3398,
      "mean_token_accuracy": 0.19905255883932113,
      "num_tokens": 57856716.0,
      "step": 28545
    },
    {
      "entropy": 6.026195621490478,
      "epoch": 2.494975093943896,
      "grad_norm": 1.375,
      "learning_rate": 0.00043761100060882486,
      "loss": 5.3097,
      "mean_token_accuracy": 0.19838134646415712,
      "num_tokens": 57867079.0,
      "step": 28550
    },
    {
      "entropy": 6.013994264602661,
      "epoch": 2.4954120422966004,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004375894621940552,
      "loss": 5.2585,
      "mean_token_accuracy": 0.20254024416208266,
      "num_tokens": 57876593.0,
      "step": 28555
    },
    {
      "entropy": 6.006065464019775,
      "epoch": 2.4958489906493053,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004375679206607344,
      "loss": 5.2578,
      "mean_token_accuracy": 0.20572003424167634,
      "num_tokens": 57886834.0,
      "step": 28560
    },
    {
      "entropy": 5.960520601272583,
      "epoch": 2.49628593900201,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004375463760092758,
      "loss": 5.2403,
      "mean_token_accuracy": 0.21278124898672104,
      "num_tokens": 57896766.0,
      "step": 28565
    },
    {
      "entropy": 5.997460412979126,
      "epoch": 2.4967228873547147,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004375248282400925,
      "loss": 5.287,
      "mean_token_accuracy": 0.20069405436515808,
      "num_tokens": 57907259.0,
      "step": 28570
    },
    {
      "entropy": 6.048718643188477,
      "epoch": 2.4971598357074196,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004375032773535977,
      "loss": 5.4021,
      "mean_token_accuracy": 0.18871962726116182,
      "num_tokens": 57917940.0,
      "step": 28575
    },
    {
      "entropy": 6.039482450485229,
      "epoch": 2.497596784060124,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000437481723350205,
      "loss": 5.2887,
      "mean_token_accuracy": 0.20111142694950104,
      "num_tokens": 57927256.0,
      "step": 28580
    },
    {
      "entropy": 5.981166553497315,
      "epoch": 2.4980337324128286,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043746016623032776,
      "loss": 5.2618,
      "mean_token_accuracy": 0.2018311396241188,
      "num_tokens": 57936718.0,
      "step": 28585
    },
    {
      "entropy": 6.040131998062134,
      "epoch": 2.4984706807655335,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004374386059943794,
      "loss": 5.2688,
      "mean_token_accuracy": 0.20033760219812394,
      "num_tokens": 57947548.0,
      "step": 28590
    },
    {
      "entropy": 5.991018867492675,
      "epoch": 2.4989076291182384,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004374170426427734,
      "loss": 5.2239,
      "mean_token_accuracy": 0.20910482257604598,
      "num_tokens": 57957598.0,
      "step": 28595
    },
    {
      "entropy": 6.009394311904908,
      "epoch": 2.499344577470943,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043739547617592344,
      "loss": 5.3476,
      "mean_token_accuracy": 0.19960928410291673,
      "num_tokens": 57967790.0,
      "step": 28600
    },
    {
      "entropy": 6.001416730880737,
      "epoch": 2.4997815258236478,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043737390659424316,
      "loss": 5.2694,
      "mean_token_accuracy": 0.2059193179011345,
      "num_tokens": 57977904.0,
      "step": 28605
    },
    {
      "entropy": 5.9319658279418945,
      "epoch": 2.5002184741763522,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043735233389814623,
      "loss": 5.1689,
      "mean_token_accuracy": 0.2143913120031357,
      "num_tokens": 57987882.0,
      "step": 28610
    },
    {
      "entropy": 5.93975830078125,
      "epoch": 2.500655422529057,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043733075808804643,
      "loss": 5.2876,
      "mean_token_accuracy": 0.2108922317624092,
      "num_tokens": 57997376.0,
      "step": 28615
    },
    {
      "entropy": 5.922978639602661,
      "epoch": 2.5010923708817616,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043730917916435776,
      "loss": 5.214,
      "mean_token_accuracy": 0.2086710274219513,
      "num_tokens": 58008115.0,
      "step": 28620
    },
    {
      "entropy": 6.04396333694458,
      "epoch": 2.5015293192344665,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004372875971274939,
      "loss": 5.3551,
      "mean_token_accuracy": 0.19359831809997557,
      "num_tokens": 58017922.0,
      "step": 28625
    },
    {
      "entropy": 6.009918260574341,
      "epoch": 2.5019662675871714,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004372660119778689,
      "loss": 5.3002,
      "mean_token_accuracy": 0.1996999740600586,
      "num_tokens": 58028208.0,
      "step": 28630
    },
    {
      "entropy": 6.031697702407837,
      "epoch": 2.502403215939876,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004372444237158968,
      "loss": 5.3439,
      "mean_token_accuracy": 0.19349271059036255,
      "num_tokens": 58037237.0,
      "step": 28635
    },
    {
      "entropy": 5.96979718208313,
      "epoch": 2.5028401642925804,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004372228323419916,
      "loss": 5.2884,
      "mean_token_accuracy": 0.2032451957464218,
      "num_tokens": 58048083.0,
      "step": 28640
    },
    {
      "entropy": 6.027090167999267,
      "epoch": 2.5032771126452853,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004372012378565674,
      "loss": 5.3409,
      "mean_token_accuracy": 0.19219769537448883,
      "num_tokens": 58058593.0,
      "step": 28645
    },
    {
      "entropy": 6.037117910385132,
      "epoch": 2.50371406099799,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043717964026003856,
      "loss": 5.2825,
      "mean_token_accuracy": 0.20567082166671752,
      "num_tokens": 58068996.0,
      "step": 28650
    },
    {
      "entropy": 6.0176819324493405,
      "epoch": 2.5041510093506947,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004371580395528192,
      "loss": 5.2897,
      "mean_token_accuracy": 0.19651248753070832,
      "num_tokens": 58078546.0,
      "step": 28655
    },
    {
      "entropy": 6.067715740203857,
      "epoch": 2.5045879577033996,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043713643573532377,
      "loss": 5.3782,
      "mean_token_accuracy": 0.18906777799129487,
      "num_tokens": 58087824.0,
      "step": 28660
    },
    {
      "entropy": 5.969485759735107,
      "epoch": 2.505024906056104,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043711482880796643,
      "loss": 5.3175,
      "mean_token_accuracy": 0.2084141954779625,
      "num_tokens": 58097940.0,
      "step": 28665
    },
    {
      "entropy": 6.019999742507935,
      "epoch": 2.505461854408809,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004370932187711618,
      "loss": 5.2995,
      "mean_token_accuracy": 0.19574492424726486,
      "num_tokens": 58109274.0,
      "step": 28670
    },
    {
      "entropy": 6.0234791278839115,
      "epoch": 2.5058988027615134,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004370716056253242,
      "loss": 5.2829,
      "mean_token_accuracy": 0.2032615438103676,
      "num_tokens": 58119061.0,
      "step": 28675
    },
    {
      "entropy": 5.992642259597778,
      "epoch": 2.5063357511142184,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043704998937086844,
      "loss": 5.2765,
      "mean_token_accuracy": 0.2003870725631714,
      "num_tokens": 58128949.0,
      "step": 28680
    },
    {
      "entropy": 5.9403791427612305,
      "epoch": 2.5067726994669233,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004370283700082088,
      "loss": 5.1865,
      "mean_token_accuracy": 0.21150230318307878,
      "num_tokens": 58139807.0,
      "step": 28685
    },
    {
      "entropy": 6.006299638748169,
      "epoch": 2.5072096478196277,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004370067475377602,
      "loss": 5.3033,
      "mean_token_accuracy": 0.20368543565273284,
      "num_tokens": 58150878.0,
      "step": 28690
    },
    {
      "entropy": 5.965108871459961,
      "epoch": 2.507646596172332,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004369851219599373,
      "loss": 5.2854,
      "mean_token_accuracy": 0.20590656399726867,
      "num_tokens": 58160805.0,
      "step": 28695
    },
    {
      "entropy": 6.024855089187622,
      "epoch": 2.508083544525037,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043696349327515484,
      "loss": 5.2267,
      "mean_token_accuracy": 0.20058587491512297,
      "num_tokens": 58170298.0,
      "step": 28700
    },
    {
      "entropy": 5.934194707870484,
      "epoch": 2.508520492877742,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043694186148382773,
      "loss": 5.2628,
      "mean_token_accuracy": 0.21212013214826583,
      "num_tokens": 58180497.0,
      "step": 28705
    },
    {
      "entropy": 6.042874765396118,
      "epoch": 2.5089574412304465,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043692022658637085,
      "loss": 5.3295,
      "mean_token_accuracy": 0.2003206804394722,
      "num_tokens": 58190871.0,
      "step": 28710
    },
    {
      "entropy": 5.996347188949585,
      "epoch": 2.5093943895831514,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043689858858319923,
      "loss": 5.272,
      "mean_token_accuracy": 0.20302024036645888,
      "num_tokens": 58201654.0,
      "step": 28715
    },
    {
      "entropy": 6.018990182876587,
      "epoch": 2.509831337935856,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043687694747472785,
      "loss": 5.3499,
      "mean_token_accuracy": 0.20431492030620574,
      "num_tokens": 58211647.0,
      "step": 28720
    },
    {
      "entropy": 5.967149305343628,
      "epoch": 2.510268286288561,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043685530326137173,
      "loss": 5.2724,
      "mean_token_accuracy": 0.20427832156419753,
      "num_tokens": 58221945.0,
      "step": 28725
    },
    {
      "entropy": 6.008641147613526,
      "epoch": 2.5107052346412653,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043683365594354614,
      "loss": 5.3147,
      "mean_token_accuracy": 0.1994442656636238,
      "num_tokens": 58231702.0,
      "step": 28730
    },
    {
      "entropy": 6.008039331436157,
      "epoch": 2.51114218299397,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004368120055216662,
      "loss": 5.3534,
      "mean_token_accuracy": 0.19244770109653472,
      "num_tokens": 58242822.0,
      "step": 28735
    },
    {
      "entropy": 5.977169418334961,
      "epoch": 2.511579131346675,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004367903519961472,
      "loss": 5.2354,
      "mean_token_accuracy": 0.2115867793560028,
      "num_tokens": 58253221.0,
      "step": 28740
    },
    {
      "entropy": 6.0088661193847654,
      "epoch": 2.5120160796993796,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004367686953674045,
      "loss": 5.2268,
      "mean_token_accuracy": 0.20640773326158524,
      "num_tokens": 58263739.0,
      "step": 28745
    },
    {
      "entropy": 5.955376148223877,
      "epoch": 2.512453028052084,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043674703563585345,
      "loss": 5.2473,
      "mean_token_accuracy": 0.21004377901554108,
      "num_tokens": 58273526.0,
      "step": 28750
    },
    {
      "entropy": 5.907234859466553,
      "epoch": 2.512889976404789,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004367253728019095,
      "loss": 5.2795,
      "mean_token_accuracy": 0.2079811468720436,
      "num_tokens": 58283373.0,
      "step": 28755
    },
    {
      "entropy": 6.013519239425659,
      "epoch": 2.513326924757494,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043670370686598814,
      "loss": 5.3358,
      "mean_token_accuracy": 0.20489245653152466,
      "num_tokens": 58293365.0,
      "step": 28760
    },
    {
      "entropy": 5.992678213119507,
      "epoch": 2.5137638731101983,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043668203782850495,
      "loss": 5.2645,
      "mean_token_accuracy": 0.20818962007761002,
      "num_tokens": 58303750.0,
      "step": 28765
    },
    {
      "entropy": 6.012611532211304,
      "epoch": 2.5142008214629032,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004366603656898755,
      "loss": 5.2989,
      "mean_token_accuracy": 0.20458375364542009,
      "num_tokens": 58313209.0,
      "step": 28770
    },
    {
      "entropy": 5.990968704223633,
      "epoch": 2.5146377698156077,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043663869045051565,
      "loss": 5.3036,
      "mean_token_accuracy": 0.20619731098413469,
      "num_tokens": 58322641.0,
      "step": 28775
    },
    {
      "entropy": 6.010357093811035,
      "epoch": 2.5150747181683126,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043661701211084096,
      "loss": 5.1835,
      "mean_token_accuracy": 0.21533506363630295,
      "num_tokens": 58331865.0,
      "step": 28780
    },
    {
      "entropy": 5.999076414108276,
      "epoch": 2.515511666521017,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043659533067126735,
      "loss": 5.3534,
      "mean_token_accuracy": 0.19763250201940535,
      "num_tokens": 58341188.0,
      "step": 28785
    },
    {
      "entropy": 5.968057441711426,
      "epoch": 2.515948614873722,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004365736461322105,
      "loss": 5.3296,
      "mean_token_accuracy": 0.19778048247098923,
      "num_tokens": 58351856.0,
      "step": 28790
    },
    {
      "entropy": 5.982862710952759,
      "epoch": 2.5163855632264265,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004365519584940866,
      "loss": 5.3207,
      "mean_token_accuracy": 0.19849742501974105,
      "num_tokens": 58363066.0,
      "step": 28795
    },
    {
      "entropy": 6.0436324119567875,
      "epoch": 2.5168225115791314,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004365302677573114,
      "loss": 5.2795,
      "mean_token_accuracy": 0.19834593087434768,
      "num_tokens": 58374646.0,
      "step": 28800
    },
    {
      "entropy": 5.978626441955567,
      "epoch": 2.517259459931836,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000436508573922301,
      "loss": 5.2653,
      "mean_token_accuracy": 0.20275369584560393,
      "num_tokens": 58385192.0,
      "step": 28805
    },
    {
      "entropy": 6.014966201782227,
      "epoch": 2.5176964082845408,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043648687698947153,
      "loss": 5.2685,
      "mean_token_accuracy": 0.2105432242155075,
      "num_tokens": 58394237.0,
      "step": 28810
    },
    {
      "entropy": 5.946484613418579,
      "epoch": 2.5181333566372457,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043646517695923923,
      "loss": 5.2472,
      "mean_token_accuracy": 0.21401653736829757,
      "num_tokens": 58404229.0,
      "step": 28815
    },
    {
      "entropy": 5.954684448242188,
      "epoch": 2.51857030498995,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004364434738320201,
      "loss": 5.302,
      "mean_token_accuracy": 0.20287141501903533,
      "num_tokens": 58414706.0,
      "step": 28820
    },
    {
      "entropy": 6.044156837463379,
      "epoch": 2.5190072533426546,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043642176760823066,
      "loss": 5.3246,
      "mean_token_accuracy": 0.19689682275056838,
      "num_tokens": 58425699.0,
      "step": 28825
    },
    {
      "entropy": 6.056179666519165,
      "epoch": 2.5194442016953595,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000436400058288287,
      "loss": 5.3184,
      "mean_token_accuracy": 0.19560938626527785,
      "num_tokens": 58435923.0,
      "step": 28830
    },
    {
      "entropy": 5.980951309204102,
      "epoch": 2.5198811500480645,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004363783458726058,
      "loss": 5.2837,
      "mean_token_accuracy": 0.20322039276361464,
      "num_tokens": 58445569.0,
      "step": 28835
    },
    {
      "entropy": 6.006072568893432,
      "epoch": 2.520318098400769,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043635663036160324,
      "loss": 5.264,
      "mean_token_accuracy": 0.20772574990987777,
      "num_tokens": 58456686.0,
      "step": 28840
    },
    {
      "entropy": 5.937897872924805,
      "epoch": 2.520755046753474,
      "grad_norm": 1.1875,
      "learning_rate": 0.000436334911755696,
      "loss": 5.2103,
      "mean_token_accuracy": 0.20926159173250197,
      "num_tokens": 58466148.0,
      "step": 28845
    },
    {
      "entropy": 6.021105337142944,
      "epoch": 2.5211919951061783,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004363131900553006,
      "loss": 5.2807,
      "mean_token_accuracy": 0.20273209363222122,
      "num_tokens": 58475161.0,
      "step": 28850
    },
    {
      "entropy": 6.035284423828125,
      "epoch": 2.521628943458883,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004362914652608338,
      "loss": 5.2405,
      "mean_token_accuracy": 0.20877001136541368,
      "num_tokens": 58483631.0,
      "step": 28855
    },
    {
      "entropy": 5.926829290390015,
      "epoch": 2.5220658918115877,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004362697373727121,
      "loss": 5.2607,
      "mean_token_accuracy": 0.20477329045534134,
      "num_tokens": 58493063.0,
      "step": 28860
    },
    {
      "entropy": 6.002942371368408,
      "epoch": 2.5225028401642926,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004362480063913524,
      "loss": 5.3672,
      "mean_token_accuracy": 0.19348460733890532,
      "num_tokens": 58503658.0,
      "step": 28865
    },
    {
      "entropy": 5.997269868850708,
      "epoch": 2.5229397885169975,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043622627231717126,
      "loss": 5.2526,
      "mean_token_accuracy": 0.20597067028284072,
      "num_tokens": 58515083.0,
      "step": 28870
    },
    {
      "entropy": 6.075941038131714,
      "epoch": 2.523376736869702,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004362045351505859,
      "loss": 5.317,
      "mean_token_accuracy": 0.204978783428669,
      "num_tokens": 58524984.0,
      "step": 28875
    },
    {
      "entropy": 6.119864749908447,
      "epoch": 2.5238136852224065,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004361827948920131,
      "loss": 5.3706,
      "mean_token_accuracy": 0.19433972388505935,
      "num_tokens": 58535651.0,
      "step": 28880
    },
    {
      "entropy": 6.012279033660889,
      "epoch": 2.5242506335751114,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043616105154186974,
      "loss": 5.3619,
      "mean_token_accuracy": 0.20019068866968154,
      "num_tokens": 58546380.0,
      "step": 28885
    },
    {
      "entropy": 5.980023622512817,
      "epoch": 2.5246875819278163,
      "grad_norm": 1.25,
      "learning_rate": 0.00043613930510057305,
      "loss": 5.2587,
      "mean_token_accuracy": 0.2065039187669754,
      "num_tokens": 58555639.0,
      "step": 28890
    },
    {
      "entropy": 5.984524488449097,
      "epoch": 2.5251245302805208,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00043611755556853993,
      "loss": 5.2997,
      "mean_token_accuracy": 0.20355966538190842,
      "num_tokens": 58564948.0,
      "step": 28895
    },
    {
      "entropy": 6.038009738922119,
      "epoch": 2.5255614786332257,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004360958029461878,
      "loss": 5.3662,
      "mean_token_accuracy": 0.19853207617998123,
      "num_tokens": 58574146.0,
      "step": 28900
    },
    {
      "entropy": 6.057765483856201,
      "epoch": 2.52599842698593,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004360740472339336,
      "loss": 5.254,
      "mean_token_accuracy": 0.20740454643964767,
      "num_tokens": 58583438.0,
      "step": 28905
    },
    {
      "entropy": 5.965974378585815,
      "epoch": 2.526435375338635,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043605228843219494,
      "loss": 5.2444,
      "mean_token_accuracy": 0.20607746243476868,
      "num_tokens": 58594552.0,
      "step": 28910
    },
    {
      "entropy": 6.002703619003296,
      "epoch": 2.5268723236913395,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004360305265413889,
      "loss": 5.261,
      "mean_token_accuracy": 0.2007711187005043,
      "num_tokens": 58604162.0,
      "step": 28915
    },
    {
      "entropy": 6.001442956924438,
      "epoch": 2.5273092720440444,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004360087615619331,
      "loss": 5.2249,
      "mean_token_accuracy": 0.2075551152229309,
      "num_tokens": 58613487.0,
      "step": 28920
    },
    {
      "entropy": 5.913831520080566,
      "epoch": 2.5277462203967493,
      "grad_norm": 1.28125,
      "learning_rate": 0.00043598699349424474,
      "loss": 5.1571,
      "mean_token_accuracy": 0.22460875362157823,
      "num_tokens": 58623252.0,
      "step": 28925
    },
    {
      "entropy": 6.04638638496399,
      "epoch": 2.528183168749454,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004359652223387415,
      "loss": 5.3532,
      "mean_token_accuracy": 0.19211871027946473,
      "num_tokens": 58633585.0,
      "step": 28930
    },
    {
      "entropy": 5.971201229095459,
      "epoch": 2.5286201171021583,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004359434480958411,
      "loss": 5.2893,
      "mean_token_accuracy": 0.20124495327472686,
      "num_tokens": 58643202.0,
      "step": 28935
    },
    {
      "entropy": 5.912626886367798,
      "epoch": 2.529057065454863,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004359216707659608,
      "loss": 5.22,
      "mean_token_accuracy": 0.21215977370738984,
      "num_tokens": 58653153.0,
      "step": 28940
    },
    {
      "entropy": 5.976848459243774,
      "epoch": 2.529494013807568,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004358998903495188,
      "loss": 5.162,
      "mean_token_accuracy": 0.2151267036795616,
      "num_tokens": 58663835.0,
      "step": 28945
    },
    {
      "entropy": 6.036446523666382,
      "epoch": 2.5299309621602726,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004358781068469324,
      "loss": 5.2643,
      "mean_token_accuracy": 0.20489258468151092,
      "num_tokens": 58673114.0,
      "step": 28950
    },
    {
      "entropy": 6.0173276424407955,
      "epoch": 2.5303679105129775,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043585632025861963,
      "loss": 5.2583,
      "mean_token_accuracy": 0.20608681738376616,
      "num_tokens": 58683121.0,
      "step": 28955
    },
    {
      "entropy": 5.996326637268067,
      "epoch": 2.530804858865682,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004358345305849984,
      "loss": 5.3138,
      "mean_token_accuracy": 0.19838588684797287,
      "num_tokens": 58693656.0,
      "step": 28960
    },
    {
      "entropy": 5.935864210128784,
      "epoch": 2.531241807218387,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043581273782648664,
      "loss": 5.2756,
      "mean_token_accuracy": 0.21135860234498977,
      "num_tokens": 58703061.0,
      "step": 28965
    },
    {
      "entropy": 6.029773092269897,
      "epoch": 2.5316787555710913,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00043579094198350224,
      "loss": 5.2969,
      "mean_token_accuracy": 0.20262625217437744,
      "num_tokens": 58712463.0,
      "step": 28970
    },
    {
      "entropy": 6.035421085357666,
      "epoch": 2.5321157039237963,
      "grad_norm": 1.453125,
      "learning_rate": 0.00043576914305646334,
      "loss": 5.2258,
      "mean_token_accuracy": 0.20511217266321183,
      "num_tokens": 58722214.0,
      "step": 28975
    },
    {
      "entropy": 5.997640562057495,
      "epoch": 2.532552652276501,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000435747341045788,
      "loss": 5.3327,
      "mean_token_accuracy": 0.1971171885728836,
      "num_tokens": 58732042.0,
      "step": 28980
    },
    {
      "entropy": 6.004339599609375,
      "epoch": 2.5329896006292056,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043572553595189444,
      "loss": 5.3849,
      "mean_token_accuracy": 0.19019940942525865,
      "num_tokens": 58743139.0,
      "step": 28985
    },
    {
      "entropy": 6.087764644622803,
      "epoch": 2.53342654898191,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043570372777520096,
      "loss": 5.3322,
      "mean_token_accuracy": 0.1949450343847275,
      "num_tokens": 58752994.0,
      "step": 28990
    },
    {
      "entropy": 6.004376697540283,
      "epoch": 2.533863497334615,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043568191651612573,
      "loss": 5.3183,
      "mean_token_accuracy": 0.19524784684181212,
      "num_tokens": 58762664.0,
      "step": 28995
    },
    {
      "entropy": 6.00140414237976,
      "epoch": 2.53430044568732,
      "grad_norm": 1.125,
      "learning_rate": 0.00043566010217508706,
      "loss": 5.3165,
      "mean_token_accuracy": 0.20162206888198853,
      "num_tokens": 58772992.0,
      "step": 29000
    },
    {
      "entropy": 6.001733875274658,
      "epoch": 2.5347373940400244,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004356382847525035,
      "loss": 5.2994,
      "mean_token_accuracy": 0.20336807072162627,
      "num_tokens": 58783157.0,
      "step": 29005
    },
    {
      "entropy": 6.066930437088013,
      "epoch": 2.5351743423927293,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004356164642487935,
      "loss": 5.3485,
      "mean_token_accuracy": 0.19639089405536653,
      "num_tokens": 58794020.0,
      "step": 29010
    },
    {
      "entropy": 5.931814670562744,
      "epoch": 2.535611290745434,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004355946406643755,
      "loss": 5.2373,
      "mean_token_accuracy": 0.2106218159198761,
      "num_tokens": 58804524.0,
      "step": 29015
    },
    {
      "entropy": 5.9624279022216795,
      "epoch": 2.5360482390981387,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043557281399966814,
      "loss": 5.3068,
      "mean_token_accuracy": 0.19757490158081054,
      "num_tokens": 58814278.0,
      "step": 29020
    },
    {
      "entropy": 5.989673471450805,
      "epoch": 2.536485187450843,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004355509842550901,
      "loss": 5.2183,
      "mean_token_accuracy": 0.20783993005752563,
      "num_tokens": 58824258.0,
      "step": 29025
    },
    {
      "entropy": 5.973184728622437,
      "epoch": 2.536922135803548,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004355291514310601,
      "loss": 5.3184,
      "mean_token_accuracy": 0.20690534114837647,
      "num_tokens": 58834316.0,
      "step": 29030
    },
    {
      "entropy": 5.996721935272217,
      "epoch": 2.537359084156253,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043550731552799676,
      "loss": 5.3078,
      "mean_token_accuracy": 0.20334005057811738,
      "num_tokens": 58844201.0,
      "step": 29035
    },
    {
      "entropy": 6.0121218204498295,
      "epoch": 2.5377960325089575,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043548547654631917,
      "loss": 5.2395,
      "mean_token_accuracy": 0.2111138179898262,
      "num_tokens": 58853675.0,
      "step": 29040
    },
    {
      "entropy": 6.010968351364136,
      "epoch": 2.538232980861662,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004354636344864459,
      "loss": 5.2828,
      "mean_token_accuracy": 0.20267933011054992,
      "num_tokens": 58863316.0,
      "step": 29045
    },
    {
      "entropy": 6.013712167739868,
      "epoch": 2.538669929214367,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004354417893487961,
      "loss": 5.2295,
      "mean_token_accuracy": 0.20431957244873047,
      "num_tokens": 58873479.0,
      "step": 29050
    },
    {
      "entropy": 5.977614831924439,
      "epoch": 2.5391068775670718,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004354199411337886,
      "loss": 5.2721,
      "mean_token_accuracy": 0.20235940665006638,
      "num_tokens": 58882540.0,
      "step": 29055
    },
    {
      "entropy": 6.000927352905274,
      "epoch": 2.5395438259197762,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004353980898418427,
      "loss": 5.2716,
      "mean_token_accuracy": 0.19997825473546982,
      "num_tokens": 58891565.0,
      "step": 29060
    },
    {
      "entropy": 5.995136547088623,
      "epoch": 2.539980774272481,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004353762354733773,
      "loss": 5.2746,
      "mean_token_accuracy": 0.1975442036986351,
      "num_tokens": 58900689.0,
      "step": 29065
    },
    {
      "entropy": 6.004953098297119,
      "epoch": 2.5404177226251856,
      "grad_norm": 1.25,
      "learning_rate": 0.00043535437802881175,
      "loss": 5.2632,
      "mean_token_accuracy": 0.20850737094879152,
      "num_tokens": 58910588.0,
      "step": 29070
    },
    {
      "entropy": 6.0128782272338865,
      "epoch": 2.5408546709778905,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004353325175085651,
      "loss": 5.3586,
      "mean_token_accuracy": 0.19596395790576934,
      "num_tokens": 58921189.0,
      "step": 29075
    },
    {
      "entropy": 6.0282622337341305,
      "epoch": 2.541291619330595,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043531065391305685,
      "loss": 5.2639,
      "mean_token_accuracy": 0.20279860645532607,
      "num_tokens": 58930941.0,
      "step": 29080
    },
    {
      "entropy": 6.0368812561035154,
      "epoch": 2.5417285676833,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004352887872427063,
      "loss": 5.4014,
      "mean_token_accuracy": 0.20313200056552888,
      "num_tokens": 58940371.0,
      "step": 29085
    },
    {
      "entropy": 5.985775184631348,
      "epoch": 2.5421655160360044,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043526691749793273,
      "loss": 5.2813,
      "mean_token_accuracy": 0.19663693904876708,
      "num_tokens": 58950446.0,
      "step": 29090
    },
    {
      "entropy": 5.979206609725952,
      "epoch": 2.5426024643887093,
      "grad_norm": 1.25,
      "learning_rate": 0.0004352450446791557,
      "loss": 5.2495,
      "mean_token_accuracy": 0.20209547728300095,
      "num_tokens": 58960757.0,
      "step": 29095
    },
    {
      "entropy": 6.030920791625976,
      "epoch": 2.5430394127414138,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043522316878679475,
      "loss": 5.3211,
      "mean_token_accuracy": 0.2020957127213478,
      "num_tokens": 58971621.0,
      "step": 29100
    },
    {
      "entropy": 6.069017362594605,
      "epoch": 2.5434763610941187,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043520128982126944,
      "loss": 5.3114,
      "mean_token_accuracy": 0.20377016067504883,
      "num_tokens": 58982505.0,
      "step": 29105
    },
    {
      "entropy": 5.977510643005371,
      "epoch": 2.5439133094468236,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004351794077829995,
      "loss": 5.2215,
      "mean_token_accuracy": 0.2095091611146927,
      "num_tokens": 58992270.0,
      "step": 29110
    },
    {
      "entropy": 5.8772499561309814,
      "epoch": 2.544350257799528,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004351575226724046,
      "loss": 5.1789,
      "mean_token_accuracy": 0.21208302527666092,
      "num_tokens": 59002510.0,
      "step": 29115
    },
    {
      "entropy": 5.9878486633300785,
      "epoch": 2.544787206152233,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004351356344899045,
      "loss": 5.27,
      "mean_token_accuracy": 0.2083282694220543,
      "num_tokens": 59012942.0,
      "step": 29120
    },
    {
      "entropy": 6.045032358169555,
      "epoch": 2.5452241545049374,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043511374323591904,
      "loss": 5.3829,
      "mean_token_accuracy": 0.19508789032697677,
      "num_tokens": 59023509.0,
      "step": 29125
    },
    {
      "entropy": 6.047530221939087,
      "epoch": 2.5456611028576424,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043509184891086795,
      "loss": 5.2576,
      "mean_token_accuracy": 0.2093736633658409,
      "num_tokens": 59033782.0,
      "step": 29130
    },
    {
      "entropy": 5.993314123153686,
      "epoch": 2.546098051210347,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043506995151517135,
      "loss": 5.3365,
      "mean_token_accuracy": 0.1928073912858963,
      "num_tokens": 59045791.0,
      "step": 29135
    },
    {
      "entropy": 5.991881084442139,
      "epoch": 2.5465349995630517,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00043504805104924925,
      "loss": 5.2527,
      "mean_token_accuracy": 0.20697390735149385,
      "num_tokens": 59056578.0,
      "step": 29140
    },
    {
      "entropy": 6.044497013092041,
      "epoch": 2.546971947915756,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004350261475135217,
      "loss": 5.3137,
      "mean_token_accuracy": 0.1981563687324524,
      "num_tokens": 59066712.0,
      "step": 29145
    },
    {
      "entropy": 5.993666934967041,
      "epoch": 2.547408896268461,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004350042409084088,
      "loss": 5.2345,
      "mean_token_accuracy": 0.21100791692733764,
      "num_tokens": 59077557.0,
      "step": 29150
    },
    {
      "entropy": 5.962836456298828,
      "epoch": 2.5478458446211656,
      "grad_norm": 1.25,
      "learning_rate": 0.00043498233123433064,
      "loss": 5.2737,
      "mean_token_accuracy": 0.20565235316753389,
      "num_tokens": 59086649.0,
      "step": 29155
    },
    {
      "entropy": 5.993390655517578,
      "epoch": 2.5482827929738705,
      "grad_norm": 1.25,
      "learning_rate": 0.00043496041849170757,
      "loss": 5.2552,
      "mean_token_accuracy": 0.2068120077252388,
      "num_tokens": 59096838.0,
      "step": 29160
    },
    {
      "entropy": 6.023121976852417,
      "epoch": 2.5487197413265754,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004349385026809599,
      "loss": 5.2845,
      "mean_token_accuracy": 0.20792946815490723,
      "num_tokens": 59107884.0,
      "step": 29165
    },
    {
      "entropy": 5.993524360656738,
      "epoch": 2.54915668967928,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043491658380250784,
      "loss": 5.316,
      "mean_token_accuracy": 0.19573152363300322,
      "num_tokens": 59118315.0,
      "step": 29170
    },
    {
      "entropy": 5.933906984329224,
      "epoch": 2.5495936380319844,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043489466185677193,
      "loss": 5.2672,
      "mean_token_accuracy": 0.20706234276294708,
      "num_tokens": 59128331.0,
      "step": 29175
    },
    {
      "entropy": 5.984966135025024,
      "epoch": 2.5500305863846893,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004348727368441727,
      "loss": 5.2531,
      "mean_token_accuracy": 0.20602380484342575,
      "num_tokens": 59138643.0,
      "step": 29180
    },
    {
      "entropy": 6.002346801757812,
      "epoch": 2.550467534737394,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004348508087651306,
      "loss": 5.2376,
      "mean_token_accuracy": 0.2081763818860054,
      "num_tokens": 59148364.0,
      "step": 29185
    },
    {
      "entropy": 5.985355806350708,
      "epoch": 2.5509044830900987,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043482887762006614,
      "loss": 5.29,
      "mean_token_accuracy": 0.20645319372415544,
      "num_tokens": 59158317.0,
      "step": 29190
    },
    {
      "entropy": 6.053474855422974,
      "epoch": 2.5513414314428036,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043480694340940017,
      "loss": 5.363,
      "mean_token_accuracy": 0.19882579147815704,
      "num_tokens": 59168879.0,
      "step": 29195
    },
    {
      "entropy": 6.046101093292236,
      "epoch": 2.551778379795508,
      "grad_norm": 1.25,
      "learning_rate": 0.0004347850061335532,
      "loss": 5.2465,
      "mean_token_accuracy": 0.20634061694145203,
      "num_tokens": 59179248.0,
      "step": 29200
    },
    {
      "entropy": 5.910547685623169,
      "epoch": 2.552215328148213,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004347630657929461,
      "loss": 5.1843,
      "mean_token_accuracy": 0.21426410973072052,
      "num_tokens": 59189217.0,
      "step": 29205
    },
    {
      "entropy": 6.003809642791748,
      "epoch": 2.5526522765009174,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043474112238799974,
      "loss": 5.2884,
      "mean_token_accuracy": 0.20661652684211732,
      "num_tokens": 59199857.0,
      "step": 29210
    },
    {
      "entropy": 6.035441780090332,
      "epoch": 2.5530892248536223,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043471917591913486,
      "loss": 5.2917,
      "mean_token_accuracy": 0.20399415940046312,
      "num_tokens": 59210119.0,
      "step": 29215
    },
    {
      "entropy": 5.973721504211426,
      "epoch": 2.5535261732063272,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043469722638677256,
      "loss": 5.2251,
      "mean_token_accuracy": 0.2067856341600418,
      "num_tokens": 59219258.0,
      "step": 29220
    },
    {
      "entropy": 5.956931257247925,
      "epoch": 2.5539631215590317,
      "grad_norm": 1.796875,
      "learning_rate": 0.00043467527379133377,
      "loss": 5.317,
      "mean_token_accuracy": 0.2007964074611664,
      "num_tokens": 59229664.0,
      "step": 29225
    },
    {
      "entropy": 6.030935478210449,
      "epoch": 2.554400069911736,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004346533181332395,
      "loss": 5.3142,
      "mean_token_accuracy": 0.20140854716300965,
      "num_tokens": 59239699.0,
      "step": 29230
    },
    {
      "entropy": 6.02437195777893,
      "epoch": 2.554837018264441,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043463135941291105,
      "loss": 5.2429,
      "mean_token_accuracy": 0.20445796251296997,
      "num_tokens": 59249891.0,
      "step": 29235
    },
    {
      "entropy": 5.992830801010132,
      "epoch": 2.555273966617146,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004346093976307694,
      "loss": 5.3023,
      "mean_token_accuracy": 0.20125265866518022,
      "num_tokens": 59260092.0,
      "step": 29240
    },
    {
      "entropy": 5.962039375305176,
      "epoch": 2.5557109149698505,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004345874327872359,
      "loss": 5.339,
      "mean_token_accuracy": 0.20104550272226335,
      "num_tokens": 59268844.0,
      "step": 29245
    },
    {
      "entropy": 6.031727886199951,
      "epoch": 2.5561478633225554,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004345654648827318,
      "loss": 5.3794,
      "mean_token_accuracy": 0.18991228938102722,
      "num_tokens": 59279275.0,
      "step": 29250
    },
    {
      "entropy": 6.039046573638916,
      "epoch": 2.55658481167526,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004345434939176784,
      "loss": 5.2855,
      "mean_token_accuracy": 0.20812391638755798,
      "num_tokens": 59290596.0,
      "step": 29255
    },
    {
      "entropy": 5.985118675231933,
      "epoch": 2.5570217600279648,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043452151989249727,
      "loss": 5.217,
      "mean_token_accuracy": 0.21255007088184358,
      "num_tokens": 59300642.0,
      "step": 29260
    },
    {
      "entropy": 5.947776365280151,
      "epoch": 2.5574587083806692,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004344995428076098,
      "loss": 5.3149,
      "mean_token_accuracy": 0.20589704364538192,
      "num_tokens": 59310064.0,
      "step": 29265
    },
    {
      "entropy": 6.062455463409424,
      "epoch": 2.557895656733374,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043447756266343755,
      "loss": 5.2976,
      "mean_token_accuracy": 0.1990237608551979,
      "num_tokens": 59319114.0,
      "step": 29270
    },
    {
      "entropy": 6.006913852691651,
      "epoch": 2.558332605086079,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043445557946040207,
      "loss": 5.3438,
      "mean_token_accuracy": 0.1968172609806061,
      "num_tokens": 59329414.0,
      "step": 29275
    },
    {
      "entropy": 6.011691856384277,
      "epoch": 2.5587695534387835,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043443359319892495,
      "loss": 5.3462,
      "mean_token_accuracy": 0.19751116782426834,
      "num_tokens": 59339980.0,
      "step": 29280
    },
    {
      "entropy": 6.050130844116211,
      "epoch": 2.559206501791488,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000434411603879428,
      "loss": 5.3369,
      "mean_token_accuracy": 0.19562319964170455,
      "num_tokens": 59349754.0,
      "step": 29285
    },
    {
      "entropy": 5.977090311050415,
      "epoch": 2.559643450144193,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000434389611502333,
      "loss": 5.139,
      "mean_token_accuracy": 0.22121379822492598,
      "num_tokens": 59358542.0,
      "step": 29290
    },
    {
      "entropy": 5.991571426391602,
      "epoch": 2.560080398496898,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004343676160680617,
      "loss": 5.2756,
      "mean_token_accuracy": 0.2049901232123375,
      "num_tokens": 59367743.0,
      "step": 29295
    },
    {
      "entropy": 5.96758623123169,
      "epoch": 2.5605173468496023,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000434345617577036,
      "loss": 5.2977,
      "mean_token_accuracy": 0.19828346520662307,
      "num_tokens": 59377622.0,
      "step": 29300
    },
    {
      "entropy": 5.966783237457276,
      "epoch": 2.560954295202307,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043432361602967787,
      "loss": 5.2093,
      "mean_token_accuracy": 0.2038247525691986,
      "num_tokens": 59388661.0,
      "step": 29305
    },
    {
      "entropy": 6.00912914276123,
      "epoch": 2.5613912435550117,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043430161142640934,
      "loss": 5.3162,
      "mean_token_accuracy": 0.1997107058763504,
      "num_tokens": 59399935.0,
      "step": 29310
    },
    {
      "entropy": 5.995010280609131,
      "epoch": 2.5618281919077166,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004342796037676524,
      "loss": 5.2662,
      "mean_token_accuracy": 0.20855874121189116,
      "num_tokens": 59409513.0,
      "step": 29315
    },
    {
      "entropy": 5.98572940826416,
      "epoch": 2.562265140260421,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004342575930538292,
      "loss": 5.2455,
      "mean_token_accuracy": 0.20627597570419312,
      "num_tokens": 59420158.0,
      "step": 29320
    },
    {
      "entropy": 5.989929437637329,
      "epoch": 2.562702088613126,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043423557928536197,
      "loss": 5.3331,
      "mean_token_accuracy": 0.20262201577425004,
      "num_tokens": 59429440.0,
      "step": 29325
    },
    {
      "entropy": 5.930717325210571,
      "epoch": 2.563139036965831,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004342135624626728,
      "loss": 5.2159,
      "mean_token_accuracy": 0.20798189342021942,
      "num_tokens": 59439014.0,
      "step": 29330
    },
    {
      "entropy": 5.954394578933716,
      "epoch": 2.5635759853185354,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004341915425861841,
      "loss": 5.2277,
      "mean_token_accuracy": 0.209067901968956,
      "num_tokens": 59448714.0,
      "step": 29335
    },
    {
      "entropy": 5.982610511779785,
      "epoch": 2.56401293367124,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004341695196563183,
      "loss": 5.2906,
      "mean_token_accuracy": 0.20615312606096267,
      "num_tokens": 59459577.0,
      "step": 29340
    },
    {
      "entropy": 6.0026336193084715,
      "epoch": 2.5644498820239447,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043414749367349755,
      "loss": 5.2721,
      "mean_token_accuracy": 0.20476330071687698,
      "num_tokens": 59470313.0,
      "step": 29345
    },
    {
      "entropy": 6.014884948730469,
      "epoch": 2.5648868303766497,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004341254646381446,
      "loss": 5.3131,
      "mean_token_accuracy": 0.2071398451924324,
      "num_tokens": 59479484.0,
      "step": 29350
    },
    {
      "entropy": 6.015931081771851,
      "epoch": 2.565323778729354,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043410343255068177,
      "loss": 5.2703,
      "mean_token_accuracy": 0.19399604797363282,
      "num_tokens": 59490290.0,
      "step": 29355
    },
    {
      "entropy": 6.009041261672974,
      "epoch": 2.565760727082059,
      "grad_norm": 1.28125,
      "learning_rate": 0.00043408139741153175,
      "loss": 5.2934,
      "mean_token_accuracy": 0.20774788409471512,
      "num_tokens": 59501080.0,
      "step": 29360
    },
    {
      "entropy": 5.956488847732544,
      "epoch": 2.5661976754347635,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043405935922111725,
      "loss": 5.1884,
      "mean_token_accuracy": 0.21616220027208327,
      "num_tokens": 59511453.0,
      "step": 29365
    },
    {
      "entropy": 5.9766443252563475,
      "epoch": 2.5666346237874684,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004340373179798608,
      "loss": 5.3004,
      "mean_token_accuracy": 0.20873180776834488,
      "num_tokens": 59521438.0,
      "step": 29370
    },
    {
      "entropy": 5.990381050109863,
      "epoch": 2.567071572140173,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004340152736881853,
      "loss": 5.3719,
      "mean_token_accuracy": 0.19460136741399764,
      "num_tokens": 59531763.0,
      "step": 29375
    },
    {
      "entropy": 6.097714138031006,
      "epoch": 2.567508520492878,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043399322634651345,
      "loss": 5.407,
      "mean_token_accuracy": 0.1887618914246559,
      "num_tokens": 59542857.0,
      "step": 29380
    },
    {
      "entropy": 6.064117002487182,
      "epoch": 2.5679454688455827,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004339711759552684,
      "loss": 5.2708,
      "mean_token_accuracy": 0.20344517081975938,
      "num_tokens": 59552353.0,
      "step": 29385
    },
    {
      "entropy": 6.033031463623047,
      "epoch": 2.568382417198287,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004339491225148727,
      "loss": 5.3276,
      "mean_token_accuracy": 0.20224992036819459,
      "num_tokens": 59562377.0,
      "step": 29390
    },
    {
      "entropy": 6.024992561340332,
      "epoch": 2.5688193655509917,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043392706602574967,
      "loss": 5.2956,
      "mean_token_accuracy": 0.2003645732998848,
      "num_tokens": 59572141.0,
      "step": 29395
    },
    {
      "entropy": 6.0091430187225345,
      "epoch": 2.5692563139036966,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043390500648832214,
      "loss": 5.2472,
      "mean_token_accuracy": 0.19808532446622848,
      "num_tokens": 59581883.0,
      "step": 29400
    },
    {
      "entropy": 6.022067022323609,
      "epoch": 2.5696932622564015,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004338829439030134,
      "loss": 5.3347,
      "mean_token_accuracy": 0.19546492248773575,
      "num_tokens": 59591798.0,
      "step": 29405
    },
    {
      "entropy": 6.010311317443848,
      "epoch": 2.570130210609106,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004338608782702465,
      "loss": 5.2155,
      "mean_token_accuracy": 0.21313142627477646,
      "num_tokens": 59601525.0,
      "step": 29410
    },
    {
      "entropy": 5.994449949264526,
      "epoch": 2.570567158961811,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004338388095904447,
      "loss": 5.3745,
      "mean_token_accuracy": 0.1997596576809883,
      "num_tokens": 59611335.0,
      "step": 29415
    },
    {
      "entropy": 6.022313356399536,
      "epoch": 2.5710041073145153,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004338167378640314,
      "loss": 5.3332,
      "mean_token_accuracy": 0.20154132843017578,
      "num_tokens": 59620962.0,
      "step": 29420
    },
    {
      "entropy": 6.008788776397705,
      "epoch": 2.5714410556672203,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004337946630914297,
      "loss": 5.2662,
      "mean_token_accuracy": 0.19946406781673431,
      "num_tokens": 59631568.0,
      "step": 29425
    },
    {
      "entropy": 5.943725728988648,
      "epoch": 2.5718780040199247,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043377258527306316,
      "loss": 5.3151,
      "mean_token_accuracy": 0.20238372534513474,
      "num_tokens": 59642249.0,
      "step": 29430
    },
    {
      "entropy": 6.006377172470093,
      "epoch": 2.5723149523726296,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004337505044093554,
      "loss": 5.3018,
      "mean_token_accuracy": 0.1972529634833336,
      "num_tokens": 59652516.0,
      "step": 29435
    },
    {
      "entropy": 5.936118173599243,
      "epoch": 2.572751900725334,
      "grad_norm": 1.25,
      "learning_rate": 0.00043372842050072965,
      "loss": 5.159,
      "mean_token_accuracy": 0.21554419845342637,
      "num_tokens": 59662487.0,
      "step": 29440
    },
    {
      "entropy": 5.949368190765381,
      "epoch": 2.573188849078039,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004337063335476097,
      "loss": 5.3164,
      "mean_token_accuracy": 0.1989164099097252,
      "num_tokens": 59674061.0,
      "step": 29445
    },
    {
      "entropy": 6.007465744018555,
      "epoch": 2.5736257974307435,
      "grad_norm": 1.234375,
      "learning_rate": 0.000433684243550419,
      "loss": 5.3337,
      "mean_token_accuracy": 0.20291101038455964,
      "num_tokens": 59684159.0,
      "step": 29450
    },
    {
      "entropy": 6.0630879402160645,
      "epoch": 2.5740627457834484,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004336621505095815,
      "loss": 5.3384,
      "mean_token_accuracy": 0.19818453192710878,
      "num_tokens": 59693742.0,
      "step": 29455
    },
    {
      "entropy": 6.074492692947388,
      "epoch": 2.5744996941361533,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043364005442552073,
      "loss": 5.3742,
      "mean_token_accuracy": 0.1941719502210617,
      "num_tokens": 59703613.0,
      "step": 29460
    },
    {
      "entropy": 6.009429025650024,
      "epoch": 2.574936642488858,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004336179552986606,
      "loss": 5.2439,
      "mean_token_accuracy": 0.20496635138988495,
      "num_tokens": 59713409.0,
      "step": 29465
    },
    {
      "entropy": 5.940529870986938,
      "epoch": 2.5753735908415623,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00043359585312942496,
      "loss": 5.2044,
      "mean_token_accuracy": 0.21223401129245759,
      "num_tokens": 59723502.0,
      "step": 29470
    },
    {
      "entropy": 6.011353015899658,
      "epoch": 2.575810539194267,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043357374791823784,
      "loss": 5.3405,
      "mean_token_accuracy": 0.19884693026542663,
      "num_tokens": 59734069.0,
      "step": 29475
    },
    {
      "entropy": 6.02239842414856,
      "epoch": 2.576247487546972,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043355163966552303,
      "loss": 5.3146,
      "mean_token_accuracy": 0.19414470642805098,
      "num_tokens": 59743793.0,
      "step": 29480
    },
    {
      "entropy": 5.960443544387817,
      "epoch": 2.5766844358996766,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043352952837170467,
      "loss": 5.3448,
      "mean_token_accuracy": 0.19822800159454346,
      "num_tokens": 59754524.0,
      "step": 29485
    },
    {
      "entropy": 6.018378400802613,
      "epoch": 2.5771213842523815,
      "grad_norm": 1.15625,
      "learning_rate": 0.000433507414037207,
      "loss": 5.2989,
      "mean_token_accuracy": 0.20562509298324586,
      "num_tokens": 59765437.0,
      "step": 29490
    },
    {
      "entropy": 6.024921894073486,
      "epoch": 2.577558332605086,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000433485296662454,
      "loss": 5.2997,
      "mean_token_accuracy": 0.2021055966615677,
      "num_tokens": 59776136.0,
      "step": 29495
    },
    {
      "entropy": 5.959860849380493,
      "epoch": 2.577995280957791,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043346317624787007,
      "loss": 5.2519,
      "mean_token_accuracy": 0.20705827027559282,
      "num_tokens": 59786375.0,
      "step": 29500
    },
    {
      "entropy": 5.891839981079102,
      "epoch": 2.5784322293104953,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004334410527938793,
      "loss": 5.2025,
      "mean_token_accuracy": 0.2094683140516281,
      "num_tokens": 59796937.0,
      "step": 29505
    },
    {
      "entropy": 6.020949363708496,
      "epoch": 2.5788691776632002,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043341892630090607,
      "loss": 5.2932,
      "mean_token_accuracy": 0.20362658500671388,
      "num_tokens": 59806625.0,
      "step": 29510
    },
    {
      "entropy": 5.996029710769653,
      "epoch": 2.579306126015905,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004333967967693749,
      "loss": 5.3118,
      "mean_token_accuracy": 0.2018146812915802,
      "num_tokens": 59816618.0,
      "step": 29515
    },
    {
      "entropy": 6.014608383178711,
      "epoch": 2.5797430743686096,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004333746641997102,
      "loss": 5.3074,
      "mean_token_accuracy": 0.20380100458860398,
      "num_tokens": 59827666.0,
      "step": 29520
    },
    {
      "entropy": 6.013907814025879,
      "epoch": 2.580180022721314,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004333525285923363,
      "loss": 5.2769,
      "mean_token_accuracy": 0.20710485726594924,
      "num_tokens": 59837704.0,
      "step": 29525
    },
    {
      "entropy": 5.998371410369873,
      "epoch": 2.580616971074019,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004333303899476781,
      "loss": 5.2284,
      "mean_token_accuracy": 0.2019607752561569,
      "num_tokens": 59848186.0,
      "step": 29530
    },
    {
      "entropy": 6.008380174636841,
      "epoch": 2.581053919426724,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043330824826616,
      "loss": 5.2499,
      "mean_token_accuracy": 0.19954028576612473,
      "num_tokens": 59858453.0,
      "step": 29535
    },
    {
      "entropy": 5.978707313537598,
      "epoch": 2.5814908677794284,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004332861035482067,
      "loss": 5.2862,
      "mean_token_accuracy": 0.1978110522031784,
      "num_tokens": 59867863.0,
      "step": 29540
    },
    {
      "entropy": 5.948050308227539,
      "epoch": 2.5819278161321333,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000433263955794243,
      "loss": 5.2506,
      "mean_token_accuracy": 0.21248064935207367,
      "num_tokens": 59879932.0,
      "step": 29545
    },
    {
      "entropy": 6.08480052947998,
      "epoch": 2.5823647644848378,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004332418050046937,
      "loss": 5.2863,
      "mean_token_accuracy": 0.2002052292227745,
      "num_tokens": 59889586.0,
      "step": 29550
    },
    {
      "entropy": 5.981526470184326,
      "epoch": 2.5828017128375427,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004332196511799837,
      "loss": 5.2267,
      "mean_token_accuracy": 0.20712166726589204,
      "num_tokens": 59898789.0,
      "step": 29555
    },
    {
      "entropy": 5.95723295211792,
      "epoch": 2.583238661190247,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004331974943205379,
      "loss": 5.3056,
      "mean_token_accuracy": 0.20036334544420242,
      "num_tokens": 59909346.0,
      "step": 29560
    },
    {
      "entropy": 5.988166809082031,
      "epoch": 2.583675609542952,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043317533442678124,
      "loss": 5.2114,
      "mean_token_accuracy": 0.2166819006204605,
      "num_tokens": 59919979.0,
      "step": 29565
    },
    {
      "entropy": 6.034021806716919,
      "epoch": 2.584112557895657,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043315317149913876,
      "loss": 5.259,
      "mean_token_accuracy": 0.20611039996147157,
      "num_tokens": 59929855.0,
      "step": 29570
    },
    {
      "entropy": 5.965359687805176,
      "epoch": 2.5845495062483614,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043313100553803564,
      "loss": 5.2792,
      "mean_token_accuracy": 0.2031888946890831,
      "num_tokens": 59939148.0,
      "step": 29575
    },
    {
      "entropy": 5.911592197418213,
      "epoch": 2.584986454601066,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004331088365438969,
      "loss": 5.2042,
      "mean_token_accuracy": 0.2055797576904297,
      "num_tokens": 59948645.0,
      "step": 29580
    },
    {
      "entropy": 6.000574064254761,
      "epoch": 2.585423402953771,
      "grad_norm": 1.125,
      "learning_rate": 0.0004330866645171479,
      "loss": 5.2717,
      "mean_token_accuracy": 0.20445573031902314,
      "num_tokens": 59958777.0,
      "step": 29585
    },
    {
      "entropy": 5.985386657714844,
      "epoch": 2.5858603513064757,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004330644894582138,
      "loss": 5.2237,
      "mean_token_accuracy": 0.20825983434915543,
      "num_tokens": 59969176.0,
      "step": 29590
    },
    {
      "entropy": 6.012767171859741,
      "epoch": 2.58629729965918,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043304231136751995,
      "loss": 5.3044,
      "mean_token_accuracy": 0.19734978824853897,
      "num_tokens": 59979263.0,
      "step": 29595
    },
    {
      "entropy": 6.009013319015503,
      "epoch": 2.586734248011885,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043302013024549174,
      "loss": 5.2308,
      "mean_token_accuracy": 0.2137300819158554,
      "num_tokens": 59989313.0,
      "step": 29600
    },
    {
      "entropy": 6.005420160293579,
      "epoch": 2.5871711963645896,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043299794609255474,
      "loss": 5.2808,
      "mean_token_accuracy": 0.20538743138313292,
      "num_tokens": 59999410.0,
      "step": 29605
    },
    {
      "entropy": 5.934671258926391,
      "epoch": 2.5876081447172945,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004329757589091342,
      "loss": 5.2101,
      "mean_token_accuracy": 0.21224396526813508,
      "num_tokens": 60010206.0,
      "step": 29610
    },
    {
      "entropy": 5.887290143966675,
      "epoch": 2.588045093069999,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00043295356869565586,
      "loss": 5.1447,
      "mean_token_accuracy": 0.2166917309165001,
      "num_tokens": 60018692.0,
      "step": 29615
    },
    {
      "entropy": 6.017224264144898,
      "epoch": 2.588482041422704,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043293137545254536,
      "loss": 5.3154,
      "mean_token_accuracy": 0.19787376075983049,
      "num_tokens": 60028486.0,
      "step": 29620
    },
    {
      "entropy": 6.0165502548217775,
      "epoch": 2.588918989775409,
      "grad_norm": 1.125,
      "learning_rate": 0.0004329091791802283,
      "loss": 5.3051,
      "mean_token_accuracy": 0.19959578961133956,
      "num_tokens": 60039347.0,
      "step": 29625
    },
    {
      "entropy": 5.971807193756104,
      "epoch": 2.5893559381281133,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004328869798791305,
      "loss": 5.1982,
      "mean_token_accuracy": 0.21677626222372054,
      "num_tokens": 60049891.0,
      "step": 29630
    },
    {
      "entropy": 6.014204025268555,
      "epoch": 2.5897928864808177,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043286477754967757,
      "loss": 5.272,
      "mean_token_accuracy": 0.20339018255472183,
      "num_tokens": 60061264.0,
      "step": 29635
    },
    {
      "entropy": 5.96911392211914,
      "epoch": 2.5902298348335226,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043284257219229557,
      "loss": 5.2945,
      "mean_token_accuracy": 0.20241996347904206,
      "num_tokens": 60072453.0,
      "step": 29640
    },
    {
      "entropy": 5.915167999267578,
      "epoch": 2.5906667831862276,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043282036380741035,
      "loss": 5.2959,
      "mean_token_accuracy": 0.19728844612836838,
      "num_tokens": 60083619.0,
      "step": 29645
    },
    {
      "entropy": 6.067513465881348,
      "epoch": 2.591103731538932,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004327981523954478,
      "loss": 5.2198,
      "mean_token_accuracy": 0.20926510244607927,
      "num_tokens": 60093231.0,
      "step": 29650
    },
    {
      "entropy": 6.014829349517822,
      "epoch": 2.591540679891637,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000432775937956834,
      "loss": 5.2203,
      "mean_token_accuracy": 0.20914014130830766,
      "num_tokens": 60103017.0,
      "step": 29655
    },
    {
      "entropy": 5.980142974853516,
      "epoch": 2.5919776282443414,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004327537204919951,
      "loss": 5.2689,
      "mean_token_accuracy": 0.20064816027879714,
      "num_tokens": 60113095.0,
      "step": 29660
    },
    {
      "entropy": 6.026646518707276,
      "epoch": 2.5924145765970463,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004327315000013571,
      "loss": 5.3366,
      "mean_token_accuracy": 0.20287289172410966,
      "num_tokens": 60123506.0,
      "step": 29665
    },
    {
      "entropy": 6.022109270095825,
      "epoch": 2.592851524949751,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004327092764853463,
      "loss": 5.3474,
      "mean_token_accuracy": 0.194853013753891,
      "num_tokens": 60134349.0,
      "step": 29670
    },
    {
      "entropy": 6.0086729526519775,
      "epoch": 2.5932884733024557,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043268704994438897,
      "loss": 5.2704,
      "mean_token_accuracy": 0.19964538514614105,
      "num_tokens": 60143970.0,
      "step": 29675
    },
    {
      "entropy": 5.949227619171142,
      "epoch": 2.5937254216551606,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004326648203789113,
      "loss": 5.2623,
      "mean_token_accuracy": 0.20151786506175995,
      "num_tokens": 60154904.0,
      "step": 29680
    },
    {
      "entropy": 5.975140523910523,
      "epoch": 2.594162370007865,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004326425877893399,
      "loss": 5.2575,
      "mean_token_accuracy": 0.20885456800460817,
      "num_tokens": 60164747.0,
      "step": 29685
    },
    {
      "entropy": 6.037966012954712,
      "epoch": 2.5945993183605696,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043262035217610104,
      "loss": 5.2691,
      "mean_token_accuracy": 0.20672134906053544,
      "num_tokens": 60175487.0,
      "step": 29690
    },
    {
      "entropy": 6.013470649719238,
      "epoch": 2.5950362667132745,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043259811353962114,
      "loss": 5.2731,
      "mean_token_accuracy": 0.19828028678894044,
      "num_tokens": 60185216.0,
      "step": 29695
    },
    {
      "entropy": 6.003086185455322,
      "epoch": 2.5954732150659794,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004325758718803269,
      "loss": 5.2673,
      "mean_token_accuracy": 0.2027199998497963,
      "num_tokens": 60195495.0,
      "step": 29700
    },
    {
      "entropy": 6.052233505249023,
      "epoch": 2.595910163418684,
      "grad_norm": 1.25,
      "learning_rate": 0.00043255362719864487,
      "loss": 5.3339,
      "mean_token_accuracy": 0.1932450443506241,
      "num_tokens": 60205556.0,
      "step": 29705
    },
    {
      "entropy": 6.006272411346435,
      "epoch": 2.5963471117713888,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043253137949500166,
      "loss": 5.2452,
      "mean_token_accuracy": 0.2059349536895752,
      "num_tokens": 60215904.0,
      "step": 29710
    },
    {
      "entropy": 5.934910774230957,
      "epoch": 2.5967840601240932,
      "grad_norm": 1.328125,
      "learning_rate": 0.00043250912876982415,
      "loss": 5.2174,
      "mean_token_accuracy": 0.2120518371462822,
      "num_tokens": 60224762.0,
      "step": 29715
    },
    {
      "entropy": 5.906230878829956,
      "epoch": 2.597221008476798,
      "grad_norm": 1.140625,
      "learning_rate": 0.000432486875023539,
      "loss": 5.2432,
      "mean_token_accuracy": 0.20434350073337554,
      "num_tokens": 60235284.0,
      "step": 29720
    },
    {
      "entropy": 6.018495845794678,
      "epoch": 2.5976579568295026,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000432464618256573,
      "loss": 5.3086,
      "mean_token_accuracy": 0.20440756380558014,
      "num_tokens": 60245939.0,
      "step": 29725
    },
    {
      "entropy": 6.084886503219605,
      "epoch": 2.5980949051822075,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004324423584693531,
      "loss": 5.2946,
      "mean_token_accuracy": 0.19819204956293107,
      "num_tokens": 60256903.0,
      "step": 29730
    },
    {
      "entropy": 6.00913257598877,
      "epoch": 2.5985318535349125,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004324200956623063,
      "loss": 5.2601,
      "mean_token_accuracy": 0.20439915955066681,
      "num_tokens": 60267046.0,
      "step": 29735
    },
    {
      "entropy": 5.981009578704834,
      "epoch": 2.598968801887617,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043239782983585964,
      "loss": 5.3116,
      "mean_token_accuracy": 0.2030423417687416,
      "num_tokens": 60276858.0,
      "step": 29740
    },
    {
      "entropy": 6.0304279804229735,
      "epoch": 2.5994057502403214,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043237556099044,
      "loss": 5.2919,
      "mean_token_accuracy": 0.20099757015705108,
      "num_tokens": 60285874.0,
      "step": 29745
    },
    {
      "entropy": 6.004197883605957,
      "epoch": 2.5998426985930263,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004323532891264748,
      "loss": 5.325,
      "mean_token_accuracy": 0.206124883890152,
      "num_tokens": 60295580.0,
      "step": 29750
    },
    {
      "entropy": 6.007757186889648,
      "epoch": 2.600279646945731,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004323310142443909,
      "loss": 5.2897,
      "mean_token_accuracy": 0.2064823940396309,
      "num_tokens": 60305425.0,
      "step": 29755
    },
    {
      "entropy": 5.994593000411987,
      "epoch": 2.6007165952984357,
      "grad_norm": 1.28125,
      "learning_rate": 0.00043230873634461577,
      "loss": 5.2529,
      "mean_token_accuracy": 0.2015625923871994,
      "num_tokens": 60315578.0,
      "step": 29760
    },
    {
      "entropy": 5.9934381484985355,
      "epoch": 2.6011535436511406,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004322864554275767,
      "loss": 5.2749,
      "mean_token_accuracy": 0.2081529140472412,
      "num_tokens": 60325345.0,
      "step": 29765
    },
    {
      "entropy": 6.0086511135101315,
      "epoch": 2.601590492003845,
      "grad_norm": 1.25,
      "learning_rate": 0.00043226417149370093,
      "loss": 5.2482,
      "mean_token_accuracy": 0.2038104295730591,
      "num_tokens": 60334608.0,
      "step": 29770
    },
    {
      "entropy": 5.959338521957397,
      "epoch": 2.60202744035655,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043224188454341593,
      "loss": 5.2424,
      "mean_token_accuracy": 0.20804283767938614,
      "num_tokens": 60346909.0,
      "step": 29775
    },
    {
      "entropy": 5.930663585662842,
      "epoch": 2.6024643887092545,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004322195945771493,
      "loss": 5.238,
      "mean_token_accuracy": 0.21625897586345671,
      "num_tokens": 60355829.0,
      "step": 29780
    },
    {
      "entropy": 5.956204795837403,
      "epoch": 2.6029013370619594,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004321973015953284,
      "loss": 5.2455,
      "mean_token_accuracy": 0.20173141062259675,
      "num_tokens": 60364954.0,
      "step": 29785
    },
    {
      "entropy": 6.039556550979614,
      "epoch": 2.603338285414664,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043217500559838084,
      "loss": 5.3837,
      "mean_token_accuracy": 0.1932249680161476,
      "num_tokens": 60375575.0,
      "step": 29790
    },
    {
      "entropy": 6.041841793060303,
      "epoch": 2.6037752337673687,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043215270658673437,
      "loss": 5.313,
      "mean_token_accuracy": 0.20047989934682847,
      "num_tokens": 60385534.0,
      "step": 29795
    },
    {
      "entropy": 5.970596837997436,
      "epoch": 2.604212182120073,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004321304045608166,
      "loss": 5.2958,
      "mean_token_accuracy": 0.20092400461435317,
      "num_tokens": 60395517.0,
      "step": 29800
    },
    {
      "entropy": 6.0399878978729244,
      "epoch": 2.604649130472778,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004321080995210554,
      "loss": 5.3705,
      "mean_token_accuracy": 0.20214051008224487,
      "num_tokens": 60404781.0,
      "step": 29805
    },
    {
      "entropy": 6.005969715118408,
      "epoch": 2.605086078825483,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004320857914678785,
      "loss": 5.3133,
      "mean_token_accuracy": 0.2059020295739174,
      "num_tokens": 60415323.0,
      "step": 29810
    },
    {
      "entropy": 6.083399963378906,
      "epoch": 2.6055230271781875,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004320634804017138,
      "loss": 5.4161,
      "mean_token_accuracy": 0.1922672435641289,
      "num_tokens": 60425328.0,
      "step": 29815
    },
    {
      "entropy": 5.9703210353851315,
      "epoch": 2.605959975530892,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004320411663229892,
      "loss": 5.2792,
      "mean_token_accuracy": 0.20410494953393937,
      "num_tokens": 60435391.0,
      "step": 29820
    },
    {
      "entropy": 6.0339967727661135,
      "epoch": 2.606396923883597,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004320188492321328,
      "loss": 5.3248,
      "mean_token_accuracy": 0.19691327959299088,
      "num_tokens": 60445155.0,
      "step": 29825
    },
    {
      "entropy": 6.026936388015747,
      "epoch": 2.606833872236302,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004319965291295726,
      "loss": 5.2555,
      "mean_token_accuracy": 0.20194342881441116,
      "num_tokens": 60455655.0,
      "step": 29830
    },
    {
      "entropy": 5.965675783157349,
      "epoch": 2.6072708205890063,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043197420601573665,
      "loss": 5.1628,
      "mean_token_accuracy": 0.2117088958621025,
      "num_tokens": 60465080.0,
      "step": 29835
    },
    {
      "entropy": 6.00243067741394,
      "epoch": 2.607707768941711,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043195187989105327,
      "loss": 5.2746,
      "mean_token_accuracy": 0.20894624888896943,
      "num_tokens": 60474970.0,
      "step": 29840
    },
    {
      "entropy": 5.9962393283844,
      "epoch": 2.6081447172944157,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043192955075595045,
      "loss": 5.2292,
      "mean_token_accuracy": 0.20819391906261445,
      "num_tokens": 60486599.0,
      "step": 29845
    },
    {
      "entropy": 6.015368556976318,
      "epoch": 2.6085816656471206,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004319072186108567,
      "loss": 5.328,
      "mean_token_accuracy": 0.19772133678197862,
      "num_tokens": 60496849.0,
      "step": 29850
    },
    {
      "entropy": 6.039522790908814,
      "epoch": 2.609018613999825,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004318848834562003,
      "loss": 5.3005,
      "mean_token_accuracy": 0.19326543658971787,
      "num_tokens": 60506708.0,
      "step": 29855
    },
    {
      "entropy": 5.869907474517822,
      "epoch": 2.60945556235253,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043186254529240957,
      "loss": 5.1803,
      "mean_token_accuracy": 0.22208037674427034,
      "num_tokens": 60516538.0,
      "step": 29860
    },
    {
      "entropy": 5.990803289413452,
      "epoch": 2.609892510705235,
      "grad_norm": 1.25,
      "learning_rate": 0.00043184020411991315,
      "loss": 5.2208,
      "mean_token_accuracy": 0.21288235485553741,
      "num_tokens": 60526206.0,
      "step": 29865
    },
    {
      "entropy": 6.030023288726807,
      "epoch": 2.6103294590579393,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004318178599391393,
      "loss": 5.3929,
      "mean_token_accuracy": 0.19236666560173035,
      "num_tokens": 60536878.0,
      "step": 29870
    },
    {
      "entropy": 6.036779737472534,
      "epoch": 2.610766407410644,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043179551275051675,
      "loss": 5.4409,
      "mean_token_accuracy": 0.19536875039339066,
      "num_tokens": 60547709.0,
      "step": 29875
    },
    {
      "entropy": 6.037413883209228,
      "epoch": 2.6112033557633487,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004317731625544742,
      "loss": 5.2047,
      "mean_token_accuracy": 0.2088867247104645,
      "num_tokens": 60557257.0,
      "step": 29880
    },
    {
      "entropy": 6.034703207015991,
      "epoch": 2.6116403041160536,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043175080935144015,
      "loss": 5.3047,
      "mean_token_accuracy": 0.21289920806884766,
      "num_tokens": 60568663.0,
      "step": 29885
    },
    {
      "entropy": 5.978660249710083,
      "epoch": 2.612077252468758,
      "grad_norm": 1.125,
      "learning_rate": 0.0004317284531418434,
      "loss": 5.247,
      "mean_token_accuracy": 0.20548469126224517,
      "num_tokens": 60579168.0,
      "step": 29890
    },
    {
      "entropy": 6.044912052154541,
      "epoch": 2.612514200821463,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043170609392611294,
      "loss": 5.3279,
      "mean_token_accuracy": 0.1951688975095749,
      "num_tokens": 60589893.0,
      "step": 29895
    },
    {
      "entropy": 5.952316045761108,
      "epoch": 2.6129511491741675,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043168373170467735,
      "loss": 5.2391,
      "mean_token_accuracy": 0.20490960031747818,
      "num_tokens": 60599406.0,
      "step": 29900
    },
    {
      "entropy": 5.926806259155273,
      "epoch": 2.6133880975268724,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043166136647796577,
      "loss": 5.3201,
      "mean_token_accuracy": 0.21073544025421143,
      "num_tokens": 60610435.0,
      "step": 29905
    },
    {
      "entropy": 6.039477109909058,
      "epoch": 2.613825045879577,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000431638998246407,
      "loss": 5.2694,
      "mean_token_accuracy": 0.20554879307746887,
      "num_tokens": 60621405.0,
      "step": 29910
    },
    {
      "entropy": 6.005379295349121,
      "epoch": 2.614261994232282,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004316166270104303,
      "loss": 5.2542,
      "mean_token_accuracy": 0.20394774973392488,
      "num_tokens": 60632119.0,
      "step": 29915
    },
    {
      "entropy": 6.015120315551758,
      "epoch": 2.6146989425849867,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043159425277046454,
      "loss": 5.2384,
      "mean_token_accuracy": 0.206350177526474,
      "num_tokens": 60642163.0,
      "step": 29920
    },
    {
      "entropy": 5.95134654045105,
      "epoch": 2.615135890937691,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000431571875526939,
      "loss": 5.2572,
      "mean_token_accuracy": 0.2055269181728363,
      "num_tokens": 60652050.0,
      "step": 29925
    },
    {
      "entropy": 5.954430103302002,
      "epoch": 2.6155728392903956,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004315494952802828,
      "loss": 5.2572,
      "mean_token_accuracy": 0.20358797907829285,
      "num_tokens": 60661843.0,
      "step": 29930
    },
    {
      "entropy": 5.949921321868897,
      "epoch": 2.6160097876431005,
      "grad_norm": 1.296875,
      "learning_rate": 0.00043152711203092524,
      "loss": 5.1941,
      "mean_token_accuracy": 0.2121392771601677,
      "num_tokens": 60671618.0,
      "step": 29935
    },
    {
      "entropy": 6.017131996154785,
      "epoch": 2.6164467359958055,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004315047257792957,
      "loss": 5.376,
      "mean_token_accuracy": 0.2001223161816597,
      "num_tokens": 60682930.0,
      "step": 29940
    },
    {
      "entropy": 6.032282447814941,
      "epoch": 2.61688368434851,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004314823365258235,
      "loss": 5.2925,
      "mean_token_accuracy": 0.20411135107278824,
      "num_tokens": 60692864.0,
      "step": 29945
    },
    {
      "entropy": 6.064271068572998,
      "epoch": 2.617320632701215,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043145994427093806,
      "loss": 5.2964,
      "mean_token_accuracy": 0.20129460394382476,
      "num_tokens": 60702934.0,
      "step": 29950
    },
    {
      "entropy": 6.061111545562744,
      "epoch": 2.6177575810539193,
      "grad_norm": 1.296875,
      "learning_rate": 0.00043143754901506893,
      "loss": 5.248,
      "mean_token_accuracy": 0.20200653821229936,
      "num_tokens": 60712289.0,
      "step": 29955
    },
    {
      "entropy": 5.905441188812256,
      "epoch": 2.6181945294066242,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043141515075864563,
      "loss": 5.2157,
      "mean_token_accuracy": 0.21299105733633042,
      "num_tokens": 60723046.0,
      "step": 29960
    },
    {
      "entropy": 6.026308012008667,
      "epoch": 2.6186314777593287,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00043139274950209776,
      "loss": 5.3067,
      "mean_token_accuracy": 0.1984963595867157,
      "num_tokens": 60731969.0,
      "step": 29965
    },
    {
      "entropy": 6.077915334701538,
      "epoch": 2.6190684261120336,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043137034524585506,
      "loss": 5.365,
      "mean_token_accuracy": 0.20544098913669587,
      "num_tokens": 60743021.0,
      "step": 29970
    },
    {
      "entropy": 6.021057605743408,
      "epoch": 2.6195053744647385,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004313479379903472,
      "loss": 5.3148,
      "mean_token_accuracy": 0.20147286206483841,
      "num_tokens": 60751854.0,
      "step": 29975
    },
    {
      "entropy": 6.061172342300415,
      "epoch": 2.619942322817443,
      "grad_norm": 1.25,
      "learning_rate": 0.0004313255277360039,
      "loss": 5.2762,
      "mean_token_accuracy": 0.19890448004007338,
      "num_tokens": 60761572.0,
      "step": 29980
    },
    {
      "entropy": 5.969974422454834,
      "epoch": 2.6203792711701475,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004313031144832551,
      "loss": 5.23,
      "mean_token_accuracy": 0.2057502821087837,
      "num_tokens": 60771618.0,
      "step": 29985
    },
    {
      "entropy": 5.99564299583435,
      "epoch": 2.6208162195228524,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043128069823253064,
      "loss": 5.2332,
      "mean_token_accuracy": 0.20936697125434875,
      "num_tokens": 60781038.0,
      "step": 29990
    },
    {
      "entropy": 5.987120485305786,
      "epoch": 2.6212531678755573,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043125827898426056,
      "loss": 5.2889,
      "mean_token_accuracy": 0.20333926230669022,
      "num_tokens": 60792433.0,
      "step": 29995
    },
    {
      "entropy": 6.01448426246643,
      "epoch": 2.6216901162282618,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043123585673887475,
      "loss": 5.2483,
      "mean_token_accuracy": 0.2015527918934822,
      "num_tokens": 60801995.0,
      "step": 30000
    },
    {
      "epoch": 2.6216901162282618,
      "eval_entropy": 5.73899938551633,
      "eval_loss": 5.314946174621582,
      "eval_mean_token_accuracy": 0.2109865663564692,
      "eval_num_tokens": 60801995.0,
      "eval_runtime": 21.6906,
      "eval_samples_per_second": 1706.455,
      "eval_steps_per_second": 213.318,
      "step": 30000
    },
    {
      "entropy": 6.015504360198975,
      "epoch": 2.6221270645809667,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004312134314968034,
      "loss": 5.2982,
      "mean_token_accuracy": 0.20089961290359498,
      "num_tokens": 60811784.0,
      "step": 30005
    },
    {
      "entropy": 6.003088426589966,
      "epoch": 2.622564012933671,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043119100325847654,
      "loss": 5.2365,
      "mean_token_accuracy": 0.21292226910591125,
      "num_tokens": 60822018.0,
      "step": 30010
    },
    {
      "entropy": 5.95697512626648,
      "epoch": 2.623000961286376,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004311685720243243,
      "loss": 5.2841,
      "mean_token_accuracy": 0.2014918729662895,
      "num_tokens": 60832204.0,
      "step": 30015
    },
    {
      "entropy": 5.939129209518432,
      "epoch": 2.6234379096390805,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004311461377947772,
      "loss": 5.2663,
      "mean_token_accuracy": 0.20604983419179917,
      "num_tokens": 60843293.0,
      "step": 30020
    },
    {
      "entropy": 6.036695671081543,
      "epoch": 2.6238748579917854,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004311237005702652,
      "loss": 5.2623,
      "mean_token_accuracy": 0.19954549223184587,
      "num_tokens": 60852323.0,
      "step": 30025
    },
    {
      "entropy": 5.973034763336182,
      "epoch": 2.6243118063444903,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004311012603512189,
      "loss": 5.2583,
      "mean_token_accuracy": 0.2050105780363083,
      "num_tokens": 60862788.0,
      "step": 30030
    },
    {
      "entropy": 5.992052459716797,
      "epoch": 2.624748754697195,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043107881713806863,
      "loss": 5.3022,
      "mean_token_accuracy": 0.202061727643013,
      "num_tokens": 60872793.0,
      "step": 30035
    },
    {
      "entropy": 5.967383909225464,
      "epoch": 2.6251857030498993,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043105637093124484,
      "loss": 5.2413,
      "mean_token_accuracy": 0.20355900675058364,
      "num_tokens": 60882712.0,
      "step": 30040
    },
    {
      "entropy": 5.98784327507019,
      "epoch": 2.625622651402604,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043103392173117814,
      "loss": 5.3065,
      "mean_token_accuracy": 0.20304392725229264,
      "num_tokens": 60892400.0,
      "step": 30045
    },
    {
      "entropy": 5.989512825012207,
      "epoch": 2.626059599755309,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043101146953829895,
      "loss": 5.2211,
      "mean_token_accuracy": 0.20851853638887405,
      "num_tokens": 60902480.0,
      "step": 30050
    },
    {
      "entropy": 5.999860143661499,
      "epoch": 2.6264965481080136,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004309890143530382,
      "loss": 5.3012,
      "mean_token_accuracy": 0.20415108799934387,
      "num_tokens": 60912183.0,
      "step": 30055
    },
    {
      "entropy": 6.030772495269775,
      "epoch": 2.6269334964607185,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004309665561758262,
      "loss": 5.2877,
      "mean_token_accuracy": 0.20136035978794098,
      "num_tokens": 60922166.0,
      "step": 30060
    },
    {
      "entropy": 5.96727442741394,
      "epoch": 2.627370444813423,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043094409500709404,
      "loss": 5.2323,
      "mean_token_accuracy": 0.20491972416639329,
      "num_tokens": 60931980.0,
      "step": 30065
    },
    {
      "entropy": 5.944014310836792,
      "epoch": 2.627807393166128,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004309216308472725,
      "loss": 5.4208,
      "mean_token_accuracy": 0.1996493235230446,
      "num_tokens": 60943382.0,
      "step": 30070
    },
    {
      "entropy": 6.021462106704712,
      "epoch": 2.6282443415188323,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004308991636967923,
      "loss": 5.3011,
      "mean_token_accuracy": 0.19853492677211762,
      "num_tokens": 60953505.0,
      "step": 30075
    },
    {
      "entropy": 6.070501327514648,
      "epoch": 2.6286812898715373,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004308766935560844,
      "loss": 5.2976,
      "mean_token_accuracy": 0.2032916083931923,
      "num_tokens": 60963851.0,
      "step": 30080
    },
    {
      "entropy": 6.02342586517334,
      "epoch": 2.6291182382242417,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043085422042557985,
      "loss": 5.3057,
      "mean_token_accuracy": 0.20358271896839142,
      "num_tokens": 60974919.0,
      "step": 30085
    },
    {
      "entropy": 6.026953935623169,
      "epoch": 2.6295551865769466,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043083174430570974,
      "loss": 5.3318,
      "mean_token_accuracy": 0.19521324187517167,
      "num_tokens": 60983866.0,
      "step": 30090
    },
    {
      "entropy": 5.951289939880371,
      "epoch": 2.629992134929651,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043080926519690517,
      "loss": 5.2435,
      "mean_token_accuracy": 0.20585519522428514,
      "num_tokens": 60994393.0,
      "step": 30095
    },
    {
      "entropy": 6.020531702041626,
      "epoch": 2.630429083282356,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043078678309959716,
      "loss": 5.3141,
      "mean_token_accuracy": 0.2054426684975624,
      "num_tokens": 61004634.0,
      "step": 30100
    },
    {
      "entropy": 5.9858856201171875,
      "epoch": 2.630866031635061,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004307642980142171,
      "loss": 5.2704,
      "mean_token_accuracy": 0.20556087791919708,
      "num_tokens": 61014771.0,
      "step": 30105
    },
    {
      "entropy": 5.98916802406311,
      "epoch": 2.6313029799877654,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043074180994119604,
      "loss": 5.3043,
      "mean_token_accuracy": 0.20414473563432695,
      "num_tokens": 61025213.0,
      "step": 30110
    },
    {
      "entropy": 6.002563238143921,
      "epoch": 2.63173992834047,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004307193188809656,
      "loss": 5.1987,
      "mean_token_accuracy": 0.21310463100671767,
      "num_tokens": 61035601.0,
      "step": 30115
    },
    {
      "entropy": 5.958721590042114,
      "epoch": 2.632176876693175,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043069682483395694,
      "loss": 5.2588,
      "mean_token_accuracy": 0.20298811793327332,
      "num_tokens": 61045862.0,
      "step": 30120
    },
    {
      "entropy": 5.988091230392456,
      "epoch": 2.6326138250458797,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043067432780060163,
      "loss": 5.2573,
      "mean_token_accuracy": 0.21078300774097442,
      "num_tokens": 61056702.0,
      "step": 30125
    },
    {
      "entropy": 6.011309003829956,
      "epoch": 2.633050773398584,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043065182778133116,
      "loss": 5.3442,
      "mean_token_accuracy": 0.19755907058715821,
      "num_tokens": 61066353.0,
      "step": 30130
    },
    {
      "entropy": 6.000275564193726,
      "epoch": 2.633487721751289,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000430629324776577,
      "loss": 5.2656,
      "mean_token_accuracy": 0.2016359329223633,
      "num_tokens": 61076534.0,
      "step": 30135
    },
    {
      "entropy": 6.001124429702759,
      "epoch": 2.6339246701039936,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043060681878677096,
      "loss": 5.1738,
      "mean_token_accuracy": 0.2119273290038109,
      "num_tokens": 61086635.0,
      "step": 30140
    },
    {
      "entropy": 5.994639778137207,
      "epoch": 2.6343616184566985,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043058430981234453,
      "loss": 5.2965,
      "mean_token_accuracy": 0.19716933071613313,
      "num_tokens": 61096605.0,
      "step": 30145
    },
    {
      "entropy": 5.946798849105835,
      "epoch": 2.634798566809403,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004305617978537295,
      "loss": 5.2517,
      "mean_token_accuracy": 0.20908751487731933,
      "num_tokens": 61106465.0,
      "step": 30150
    },
    {
      "entropy": 6.012409639358521,
      "epoch": 2.635235515162108,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004305392829113577,
      "loss": 5.3492,
      "mean_token_accuracy": 0.20288234055042267,
      "num_tokens": 61116063.0,
      "step": 30155
    },
    {
      "entropy": 6.082350158691407,
      "epoch": 2.6356724635148128,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043051676498566095,
      "loss": 5.3169,
      "mean_token_accuracy": 0.19425660073757173,
      "num_tokens": 61125794.0,
      "step": 30160
    },
    {
      "entropy": 5.9913678646087645,
      "epoch": 2.6361094118675172,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004304942440770712,
      "loss": 5.3017,
      "mean_token_accuracy": 0.2000654086470604,
      "num_tokens": 61135734.0,
      "step": 30165
    },
    {
      "entropy": 5.935859155654907,
      "epoch": 2.6365463602202217,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004304717201860203,
      "loss": 5.2624,
      "mean_token_accuracy": 0.20574834793806077,
      "num_tokens": 61146071.0,
      "step": 30170
    },
    {
      "entropy": 5.999017906188965,
      "epoch": 2.6369833085729266,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004304491933129404,
      "loss": 5.2463,
      "mean_token_accuracy": 0.20599700063467025,
      "num_tokens": 61154931.0,
      "step": 30175
    },
    {
      "entropy": 5.998206996917725,
      "epoch": 2.6374202569256315,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004304266634582634,
      "loss": 5.2901,
      "mean_token_accuracy": 0.208437380194664,
      "num_tokens": 61164464.0,
      "step": 30180
    },
    {
      "entropy": 6.011873388290406,
      "epoch": 2.637857205278336,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004304041306224217,
      "loss": 5.2993,
      "mean_token_accuracy": 0.20028140842914582,
      "num_tokens": 61175379.0,
      "step": 30185
    },
    {
      "entropy": 5.989103317260742,
      "epoch": 2.638294153631041,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043038159480584723,
      "loss": 5.2166,
      "mean_token_accuracy": 0.21137642711400986,
      "num_tokens": 61186340.0,
      "step": 30190
    },
    {
      "entropy": 5.980472660064697,
      "epoch": 2.6387311019837454,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043035905600897237,
      "loss": 5.231,
      "mean_token_accuracy": 0.2086677685379982,
      "num_tokens": 61196495.0,
      "step": 30195
    },
    {
      "entropy": 6.019885015487671,
      "epoch": 2.6391680503364503,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043033651423222947,
      "loss": 5.3059,
      "mean_token_accuracy": 0.2038327008485794,
      "num_tokens": 61206555.0,
      "step": 30200
    },
    {
      "entropy": 6.05070948600769,
      "epoch": 2.6396049986891548,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004303139694760508,
      "loss": 5.3709,
      "mean_token_accuracy": 0.1972925469279289,
      "num_tokens": 61215818.0,
      "step": 30205
    },
    {
      "entropy": 6.015393781661987,
      "epoch": 2.6400419470418597,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043029142174086875,
      "loss": 5.261,
      "mean_token_accuracy": 0.20671987533569336,
      "num_tokens": 61226325.0,
      "step": 30210
    },
    {
      "entropy": 6.04637336730957,
      "epoch": 2.6404788953945646,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004302688710271159,
      "loss": 5.2862,
      "mean_token_accuracy": 0.20240240842103957,
      "num_tokens": 61236179.0,
      "step": 30215
    },
    {
      "entropy": 5.968053102493286,
      "epoch": 2.640915843747269,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00043024631733522484,
      "loss": 5.2227,
      "mean_token_accuracy": 0.2100789189338684,
      "num_tokens": 61245680.0,
      "step": 30220
    },
    {
      "entropy": 5.99322543144226,
      "epoch": 2.6413527920999735,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000430223760665628,
      "loss": 5.3428,
      "mean_token_accuracy": 0.19779606461524962,
      "num_tokens": 61256369.0,
      "step": 30225
    },
    {
      "entropy": 5.9506477355957035,
      "epoch": 2.6417897404526784,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043020120101875807,
      "loss": 5.2014,
      "mean_token_accuracy": 0.20880543291568757,
      "num_tokens": 61266395.0,
      "step": 30230
    },
    {
      "entropy": 6.027905225753784,
      "epoch": 2.6422266888053834,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043017863839504784,
      "loss": 5.3229,
      "mean_token_accuracy": 0.20610193759202958,
      "num_tokens": 61276442.0,
      "step": 30235
    },
    {
      "entropy": 6.054361200332641,
      "epoch": 2.642663637158088,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043015607279493,
      "loss": 5.3696,
      "mean_token_accuracy": 0.1951817512512207,
      "num_tokens": 61286230.0,
      "step": 30240
    },
    {
      "entropy": 5.965408658981323,
      "epoch": 2.6431005855107927,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004301335042188373,
      "loss": 5.2269,
      "mean_token_accuracy": 0.21073378771543502,
      "num_tokens": 61297853.0,
      "step": 30245
    },
    {
      "entropy": 5.966507768630981,
      "epoch": 2.643537533863497,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043011093266720273,
      "loss": 5.1977,
      "mean_token_accuracy": 0.20891975313425065,
      "num_tokens": 61307976.0,
      "step": 30250
    },
    {
      "entropy": 6.031795501708984,
      "epoch": 2.643974482216202,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043008835814045926,
      "loss": 5.3034,
      "mean_token_accuracy": 0.197227318584919,
      "num_tokens": 61319101.0,
      "step": 30255
    },
    {
      "entropy": 6.019554471969604,
      "epoch": 2.6444114305689066,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004300657806390398,
      "loss": 5.2876,
      "mean_token_accuracy": 0.20317939817905425,
      "num_tokens": 61329492.0,
      "step": 30260
    },
    {
      "entropy": 6.016564273834229,
      "epoch": 2.6448483789216115,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043004320016337737,
      "loss": 5.3239,
      "mean_token_accuracy": 0.20554356276988983,
      "num_tokens": 61340480.0,
      "step": 30265
    },
    {
      "entropy": 6.02332181930542,
      "epoch": 2.6452853272743164,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043002061671390513,
      "loss": 5.2407,
      "mean_token_accuracy": 0.19897794276475905,
      "num_tokens": 61350572.0,
      "step": 30270
    },
    {
      "entropy": 6.024818849563599,
      "epoch": 2.645722275627021,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004299980302910563,
      "loss": 5.2814,
      "mean_token_accuracy": 0.20571965724229813,
      "num_tokens": 61361329.0,
      "step": 30275
    },
    {
      "entropy": 6.036417150497437,
      "epoch": 2.6461592239797254,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004299754408952639,
      "loss": 5.3422,
      "mean_token_accuracy": 0.19764477014541626,
      "num_tokens": 61372618.0,
      "step": 30280
    },
    {
      "entropy": 6.098924398422241,
      "epoch": 2.6465961723324303,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004299528485269614,
      "loss": 5.3732,
      "mean_token_accuracy": 0.193365441262722,
      "num_tokens": 61383831.0,
      "step": 30285
    },
    {
      "entropy": 6.00874605178833,
      "epoch": 2.647033120685135,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004299302531865821,
      "loss": 5.2672,
      "mean_token_accuracy": 0.20545354336500168,
      "num_tokens": 61394527.0,
      "step": 30290
    },
    {
      "entropy": 6.02073392868042,
      "epoch": 2.6474700690378397,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042990765487455935,
      "loss": 5.3714,
      "mean_token_accuracy": 0.1972415715456009,
      "num_tokens": 61404787.0,
      "step": 30295
    },
    {
      "entropy": 5.944258356094361,
      "epoch": 2.6479070173905446,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004298850535913266,
      "loss": 5.1747,
      "mean_token_accuracy": 0.21832630634307862,
      "num_tokens": 61413922.0,
      "step": 30300
    },
    {
      "entropy": 6.01386399269104,
      "epoch": 2.648343965743249,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004298624493373174,
      "loss": 5.282,
      "mean_token_accuracy": 0.20268866568803787,
      "num_tokens": 61424750.0,
      "step": 30305
    },
    {
      "entropy": 5.9203613758087155,
      "epoch": 2.648780914095954,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004298398421129653,
      "loss": 5.2392,
      "mean_token_accuracy": 0.20966076701879502,
      "num_tokens": 61435222.0,
      "step": 30310
    },
    {
      "entropy": 5.982348442077637,
      "epoch": 2.6492178624486584,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004298172319187038,
      "loss": 5.2881,
      "mean_token_accuracy": 0.20707590281963348,
      "num_tokens": 61444230.0,
      "step": 30315
    },
    {
      "entropy": 6.003763103485108,
      "epoch": 2.6496548108013633,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004297946187549667,
      "loss": 5.2874,
      "mean_token_accuracy": 0.19809106439352037,
      "num_tokens": 61454975.0,
      "step": 30320
    },
    {
      "entropy": 6.0037836074829105,
      "epoch": 2.6500917591540682,
      "grad_norm": 1.328125,
      "learning_rate": 0.00042977200262218776,
      "loss": 5.3713,
      "mean_token_accuracy": 0.19878738522529601,
      "num_tokens": 61465563.0,
      "step": 30325
    },
    {
      "entropy": 6.021136522293091,
      "epoch": 2.6505287075067727,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004297493835208006,
      "loss": 5.2982,
      "mean_token_accuracy": 0.2032650500535965,
      "num_tokens": 61475330.0,
      "step": 30330
    },
    {
      "entropy": 6.0402415752410885,
      "epoch": 2.650965655859477,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042972676145123926,
      "loss": 5.3118,
      "mean_token_accuracy": 0.20060920417308808,
      "num_tokens": 61485577.0,
      "step": 30335
    },
    {
      "entropy": 6.018212890625,
      "epoch": 2.651402604212182,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042970413641393754,
      "loss": 5.2676,
      "mean_token_accuracy": 0.2010944291949272,
      "num_tokens": 61495502.0,
      "step": 30340
    },
    {
      "entropy": 5.9793455600738525,
      "epoch": 2.651839552564887,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004296815084093294,
      "loss": 5.3092,
      "mean_token_accuracy": 0.2000475585460663,
      "num_tokens": 61505383.0,
      "step": 30345
    },
    {
      "entropy": 6.001484203338623,
      "epoch": 2.6522765009175915,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042965887743784886,
      "loss": 5.2896,
      "mean_token_accuracy": 0.2044094666838646,
      "num_tokens": 61515052.0,
      "step": 30350
    },
    {
      "entropy": 6.026920032501221,
      "epoch": 2.6527134492702964,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004296362434999301,
      "loss": 5.2295,
      "mean_token_accuracy": 0.2101409763097763,
      "num_tokens": 61525139.0,
      "step": 30355
    },
    {
      "entropy": 5.958214664459229,
      "epoch": 2.653150397623001,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042961360659600706,
      "loss": 5.1796,
      "mean_token_accuracy": 0.21144817471504213,
      "num_tokens": 61534901.0,
      "step": 30360
    },
    {
      "entropy": 5.982995414733887,
      "epoch": 2.653587345975706,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004295909667265141,
      "loss": 5.2848,
      "mean_token_accuracy": 0.1976694405078888,
      "num_tokens": 61543258.0,
      "step": 30365
    },
    {
      "entropy": 5.980223083496094,
      "epoch": 2.6540242943284102,
      "grad_norm": 1.328125,
      "learning_rate": 0.00042956832389188534,
      "loss": 5.2641,
      "mean_token_accuracy": 0.20465803295373916,
      "num_tokens": 61553294.0,
      "step": 30370
    },
    {
      "entropy": 6.025187826156616,
      "epoch": 2.654461242681115,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004295456780925551,
      "loss": 5.2562,
      "mean_token_accuracy": 0.21040416210889817,
      "num_tokens": 61562986.0,
      "step": 30375
    },
    {
      "entropy": 5.995108652114868,
      "epoch": 2.65489819103382,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042952302932895784,
      "loss": 5.3161,
      "mean_token_accuracy": 0.20604764819145202,
      "num_tokens": 61573994.0,
      "step": 30380
    },
    {
      "entropy": 5.958453226089477,
      "epoch": 2.6553351393865245,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042950037760152793,
      "loss": 5.2605,
      "mean_token_accuracy": 0.20767159014940262,
      "num_tokens": 61584279.0,
      "step": 30385
    },
    {
      "entropy": 6.09016523361206,
      "epoch": 2.655772087739229,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004294777229106998,
      "loss": 5.3263,
      "mean_token_accuracy": 0.20073698163032533,
      "num_tokens": 61595188.0,
      "step": 30390
    },
    {
      "entropy": 5.977011632919312,
      "epoch": 2.656209036091934,
      "grad_norm": 1.171875,
      "learning_rate": 0.000429455065256908,
      "loss": 5.3673,
      "mean_token_accuracy": 0.2008710876107216,
      "num_tokens": 61605321.0,
      "step": 30395
    },
    {
      "entropy": 6.062139081954956,
      "epoch": 2.656645984444639,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004294324046405871,
      "loss": 5.305,
      "mean_token_accuracy": 0.19694778621196746,
      "num_tokens": 61615137.0,
      "step": 30400
    },
    {
      "entropy": 5.997072696685791,
      "epoch": 2.6570829327973433,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042940974106217176,
      "loss": 5.1866,
      "mean_token_accuracy": 0.20998602360486984,
      "num_tokens": 61625912.0,
      "step": 30405
    },
    {
      "entropy": 5.990564489364624,
      "epoch": 2.6575198811500482,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004293870745220967,
      "loss": 5.2302,
      "mean_token_accuracy": 0.2056242361664772,
      "num_tokens": 61635836.0,
      "step": 30410
    },
    {
      "entropy": 6.002114772796631,
      "epoch": 2.6579568295027527,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004293644050207966,
      "loss": 5.3045,
      "mean_token_accuracy": 0.20118673890829086,
      "num_tokens": 61645151.0,
      "step": 30415
    },
    {
      "entropy": 6.059758281707763,
      "epoch": 2.6583937778554576,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004293417325587064,
      "loss": 5.3359,
      "mean_token_accuracy": 0.19895196110010147,
      "num_tokens": 61654266.0,
      "step": 30420
    },
    {
      "entropy": 6.07839822769165,
      "epoch": 2.658830726208162,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004293190571362608,
      "loss": 5.42,
      "mean_token_accuracy": 0.19072960466146469,
      "num_tokens": 61664355.0,
      "step": 30425
    },
    {
      "entropy": 6.0852165699005125,
      "epoch": 2.659267674560867,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00042929637875389486,
      "loss": 5.3412,
      "mean_token_accuracy": 0.20361415147781373,
      "num_tokens": 61674130.0,
      "step": 30430
    },
    {
      "entropy": 6.102247524261474,
      "epoch": 2.6597046229135715,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042927369741204357,
      "loss": 5.3526,
      "mean_token_accuracy": 0.199840147793293,
      "num_tokens": 61683612.0,
      "step": 30435
    },
    {
      "entropy": 6.001351404190063,
      "epoch": 2.6601415712662764,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004292510131111419,
      "loss": 5.3261,
      "mean_token_accuracy": 0.20720780789852142,
      "num_tokens": 61693156.0,
      "step": 30440
    },
    {
      "entropy": 5.956441640853882,
      "epoch": 2.660578519618981,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042922832585162496,
      "loss": 5.2004,
      "mean_token_accuracy": 0.2064582109451294,
      "num_tokens": 61701925.0,
      "step": 30445
    },
    {
      "entropy": 6.020820426940918,
      "epoch": 2.6610154679716858,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042920563563392803,
      "loss": 5.2528,
      "mean_token_accuracy": 0.20707642734050752,
      "num_tokens": 61712388.0,
      "step": 30450
    },
    {
      "entropy": 5.98899736404419,
      "epoch": 2.6614524163243907,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042918294245848605,
      "loss": 5.2455,
      "mean_token_accuracy": 0.20204924792051315,
      "num_tokens": 61722402.0,
      "step": 30455
    },
    {
      "entropy": 6.056678533554077,
      "epoch": 2.661889364677095,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042916024632573454,
      "loss": 5.3487,
      "mean_token_accuracy": 0.19346551448106766,
      "num_tokens": 61733160.0,
      "step": 30460
    },
    {
      "entropy": 6.013279914855957,
      "epoch": 2.6623263130297996,
      "grad_norm": 1.25,
      "learning_rate": 0.00042913754723610865,
      "loss": 5.2652,
      "mean_token_accuracy": 0.2070107489824295,
      "num_tokens": 61742536.0,
      "step": 30465
    },
    {
      "entropy": 5.993255949020385,
      "epoch": 2.6627632613825045,
      "grad_norm": 1.125,
      "learning_rate": 0.0004291148451900439,
      "loss": 5.2961,
      "mean_token_accuracy": 0.2001469388604164,
      "num_tokens": 61753439.0,
      "step": 30470
    },
    {
      "entropy": 5.998438358306885,
      "epoch": 2.6632002097352094,
      "grad_norm": 1.359375,
      "learning_rate": 0.00042909214018797565,
      "loss": 5.2789,
      "mean_token_accuracy": 0.20998039990663528,
      "num_tokens": 61763994.0,
      "step": 30475
    },
    {
      "entropy": 6.027273654937744,
      "epoch": 2.663637158087914,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004290694322303394,
      "loss": 5.3098,
      "mean_token_accuracy": 0.2023107200860977,
      "num_tokens": 61774082.0,
      "step": 30480
    },
    {
      "entropy": 5.977960872650146,
      "epoch": 2.664074106440619,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042904672131757065,
      "loss": 5.2457,
      "mean_token_accuracy": 0.21235260516405105,
      "num_tokens": 61784528.0,
      "step": 30485
    },
    {
      "entropy": 6.016061925888062,
      "epoch": 2.6645110547933233,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004290240074501051,
      "loss": 5.287,
      "mean_token_accuracy": 0.20305137783288957,
      "num_tokens": 61794267.0,
      "step": 30490
    },
    {
      "entropy": 6.0343403816223145,
      "epoch": 2.664948003146028,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004290012906283784,
      "loss": 5.2725,
      "mean_token_accuracy": 0.20783269554376602,
      "num_tokens": 61804842.0,
      "step": 30495
    },
    {
      "entropy": 6.043171548843384,
      "epoch": 2.6653849514987327,
      "grad_norm": 1.3125,
      "learning_rate": 0.00042897857085282627,
      "loss": 5.2993,
      "mean_token_accuracy": 0.2001101404428482,
      "num_tokens": 61814316.0,
      "step": 30500
    },
    {
      "entropy": 6.013114976882934,
      "epoch": 2.6658218998514376,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042895584812388445,
      "loss": 5.2938,
      "mean_token_accuracy": 0.20395727902650834,
      "num_tokens": 61823828.0,
      "step": 30505
    },
    {
      "entropy": 6.027829980850219,
      "epoch": 2.6662588482041425,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004289331224419888,
      "loss": 5.3035,
      "mean_token_accuracy": 0.20455261468887329,
      "num_tokens": 61833878.0,
      "step": 30510
    },
    {
      "entropy": 5.989151573181152,
      "epoch": 2.666695796556847,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042891039380757516,
      "loss": 5.2197,
      "mean_token_accuracy": 0.2070239946246147,
      "num_tokens": 61844591.0,
      "step": 30515
    },
    {
      "entropy": 6.000449895858765,
      "epoch": 2.6671327449095514,
      "grad_norm": 1.125,
      "learning_rate": 0.00042888766222107954,
      "loss": 5.3022,
      "mean_token_accuracy": 0.20176043212413788,
      "num_tokens": 61854791.0,
      "step": 30520
    },
    {
      "entropy": 5.985416984558105,
      "epoch": 2.6675696932622563,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000428864927682938,
      "loss": 5.327,
      "mean_token_accuracy": 0.2028730407357216,
      "num_tokens": 61865194.0,
      "step": 30525
    },
    {
      "entropy": 5.984064245223999,
      "epoch": 2.6680066416149613,
      "grad_norm": 1.34375,
      "learning_rate": 0.00042884219019358644,
      "loss": 5.2377,
      "mean_token_accuracy": 0.20179056525230407,
      "num_tokens": 61874827.0,
      "step": 30530
    },
    {
      "entropy": 6.059687232971191,
      "epoch": 2.6684435899676657,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004288194497534611,
      "loss": 5.3074,
      "mean_token_accuracy": 0.19915589839220046,
      "num_tokens": 61885596.0,
      "step": 30535
    },
    {
      "entropy": 5.974049377441406,
      "epoch": 2.6688805383203706,
      "grad_norm": 1.296875,
      "learning_rate": 0.00042879670636299805,
      "loss": 5.1581,
      "mean_token_accuracy": 0.21156062334775924,
      "num_tokens": 61895732.0,
      "step": 30540
    },
    {
      "entropy": 5.967705392837525,
      "epoch": 2.669317486673075,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042877396002263357,
      "loss": 5.277,
      "mean_token_accuracy": 0.2007494330406189,
      "num_tokens": 61905581.0,
      "step": 30545
    },
    {
      "entropy": 5.971679019927978,
      "epoch": 2.66975443502578,
      "grad_norm": 1.3125,
      "learning_rate": 0.00042875121073280404,
      "loss": 5.2285,
      "mean_token_accuracy": 0.2090811014175415,
      "num_tokens": 61914926.0,
      "step": 30550
    },
    {
      "entropy": 5.918217611312866,
      "epoch": 2.6701913833784845,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042872845849394563,
      "loss": 5.2698,
      "mean_token_accuracy": 0.20736827701330185,
      "num_tokens": 61926501.0,
      "step": 30555
    },
    {
      "entropy": 6.016000127792358,
      "epoch": 2.6706283317311894,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042870570330649495,
      "loss": 5.3375,
      "mean_token_accuracy": 0.19726160019636155,
      "num_tokens": 61937161.0,
      "step": 30560
    },
    {
      "entropy": 5.986431837081909,
      "epoch": 2.6710652800838943,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004286829451708883,
      "loss": 5.2363,
      "mean_token_accuracy": 0.20402432829141617,
      "num_tokens": 61949125.0,
      "step": 30565
    },
    {
      "entropy": 5.955962896347046,
      "epoch": 2.671502228436599,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004286601840875622,
      "loss": 5.1504,
      "mean_token_accuracy": 0.20972481071949006,
      "num_tokens": 61958110.0,
      "step": 30570
    },
    {
      "entropy": 5.962193298339844,
      "epoch": 2.6719391767893033,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004286374200569533,
      "loss": 5.2691,
      "mean_token_accuracy": 0.20553994327783584,
      "num_tokens": 61967805.0,
      "step": 30575
    },
    {
      "entropy": 5.974402809143067,
      "epoch": 2.672376125142008,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042861465307949815,
      "loss": 5.2759,
      "mean_token_accuracy": 0.20807578712701796,
      "num_tokens": 61977189.0,
      "step": 30580
    },
    {
      "entropy": 6.016959571838379,
      "epoch": 2.672813073494713,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042859188315563345,
      "loss": 5.3701,
      "mean_token_accuracy": 0.2034279227256775,
      "num_tokens": 61988568.0,
      "step": 30585
    },
    {
      "entropy": 6.020184946060181,
      "epoch": 2.6732500218474176,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042856911028579597,
      "loss": 5.2143,
      "mean_token_accuracy": 0.20717351734638215,
      "num_tokens": 61999316.0,
      "step": 30590
    },
    {
      "entropy": 6.007657194137574,
      "epoch": 2.6736869702001225,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042854633447042247,
      "loss": 5.3189,
      "mean_token_accuracy": 0.2039225369691849,
      "num_tokens": 62009345.0,
      "step": 30595
    },
    {
      "entropy": 5.939137125015259,
      "epoch": 2.674123918552827,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042852355570994983,
      "loss": 5.2415,
      "mean_token_accuracy": 0.2106005311012268,
      "num_tokens": 62020847.0,
      "step": 30600
    },
    {
      "entropy": 5.966865921020508,
      "epoch": 2.674560866905532,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00042850077400481495,
      "loss": 5.2286,
      "mean_token_accuracy": 0.20964713394641876,
      "num_tokens": 62030506.0,
      "step": 30605
    },
    {
      "entropy": 6.071293210983276,
      "epoch": 2.6749978152582363,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042847798935545484,
      "loss": 5.3692,
      "mean_token_accuracy": 0.19797634929418564,
      "num_tokens": 62040238.0,
      "step": 30610
    },
    {
      "entropy": 5.967975044250489,
      "epoch": 2.6754347636109412,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004284552017623064,
      "loss": 5.2544,
      "mean_token_accuracy": 0.2064968466758728,
      "num_tokens": 62050192.0,
      "step": 30615
    },
    {
      "entropy": 5.975988149642944,
      "epoch": 2.675871711963646,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004284324112258068,
      "loss": 5.2899,
      "mean_token_accuracy": 0.20407289415597915,
      "num_tokens": 62059475.0,
      "step": 30620
    },
    {
      "entropy": 6.003735971450806,
      "epoch": 2.6763086603163506,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042840961774639314,
      "loss": 5.2644,
      "mean_token_accuracy": 0.20079730451107025,
      "num_tokens": 62068799.0,
      "step": 30625
    },
    {
      "entropy": 5.988340854644775,
      "epoch": 2.676745608669055,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004283868213245026,
      "loss": 5.3348,
      "mean_token_accuracy": 0.19926852136850357,
      "num_tokens": 62078205.0,
      "step": 30630
    },
    {
      "entropy": 5.946833753585816,
      "epoch": 2.67718255702176,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042836402196057255,
      "loss": 5.2068,
      "mean_token_accuracy": 0.20728017687797545,
      "num_tokens": 62088084.0,
      "step": 30635
    },
    {
      "entropy": 5.9832985401153564,
      "epoch": 2.677619505374465,
      "grad_norm": 1.25,
      "learning_rate": 0.00042834121965504014,
      "loss": 5.2385,
      "mean_token_accuracy": 0.20814571529626846,
      "num_tokens": 62098709.0,
      "step": 30640
    },
    {
      "entropy": 5.9156567573547365,
      "epoch": 2.6780564537271694,
      "grad_norm": 1.25,
      "learning_rate": 0.0004283184144083428,
      "loss": 5.2011,
      "mean_token_accuracy": 0.208390673995018,
      "num_tokens": 62108949.0,
      "step": 30645
    },
    {
      "entropy": 5.97188024520874,
      "epoch": 2.6784934020798743,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042829560622091795,
      "loss": 5.2591,
      "mean_token_accuracy": 0.209261517226696,
      "num_tokens": 62118922.0,
      "step": 30650
    },
    {
      "entropy": 5.977369499206543,
      "epoch": 2.6789303504325788,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042827279509320297,
      "loss": 5.2573,
      "mean_token_accuracy": 0.20499834269285203,
      "num_tokens": 62128396.0,
      "step": 30655
    },
    {
      "entropy": 6.053115892410278,
      "epoch": 2.6793672987852837,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004282499810256356,
      "loss": 5.3454,
      "mean_token_accuracy": 0.19225690364837647,
      "num_tokens": 62139223.0,
      "step": 30660
    },
    {
      "entropy": 6.000826597213745,
      "epoch": 2.679804247137988,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042822716401865314,
      "loss": 5.2483,
      "mean_token_accuracy": 0.20671714842319489,
      "num_tokens": 62149432.0,
      "step": 30665
    },
    {
      "entropy": 5.999237728118897,
      "epoch": 2.680241195490693,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042820434407269347,
      "loss": 5.2713,
      "mean_token_accuracy": 0.2032574877142906,
      "num_tokens": 62160945.0,
      "step": 30670
    },
    {
      "entropy": 6.009237003326416,
      "epoch": 2.680678143843398,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004281815211881942,
      "loss": 5.2734,
      "mean_token_accuracy": 0.19804950952529907,
      "num_tokens": 62171514.0,
      "step": 30675
    },
    {
      "entropy": 6.0145498752594,
      "epoch": 2.6811150921961024,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004281586953655931,
      "loss": 5.3312,
      "mean_token_accuracy": 0.19671573042869567,
      "num_tokens": 62181011.0,
      "step": 30680
    },
    {
      "entropy": 5.967332649230957,
      "epoch": 2.681552040548807,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004281358666053279,
      "loss": 5.2551,
      "mean_token_accuracy": 0.20736487209796906,
      "num_tokens": 62190915.0,
      "step": 30685
    },
    {
      "entropy": 6.000997591018677,
      "epoch": 2.681988988901512,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004281130349078366,
      "loss": 5.2604,
      "mean_token_accuracy": 0.20205869078636168,
      "num_tokens": 62200660.0,
      "step": 30690
    },
    {
      "entropy": 5.989899206161499,
      "epoch": 2.6824259372542167,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000428090200273557,
      "loss": 5.2851,
      "mean_token_accuracy": 0.20300278812646866,
      "num_tokens": 62210425.0,
      "step": 30695
    },
    {
      "entropy": 5.977500534057617,
      "epoch": 2.682862885606921,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004280673627029272,
      "loss": 5.2238,
      "mean_token_accuracy": 0.21200932264328004,
      "num_tokens": 62220921.0,
      "step": 30700
    },
    {
      "entropy": 5.993402004241943,
      "epoch": 2.683299833959626,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004280445221963852,
      "loss": 5.2786,
      "mean_token_accuracy": 0.2026844561100006,
      "num_tokens": 62230561.0,
      "step": 30705
    },
    {
      "entropy": 6.005105018615723,
      "epoch": 2.6837367823123306,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000428021678754369,
      "loss": 5.317,
      "mean_token_accuracy": 0.19888503551483155,
      "num_tokens": 62240744.0,
      "step": 30710
    },
    {
      "entropy": 5.956674003601075,
      "epoch": 2.6841737306650355,
      "grad_norm": 1.125,
      "learning_rate": 0.00042799883237731683,
      "loss": 5.2125,
      "mean_token_accuracy": 0.20805733203887938,
      "num_tokens": 62250902.0,
      "step": 30715
    },
    {
      "entropy": 6.029712915420532,
      "epoch": 2.68461067901774,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042797598306566694,
      "loss": 5.322,
      "mean_token_accuracy": 0.2011132135987282,
      "num_tokens": 62260459.0,
      "step": 30720
    },
    {
      "entropy": 6.00959267616272,
      "epoch": 2.685047627370445,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004279531308198575,
      "loss": 5.1886,
      "mean_token_accuracy": 0.212527534365654,
      "num_tokens": 62269001.0,
      "step": 30725
    },
    {
      "entropy": 5.9595178127288815,
      "epoch": 2.6854845757231494,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042793027564032684,
      "loss": 5.2154,
      "mean_token_accuracy": 0.2049117773771286,
      "num_tokens": 62279197.0,
      "step": 30730
    },
    {
      "entropy": 5.961982488632202,
      "epoch": 2.6859215240758543,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042790741752751337,
      "loss": 5.2337,
      "mean_token_accuracy": 0.20392071157693864,
      "num_tokens": 62289229.0,
      "step": 30735
    },
    {
      "entropy": 5.936188268661499,
      "epoch": 2.6863584724285587,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042788455648185556,
      "loss": 5.2459,
      "mean_token_accuracy": 0.21129808723926544,
      "num_tokens": 62298819.0,
      "step": 30740
    },
    {
      "entropy": 6.031307554244995,
      "epoch": 2.6867954207812637,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042786169250379175,
      "loss": 5.2351,
      "mean_token_accuracy": 0.20629674792289734,
      "num_tokens": 62308567.0,
      "step": 30745
    },
    {
      "entropy": 6.0218284130096436,
      "epoch": 2.6872323691339686,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042783882559376067,
      "loss": 5.2932,
      "mean_token_accuracy": 0.20481058210134506,
      "num_tokens": 62318552.0,
      "step": 30750
    },
    {
      "entropy": 5.994798040390014,
      "epoch": 2.687669317486673,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042781595575220086,
      "loss": 5.2309,
      "mean_token_accuracy": 0.20902583748102188,
      "num_tokens": 62328430.0,
      "step": 30755
    },
    {
      "entropy": 6.027747774124146,
      "epoch": 2.688106265839378,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042779308297955084,
      "loss": 5.3334,
      "mean_token_accuracy": 0.19721265882253647,
      "num_tokens": 62338540.0,
      "step": 30760
    },
    {
      "entropy": 5.933374071121216,
      "epoch": 2.6885432141920824,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042777020727624945,
      "loss": 5.2895,
      "mean_token_accuracy": 0.2005622237920761,
      "num_tokens": 62348161.0,
      "step": 30765
    },
    {
      "entropy": 6.050360250473022,
      "epoch": 2.6889801625447873,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004277473286427354,
      "loss": 5.3636,
      "mean_token_accuracy": 0.1967689126729965,
      "num_tokens": 62358217.0,
      "step": 30770
    },
    {
      "entropy": 5.994146108627319,
      "epoch": 2.689417110897492,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00042772444707944767,
      "loss": 5.2363,
      "mean_token_accuracy": 0.2022579401731491,
      "num_tokens": 62368801.0,
      "step": 30775
    },
    {
      "entropy": 6.025154161453247,
      "epoch": 2.6898540592501967,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004277015625868249,
      "loss": 5.273,
      "mean_token_accuracy": 0.20342935472726822,
      "num_tokens": 62377632.0,
      "step": 30780
    },
    {
      "entropy": 6.024839258193969,
      "epoch": 2.690291007602901,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004276786751653062,
      "loss": 5.3025,
      "mean_token_accuracy": 0.20264742374420167,
      "num_tokens": 62388056.0,
      "step": 30785
    },
    {
      "entropy": 6.0284685611724855,
      "epoch": 2.690727955955606,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004276557848153305,
      "loss": 5.2546,
      "mean_token_accuracy": 0.20290948897600175,
      "num_tokens": 62398015.0,
      "step": 30790
    },
    {
      "entropy": 6.013485431671143,
      "epoch": 2.6911649043083106,
      "grad_norm": 1.25,
      "learning_rate": 0.00042763289153733685,
      "loss": 5.3177,
      "mean_token_accuracy": 0.19836321473121643,
      "num_tokens": 62408174.0,
      "step": 30795
    },
    {
      "entropy": 6.049573087692261,
      "epoch": 2.6916018526610155,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042760999533176435,
      "loss": 5.2722,
      "mean_token_accuracy": 0.19843191355466844,
      "num_tokens": 62417571.0,
      "step": 30800
    },
    {
      "entropy": 5.985652828216553,
      "epoch": 2.6920388010137204,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004275870961990523,
      "loss": 5.308,
      "mean_token_accuracy": 0.20871504694223403,
      "num_tokens": 62427734.0,
      "step": 30805
    },
    {
      "entropy": 6.041678714752197,
      "epoch": 2.692475749366425,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042756419413963965,
      "loss": 5.2646,
      "mean_token_accuracy": 0.20149241238832474,
      "num_tokens": 62437109.0,
      "step": 30810
    },
    {
      "entropy": 5.969500589370727,
      "epoch": 2.6929126977191293,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042754128915396583,
      "loss": 5.257,
      "mean_token_accuracy": 0.20446482449769973,
      "num_tokens": 62448163.0,
      "step": 30815
    },
    {
      "entropy": 5.990156507492065,
      "epoch": 2.6933496460718342,
      "grad_norm": 1.25,
      "learning_rate": 0.00042751838124247015,
      "loss": 5.2958,
      "mean_token_accuracy": 0.20004887729883195,
      "num_tokens": 62457993.0,
      "step": 30820
    },
    {
      "entropy": 6.047331619262695,
      "epoch": 2.693786594424539,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042749547040559204,
      "loss": 5.3617,
      "mean_token_accuracy": 0.2020869418978691,
      "num_tokens": 62468554.0,
      "step": 30825
    },
    {
      "entropy": 6.015768527984619,
      "epoch": 2.6942235427772436,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004274725566437709,
      "loss": 5.2498,
      "mean_token_accuracy": 0.20585893839597702,
      "num_tokens": 62479204.0,
      "step": 30830
    },
    {
      "entropy": 5.965945291519165,
      "epoch": 2.6946604911299485,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004274496399574463,
      "loss": 5.2464,
      "mean_token_accuracy": 0.20526932179927826,
      "num_tokens": 62491391.0,
      "step": 30835
    },
    {
      "entropy": 5.970257711410523,
      "epoch": 2.695097439482653,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004274267203470576,
      "loss": 5.2087,
      "mean_token_accuracy": 0.21198890805244447,
      "num_tokens": 62500796.0,
      "step": 30840
    },
    {
      "entropy": 5.997080230712891,
      "epoch": 2.695534387835358,
      "grad_norm": 1.28125,
      "learning_rate": 0.00042740379781304455,
      "loss": 5.2496,
      "mean_token_accuracy": 0.2072827026247978,
      "num_tokens": 62509996.0,
      "step": 30845
    },
    {
      "entropy": 6.016570711135865,
      "epoch": 2.6959713361880624,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042738087235584684,
      "loss": 5.2521,
      "mean_token_accuracy": 0.204369980096817,
      "num_tokens": 62519676.0,
      "step": 30850
    },
    {
      "entropy": 6.026767015457153,
      "epoch": 2.6964082845407673,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042735794397590407,
      "loss": 5.2642,
      "mean_token_accuracy": 0.20784118473529817,
      "num_tokens": 62530775.0,
      "step": 30855
    },
    {
      "entropy": 5.960062551498413,
      "epoch": 2.696845232893472,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004273350126736562,
      "loss": 5.2457,
      "mean_token_accuracy": 0.20863468796014786,
      "num_tokens": 62540643.0,
      "step": 30860
    },
    {
      "entropy": 5.988303565979004,
      "epoch": 2.6972821812461767,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004273120784495429,
      "loss": 5.2141,
      "mean_token_accuracy": 0.21105452328920365,
      "num_tokens": 62549766.0,
      "step": 30865
    },
    {
      "entropy": 5.974072074890136,
      "epoch": 2.697719129598881,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004272891413040042,
      "loss": 5.2023,
      "mean_token_accuracy": 0.21110741943120956,
      "num_tokens": 62560130.0,
      "step": 30870
    },
    {
      "entropy": 5.995970058441162,
      "epoch": 2.698156077951586,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004272662012374799,
      "loss": 5.2971,
      "mean_token_accuracy": 0.20238592922687532,
      "num_tokens": 62571256.0,
      "step": 30875
    },
    {
      "entropy": 5.966144561767578,
      "epoch": 2.698593026304291,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004272432582504101,
      "loss": 5.1972,
      "mean_token_accuracy": 0.21408095955848694,
      "num_tokens": 62582269.0,
      "step": 30880
    },
    {
      "entropy": 5.95723147392273,
      "epoch": 2.6990299746569955,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042722031234323485,
      "loss": 5.2408,
      "mean_token_accuracy": 0.2086462765932083,
      "num_tokens": 62591875.0,
      "step": 30885
    },
    {
      "entropy": 6.080978107452393,
      "epoch": 2.6994669230097004,
      "grad_norm": 1.375,
      "learning_rate": 0.00042719736351639416,
      "loss": 5.3456,
      "mean_token_accuracy": 0.19914751052856444,
      "num_tokens": 62602467.0,
      "step": 30890
    },
    {
      "entropy": 6.011315965652466,
      "epoch": 2.699903871362405,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004271744117703284,
      "loss": 5.2271,
      "mean_token_accuracy": 0.20629620552062988,
      "num_tokens": 62611191.0,
      "step": 30895
    },
    {
      "entropy": 5.932937097549439,
      "epoch": 2.7003408197151098,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042715145710547767,
      "loss": 5.3121,
      "mean_token_accuracy": 0.20751658976078033,
      "num_tokens": 62620961.0,
      "step": 30900
    },
    {
      "entropy": 5.9835106372833256,
      "epoch": 2.700777768067814,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004271284995222822,
      "loss": 5.2575,
      "mean_token_accuracy": 0.20876320600509643,
      "num_tokens": 62631972.0,
      "step": 30905
    },
    {
      "entropy": 6.023672962188721,
      "epoch": 2.701214716420519,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004271055390211824,
      "loss": 5.2149,
      "mean_token_accuracy": 0.2129994347691536,
      "num_tokens": 62641073.0,
      "step": 30910
    },
    {
      "entropy": 5.994776105880737,
      "epoch": 2.701651664773224,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004270825756026187,
      "loss": 5.2534,
      "mean_token_accuracy": 0.20479032695293425,
      "num_tokens": 62650364.0,
      "step": 30915
    },
    {
      "entropy": 5.9826812744140625,
      "epoch": 2.7020886131259285,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004270596092670315,
      "loss": 5.2929,
      "mean_token_accuracy": 0.20629042237997056,
      "num_tokens": 62659846.0,
      "step": 30920
    },
    {
      "entropy": 6.002386236190796,
      "epoch": 2.702525561478633,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004270366400148613,
      "loss": 5.292,
      "mean_token_accuracy": 0.20634525865316392,
      "num_tokens": 62670466.0,
      "step": 30925
    },
    {
      "entropy": 5.968366241455078,
      "epoch": 2.702962509831338,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004270136678465487,
      "loss": 5.1991,
      "mean_token_accuracy": 0.20876386016607285,
      "num_tokens": 62680330.0,
      "step": 30930
    },
    {
      "entropy": 6.070248413085937,
      "epoch": 2.703399458184043,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042699069276253427,
      "loss": 5.3432,
      "mean_token_accuracy": 0.1960307851433754,
      "num_tokens": 62690177.0,
      "step": 30935
    },
    {
      "entropy": 6.033216762542724,
      "epoch": 2.7038364065367473,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004269677147632587,
      "loss": 5.2685,
      "mean_token_accuracy": 0.20284547954797744,
      "num_tokens": 62699365.0,
      "step": 30940
    },
    {
      "entropy": 5.9078325748443605,
      "epoch": 2.704273354889452,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004269447338491628,
      "loss": 5.2199,
      "mean_token_accuracy": 0.20928723365068436,
      "num_tokens": 62709825.0,
      "step": 30945
    },
    {
      "entropy": 6.025191020965576,
      "epoch": 2.7047103032421567,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004269217500206873,
      "loss": 5.3084,
      "mean_token_accuracy": 0.20953065603971482,
      "num_tokens": 62719848.0,
      "step": 30950
    },
    {
      "entropy": 6.027315378189087,
      "epoch": 2.7051472515948616,
      "grad_norm": 1.3125,
      "learning_rate": 0.00042689876327827293,
      "loss": 5.3584,
      "mean_token_accuracy": 0.19880958795547485,
      "num_tokens": 62730405.0,
      "step": 30955
    },
    {
      "entropy": 5.9869043827056885,
      "epoch": 2.705584199947566,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042687577362236073,
      "loss": 5.2986,
      "mean_token_accuracy": 0.2001892238855362,
      "num_tokens": 62740275.0,
      "step": 30960
    },
    {
      "entropy": 6.003747081756591,
      "epoch": 2.706021148300271,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004268527810533917,
      "loss": 5.2747,
      "mean_token_accuracy": 0.2106364354491234,
      "num_tokens": 62750847.0,
      "step": 30965
    },
    {
      "entropy": 6.040261125564575,
      "epoch": 2.706458096652976,
      "grad_norm": 1.125,
      "learning_rate": 0.00042682978557180665,
      "loss": 5.3327,
      "mean_token_accuracy": 0.1962918221950531,
      "num_tokens": 62761688.0,
      "step": 30970
    },
    {
      "entropy": 6.034981679916382,
      "epoch": 2.7068950450056803,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00042680678717804676,
      "loss": 5.3191,
      "mean_token_accuracy": 0.20038652271032334,
      "num_tokens": 62771168.0,
      "step": 30975
    },
    {
      "entropy": 5.935457372665406,
      "epoch": 2.707331993358385,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00042678378587255315,
      "loss": 5.1699,
      "mean_token_accuracy": 0.21506573855876923,
      "num_tokens": 62780412.0,
      "step": 30980
    },
    {
      "entropy": 6.044362735748291,
      "epoch": 2.7077689417110897,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042676078165576706,
      "loss": 5.2881,
      "mean_token_accuracy": 0.20519735366106034,
      "num_tokens": 62791673.0,
      "step": 30985
    },
    {
      "entropy": 6.074352693557739,
      "epoch": 2.7082058900637946,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004267377745281296,
      "loss": 5.2909,
      "mean_token_accuracy": 0.19873027205467225,
      "num_tokens": 62800980.0,
      "step": 30990
    },
    {
      "entropy": 5.982162761688232,
      "epoch": 2.708642838416499,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042671476449008205,
      "loss": 5.2449,
      "mean_token_accuracy": 0.20515130162239076,
      "num_tokens": 62813183.0,
      "step": 30995
    },
    {
      "entropy": 5.984153461456299,
      "epoch": 2.709079786769204,
      "grad_norm": 1.21875,
      "learning_rate": 0.000426691751542066,
      "loss": 5.2878,
      "mean_token_accuracy": 0.20436179041862487,
      "num_tokens": 62823170.0,
      "step": 31000
    },
    {
      "entropy": 6.004971933364868,
      "epoch": 2.7095167351219085,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004266687356845225,
      "loss": 5.2573,
      "mean_token_accuracy": 0.2027103751897812,
      "num_tokens": 62833024.0,
      "step": 31005
    },
    {
      "entropy": 5.9675671577453615,
      "epoch": 2.7099536834746134,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042664571691789323,
      "loss": 5.2926,
      "mean_token_accuracy": 0.20163017511367798,
      "num_tokens": 62842533.0,
      "step": 31010
    },
    {
      "entropy": 6.039025926589966,
      "epoch": 2.710390631827318,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004266226952426197,
      "loss": 5.3068,
      "mean_token_accuracy": 0.20112222284078599,
      "num_tokens": 62852984.0,
      "step": 31015
    },
    {
      "entropy": 6.0505678176879885,
      "epoch": 2.710827580180023,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004265996706591434,
      "loss": 5.3086,
      "mean_token_accuracy": 0.19885964840650558,
      "num_tokens": 62861926.0,
      "step": 31020
    },
    {
      "entropy": 6.0320014476776125,
      "epoch": 2.7112645285327277,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042657664316790594,
      "loss": 5.204,
      "mean_token_accuracy": 0.21198493391275405,
      "num_tokens": 62872194.0,
      "step": 31025
    },
    {
      "entropy": 5.994629955291748,
      "epoch": 2.711701476885432,
      "grad_norm": 1.25,
      "learning_rate": 0.000426553612769349,
      "loss": 5.2022,
      "mean_token_accuracy": 0.20919970124959947,
      "num_tokens": 62882887.0,
      "step": 31030
    },
    {
      "entropy": 5.9377789974212645,
      "epoch": 2.7121384252381366,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042653057946391446,
      "loss": 5.261,
      "mean_token_accuracy": 0.2024497702717781,
      "num_tokens": 62893563.0,
      "step": 31035
    },
    {
      "entropy": 6.0182267189025875,
      "epoch": 2.7125753735908416,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042650754325204395,
      "loss": 5.2616,
      "mean_token_accuracy": 0.2043445497751236,
      "num_tokens": 62902691.0,
      "step": 31040
    },
    {
      "entropy": 5.944996452331543,
      "epoch": 2.7130123219435465,
      "grad_norm": 1.28125,
      "learning_rate": 0.00042648450413417937,
      "loss": 5.2023,
      "mean_token_accuracy": 0.21194286942481994,
      "num_tokens": 62912261.0,
      "step": 31045
    },
    {
      "entropy": 6.038604497909546,
      "epoch": 2.713449270296251,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042646146211076263,
      "loss": 5.3368,
      "mean_token_accuracy": 0.20153747498989105,
      "num_tokens": 62922438.0,
      "step": 31050
    },
    {
      "entropy": 5.991335201263428,
      "epoch": 2.713886218648956,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004264384171822357,
      "loss": 5.2062,
      "mean_token_accuracy": 0.20730459094047546,
      "num_tokens": 62932378.0,
      "step": 31055
    },
    {
      "entropy": 6.049380731582642,
      "epoch": 2.7143231670016603,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00042641536934904064,
      "loss": 5.3698,
      "mean_token_accuracy": 0.19440572708845139,
      "num_tokens": 62943590.0,
      "step": 31060
    },
    {
      "entropy": 5.974207735061645,
      "epoch": 2.7147601153543652,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004263923186116193,
      "loss": 5.2588,
      "mean_token_accuracy": 0.20505766719579696,
      "num_tokens": 62953393.0,
      "step": 31065
    },
    {
      "entropy": 6.028953266143799,
      "epoch": 2.7151970637070697,
      "grad_norm": 1.125,
      "learning_rate": 0.00042636926497041406,
      "loss": 5.2675,
      "mean_token_accuracy": 0.20517832338809966,
      "num_tokens": 62964407.0,
      "step": 31070
    },
    {
      "entropy": 5.984571695327759,
      "epoch": 2.7156340120597746,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000426346208425867,
      "loss": 5.3101,
      "mean_token_accuracy": 0.20057284981012344,
      "num_tokens": 62974968.0,
      "step": 31075
    },
    {
      "entropy": 6.022478914260864,
      "epoch": 2.716070960412479,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004263231489784203,
      "loss": 5.2225,
      "mean_token_accuracy": 0.20495116263628005,
      "num_tokens": 62984978.0,
      "step": 31080
    },
    {
      "entropy": 5.9919610023498535,
      "epoch": 2.716507908765184,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004263000866285164,
      "loss": 5.3199,
      "mean_token_accuracy": 0.20099313110113143,
      "num_tokens": 62995139.0,
      "step": 31085
    },
    {
      "entropy": 5.9910320281982425,
      "epoch": 2.7169448571178885,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042627702137659746,
      "loss": 5.3138,
      "mean_token_accuracy": 0.20262989699840545,
      "num_tokens": 63006127.0,
      "step": 31090
    },
    {
      "entropy": 6.081096601486206,
      "epoch": 2.7173818054705934,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042625395322310605,
      "loss": 5.3357,
      "mean_token_accuracy": 0.19811951220035554,
      "num_tokens": 63015844.0,
      "step": 31095
    },
    {
      "entropy": 6.058301019668579,
      "epoch": 2.7178187538232983,
      "grad_norm": 1.296875,
      "learning_rate": 0.00042623088216848446,
      "loss": 5.2594,
      "mean_token_accuracy": 0.20553119629621505,
      "num_tokens": 63025815.0,
      "step": 31100
    },
    {
      "entropy": 5.989779138565064,
      "epoch": 2.7182557021760028,
      "grad_norm": 1.125,
      "learning_rate": 0.00042620780821317543,
      "loss": 5.3467,
      "mean_token_accuracy": 0.19940211921930312,
      "num_tokens": 63036108.0,
      "step": 31105
    },
    {
      "entropy": 5.9457896709442135,
      "epoch": 2.7186926505287072,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004261847313576214,
      "loss": 5.2283,
      "mean_token_accuracy": 0.21315347254276276,
      "num_tokens": 63046305.0,
      "step": 31110
    },
    {
      "entropy": 6.085269498825073,
      "epoch": 2.719129598881412,
      "grad_norm": 1.109375,
      "learning_rate": 0.000426161651602265,
      "loss": 5.3932,
      "mean_token_accuracy": 0.19320755451917648,
      "num_tokens": 63057115.0,
      "step": 31115
    },
    {
      "entropy": 6.018246126174927,
      "epoch": 2.719566547234117,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042613856894754886,
      "loss": 5.2361,
      "mean_token_accuracy": 0.20948751866817475,
      "num_tokens": 63067730.0,
      "step": 31120
    },
    {
      "entropy": 5.957409906387329,
      "epoch": 2.7200034955868215,
      "grad_norm": 1.3125,
      "learning_rate": 0.00042611548339391574,
      "loss": 5.2581,
      "mean_token_accuracy": 0.205299574136734,
      "num_tokens": 63077510.0,
      "step": 31125
    },
    {
      "entropy": 5.942522048950195,
      "epoch": 2.7204404439395264,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042609239494180857,
      "loss": 5.2129,
      "mean_token_accuracy": 0.20849209576845168,
      "num_tokens": 63087160.0,
      "step": 31130
    },
    {
      "entropy": 6.028599452972412,
      "epoch": 2.720877392292231,
      "grad_norm": 1.125,
      "learning_rate": 0.0004260693035916701,
      "loss": 5.2941,
      "mean_token_accuracy": 0.20317198485136032,
      "num_tokens": 63097284.0,
      "step": 31135
    },
    {
      "entropy": 6.0514050960540775,
      "epoch": 2.721314340644936,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00042604620934394313,
      "loss": 5.3021,
      "mean_token_accuracy": 0.20166370421648025,
      "num_tokens": 63106910.0,
      "step": 31140
    },
    {
      "entropy": 6.004741573333741,
      "epoch": 2.7217512889976403,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042602311219907076,
      "loss": 5.1969,
      "mean_token_accuracy": 0.208535735309124,
      "num_tokens": 63116129.0,
      "step": 31145
    },
    {
      "entropy": 5.978310537338257,
      "epoch": 2.722188237350345,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042600001215749593,
      "loss": 5.2753,
      "mean_token_accuracy": 0.20168716460466385,
      "num_tokens": 63125602.0,
      "step": 31150
    },
    {
      "entropy": 6.01067271232605,
      "epoch": 2.72262518570305,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004259769092196618,
      "loss": 5.2538,
      "mean_token_accuracy": 0.20698297321796416,
      "num_tokens": 63136592.0,
      "step": 31155
    },
    {
      "entropy": 6.037215232849121,
      "epoch": 2.7230621340557546,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042595380338601144,
      "loss": 5.2624,
      "mean_token_accuracy": 0.20562649816274642,
      "num_tokens": 63147040.0,
      "step": 31160
    },
    {
      "entropy": 5.960118961334229,
      "epoch": 2.723499082408459,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000425930694656988,
      "loss": 5.2564,
      "mean_token_accuracy": 0.19970947951078416,
      "num_tokens": 63156766.0,
      "step": 31165
    },
    {
      "entropy": 5.984361028671264,
      "epoch": 2.723936030761164,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004259075830330348,
      "loss": 5.2657,
      "mean_token_accuracy": 0.2010002389550209,
      "num_tokens": 63166253.0,
      "step": 31170
    },
    {
      "entropy": 6.025125551223755,
      "epoch": 2.724372979113869,
      "grad_norm": 1.296875,
      "learning_rate": 0.000425884468514595,
      "loss": 5.3435,
      "mean_token_accuracy": 0.1994685247540474,
      "num_tokens": 63177352.0,
      "step": 31175
    },
    {
      "entropy": 6.0443751335144045,
      "epoch": 2.7248099274665734,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004258613511021121,
      "loss": 5.3538,
      "mean_token_accuracy": 0.1936429336667061,
      "num_tokens": 63188092.0,
      "step": 31180
    },
    {
      "entropy": 5.980490970611572,
      "epoch": 2.7252468758192783,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004258382307960294,
      "loss": 5.2192,
      "mean_token_accuracy": 0.21233890801668168,
      "num_tokens": 63197677.0,
      "step": 31185
    },
    {
      "entropy": 6.033487701416016,
      "epoch": 2.7256838241719827,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004258151075967904,
      "loss": 5.3865,
      "mean_token_accuracy": 0.19550403207540512,
      "num_tokens": 63208536.0,
      "step": 31190
    },
    {
      "entropy": 5.990556097030639,
      "epoch": 2.7261207725246877,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004257919815048387,
      "loss": 5.2359,
      "mean_token_accuracy": 0.20670511871576308,
      "num_tokens": 63217966.0,
      "step": 31195
    },
    {
      "entropy": 6.02513689994812,
      "epoch": 2.726557720877392,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004257688525206176,
      "loss": 5.2629,
      "mean_token_accuracy": 0.20422470420598984,
      "num_tokens": 63227420.0,
      "step": 31200
    },
    {
      "entropy": 6.057897615432739,
      "epoch": 2.726994669230097,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004257457206445711,
      "loss": 5.3354,
      "mean_token_accuracy": 0.20006177127361296,
      "num_tokens": 63237887.0,
      "step": 31205
    },
    {
      "entropy": 5.987963342666626,
      "epoch": 2.727431617582802,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00042572258587714255,
      "loss": 5.2578,
      "mean_token_accuracy": 0.20737747550010682,
      "num_tokens": 63247539.0,
      "step": 31210
    },
    {
      "entropy": 6.0033355236053465,
      "epoch": 2.7278685659355064,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004256994482187759,
      "loss": 5.1996,
      "mean_token_accuracy": 0.2142936959862709,
      "num_tokens": 63257839.0,
      "step": 31215
    },
    {
      "entropy": 5.972794532775879,
      "epoch": 2.728305514288211,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042567630766991477,
      "loss": 5.3188,
      "mean_token_accuracy": 0.20297221094369888,
      "num_tokens": 63268667.0,
      "step": 31220
    },
    {
      "entropy": 6.057065868377686,
      "epoch": 2.728742462640916,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004256531642310032,
      "loss": 5.3181,
      "mean_token_accuracy": 0.20439847111701964,
      "num_tokens": 63277413.0,
      "step": 31225
    },
    {
      "entropy": 6.050063753128052,
      "epoch": 2.7291794109936207,
      "grad_norm": 1.21875,
      "learning_rate": 0.000425630017902485,
      "loss": 5.289,
      "mean_token_accuracy": 0.20945686399936675,
      "num_tokens": 63287225.0,
      "step": 31230
    },
    {
      "entropy": 5.990861082077027,
      "epoch": 2.729616359346325,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004256068686848041,
      "loss": 5.2672,
      "mean_token_accuracy": 0.20744854509830474,
      "num_tokens": 63297211.0,
      "step": 31235
    },
    {
      "entropy": 5.935460758209229,
      "epoch": 2.73005330769903,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004255837165784045,
      "loss": 5.249,
      "mean_token_accuracy": 0.2145688459277153,
      "num_tokens": 63307443.0,
      "step": 31240
    },
    {
      "entropy": 5.979538774490356,
      "epoch": 2.7304902560517346,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004255605615837304,
      "loss": 5.3181,
      "mean_token_accuracy": 0.20457523316144943,
      "num_tokens": 63317951.0,
      "step": 31245
    },
    {
      "entropy": 6.012710618972778,
      "epoch": 2.7309272044044395,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042553740370122574,
      "loss": 5.2574,
      "mean_token_accuracy": 0.20919235944747924,
      "num_tokens": 63328138.0,
      "step": 31250
    },
    {
      "entropy": 5.968909645080567,
      "epoch": 2.731364152757144,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004255142429313348,
      "loss": 5.161,
      "mean_token_accuracy": 0.21797444820404052,
      "num_tokens": 63338229.0,
      "step": 31255
    },
    {
      "entropy": 6.031705093383789,
      "epoch": 2.731801101109849,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004254910792745019,
      "loss": 5.3118,
      "mean_token_accuracy": 0.2055996224284172,
      "num_tokens": 63348454.0,
      "step": 31260
    },
    {
      "entropy": 6.009502315521241,
      "epoch": 2.7322380494625538,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042546791273117125,
      "loss": 5.1597,
      "mean_token_accuracy": 0.21219008415937424,
      "num_tokens": 63357346.0,
      "step": 31265
    },
    {
      "entropy": 5.97611813545227,
      "epoch": 2.7326749978152582,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004254447433017871,
      "loss": 5.3179,
      "mean_token_accuracy": 0.1989498883485794,
      "num_tokens": 63368329.0,
      "step": 31270
    },
    {
      "entropy": 6.040006113052368,
      "epoch": 2.7331119461679627,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042542157098679397,
      "loss": 5.3464,
      "mean_token_accuracy": 0.1973353073000908,
      "num_tokens": 63378815.0,
      "step": 31275
    },
    {
      "entropy": 6.011323404312134,
      "epoch": 2.7335488945206676,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004253983957866364,
      "loss": 5.2318,
      "mean_token_accuracy": 0.20592664629220964,
      "num_tokens": 63390099.0,
      "step": 31280
    },
    {
      "entropy": 6.008476209640503,
      "epoch": 2.7339858428733725,
      "grad_norm": 1.40625,
      "learning_rate": 0.00042537521770175856,
      "loss": 5.3257,
      "mean_token_accuracy": 0.1983812227845192,
      "num_tokens": 63401244.0,
      "step": 31285
    },
    {
      "entropy": 5.968120002746582,
      "epoch": 2.734422791226077,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00042535203673260544,
      "loss": 5.2625,
      "mean_token_accuracy": 0.20962802618741988,
      "num_tokens": 63410750.0,
      "step": 31290
    },
    {
      "entropy": 5.98485918045044,
      "epoch": 2.734859739578782,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042532885287962146,
      "loss": 5.2391,
      "mean_token_accuracy": 0.2048269346356392,
      "num_tokens": 63420936.0,
      "step": 31295
    },
    {
      "entropy": 5.963919019699096,
      "epoch": 2.7352966879314864,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042530566614325125,
      "loss": 5.2716,
      "mean_token_accuracy": 0.21321940869092942,
      "num_tokens": 63431614.0,
      "step": 31300
    },
    {
      "entropy": 5.9820719242095945,
      "epoch": 2.7357336362841913,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042528247652393964,
      "loss": 5.2241,
      "mean_token_accuracy": 0.20466772615909576,
      "num_tokens": 63441298.0,
      "step": 31305
    },
    {
      "entropy": 5.995867013931274,
      "epoch": 2.7361705846368958,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00042525928402213127,
      "loss": 5.2409,
      "mean_token_accuracy": 0.20196966230869293,
      "num_tokens": 63451542.0,
      "step": 31310
    },
    {
      "entropy": 6.035914564132691,
      "epoch": 2.7366075329896007,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004252360886382712,
      "loss": 5.2989,
      "mean_token_accuracy": 0.2024863764643669,
      "num_tokens": 63460863.0,
      "step": 31315
    },
    {
      "entropy": 5.996630525588989,
      "epoch": 2.7370444813423056,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042521289037280425,
      "loss": 5.3196,
      "mean_token_accuracy": 0.19715805500745773,
      "num_tokens": 63470499.0,
      "step": 31320
    },
    {
      "entropy": 6.046760082244873,
      "epoch": 2.73748142969501,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042518968922617534,
      "loss": 5.294,
      "mean_token_accuracy": 0.20558284223079681,
      "num_tokens": 63480803.0,
      "step": 31325
    },
    {
      "entropy": 5.980564117431641,
      "epoch": 2.7379183780477145,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004251664851988295,
      "loss": 5.2298,
      "mean_token_accuracy": 0.21001771539449693,
      "num_tokens": 63491544.0,
      "step": 31330
    },
    {
      "entropy": 5.957389307022095,
      "epoch": 2.7383553264004195,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042514327829121175,
      "loss": 5.1384,
      "mean_token_accuracy": 0.21145715713500976,
      "num_tokens": 63501086.0,
      "step": 31335
    },
    {
      "entropy": 5.986258506774902,
      "epoch": 2.7387922747531244,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042512006850376724,
      "loss": 5.268,
      "mean_token_accuracy": 0.20518275946378708,
      "num_tokens": 63512408.0,
      "step": 31340
    },
    {
      "entropy": 6.029832935333252,
      "epoch": 2.739229223105829,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00042509685583694116,
      "loss": 5.2288,
      "mean_token_accuracy": 0.20264763236045838,
      "num_tokens": 63521735.0,
      "step": 31345
    },
    {
      "entropy": 5.988440132141113,
      "epoch": 2.7396661714585337,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004250736402911788,
      "loss": 5.2355,
      "mean_token_accuracy": 0.21149661093950273,
      "num_tokens": 63532045.0,
      "step": 31350
    },
    {
      "entropy": 5.964945983886719,
      "epoch": 2.740103119811238,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004250504218669253,
      "loss": 5.2391,
      "mean_token_accuracy": 0.20617637187242507,
      "num_tokens": 63541286.0,
      "step": 31355
    },
    {
      "entropy": 5.988555145263672,
      "epoch": 2.740540068163943,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004250272005646261,
      "loss": 5.2218,
      "mean_token_accuracy": 0.20904979407787322,
      "num_tokens": 63552195.0,
      "step": 31360
    },
    {
      "entropy": 5.978481578826904,
      "epoch": 2.7409770165166476,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004250039763847266,
      "loss": 5.2267,
      "mean_token_accuracy": 0.2135532394051552,
      "num_tokens": 63561318.0,
      "step": 31365
    },
    {
      "entropy": 5.944555473327637,
      "epoch": 2.7414139648693525,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042498074932767227,
      "loss": 5.279,
      "mean_token_accuracy": 0.20391665995121003,
      "num_tokens": 63571637.0,
      "step": 31370
    },
    {
      "entropy": 6.043389177322387,
      "epoch": 2.7418509132220574,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042495751939390853,
      "loss": 5.3154,
      "mean_token_accuracy": 0.20062548220157622,
      "num_tokens": 63581618.0,
      "step": 31375
    },
    {
      "entropy": 5.936142683029175,
      "epoch": 2.742287861574762,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042493428658388094,
      "loss": 5.1531,
      "mean_token_accuracy": 0.21915891021490097,
      "num_tokens": 63592129.0,
      "step": 31380
    },
    {
      "entropy": 6.009351634979248,
      "epoch": 2.7427248099274664,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004249110508980353,
      "loss": 5.3195,
      "mean_token_accuracy": 0.20445786267518998,
      "num_tokens": 63602861.0,
      "step": 31385
    },
    {
      "entropy": 6.0111494064331055,
      "epoch": 2.7431617582801713,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042488781233681705,
      "loss": 5.2482,
      "mean_token_accuracy": 0.20658016204833984,
      "num_tokens": 63612423.0,
      "step": 31390
    },
    {
      "entropy": 5.972311210632324,
      "epoch": 2.743598706632876,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042486457090067206,
      "loss": 5.3058,
      "mean_token_accuracy": 0.198809914290905,
      "num_tokens": 63622351.0,
      "step": 31395
    },
    {
      "entropy": 5.968535709381103,
      "epoch": 2.7440356549855807,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00042484132659004604,
      "loss": 5.2178,
      "mean_token_accuracy": 0.2099035993218422,
      "num_tokens": 63632024.0,
      "step": 31400
    },
    {
      "entropy": 5.979084920883179,
      "epoch": 2.7444726033382856,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004248180794053849,
      "loss": 5.2628,
      "mean_token_accuracy": 0.20227562487125397,
      "num_tokens": 63641962.0,
      "step": 31405
    },
    {
      "entropy": 5.95117220878601,
      "epoch": 2.74490955169099,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042479482934713446,
      "loss": 5.281,
      "mean_token_accuracy": 0.2080460622906685,
      "num_tokens": 63652095.0,
      "step": 31410
    },
    {
      "entropy": 6.034651851654052,
      "epoch": 2.745346500043695,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004247715764157407,
      "loss": 5.3487,
      "mean_token_accuracy": 0.1993912637233734,
      "num_tokens": 63661392.0,
      "step": 31415
    },
    {
      "entropy": 5.994488191604614,
      "epoch": 2.7457834483963994,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042474832061164966,
      "loss": 5.2432,
      "mean_token_accuracy": 0.20243531912565232,
      "num_tokens": 63672383.0,
      "step": 31420
    },
    {
      "entropy": 6.00038890838623,
      "epoch": 2.7462203967491043,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042472506193530734,
      "loss": 5.2676,
      "mean_token_accuracy": 0.20133698582649232,
      "num_tokens": 63682522.0,
      "step": 31425
    },
    {
      "entropy": 5.976169872283935,
      "epoch": 2.746657345101809,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004247018003871599,
      "loss": 5.2293,
      "mean_token_accuracy": 0.20917905420064925,
      "num_tokens": 63692443.0,
      "step": 31430
    },
    {
      "entropy": 5.971882629394531,
      "epoch": 2.7470942934545137,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042467853596765345,
      "loss": 5.2464,
      "mean_token_accuracy": 0.2093265473842621,
      "num_tokens": 63702268.0,
      "step": 31435
    },
    {
      "entropy": 5.920893049240112,
      "epoch": 2.747531241807218,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042465526867723423,
      "loss": 5.1513,
      "mean_token_accuracy": 0.21600884199142456,
      "num_tokens": 63712621.0,
      "step": 31440
    },
    {
      "entropy": 6.02039065361023,
      "epoch": 2.747968190159923,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004246319985163487,
      "loss": 5.3154,
      "mean_token_accuracy": 0.20215138792991638,
      "num_tokens": 63722319.0,
      "step": 31445
    },
    {
      "entropy": 6.000820207595825,
      "epoch": 2.748405138512628,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004246087254854428,
      "loss": 5.3129,
      "mean_token_accuracy": 0.20619625151157378,
      "num_tokens": 63732726.0,
      "step": 31450
    },
    {
      "entropy": 5.980285024642944,
      "epoch": 2.7488420868653325,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042458544958496333,
      "loss": 5.2103,
      "mean_token_accuracy": 0.21011082381010054,
      "num_tokens": 63741197.0,
      "step": 31455
    },
    {
      "entropy": 5.969593620300293,
      "epoch": 2.749279035218037,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042456217081535645,
      "loss": 5.229,
      "mean_token_accuracy": 0.21185342967510223,
      "num_tokens": 63751316.0,
      "step": 31460
    },
    {
      "entropy": 6.026042222976685,
      "epoch": 2.749715983570742,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042453888917706885,
      "loss": 5.3,
      "mean_token_accuracy": 0.2005605846643448,
      "num_tokens": 63761246.0,
      "step": 31465
    },
    {
      "entropy": 6.044060754776001,
      "epoch": 2.750152931923447,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042451560467054696,
      "loss": 5.3072,
      "mean_token_accuracy": 0.1998155012726784,
      "num_tokens": 63771442.0,
      "step": 31470
    },
    {
      "entropy": 6.0313684940338135,
      "epoch": 2.7505898802761513,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00042449231729623735,
      "loss": 5.3521,
      "mean_token_accuracy": 0.19796697050333023,
      "num_tokens": 63782840.0,
      "step": 31475
    },
    {
      "entropy": 5.999184608459473,
      "epoch": 2.751026828628856,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042446902705458683,
      "loss": 5.2056,
      "mean_token_accuracy": 0.20885823369026185,
      "num_tokens": 63793089.0,
      "step": 31480
    },
    {
      "entropy": 6.011334037780761,
      "epoch": 2.7514637769815606,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004244457339460421,
      "loss": 5.3252,
      "mean_token_accuracy": 0.19500420540571212,
      "num_tokens": 63804125.0,
      "step": 31485
    },
    {
      "entropy": 5.980732727050781,
      "epoch": 2.7519007253342656,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042442243797104976,
      "loss": 5.294,
      "mean_token_accuracy": 0.20389412492513656,
      "num_tokens": 63814713.0,
      "step": 31490
    },
    {
      "entropy": 6.060719585418701,
      "epoch": 2.75233767368697,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004243991391300569,
      "loss": 5.3652,
      "mean_token_accuracy": 0.19336203038692473,
      "num_tokens": 63824403.0,
      "step": 31495
    },
    {
      "entropy": 6.043661642074585,
      "epoch": 2.752774622039675,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004243758374235102,
      "loss": 5.3084,
      "mean_token_accuracy": 0.2042991802096367,
      "num_tokens": 63834182.0,
      "step": 31500
    },
    {
      "entropy": 5.978044271469116,
      "epoch": 2.75321157039238,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004243525328518567,
      "loss": 5.2865,
      "mean_token_accuracy": 0.20249486565589905,
      "num_tokens": 63844282.0,
      "step": 31505
    },
    {
      "entropy": 5.997984075546265,
      "epoch": 2.7536485187450843,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004243292254155433,
      "loss": 5.261,
      "mean_token_accuracy": 0.20485454201698303,
      "num_tokens": 63854898.0,
      "step": 31510
    },
    {
      "entropy": 5.944573450088501,
      "epoch": 2.754085467097789,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042430591511501717,
      "loss": 5.1756,
      "mean_token_accuracy": 0.21876434683799745,
      "num_tokens": 63864501.0,
      "step": 31515
    },
    {
      "entropy": 5.977957439422608,
      "epoch": 2.7545224154504937,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042428260195072527,
      "loss": 5.288,
      "mean_token_accuracy": 0.20476389527320862,
      "num_tokens": 63874812.0,
      "step": 31520
    },
    {
      "entropy": 6.051693534851074,
      "epoch": 2.7549593638031986,
      "grad_norm": 1.25,
      "learning_rate": 0.0004242592859231149,
      "loss": 5.2687,
      "mean_token_accuracy": 0.20250968039035797,
      "num_tokens": 63884028.0,
      "step": 31525
    },
    {
      "entropy": 5.9982129573822025,
      "epoch": 2.755396312155903,
      "grad_norm": 1.34375,
      "learning_rate": 0.00042423596703263324,
      "loss": 5.2741,
      "mean_token_accuracy": 0.20669147223234177,
      "num_tokens": 63893991.0,
      "step": 31530
    },
    {
      "entropy": 6.013180303573608,
      "epoch": 2.755833260508608,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004242126452797275,
      "loss": 5.3227,
      "mean_token_accuracy": 0.20049135982990265,
      "num_tokens": 63904168.0,
      "step": 31535
    },
    {
      "entropy": 6.0149250507354735,
      "epoch": 2.7562702088613125,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042418932066484505,
      "loss": 5.2952,
      "mean_token_accuracy": 0.19625525325536727,
      "num_tokens": 63913881.0,
      "step": 31540
    },
    {
      "entropy": 6.017817878723145,
      "epoch": 2.7567071572140174,
      "grad_norm": 1.25,
      "learning_rate": 0.00042416599318843324,
      "loss": 5.3755,
      "mean_token_accuracy": 0.19797254353761673,
      "num_tokens": 63924521.0,
      "step": 31545
    },
    {
      "entropy": 6.069774770736695,
      "epoch": 2.757144105566722,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004241426628509395,
      "loss": 5.3097,
      "mean_token_accuracy": 0.19389386773109435,
      "num_tokens": 63935046.0,
      "step": 31550
    },
    {
      "entropy": 6.022601842880249,
      "epoch": 2.7575810539194268,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042411932965281143,
      "loss": 5.2825,
      "mean_token_accuracy": 0.1984374701976776,
      "num_tokens": 63944048.0,
      "step": 31555
    },
    {
      "entropy": 5.994948196411133,
      "epoch": 2.7580180022721317,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004240959935944964,
      "loss": 5.2525,
      "mean_token_accuracy": 0.2091617077589035,
      "num_tokens": 63954943.0,
      "step": 31560
    },
    {
      "entropy": 5.959551000595093,
      "epoch": 2.758454950624836,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004240726546764421,
      "loss": 5.247,
      "mean_token_accuracy": 0.2070344626903534,
      "num_tokens": 63965929.0,
      "step": 31565
    },
    {
      "entropy": 5.979798221588135,
      "epoch": 2.7588918989775406,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042404931289909614,
      "loss": 5.2952,
      "mean_token_accuracy": 0.19604102075099944,
      "num_tokens": 63976353.0,
      "step": 31570
    },
    {
      "entropy": 6.052369785308838,
      "epoch": 2.7593288473302455,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042402596826290634,
      "loss": 5.2969,
      "mean_token_accuracy": 0.20043713301420213,
      "num_tokens": 63987073.0,
      "step": 31575
    },
    {
      "entropy": 6.026928091049195,
      "epoch": 2.7597657956829504,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004240026207683203,
      "loss": 5.2845,
      "mean_token_accuracy": 0.20842959135770797,
      "num_tokens": 63996631.0,
      "step": 31580
    },
    {
      "entropy": 5.988203811645508,
      "epoch": 2.760202744035655,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042397927041578596,
      "loss": 5.3039,
      "mean_token_accuracy": 0.20294716358184814,
      "num_tokens": 64007662.0,
      "step": 31585
    },
    {
      "entropy": 5.9358968257904055,
      "epoch": 2.76063969238836,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042395591720575113,
      "loss": 5.2319,
      "mean_token_accuracy": 0.20998739451169968,
      "num_tokens": 64017868.0,
      "step": 31590
    },
    {
      "entropy": 6.043112182617188,
      "epoch": 2.7610766407410643,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042393256113866377,
      "loss": 5.2972,
      "mean_token_accuracy": 0.19394908249378204,
      "num_tokens": 64027464.0,
      "step": 31595
    },
    {
      "entropy": 5.985465717315674,
      "epoch": 2.761513589093769,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004239092022149718,
      "loss": 5.3029,
      "mean_token_accuracy": 0.19793476164340973,
      "num_tokens": 64037795.0,
      "step": 31600
    },
    {
      "entropy": 6.058812665939331,
      "epoch": 2.7619505374464737,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042388584043512333,
      "loss": 5.3139,
      "mean_token_accuracy": 0.20105184614658356,
      "num_tokens": 64048055.0,
      "step": 31605
    },
    {
      "entropy": 6.025070333480835,
      "epoch": 2.7623874857991786,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004238624757995664,
      "loss": 5.1989,
      "mean_token_accuracy": 0.2106286734342575,
      "num_tokens": 64057531.0,
      "step": 31610
    },
    {
      "entropy": 5.972158193588257,
      "epoch": 2.7628244341518835,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004238391083087493,
      "loss": 5.2565,
      "mean_token_accuracy": 0.2039350301027298,
      "num_tokens": 64066476.0,
      "step": 31615
    },
    {
      "entropy": 5.949263381958008,
      "epoch": 2.763261382504588,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042381573796312,
      "loss": 5.2272,
      "mean_token_accuracy": 0.21858266443014146,
      "num_tokens": 64077069.0,
      "step": 31620
    },
    {
      "entropy": 5.97817063331604,
      "epoch": 2.7636983308572924,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004237923647631269,
      "loss": 5.2682,
      "mean_token_accuracy": 0.20783480852842331,
      "num_tokens": 64087325.0,
      "step": 31625
    },
    {
      "entropy": 6.0147223472595215,
      "epoch": 2.7641352792099974,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004237689887092183,
      "loss": 5.2817,
      "mean_token_accuracy": 0.20651257783174515,
      "num_tokens": 64097040.0,
      "step": 31630
    },
    {
      "entropy": 5.977636671066284,
      "epoch": 2.7645722275627023,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042374560980184254,
      "loss": 5.2735,
      "mean_token_accuracy": 0.2062533140182495,
      "num_tokens": 64107798.0,
      "step": 31635
    },
    {
      "entropy": 6.043969392776489,
      "epoch": 2.7650091759154067,
      "grad_norm": 1.140625,
      "learning_rate": 0.000423722228041448,
      "loss": 5.3088,
      "mean_token_accuracy": 0.20485418289899826,
      "num_tokens": 64118395.0,
      "step": 31640
    },
    {
      "entropy": 6.036913537979126,
      "epoch": 2.7654461242681116,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042369884342848327,
      "loss": 5.347,
      "mean_token_accuracy": 0.1961314171552658,
      "num_tokens": 64128040.0,
      "step": 31645
    },
    {
      "entropy": 6.010632419586182,
      "epoch": 2.765883072620816,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042367545596339674,
      "loss": 5.268,
      "mean_token_accuracy": 0.20781236588954927,
      "num_tokens": 64137660.0,
      "step": 31650
    },
    {
      "entropy": 5.962187719345093,
      "epoch": 2.766320020973521,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042365206564663705,
      "loss": 5.281,
      "mean_token_accuracy": 0.2034202143549919,
      "num_tokens": 64147632.0,
      "step": 31655
    },
    {
      "entropy": 5.962792253494262,
      "epoch": 2.7667569693262255,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004236286724786529,
      "loss": 5.2743,
      "mean_token_accuracy": 0.2036137744784355,
      "num_tokens": 64157604.0,
      "step": 31660
    },
    {
      "entropy": 6.050580787658691,
      "epoch": 2.7671939176789304,
      "grad_norm": 1.296875,
      "learning_rate": 0.00042360527645989287,
      "loss": 5.3441,
      "mean_token_accuracy": 0.19531762152910231,
      "num_tokens": 64167626.0,
      "step": 31665
    },
    {
      "entropy": 6.070722007751465,
      "epoch": 2.7676308660316353,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004235818775908059,
      "loss": 5.3244,
      "mean_token_accuracy": 0.20311129838228226,
      "num_tokens": 64178436.0,
      "step": 31670
    },
    {
      "entropy": 5.919925546646118,
      "epoch": 2.76806781438434,
      "grad_norm": 1.125,
      "learning_rate": 0.00042355847587184056,
      "loss": 5.2261,
      "mean_token_accuracy": 0.21584825068712235,
      "num_tokens": 64188911.0,
      "step": 31675
    },
    {
      "entropy": 5.939052629470825,
      "epoch": 2.7685047627370443,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042353507130344584,
      "loss": 5.1883,
      "mean_token_accuracy": 0.21868192106485368,
      "num_tokens": 64198539.0,
      "step": 31680
    },
    {
      "entropy": 6.030122566223144,
      "epoch": 2.768941711089749,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004235116638860705,
      "loss": 5.2909,
      "mean_token_accuracy": 0.20957937091588974,
      "num_tokens": 64209121.0,
      "step": 31685
    },
    {
      "entropy": 6.081256771087647,
      "epoch": 2.769378659442454,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004234882536201638,
      "loss": 5.2904,
      "mean_token_accuracy": 0.20142867267131806,
      "num_tokens": 64218610.0,
      "step": 31690
    },
    {
      "entropy": 5.977560997009277,
      "epoch": 2.7698156077951586,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004234648405061745,
      "loss": 5.264,
      "mean_token_accuracy": 0.2060674771666527,
      "num_tokens": 64228626.0,
      "step": 31695
    },
    {
      "entropy": 5.9535621166229244,
      "epoch": 2.7702525561478635,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042344142454455174,
      "loss": 5.2391,
      "mean_token_accuracy": 0.20881624668836593,
      "num_tokens": 64238575.0,
      "step": 31700
    },
    {
      "entropy": 6.006302833557129,
      "epoch": 2.770689504500568,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004234180057357448,
      "loss": 5.2895,
      "mean_token_accuracy": 0.20205435305833816,
      "num_tokens": 64248551.0,
      "step": 31705
    },
    {
      "entropy": 6.035488033294678,
      "epoch": 2.771126452853273,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042339458408020267,
      "loss": 5.2804,
      "mean_token_accuracy": 0.2015762984752655,
      "num_tokens": 64259315.0,
      "step": 31710
    },
    {
      "entropy": 6.048279333114624,
      "epoch": 2.7715634012059773,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004233711595783746,
      "loss": 5.2823,
      "mean_token_accuracy": 0.20330921858549117,
      "num_tokens": 64269742.0,
      "step": 31715
    },
    {
      "entropy": 6.040583944320678,
      "epoch": 2.7720003495586822,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00042334773223071,
      "loss": 5.2827,
      "mean_token_accuracy": 0.20064716935157775,
      "num_tokens": 64279266.0,
      "step": 31720
    },
    {
      "entropy": 5.99870753288269,
      "epoch": 2.7724372979113867,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042332430203765816,
      "loss": 5.269,
      "mean_token_accuracy": 0.20479351580142974,
      "num_tokens": 64289567.0,
      "step": 31725
    },
    {
      "entropy": 5.980914640426636,
      "epoch": 2.7728742462640916,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004233008689996685,
      "loss": 5.3567,
      "mean_token_accuracy": 0.20226570814847947,
      "num_tokens": 64299126.0,
      "step": 31730
    },
    {
      "entropy": 5.991626596450805,
      "epoch": 2.773311194616796,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004232774331171904,
      "loss": 5.2233,
      "mean_token_accuracy": 0.2169140487909317,
      "num_tokens": 64308964.0,
      "step": 31735
    },
    {
      "entropy": 6.068089294433594,
      "epoch": 2.773748142969501,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042325399439067354,
      "loss": 5.3022,
      "mean_token_accuracy": 0.20346161127090454,
      "num_tokens": 64318463.0,
      "step": 31740
    },
    {
      "entropy": 6.063591146469117,
      "epoch": 2.774185091322206,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004232305528205673,
      "loss": 5.4154,
      "mean_token_accuracy": 0.19552943110466003,
      "num_tokens": 64328841.0,
      "step": 31745
    },
    {
      "entropy": 6.025351190567017,
      "epoch": 2.7746220396749104,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004232071084073214,
      "loss": 5.2796,
      "mean_token_accuracy": 0.21116833537817,
      "num_tokens": 64339037.0,
      "step": 31750
    },
    {
      "entropy": 6.049991655349731,
      "epoch": 2.775058988027615,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042318366115138553,
      "loss": 5.4126,
      "mean_token_accuracy": 0.19724111258983612,
      "num_tokens": 64350048.0,
      "step": 31755
    },
    {
      "entropy": 6.0629109859466555,
      "epoch": 2.7754959363803198,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042316021105320937,
      "loss": 5.3604,
      "mean_token_accuracy": 0.19694602787494658,
      "num_tokens": 64361066.0,
      "step": 31760
    },
    {
      "entropy": 6.064206743240357,
      "epoch": 2.7759328847330247,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042313675811324276,
      "loss": 5.3318,
      "mean_token_accuracy": 0.19681198596954347,
      "num_tokens": 64371064.0,
      "step": 31765
    },
    {
      "entropy": 5.990127944946289,
      "epoch": 2.776369833085729,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042311330233193545,
      "loss": 5.2053,
      "mean_token_accuracy": 0.21297587603330612,
      "num_tokens": 64380820.0,
      "step": 31770
    },
    {
      "entropy": 5.9799117088317875,
      "epoch": 2.776806781438434,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004230898437097374,
      "loss": 5.2075,
      "mean_token_accuracy": 0.20918107330799102,
      "num_tokens": 64389777.0,
      "step": 31775
    },
    {
      "entropy": 6.012884855270386,
      "epoch": 2.7772437297911385,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004230663822470986,
      "loss": 5.366,
      "mean_token_accuracy": 0.19894403517246245,
      "num_tokens": 64399542.0,
      "step": 31780
    },
    {
      "entropy": 5.993813943862915,
      "epoch": 2.7776806781438435,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004230429179444689,
      "loss": 5.2737,
      "mean_token_accuracy": 0.19972531199455262,
      "num_tokens": 64410254.0,
      "step": 31785
    },
    {
      "entropy": 5.972187614440918,
      "epoch": 2.778117626496548,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004230194508022986,
      "loss": 5.2612,
      "mean_token_accuracy": 0.20663903802633285,
      "num_tokens": 64420487.0,
      "step": 31790
    },
    {
      "entropy": 5.998356580734253,
      "epoch": 2.778554574849253,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042299598082103757,
      "loss": 5.2953,
      "mean_token_accuracy": 0.20521116852760315,
      "num_tokens": 64430171.0,
      "step": 31795
    },
    {
      "entropy": 6.020726823806763,
      "epoch": 2.7789915232019577,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00042297250800113614,
      "loss": 5.3575,
      "mean_token_accuracy": 0.192211377620697,
      "num_tokens": 64439960.0,
      "step": 31800
    },
    {
      "entropy": 6.080677080154419,
      "epoch": 2.779428471554662,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004229490323430444,
      "loss": 5.2988,
      "mean_token_accuracy": 0.2029607191681862,
      "num_tokens": 64450238.0,
      "step": 31805
    },
    {
      "entropy": 6.031198596954345,
      "epoch": 2.7798654199073667,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042292555384721277,
      "loss": 5.2647,
      "mean_token_accuracy": 0.2036583036184311,
      "num_tokens": 64460137.0,
      "step": 31810
    },
    {
      "entropy": 5.996274518966675,
      "epoch": 2.7803023682600716,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004229020725140915,
      "loss": 5.2295,
      "mean_token_accuracy": 0.21221008747816086,
      "num_tokens": 64470138.0,
      "step": 31815
    },
    {
      "entropy": 5.985234546661377,
      "epoch": 2.7807393166127765,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004228785883441309,
      "loss": 5.3428,
      "mean_token_accuracy": 0.19719249457120896,
      "num_tokens": 64480536.0,
      "step": 31820
    },
    {
      "entropy": 6.042021083831787,
      "epoch": 2.781176264965481,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004228551013377816,
      "loss": 5.2607,
      "mean_token_accuracy": 0.20955720990896226,
      "num_tokens": 64489899.0,
      "step": 31825
    },
    {
      "entropy": 5.9837347030639645,
      "epoch": 2.781613213318186,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042283161149549387,
      "loss": 5.2324,
      "mean_token_accuracy": 0.21127953827381135,
      "num_tokens": 64500272.0,
      "step": 31830
    },
    {
      "entropy": 5.9872674465179445,
      "epoch": 2.7820501616708904,
      "grad_norm": 1.34375,
      "learning_rate": 0.00042280811881771846,
      "loss": 5.2937,
      "mean_token_accuracy": 0.20708418190479277,
      "num_tokens": 64510507.0,
      "step": 31835
    },
    {
      "entropy": 5.945323133468628,
      "epoch": 2.7824871100235953,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042278462330490584,
      "loss": 5.2176,
      "mean_token_accuracy": 0.21245792806148528,
      "num_tokens": 64521105.0,
      "step": 31840
    },
    {
      "entropy": 5.97895884513855,
      "epoch": 2.7829240583762997,
      "grad_norm": 1.296875,
      "learning_rate": 0.00042276112495750676,
      "loss": 5.2753,
      "mean_token_accuracy": 0.20727930963039398,
      "num_tokens": 64531828.0,
      "step": 31845
    },
    {
      "entropy": 5.988350296020508,
      "epoch": 2.7833610067290047,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004227376237759718,
      "loss": 5.1666,
      "mean_token_accuracy": 0.2091330349445343,
      "num_tokens": 64540783.0,
      "step": 31850
    },
    {
      "entropy": 5.9985393524169925,
      "epoch": 2.7837979550817096,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004227141197607518,
      "loss": 5.2687,
      "mean_token_accuracy": 0.20114681571722032,
      "num_tokens": 64550958.0,
      "step": 31855
    },
    {
      "entropy": 5.940800285339355,
      "epoch": 2.784234903434414,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00042269061291229766,
      "loss": 5.2562,
      "mean_token_accuracy": 0.2090567797422409,
      "num_tokens": 64561666.0,
      "step": 31860
    },
    {
      "entropy": 6.0327033519744875,
      "epoch": 2.7846718517871185,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042266710323106004,
      "loss": 5.3063,
      "mean_token_accuracy": 0.19566449671983718,
      "num_tokens": 64570908.0,
      "step": 31865
    },
    {
      "entropy": 6.0066667079925535,
      "epoch": 2.7851088001398234,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042264359071749017,
      "loss": 5.2298,
      "mean_token_accuracy": 0.20780181735754014,
      "num_tokens": 64580981.0,
      "step": 31870
    },
    {
      "entropy": 6.019966125488281,
      "epoch": 2.7855457484925283,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042262007537203876,
      "loss": 5.2938,
      "mean_token_accuracy": 0.2015400230884552,
      "num_tokens": 64591776.0,
      "step": 31875
    },
    {
      "entropy": 6.007081604003906,
      "epoch": 2.785982696845233,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042259655719515706,
      "loss": 5.2222,
      "mean_token_accuracy": 0.21226180642843245,
      "num_tokens": 64601201.0,
      "step": 31880
    },
    {
      "entropy": 6.003474569320678,
      "epoch": 2.7864196451979377,
      "grad_norm": 1.25,
      "learning_rate": 0.0004225730361872959,
      "loss": 5.2612,
      "mean_token_accuracy": 0.20842698812484742,
      "num_tokens": 64610998.0,
      "step": 31885
    },
    {
      "entropy": 6.076160526275634,
      "epoch": 2.786856593550642,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004225495123489067,
      "loss": 5.3674,
      "mean_token_accuracy": 0.19331634044647217,
      "num_tokens": 64620503.0,
      "step": 31890
    },
    {
      "entropy": 6.0040223598480225,
      "epoch": 2.787293541903347,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004225259856804405,
      "loss": 5.2248,
      "mean_token_accuracy": 0.2161240592598915,
      "num_tokens": 64630172.0,
      "step": 31895
    },
    {
      "entropy": 5.99509859085083,
      "epoch": 2.7877304902560516,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004225024561823486,
      "loss": 5.2414,
      "mean_token_accuracy": 0.21097314953804017,
      "num_tokens": 64639737.0,
      "step": 31900
    },
    {
      "entropy": 5.958954000473023,
      "epoch": 2.7881674386087565,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004224789238550823,
      "loss": 5.1661,
      "mean_token_accuracy": 0.20997340828180314,
      "num_tokens": 64650324.0,
      "step": 31905
    },
    {
      "entropy": 6.041666698455811,
      "epoch": 2.7886043869614614,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004224553886990929,
      "loss": 5.3543,
      "mean_token_accuracy": 0.19570692777633666,
      "num_tokens": 64661938.0,
      "step": 31910
    },
    {
      "entropy": 5.980074501037597,
      "epoch": 2.789041335314166,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042243185071483193,
      "loss": 5.2203,
      "mean_token_accuracy": 0.21057890504598617,
      "num_tokens": 64672042.0,
      "step": 31915
    },
    {
      "entropy": 5.992918062210083,
      "epoch": 2.7894782836668703,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042240830990275085,
      "loss": 5.2809,
      "mean_token_accuracy": 0.20665229707956315,
      "num_tokens": 64681798.0,
      "step": 31920
    },
    {
      "entropy": 6.012958526611328,
      "epoch": 2.7899152320195753,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000422384766263301,
      "loss": 5.3064,
      "mean_token_accuracy": 0.20067310333251953,
      "num_tokens": 64692329.0,
      "step": 31925
    },
    {
      "entropy": 6.025276231765747,
      "epoch": 2.79035218037228,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042236121979693424,
      "loss": 5.2603,
      "mean_token_accuracy": 0.2068995088338852,
      "num_tokens": 64701605.0,
      "step": 31930
    },
    {
      "entropy": 6.048765563964844,
      "epoch": 2.7907891287249846,
      "grad_norm": 1.25,
      "learning_rate": 0.00042233767050410206,
      "loss": 5.3473,
      "mean_token_accuracy": 0.2000545307993889,
      "num_tokens": 64711649.0,
      "step": 31935
    },
    {
      "entropy": 6.011166572570801,
      "epoch": 2.7912260770776895,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004223141183852562,
      "loss": 5.2911,
      "mean_token_accuracy": 0.20042625516653062,
      "num_tokens": 64721451.0,
      "step": 31940
    },
    {
      "entropy": 6.0198756694793705,
      "epoch": 2.791663025430394,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004222905634408482,
      "loss": 5.2431,
      "mean_token_accuracy": 0.20685170888900756,
      "num_tokens": 64731725.0,
      "step": 31945
    },
    {
      "entropy": 6.023550796508789,
      "epoch": 2.792099973783099,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042226700567133015,
      "loss": 5.181,
      "mean_token_accuracy": 0.21415599733591079,
      "num_tokens": 64741731.0,
      "step": 31950
    },
    {
      "entropy": 5.965267086029053,
      "epoch": 2.7925369221358034,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004222434450771537,
      "loss": 5.2996,
      "mean_token_accuracy": 0.20227378755807876,
      "num_tokens": 64753324.0,
      "step": 31955
    },
    {
      "entropy": 5.953118181228637,
      "epoch": 2.7929738704885083,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042221988165877083,
      "loss": 5.2248,
      "mean_token_accuracy": 0.21284350603818894,
      "num_tokens": 64763209.0,
      "step": 31960
    },
    {
      "entropy": 5.998643827438355,
      "epoch": 2.7934108188412132,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004221963154166335,
      "loss": 5.2831,
      "mean_token_accuracy": 0.2048606902360916,
      "num_tokens": 64773867.0,
      "step": 31965
    },
    {
      "entropy": 6.029735708236695,
      "epoch": 2.7938477671939177,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004221727463511937,
      "loss": 5.2661,
      "mean_token_accuracy": 0.19955476820468904,
      "num_tokens": 64785242.0,
      "step": 31970
    },
    {
      "entropy": 5.95025863647461,
      "epoch": 2.794284715546622,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004221491744629035,
      "loss": 5.236,
      "mean_token_accuracy": 0.20756880789995194,
      "num_tokens": 64795852.0,
      "step": 31975
    },
    {
      "entropy": 5.987711238861084,
      "epoch": 2.794721663899327,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042212559975221504,
      "loss": 5.1724,
      "mean_token_accuracy": 0.21679575443267823,
      "num_tokens": 64805982.0,
      "step": 31980
    },
    {
      "entropy": 5.979715156555176,
      "epoch": 2.795158612252032,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042210202221958055,
      "loss": 5.3027,
      "mean_token_accuracy": 0.19877956062555313,
      "num_tokens": 64816997.0,
      "step": 31985
    },
    {
      "entropy": 6.01060209274292,
      "epoch": 2.7955955606047365,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004220784418654521,
      "loss": 5.2585,
      "mean_token_accuracy": 0.20050954520702363,
      "num_tokens": 64826597.0,
      "step": 31990
    },
    {
      "entropy": 5.952537727355957,
      "epoch": 2.7960325089574414,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042205485869028207,
      "loss": 5.2298,
      "mean_token_accuracy": 0.2105687752366066,
      "num_tokens": 64835917.0,
      "step": 31995
    },
    {
      "entropy": 5.975584936141968,
      "epoch": 2.796469457310146,
      "grad_norm": 1.46875,
      "learning_rate": 0.00042203127269452284,
      "loss": 5.2382,
      "mean_token_accuracy": 0.20388528406620027,
      "num_tokens": 64846889.0,
      "step": 32000
    },
    {
      "entropy": 6.041873931884766,
      "epoch": 2.7969064056628508,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042200768387862676,
      "loss": 5.3335,
      "mean_token_accuracy": 0.20327913910150527,
      "num_tokens": 64857321.0,
      "step": 32005
    },
    {
      "entropy": 5.969651126861573,
      "epoch": 2.7973433540155552,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042198409224304624,
      "loss": 5.2903,
      "mean_token_accuracy": 0.2061345472931862,
      "num_tokens": 64867255.0,
      "step": 32010
    },
    {
      "entropy": 6.041647005081177,
      "epoch": 2.79778030236826,
      "grad_norm": 1.25,
      "learning_rate": 0.0004219604977882338,
      "loss": 5.2492,
      "mean_token_accuracy": 0.20474007725715637,
      "num_tokens": 64876179.0,
      "step": 32015
    },
    {
      "entropy": 6.061096334457398,
      "epoch": 2.798217250720965,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042193690051464206,
      "loss": 5.3452,
      "mean_token_accuracy": 0.20004326552152635,
      "num_tokens": 64886387.0,
      "step": 32020
    },
    {
      "entropy": 5.9827783584594725,
      "epoch": 2.7986541990736695,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042191330042272353,
      "loss": 5.2976,
      "mean_token_accuracy": 0.2029422789812088,
      "num_tokens": 64896816.0,
      "step": 32025
    },
    {
      "entropy": 5.940207672119141,
      "epoch": 2.799091147426374,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042188969751293093,
      "loss": 5.237,
      "mean_token_accuracy": 0.21271032243967056,
      "num_tokens": 64907747.0,
      "step": 32030
    },
    {
      "entropy": 5.983662414550781,
      "epoch": 2.799528095779079,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042186609178571696,
      "loss": 5.2925,
      "mean_token_accuracy": 0.19990204274654388,
      "num_tokens": 64918066.0,
      "step": 32035
    },
    {
      "entropy": 6.0385292053222654,
      "epoch": 2.799965044131784,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004218424832415344,
      "loss": 5.3425,
      "mean_token_accuracy": 0.19785641133785248,
      "num_tokens": 64927021.0,
      "step": 32040
    },
    {
      "entropy": 5.9555017948150635,
      "epoch": 2.8004019924844883,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042181887188083605,
      "loss": 5.1387,
      "mean_token_accuracy": 0.2180398717522621,
      "num_tokens": 64937477.0,
      "step": 32045
    },
    {
      "entropy": 5.933188533782959,
      "epoch": 2.800838940837193,
      "grad_norm": 1.25,
      "learning_rate": 0.00042179525770407486,
      "loss": 5.2546,
      "mean_token_accuracy": 0.2042672872543335,
      "num_tokens": 64947178.0,
      "step": 32050
    },
    {
      "entropy": 6.012243318557739,
      "epoch": 2.8012758891898977,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042177164071170373,
      "loss": 5.3101,
      "mean_token_accuracy": 0.19558767974376678,
      "num_tokens": 64956921.0,
      "step": 32055
    },
    {
      "entropy": 6.009689092636108,
      "epoch": 2.8017128375426026,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004217480209041755,
      "loss": 5.2138,
      "mean_token_accuracy": 0.20803750455379486,
      "num_tokens": 64966688.0,
      "step": 32060
    },
    {
      "entropy": 5.981153917312622,
      "epoch": 2.802149785895307,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004217243982819434,
      "loss": 5.2099,
      "mean_token_accuracy": 0.20872134268283843,
      "num_tokens": 64975577.0,
      "step": 32065
    },
    {
      "entropy": 5.9854902744293215,
      "epoch": 2.802586734248012,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004217007728454605,
      "loss": 5.2823,
      "mean_token_accuracy": 0.20360124856233597,
      "num_tokens": 64986121.0,
      "step": 32070
    },
    {
      "entropy": 6.056390571594238,
      "epoch": 2.8030236826007164,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004216771445951798,
      "loss": 5.306,
      "mean_token_accuracy": 0.2009967550635338,
      "num_tokens": 64996623.0,
      "step": 32075
    },
    {
      "entropy": 6.0066056728363035,
      "epoch": 2.8034606309534214,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042165351353155476,
      "loss": 5.2359,
      "mean_token_accuracy": 0.20263925641775132,
      "num_tokens": 65006756.0,
      "step": 32080
    },
    {
      "entropy": 5.980826044082642,
      "epoch": 2.803897579306126,
      "grad_norm": 1.25,
      "learning_rate": 0.00042162987965503834,
      "loss": 5.2765,
      "mean_token_accuracy": 0.20121686458587645,
      "num_tokens": 65017453.0,
      "step": 32085
    },
    {
      "entropy": 6.020394134521484,
      "epoch": 2.8043345276588307,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042160624296608403,
      "loss": 5.3018,
      "mean_token_accuracy": 0.19720491468906404,
      "num_tokens": 65028090.0,
      "step": 32090
    },
    {
      "entropy": 6.061325454711914,
      "epoch": 2.8047714760115356,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004215826034651452,
      "loss": 5.385,
      "mean_token_accuracy": 0.19145025163888932,
      "num_tokens": 65038422.0,
      "step": 32095
    },
    {
      "entropy": 6.014114141464233,
      "epoch": 2.80520842436424,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004215589611526752,
      "loss": 5.1798,
      "mean_token_accuracy": 0.21582283079624176,
      "num_tokens": 65048244.0,
      "step": 32100
    },
    {
      "entropy": 6.004331016540528,
      "epoch": 2.8056453727169446,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004215353160291275,
      "loss": 5.3253,
      "mean_token_accuracy": 0.20137124359607697,
      "num_tokens": 65058411.0,
      "step": 32105
    },
    {
      "entropy": 5.9478028297424315,
      "epoch": 2.8060823210696495,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004215116680949558,
      "loss": 5.2339,
      "mean_token_accuracy": 0.20616828054189681,
      "num_tokens": 65068148.0,
      "step": 32110
    },
    {
      "entropy": 6.032308626174927,
      "epoch": 2.8065192694223544,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004214880173506133,
      "loss": 5.2667,
      "mean_token_accuracy": 0.1994102567434311,
      "num_tokens": 65078212.0,
      "step": 32115
    },
    {
      "entropy": 6.000056886672974,
      "epoch": 2.806956217775059,
      "grad_norm": 1.34375,
      "learning_rate": 0.00042146436379655405,
      "loss": 5.2517,
      "mean_token_accuracy": 0.21047764271497726,
      "num_tokens": 65088133.0,
      "step": 32120
    },
    {
      "entropy": 6.007868051528931,
      "epoch": 2.807393166127764,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042144070743323155,
      "loss": 5.2506,
      "mean_token_accuracy": 0.2046002984046936,
      "num_tokens": 65097671.0,
      "step": 32125
    },
    {
      "entropy": 6.04415717124939,
      "epoch": 2.8078301144804683,
      "grad_norm": 1.25,
      "learning_rate": 0.0004214170482610995,
      "loss": 5.3614,
      "mean_token_accuracy": 0.19480465352535248,
      "num_tokens": 65107833.0,
      "step": 32130
    },
    {
      "entropy": 5.9715618133544925,
      "epoch": 2.808267062833173,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042139338628061177,
      "loss": 5.2361,
      "mean_token_accuracy": 0.20684728324413298,
      "num_tokens": 65118414.0,
      "step": 32135
    },
    {
      "entropy": 6.05613751411438,
      "epoch": 2.8087040111858776,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042136972149222203,
      "loss": 5.3559,
      "mean_token_accuracy": 0.19858960062265396,
      "num_tokens": 65128920.0,
      "step": 32140
    },
    {
      "entropy": 5.943717098236084,
      "epoch": 2.8091409595385826,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004213460538963845,
      "loss": 5.2322,
      "mean_token_accuracy": 0.20717114210128784,
      "num_tokens": 65138869.0,
      "step": 32145
    },
    {
      "entropy": 6.004160785675049,
      "epoch": 2.8095779078912875,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004213223834935529,
      "loss": 5.2854,
      "mean_token_accuracy": 0.20483725965023042,
      "num_tokens": 65148737.0,
      "step": 32150
    },
    {
      "entropy": 6.054381227493286,
      "epoch": 2.810014856243992,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042129871028418137,
      "loss": 5.3375,
      "mean_token_accuracy": 0.19816495180130006,
      "num_tokens": 65158075.0,
      "step": 32155
    },
    {
      "entropy": 6.070602178573608,
      "epoch": 2.8104518045966964,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004212750342687239,
      "loss": 5.357,
      "mean_token_accuracy": 0.19642313718795776,
      "num_tokens": 65167788.0,
      "step": 32160
    },
    {
      "entropy": 5.983587789535522,
      "epoch": 2.8108887529494013,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004212513554476346,
      "loss": 5.2977,
      "mean_token_accuracy": 0.20381184220314025,
      "num_tokens": 65177581.0,
      "step": 32165
    },
    {
      "entropy": 6.031342458724976,
      "epoch": 2.8113257013021062,
      "grad_norm": 1.25,
      "learning_rate": 0.00042122767382136766,
      "loss": 5.2551,
      "mean_token_accuracy": 0.22062018364667893,
      "num_tokens": 65186546.0,
      "step": 32170
    },
    {
      "entropy": 5.904427719116211,
      "epoch": 2.8117626496548107,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004212039893903774,
      "loss": 5.1859,
      "mean_token_accuracy": 0.22197202146053313,
      "num_tokens": 65195717.0,
      "step": 32175
    },
    {
      "entropy": 5.997617864608765,
      "epoch": 2.8121995980075156,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042118030215511793,
      "loss": 5.3116,
      "mean_token_accuracy": 0.20209171772003173,
      "num_tokens": 65206090.0,
      "step": 32180
    },
    {
      "entropy": 5.982721376419067,
      "epoch": 2.81263654636022,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042115661211604373,
      "loss": 5.3157,
      "mean_token_accuracy": 0.20374356657266618,
      "num_tokens": 65217194.0,
      "step": 32185
    },
    {
      "entropy": 6.1002589702606205,
      "epoch": 2.813073494712925,
      "grad_norm": 1.25,
      "learning_rate": 0.000421132919273609,
      "loss": 5.3439,
      "mean_token_accuracy": 0.2045903578400612,
      "num_tokens": 65228906.0,
      "step": 32190
    },
    {
      "entropy": 6.065336751937866,
      "epoch": 2.8135104430656295,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004211092236282685,
      "loss": 5.312,
      "mean_token_accuracy": 0.1994745045900345,
      "num_tokens": 65238606.0,
      "step": 32195
    },
    {
      "entropy": 5.9854407787323,
      "epoch": 2.8139473914183344,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042108552518047644,
      "loss": 5.2301,
      "mean_token_accuracy": 0.21144289672374725,
      "num_tokens": 65248259.0,
      "step": 32200
    },
    {
      "entropy": 6.007139492034912,
      "epoch": 2.8143843397710393,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004210618239306874,
      "loss": 5.3205,
      "mean_token_accuracy": 0.20012564361095428,
      "num_tokens": 65259294.0,
      "step": 32205
    },
    {
      "entropy": 6.109768486022949,
      "epoch": 2.8148212881237438,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004210381198793561,
      "loss": 5.3493,
      "mean_token_accuracy": 0.1958491250872612,
      "num_tokens": 65269498.0,
      "step": 32210
    },
    {
      "entropy": 6.068432712554932,
      "epoch": 2.8152582364764482,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004210144130269371,
      "loss": 5.3685,
      "mean_token_accuracy": 0.19408551156520842,
      "num_tokens": 65280027.0,
      "step": 32215
    },
    {
      "entropy": 5.982158470153808,
      "epoch": 2.815695184829153,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042099070337388517,
      "loss": 5.2833,
      "mean_token_accuracy": 0.1985692635178566,
      "num_tokens": 65290451.0,
      "step": 32220
    },
    {
      "entropy": 6.005788850784302,
      "epoch": 2.816132133181858,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042096699092065504,
      "loss": 5.2561,
      "mean_token_accuracy": 0.2027798905968666,
      "num_tokens": 65299906.0,
      "step": 32225
    },
    {
      "entropy": 6.015273332595825,
      "epoch": 2.8165690815345625,
      "grad_norm": 1.25,
      "learning_rate": 0.0004209432756677015,
      "loss": 5.2767,
      "mean_token_accuracy": 0.2030091866850853,
      "num_tokens": 65310197.0,
      "step": 32230
    },
    {
      "entropy": 5.997279691696167,
      "epoch": 2.8170060298872674,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042091955761547963,
      "loss": 5.3188,
      "mean_token_accuracy": 0.2040838658809662,
      "num_tokens": 65320779.0,
      "step": 32235
    },
    {
      "entropy": 5.916705513000489,
      "epoch": 2.817442978239972,
      "grad_norm": 1.125,
      "learning_rate": 0.000420895836764444,
      "loss": 5.195,
      "mean_token_accuracy": 0.21019417941570281,
      "num_tokens": 65330578.0,
      "step": 32240
    },
    {
      "entropy": 6.010801076889038,
      "epoch": 2.817879926592677,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042087211311504986,
      "loss": 5.2884,
      "mean_token_accuracy": 0.19619310051202773,
      "num_tokens": 65341202.0,
      "step": 32245
    },
    {
      "entropy": 6.039598512649536,
      "epoch": 2.8183168749453813,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004208483866677521,
      "loss": 5.2536,
      "mean_token_accuracy": 0.2117109015583992,
      "num_tokens": 65352584.0,
      "step": 32250
    },
    {
      "entropy": 6.055736351013183,
      "epoch": 2.818753823298086,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004208246574230059,
      "loss": 5.2304,
      "mean_token_accuracy": 0.2040405184030533,
      "num_tokens": 65363191.0,
      "step": 32255
    },
    {
      "entropy": 5.9583470821380615,
      "epoch": 2.819190771650791,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042080092538126633,
      "loss": 5.2601,
      "mean_token_accuracy": 0.2038148745894432,
      "num_tokens": 65373880.0,
      "step": 32260
    },
    {
      "entropy": 5.979863262176513,
      "epoch": 2.8196277200034956,
      "grad_norm": 1.3125,
      "learning_rate": 0.00042077719054298867,
      "loss": 5.3364,
      "mean_token_accuracy": 0.20224220156669617,
      "num_tokens": 65383149.0,
      "step": 32265
    },
    {
      "entropy": 5.997218084335327,
      "epoch": 2.8200646683562,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004207534529086281,
      "loss": 5.2296,
      "mean_token_accuracy": 0.2164700895547867,
      "num_tokens": 65393005.0,
      "step": 32270
    },
    {
      "entropy": 6.032480669021607,
      "epoch": 2.820501616708905,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042072971247863984,
      "loss": 5.2999,
      "mean_token_accuracy": 0.200030680000782,
      "num_tokens": 65402640.0,
      "step": 32275
    },
    {
      "entropy": 6.059476613998413,
      "epoch": 2.82093856506161,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004207059692534794,
      "loss": 5.3243,
      "mean_token_accuracy": 0.1967906415462494,
      "num_tokens": 65412863.0,
      "step": 32280
    },
    {
      "entropy": 5.959573364257812,
      "epoch": 2.8213755134143144,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042068222323360214,
      "loss": 5.3215,
      "mean_token_accuracy": 0.20016685873270035,
      "num_tokens": 65422581.0,
      "step": 32285
    },
    {
      "entropy": 6.017776155471802,
      "epoch": 2.8218124617670193,
      "grad_norm": 1.125,
      "learning_rate": 0.00042065847441946354,
      "loss": 5.2899,
      "mean_token_accuracy": 0.20505315363407134,
      "num_tokens": 65433604.0,
      "step": 32290
    },
    {
      "entropy": 5.980579996109009,
      "epoch": 2.8222494101197237,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000420634722811519,
      "loss": 5.2118,
      "mean_token_accuracy": 0.20811801850795747,
      "num_tokens": 65442956.0,
      "step": 32295
    },
    {
      "entropy": 6.046601963043213,
      "epoch": 2.8226863584724287,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042061096841022416,
      "loss": 5.2794,
      "mean_token_accuracy": 0.2020423471927643,
      "num_tokens": 65452441.0,
      "step": 32300
    },
    {
      "entropy": 6.034290170669555,
      "epoch": 2.823123306825133,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00042058721121603467,
      "loss": 5.2936,
      "mean_token_accuracy": 0.20820293426513672,
      "num_tokens": 65463628.0,
      "step": 32305
    },
    {
      "entropy": 5.9864805221557615,
      "epoch": 2.823560255177838,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004205634512294062,
      "loss": 5.2719,
      "mean_token_accuracy": 0.20615077167749404,
      "num_tokens": 65472771.0,
      "step": 32310
    },
    {
      "entropy": 6.030407094955445,
      "epoch": 2.823997203530543,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042053968845079436,
      "loss": 5.3216,
      "mean_token_accuracy": 0.19744263738393783,
      "num_tokens": 65482678.0,
      "step": 32315
    },
    {
      "entropy": 6.08578634262085,
      "epoch": 2.8244341518832474,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042051592288065514,
      "loss": 5.3151,
      "mean_token_accuracy": 0.19757640659809111,
      "num_tokens": 65493127.0,
      "step": 32320
    },
    {
      "entropy": 6.008351182937622,
      "epoch": 2.824871100235952,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004204921545194443,
      "loss": 5.3182,
      "mean_token_accuracy": 0.20401381999254226,
      "num_tokens": 65502508.0,
      "step": 32325
    },
    {
      "entropy": 6.016004180908203,
      "epoch": 2.825308048588657,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004204683833676176,
      "loss": 5.3246,
      "mean_token_accuracy": 0.20115489065647124,
      "num_tokens": 65512336.0,
      "step": 32330
    },
    {
      "entropy": 6.031457901000977,
      "epoch": 2.8257449969413617,
      "grad_norm": 1.25,
      "learning_rate": 0.0004204446094256312,
      "loss": 5.2617,
      "mean_token_accuracy": 0.19926098734140396,
      "num_tokens": 65523139.0,
      "step": 32335
    },
    {
      "entropy": 5.956889867782593,
      "epoch": 2.826181945294066,
      "grad_norm": 1.28125,
      "learning_rate": 0.00042042083269394083,
      "loss": 5.2537,
      "mean_token_accuracy": 0.20612742006778717,
      "num_tokens": 65532846.0,
      "step": 32340
    },
    {
      "entropy": 6.038541984558106,
      "epoch": 2.826618893646771,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004203970531730028,
      "loss": 5.2517,
      "mean_token_accuracy": 0.20068663209676743,
      "num_tokens": 65544493.0,
      "step": 32345
    },
    {
      "entropy": 6.055978775024414,
      "epoch": 2.8270558419994756,
      "grad_norm": 1.3125,
      "learning_rate": 0.00042037327086327304,
      "loss": 5.3533,
      "mean_token_accuracy": 0.2005913197994232,
      "num_tokens": 65554183.0,
      "step": 32350
    },
    {
      "entropy": 6.006471920013428,
      "epoch": 2.8274927903521805,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004203494857652078,
      "loss": 5.2565,
      "mean_token_accuracy": 0.2089239090681076,
      "num_tokens": 65564068.0,
      "step": 32355
    },
    {
      "entropy": 5.967813014984131,
      "epoch": 2.827929738704885,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042032569787926326,
      "loss": 5.2132,
      "mean_token_accuracy": 0.2094124212861061,
      "num_tokens": 65574272.0,
      "step": 32360
    },
    {
      "entropy": 5.9533874034881595,
      "epoch": 2.82836668705759,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042030190720589573,
      "loss": 5.2728,
      "mean_token_accuracy": 0.20578764826059343,
      "num_tokens": 65584890.0,
      "step": 32365
    },
    {
      "entropy": 6.028859281539917,
      "epoch": 2.8288036354102943,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004202781137455614,
      "loss": 5.2949,
      "mean_token_accuracy": 0.20458595603704452,
      "num_tokens": 65595776.0,
      "step": 32370
    },
    {
      "entropy": 5.991361427307129,
      "epoch": 2.8292405837629993,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004202543174987168,
      "loss": 5.2642,
      "mean_token_accuracy": 0.20692006051540374,
      "num_tokens": 65606234.0,
      "step": 32375
    },
    {
      "entropy": 6.0736000537872314,
      "epoch": 2.8296775321157037,
      "grad_norm": 1.34375,
      "learning_rate": 0.00042023051846581824,
      "loss": 5.4007,
      "mean_token_accuracy": 0.19740311652421952,
      "num_tokens": 65616576.0,
      "step": 32380
    },
    {
      "entropy": 5.991497230529785,
      "epoch": 2.8301144804684086,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004202067166473222,
      "loss": 5.2146,
      "mean_token_accuracy": 0.21413390040397645,
      "num_tokens": 65626792.0,
      "step": 32385
    },
    {
      "entropy": 6.021885871887207,
      "epoch": 2.8305514288211135,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004201829120436854,
      "loss": 5.2307,
      "mean_token_accuracy": 0.20672391057014466,
      "num_tokens": 65637041.0,
      "step": 32390
    },
    {
      "entropy": 5.941710186004639,
      "epoch": 2.830988377173818,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004201591046553641,
      "loss": 5.3333,
      "mean_token_accuracy": 0.2037088543176651,
      "num_tokens": 65647132.0,
      "step": 32395
    },
    {
      "entropy": 6.061775350570679,
      "epoch": 2.831425325526523,
      "grad_norm": 1.25,
      "learning_rate": 0.00042013529448281517,
      "loss": 5.3526,
      "mean_token_accuracy": 0.19709531217813492,
      "num_tokens": 65657372.0,
      "step": 32400
    },
    {
      "entropy": 6.048960494995117,
      "epoch": 2.8318622738792274,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004201114815264952,
      "loss": 5.2866,
      "mean_token_accuracy": 0.2105114072561264,
      "num_tokens": 65667118.0,
      "step": 32405
    },
    {
      "entropy": 6.041401290893555,
      "epoch": 2.8322992222319323,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000420087665786861,
      "loss": 5.3219,
      "mean_token_accuracy": 0.19997551888227463,
      "num_tokens": 65677830.0,
      "step": 32410
    },
    {
      "entropy": 5.926181316375732,
      "epoch": 2.832736170584637,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00042006384726436943,
      "loss": 5.268,
      "mean_token_accuracy": 0.2100610002875328,
      "num_tokens": 65687238.0,
      "step": 32415
    },
    {
      "entropy": 5.9444833278656,
      "epoch": 2.8331731189373417,
      "grad_norm": 1.328125,
      "learning_rate": 0.00042004002595947713,
      "loss": 5.2373,
      "mean_token_accuracy": 0.2105489954352379,
      "num_tokens": 65697547.0,
      "step": 32420
    },
    {
      "entropy": 5.9918313980102536,
      "epoch": 2.833610067290046,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004200162018726412,
      "loss": 5.3734,
      "mean_token_accuracy": 0.2052214562892914,
      "num_tokens": 65709272.0,
      "step": 32425
    },
    {
      "entropy": 6.025968265533447,
      "epoch": 2.834047015642751,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004199923750043185,
      "loss": 5.3288,
      "mean_token_accuracy": 0.20031205266714097,
      "num_tokens": 65720986.0,
      "step": 32430
    },
    {
      "entropy": 6.101785516738891,
      "epoch": 2.8344839639954555,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041996854535496616,
      "loss": 5.3095,
      "mean_token_accuracy": 0.20010914504528046,
      "num_tokens": 65731339.0,
      "step": 32435
    },
    {
      "entropy": 6.0736290454864506,
      "epoch": 2.8349209123481605,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004199447129250411,
      "loss": 5.3119,
      "mean_token_accuracy": 0.20525358319282533,
      "num_tokens": 65741826.0,
      "step": 32440
    },
    {
      "entropy": 6.01105489730835,
      "epoch": 2.8353578607008654,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004199208777150004,
      "loss": 5.3355,
      "mean_token_accuracy": 0.19597766697406768,
      "num_tokens": 65752925.0,
      "step": 32445
    },
    {
      "entropy": 6.063316822052002,
      "epoch": 2.83579480905357,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004198970397253015,
      "loss": 5.3106,
      "mean_token_accuracy": 0.1987650215625763,
      "num_tokens": 65763948.0,
      "step": 32450
    },
    {
      "entropy": 6.046033334732056,
      "epoch": 2.8362317574062743,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004198731989564014,
      "loss": 5.2864,
      "mean_token_accuracy": 0.20511841773986816,
      "num_tokens": 65774086.0,
      "step": 32455
    },
    {
      "entropy": 5.952186155319214,
      "epoch": 2.8366687057589792,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004198493554087574,
      "loss": 5.1801,
      "mean_token_accuracy": 0.2113691449165344,
      "num_tokens": 65783630.0,
      "step": 32460
    },
    {
      "entropy": 5.9993024349212645,
      "epoch": 2.837105654111684,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041982550908282686,
      "loss": 5.2824,
      "mean_token_accuracy": 0.20943133682012557,
      "num_tokens": 65793903.0,
      "step": 32465
    },
    {
      "entropy": 5.999702739715576,
      "epoch": 2.8375426024643886,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041980165997906714,
      "loss": 5.2168,
      "mean_token_accuracy": 0.21530838012695314,
      "num_tokens": 65803483.0,
      "step": 32470
    },
    {
      "entropy": 6.0097252368927006,
      "epoch": 2.8379795508170935,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004197778080979358,
      "loss": 5.2815,
      "mean_token_accuracy": 0.20575494915246964,
      "num_tokens": 65814336.0,
      "step": 32475
    },
    {
      "entropy": 6.000103855133057,
      "epoch": 2.838416499169798,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041975395343989017,
      "loss": 5.3282,
      "mean_token_accuracy": 0.19846330881118773,
      "num_tokens": 65825525.0,
      "step": 32480
    },
    {
      "entropy": 6.012115573883056,
      "epoch": 2.838853447522503,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004197300960053879,
      "loss": 5.2547,
      "mean_token_accuracy": 0.21389498859643935,
      "num_tokens": 65836276.0,
      "step": 32485
    },
    {
      "entropy": 6.047510433197021,
      "epoch": 2.8392903958752074,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004197062357948865,
      "loss": 5.3201,
      "mean_token_accuracy": 0.19699016809463502,
      "num_tokens": 65846164.0,
      "step": 32490
    },
    {
      "entropy": 6.037675762176514,
      "epoch": 2.8397273442279123,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041968237280884364,
      "loss": 5.2758,
      "mean_token_accuracy": 0.2032027766108513,
      "num_tokens": 65855759.0,
      "step": 32495
    },
    {
      "entropy": 5.925498676300049,
      "epoch": 2.840164292580617,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004196585070477172,
      "loss": 5.15,
      "mean_token_accuracy": 0.21798329949378967,
      "num_tokens": 65866009.0,
      "step": 32500
    },
    {
      "entropy": 6.002307796478272,
      "epoch": 2.8406012409333217,
      "grad_norm": 1.296875,
      "learning_rate": 0.00041963463851196466,
      "loss": 5.2435,
      "mean_token_accuracy": 0.20608574897050858,
      "num_tokens": 65875707.0,
      "step": 32505
    },
    {
      "entropy": 6.039607238769531,
      "epoch": 2.841038189286026,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041961076720204403,
      "loss": 5.2651,
      "mean_token_accuracy": 0.2106424018740654,
      "num_tokens": 65884396.0,
      "step": 32510
    },
    {
      "entropy": 6.020926141738892,
      "epoch": 2.841475137638731,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004195868931184131,
      "loss": 5.3069,
      "mean_token_accuracy": 0.20107924938201904,
      "num_tokens": 65894278.0,
      "step": 32515
    },
    {
      "entropy": 5.978949546813965,
      "epoch": 2.841912085991436,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041956301626152974,
      "loss": 5.2794,
      "mean_token_accuracy": 0.20765581876039504,
      "num_tokens": 65903863.0,
      "step": 32520
    },
    {
      "entropy": 5.972645902633667,
      "epoch": 2.8423490343441404,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041953913663185207,
      "loss": 5.1566,
      "mean_token_accuracy": 0.2141909420490265,
      "num_tokens": 65913996.0,
      "step": 32525
    },
    {
      "entropy": 5.92564115524292,
      "epoch": 2.8427859826968453,
      "grad_norm": 1.359375,
      "learning_rate": 0.000419515254229838,
      "loss": 5.2231,
      "mean_token_accuracy": 0.21735632717609404,
      "num_tokens": 65923457.0,
      "step": 32530
    },
    {
      "entropy": 6.045251893997192,
      "epoch": 2.84322293104955,
      "grad_norm": 1.3125,
      "learning_rate": 0.00041949136905594555,
      "loss": 5.2701,
      "mean_token_accuracy": 0.20515583008527755,
      "num_tokens": 65933132.0,
      "step": 32535
    },
    {
      "entropy": 6.016510963439941,
      "epoch": 2.8436598794022547,
      "grad_norm": 1.296875,
      "learning_rate": 0.000419467481110633,
      "loss": 5.2646,
      "mean_token_accuracy": 0.20853327810764313,
      "num_tokens": 65942784.0,
      "step": 32540
    },
    {
      "entropy": 5.970201587677002,
      "epoch": 2.844096827754959,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004194435903943585,
      "loss": 5.2171,
      "mean_token_accuracy": 0.2086588606238365,
      "num_tokens": 65952734.0,
      "step": 32545
    },
    {
      "entropy": 5.930238533020019,
      "epoch": 2.844533776107664,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041941969690758023,
      "loss": 5.1164,
      "mean_token_accuracy": 0.21986257135868073,
      "num_tokens": 65962732.0,
      "step": 32550
    },
    {
      "entropy": 6.006203889846802,
      "epoch": 2.844970724460369,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041939580065075645,
      "loss": 5.3196,
      "mean_token_accuracy": 0.210242760181427,
      "num_tokens": 65973287.0,
      "step": 32555
    },
    {
      "entropy": 6.018118667602539,
      "epoch": 2.8454076728130735,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004193719016243456,
      "loss": 5.2612,
      "mean_token_accuracy": 0.19899605810642243,
      "num_tokens": 65982034.0,
      "step": 32560
    },
    {
      "entropy": 6.003082990646362,
      "epoch": 2.845844621165778,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041934799982880594,
      "loss": 5.2529,
      "mean_token_accuracy": 0.2058311030268669,
      "num_tokens": 65992612.0,
      "step": 32565
    },
    {
      "entropy": 5.999074220657349,
      "epoch": 2.846281569518483,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041932409526459607,
      "loss": 5.2326,
      "mean_token_accuracy": 0.20729840695858,
      "num_tokens": 66002788.0,
      "step": 32570
    },
    {
      "entropy": 6.0286369800567625,
      "epoch": 2.846718517871188,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041930018793217443,
      "loss": 5.3064,
      "mean_token_accuracy": 0.20420503616333008,
      "num_tokens": 66012559.0,
      "step": 32575
    },
    {
      "entropy": 5.9782154083251955,
      "epoch": 2.8471554662238923,
      "grad_norm": 1.125,
      "learning_rate": 0.0004192762778319996,
      "loss": 5.3184,
      "mean_token_accuracy": 0.1956767588853836,
      "num_tokens": 66023230.0,
      "step": 32580
    },
    {
      "entropy": 6.0127969741821286,
      "epoch": 2.847592414576597,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041925236496453006,
      "loss": 5.3283,
      "mean_token_accuracy": 0.20516575425863265,
      "num_tokens": 66033556.0,
      "step": 32585
    },
    {
      "entropy": 6.08321623802185,
      "epoch": 2.8480293629293016,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00041922844933022467,
      "loss": 5.3176,
      "mean_token_accuracy": 0.19892172813415526,
      "num_tokens": 66042495.0,
      "step": 32590
    },
    {
      "entropy": 5.995156145095825,
      "epoch": 2.8484663112820066,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041920453092954204,
      "loss": 5.2618,
      "mean_token_accuracy": 0.20617987513542174,
      "num_tokens": 66051933.0,
      "step": 32595
    },
    {
      "entropy": 6.014125204086303,
      "epoch": 2.848903259634711,
      "grad_norm": 1.21875,
      "learning_rate": 0.000419180609762941,
      "loss": 5.2229,
      "mean_token_accuracy": 0.20924472510814668,
      "num_tokens": 66060964.0,
      "step": 32600
    },
    {
      "entropy": 6.0485039234161375,
      "epoch": 2.849340207987416,
      "grad_norm": 1.328125,
      "learning_rate": 0.00041915668583088016,
      "loss": 5.2708,
      "mean_token_accuracy": 0.20241478085517883,
      "num_tokens": 66070210.0,
      "step": 32605
    },
    {
      "entropy": 6.0490867614746096,
      "epoch": 2.849777156340121,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041913275913381864,
      "loss": 5.3277,
      "mean_token_accuracy": 0.2049719884991646,
      "num_tokens": 66081010.0,
      "step": 32610
    },
    {
      "entropy": 6.021329307556153,
      "epoch": 2.8502141046928253,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004191088296722153,
      "loss": 5.227,
      "mean_token_accuracy": 0.20677312910556794,
      "num_tokens": 66090448.0,
      "step": 32615
    },
    {
      "entropy": 6.051880502700806,
      "epoch": 2.85065105304553,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000419084897446529,
      "loss": 5.3222,
      "mean_token_accuracy": 0.19955717027187347,
      "num_tokens": 66100357.0,
      "step": 32620
    },
    {
      "entropy": 6.0054931640625,
      "epoch": 2.8510880013982347,
      "grad_norm": 1.25,
      "learning_rate": 0.000419060962457219,
      "loss": 5.2544,
      "mean_token_accuracy": 0.20606849044561387,
      "num_tokens": 66111068.0,
      "step": 32625
    },
    {
      "entropy": 6.0348835468292235,
      "epoch": 2.8515249497509396,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00041903702470474414,
      "loss": 5.2605,
      "mean_token_accuracy": 0.206336510181427,
      "num_tokens": 66122185.0,
      "step": 32630
    },
    {
      "entropy": 6.014275360107422,
      "epoch": 2.851961898103644,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004190130841895637,
      "loss": 5.2583,
      "mean_token_accuracy": 0.20306517332792282,
      "num_tokens": 66133192.0,
      "step": 32635
    },
    {
      "entropy": 6.0151602268219,
      "epoch": 2.852398846456349,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004189891409121369,
      "loss": 5.3285,
      "mean_token_accuracy": 0.20262817591428756,
      "num_tokens": 66143381.0,
      "step": 32640
    },
    {
      "entropy": 6.023988580703735,
      "epoch": 2.8528357948090535,
      "grad_norm": 1.25,
      "learning_rate": 0.00041896519487292285,
      "loss": 5.2775,
      "mean_token_accuracy": 0.20806159377098082,
      "num_tokens": 66152883.0,
      "step": 32645
    },
    {
      "entropy": 6.019995069503784,
      "epoch": 2.8532727431617584,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041894124607238104,
      "loss": 5.1997,
      "mean_token_accuracy": 0.2090981885790825,
      "num_tokens": 66162725.0,
      "step": 32650
    },
    {
      "entropy": 6.054768848419189,
      "epoch": 2.853709691514463,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004189172945109706,
      "loss": 5.3449,
      "mean_token_accuracy": 0.1989330142736435,
      "num_tokens": 66171478.0,
      "step": 32655
    },
    {
      "entropy": 5.923309850692749,
      "epoch": 2.8541466398671678,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004188933401891511,
      "loss": 5.167,
      "mean_token_accuracy": 0.21423859596252443,
      "num_tokens": 66181044.0,
      "step": 32660
    },
    {
      "entropy": 5.944503974914551,
      "epoch": 2.8545835882198727,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004188693831073818,
      "loss": 5.2734,
      "mean_token_accuracy": 0.20977447628974916,
      "num_tokens": 66190848.0,
      "step": 32665
    },
    {
      "entropy": 6.016620206832886,
      "epoch": 2.855020536572577,
      "grad_norm": 1.421875,
      "learning_rate": 0.00041884542326612245,
      "loss": 5.3147,
      "mean_token_accuracy": 0.19422310441732407,
      "num_tokens": 66200756.0,
      "step": 32670
    },
    {
      "entropy": 6.088240718841552,
      "epoch": 2.8554574849252816,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00041882146066583257,
      "loss": 5.3204,
      "mean_token_accuracy": 0.2005050227046013,
      "num_tokens": 66211463.0,
      "step": 32675
    },
    {
      "entropy": 6.099939298629761,
      "epoch": 2.8558944332779865,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041879749530697165,
      "loss": 5.3134,
      "mean_token_accuracy": 0.1993303641676903,
      "num_tokens": 66222522.0,
      "step": 32680
    },
    {
      "entropy": 6.055868625640869,
      "epoch": 2.8563313816306914,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004187735271899994,
      "loss": 5.332,
      "mean_token_accuracy": 0.20130694210529326,
      "num_tokens": 66232741.0,
      "step": 32685
    },
    {
      "entropy": 6.030843448638916,
      "epoch": 2.856768329983396,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041874955631537555,
      "loss": 5.2145,
      "mean_token_accuracy": 0.2056140571832657,
      "num_tokens": 66242845.0,
      "step": 32690
    },
    {
      "entropy": 5.967187070846558,
      "epoch": 2.857205278336101,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041872558268355993,
      "loss": 5.2207,
      "mean_token_accuracy": 0.2076162353157997,
      "num_tokens": 66252697.0,
      "step": 32695
    },
    {
      "entropy": 5.972565412521362,
      "epoch": 2.8576422266888053,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041870160629501224,
      "loss": 5.2938,
      "mean_token_accuracy": 0.19986281394958497,
      "num_tokens": 66263607.0,
      "step": 32700
    },
    {
      "entropy": 5.999087476730347,
      "epoch": 2.85807917504151,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041867762715019257,
      "loss": 5.222,
      "mean_token_accuracy": 0.2101820394396782,
      "num_tokens": 66274352.0,
      "step": 32705
    },
    {
      "entropy": 6.023151493072509,
      "epoch": 2.8585161233942147,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041865364524956066,
      "loss": 5.2883,
      "mean_token_accuracy": 0.2082671821117401,
      "num_tokens": 66283972.0,
      "step": 32710
    },
    {
      "entropy": 6.014147663116455,
      "epoch": 2.8589530717469196,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041862966059357665,
      "loss": 5.3241,
      "mean_token_accuracy": 0.20307697653770446,
      "num_tokens": 66295179.0,
      "step": 32715
    },
    {
      "entropy": 6.043640613555908,
      "epoch": 2.859390020099624,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004186056731827003,
      "loss": 5.3346,
      "mean_token_accuracy": 0.20053403526544572,
      "num_tokens": 66306065.0,
      "step": 32720
    },
    {
      "entropy": 6.0142964839935305,
      "epoch": 2.859826968452329,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000418581683017392,
      "loss": 5.241,
      "mean_token_accuracy": 0.20296860635280609,
      "num_tokens": 66315491.0,
      "step": 32725
    },
    {
      "entropy": 5.993583679199219,
      "epoch": 2.8602639168050334,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004185576900981119,
      "loss": 5.2228,
      "mean_token_accuracy": 0.2095915660262108,
      "num_tokens": 66324327.0,
      "step": 32730
    },
    {
      "entropy": 6.0136330127716064,
      "epoch": 2.8607008651577384,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004185336944253199,
      "loss": 5.2122,
      "mean_token_accuracy": 0.21118274480104446,
      "num_tokens": 66334201.0,
      "step": 32735
    },
    {
      "entropy": 6.043694639205933,
      "epoch": 2.8611378135104433,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041850969599947655,
      "loss": 5.2602,
      "mean_token_accuracy": 0.20932846516370773,
      "num_tokens": 66344745.0,
      "step": 32740
    },
    {
      "entropy": 5.941946649551392,
      "epoch": 2.8615747618631477,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041848569482104197,
      "loss": 5.2589,
      "mean_token_accuracy": 0.2101159244775772,
      "num_tokens": 66354815.0,
      "step": 32745
    },
    {
      "entropy": 6.019738101959229,
      "epoch": 2.862011710215852,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004184616908904765,
      "loss": 5.286,
      "mean_token_accuracy": 0.20133539587259291,
      "num_tokens": 66364744.0,
      "step": 32750
    },
    {
      "entropy": 6.012690591812134,
      "epoch": 2.862448658568557,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004184376842082408,
      "loss": 5.3331,
      "mean_token_accuracy": 0.20257275849580764,
      "num_tokens": 66376363.0,
      "step": 32755
    },
    {
      "entropy": 6.011279773712158,
      "epoch": 2.862885606921262,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00041841367477479503,
      "loss": 5.294,
      "mean_token_accuracy": 0.19836761206388473,
      "num_tokens": 66386149.0,
      "step": 32760
    },
    {
      "entropy": 6.0304762840271,
      "epoch": 2.8633225552739665,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004183896625905999,
      "loss": 5.2424,
      "mean_token_accuracy": 0.2069535046815872,
      "num_tokens": 66396874.0,
      "step": 32765
    },
    {
      "entropy": 6.018889093399048,
      "epoch": 2.8637595036266714,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041836564765611583,
      "loss": 5.2196,
      "mean_token_accuracy": 0.20635587722063065,
      "num_tokens": 66406761.0,
      "step": 32770
    },
    {
      "entropy": 6.034934186935425,
      "epoch": 2.864196451979376,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041834162997180364,
      "loss": 5.3567,
      "mean_token_accuracy": 0.2013703018426895,
      "num_tokens": 66416947.0,
      "step": 32775
    },
    {
      "entropy": 5.950884580612183,
      "epoch": 2.864633400332081,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004183176095381238,
      "loss": 5.2617,
      "mean_token_accuracy": 0.2083984926342964,
      "num_tokens": 66427980.0,
      "step": 32780
    },
    {
      "entropy": 5.9474372386932375,
      "epoch": 2.8650703486847853,
      "grad_norm": 1.25,
      "learning_rate": 0.00041829358635553707,
      "loss": 5.1567,
      "mean_token_accuracy": 0.21359120160341263,
      "num_tokens": 66437226.0,
      "step": 32785
    },
    {
      "entropy": 5.978408765792847,
      "epoch": 2.86550729703749,
      "grad_norm": 1.328125,
      "learning_rate": 0.00041826956042450436,
      "loss": 5.3222,
      "mean_token_accuracy": 0.20146464705467224,
      "num_tokens": 66447820.0,
      "step": 32790
    },
    {
      "entropy": 5.9895905494689945,
      "epoch": 2.865944245390195,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004182455317454863,
      "loss": 5.2807,
      "mean_token_accuracy": 0.1984696567058563,
      "num_tokens": 66459727.0,
      "step": 32795
    },
    {
      "entropy": 5.9978326797485355,
      "epoch": 2.8663811937428996,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041822150031894397,
      "loss": 5.2406,
      "mean_token_accuracy": 0.20002793818712233,
      "num_tokens": 66469743.0,
      "step": 32800
    },
    {
      "entropy": 6.0189924240112305,
      "epoch": 2.866818142095604,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00041819746614533824,
      "loss": 5.329,
      "mean_token_accuracy": 0.199495068192482,
      "num_tokens": 66479974.0,
      "step": 32805
    },
    {
      "entropy": 5.962000942230224,
      "epoch": 2.867255090448309,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041817342922513,
      "loss": 5.1897,
      "mean_token_accuracy": 0.21754330694675444,
      "num_tokens": 66489605.0,
      "step": 32810
    },
    {
      "entropy": 6.012280225753784,
      "epoch": 2.867692038801014,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00041814938955878035,
      "loss": 5.3033,
      "mean_token_accuracy": 0.20736389309167863,
      "num_tokens": 66499454.0,
      "step": 32815
    },
    {
      "entropy": 6.051541471481324,
      "epoch": 2.8681289871537183,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041812534714675046,
      "loss": 5.3205,
      "mean_token_accuracy": 0.20706312954425812,
      "num_tokens": 66510076.0,
      "step": 32820
    },
    {
      "entropy": 6.022870111465454,
      "epoch": 2.8685659355064232,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041810130198950137,
      "loss": 5.2957,
      "mean_token_accuracy": 0.20373703688383102,
      "num_tokens": 66520659.0,
      "step": 32825
    },
    {
      "entropy": 6.051510524749756,
      "epoch": 2.8690028838591277,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004180772540874943,
      "loss": 5.3067,
      "mean_token_accuracy": 0.2061956375837326,
      "num_tokens": 66531387.0,
      "step": 32830
    },
    {
      "entropy": 6.047219514846802,
      "epoch": 2.8694398322118326,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041805320344119064,
      "loss": 5.2441,
      "mean_token_accuracy": 0.20772063732147217,
      "num_tokens": 66540654.0,
      "step": 32835
    },
    {
      "entropy": 6.012439632415772,
      "epoch": 2.869876780564537,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004180291500510515,
      "loss": 5.2643,
      "mean_token_accuracy": 0.2078496128320694,
      "num_tokens": 66550966.0,
      "step": 32840
    },
    {
      "entropy": 6.045135831832885,
      "epoch": 2.870313728917242,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041800509391753834,
      "loss": 5.377,
      "mean_token_accuracy": 0.19685176610946656,
      "num_tokens": 66561550.0,
      "step": 32845
    },
    {
      "entropy": 6.005746173858642,
      "epoch": 2.870750677269947,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004179810350411125,
      "loss": 5.2382,
      "mean_token_accuracy": 0.21031394153833388,
      "num_tokens": 66571477.0,
      "step": 32850
    },
    {
      "entropy": 6.0245341777801515,
      "epoch": 2.8711876256226514,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004179569734222355,
      "loss": 5.296,
      "mean_token_accuracy": 0.2039724811911583,
      "num_tokens": 66581976.0,
      "step": 32855
    },
    {
      "entropy": 6.05712480545044,
      "epoch": 2.871624573975356,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041793290906136886,
      "loss": 5.2859,
      "mean_token_accuracy": 0.2026943936944008,
      "num_tokens": 66591747.0,
      "step": 32860
    },
    {
      "entropy": 6.07157678604126,
      "epoch": 2.872061522328061,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004179088419589741,
      "loss": 5.2585,
      "mean_token_accuracy": 0.19850103110074996,
      "num_tokens": 66602681.0,
      "step": 32865
    },
    {
      "entropy": 5.995519685745239,
      "epoch": 2.8724984706807657,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041788477211551295,
      "loss": 5.2952,
      "mean_token_accuracy": 0.2079764649271965,
      "num_tokens": 66612887.0,
      "step": 32870
    },
    {
      "entropy": 6.000078916549683,
      "epoch": 2.87293541903347,
      "grad_norm": 1.25,
      "learning_rate": 0.00041786069953144697,
      "loss": 5.223,
      "mean_token_accuracy": 0.21038988083600998,
      "num_tokens": 66622418.0,
      "step": 32875
    },
    {
      "entropy": 6.041340970993042,
      "epoch": 2.873372367386175,
      "grad_norm": 1.125,
      "learning_rate": 0.00041783662420723786,
      "loss": 5.3263,
      "mean_token_accuracy": 0.20266883224248886,
      "num_tokens": 66633457.0,
      "step": 32880
    },
    {
      "entropy": 6.024956607818604,
      "epoch": 2.8738093157388795,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004178125461433475,
      "loss": 5.2199,
      "mean_token_accuracy": 0.20601361095905305,
      "num_tokens": 66643691.0,
      "step": 32885
    },
    {
      "entropy": 6.0145121097564695,
      "epoch": 2.8742462640915845,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041778846534023763,
      "loss": 5.2685,
      "mean_token_accuracy": 0.2036000058054924,
      "num_tokens": 66653745.0,
      "step": 32890
    },
    {
      "entropy": 5.978254699707032,
      "epoch": 2.874683212444289,
      "grad_norm": 1.296875,
      "learning_rate": 0.00041776438179837026,
      "loss": 5.2494,
      "mean_token_accuracy": 0.21211353838443756,
      "num_tokens": 66664093.0,
      "step": 32895
    },
    {
      "entropy": 6.003677845001221,
      "epoch": 2.875120160796994,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004177402955182071,
      "loss": 5.3189,
      "mean_token_accuracy": 0.20184837877750397,
      "num_tokens": 66675204.0,
      "step": 32900
    },
    {
      "entropy": 6.071339893341064,
      "epoch": 2.8755571091496988,
      "grad_norm": 1.25,
      "learning_rate": 0.00041771620650021043,
      "loss": 5.3211,
      "mean_token_accuracy": 0.2071332886815071,
      "num_tokens": 66684789.0,
      "step": 32905
    },
    {
      "entropy": 6.066742086410523,
      "epoch": 2.8759940575024032,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041769211474484204,
      "loss": 5.3249,
      "mean_token_accuracy": 0.20239912569522858,
      "num_tokens": 66694734.0,
      "step": 32910
    },
    {
      "entropy": 6.034163331985473,
      "epoch": 2.8764310058551077,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004176680202525641,
      "loss": 5.3253,
      "mean_token_accuracy": 0.19531173408031463,
      "num_tokens": 66705231.0,
      "step": 32915
    },
    {
      "entropy": 5.964741468429565,
      "epoch": 2.8768679542078126,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00041764392302383893,
      "loss": 5.2563,
      "mean_token_accuracy": 0.20677836537361144,
      "num_tokens": 66715627.0,
      "step": 32920
    },
    {
      "entropy": 6.018080615997315,
      "epoch": 2.8773049025605175,
      "grad_norm": 1.25,
      "learning_rate": 0.0004176198230591284,
      "loss": 5.2455,
      "mean_token_accuracy": 0.2055909067392349,
      "num_tokens": 66725525.0,
      "step": 32925
    },
    {
      "entropy": 6.052545881271362,
      "epoch": 2.877741850913222,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000417595720358895,
      "loss": 5.3042,
      "mean_token_accuracy": 0.19653818905353546,
      "num_tokens": 66736352.0,
      "step": 32930
    },
    {
      "entropy": 6.04739408493042,
      "epoch": 2.878178799265927,
      "grad_norm": 1.296875,
      "learning_rate": 0.00041757161492360096,
      "loss": 5.3721,
      "mean_token_accuracy": 0.19527708739042282,
      "num_tokens": 66745961.0,
      "step": 32935
    },
    {
      "entropy": 6.003965139389038,
      "epoch": 2.8786157476186314,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041754750675370865,
      "loss": 5.2229,
      "mean_token_accuracy": 0.21083780229091645,
      "num_tokens": 66755295.0,
      "step": 32940
    },
    {
      "entropy": 5.995307064056396,
      "epoch": 2.8790526959713363,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004175233958496804,
      "loss": 5.2374,
      "mean_token_accuracy": 0.20514047741889954,
      "num_tokens": 66765089.0,
      "step": 32945
    },
    {
      "entropy": 6.036339998245239,
      "epoch": 2.8794896443240408,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004174992822119788,
      "loss": 5.2595,
      "mean_token_accuracy": 0.19907354861497878,
      "num_tokens": 66774995.0,
      "step": 32950
    },
    {
      "entropy": 6.013517093658447,
      "epoch": 2.8799265926767457,
      "grad_norm": 1.125,
      "learning_rate": 0.0004174751658410663,
      "loss": 5.2893,
      "mean_token_accuracy": 0.20796497613191606,
      "num_tokens": 66784584.0,
      "step": 32955
    },
    {
      "entropy": 5.994872188568115,
      "epoch": 2.8803635410294506,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041745104673740545,
      "loss": 5.2545,
      "mean_token_accuracy": 0.20418643057346345,
      "num_tokens": 66794747.0,
      "step": 32960
    },
    {
      "entropy": 6.06028995513916,
      "epoch": 2.880800489382155,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00041742692490145895,
      "loss": 5.2674,
      "mean_token_accuracy": 0.20409043729305268,
      "num_tokens": 66804632.0,
      "step": 32965
    },
    {
      "entropy": 5.963405752182007,
      "epoch": 2.8812374377348595,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004174028003336893,
      "loss": 5.2168,
      "mean_token_accuracy": 0.2035355731844902,
      "num_tokens": 66814534.0,
      "step": 32970
    },
    {
      "entropy": 5.983456373214722,
      "epoch": 2.8816743860875644,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004173786730345594,
      "loss": 5.3555,
      "mean_token_accuracy": 0.19544609040021896,
      "num_tokens": 66824629.0,
      "step": 32975
    },
    {
      "entropy": 6.034471321105957,
      "epoch": 2.8821113344402693,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004173545430045319,
      "loss": 5.3,
      "mean_token_accuracy": 0.20635754019021987,
      "num_tokens": 66834345.0,
      "step": 32980
    },
    {
      "entropy": 6.032412338256836,
      "epoch": 2.882548282792974,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041733041024406966,
      "loss": 5.2831,
      "mean_token_accuracy": 0.2019471362233162,
      "num_tokens": 66845272.0,
      "step": 32985
    },
    {
      "entropy": 6.071372699737549,
      "epoch": 2.8829852311456787,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004173062747536356,
      "loss": 5.2986,
      "mean_token_accuracy": 0.20434212535619736,
      "num_tokens": 66855022.0,
      "step": 32990
    },
    {
      "entropy": 6.004257249832153,
      "epoch": 2.883422179498383,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004172821365336926,
      "loss": 5.2099,
      "mean_token_accuracy": 0.20542179495096208,
      "num_tokens": 66864895.0,
      "step": 32995
    },
    {
      "entropy": 5.947495985031128,
      "epoch": 2.883859127851088,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004172579955847037,
      "loss": 5.2121,
      "mean_token_accuracy": 0.20906040519475938,
      "num_tokens": 66876134.0,
      "step": 33000
    },
    {
      "epoch": 2.883859127851088,
      "eval_entropy": 5.710412291025327,
      "eval_loss": 5.299934387207031,
      "eval_mean_token_accuracy": 0.21328549647652642,
      "eval_num_tokens": 66876134.0,
      "eval_runtime": 21.6213,
      "eval_samples_per_second": 1711.925,
      "eval_steps_per_second": 214.002,
      "step": 33000
    },
    {
      "entropy": 5.9968067646026615,
      "epoch": 2.8842960762037926,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004172338519071318,
      "loss": 5.2333,
      "mean_token_accuracy": 0.2055296167731285,
      "num_tokens": 66886089.0,
      "step": 33005
    },
    {
      "entropy": 5.872168397903442,
      "epoch": 2.8847330245564975,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041720970550144024,
      "loss": 5.163,
      "mean_token_accuracy": 0.21666219234466552,
      "num_tokens": 66896072.0,
      "step": 33010
    },
    {
      "entropy": 6.042267465591431,
      "epoch": 2.8851699729092024,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041718555636809194,
      "loss": 5.3771,
      "mean_token_accuracy": 0.1917472794651985,
      "num_tokens": 66906030.0,
      "step": 33015
    },
    {
      "entropy": 6.061028051376343,
      "epoch": 2.885606921261907,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004171614045075501,
      "loss": 5.2442,
      "mean_token_accuracy": 0.20356094241142272,
      "num_tokens": 66916714.0,
      "step": 33020
    },
    {
      "entropy": 6.007525205612183,
      "epoch": 2.8860438696146113,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004171372499202781,
      "loss": 5.3091,
      "mean_token_accuracy": 0.20284245312213897,
      "num_tokens": 66926570.0,
      "step": 33025
    },
    {
      "entropy": 6.062015342712402,
      "epoch": 2.8864808179673163,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004171130926067391,
      "loss": 5.3718,
      "mean_token_accuracy": 0.19931234717369078,
      "num_tokens": 66937565.0,
      "step": 33030
    },
    {
      "entropy": 6.042582893371582,
      "epoch": 2.886917766320021,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004170889325673967,
      "loss": 5.3258,
      "mean_token_accuracy": 0.20359238982200623,
      "num_tokens": 66948363.0,
      "step": 33035
    },
    {
      "entropy": 5.9300943374633786,
      "epoch": 2.8873547146727256,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041706476980271395,
      "loss": 5.2127,
      "mean_token_accuracy": 0.2107132628560066,
      "num_tokens": 66958312.0,
      "step": 33040
    },
    {
      "entropy": 5.988221597671509,
      "epoch": 2.8877916630254306,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041704060431315456,
      "loss": 5.1804,
      "mean_token_accuracy": 0.21377159655094147,
      "num_tokens": 66967880.0,
      "step": 33045
    },
    {
      "entropy": 5.9341065883636475,
      "epoch": 2.888228611378135,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041701643609918187,
      "loss": 5.1879,
      "mean_token_accuracy": 0.2135532483458519,
      "num_tokens": 66977930.0,
      "step": 33050
    },
    {
      "entropy": 6.002468395233154,
      "epoch": 2.88866555973084,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004169922651612596,
      "loss": 5.3209,
      "mean_token_accuracy": 0.20633059293031691,
      "num_tokens": 66987780.0,
      "step": 33055
    },
    {
      "entropy": 6.0878900527954105,
      "epoch": 2.8891025080835444,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041696809149985124,
      "loss": 5.2908,
      "mean_token_accuracy": 0.1980967938899994,
      "num_tokens": 66997910.0,
      "step": 33060
    },
    {
      "entropy": 6.021241950988769,
      "epoch": 2.8895394564362493,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041694391511542053,
      "loss": 5.2157,
      "mean_token_accuracy": 0.21308056712150575,
      "num_tokens": 67007365.0,
      "step": 33065
    },
    {
      "entropy": 5.933074474334717,
      "epoch": 2.889976404788954,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004169197360084311,
      "loss": 5.2022,
      "mean_token_accuracy": 0.2152461513876915,
      "num_tokens": 67016438.0,
      "step": 33070
    },
    {
      "entropy": 5.901926136016845,
      "epoch": 2.8904133531416587,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041689555417934675,
      "loss": 5.0899,
      "mean_token_accuracy": 0.22320399582386016,
      "num_tokens": 67026842.0,
      "step": 33075
    },
    {
      "entropy": 5.979216814041138,
      "epoch": 2.890850301494363,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041687136962863144,
      "loss": 5.3284,
      "mean_token_accuracy": 0.20300541073083878,
      "num_tokens": 67037343.0,
      "step": 33080
    },
    {
      "entropy": 6.01785979270935,
      "epoch": 2.891287249847068,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041684718235674887,
      "loss": 5.2765,
      "mean_token_accuracy": 0.19994231462478637,
      "num_tokens": 67047352.0,
      "step": 33085
    },
    {
      "entropy": 6.038596248626709,
      "epoch": 2.891724198199773,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004168229923641629,
      "loss": 5.3472,
      "mean_token_accuracy": 0.19190701246261596,
      "num_tokens": 67058717.0,
      "step": 33090
    },
    {
      "entropy": 6.054547882080078,
      "epoch": 2.8921611465524775,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004167987996513378,
      "loss": 5.2711,
      "mean_token_accuracy": 0.205708047747612,
      "num_tokens": 67068124.0,
      "step": 33095
    },
    {
      "entropy": 6.009286737442016,
      "epoch": 2.892598094905182,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004167746042187373,
      "loss": 5.3066,
      "mean_token_accuracy": 0.20303783118724822,
      "num_tokens": 67078906.0,
      "step": 33100
    },
    {
      "entropy": 5.937557220458984,
      "epoch": 2.893035043257887,
      "grad_norm": 1.8046875,
      "learning_rate": 0.00041675040606682564,
      "loss": 5.2441,
      "mean_token_accuracy": 0.20941648483276368,
      "num_tokens": 67090424.0,
      "step": 33105
    },
    {
      "entropy": 6.067427921295166,
      "epoch": 2.8934719916105918,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041672620519606687,
      "loss": 5.2925,
      "mean_token_accuracy": 0.2021653473377228,
      "num_tokens": 67100311.0,
      "step": 33110
    },
    {
      "entropy": 6.050157260894776,
      "epoch": 2.8939089399632962,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004167020016069253,
      "loss": 5.3035,
      "mean_token_accuracy": 0.20201400220394133,
      "num_tokens": 67109337.0,
      "step": 33115
    },
    {
      "entropy": 5.992065954208374,
      "epoch": 2.894345888316001,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041667779529986504,
      "loss": 5.2565,
      "mean_token_accuracy": 0.19867612719535827,
      "num_tokens": 67119795.0,
      "step": 33120
    },
    {
      "entropy": 5.982805013656616,
      "epoch": 2.8947828366687056,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041665358627535043,
      "loss": 5.2883,
      "mean_token_accuracy": 0.19841933697462083,
      "num_tokens": 67129444.0,
      "step": 33125
    },
    {
      "entropy": 6.039349603652954,
      "epoch": 2.8952197850214105,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004166293745338458,
      "loss": 5.2947,
      "mean_token_accuracy": 0.20937032550573348,
      "num_tokens": 67138868.0,
      "step": 33130
    },
    {
      "entropy": 6.061404275894165,
      "epoch": 2.895656733374115,
      "grad_norm": 1.25,
      "learning_rate": 0.0004166051600758156,
      "loss": 5.3496,
      "mean_token_accuracy": 0.19816589653491973,
      "num_tokens": 67149615.0,
      "step": 33135
    },
    {
      "entropy": 5.974964189529419,
      "epoch": 2.89609368172682,
      "grad_norm": 1.125,
      "learning_rate": 0.00041658094290172415,
      "loss": 5.2872,
      "mean_token_accuracy": 0.20818630158901213,
      "num_tokens": 67159745.0,
      "step": 33140
    },
    {
      "entropy": 6.054484462738037,
      "epoch": 2.896530630079525,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004165567230120362,
      "loss": 5.2348,
      "mean_token_accuracy": 0.20226828157901763,
      "num_tokens": 67169900.0,
      "step": 33145
    },
    {
      "entropy": 5.957871150970459,
      "epoch": 2.8969675784322293,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004165325004072159,
      "loss": 5.2761,
      "mean_token_accuracy": 0.20971225053071976,
      "num_tokens": 67180752.0,
      "step": 33150
    },
    {
      "entropy": 5.995830965042114,
      "epoch": 2.8974045267849338,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041650827508772826,
      "loss": 5.3117,
      "mean_token_accuracy": 0.20402443557977676,
      "num_tokens": 67191466.0,
      "step": 33155
    },
    {
      "entropy": 5.993710041046143,
      "epoch": 2.8978414751376387,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041648404705403765,
      "loss": 5.2459,
      "mean_token_accuracy": 0.20072269439697266,
      "num_tokens": 67202813.0,
      "step": 33160
    },
    {
      "entropy": 6.009892225265503,
      "epoch": 2.8982784234903436,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041645981630660894,
      "loss": 5.1832,
      "mean_token_accuracy": 0.21337340772151947,
      "num_tokens": 67213686.0,
      "step": 33165
    },
    {
      "entropy": 6.024523115158081,
      "epoch": 2.898715371843048,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004164355828459068,
      "loss": 5.2852,
      "mean_token_accuracy": 0.20156981348991393,
      "num_tokens": 67224955.0,
      "step": 33170
    },
    {
      "entropy": 5.9702949047088625,
      "epoch": 2.899152320195753,
      "grad_norm": 1.28125,
      "learning_rate": 0.000416411346672396,
      "loss": 5.2555,
      "mean_token_accuracy": 0.20161638110876084,
      "num_tokens": 67234438.0,
      "step": 33175
    },
    {
      "entropy": 6.021363210678101,
      "epoch": 2.8995892685484574,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041638710778654157,
      "loss": 5.3193,
      "mean_token_accuracy": 0.1987764224410057,
      "num_tokens": 67245151.0,
      "step": 33180
    },
    {
      "entropy": 6.012841081619262,
      "epoch": 2.9000262169011624,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004163628661888083,
      "loss": 5.2759,
      "mean_token_accuracy": 0.20579211115837098,
      "num_tokens": 67254942.0,
      "step": 33185
    },
    {
      "entropy": 6.0805881977081295,
      "epoch": 2.900463165253867,
      "grad_norm": 1.34375,
      "learning_rate": 0.00041633862187966125,
      "loss": 5.2989,
      "mean_token_accuracy": 0.20214019864797592,
      "num_tokens": 67264435.0,
      "step": 33190
    },
    {
      "entropy": 5.98910002708435,
      "epoch": 2.9009001136065717,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004163143748595653,
      "loss": 5.2478,
      "mean_token_accuracy": 0.20529916733503342,
      "num_tokens": 67273304.0,
      "step": 33195
    },
    {
      "entropy": 5.990701341629029,
      "epoch": 2.9013370619592767,
      "grad_norm": 1.296875,
      "learning_rate": 0.00041629012512898556,
      "loss": 5.3377,
      "mean_token_accuracy": 0.19726879447698592,
      "num_tokens": 67283713.0,
      "step": 33200
    },
    {
      "entropy": 6.016347217559814,
      "epoch": 2.901774010311981,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004162658726883873,
      "loss": 5.1608,
      "mean_token_accuracy": 0.21736272275447846,
      "num_tokens": 67292420.0,
      "step": 33205
    },
    {
      "entropy": 6.029789972305298,
      "epoch": 2.9022109586646856,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004162416175382354,
      "loss": 5.263,
      "mean_token_accuracy": 0.2088572710752487,
      "num_tokens": 67302138.0,
      "step": 33210
    },
    {
      "entropy": 5.981328392028809,
      "epoch": 2.9026479070173905,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004162173596789954,
      "loss": 5.2642,
      "mean_token_accuracy": 0.20926377773284913,
      "num_tokens": 67311776.0,
      "step": 33215
    },
    {
      "entropy": 6.011218786239624,
      "epoch": 2.9030848553700954,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004161930991111325,
      "loss": 5.2976,
      "mean_token_accuracy": 0.2033473491668701,
      "num_tokens": 67322296.0,
      "step": 33220
    },
    {
      "entropy": 6.0286918640136715,
      "epoch": 2.9035218037228,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004161688358351118,
      "loss": 5.2395,
      "mean_token_accuracy": 0.20261939465999604,
      "num_tokens": 67332173.0,
      "step": 33225
    },
    {
      "entropy": 6.023821830749512,
      "epoch": 2.903958752075505,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000416144569851399,
      "loss": 5.2585,
      "mean_token_accuracy": 0.20949587225914001,
      "num_tokens": 67342156.0,
      "step": 33230
    },
    {
      "entropy": 5.991478061676025,
      "epoch": 2.9043957004282093,
      "grad_norm": 1.25,
      "learning_rate": 0.00041612030116045933,
      "loss": 5.3007,
      "mean_token_accuracy": 0.20266005098819734,
      "num_tokens": 67353208.0,
      "step": 33235
    },
    {
      "entropy": 5.998099994659424,
      "epoch": 2.904832648780914,
      "grad_norm": 1.3125,
      "learning_rate": 0.00041609602976275836,
      "loss": 5.2148,
      "mean_token_accuracy": 0.21421699970960617,
      "num_tokens": 67362577.0,
      "step": 33240
    },
    {
      "entropy": 6.011343908309937,
      "epoch": 2.9052695971336187,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004160717556587616,
      "loss": 5.1896,
      "mean_token_accuracy": 0.21144066154956817,
      "num_tokens": 67371996.0,
      "step": 33245
    },
    {
      "entropy": 5.933028984069824,
      "epoch": 2.9057065454863236,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004160474788489346,
      "loss": 5.2207,
      "mean_token_accuracy": 0.21410810947418213,
      "num_tokens": 67381714.0,
      "step": 33250
    },
    {
      "entropy": 5.977512502670288,
      "epoch": 2.9061434938390285,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00041602319933374305,
      "loss": 5.2394,
      "mean_token_accuracy": 0.21463295966386794,
      "num_tokens": 67391551.0,
      "step": 33255
    },
    {
      "entropy": 6.035882139205933,
      "epoch": 2.906580442191733,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004159989171136527,
      "loss": 5.2729,
      "mean_token_accuracy": 0.19876564741134645,
      "num_tokens": 67400816.0,
      "step": 33260
    },
    {
      "entropy": 6.001716661453247,
      "epoch": 2.9070173905444374,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004159746321891292,
      "loss": 5.2683,
      "mean_token_accuracy": 0.20592522025108337,
      "num_tokens": 67410633.0,
      "step": 33265
    },
    {
      "entropy": 6.001225185394287,
      "epoch": 2.9074543388971423,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004159503445606385,
      "loss": 5.2684,
      "mean_token_accuracy": 0.2103530690073967,
      "num_tokens": 67420588.0,
      "step": 33270
    },
    {
      "entropy": 6.0290709972381595,
      "epoch": 2.9078912872498472,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041592605422864615,
      "loss": 5.2576,
      "mean_token_accuracy": 0.20651746094226836,
      "num_tokens": 67430494.0,
      "step": 33275
    },
    {
      "entropy": 5.922484302520752,
      "epoch": 2.9083282356025517,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004159017611936183,
      "loss": 5.2042,
      "mean_token_accuracy": 0.21678190380334855,
      "num_tokens": 67440692.0,
      "step": 33280
    },
    {
      "entropy": 6.014947509765625,
      "epoch": 2.9087651839552566,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041587746545602094,
      "loss": 5.2611,
      "mean_token_accuracy": 0.20387590527534485,
      "num_tokens": 67450936.0,
      "step": 33285
    },
    {
      "entropy": 5.998129653930664,
      "epoch": 2.909202132307961,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004158531670163199,
      "loss": 5.2365,
      "mean_token_accuracy": 0.21297802925109863,
      "num_tokens": 67460997.0,
      "step": 33290
    },
    {
      "entropy": 5.962801313400268,
      "epoch": 2.909639080660666,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004158288658749813,
      "loss": 5.2961,
      "mean_token_accuracy": 0.20730030834674834,
      "num_tokens": 67470685.0,
      "step": 33295
    },
    {
      "entropy": 5.9782813549041744,
      "epoch": 2.9100760290133705,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041580456203247127,
      "loss": 5.2502,
      "mean_token_accuracy": 0.21010536849498748,
      "num_tokens": 67481734.0,
      "step": 33300
    },
    {
      "entropy": 5.8781991481781,
      "epoch": 2.9105129773660754,
      "grad_norm": 1.21875,
      "learning_rate": 0.000415780255489256,
      "loss": 5.0694,
      "mean_token_accuracy": 0.2281798779964447,
      "num_tokens": 67491290.0,
      "step": 33305
    },
    {
      "entropy": 6.001952743530273,
      "epoch": 2.9109499257187803,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004157559462458016,
      "loss": 5.3193,
      "mean_token_accuracy": 0.2041889637708664,
      "num_tokens": 67501758.0,
      "step": 33310
    },
    {
      "entropy": 6.001793909072876,
      "epoch": 2.9113868740714848,
      "grad_norm": 1.328125,
      "learning_rate": 0.00041573163430257435,
      "loss": 5.2804,
      "mean_token_accuracy": 0.20418193340301513,
      "num_tokens": 67511435.0,
      "step": 33315
    },
    {
      "entropy": 6.025907516479492,
      "epoch": 2.9118238224241892,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041570731966004065,
      "loss": 5.3155,
      "mean_token_accuracy": 0.20011704415082932,
      "num_tokens": 67521408.0,
      "step": 33320
    },
    {
      "entropy": 6.029778051376343,
      "epoch": 2.912260770776894,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004156830023186669,
      "loss": 5.3436,
      "mean_token_accuracy": 0.2001993790268898,
      "num_tokens": 67531631.0,
      "step": 33325
    },
    {
      "entropy": 6.048289251327515,
      "epoch": 2.912697719129599,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004156586822789194,
      "loss": 5.313,
      "mean_token_accuracy": 0.1913161411881447,
      "num_tokens": 67542265.0,
      "step": 33330
    },
    {
      "entropy": 6.006251430511474,
      "epoch": 2.9131346674823035,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041563435954126467,
      "loss": 5.3085,
      "mean_token_accuracy": 0.20221371352672576,
      "num_tokens": 67553664.0,
      "step": 33335
    },
    {
      "entropy": 6.000982618331909,
      "epoch": 2.9135716158350085,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004156100341061693,
      "loss": 5.2249,
      "mean_token_accuracy": 0.20414482206106185,
      "num_tokens": 67564500.0,
      "step": 33340
    },
    {
      "entropy": 5.954046154022217,
      "epoch": 2.914008564187713,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004155857059740997,
      "loss": 5.1933,
      "mean_token_accuracy": 0.21279112696647645,
      "num_tokens": 67573611.0,
      "step": 33345
    },
    {
      "entropy": 5.961113357543946,
      "epoch": 2.914445512540418,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004155613751455226,
      "loss": 5.2329,
      "mean_token_accuracy": 0.2145821899175644,
      "num_tokens": 67583721.0,
      "step": 33350
    },
    {
      "entropy": 6.016874647140503,
      "epoch": 2.9148824608931223,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00041553704162090475,
      "loss": 5.2865,
      "mean_token_accuracy": 0.20897402614355087,
      "num_tokens": 67593381.0,
      "step": 33355
    },
    {
      "entropy": 6.0290509223937985,
      "epoch": 2.915319409245827,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004155127054007127,
      "loss": 5.2787,
      "mean_token_accuracy": 0.2038653686642647,
      "num_tokens": 67603799.0,
      "step": 33360
    },
    {
      "entropy": 6.0469568252563475,
      "epoch": 2.9157563575985317,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041548836648541346,
      "loss": 5.2979,
      "mean_token_accuracy": 0.20378226637840272,
      "num_tokens": 67613543.0,
      "step": 33365
    },
    {
      "entropy": 6.023906183242798,
      "epoch": 2.9161933059512366,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00041546402487547363,
      "loss": 5.2892,
      "mean_token_accuracy": 0.20547119230031968,
      "num_tokens": 67623302.0,
      "step": 33370
    },
    {
      "entropy": 5.995791053771972,
      "epoch": 2.916630254303941,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004154396805713602,
      "loss": 5.177,
      "mean_token_accuracy": 0.2198203518986702,
      "num_tokens": 67633655.0,
      "step": 33375
    },
    {
      "entropy": 5.965502977371216,
      "epoch": 2.917067202656646,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041541533357354013,
      "loss": 5.2943,
      "mean_token_accuracy": 0.2068890869617462,
      "num_tokens": 67644832.0,
      "step": 33380
    },
    {
      "entropy": 6.066838598251342,
      "epoch": 2.917504151009351,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004153909838824804,
      "loss": 5.2647,
      "mean_token_accuracy": 0.20715331733226777,
      "num_tokens": 67655290.0,
      "step": 33385
    },
    {
      "entropy": 6.003138828277588,
      "epoch": 2.9179410993620554,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000415366631498648,
      "loss": 5.207,
      "mean_token_accuracy": 0.2087895542383194,
      "num_tokens": 67664252.0,
      "step": 33390
    },
    {
      "entropy": 5.992795467376709,
      "epoch": 2.9183780477147603,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004153422764225101,
      "loss": 5.267,
      "mean_token_accuracy": 0.2057493358850479,
      "num_tokens": 67674655.0,
      "step": 33395
    },
    {
      "entropy": 5.959468412399292,
      "epoch": 2.9188149960674648,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004153179186545338,
      "loss": 5.1951,
      "mean_token_accuracy": 0.2109319195151329,
      "num_tokens": 67685126.0,
      "step": 33400
    },
    {
      "entropy": 5.942548990249634,
      "epoch": 2.9192519444201697,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041529355819518626,
      "loss": 5.2249,
      "mean_token_accuracy": 0.20752532482147218,
      "num_tokens": 67694111.0,
      "step": 33405
    },
    {
      "entropy": 6.038792562484741,
      "epoch": 2.919688892772874,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041526919504493475,
      "loss": 5.2849,
      "mean_token_accuracy": 0.20720969885587692,
      "num_tokens": 67704229.0,
      "step": 33410
    },
    {
      "entropy": 6.016286420822143,
      "epoch": 2.920125841125579,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00041524482920424666,
      "loss": 5.2794,
      "mean_token_accuracy": 0.2090599298477173,
      "num_tokens": 67714625.0,
      "step": 33415
    },
    {
      "entropy": 5.992089796066284,
      "epoch": 2.9205627894782835,
      "grad_norm": 1.328125,
      "learning_rate": 0.00041522046067358917,
      "loss": 5.2855,
      "mean_token_accuracy": 0.20004281401634216,
      "num_tokens": 67724094.0,
      "step": 33420
    },
    {
      "entropy": 5.971501398086548,
      "epoch": 2.9209997378309884,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041519608945342985,
      "loss": 5.2825,
      "mean_token_accuracy": 0.20857605934143067,
      "num_tokens": 67733233.0,
      "step": 33425
    },
    {
      "entropy": 5.983538246154785,
      "epoch": 2.921436686183693,
      "grad_norm": 1.25,
      "learning_rate": 0.000415171715544236,
      "loss": 5.202,
      "mean_token_accuracy": 0.212835656106472,
      "num_tokens": 67742781.0,
      "step": 33430
    },
    {
      "entropy": 6.000315809249878,
      "epoch": 2.921873634536398,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00041514733894647526,
      "loss": 5.2207,
      "mean_token_accuracy": 0.21088994592428206,
      "num_tokens": 67753166.0,
      "step": 33435
    },
    {
      "entropy": 6.026533412933349,
      "epoch": 2.9223105828891027,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004151229596606151,
      "loss": 5.3277,
      "mean_token_accuracy": 0.20184507966041565,
      "num_tokens": 67762599.0,
      "step": 33440
    },
    {
      "entropy": 5.981282329559326,
      "epoch": 2.922747531241807,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004150985776871231,
      "loss": 5.2476,
      "mean_token_accuracy": 0.20724800527095794,
      "num_tokens": 67772620.0,
      "step": 33445
    },
    {
      "entropy": 6.044471549987793,
      "epoch": 2.9231844795945117,
      "grad_norm": 1.21875,
      "learning_rate": 0.000415074193026467,
      "loss": 5.3104,
      "mean_token_accuracy": 0.20445686727762222,
      "num_tokens": 67782962.0,
      "step": 33450
    },
    {
      "entropy": 6.051120710372925,
      "epoch": 2.9236214279472166,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00041504980567911453,
      "loss": 5.2921,
      "mean_token_accuracy": 0.20303470641374588,
      "num_tokens": 67793168.0,
      "step": 33455
    },
    {
      "entropy": 6.015960645675659,
      "epoch": 2.9240583762999215,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041502541564553336,
      "loss": 5.3142,
      "mean_token_accuracy": 0.2048271968960762,
      "num_tokens": 67803317.0,
      "step": 33460
    },
    {
      "entropy": 6.000437498092651,
      "epoch": 2.924495324652626,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041500102292619135,
      "loss": 5.2509,
      "mean_token_accuracy": 0.2006200760602951,
      "num_tokens": 67812594.0,
      "step": 33465
    },
    {
      "entropy": 6.024224805831909,
      "epoch": 2.924932273005331,
      "grad_norm": 1.359375,
      "learning_rate": 0.00041497662752155637,
      "loss": 5.2282,
      "mean_token_accuracy": 0.2103202149271965,
      "num_tokens": 67822277.0,
      "step": 33470
    },
    {
      "entropy": 6.024864912033081,
      "epoch": 2.9253692213580353,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004149522294320963,
      "loss": 5.2794,
      "mean_token_accuracy": 0.20494387447834014,
      "num_tokens": 67832948.0,
      "step": 33475
    },
    {
      "entropy": 5.957875442504883,
      "epoch": 2.9258061697107403,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041492782865827914,
      "loss": 5.2033,
      "mean_token_accuracy": 0.20819002985954285,
      "num_tokens": 67843965.0,
      "step": 33480
    },
    {
      "entropy": 5.926089191436768,
      "epoch": 2.9262431180634447,
      "grad_norm": 1.328125,
      "learning_rate": 0.00041490342520057293,
      "loss": 5.2697,
      "mean_token_accuracy": 0.20950830578804017,
      "num_tokens": 67854079.0,
      "step": 33485
    },
    {
      "entropy": 6.012676763534546,
      "epoch": 2.9266800664161496,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041487901905944575,
      "loss": 5.2508,
      "mean_token_accuracy": 0.21112090051174165,
      "num_tokens": 67865218.0,
      "step": 33490
    },
    {
      "entropy": 6.039904499053955,
      "epoch": 2.9271170147688546,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041485461023536564,
      "loss": 5.1991,
      "mean_token_accuracy": 0.2173859566450119,
      "num_tokens": 67875229.0,
      "step": 33495
    },
    {
      "entropy": 6.021799230575562,
      "epoch": 2.927553963121559,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004148301987288008,
      "loss": 5.3364,
      "mean_token_accuracy": 0.20534197688102723,
      "num_tokens": 67885795.0,
      "step": 33500
    },
    {
      "entropy": 5.9713624000549315,
      "epoch": 2.9279909114742635,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004148057845402195,
      "loss": 5.2465,
      "mean_token_accuracy": 0.21090805530548096,
      "num_tokens": 67896668.0,
      "step": 33505
    },
    {
      "entropy": 5.9723726272583,
      "epoch": 2.9284278598269684,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041478136767009004,
      "loss": 5.208,
      "mean_token_accuracy": 0.2172871395945549,
      "num_tokens": 67907082.0,
      "step": 33510
    },
    {
      "entropy": 6.0331559658050535,
      "epoch": 2.9288648081796733,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004147569481188806,
      "loss": 5.2908,
      "mean_token_accuracy": 0.20605139136314393,
      "num_tokens": 67917517.0,
      "step": 33515
    },
    {
      "entropy": 6.100778579711914,
      "epoch": 2.929301756532378,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041473252588705977,
      "loss": 5.3361,
      "mean_token_accuracy": 0.2000483825802803,
      "num_tokens": 67927332.0,
      "step": 33520
    },
    {
      "entropy": 6.013118839263916,
      "epoch": 2.9297387048850827,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00041470810097509587,
      "loss": 5.2793,
      "mean_token_accuracy": 0.2019883617758751,
      "num_tokens": 67936939.0,
      "step": 33525
    },
    {
      "entropy": 5.954524326324463,
      "epoch": 2.930175653237787,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004146836733834573,
      "loss": 5.267,
      "mean_token_accuracy": 0.21201395541429519,
      "num_tokens": 67947553.0,
      "step": 33530
    },
    {
      "entropy": 5.962177371978759,
      "epoch": 2.930612601590492,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004146592431126128,
      "loss": 5.2246,
      "mean_token_accuracy": 0.20260761678218842,
      "num_tokens": 67958489.0,
      "step": 33535
    },
    {
      "entropy": 6.016440725326538,
      "epoch": 2.9310495499431966,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004146348101630307,
      "loss": 5.2008,
      "mean_token_accuracy": 0.20920076668262483,
      "num_tokens": 67967305.0,
      "step": 33540
    },
    {
      "entropy": 6.005506229400635,
      "epoch": 2.9314864982959015,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004146103745351799,
      "loss": 5.2332,
      "mean_token_accuracy": 0.20222820043563844,
      "num_tokens": 67977400.0,
      "step": 33545
    },
    {
      "entropy": 5.958978223800659,
      "epoch": 2.9319234466486064,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004145859362295289,
      "loss": 5.2461,
      "mean_token_accuracy": 0.20619310885667802,
      "num_tokens": 67987275.0,
      "step": 33550
    },
    {
      "entropy": 6.080948257446289,
      "epoch": 2.932360395001311,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004145614952465465,
      "loss": 5.4066,
      "mean_token_accuracy": 0.19584145694971083,
      "num_tokens": 67999002.0,
      "step": 33555
    },
    {
      "entropy": 6.0550047874450685,
      "epoch": 2.9327973433540153,
      "grad_norm": 1.34375,
      "learning_rate": 0.00041453705158670147,
      "loss": 5.2346,
      "mean_token_accuracy": 0.21160557121038437,
      "num_tokens": 68008504.0,
      "step": 33560
    },
    {
      "entropy": 6.024871921539306,
      "epoch": 2.9332342917067202,
      "grad_norm": 1.25,
      "learning_rate": 0.00041451260525046267,
      "loss": 5.2475,
      "mean_token_accuracy": 0.20517239570617676,
      "num_tokens": 68019522.0,
      "step": 33565
    },
    {
      "entropy": 5.986284065246582,
      "epoch": 2.933671240059425,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041448815623829907,
      "loss": 5.2459,
      "mean_token_accuracy": 0.2067238748073578,
      "num_tokens": 68029991.0,
      "step": 33570
    },
    {
      "entropy": 5.963617563247681,
      "epoch": 2.9341081884121296,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004144637045506795,
      "loss": 5.3074,
      "mean_token_accuracy": 0.19901185482740402,
      "num_tokens": 68040347.0,
      "step": 33575
    },
    {
      "entropy": 6.011569833755493,
      "epoch": 2.9345451367648345,
      "grad_norm": 1.140625,
      "learning_rate": 0.000414439250188073,
      "loss": 5.2395,
      "mean_token_accuracy": 0.21394555866718293,
      "num_tokens": 68050932.0,
      "step": 33580
    },
    {
      "entropy": 6.117492055892944,
      "epoch": 2.934982085117539,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004144147931509486,
      "loss": 5.2693,
      "mean_token_accuracy": 0.20255480855703353,
      "num_tokens": 68061104.0,
      "step": 33585
    },
    {
      "entropy": 5.981932592391968,
      "epoch": 2.935419033470244,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00041439033343977546,
      "loss": 5.2509,
      "mean_token_accuracy": 0.21334815621376038,
      "num_tokens": 68070374.0,
      "step": 33590
    },
    {
      "entropy": 5.96186637878418,
      "epoch": 2.9358559818229484,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041436587105502263,
      "loss": 5.2749,
      "mean_token_accuracy": 0.20125075280666352,
      "num_tokens": 68079996.0,
      "step": 33595
    },
    {
      "entropy": 5.99584264755249,
      "epoch": 2.9362929301756533,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004143414059971594,
      "loss": 5.2869,
      "mean_token_accuracy": 0.19910017848014833,
      "num_tokens": 68090322.0,
      "step": 33600
    },
    {
      "entropy": 5.977317523956299,
      "epoch": 2.936729878528358,
      "grad_norm": 1.3125,
      "learning_rate": 0.000414316938266655,
      "loss": 5.1566,
      "mean_token_accuracy": 0.20555950552225113,
      "num_tokens": 68099482.0,
      "step": 33605
    },
    {
      "entropy": 5.971925401687622,
      "epoch": 2.9371668268810627,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004142924678639787,
      "loss": 5.2506,
      "mean_token_accuracy": 0.2074048027396202,
      "num_tokens": 68109852.0,
      "step": 33610
    },
    {
      "entropy": 5.991778755187989,
      "epoch": 2.937603775233767,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041426799478959993,
      "loss": 5.2699,
      "mean_token_accuracy": 0.20352532267570494,
      "num_tokens": 68118721.0,
      "step": 33615
    },
    {
      "entropy": 6.035291576385498,
      "epoch": 2.938040723586472,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004142435190439879,
      "loss": 5.2102,
      "mean_token_accuracy": 0.21176542788743974,
      "num_tokens": 68129080.0,
      "step": 33620
    },
    {
      "entropy": 6.037090539932251,
      "epoch": 2.938477671939177,
      "grad_norm": 1.3125,
      "learning_rate": 0.00041421904062761234,
      "loss": 5.2933,
      "mean_token_accuracy": 0.20362513065338134,
      "num_tokens": 68138154.0,
      "step": 33625
    },
    {
      "entropy": 6.014033937454224,
      "epoch": 2.9389146202918814,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004141945595409426,
      "loss": 5.3635,
      "mean_token_accuracy": 0.19716764688491822,
      "num_tokens": 68148736.0,
      "step": 33630
    },
    {
      "entropy": 5.956044673919678,
      "epoch": 2.9393515686445864,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004141700757844483,
      "loss": 5.2289,
      "mean_token_accuracy": 0.20761973410844803,
      "num_tokens": 68159041.0,
      "step": 33635
    },
    {
      "entropy": 6.024954175949096,
      "epoch": 2.939788516997291,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000414145589358599,
      "loss": 5.2813,
      "mean_token_accuracy": 0.21195410937070847,
      "num_tokens": 68168715.0,
      "step": 33640
    },
    {
      "entropy": 5.990097236633301,
      "epoch": 2.9402254653499957,
      "grad_norm": 1.25,
      "learning_rate": 0.00041412110026386436,
      "loss": 5.2766,
      "mean_token_accuracy": 0.205586276948452,
      "num_tokens": 68178863.0,
      "step": 33645
    },
    {
      "entropy": 6.040685224533081,
      "epoch": 2.9406624137027,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004140966085007142,
      "loss": 5.309,
      "mean_token_accuracy": 0.1999921351671219,
      "num_tokens": 68190131.0,
      "step": 33650
    },
    {
      "entropy": 5.980509853363037,
      "epoch": 2.941099362055405,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041407211406961815,
      "loss": 5.2325,
      "mean_token_accuracy": 0.208311028778553,
      "num_tokens": 68200193.0,
      "step": 33655
    },
    {
      "entropy": 5.986698389053345,
      "epoch": 2.94153631040811,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041404761697104605,
      "loss": 5.2455,
      "mean_token_accuracy": 0.2061616450548172,
      "num_tokens": 68210714.0,
      "step": 33660
    },
    {
      "entropy": 6.0216896533966064,
      "epoch": 2.9419732587608145,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004140231172054678,
      "loss": 5.2865,
      "mean_token_accuracy": 0.21045396625995635,
      "num_tokens": 68220809.0,
      "step": 33665
    },
    {
      "entropy": 6.062749767303467,
      "epoch": 2.942410207113519,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004139986147733534,
      "loss": 5.2902,
      "mean_token_accuracy": 0.20597892850637436,
      "num_tokens": 68229887.0,
      "step": 33670
    },
    {
      "entropy": 6.004016017913818,
      "epoch": 2.942847155466224,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041397410967517265,
      "loss": 5.239,
      "mean_token_accuracy": 0.20707114934921264,
      "num_tokens": 68240491.0,
      "step": 33675
    },
    {
      "entropy": 5.9951395988464355,
      "epoch": 2.943284103818929,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041394960191139566,
      "loss": 5.3467,
      "mean_token_accuracy": 0.20011776834726333,
      "num_tokens": 68250302.0,
      "step": 33680
    },
    {
      "entropy": 6.0107879638671875,
      "epoch": 2.9437210521716333,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004139250914824925,
      "loss": 5.1803,
      "mean_token_accuracy": 0.21092199087142943,
      "num_tokens": 68259245.0,
      "step": 33685
    },
    {
      "entropy": 5.98463077545166,
      "epoch": 2.944158000524338,
      "grad_norm": 1.359375,
      "learning_rate": 0.00041390057838893335,
      "loss": 5.2086,
      "mean_token_accuracy": 0.21218458712100982,
      "num_tokens": 68269096.0,
      "step": 33690
    },
    {
      "entropy": 6.034923362731933,
      "epoch": 2.9445949488770427,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041387606263118825,
      "loss": 5.2624,
      "mean_token_accuracy": 0.20834018141031266,
      "num_tokens": 68278949.0,
      "step": 33695
    },
    {
      "entropy": 5.970423841476441,
      "epoch": 2.9450318972297476,
      "grad_norm": 1.375,
      "learning_rate": 0.0004138515442097275,
      "loss": 5.1925,
      "mean_token_accuracy": 0.20920852422714234,
      "num_tokens": 68288526.0,
      "step": 33700
    },
    {
      "entropy": 5.987569046020508,
      "epoch": 2.945468845582452,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004138270231250214,
      "loss": 5.2028,
      "mean_token_accuracy": 0.2063024416565895,
      "num_tokens": 68297997.0,
      "step": 33705
    },
    {
      "entropy": 5.971465587615967,
      "epoch": 2.945905793935157,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004138024993775403,
      "loss": 5.3244,
      "mean_token_accuracy": 0.20451106429100036,
      "num_tokens": 68307316.0,
      "step": 33710
    },
    {
      "entropy": 5.927141809463501,
      "epoch": 2.9463427422878614,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004137779729677544,
      "loss": 5.0863,
      "mean_token_accuracy": 0.22664965987205504,
      "num_tokens": 68317124.0,
      "step": 33715
    },
    {
      "entropy": 5.979611396789551,
      "epoch": 2.9467796906405663,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004137534438961344,
      "loss": 5.2603,
      "mean_token_accuracy": 0.20574887245893478,
      "num_tokens": 68326576.0,
      "step": 33720
    },
    {
      "entropy": 5.989595222473144,
      "epoch": 2.947216638993271,
      "grad_norm": 1.328125,
      "learning_rate": 0.00041372891216315055,
      "loss": 5.2265,
      "mean_token_accuracy": 0.21331822872161865,
      "num_tokens": 68336999.0,
      "step": 33725
    },
    {
      "entropy": 6.056910514831543,
      "epoch": 2.9476535873459757,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041370437776927347,
      "loss": 5.2464,
      "mean_token_accuracy": 0.21057202965021132,
      "num_tokens": 68347084.0,
      "step": 33730
    },
    {
      "entropy": 5.974510288238525,
      "epoch": 2.9480905356986806,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004136798407149738,
      "loss": 5.2359,
      "mean_token_accuracy": 0.20549483150243758,
      "num_tokens": 68357136.0,
      "step": 33735
    },
    {
      "entropy": 6.0333404541015625,
      "epoch": 2.948527484051385,
      "grad_norm": 1.296875,
      "learning_rate": 0.00041365530100072215,
      "loss": 5.3348,
      "mean_token_accuracy": 0.19943817108869552,
      "num_tokens": 68367020.0,
      "step": 33740
    },
    {
      "entropy": 6.015858793258667,
      "epoch": 2.9489644324040896,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004136307586269891,
      "loss": 5.2285,
      "mean_token_accuracy": 0.2137375757098198,
      "num_tokens": 68378436.0,
      "step": 33745
    },
    {
      "entropy": 6.002866172790528,
      "epoch": 2.9494013807567945,
      "grad_norm": 1.328125,
      "learning_rate": 0.00041360621359424544,
      "loss": 5.2067,
      "mean_token_accuracy": 0.20889655500650406,
      "num_tokens": 68387777.0,
      "step": 33750
    },
    {
      "entropy": 5.966463470458985,
      "epoch": 2.9498383291094994,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041358166590296205,
      "loss": 5.2439,
      "mean_token_accuracy": 0.20423608124256135,
      "num_tokens": 68397648.0,
      "step": 33755
    },
    {
      "entropy": 6.000646781921387,
      "epoch": 2.950275277462204,
      "grad_norm": 1.25,
      "learning_rate": 0.00041355711555360965,
      "loss": 5.3884,
      "mean_token_accuracy": 0.19753096997737885,
      "num_tokens": 68409539.0,
      "step": 33760
    },
    {
      "entropy": 6.031875419616699,
      "epoch": 2.9507122258149088,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041353256254665914,
      "loss": 5.2992,
      "mean_token_accuracy": 0.20105440616607667,
      "num_tokens": 68419626.0,
      "step": 33765
    },
    {
      "entropy": 6.0214732646942135,
      "epoch": 2.9511491741676132,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041350800688258157,
      "loss": 5.2307,
      "mean_token_accuracy": 0.21319443583488465,
      "num_tokens": 68429289.0,
      "step": 33770
    },
    {
      "entropy": 6.0356330394744875,
      "epoch": 2.951586122520318,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041348344856184785,
      "loss": 5.2769,
      "mean_token_accuracy": 0.20982512533664704,
      "num_tokens": 68440186.0,
      "step": 33775
    },
    {
      "entropy": 6.018475198745728,
      "epoch": 2.9520230708730226,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041345888758492896,
      "loss": 5.2073,
      "mean_token_accuracy": 0.21306423097848892,
      "num_tokens": 68450292.0,
      "step": 33780
    },
    {
      "entropy": 6.039930963516236,
      "epoch": 2.9524600192257275,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041343432395229605,
      "loss": 5.2761,
      "mean_token_accuracy": 0.20501280277967454,
      "num_tokens": 68460419.0,
      "step": 33785
    },
    {
      "entropy": 5.964353370666504,
      "epoch": 2.9528969675784325,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004134097576644203,
      "loss": 5.2443,
      "mean_token_accuracy": 0.21088800579309464,
      "num_tokens": 68470488.0,
      "step": 33790
    },
    {
      "entropy": 5.994617652893067,
      "epoch": 2.953333915931137,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041338518872177286,
      "loss": 5.2626,
      "mean_token_accuracy": 0.2072037249803543,
      "num_tokens": 68479872.0,
      "step": 33795
    },
    {
      "entropy": 6.046322870254516,
      "epoch": 2.9537708642838414,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041336061712482497,
      "loss": 5.317,
      "mean_token_accuracy": 0.1992254838347435,
      "num_tokens": 68491086.0,
      "step": 33800
    },
    {
      "entropy": 6.087268924713134,
      "epoch": 2.9542078126365463,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00041333604287404805,
      "loss": 5.3434,
      "mean_token_accuracy": 0.19161796271800996,
      "num_tokens": 68501136.0,
      "step": 33805
    },
    {
      "entropy": 5.957855749130249,
      "epoch": 2.954644760989251,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004133114659699132,
      "loss": 5.2158,
      "mean_token_accuracy": 0.20944413989782334,
      "num_tokens": 68511579.0,
      "step": 33810
    },
    {
      "entropy": 5.982857418060303,
      "epoch": 2.9550817093419557,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041328688641289205,
      "loss": 5.2413,
      "mean_token_accuracy": 0.21045560538768768,
      "num_tokens": 68521716.0,
      "step": 33815
    },
    {
      "entropy": 6.044832944869995,
      "epoch": 2.9555186576946606,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041326230420345584,
      "loss": 5.2885,
      "mean_token_accuracy": 0.20730263739824295,
      "num_tokens": 68531814.0,
      "step": 33820
    },
    {
      "entropy": 5.940910387039184,
      "epoch": 2.955955606047365,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004132377193420763,
      "loss": 5.1881,
      "mean_token_accuracy": 0.21761136502027512,
      "num_tokens": 68541388.0,
      "step": 33825
    },
    {
      "entropy": 6.00835862159729,
      "epoch": 2.95639255440007,
      "grad_norm": 1.328125,
      "learning_rate": 0.00041321313182922484,
      "loss": 5.3781,
      "mean_token_accuracy": 0.19503850489854813,
      "num_tokens": 68552259.0,
      "step": 33830
    },
    {
      "entropy": 6.0339867115020756,
      "epoch": 2.9568295027527745,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004131885416653731,
      "loss": 5.2338,
      "mean_token_accuracy": 0.21050545573234558,
      "num_tokens": 68562001.0,
      "step": 33835
    },
    {
      "entropy": 6.055719423294067,
      "epoch": 2.9572664511054794,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004131639488509927,
      "loss": 5.2381,
      "mean_token_accuracy": 0.2116050511598587,
      "num_tokens": 68571636.0,
      "step": 33840
    },
    {
      "entropy": 6.040849065780639,
      "epoch": 2.9577033994581843,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004131393533865554,
      "loss": 5.3896,
      "mean_token_accuracy": 0.19506694972515107,
      "num_tokens": 68581448.0,
      "step": 33845
    },
    {
      "entropy": 6.077304220199585,
      "epoch": 2.9581403478108887,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041311475527253284,
      "loss": 5.3675,
      "mean_token_accuracy": 0.19775112420320512,
      "num_tokens": 68591880.0,
      "step": 33850
    },
    {
      "entropy": 6.080837726593018,
      "epoch": 2.958577296163593,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041309015450939694,
      "loss": 5.3371,
      "mean_token_accuracy": 0.20149960666894912,
      "num_tokens": 68603025.0,
      "step": 33855
    },
    {
      "entropy": 6.065456914901733,
      "epoch": 2.959014244516298,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004130655510976196,
      "loss": 5.3384,
      "mean_token_accuracy": 0.20070160925388336,
      "num_tokens": 68613580.0,
      "step": 33860
    },
    {
      "entropy": 6.059202575683594,
      "epoch": 2.959451192869003,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041304094503767256,
      "loss": 5.3364,
      "mean_token_accuracy": 0.20418660789728166,
      "num_tokens": 68624451.0,
      "step": 33865
    },
    {
      "entropy": 5.974252319335937,
      "epoch": 2.9598881412217075,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00041301633633002784,
      "loss": 5.2678,
      "mean_token_accuracy": 0.2052317038178444,
      "num_tokens": 68634408.0,
      "step": 33870
    },
    {
      "entropy": 5.967637968063355,
      "epoch": 2.9603250895744124,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004129917249751575,
      "loss": 5.2356,
      "mean_token_accuracy": 0.20699929296970368,
      "num_tokens": 68643795.0,
      "step": 33875
    },
    {
      "entropy": 6.010975313186646,
      "epoch": 2.960762037927117,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004129671109735336,
      "loss": 5.2195,
      "mean_token_accuracy": 0.20747653692960738,
      "num_tokens": 68653669.0,
      "step": 33880
    },
    {
      "entropy": 6.084629487991333,
      "epoch": 2.961198986279822,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041294249432562825,
      "loss": 5.309,
      "mean_token_accuracy": 0.20234962850809096,
      "num_tokens": 68663094.0,
      "step": 33885
    },
    {
      "entropy": 5.949597358703613,
      "epoch": 2.9616359346325263,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00041291787503191347,
      "loss": 5.2134,
      "mean_token_accuracy": 0.21122609078884125,
      "num_tokens": 68673561.0,
      "step": 33890
    },
    {
      "entropy": 6.021357250213623,
      "epoch": 2.962072882985231,
      "grad_norm": 1.296875,
      "learning_rate": 0.00041289325309286167,
      "loss": 5.3647,
      "mean_token_accuracy": 0.18820168673992158,
      "num_tokens": 68683624.0,
      "step": 33895
    },
    {
      "entropy": 6.021961784362793,
      "epoch": 2.962509831337936,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041286862850894497,
      "loss": 5.2799,
      "mean_token_accuracy": 0.21197775900363922,
      "num_tokens": 68693805.0,
      "step": 33900
    },
    {
      "entropy": 6.050325584411621,
      "epoch": 2.9629467796906406,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004128440012806357,
      "loss": 5.3044,
      "mean_token_accuracy": 0.208068411052227,
      "num_tokens": 68704020.0,
      "step": 33905
    },
    {
      "entropy": 6.004342317581177,
      "epoch": 2.963383728043345,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041281937140840634,
      "loss": 5.2709,
      "mean_token_accuracy": 0.19828408658504487,
      "num_tokens": 68714486.0,
      "step": 33910
    },
    {
      "entropy": 6.013730478286743,
      "epoch": 2.96382067639605,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004127947388927292,
      "loss": 5.225,
      "mean_token_accuracy": 0.20857815742492675,
      "num_tokens": 68724759.0,
      "step": 33915
    },
    {
      "entropy": 5.962645149230957,
      "epoch": 2.964257624748755,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004127701037340767,
      "loss": 5.2004,
      "mean_token_accuracy": 0.21483092308044432,
      "num_tokens": 68734826.0,
      "step": 33920
    },
    {
      "entropy": 5.999151992797851,
      "epoch": 2.9646945731014593,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041274546593292143,
      "loss": 5.2944,
      "mean_token_accuracy": 0.20264507681131363,
      "num_tokens": 68744786.0,
      "step": 33925
    },
    {
      "entropy": 6.053039455413819,
      "epoch": 2.9651315214541643,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004127208254897361,
      "loss": 5.3223,
      "mean_token_accuracy": 0.19674134999513626,
      "num_tokens": 68754751.0,
      "step": 33930
    },
    {
      "entropy": 6.048620414733887,
      "epoch": 2.9655684698068687,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041269618240499296,
      "loss": 5.2636,
      "mean_token_accuracy": 0.20303929150104522,
      "num_tokens": 68765652.0,
      "step": 33935
    },
    {
      "entropy": 6.039805173873901,
      "epoch": 2.9660054181595736,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041267153667916506,
      "loss": 5.2358,
      "mean_token_accuracy": 0.2126783937215805,
      "num_tokens": 68775587.0,
      "step": 33940
    },
    {
      "entropy": 6.039570283889771,
      "epoch": 2.966442366512278,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00041264688831272483,
      "loss": 5.2669,
      "mean_token_accuracy": 0.20692069232463836,
      "num_tokens": 68786678.0,
      "step": 33945
    },
    {
      "entropy": 5.9481010913848875,
      "epoch": 2.966879314864983,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004126222373061452,
      "loss": 5.1375,
      "mean_token_accuracy": 0.21757036596536636,
      "num_tokens": 68796108.0,
      "step": 33950
    },
    {
      "entropy": 5.948995685577392,
      "epoch": 2.967316263217688,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041259758365989896,
      "loss": 5.2363,
      "mean_token_accuracy": 0.2102041110396385,
      "num_tokens": 68806364.0,
      "step": 33955
    },
    {
      "entropy": 6.013642454147339,
      "epoch": 2.9677532115703924,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004125729273744589,
      "loss": 5.2909,
      "mean_token_accuracy": 0.20331107527017594,
      "num_tokens": 68816099.0,
      "step": 33960
    },
    {
      "entropy": 6.019688558578491,
      "epoch": 2.968190159923097,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041254826845029806,
      "loss": 5.3205,
      "mean_token_accuracy": 0.19717129468917846,
      "num_tokens": 68826480.0,
      "step": 33965
    },
    {
      "entropy": 5.986487579345703,
      "epoch": 2.968627108275802,
      "grad_norm": 1.296875,
      "learning_rate": 0.00041252360688788937,
      "loss": 5.2487,
      "mean_token_accuracy": 0.20860457867383958,
      "num_tokens": 68835657.0,
      "step": 33970
    },
    {
      "entropy": 6.0174211978912355,
      "epoch": 2.9690640566285067,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004124989426877058,
      "loss": 5.301,
      "mean_token_accuracy": 0.20654910951852798,
      "num_tokens": 68845115.0,
      "step": 33975
    },
    {
      "entropy": 6.0158556461334225,
      "epoch": 2.969501004981211,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004124742758502205,
      "loss": 5.2666,
      "mean_token_accuracy": 0.20842056721448898,
      "num_tokens": 68854562.0,
      "step": 33980
    },
    {
      "entropy": 6.065128469467163,
      "epoch": 2.969937953333916,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004124496063759066,
      "loss": 5.3297,
      "mean_token_accuracy": 0.19694943875074386,
      "num_tokens": 68864787.0,
      "step": 33985
    },
    {
      "entropy": 5.96388611793518,
      "epoch": 2.9703749016866206,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004124249342652371,
      "loss": 5.2418,
      "mean_token_accuracy": 0.20519611686468125,
      "num_tokens": 68874518.0,
      "step": 33990
    },
    {
      "entropy": 6.040971374511718,
      "epoch": 2.9708118500393255,
      "grad_norm": 1.40625,
      "learning_rate": 0.00041240025951868545,
      "loss": 5.2392,
      "mean_token_accuracy": 0.19265319854021073,
      "num_tokens": 68883765.0,
      "step": 33995
    },
    {
      "entropy": 6.0502190589904785,
      "epoch": 2.97124879839203,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041237558213672477,
      "loss": 5.2712,
      "mean_token_accuracy": 0.2075868010520935,
      "num_tokens": 68894191.0,
      "step": 34000
    },
    {
      "entropy": 6.029387140274048,
      "epoch": 2.971685746744735,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041235090211982846,
      "loss": 5.2557,
      "mean_token_accuracy": 0.2041458711028099,
      "num_tokens": 68905136.0,
      "step": 34005
    },
    {
      "entropy": 5.997659492492676,
      "epoch": 2.9721226950974398,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004123262194684699,
      "loss": 5.2603,
      "mean_token_accuracy": 0.20905065089464187,
      "num_tokens": 68914620.0,
      "step": 34010
    },
    {
      "entropy": 6.03222885131836,
      "epoch": 2.9725596434501442,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00041230153418312244,
      "loss": 5.2434,
      "mean_token_accuracy": 0.2088375359773636,
      "num_tokens": 68923781.0,
      "step": 34015
    },
    {
      "entropy": 5.976896476745606,
      "epoch": 2.9729965918028487,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004122768462642596,
      "loss": 5.3031,
      "mean_token_accuracy": 0.20338049978017808,
      "num_tokens": 68934283.0,
      "step": 34020
    },
    {
      "entropy": 5.980893802642822,
      "epoch": 2.9734335401555536,
      "grad_norm": 1.234375,
      "learning_rate": 0.000412252155712355,
      "loss": 5.2161,
      "mean_token_accuracy": 0.21238497495651246,
      "num_tokens": 68943799.0,
      "step": 34025
    },
    {
      "entropy": 6.068899917602539,
      "epoch": 2.9738704885082585,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041222746252788213,
      "loss": 5.3032,
      "mean_token_accuracy": 0.20190735310316085,
      "num_tokens": 68953733.0,
      "step": 34030
    },
    {
      "entropy": 5.996960687637329,
      "epoch": 2.974307436860963,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004122027667113146,
      "loss": 5.2979,
      "mean_token_accuracy": 0.2013537034392357,
      "num_tokens": 68963888.0,
      "step": 34035
    },
    {
      "entropy": 6.091354703903198,
      "epoch": 2.974744385213668,
      "grad_norm": 1.21875,
      "learning_rate": 0.000412178068263126,
      "loss": 5.3113,
      "mean_token_accuracy": 0.19982012510299682,
      "num_tokens": 68974652.0,
      "step": 34040
    },
    {
      "entropy": 6.064648914337158,
      "epoch": 2.9751813335663724,
      "grad_norm": 1.25,
      "learning_rate": 0.0004121533671837903,
      "loss": 5.206,
      "mean_token_accuracy": 0.2078012779355049,
      "num_tokens": 68984583.0,
      "step": 34045
    },
    {
      "entropy": 5.969772005081177,
      "epoch": 2.9756182819190773,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004121286634737811,
      "loss": 5.31,
      "mean_token_accuracy": 0.20302751809358596,
      "num_tokens": 68995790.0,
      "step": 34050
    },
    {
      "entropy": 6.039069509506225,
      "epoch": 2.9760552302717818,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004121039571335723,
      "loss": 5.3314,
      "mean_token_accuracy": 0.1983987346291542,
      "num_tokens": 69006695.0,
      "step": 34055
    },
    {
      "entropy": 6.008229207992554,
      "epoch": 2.9764921786244867,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004120792481636378,
      "loss": 5.2162,
      "mean_token_accuracy": 0.21891285479068756,
      "num_tokens": 69017504.0,
      "step": 34060
    },
    {
      "entropy": 6.001964712142945,
      "epoch": 2.976929126977191,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004120545365644515,
      "loss": 5.3609,
      "mean_token_accuracy": 0.19650242626667022,
      "num_tokens": 69026938.0,
      "step": 34065
    },
    {
      "entropy": 5.998024225234985,
      "epoch": 2.977366075329896,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004120298223364872,
      "loss": 5.2472,
      "mean_token_accuracy": 0.20695810317993163,
      "num_tokens": 69037404.0,
      "step": 34070
    },
    {
      "entropy": 6.022457790374756,
      "epoch": 2.9778030236826005,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041200510548021935,
      "loss": 5.229,
      "mean_token_accuracy": 0.2102394312620163,
      "num_tokens": 69047507.0,
      "step": 34075
    },
    {
      "entropy": 5.980376291275024,
      "epoch": 2.9782399720353054,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041198038599612156,
      "loss": 5.2474,
      "mean_token_accuracy": 0.2094618171453476,
      "num_tokens": 69057626.0,
      "step": 34080
    },
    {
      "entropy": 6.027951717376709,
      "epoch": 2.9786769203880104,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041195566388466836,
      "loss": 5.2926,
      "mean_token_accuracy": 0.20470763742923737,
      "num_tokens": 69068142.0,
      "step": 34085
    },
    {
      "entropy": 6.069239568710327,
      "epoch": 2.979113868740715,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041193093914633364,
      "loss": 5.3062,
      "mean_token_accuracy": 0.20382058918476104,
      "num_tokens": 69077935.0,
      "step": 34090
    },
    {
      "entropy": 6.053731489181518,
      "epoch": 2.9795508170934193,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004119062117815918,
      "loss": 5.3198,
      "mean_token_accuracy": 0.19745587259531022,
      "num_tokens": 69087397.0,
      "step": 34095
    },
    {
      "entropy": 6.02899866104126,
      "epoch": 2.979987765446124,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004118814817909171,
      "loss": 5.3376,
      "mean_token_accuracy": 0.20749568790197373,
      "num_tokens": 69097666.0,
      "step": 34100
    },
    {
      "entropy": 6.010295152664185,
      "epoch": 2.980424713798829,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004118567491747839,
      "loss": 5.2738,
      "mean_token_accuracy": 0.20318912118673324,
      "num_tokens": 69107516.0,
      "step": 34105
    },
    {
      "entropy": 6.027885293960571,
      "epoch": 2.9808616621515336,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004118320139336665,
      "loss": 5.245,
      "mean_token_accuracy": 0.20498040467500686,
      "num_tokens": 69117462.0,
      "step": 34110
    },
    {
      "entropy": 6.011681604385376,
      "epoch": 2.9812986105042385,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004118072760680393,
      "loss": 5.2227,
      "mean_token_accuracy": 0.2164926752448082,
      "num_tokens": 69128171.0,
      "step": 34115
    },
    {
      "entropy": 6.004879426956177,
      "epoch": 2.981735558856943,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041178253557837697,
      "loss": 5.2757,
      "mean_token_accuracy": 0.2043527513742447,
      "num_tokens": 69138169.0,
      "step": 34120
    },
    {
      "entropy": 6.003472518920899,
      "epoch": 2.982172507209648,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041175779246515384,
      "loss": 5.2211,
      "mean_token_accuracy": 0.2125873386859894,
      "num_tokens": 69147518.0,
      "step": 34125
    },
    {
      "entropy": 6.026022577285767,
      "epoch": 2.9826094555623524,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004117330467288447,
      "loss": 5.2293,
      "mean_token_accuracy": 0.20196057260036468,
      "num_tokens": 69157049.0,
      "step": 34130
    },
    {
      "entropy": 5.973409795761109,
      "epoch": 2.9830464039150573,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041170829836992403,
      "loss": 5.2679,
      "mean_token_accuracy": 0.20845962464809417,
      "num_tokens": 69166502.0,
      "step": 34135
    },
    {
      "entropy": 6.0321320533752445,
      "epoch": 2.983483352267762,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004116835473888665,
      "loss": 5.2229,
      "mean_token_accuracy": 0.2077076643705368,
      "num_tokens": 69176566.0,
      "step": 34140
    },
    {
      "entropy": 6.054804897308349,
      "epoch": 2.9839203006204666,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004116587937861469,
      "loss": 5.264,
      "mean_token_accuracy": 0.19997065514326096,
      "num_tokens": 69186677.0,
      "step": 34145
    },
    {
      "entropy": 6.032770013809204,
      "epoch": 2.984357248973171,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004116340375622402,
      "loss": 5.3347,
      "mean_token_accuracy": 0.20007654279470444,
      "num_tokens": 69197595.0,
      "step": 34150
    },
    {
      "entropy": 5.994155311584473,
      "epoch": 2.984794197325876,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004116092787176209,
      "loss": 5.2607,
      "mean_token_accuracy": 0.21605707406997682,
      "num_tokens": 69207384.0,
      "step": 34155
    },
    {
      "entropy": 6.000405883789062,
      "epoch": 2.985231145678581,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041158451725276405,
      "loss": 5.1787,
      "mean_token_accuracy": 0.21364200115203857,
      "num_tokens": 69217019.0,
      "step": 34160
    },
    {
      "entropy": 6.103455018997193,
      "epoch": 2.9856680940312854,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004115597531681447,
      "loss": 5.4075,
      "mean_token_accuracy": 0.18868394941091537,
      "num_tokens": 69229076.0,
      "step": 34165
    },
    {
      "entropy": 6.032202482223511,
      "epoch": 2.9861050423839903,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041153498646423763,
      "loss": 5.3394,
      "mean_token_accuracy": 0.19999389946460724,
      "num_tokens": 69238771.0,
      "step": 34170
    },
    {
      "entropy": 6.009576272964478,
      "epoch": 2.986541990736695,
      "grad_norm": 1.328125,
      "learning_rate": 0.000411510217141518,
      "loss": 5.2844,
      "mean_token_accuracy": 0.2010635256767273,
      "num_tokens": 69248864.0,
      "step": 34175
    },
    {
      "entropy": 6.044854545593262,
      "epoch": 2.9869789390893997,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041148544520046087,
      "loss": 5.3278,
      "mean_token_accuracy": 0.19866156578063965,
      "num_tokens": 69259372.0,
      "step": 34180
    },
    {
      "entropy": 6.039796352386475,
      "epoch": 2.987415887442104,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004114606706415415,
      "loss": 5.2179,
      "mean_token_accuracy": 0.2080077424645424,
      "num_tokens": 69270079.0,
      "step": 34185
    },
    {
      "entropy": 6.049261999130249,
      "epoch": 2.987852835794809,
      "grad_norm": 1.25,
      "learning_rate": 0.0004114358934652349,
      "loss": 5.2404,
      "mean_token_accuracy": 0.2130574628710747,
      "num_tokens": 69279909.0,
      "step": 34190
    },
    {
      "entropy": 5.953661203384399,
      "epoch": 2.988289784147514,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004114111136720163,
      "loss": 5.1967,
      "mean_token_accuracy": 0.21282802969217302,
      "num_tokens": 69289598.0,
      "step": 34195
    },
    {
      "entropy": 6.045504713058472,
      "epoch": 2.9887267325002185,
      "grad_norm": 1.125,
      "learning_rate": 0.0004113863312623611,
      "loss": 5.3001,
      "mean_token_accuracy": 0.19988285452127458,
      "num_tokens": 69301305.0,
      "step": 34200
    },
    {
      "entropy": 5.9949657917022705,
      "epoch": 2.989163680852923,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004113615462367446,
      "loss": 5.2439,
      "mean_token_accuracy": 0.20977806746959687,
      "num_tokens": 69310560.0,
      "step": 34205
    },
    {
      "entropy": 5.9953062534332275,
      "epoch": 2.989600629205628,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041133675859564213,
      "loss": 5.2876,
      "mean_token_accuracy": 0.2011028215289116,
      "num_tokens": 69320286.0,
      "step": 34210
    },
    {
      "entropy": 6.008439064025879,
      "epoch": 2.9900375775583328,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041131196833952925,
      "loss": 5.229,
      "mean_token_accuracy": 0.20292495042085648,
      "num_tokens": 69329895.0,
      "step": 34215
    },
    {
      "entropy": 6.056705141067505,
      "epoch": 2.9904745259110372,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004112871754688814,
      "loss": 5.2406,
      "mean_token_accuracy": 0.20474981516599655,
      "num_tokens": 69340122.0,
      "step": 34220
    },
    {
      "entropy": 6.016272258758545,
      "epoch": 2.990911474263742,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041126237998417414,
      "loss": 5.2714,
      "mean_token_accuracy": 0.20438892990350724,
      "num_tokens": 69349354.0,
      "step": 34225
    },
    {
      "entropy": 6.009840297698974,
      "epoch": 2.9913484226164466,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00041123758188588296,
      "loss": 5.3219,
      "mean_token_accuracy": 0.20061679184436798,
      "num_tokens": 69360236.0,
      "step": 34230
    },
    {
      "entropy": 6.065124797821045,
      "epoch": 2.9917853709691515,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00041121278117448365,
      "loss": 5.3253,
      "mean_token_accuracy": 0.20245359390974044,
      "num_tokens": 69369362.0,
      "step": 34235
    },
    {
      "entropy": 6.041576910018921,
      "epoch": 2.992222319321856,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041118797785045174,
      "loss": 5.2794,
      "mean_token_accuracy": 0.19843100607395173,
      "num_tokens": 69379403.0,
      "step": 34240
    },
    {
      "entropy": 5.980595541000366,
      "epoch": 2.992659267674561,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004111631719142631,
      "loss": 5.3132,
      "mean_token_accuracy": 0.20556840300559998,
      "num_tokens": 69388667.0,
      "step": 34245
    },
    {
      "entropy": 6.030351257324218,
      "epoch": 2.993096216027266,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041113836336639347,
      "loss": 5.2525,
      "mean_token_accuracy": 0.2030839890241623,
      "num_tokens": 69399552.0,
      "step": 34250
    },
    {
      "entropy": 6.006982851028442,
      "epoch": 2.9935331643799703,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004111135522073187,
      "loss": 5.1764,
      "mean_token_accuracy": 0.21352204382419587,
      "num_tokens": 69409323.0,
      "step": 34255
    },
    {
      "entropy": 6.065612697601319,
      "epoch": 2.9939701127326748,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004110887384375147,
      "loss": 5.3106,
      "mean_token_accuracy": 0.19831886440515517,
      "num_tokens": 69418696.0,
      "step": 34260
    },
    {
      "entropy": 6.05010404586792,
      "epoch": 2.9944070610853797,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004110639220574573,
      "loss": 5.3208,
      "mean_token_accuracy": 0.19899354130029678,
      "num_tokens": 69429470.0,
      "step": 34265
    },
    {
      "entropy": 5.9626627445220945,
      "epoch": 2.9948440094380846,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041103910306762267,
      "loss": 5.2035,
      "mean_token_accuracy": 0.21010808795690536,
      "num_tokens": 69440636.0,
      "step": 34270
    },
    {
      "entropy": 6.071877336502075,
      "epoch": 2.995280957790789,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041101428146848667,
      "loss": 5.3228,
      "mean_token_accuracy": 0.20276567041873933,
      "num_tokens": 69451819.0,
      "step": 34275
    },
    {
      "entropy": 6.077642822265625,
      "epoch": 2.995717906143494,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041098945726052556,
      "loss": 5.3624,
      "mean_token_accuracy": 0.2001253679394722,
      "num_tokens": 69462690.0,
      "step": 34280
    },
    {
      "entropy": 5.975320291519165,
      "epoch": 2.9961548544961984,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004109646304442154,
      "loss": 5.2285,
      "mean_token_accuracy": 0.2112015500664711,
      "num_tokens": 69473338.0,
      "step": 34285
    },
    {
      "entropy": 6.0020973682403564,
      "epoch": 2.9965918028489034,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004109398010200323,
      "loss": 5.2705,
      "mean_token_accuracy": 0.2011418327689171,
      "num_tokens": 69482189.0,
      "step": 34290
    },
    {
      "entropy": 5.929888343811035,
      "epoch": 2.997028751201608,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004109149689884527,
      "loss": 5.2295,
      "mean_token_accuracy": 0.21237343102693557,
      "num_tokens": 69492084.0,
      "step": 34295
    },
    {
      "entropy": 6.022088050842285,
      "epoch": 2.9974656995543127,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041089013434995273,
      "loss": 5.2446,
      "mean_token_accuracy": 0.20627565085887908,
      "num_tokens": 69502886.0,
      "step": 34300
    },
    {
      "entropy": 6.054227352142334,
      "epoch": 2.9979026479070177,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004108652971050088,
      "loss": 5.2283,
      "mean_token_accuracy": 0.20705406963825226,
      "num_tokens": 69512670.0,
      "step": 34305
    },
    {
      "entropy": 5.993274784088134,
      "epoch": 2.998339596259722,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041084045725409726,
      "loss": 5.312,
      "mean_token_accuracy": 0.1996456891298294,
      "num_tokens": 69522744.0,
      "step": 34310
    },
    {
      "entropy": 5.953725719451905,
      "epoch": 2.9987765446124266,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004108156147976945,
      "loss": 5.1723,
      "mean_token_accuracy": 0.21115855276584625,
      "num_tokens": 69533195.0,
      "step": 34315
    },
    {
      "entropy": 6.04329776763916,
      "epoch": 2.9992134929651315,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004107907697362772,
      "loss": 5.334,
      "mean_token_accuracy": 0.20072345286607743,
      "num_tokens": 69543058.0,
      "step": 34320
    },
    {
      "entropy": 6.090622186660767,
      "epoch": 2.9996504413178364,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004107659220703217,
      "loss": 5.2917,
      "mean_token_accuracy": 0.2020493820309639,
      "num_tokens": 69552428.0,
      "step": 34325
    },
    {
      "entropy": 6.0521653175354,
      "epoch": 3.000087389670541,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004107410718003047,
      "loss": 5.2551,
      "mean_token_accuracy": 0.2104414537549019,
      "num_tokens": 69561735.0,
      "step": 34330
    },
    {
      "entropy": 5.9812651634216305,
      "epoch": 3.000524338023246,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004107162189267028,
      "loss": 5.2312,
      "mean_token_accuracy": 0.20704602003097533,
      "num_tokens": 69573165.0,
      "step": 34335
    },
    {
      "entropy": 5.968122243881226,
      "epoch": 3.0009612863759503,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00041069136344999273,
      "loss": 5.1803,
      "mean_token_accuracy": 0.21808178424835206,
      "num_tokens": 69583484.0,
      "step": 34340
    },
    {
      "entropy": 5.975345230102539,
      "epoch": 3.001398234728655,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004106665053706512,
      "loss": 5.2378,
      "mean_token_accuracy": 0.2071292594075203,
      "num_tokens": 69593486.0,
      "step": 34345
    },
    {
      "entropy": 6.057748937606812,
      "epoch": 3.0018351830813597,
      "grad_norm": 1.3125,
      "learning_rate": 0.000410641644689155,
      "loss": 5.3116,
      "mean_token_accuracy": 0.2097623512148857,
      "num_tokens": 69604245.0,
      "step": 34350
    },
    {
      "entropy": 6.0229075908660885,
      "epoch": 3.0022721314340646,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000410616781405981,
      "loss": 5.2298,
      "mean_token_accuracy": 0.2091633826494217,
      "num_tokens": 69614237.0,
      "step": 34355
    },
    {
      "entropy": 6.048323106765747,
      "epoch": 3.002709079786769,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004105919155216061,
      "loss": 5.2614,
      "mean_token_accuracy": 0.2073272407054901,
      "num_tokens": 69623695.0,
      "step": 34360
    },
    {
      "entropy": 6.031410789489746,
      "epoch": 3.003146028139474,
      "grad_norm": 1.359375,
      "learning_rate": 0.00041056704703650715,
      "loss": 5.1408,
      "mean_token_accuracy": 0.21792724430561067,
      "num_tokens": 69633299.0,
      "step": 34365
    },
    {
      "entropy": 5.987655591964722,
      "epoch": 3.0035829764921784,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004105421759511612,
      "loss": 5.2177,
      "mean_token_accuracy": 0.20380600690841674,
      "num_tokens": 69642853.0,
      "step": 34370
    },
    {
      "entropy": 6.053563928604126,
      "epoch": 3.0040199248448833,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004105173022660453,
      "loss": 5.2813,
      "mean_token_accuracy": 0.2011645719408989,
      "num_tokens": 69652697.0,
      "step": 34375
    },
    {
      "entropy": 5.991824150085449,
      "epoch": 3.0044568731975883,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004104924259816366,
      "loss": 5.2563,
      "mean_token_accuracy": 0.21186551451683044,
      "num_tokens": 69664261.0,
      "step": 34380
    },
    {
      "entropy": 6.0420835494995115,
      "epoch": 3.0048938215502927,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004104675470984121,
      "loss": 5.2796,
      "mean_token_accuracy": 0.21334433406591416,
      "num_tokens": 69674819.0,
      "step": 34385
    },
    {
      "entropy": 6.030449771881104,
      "epoch": 3.0053307699029976,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041044266561684914,
      "loss": 5.2248,
      "mean_token_accuracy": 0.20479593127965928,
      "num_tokens": 69685546.0,
      "step": 34390
    },
    {
      "entropy": 5.910425090789795,
      "epoch": 3.005767718255702,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004104177815374248,
      "loss": 5.1329,
      "mean_token_accuracy": 0.22010606229305268,
      "num_tokens": 69695192.0,
      "step": 34395
    },
    {
      "entropy": 6.034918785095215,
      "epoch": 3.006204666608407,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004103928948606164,
      "loss": 5.1881,
      "mean_token_accuracy": 0.20772004574537278,
      "num_tokens": 69705669.0,
      "step": 34400
    },
    {
      "entropy": 5.952854251861572,
      "epoch": 3.0066416149611115,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004103680055869015,
      "loss": 5.1766,
      "mean_token_accuracy": 0.21676891297101974,
      "num_tokens": 69716274.0,
      "step": 34405
    },
    {
      "entropy": 6.045078992843628,
      "epoch": 3.0070785633138164,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004103431137167572,
      "loss": 5.2555,
      "mean_token_accuracy": 0.2041129380464554,
      "num_tokens": 69726438.0,
      "step": 34410
    },
    {
      "entropy": 6.015651893615723,
      "epoch": 3.007515511666521,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00041031821925066115,
      "loss": 5.2289,
      "mean_token_accuracy": 0.21198931187391282,
      "num_tokens": 69735467.0,
      "step": 34415
    },
    {
      "entropy": 5.997117185592652,
      "epoch": 3.007952460019226,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041029332218909067,
      "loss": 5.2239,
      "mean_token_accuracy": 0.2094128280878067,
      "num_tokens": 69744886.0,
      "step": 34420
    },
    {
      "entropy": 5.981879615783692,
      "epoch": 3.0083894083719303,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00041026842253252333,
      "loss": 5.2582,
      "mean_token_accuracy": 0.20545212924480438,
      "num_tokens": 69754337.0,
      "step": 34425
    },
    {
      "entropy": 5.9520172595977785,
      "epoch": 3.008826356724635,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004102435202814369,
      "loss": 5.2123,
      "mean_token_accuracy": 0.2161676213145256,
      "num_tokens": 69764660.0,
      "step": 34430
    },
    {
      "entropy": 5.9664100170135494,
      "epoch": 3.00926330507734,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004102186154363088,
      "loss": 5.2033,
      "mean_token_accuracy": 0.20734989047050476,
      "num_tokens": 69774339.0,
      "step": 34435
    },
    {
      "entropy": 6.030060911178589,
      "epoch": 3.0097002534300445,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041019370799761685,
      "loss": 5.2857,
      "mean_token_accuracy": 0.2034371241927147,
      "num_tokens": 69785272.0,
      "step": 34440
    },
    {
      "entropy": 5.99529767036438,
      "epoch": 3.0101372017827495,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004101687979658386,
      "loss": 5.3274,
      "mean_token_accuracy": 0.2047102138400078,
      "num_tokens": 69796646.0,
      "step": 34445
    },
    {
      "entropy": 5.967960166931152,
      "epoch": 3.010574150135454,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041014388534145213,
      "loss": 5.1824,
      "mean_token_accuracy": 0.21534176617860795,
      "num_tokens": 69807233.0,
      "step": 34450
    },
    {
      "entropy": 6.032260847091675,
      "epoch": 3.011011098488159,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004101189701249351,
      "loss": 5.2742,
      "mean_token_accuracy": 0.20507555454969406,
      "num_tokens": 69818131.0,
      "step": 34455
    },
    {
      "entropy": 6.064706325531006,
      "epoch": 3.0114480468408633,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004100940523167654,
      "loss": 5.3157,
      "mean_token_accuracy": 0.20100643932819368,
      "num_tokens": 69829537.0,
      "step": 34460
    },
    {
      "entropy": 6.003563356399536,
      "epoch": 3.0118849951935682,
      "grad_norm": 1.25,
      "learning_rate": 0.000410069131917421,
      "loss": 5.2775,
      "mean_token_accuracy": 0.20076480656862258,
      "num_tokens": 69840964.0,
      "step": 34465
    },
    {
      "entropy": 6.018824291229248,
      "epoch": 3.0123219435462727,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041004420892737993,
      "loss": 5.1756,
      "mean_token_accuracy": 0.2078770726919174,
      "num_tokens": 69849919.0,
      "step": 34470
    },
    {
      "entropy": 5.994872617721557,
      "epoch": 3.0127588918989776,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004100192833471201,
      "loss": 5.2607,
      "mean_token_accuracy": 0.2092216655611992,
      "num_tokens": 69859832.0,
      "step": 34475
    },
    {
      "entropy": 5.9896338939666744,
      "epoch": 3.013195840251682,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004099943551771196,
      "loss": 5.1686,
      "mean_token_accuracy": 0.21408320665359498,
      "num_tokens": 69869382.0,
      "step": 34480
    },
    {
      "entropy": 5.977575588226318,
      "epoch": 3.013632788604387,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004099694244178568,
      "loss": 5.1931,
      "mean_token_accuracy": 0.21161047667264937,
      "num_tokens": 69879282.0,
      "step": 34485
    },
    {
      "entropy": 5.963607263565064,
      "epoch": 3.0140697369570915,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004099444910698096,
      "loss": 5.1565,
      "mean_token_accuracy": 0.2183496579527855,
      "num_tokens": 69889876.0,
      "step": 34490
    },
    {
      "entropy": 6.054664754867554,
      "epoch": 3.0145066853097964,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004099195551334564,
      "loss": 5.2744,
      "mean_token_accuracy": 0.199636572599411,
      "num_tokens": 69899642.0,
      "step": 34495
    },
    {
      "entropy": 5.954282617568969,
      "epoch": 3.0149436336625013,
      "grad_norm": 1.390625,
      "learning_rate": 0.00040989461660927536,
      "loss": 5.1453,
      "mean_token_accuracy": 0.2135930687189102,
      "num_tokens": 69909877.0,
      "step": 34500
    },
    {
      "entropy": 5.976737594604492,
      "epoch": 3.0153805820152058,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00040986967549774497,
      "loss": 5.1792,
      "mean_token_accuracy": 0.21731090694665908,
      "num_tokens": 69920061.0,
      "step": 34505
    },
    {
      "entropy": 5.975213670730591,
      "epoch": 3.0158175303679107,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004098447317993435,
      "loss": 5.2071,
      "mean_token_accuracy": 0.21276819556951523,
      "num_tokens": 69930739.0,
      "step": 34510
    },
    {
      "entropy": 5.917355537414551,
      "epoch": 3.016254478720615,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040981978551454947,
      "loss": 5.122,
      "mean_token_accuracy": 0.21771000027656556,
      "num_tokens": 69939825.0,
      "step": 34515
    },
    {
      "entropy": 6.082456636428833,
      "epoch": 3.01669142707332,
      "grad_norm": 1.34375,
      "learning_rate": 0.00040979483664384123,
      "loss": 5.3635,
      "mean_token_accuracy": 0.19421915858983993,
      "num_tokens": 69949662.0,
      "step": 34520
    },
    {
      "entropy": 6.007339572906494,
      "epoch": 3.0171283754260245,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00040976988518769743,
      "loss": 5.2371,
      "mean_token_accuracy": 0.20943409204483032,
      "num_tokens": 69959950.0,
      "step": 34525
    },
    {
      "entropy": 6.064197015762329,
      "epoch": 3.0175653237787294,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040974493114659654,
      "loss": 5.2749,
      "mean_token_accuracy": 0.19860604107379914,
      "num_tokens": 69970451.0,
      "step": 34530
    },
    {
      "entropy": 6.013507318496704,
      "epoch": 3.018002272131434,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004097199745210174,
      "loss": 5.2033,
      "mean_token_accuracy": 0.20932240039110184,
      "num_tokens": 69979940.0,
      "step": 34535
    },
    {
      "entropy": 6.024255895614624,
      "epoch": 3.018439220484139,
      "grad_norm": 1.25,
      "learning_rate": 0.00040969501531143844,
      "loss": 5.2065,
      "mean_token_accuracy": 0.21147104650735854,
      "num_tokens": 69989507.0,
      "step": 34540
    },
    {
      "entropy": 6.017589807510376,
      "epoch": 3.0188761688368433,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00040967005351833863,
      "loss": 5.1916,
      "mean_token_accuracy": 0.21275194436311723,
      "num_tokens": 69999184.0,
      "step": 34545
    },
    {
      "entropy": 6.045178174972534,
      "epoch": 3.019313117189548,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004096450891421965,
      "loss": 5.2803,
      "mean_token_accuracy": 0.20306475907564164,
      "num_tokens": 70009569.0,
      "step": 34550
    },
    {
      "entropy": 6.051294946670533,
      "epoch": 3.019750065542253,
      "grad_norm": 1.25,
      "learning_rate": 0.00040962012218349105,
      "loss": 5.2432,
      "mean_token_accuracy": 0.20464086830615996,
      "num_tokens": 70019824.0,
      "step": 34555
    },
    {
      "entropy": 5.954359292984009,
      "epoch": 3.0201870138949576,
      "grad_norm": 1.375,
      "learning_rate": 0.0004095951526427011,
      "loss": 5.1258,
      "mean_token_accuracy": 0.21774054020643235,
      "num_tokens": 70029043.0,
      "step": 34560
    },
    {
      "entropy": 6.0162248611450195,
      "epoch": 3.0206239622476625,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004095701805203056,
      "loss": 5.2377,
      "mean_token_accuracy": 0.21021708846092224,
      "num_tokens": 70039794.0,
      "step": 34565
    },
    {
      "entropy": 6.0354063510894775,
      "epoch": 3.021060910600367,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004095452058167837,
      "loss": 5.2541,
      "mean_token_accuracy": 0.20641523152589797,
      "num_tokens": 70049866.0,
      "step": 34570
    },
    {
      "entropy": 5.974526596069336,
      "epoch": 3.021497858953072,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000409520228532614,
      "loss": 5.1587,
      "mean_token_accuracy": 0.2160967245697975,
      "num_tokens": 70059471.0,
      "step": 34575
    },
    {
      "entropy": 5.967530393600464,
      "epoch": 3.0219348073057763,
      "grad_norm": 1.375,
      "learning_rate": 0.000409495248668276,
      "loss": 5.1932,
      "mean_token_accuracy": 0.2136614054441452,
      "num_tokens": 70069003.0,
      "step": 34580
    },
    {
      "entropy": 6.015859127044678,
      "epoch": 3.0223717556584813,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040947026622424853,
      "loss": 5.2769,
      "mean_token_accuracy": 0.20294377356767654,
      "num_tokens": 70078650.0,
      "step": 34585
    },
    {
      "entropy": 6.027313232421875,
      "epoch": 3.0228087040111857,
      "grad_norm": 1.375,
      "learning_rate": 0.00040944528120101097,
      "loss": 5.1889,
      "mean_token_accuracy": 0.21558879613876342,
      "num_tokens": 70087361.0,
      "step": 34590
    },
    {
      "entropy": 5.991868352890014,
      "epoch": 3.0232456523638906,
      "grad_norm": 1.28125,
      "learning_rate": 0.00040942029359904245,
      "loss": 5.2646,
      "mean_token_accuracy": 0.2103925257921219,
      "num_tokens": 70097289.0,
      "step": 34595
    },
    {
      "entropy": 5.968020582199097,
      "epoch": 3.023682600716595,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040939530341882225,
      "loss": 5.2046,
      "mean_token_accuracy": 0.2183875098824501,
      "num_tokens": 70108099.0,
      "step": 34600
    },
    {
      "entropy": 6.0297472953796385,
      "epoch": 3.0241195490693,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00040937031066082974,
      "loss": 5.2971,
      "mean_token_accuracy": 0.20436987578868865,
      "num_tokens": 70119243.0,
      "step": 34605
    },
    {
      "entropy": 6.047988319396973,
      "epoch": 3.0245564974220045,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004093453153255442,
      "loss": 5.2084,
      "mean_token_accuracy": 0.21217515021562577,
      "num_tokens": 70129073.0,
      "step": 34610
    },
    {
      "entropy": 5.965170907974243,
      "epoch": 3.0249934457747094,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004093203174134452,
      "loss": 5.1568,
      "mean_token_accuracy": 0.21713462471961975,
      "num_tokens": 70139778.0,
      "step": 34615
    },
    {
      "entropy": 6.030516576766968,
      "epoch": 3.0254303941274143,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040929531692501206,
      "loss": 5.3375,
      "mean_token_accuracy": 0.20379277020692826,
      "num_tokens": 70151088.0,
      "step": 34620
    },
    {
      "entropy": 5.96963357925415,
      "epoch": 3.025867342480119,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040927031386072435,
      "loss": 5.1861,
      "mean_token_accuracy": 0.21700191199779512,
      "num_tokens": 70160133.0,
      "step": 34625
    },
    {
      "entropy": 6.029078340530395,
      "epoch": 3.0263042908328237,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040924530822106177,
      "loss": 5.1945,
      "mean_token_accuracy": 0.20688151568174362,
      "num_tokens": 70171176.0,
      "step": 34630
    },
    {
      "entropy": 6.035934972763061,
      "epoch": 3.026741239185528,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040922030000650376,
      "loss": 5.2321,
      "mean_token_accuracy": 0.20382107496261598,
      "num_tokens": 70181413.0,
      "step": 34635
    },
    {
      "entropy": 6.044013738632202,
      "epoch": 3.027178187538233,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004091952892175301,
      "loss": 5.2497,
      "mean_token_accuracy": 0.20493296831846236,
      "num_tokens": 70190535.0,
      "step": 34640
    },
    {
      "entropy": 5.98689866065979,
      "epoch": 3.0276151358909376,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00040917027585462044,
      "loss": 5.2225,
      "mean_token_accuracy": 0.21183291971683502,
      "num_tokens": 70199440.0,
      "step": 34645
    },
    {
      "entropy": 5.991759967803955,
      "epoch": 3.0280520842436425,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004091452599182546,
      "loss": 5.1749,
      "mean_token_accuracy": 0.21359325349330902,
      "num_tokens": 70209846.0,
      "step": 34650
    },
    {
      "entropy": 6.022160148620605,
      "epoch": 3.028489032596347,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040912024140891235,
      "loss": 5.2846,
      "mean_token_accuracy": 0.2021244153380394,
      "num_tokens": 70219910.0,
      "step": 34655
    },
    {
      "entropy": 6.014825439453125,
      "epoch": 3.028925980949052,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004090952203270737,
      "loss": 5.2165,
      "mean_token_accuracy": 0.20634110867977143,
      "num_tokens": 70230937.0,
      "step": 34660
    },
    {
      "entropy": 5.9801976680755615,
      "epoch": 3.0293629293017563,
      "grad_norm": 1.28125,
      "learning_rate": 0.00040907019667321833,
      "loss": 5.2192,
      "mean_token_accuracy": 0.2116853892803192,
      "num_tokens": 70240791.0,
      "step": 34665
    },
    {
      "entropy": 6.020701789855957,
      "epoch": 3.0297998776544612,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004090451704478264,
      "loss": 5.2255,
      "mean_token_accuracy": 0.21215442270040513,
      "num_tokens": 70251231.0,
      "step": 34670
    },
    {
      "entropy": 5.9882417678833,
      "epoch": 3.030236826007166,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040902014165137787,
      "loss": 5.1533,
      "mean_token_accuracy": 0.21622912883758544,
      "num_tokens": 70260993.0,
      "step": 34675
    },
    {
      "entropy": 6.026395654678344,
      "epoch": 3.0306737743598706,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004089951102843528,
      "loss": 5.3168,
      "mean_token_accuracy": 0.2053797349333763,
      "num_tokens": 70271380.0,
      "step": 34680
    },
    {
      "entropy": 6.0411242008209225,
      "epoch": 3.0311107227125755,
      "grad_norm": 1.234375,
      "learning_rate": 0.00040897007634723125,
      "loss": 5.2692,
      "mean_token_accuracy": 0.206558720767498,
      "num_tokens": 70282614.0,
      "step": 34685
    },
    {
      "entropy": 6.062633991241455,
      "epoch": 3.03154767106528,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004089450398404935,
      "loss": 5.2099,
      "mean_token_accuracy": 0.21422804445028304,
      "num_tokens": 70292084.0,
      "step": 34690
    },
    {
      "entropy": 5.937966537475586,
      "epoch": 3.031984619417985,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00040892000076461966,
      "loss": 5.1297,
      "mean_token_accuracy": 0.21677112728357315,
      "num_tokens": 70302205.0,
      "step": 34695
    },
    {
      "entropy": 5.9848284244537355,
      "epoch": 3.0324215677706894,
      "grad_norm": 1.125,
      "learning_rate": 0.00040889495912009,
      "loss": 5.2873,
      "mean_token_accuracy": 0.20815921425819398,
      "num_tokens": 70312264.0,
      "step": 34700
    },
    {
      "entropy": 5.996008253097534,
      "epoch": 3.0328585161233943,
      "grad_norm": 1.28125,
      "learning_rate": 0.00040886991490738495,
      "loss": 5.1818,
      "mean_token_accuracy": 0.21225829273462296,
      "num_tokens": 70321733.0,
      "step": 34705
    },
    {
      "entropy": 6.016478252410889,
      "epoch": 3.0332954644760988,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040884486812698475,
      "loss": 5.3035,
      "mean_token_accuracy": 0.2013198584318161,
      "num_tokens": 70331621.0,
      "step": 34710
    },
    {
      "entropy": 6.004620170593261,
      "epoch": 3.0337324128288037,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040881981877936987,
      "loss": 5.2243,
      "mean_token_accuracy": 0.2032810628414154,
      "num_tokens": 70342247.0,
      "step": 34715
    },
    {
      "entropy": 5.990600681304931,
      "epoch": 3.034169361181508,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004087947668650207,
      "loss": 5.198,
      "mean_token_accuracy": 0.2106366127729416,
      "num_tokens": 70352423.0,
      "step": 34720
    },
    {
      "entropy": 6.013003730773926,
      "epoch": 3.034606309534213,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004087697123844178,
      "loss": 5.2719,
      "mean_token_accuracy": 0.20733019560575486,
      "num_tokens": 70362740.0,
      "step": 34725
    },
    {
      "entropy": 5.988236045837402,
      "epoch": 3.035043257886918,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004087446553380417,
      "loss": 5.1666,
      "mean_token_accuracy": 0.21428755968809127,
      "num_tokens": 70372456.0,
      "step": 34730
    },
    {
      "entropy": 5.99922890663147,
      "epoch": 3.0354802062396224,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004087195957263731,
      "loss": 5.1991,
      "mean_token_accuracy": 0.2205453634262085,
      "num_tokens": 70382741.0,
      "step": 34735
    },
    {
      "entropy": 5.957730340957641,
      "epoch": 3.0359171545923274,
      "grad_norm": 1.3125,
      "learning_rate": 0.00040869453354989255,
      "loss": 5.2113,
      "mean_token_accuracy": 0.2100470706820488,
      "num_tokens": 70392712.0,
      "step": 34740
    },
    {
      "entropy": 5.9996177673339846,
      "epoch": 3.036354102945032,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040866946880908076,
      "loss": 5.2761,
      "mean_token_accuracy": 0.2068618729710579,
      "num_tokens": 70403655.0,
      "step": 34745
    },
    {
      "entropy": 6.086016082763672,
      "epoch": 3.0367910512977367,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004086444015044186,
      "loss": 5.3106,
      "mean_token_accuracy": 0.2012304186820984,
      "num_tokens": 70413720.0,
      "step": 34750
    },
    {
      "entropy": 6.0113132953643795,
      "epoch": 3.037227999650441,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004086193316363867,
      "loss": 5.2344,
      "mean_token_accuracy": 0.2052200362086296,
      "num_tokens": 70423910.0,
      "step": 34755
    },
    {
      "entropy": 5.958978223800659,
      "epoch": 3.037664948003146,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040859425920546604,
      "loss": 5.233,
      "mean_token_accuracy": 0.2085111543536186,
      "num_tokens": 70434593.0,
      "step": 34760
    },
    {
      "entropy": 5.986016464233399,
      "epoch": 3.0381018963558506,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004085691842121375,
      "loss": 5.2234,
      "mean_token_accuracy": 0.20586353540420532,
      "num_tokens": 70443967.0,
      "step": 34765
    },
    {
      "entropy": 6.070131587982178,
      "epoch": 3.0385388447085555,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000408544106656882,
      "loss": 5.2429,
      "mean_token_accuracy": 0.20312355309724808,
      "num_tokens": 70454098.0,
      "step": 34770
    },
    {
      "entropy": 5.993809223175049,
      "epoch": 3.03897579306126,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040851902654018057,
      "loss": 5.2344,
      "mean_token_accuracy": 0.20691446214914322,
      "num_tokens": 70463946.0,
      "step": 34775
    },
    {
      "entropy": 6.022187376022339,
      "epoch": 3.039412741413965,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040849394386251425,
      "loss": 5.2452,
      "mean_token_accuracy": 0.20947476476430893,
      "num_tokens": 70474805.0,
      "step": 34780
    },
    {
      "entropy": 5.95706524848938,
      "epoch": 3.03984968976667,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004084688586243641,
      "loss": 5.1699,
      "mean_token_accuracy": 0.21536603271961213,
      "num_tokens": 70484585.0,
      "step": 34785
    },
    {
      "entropy": 6.027737712860107,
      "epoch": 3.0402866381193743,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004084437708262113,
      "loss": 5.2442,
      "mean_token_accuracy": 0.20466963201761246,
      "num_tokens": 70494811.0,
      "step": 34790
    },
    {
      "entropy": 6.041440343856811,
      "epoch": 3.040723586472079,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004084186804685371,
      "loss": 5.259,
      "mean_token_accuracy": 0.20824200809001922,
      "num_tokens": 70505392.0,
      "step": 34795
    },
    {
      "entropy": 5.986456060409546,
      "epoch": 3.0411605348247837,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004083935875518227,
      "loss": 5.2069,
      "mean_token_accuracy": 0.21650601029396058,
      "num_tokens": 70515066.0,
      "step": 34800
    },
    {
      "entropy": 6.042418193817139,
      "epoch": 3.0415974831774886,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00040836849207654937,
      "loss": 5.2091,
      "mean_token_accuracy": 0.20878828316926956,
      "num_tokens": 70525100.0,
      "step": 34805
    },
    {
      "entropy": 6.009033155441284,
      "epoch": 3.042034431530193,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004083433940431985,
      "loss": 5.2271,
      "mean_token_accuracy": 0.20884938538074493,
      "num_tokens": 70535286.0,
      "step": 34810
    },
    {
      "entropy": 6.009340238571167,
      "epoch": 3.042471379882898,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004083182934522514,
      "loss": 5.2625,
      "mean_token_accuracy": 0.211899296939373,
      "num_tokens": 70545724.0,
      "step": 34815
    },
    {
      "entropy": 6.013680648803711,
      "epoch": 3.0429083282356024,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040829319030418967,
      "loss": 5.2173,
      "mean_token_accuracy": 0.21008520126342772,
      "num_tokens": 70555179.0,
      "step": 34820
    },
    {
      "entropy": 6.067273569107056,
      "epoch": 3.0433452765883073,
      "grad_norm": 1.25,
      "learning_rate": 0.0004082680845994946,
      "loss": 5.2673,
      "mean_token_accuracy": 0.2103591278195381,
      "num_tokens": 70564355.0,
      "step": 34825
    },
    {
      "entropy": 5.948031759262085,
      "epoch": 3.043782224941012,
      "grad_norm": 1.125,
      "learning_rate": 0.0004082429763386479,
      "loss": 5.1527,
      "mean_token_accuracy": 0.2226653814315796,
      "num_tokens": 70575367.0,
      "step": 34830
    },
    {
      "entropy": 5.959432458877563,
      "epoch": 3.0442191732937167,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00040821786552213107,
      "loss": 5.2001,
      "mean_token_accuracy": 0.2149418666958809,
      "num_tokens": 70585047.0,
      "step": 34835
    },
    {
      "entropy": 6.009858894348144,
      "epoch": 3.044656121646421,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00040819275215042586,
      "loss": 5.2333,
      "mean_token_accuracy": 0.2035510152578354,
      "num_tokens": 70594736.0,
      "step": 34840
    },
    {
      "entropy": 6.021680688858032,
      "epoch": 3.045093069999126,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040816763622401376,
      "loss": 5.2361,
      "mean_token_accuracy": 0.21068723499774933,
      "num_tokens": 70604762.0,
      "step": 34845
    },
    {
      "entropy": 6.016321182250977,
      "epoch": 3.045530018351831,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040814251774337666,
      "loss": 5.2395,
      "mean_token_accuracy": 0.2074071690440178,
      "num_tokens": 70614438.0,
      "step": 34850
    },
    {
      "entropy": 6.084048414230347,
      "epoch": 3.0459669667045355,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004081173967089963,
      "loss": 5.3505,
      "mean_token_accuracy": 0.20162802040576935,
      "num_tokens": 70624669.0,
      "step": 34855
    },
    {
      "entropy": 6.002048635482788,
      "epoch": 3.0464039150572404,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040809227312135456,
      "loss": 5.1865,
      "mean_token_accuracy": 0.21277084201574326,
      "num_tokens": 70635082.0,
      "step": 34860
    },
    {
      "entropy": 6.017177963256836,
      "epoch": 3.046840863409945,
      "grad_norm": 1.25,
      "learning_rate": 0.0004080671469809332,
      "loss": 5.2317,
      "mean_token_accuracy": 0.2173943430185318,
      "num_tokens": 70644117.0,
      "step": 34865
    },
    {
      "entropy": 6.007169198989868,
      "epoch": 3.04727781176265,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004080420182882144,
      "loss": 5.2658,
      "mean_token_accuracy": 0.20355447232723237,
      "num_tokens": 70653299.0,
      "step": 34870
    },
    {
      "entropy": 5.969123458862304,
      "epoch": 3.0477147601153542,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040801688704367975,
      "loss": 5.233,
      "mean_token_accuracy": 0.2093753084540367,
      "num_tokens": 70664477.0,
      "step": 34875
    },
    {
      "entropy": 6.029232406616211,
      "epoch": 3.048151708468059,
      "grad_norm": 1.375,
      "learning_rate": 0.00040799175324781164,
      "loss": 5.2637,
      "mean_token_accuracy": 0.19982565492391585,
      "num_tokens": 70674635.0,
      "step": 34880
    },
    {
      "entropy": 5.985063171386718,
      "epoch": 3.0485886568207636,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000407966616901092,
      "loss": 5.1958,
      "mean_token_accuracy": 0.21363237649202346,
      "num_tokens": 70684633.0,
      "step": 34885
    },
    {
      "entropy": 5.994528293609619,
      "epoch": 3.0490256051734685,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004079414780040029,
      "loss": 5.2727,
      "mean_token_accuracy": 0.20997551828622818,
      "num_tokens": 70694416.0,
      "step": 34890
    },
    {
      "entropy": 6.015211915969848,
      "epoch": 3.049462553526173,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004079163365570267,
      "loss": 5.2189,
      "mean_token_accuracy": 0.20782818794250488,
      "num_tokens": 70703373.0,
      "step": 34895
    },
    {
      "entropy": 5.988285255432129,
      "epoch": 3.049899501878878,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004078911925606455,
      "loss": 5.1989,
      "mean_token_accuracy": 0.2105731710791588,
      "num_tokens": 70713868.0,
      "step": 34900
    },
    {
      "entropy": 6.080082654953003,
      "epoch": 3.050336450231583,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004078660460153415,
      "loss": 5.3467,
      "mean_token_accuracy": 0.19486829787492752,
      "num_tokens": 70724306.0,
      "step": 34905
    },
    {
      "entropy": 6.025181293487549,
      "epoch": 3.0507733985842873,
      "grad_norm": 1.421875,
      "learning_rate": 0.00040784089692159724,
      "loss": 5.2716,
      "mean_token_accuracy": 0.2095711499452591,
      "num_tokens": 70733923.0,
      "step": 34910
    },
    {
      "entropy": 6.047804594039917,
      "epoch": 3.0512103469369922,
      "grad_norm": 1.25,
      "learning_rate": 0.0004078157452798949,
      "loss": 5.3044,
      "mean_token_accuracy": 0.20435062050819397,
      "num_tokens": 70744964.0,
      "step": 34915
    },
    {
      "entropy": 6.006855058670044,
      "epoch": 3.0516472952896967,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00040779059109071706,
      "loss": 5.179,
      "mean_token_accuracy": 0.21083289086818696,
      "num_tokens": 70754639.0,
      "step": 34920
    },
    {
      "entropy": 6.050586795806884,
      "epoch": 3.0520842436424016,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040776543435454597,
      "loss": 5.2641,
      "mean_token_accuracy": 0.20957914888858795,
      "num_tokens": 70764810.0,
      "step": 34925
    },
    {
      "entropy": 6.048526334762573,
      "epoch": 3.052521191995106,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004077402750718643,
      "loss": 5.2566,
      "mean_token_accuracy": 0.20650407075881957,
      "num_tokens": 70774698.0,
      "step": 34930
    },
    {
      "entropy": 6.011918354034424,
      "epoch": 3.052958140347811,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004077151132431546,
      "loss": 5.196,
      "mean_token_accuracy": 0.2114843964576721,
      "num_tokens": 70785522.0,
      "step": 34935
    },
    {
      "entropy": 6.001184606552124,
      "epoch": 3.0533950887005155,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004076899488688996,
      "loss": 5.2809,
      "mean_token_accuracy": 0.20535277128219603,
      "num_tokens": 70797001.0,
      "step": 34940
    },
    {
      "entropy": 5.987829256057739,
      "epoch": 3.0538320370532204,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040766478194958176,
      "loss": 5.2767,
      "mean_token_accuracy": 0.20583926737308503,
      "num_tokens": 70807247.0,
      "step": 34945
    },
    {
      "entropy": 6.049748516082763,
      "epoch": 3.054268985405925,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004076396124856839,
      "loss": 5.3164,
      "mean_token_accuracy": 0.1980672597885132,
      "num_tokens": 70818269.0,
      "step": 34950
    },
    {
      "entropy": 6.1226812362670895,
      "epoch": 3.0547059337586298,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004076144404776888,
      "loss": 5.3258,
      "mean_token_accuracy": 0.20416467040777206,
      "num_tokens": 70827759.0,
      "step": 34955
    },
    {
      "entropy": 6.01351900100708,
      "epoch": 3.0551428821113342,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004075892659260792,
      "loss": 5.2532,
      "mean_token_accuracy": 0.20221646577119828,
      "num_tokens": 70838296.0,
      "step": 34960
    },
    {
      "entropy": 5.9727788925170895,
      "epoch": 3.055579830464039,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040756408883133797,
      "loss": 5.2298,
      "mean_token_accuracy": 0.20507645905017852,
      "num_tokens": 70848364.0,
      "step": 34965
    },
    {
      "entropy": 6.0448915481567385,
      "epoch": 3.056016778816744,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004075389091939481,
      "loss": 5.2287,
      "mean_token_accuracy": 0.21136355251073838,
      "num_tokens": 70858845.0,
      "step": 34970
    },
    {
      "entropy": 6.006518316268921,
      "epoch": 3.0564537271694485,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004075137270143925,
      "loss": 5.285,
      "mean_token_accuracy": 0.21228934824466705,
      "num_tokens": 70868976.0,
      "step": 34975
    },
    {
      "entropy": 5.947638845443725,
      "epoch": 3.0568906755221534,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004074885422931542,
      "loss": 5.2174,
      "mean_token_accuracy": 0.21801482141017914,
      "num_tokens": 70878287.0,
      "step": 34980
    },
    {
      "entropy": 6.005930471420288,
      "epoch": 3.057327623874858,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040746335503071625,
      "loss": 5.2609,
      "mean_token_accuracy": 0.21024437993764877,
      "num_tokens": 70889374.0,
      "step": 34985
    },
    {
      "entropy": 5.954008674621582,
      "epoch": 3.057764572227563,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040743816522756165,
      "loss": 5.2296,
      "mean_token_accuracy": 0.21749753206968309,
      "num_tokens": 70899741.0,
      "step": 34990
    },
    {
      "entropy": 5.999426937103271,
      "epoch": 3.0582015205802673,
      "grad_norm": 1.25,
      "learning_rate": 0.00040741297288417373,
      "loss": 5.2985,
      "mean_token_accuracy": 0.205663301050663,
      "num_tokens": 70909962.0,
      "step": 34995
    },
    {
      "entropy": 6.019423484802246,
      "epoch": 3.058638468932972,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004073877780010356,
      "loss": 5.221,
      "mean_token_accuracy": 0.2104756087064743,
      "num_tokens": 70921129.0,
      "step": 35000
    },
    {
      "entropy": 6.007030200958252,
      "epoch": 3.0590754172856767,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004073625805786305,
      "loss": 5.2298,
      "mean_token_accuracy": 0.2117555946111679,
      "num_tokens": 70930492.0,
      "step": 35005
    },
    {
      "entropy": 6.0727697849273685,
      "epoch": 3.0595123656383816,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004073373806174417,
      "loss": 5.2332,
      "mean_token_accuracy": 0.20677455365657807,
      "num_tokens": 70940667.0,
      "step": 35010
    },
    {
      "entropy": 6.0346214294433596,
      "epoch": 3.059949313991086,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004073121781179527,
      "loss": 5.2013,
      "mean_token_accuracy": 0.2129579946398735,
      "num_tokens": 70949815.0,
      "step": 35015
    },
    {
      "entropy": 6.04792275428772,
      "epoch": 3.060386262343791,
      "grad_norm": 1.28125,
      "learning_rate": 0.00040728697308064674,
      "loss": 5.3304,
      "mean_token_accuracy": 0.19390431195497512,
      "num_tokens": 70959931.0,
      "step": 35020
    },
    {
      "entropy": 6.013092041015625,
      "epoch": 3.060823210696496,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004072617655060073,
      "loss": 5.216,
      "mean_token_accuracy": 0.20839598923921585,
      "num_tokens": 70970204.0,
      "step": 35025
    },
    {
      "entropy": 6.064183759689331,
      "epoch": 3.0612601590492003,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040723655539451787,
      "loss": 5.3159,
      "mean_token_accuracy": 0.20248118191957473,
      "num_tokens": 70981841.0,
      "step": 35030
    },
    {
      "entropy": 5.9978453636169435,
      "epoch": 3.0616971074019053,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004072113427466621,
      "loss": 5.2541,
      "mean_token_accuracy": 0.2088874027132988,
      "num_tokens": 70992281.0,
      "step": 35035
    },
    {
      "entropy": 5.938990688323974,
      "epoch": 3.0621340557546097,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00040718612756292347,
      "loss": 5.1834,
      "mean_token_accuracy": 0.21461491286754608,
      "num_tokens": 71002737.0,
      "step": 35040
    },
    {
      "entropy": 6.01088581085205,
      "epoch": 3.0625710041073146,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004071609098437856,
      "loss": 5.2493,
      "mean_token_accuracy": 0.20892786979675293,
      "num_tokens": 71013604.0,
      "step": 35045
    },
    {
      "entropy": 5.969170379638672,
      "epoch": 3.063007952460019,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00040713568958973225,
      "loss": 5.2268,
      "mean_token_accuracy": 0.20894430875778197,
      "num_tokens": 71023278.0,
      "step": 35050
    },
    {
      "entropy": 6.012540817260742,
      "epoch": 3.063444900812724,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00040711046680124715,
      "loss": 5.2467,
      "mean_token_accuracy": 0.20763417929410935,
      "num_tokens": 71033823.0,
      "step": 35055
    },
    {
      "entropy": 5.983752346038818,
      "epoch": 3.0638818491654285,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004070852414788141,
      "loss": 5.281,
      "mean_token_accuracy": 0.20605590045452118,
      "num_tokens": 71043802.0,
      "step": 35060
    },
    {
      "entropy": 6.0612953186035154,
      "epoch": 3.0643187975181334,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004070600136229168,
      "loss": 5.2764,
      "mean_token_accuracy": 0.2031553417444229,
      "num_tokens": 71054152.0,
      "step": 35065
    },
    {
      "entropy": 5.992659950256348,
      "epoch": 3.064755745870838,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00040703478323403944,
      "loss": 5.1898,
      "mean_token_accuracy": 0.2168510377407074,
      "num_tokens": 71065346.0,
      "step": 35070
    },
    {
      "entropy": 6.005043172836304,
      "epoch": 3.065192694223543,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00040700955031266556,
      "loss": 5.2469,
      "mean_token_accuracy": 0.20443396866321564,
      "num_tokens": 71074855.0,
      "step": 35075
    },
    {
      "entropy": 6.006979513168335,
      "epoch": 3.0656296425762477,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004069843148592794,
      "loss": 5.2068,
      "mean_token_accuracy": 0.2135735958814621,
      "num_tokens": 71084372.0,
      "step": 35080
    },
    {
      "entropy": 6.035584402084351,
      "epoch": 3.066066590928952,
      "grad_norm": 1.359375,
      "learning_rate": 0.000406959076874365,
      "loss": 5.2708,
      "mean_token_accuracy": 0.21010348945856094,
      "num_tokens": 71093365.0,
      "step": 35085
    },
    {
      "entropy": 5.930618619918823,
      "epoch": 3.066503539281657,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00040693383635840625,
      "loss": 5.1807,
      "mean_token_accuracy": 0.21513493061065675,
      "num_tokens": 71102748.0,
      "step": 35090
    },
    {
      "entropy": 6.021502923965454,
      "epoch": 3.0669404876343616,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004069085933118874,
      "loss": 5.2836,
      "mean_token_accuracy": 0.20839932560920715,
      "num_tokens": 71113878.0,
      "step": 35095
    },
    {
      "entropy": 6.102246809005737,
      "epoch": 3.0673774359870665,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004068833477352927,
      "loss": 5.3152,
      "mean_token_accuracy": 0.20122425705194474,
      "num_tokens": 71124849.0,
      "step": 35100
    },
    {
      "entropy": 6.007090997695923,
      "epoch": 3.067814384339771,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00040685809962910614,
      "loss": 5.2311,
      "mean_token_accuracy": 0.2091096132993698,
      "num_tokens": 71134064.0,
      "step": 35105
    },
    {
      "entropy": 5.932328557968139,
      "epoch": 3.068251332692476,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040683284899381225,
      "loss": 5.1193,
      "mean_token_accuracy": 0.2163292109966278,
      "num_tokens": 71144270.0,
      "step": 35110
    },
    {
      "entropy": 5.968714809417724,
      "epoch": 3.0686882810451803,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00040680759582989516,
      "loss": 5.1438,
      "mean_token_accuracy": 0.21448241621255876,
      "num_tokens": 71153949.0,
      "step": 35115
    },
    {
      "entropy": 6.0009605884552,
      "epoch": 3.0691252293978852,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040678234013783945,
      "loss": 5.2472,
      "mean_token_accuracy": 0.20700473934412003,
      "num_tokens": 71164858.0,
      "step": 35120
    },
    {
      "entropy": 5.985754728317261,
      "epoch": 3.0695621777505897,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004067570819181293,
      "loss": 5.1567,
      "mean_token_accuracy": 0.21426850706338882,
      "num_tokens": 71174832.0,
      "step": 35125
    },
    {
      "entropy": 5.996290874481201,
      "epoch": 3.0699991261032946,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004067318211712492,
      "loss": 5.2589,
      "mean_token_accuracy": 0.20684006065130234,
      "num_tokens": 71185089.0,
      "step": 35130
    },
    {
      "entropy": 6.044051694869995,
      "epoch": 3.070436074455999,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004067065578976839,
      "loss": 5.223,
      "mean_token_accuracy": 0.2047010287642479,
      "num_tokens": 71195889.0,
      "step": 35135
    },
    {
      "entropy": 6.095178651809692,
      "epoch": 3.070873022808704,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00040668129209791773,
      "loss": 5.2845,
      "mean_token_accuracy": 0.20530079454183578,
      "num_tokens": 71205825.0,
      "step": 35140
    },
    {
      "entropy": 6.013595199584961,
      "epoch": 3.071309971161409,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004066560237724354,
      "loss": 5.1679,
      "mean_token_accuracy": 0.2130955696105957,
      "num_tokens": 71215201.0,
      "step": 35145
    },
    {
      "entropy": 5.909121465682984,
      "epoch": 3.0717469195141134,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004066307529217215,
      "loss": 5.1369,
      "mean_token_accuracy": 0.2190663442015648,
      "num_tokens": 71225329.0,
      "step": 35150
    },
    {
      "entropy": 5.966169023513794,
      "epoch": 3.0721838678668183,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040660547954626076,
      "loss": 5.1258,
      "mean_token_accuracy": 0.21892655342817308,
      "num_tokens": 71235301.0,
      "step": 35155
    },
    {
      "entropy": 6.027197265625,
      "epoch": 3.0726208162195228,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040658020364653803,
      "loss": 5.2464,
      "mean_token_accuracy": 0.2022918403148651,
      "num_tokens": 71244991.0,
      "step": 35160
    },
    {
      "entropy": 6.008800411224366,
      "epoch": 3.0730577645722277,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000406554925223038,
      "loss": 5.2361,
      "mean_token_accuracy": 0.20839802473783492,
      "num_tokens": 71255480.0,
      "step": 35165
    },
    {
      "entropy": 6.097862434387207,
      "epoch": 3.073494712924932,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004065296442762456,
      "loss": 5.333,
      "mean_token_accuracy": 0.2035711884498596,
      "num_tokens": 71265996.0,
      "step": 35170
    },
    {
      "entropy": 6.084036874771118,
      "epoch": 3.073931661277637,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040650436080664576,
      "loss": 5.2197,
      "mean_token_accuracy": 0.21090051978826524,
      "num_tokens": 71277293.0,
      "step": 35175
    },
    {
      "entropy": 6.001480007171631,
      "epoch": 3.0743686096303415,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00040647907481472333,
      "loss": 5.1966,
      "mean_token_accuracy": 0.20719846338033676,
      "num_tokens": 71285903.0,
      "step": 35180
    },
    {
      "entropy": 5.944966268539429,
      "epoch": 3.0748055579830464,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004064537863009633,
      "loss": 5.1738,
      "mean_token_accuracy": 0.21692296415567397,
      "num_tokens": 71294979.0,
      "step": 35185
    },
    {
      "entropy": 5.971688365936279,
      "epoch": 3.075242506335751,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040642849526585087,
      "loss": 5.2117,
      "mean_token_accuracy": 0.2168597474694252,
      "num_tokens": 71306223.0,
      "step": 35190
    },
    {
      "entropy": 6.094396352767944,
      "epoch": 3.075679454688456,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004064032017098709,
      "loss": 5.324,
      "mean_token_accuracy": 0.19943636506795884,
      "num_tokens": 71316984.0,
      "step": 35195
    },
    {
      "entropy": 6.093851375579834,
      "epoch": 3.0761164030411607,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00040637790563350886,
      "loss": 5.2986,
      "mean_token_accuracy": 0.2124711737036705,
      "num_tokens": 71327855.0,
      "step": 35200
    },
    {
      "entropy": 6.032990264892578,
      "epoch": 3.076553351393865,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004063526070372496,
      "loss": 5.2104,
      "mean_token_accuracy": 0.21284947395324708,
      "num_tokens": 71338115.0,
      "step": 35205
    },
    {
      "entropy": 5.947810316085816,
      "epoch": 3.07699029974657,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004063273059215785,
      "loss": 5.166,
      "mean_token_accuracy": 0.2148173600435257,
      "num_tokens": 71348533.0,
      "step": 35210
    },
    {
      "entropy": 5.8932710647583,
      "epoch": 3.0774272480992746,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004063020022869809,
      "loss": 5.0323,
      "mean_token_accuracy": 0.23159716725349427,
      "num_tokens": 71358368.0,
      "step": 35215
    },
    {
      "entropy": 6.014525365829468,
      "epoch": 3.0778641964519795,
      "grad_norm": 1.296875,
      "learning_rate": 0.00040627669613394206,
      "loss": 5.2585,
      "mean_token_accuracy": 0.2048746168613434,
      "num_tokens": 71368443.0,
      "step": 35220
    },
    {
      "entropy": 6.034038734436035,
      "epoch": 3.078301144804684,
      "grad_norm": 1.34375,
      "learning_rate": 0.00040625138746294747,
      "loss": 5.2373,
      "mean_token_accuracy": 0.20478758960962296,
      "num_tokens": 71378508.0,
      "step": 35225
    },
    {
      "entropy": 6.006376695632935,
      "epoch": 3.078738093157389,
      "grad_norm": 1.25,
      "learning_rate": 0.0004062260762744825,
      "loss": 5.2231,
      "mean_token_accuracy": 0.20687633901834487,
      "num_tokens": 71388512.0,
      "step": 35230
    },
    {
      "entropy": 6.026971101760864,
      "epoch": 3.0791750415100934,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040620076256903256,
      "loss": 5.2427,
      "mean_token_accuracy": 0.20675787925720215,
      "num_tokens": 71397908.0,
      "step": 35235
    },
    {
      "entropy": 6.087830829620361,
      "epoch": 3.0796119898627983,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004061754463470833,
      "loss": 5.3157,
      "mean_token_accuracy": 0.2001950666308403,
      "num_tokens": 71408287.0,
      "step": 35240
    },
    {
      "entropy": 6.031873273849487,
      "epoch": 3.0800489382155027,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004061501276091202,
      "loss": 5.3069,
      "mean_token_accuracy": 0.20525415539741515,
      "num_tokens": 71419227.0,
      "step": 35245
    },
    {
      "entropy": 6.068632650375366,
      "epoch": 3.0804858865682077,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000406124806355629,
      "loss": 5.2439,
      "mean_token_accuracy": 0.21171698868274688,
      "num_tokens": 71429398.0,
      "step": 35250
    },
    {
      "entropy": 5.995148611068726,
      "epoch": 3.080922834920912,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040609948258709517,
      "loss": 5.2152,
      "mean_token_accuracy": 0.21361487060785295,
      "num_tokens": 71440185.0,
      "step": 35255
    },
    {
      "entropy": 5.916224765777588,
      "epoch": 3.081359783273617,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004060741563040047,
      "loss": 5.2138,
      "mean_token_accuracy": 0.21860401779413224,
      "num_tokens": 71451160.0,
      "step": 35260
    },
    {
      "entropy": 6.094847345352173,
      "epoch": 3.081796731626322,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004060488275068432,
      "loss": 5.2843,
      "mean_token_accuracy": 0.20516441464424134,
      "num_tokens": 71461459.0,
      "step": 35265
    },
    {
      "entropy": 6.014743328094482,
      "epoch": 3.0822336799790264,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00040602349619609646,
      "loss": 5.218,
      "mean_token_accuracy": 0.20733260214328766,
      "num_tokens": 71470806.0,
      "step": 35270
    },
    {
      "entropy": 5.944396114349365,
      "epoch": 3.0826706283317313,
      "grad_norm": 1.3125,
      "learning_rate": 0.00040599816237225044,
      "loss": 5.194,
      "mean_token_accuracy": 0.21230337917804717,
      "num_tokens": 71480929.0,
      "step": 35275
    },
    {
      "entropy": 6.029759931564331,
      "epoch": 3.083107576684436,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00040597282603579103,
      "loss": 5.2277,
      "mean_token_accuracy": 0.21215838491916655,
      "num_tokens": 71490360.0,
      "step": 35280
    },
    {
      "entropy": 5.998914194107056,
      "epoch": 3.0835445250371407,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040594748718720424,
      "loss": 5.2937,
      "mean_token_accuracy": 0.20499097853899,
      "num_tokens": 71500355.0,
      "step": 35285
    },
    {
      "entropy": 5.984378051757813,
      "epoch": 3.083981473389845,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040592214582697595,
      "loss": 5.2006,
      "mean_token_accuracy": 0.21517712920904158,
      "num_tokens": 71509581.0,
      "step": 35290
    },
    {
      "entropy": 5.996991395950317,
      "epoch": 3.08441842174255,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004058968019555923,
      "loss": 5.2695,
      "mean_token_accuracy": 0.20258226245641708,
      "num_tokens": 71519967.0,
      "step": 35295
    },
    {
      "entropy": 5.967708015441895,
      "epoch": 3.0848553700952546,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040587145557353943,
      "loss": 5.1869,
      "mean_token_accuracy": 0.21503883302211763,
      "num_tokens": 71530619.0,
      "step": 35300
    },
    {
      "entropy": 6.028796768188476,
      "epoch": 3.0852923184479595,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004058461066813035,
      "loss": 5.2745,
      "mean_token_accuracy": 0.20182029008865357,
      "num_tokens": 71541682.0,
      "step": 35305
    },
    {
      "entropy": 5.988647985458374,
      "epoch": 3.085729266800664,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00040582075527937066,
      "loss": 5.2164,
      "mean_token_accuracy": 0.20914653390645982,
      "num_tokens": 71552971.0,
      "step": 35310
    },
    {
      "entropy": 5.94437460899353,
      "epoch": 3.086166215153369,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004057954013682272,
      "loss": 5.228,
      "mean_token_accuracy": 0.20956174582242965,
      "num_tokens": 71562932.0,
      "step": 35315
    },
    {
      "entropy": 6.017757701873779,
      "epoch": 3.0866031635060738,
      "grad_norm": 1.421875,
      "learning_rate": 0.00040577004494835934,
      "loss": 5.1865,
      "mean_token_accuracy": 0.21306897401809693,
      "num_tokens": 71572662.0,
      "step": 35320
    },
    {
      "entropy": 6.003777551651001,
      "epoch": 3.0870401118587782,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00040574468602025355,
      "loss": 5.2069,
      "mean_token_accuracy": 0.21407117694616318,
      "num_tokens": 71582503.0,
      "step": 35325
    },
    {
      "entropy": 6.065600490570068,
      "epoch": 3.087477060211483,
      "grad_norm": 1.328125,
      "learning_rate": 0.00040571932458439617,
      "loss": 5.2561,
      "mean_token_accuracy": 0.20728277117013932,
      "num_tokens": 71592261.0,
      "step": 35330
    },
    {
      "entropy": 5.983888769149781,
      "epoch": 3.0879140085641876,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040569396064127355,
      "loss": 5.2103,
      "mean_token_accuracy": 0.2164016455411911,
      "num_tokens": 71602514.0,
      "step": 35335
    },
    {
      "entropy": 5.96472659111023,
      "epoch": 3.0883509569168925,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004056685941913724,
      "loss": 5.2296,
      "mean_token_accuracy": 0.21197912245988845,
      "num_tokens": 71612912.0,
      "step": 35340
    },
    {
      "entropy": 5.953876638412476,
      "epoch": 3.088787905269597,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00040564322523517905,
      "loss": 5.1545,
      "mean_token_accuracy": 0.21208005845546724,
      "num_tokens": 71622673.0,
      "step": 35345
    },
    {
      "entropy": 6.041848802566529,
      "epoch": 3.089224853622302,
      "grad_norm": 1.25,
      "learning_rate": 0.0004056178537731803,
      "loss": 5.2106,
      "mean_token_accuracy": 0.20804373025894166,
      "num_tokens": 71633315.0,
      "step": 35350
    },
    {
      "entropy": 6.016338300704956,
      "epoch": 3.0896618019750064,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004055924798058626,
      "loss": 5.2873,
      "mean_token_accuracy": 0.2044851914048195,
      "num_tokens": 71644159.0,
      "step": 35355
    },
    {
      "entropy": 6.006757307052612,
      "epoch": 3.0900987503277113,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040556710333371275,
      "loss": 5.2115,
      "mean_token_accuracy": 0.2094744548201561,
      "num_tokens": 71654227.0,
      "step": 35360
    },
    {
      "entropy": 5.970843696594239,
      "epoch": 3.0905356986804158,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004055417243572174,
      "loss": 5.1829,
      "mean_token_accuracy": 0.21126883774995803,
      "num_tokens": 71664711.0,
      "step": 35365
    },
    {
      "entropy": 6.065037584304809,
      "epoch": 3.0909726470331207,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040551634287686343,
      "loss": 5.2381,
      "mean_token_accuracy": 0.21225452423095703,
      "num_tokens": 71674880.0,
      "step": 35370
    },
    {
      "entropy": 6.082315731048584,
      "epoch": 3.0914095953858256,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004054909588931375,
      "loss": 5.2736,
      "mean_token_accuracy": 0.20332275629043578,
      "num_tokens": 71685386.0,
      "step": 35375
    },
    {
      "entropy": 6.047211074829102,
      "epoch": 3.09184654373853,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004054655724065268,
      "loss": 5.3581,
      "mean_token_accuracy": 0.19970939606428145,
      "num_tokens": 71696008.0,
      "step": 35380
    },
    {
      "entropy": 5.994264936447143,
      "epoch": 3.092283492091235,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004054401834175179,
      "loss": 5.1855,
      "mean_token_accuracy": 0.2093050852417946,
      "num_tokens": 71705548.0,
      "step": 35385
    },
    {
      "entropy": 5.966810512542724,
      "epoch": 3.0927204404439395,
      "grad_norm": 1.25,
      "learning_rate": 0.00040541479192659797,
      "loss": 5.2251,
      "mean_token_accuracy": 0.2104998931288719,
      "num_tokens": 71716891.0,
      "step": 35390
    },
    {
      "entropy": 6.05039381980896,
      "epoch": 3.0931573887966444,
      "grad_norm": 1.40625,
      "learning_rate": 0.00040538939793425394,
      "loss": 5.2824,
      "mean_token_accuracy": 0.20533577054739,
      "num_tokens": 71727954.0,
      "step": 35395
    },
    {
      "entropy": 5.997773551940918,
      "epoch": 3.093594337149349,
      "grad_norm": 1.203125,
      "learning_rate": 0.000405364001440973,
      "loss": 5.2873,
      "mean_token_accuracy": 0.20644728392362593,
      "num_tokens": 71738282.0,
      "step": 35400
    },
    {
      "entropy": 6.0648092269897464,
      "epoch": 3.0940312855020538,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004053386024472422,
      "loss": 5.2995,
      "mean_token_accuracy": 0.2001652494072914,
      "num_tokens": 71747895.0,
      "step": 35405
    },
    {
      "entropy": 6.007932376861572,
      "epoch": 3.094468233854758,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004053132009535487,
      "loss": 5.2492,
      "mean_token_accuracy": 0.20851772278547287,
      "num_tokens": 71758107.0,
      "step": 35410
    },
    {
      "entropy": 6.085710525512695,
      "epoch": 3.094905182207463,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004052877969603797,
      "loss": 5.3224,
      "mean_token_accuracy": 0.2002759039402008,
      "num_tokens": 71767661.0,
      "step": 35415
    },
    {
      "entropy": 6.024410438537598,
      "epoch": 3.0953421305601676,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004052623904682225,
      "loss": 5.266,
      "mean_token_accuracy": 0.2032524585723877,
      "num_tokens": 71777908.0,
      "step": 35420
    },
    {
      "entropy": 6.0093395709991455,
      "epoch": 3.0957790789128725,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00040523698147756443,
      "loss": 5.2093,
      "mean_token_accuracy": 0.21115070432424546,
      "num_tokens": 71786809.0,
      "step": 35425
    },
    {
      "entropy": 5.9380027770996096,
      "epoch": 3.0962160272655774,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040521156998889283,
      "loss": 5.233,
      "mean_token_accuracy": 0.21111720502376558,
      "num_tokens": 71797299.0,
      "step": 35430
    },
    {
      "entropy": 6.036979007720947,
      "epoch": 3.096652975618282,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00040518615600269506,
      "loss": 5.2142,
      "mean_token_accuracy": 0.21410403698682784,
      "num_tokens": 71807565.0,
      "step": 35435
    },
    {
      "entropy": 6.011176824569702,
      "epoch": 3.097089923970987,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004051607395194586,
      "loss": 5.2151,
      "mean_token_accuracy": 0.2124226361513138,
      "num_tokens": 71817288.0,
      "step": 35440
    },
    {
      "entropy": 6.02723069190979,
      "epoch": 3.0975268723236913,
      "grad_norm": 1.40625,
      "learning_rate": 0.00040513532053967086,
      "loss": 5.2123,
      "mean_token_accuracy": 0.21019705682992934,
      "num_tokens": 71827371.0,
      "step": 35445
    },
    {
      "entropy": 6.113615560531616,
      "epoch": 3.097963820676396,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040510989906381955,
      "loss": 5.3674,
      "mean_token_accuracy": 0.19996369183063506,
      "num_tokens": 71837872.0,
      "step": 35450
    },
    {
      "entropy": 5.953599882125855,
      "epoch": 3.0984007690291007,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040508447509239224,
      "loss": 5.1675,
      "mean_token_accuracy": 0.21532609909772873,
      "num_tokens": 71848507.0,
      "step": 35455
    },
    {
      "entropy": 6.0401444911956785,
      "epoch": 3.0988377173818056,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004050590486258764,
      "loss": 5.2713,
      "mean_token_accuracy": 0.20360045582056047,
      "num_tokens": 71860495.0,
      "step": 35460
    },
    {
      "entropy": 6.0204308986663815,
      "epoch": 3.09927466573451,
      "grad_norm": 1.8125,
      "learning_rate": 0.00040503361966476005,
      "loss": 5.1271,
      "mean_token_accuracy": 0.22652624398469925,
      "num_tokens": 71870081.0,
      "step": 35465
    },
    {
      "entropy": 6.023430919647216,
      "epoch": 3.099711614087215,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004050081882095307,
      "loss": 5.2421,
      "mean_token_accuracy": 0.20590901076793672,
      "num_tokens": 71880072.0,
      "step": 35470
    },
    {
      "entropy": 5.988133144378662,
      "epoch": 3.1001485624399194,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004049827542606761,
      "loss": 5.2058,
      "mean_token_accuracy": 0.20451007634401322,
      "num_tokens": 71889225.0,
      "step": 35475
    },
    {
      "entropy": 6.060214900970459,
      "epoch": 3.1005855107926243,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004049573178186842,
      "loss": 5.2435,
      "mean_token_accuracy": 0.21025344282388686,
      "num_tokens": 71898601.0,
      "step": 35480
    },
    {
      "entropy": 6.045565414428711,
      "epoch": 3.101022459145329,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040493187888404283,
      "loss": 5.251,
      "mean_token_accuracy": 0.20896627753973007,
      "num_tokens": 71909267.0,
      "step": 35485
    },
    {
      "entropy": 6.046857404708862,
      "epoch": 3.1014594074980337,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00040490643745724,
      "loss": 5.2595,
      "mean_token_accuracy": 0.20420864671468736,
      "num_tokens": 71919066.0,
      "step": 35490
    },
    {
      "entropy": 6.021708011627197,
      "epoch": 3.1018963558507386,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004048809935387635,
      "loss": 5.2088,
      "mean_token_accuracy": 0.211625537276268,
      "num_tokens": 71929067.0,
      "step": 35495
    },
    {
      "entropy": 6.029827833175659,
      "epoch": 3.102333304203443,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00040485554712910155,
      "loss": 5.1615,
      "mean_token_accuracy": 0.21215946227312088,
      "num_tokens": 71938044.0,
      "step": 35500
    },
    {
      "entropy": 5.964566230773926,
      "epoch": 3.102770252556148,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00040483009822874216,
      "loss": 5.2627,
      "mean_token_accuracy": 0.21246422976255416,
      "num_tokens": 71947714.0,
      "step": 35505
    },
    {
      "entropy": 5.969930171966553,
      "epoch": 3.1032072009088525,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004048046468381735,
      "loss": 5.2098,
      "mean_token_accuracy": 0.2128792092204094,
      "num_tokens": 71957660.0,
      "step": 35510
    },
    {
      "entropy": 5.993561935424805,
      "epoch": 3.1036441492615574,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004047791929578837,
      "loss": 5.1625,
      "mean_token_accuracy": 0.21778318136930466,
      "num_tokens": 71967063.0,
      "step": 35515
    },
    {
      "entropy": 5.968338584899902,
      "epoch": 3.104081097614262,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004047537365883609,
      "loss": 5.2572,
      "mean_token_accuracy": 0.20555483549833298,
      "num_tokens": 71977331.0,
      "step": 35520
    },
    {
      "entropy": 5.940487337112427,
      "epoch": 3.104518045966967,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00040472827773009346,
      "loss": 5.1698,
      "mean_token_accuracy": 0.21409732103347778,
      "num_tokens": 71987047.0,
      "step": 35525
    },
    {
      "entropy": 6.018299341201782,
      "epoch": 3.1049549943196713,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00040470281638356974,
      "loss": 5.1894,
      "mean_token_accuracy": 0.20704491287469864,
      "num_tokens": 71997057.0,
      "step": 35530
    },
    {
      "entropy": 6.0389340877532955,
      "epoch": 3.105391942672376,
      "grad_norm": 1.328125,
      "learning_rate": 0.000404677352549278,
      "loss": 5.2915,
      "mean_token_accuracy": 0.20600130409002304,
      "num_tokens": 72007092.0,
      "step": 35535
    },
    {
      "entropy": 6.057676601409912,
      "epoch": 3.1058288910250806,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00040465188622770663,
      "loss": 5.3308,
      "mean_token_accuracy": 0.2038910984992981,
      "num_tokens": 72016847.0,
      "step": 35540
    },
    {
      "entropy": 6.008616018295288,
      "epoch": 3.1062658393777856,
      "grad_norm": 1.25,
      "learning_rate": 0.0004046264174193443,
      "loss": 5.2172,
      "mean_token_accuracy": 0.20939266830682754,
      "num_tokens": 72026706.0,
      "step": 35545
    },
    {
      "entropy": 6.086511611938477,
      "epoch": 3.1067027877304905,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004046009461246792,
      "loss": 5.3596,
      "mean_token_accuracy": 0.1958037570118904,
      "num_tokens": 72036358.0,
      "step": 35550
    },
    {
      "entropy": 6.029675769805908,
      "epoch": 3.107139736083195,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004045754723442001,
      "loss": 5.3137,
      "mean_token_accuracy": 0.2048747718334198,
      "num_tokens": 72049156.0,
      "step": 35555
    },
    {
      "entropy": 5.995788192749023,
      "epoch": 3.1075766844359,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004045499960783955,
      "loss": 5.1868,
      "mean_token_accuracy": 0.20985940247774124,
      "num_tokens": 72058986.0,
      "step": 35560
    },
    {
      "entropy": 5.99634199142456,
      "epoch": 3.1080136327886043,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004045245173277543,
      "loss": 5.2112,
      "mean_token_accuracy": 0.21156397759914397,
      "num_tokens": 72069141.0,
      "step": 35565
    },
    {
      "entropy": 5.965301179885865,
      "epoch": 3.1084505811413092,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040449903609276474,
      "loss": 5.1671,
      "mean_token_accuracy": 0.2237307295203209,
      "num_tokens": 72079009.0,
      "step": 35570
    },
    {
      "entropy": 6.0057525634765625,
      "epoch": 3.1088875294940137,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004044735523739159,
      "loss": 5.244,
      "mean_token_accuracy": 0.20591915249824524,
      "num_tokens": 72089551.0,
      "step": 35575
    },
    {
      "entropy": 6.032135343551635,
      "epoch": 3.1093244778467186,
      "grad_norm": 1.328125,
      "learning_rate": 0.00040444806617169655,
      "loss": 5.2147,
      "mean_token_accuracy": 0.2115800216794014,
      "num_tokens": 72098965.0,
      "step": 35580
    },
    {
      "entropy": 6.027977323532104,
      "epoch": 3.109761426199423,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040442257748659537,
      "loss": 5.2495,
      "mean_token_accuracy": 0.2112022042274475,
      "num_tokens": 72109568.0,
      "step": 35585
    },
    {
      "entropy": 6.0190633773803714,
      "epoch": 3.110198374552128,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040439708631910136,
      "loss": 5.254,
      "mean_token_accuracy": 0.2139955699443817,
      "num_tokens": 72119988.0,
      "step": 35590
    },
    {
      "entropy": 5.9544816493988035,
      "epoch": 3.1106353229048325,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004043715926697034,
      "loss": 5.3159,
      "mean_token_accuracy": 0.21003491431474686,
      "num_tokens": 72130515.0,
      "step": 35595
    },
    {
      "entropy": 6.068127107620239,
      "epoch": 3.1110722712575374,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004043460965388906,
      "loss": 5.2666,
      "mean_token_accuracy": 0.20150530338287354,
      "num_tokens": 72140362.0,
      "step": 35600
    },
    {
      "entropy": 6.036306619644165,
      "epoch": 3.111509219610242,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004043205979271518,
      "loss": 5.2396,
      "mean_token_accuracy": 0.2076708897948265,
      "num_tokens": 72150726.0,
      "step": 35605
    },
    {
      "entropy": 6.000241041183472,
      "epoch": 3.1119461679629468,
      "grad_norm": 1.375,
      "learning_rate": 0.0004042950968349762,
      "loss": 5.194,
      "mean_token_accuracy": 0.2156362861394882,
      "num_tokens": 72160264.0,
      "step": 35610
    },
    {
      "entropy": 5.962995338439941,
      "epoch": 3.1123831163156517,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004042695932628529,
      "loss": 5.2493,
      "mean_token_accuracy": 0.20720372647047042,
      "num_tokens": 72171407.0,
      "step": 35615
    },
    {
      "entropy": 6.039564180374145,
      "epoch": 3.112820064668356,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004042440872112711,
      "loss": 5.2897,
      "mean_token_accuracy": 0.2054512992501259,
      "num_tokens": 72182132.0,
      "step": 35620
    },
    {
      "entropy": 6.041181135177612,
      "epoch": 3.113257013021061,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040421857868071996,
      "loss": 5.2473,
      "mean_token_accuracy": 0.20322631150484086,
      "num_tokens": 72192129.0,
      "step": 35625
    },
    {
      "entropy": 5.960925817489624,
      "epoch": 3.1136939613737655,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040419306767168874,
      "loss": 5.2149,
      "mean_token_accuracy": 0.2144980624318123,
      "num_tokens": 72202053.0,
      "step": 35630
    },
    {
      "entropy": 6.013601016998291,
      "epoch": 3.1141309097264704,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004041675541846668,
      "loss": 5.1977,
      "mean_token_accuracy": 0.20967820584774016,
      "num_tokens": 72211352.0,
      "step": 35635
    },
    {
      "entropy": 6.009832525253296,
      "epoch": 3.114567858079175,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00040414203822014343,
      "loss": 5.1982,
      "mean_token_accuracy": 0.21173278242349625,
      "num_tokens": 72220816.0,
      "step": 35640
    },
    {
      "entropy": 6.045367670059204,
      "epoch": 3.11500480643188,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040411651977860817,
      "loss": 5.2888,
      "mean_token_accuracy": 0.19765387028455733,
      "num_tokens": 72231318.0,
      "step": 35645
    },
    {
      "entropy": 6.01841459274292,
      "epoch": 3.1154417547845843,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00040409099886055044,
      "loss": 5.2352,
      "mean_token_accuracy": 0.20714884400367736,
      "num_tokens": 72241035.0,
      "step": 35650
    },
    {
      "entropy": 6.086354732513428,
      "epoch": 3.115878703137289,
      "grad_norm": 1.25,
      "learning_rate": 0.0004040654754664596,
      "loss": 5.1781,
      "mean_token_accuracy": 0.2040769562125206,
      "num_tokens": 72249967.0,
      "step": 35655
    },
    {
      "entropy": 6.044242429733276,
      "epoch": 3.1163156514899937,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040403994959682546,
      "loss": 5.2704,
      "mean_token_accuracy": 0.20872316211462022,
      "num_tokens": 72260114.0,
      "step": 35660
    },
    {
      "entropy": 6.010709190368653,
      "epoch": 3.1167525998426986,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00040401442125213735,
      "loss": 5.199,
      "mean_token_accuracy": 0.2075783908367157,
      "num_tokens": 72270257.0,
      "step": 35665
    },
    {
      "entropy": 6.005086517333984,
      "epoch": 3.1171895481954035,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00040398889043288507,
      "loss": 5.1826,
      "mean_token_accuracy": 0.20774537324905396,
      "num_tokens": 72279628.0,
      "step": 35670
    },
    {
      "entropy": 6.006109142303467,
      "epoch": 3.117626496548108,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004039633571395583,
      "loss": 5.2167,
      "mean_token_accuracy": 0.2072230577468872,
      "num_tokens": 72289493.0,
      "step": 35675
    },
    {
      "entropy": 5.952965259552002,
      "epoch": 3.118063444900813,
      "grad_norm": 1.25,
      "learning_rate": 0.00040393782137264676,
      "loss": 5.122,
      "mean_token_accuracy": 0.2219855934381485,
      "num_tokens": 72300007.0,
      "step": 35680
    },
    {
      "entropy": 5.970151567459107,
      "epoch": 3.1185003932535174,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004039122831326402,
      "loss": 5.2096,
      "mean_token_accuracy": 0.21082223057746888,
      "num_tokens": 72310596.0,
      "step": 35685
    },
    {
      "entropy": 5.99593152999878,
      "epoch": 3.1189373416062223,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004038867424200285,
      "loss": 5.2794,
      "mean_token_accuracy": 0.2057865098118782,
      "num_tokens": 72320156.0,
      "step": 35690
    },
    {
      "entropy": 6.03639850616455,
      "epoch": 3.1193742899589267,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004038611992353016,
      "loss": 5.2276,
      "mean_token_accuracy": 0.2088891163468361,
      "num_tokens": 72329608.0,
      "step": 35695
    },
    {
      "entropy": 6.018807697296142,
      "epoch": 3.1198112383116317,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004038356535789494,
      "loss": 5.1882,
      "mean_token_accuracy": 0.21465280503034592,
      "num_tokens": 72339565.0,
      "step": 35700
    },
    {
      "entropy": 5.996279096603393,
      "epoch": 3.120248186664336,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040381010545146174,
      "loss": 5.289,
      "mean_token_accuracy": 0.20607105195522307,
      "num_tokens": 72349566.0,
      "step": 35705
    },
    {
      "entropy": 6.014826154708862,
      "epoch": 3.120685135017041,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004037845548533289,
      "loss": 5.1889,
      "mean_token_accuracy": 0.20875978618860244,
      "num_tokens": 72360366.0,
      "step": 35710
    },
    {
      "entropy": 6.108172988891601,
      "epoch": 3.1211220833697455,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040375900178504066,
      "loss": 5.3445,
      "mean_token_accuracy": 0.20464396923780442,
      "num_tokens": 72369893.0,
      "step": 35715
    },
    {
      "entropy": 5.999167156219483,
      "epoch": 3.1215590317224504,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004037334462470874,
      "loss": 5.2691,
      "mean_token_accuracy": 0.19976500123739244,
      "num_tokens": 72380484.0,
      "step": 35720
    },
    {
      "entropy": 6.008948516845703,
      "epoch": 3.1219959800751553,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004037078882399591,
      "loss": 5.1799,
      "mean_token_accuracy": 0.21157199442386626,
      "num_tokens": 72391088.0,
      "step": 35725
    },
    {
      "entropy": 6.103484582901001,
      "epoch": 3.12243292842786,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004036823277641462,
      "loss": 5.2994,
      "mean_token_accuracy": 0.20080119073390962,
      "num_tokens": 72401253.0,
      "step": 35730
    },
    {
      "entropy": 6.043921756744385,
      "epoch": 3.1228698767805647,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040365676482013877,
      "loss": 5.1889,
      "mean_token_accuracy": 0.21299372464418412,
      "num_tokens": 72411235.0,
      "step": 35735
    },
    {
      "entropy": 6.005966281890869,
      "epoch": 3.123306825133269,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004036311994084271,
      "loss": 5.2955,
      "mean_token_accuracy": 0.2023037329316139,
      "num_tokens": 72421387.0,
      "step": 35740
    },
    {
      "entropy": 5.9784962177276615,
      "epoch": 3.123743773485974,
      "grad_norm": 1.296875,
      "learning_rate": 0.00040360563152950173,
      "loss": 5.2714,
      "mean_token_accuracy": 0.20463391840457917,
      "num_tokens": 72431431.0,
      "step": 35745
    },
    {
      "entropy": 6.033056640625,
      "epoch": 3.1241807218386786,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004035800611838529,
      "loss": 5.2657,
      "mean_token_accuracy": 0.21063880920410155,
      "num_tokens": 72441286.0,
      "step": 35750
    },
    {
      "entropy": 6.042086362838745,
      "epoch": 3.1246176701913835,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00040355448837197115,
      "loss": 5.249,
      "mean_token_accuracy": 0.20741797983646393,
      "num_tokens": 72451661.0,
      "step": 35755
    },
    {
      "entropy": 6.041322946548462,
      "epoch": 3.125054618544088,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040352891309434693,
      "loss": 5.254,
      "mean_token_accuracy": 0.20081848353147508,
      "num_tokens": 72462013.0,
      "step": 35760
    },
    {
      "entropy": 6.003670406341553,
      "epoch": 3.125491566896793,
      "grad_norm": 1.375,
      "learning_rate": 0.0004035033353514708,
      "loss": 5.1741,
      "mean_token_accuracy": 0.20844481885433197,
      "num_tokens": 72472804.0,
      "step": 35765
    },
    {
      "entropy": 6.027034044265747,
      "epoch": 3.1259285152494973,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040347775514383344,
      "loss": 5.2387,
      "mean_token_accuracy": 0.20494974851608277,
      "num_tokens": 72482893.0,
      "step": 35770
    },
    {
      "entropy": 5.993804788589477,
      "epoch": 3.1263654636022022,
      "grad_norm": 1.3125,
      "learning_rate": 0.00040345217247192536,
      "loss": 5.2264,
      "mean_token_accuracy": 0.21125880628824234,
      "num_tokens": 72492882.0,
      "step": 35775
    },
    {
      "entropy": 5.916068744659424,
      "epoch": 3.126802411954907,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004034265873362373,
      "loss": 5.1503,
      "mean_token_accuracy": 0.2123771771788597,
      "num_tokens": 72502322.0,
      "step": 35780
    },
    {
      "entropy": 6.030395460128784,
      "epoch": 3.1272393603076116,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00040340099973726,
      "loss": 5.2527,
      "mean_token_accuracy": 0.2104560077190399,
      "num_tokens": 72511401.0,
      "step": 35785
    },
    {
      "entropy": 6.024174451828003,
      "epoch": 3.1276763086603165,
      "grad_norm": 1.375,
      "learning_rate": 0.00040337540967548427,
      "loss": 5.2669,
      "mean_token_accuracy": 0.20992031693458557,
      "num_tokens": 72522071.0,
      "step": 35790
    },
    {
      "entropy": 6.052229928970337,
      "epoch": 3.128113257013021,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00040334981715140086,
      "loss": 5.3052,
      "mean_token_accuracy": 0.20238305777311325,
      "num_tokens": 72532899.0,
      "step": 35795
    },
    {
      "entropy": 6.044043254852295,
      "epoch": 3.128550205365726,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040332422216550067,
      "loss": 5.2925,
      "mean_token_accuracy": 0.2032042071223259,
      "num_tokens": 72543372.0,
      "step": 35800
    },
    {
      "entropy": 6.0345314025878904,
      "epoch": 3.1289871537184304,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004032986247182748,
      "loss": 5.2462,
      "mean_token_accuracy": 0.2102514088153839,
      "num_tokens": 72553014.0,
      "step": 35805
    },
    {
      "entropy": 5.966872072219848,
      "epoch": 3.1294241020711353,
      "grad_norm": 1.296875,
      "learning_rate": 0.00040327302481021394,
      "loss": 5.1703,
      "mean_token_accuracy": 0.21789804697036744,
      "num_tokens": 72562281.0,
      "step": 35810
    },
    {
      "entropy": 5.980292320251465,
      "epoch": 3.1298610504238398,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00040324742244180937,
      "loss": 5.1593,
      "mean_token_accuracy": 0.21543047130107879,
      "num_tokens": 72571817.0,
      "step": 35815
    },
    {
      "entropy": 5.954221248626709,
      "epoch": 3.1302979987765447,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040322181761355194,
      "loss": 5.2032,
      "mean_token_accuracy": 0.21761967986822128,
      "num_tokens": 72582935.0,
      "step": 35820
    },
    {
      "entropy": 6.022347736358642,
      "epoch": 3.130734947129249,
      "grad_norm": 1.25,
      "learning_rate": 0.0004031962103259329,
      "loss": 5.2377,
      "mean_token_accuracy": 0.21347364485263826,
      "num_tokens": 72593280.0,
      "step": 35825
    },
    {
      "entropy": 5.994548749923706,
      "epoch": 3.131171895481954,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004031706005794434,
      "loss": 5.1851,
      "mean_token_accuracy": 0.20855167806148528,
      "num_tokens": 72603193.0,
      "step": 35830
    },
    {
      "entropy": 5.965228605270386,
      "epoch": 3.1316088438346585,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004031449883745746,
      "loss": 5.252,
      "mean_token_accuracy": 0.21372345536947251,
      "num_tokens": 72613363.0,
      "step": 35835
    },
    {
      "entropy": 6.00995659828186,
      "epoch": 3.1320457921873635,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040311937371181774,
      "loss": 5.4002,
      "mean_token_accuracy": 0.19713767021894454,
      "num_tokens": 72624151.0,
      "step": 35840
    },
    {
      "entropy": 6.02982497215271,
      "epoch": 3.1324827405400684,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040309375659166414,
      "loss": 5.1645,
      "mean_token_accuracy": 0.21702875941991806,
      "num_tokens": 72633692.0,
      "step": 35845
    },
    {
      "entropy": 6.0130914688110355,
      "epoch": 3.132919688892773,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00040306813701460525,
      "loss": 5.2595,
      "mean_token_accuracy": 0.2065451830625534,
      "num_tokens": 72644016.0,
      "step": 35850
    },
    {
      "entropy": 5.9662927150726315,
      "epoch": 3.1333566372454777,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004030425149811324,
      "loss": 5.2564,
      "mean_token_accuracy": 0.21173144727945328,
      "num_tokens": 72654033.0,
      "step": 35855
    },
    {
      "entropy": 6.049905681610108,
      "epoch": 3.133793585598182,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004030168904917369,
      "loss": 5.2347,
      "mean_token_accuracy": 0.19869724810123443,
      "num_tokens": 72663452.0,
      "step": 35860
    },
    {
      "entropy": 6.048838138580322,
      "epoch": 3.134230533950887,
      "grad_norm": 1.25,
      "learning_rate": 0.00040299126354691043,
      "loss": 5.2712,
      "mean_token_accuracy": 0.20620719492435455,
      "num_tokens": 72674766.0,
      "step": 35865
    },
    {
      "entropy": 6.063779211044311,
      "epoch": 3.1346674823035916,
      "grad_norm": 1.3125,
      "learning_rate": 0.00040296563414714444,
      "loss": 5.2483,
      "mean_token_accuracy": 0.2077609494328499,
      "num_tokens": 72683816.0,
      "step": 35870
    },
    {
      "entropy": 5.998328113555909,
      "epoch": 3.1351044306562965,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004029400022929306,
      "loss": 5.1798,
      "mean_token_accuracy": 0.21585624516010285,
      "num_tokens": 72693604.0,
      "step": 35875
    },
    {
      "entropy": 5.996345663070679,
      "epoch": 3.135541379009001,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004029143679847604,
      "loss": 5.2426,
      "mean_token_accuracy": 0.20656104236841202,
      "num_tokens": 72703936.0,
      "step": 35880
    },
    {
      "entropy": 6.02845869064331,
      "epoch": 3.135978327361706,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004028887312231256,
      "loss": 5.21,
      "mean_token_accuracy": 0.21478404998779296,
      "num_tokens": 72713646.0,
      "step": 35885
    },
    {
      "entropy": 6.04366021156311,
      "epoch": 3.1364152757144104,
      "grad_norm": 1.28125,
      "learning_rate": 0.00040286309200851793,
      "loss": 5.1766,
      "mean_token_accuracy": 0.2195429638028145,
      "num_tokens": 72723573.0,
      "step": 35890
    },
    {
      "entropy": 6.042674446105957,
      "epoch": 3.1368522240671153,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004028374503414292,
      "loss": 5.2936,
      "mean_token_accuracy": 0.1979572892189026,
      "num_tokens": 72733805.0,
      "step": 35895
    },
    {
      "entropy": 5.9394278049469,
      "epoch": 3.1372891724198197,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004028118062223512,
      "loss": 5.1815,
      "mean_token_accuracy": 0.21578700840473175,
      "num_tokens": 72744254.0,
      "step": 35900
    },
    {
      "entropy": 6.034436416625977,
      "epoch": 3.1377261207725247,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004027861596517757,
      "loss": 5.2867,
      "mean_token_accuracy": 0.19994129687547685,
      "num_tokens": 72754267.0,
      "step": 35905
    },
    {
      "entropy": 5.907765340805054,
      "epoch": 3.1381630691252296,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004027605106301948,
      "loss": 5.1441,
      "mean_token_accuracy": 0.2264554962515831,
      "num_tokens": 72764598.0,
      "step": 35910
    },
    {
      "entropy": 6.042913484573364,
      "epoch": 3.138600017477934,
      "grad_norm": 1.28125,
      "learning_rate": 0.00040273485915810026,
      "loss": 5.2832,
      "mean_token_accuracy": 0.20580648481845856,
      "num_tokens": 72774145.0,
      "step": 35915
    },
    {
      "entropy": 6.130795335769653,
      "epoch": 3.139036965830639,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00040270920523598425,
      "loss": 5.3451,
      "mean_token_accuracy": 0.19467676132917405,
      "num_tokens": 72783766.0,
      "step": 35920
    },
    {
      "entropy": 6.039707946777344,
      "epoch": 3.1394739141833434,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004026835488643387,
      "loss": 5.227,
      "mean_token_accuracy": 0.21842605471611024,
      "num_tokens": 72793907.0,
      "step": 35925
    },
    {
      "entropy": 6.06435809135437,
      "epoch": 3.1399108625360483,
      "grad_norm": 1.40625,
      "learning_rate": 0.00040265789004365586,
      "loss": 5.3478,
      "mean_token_accuracy": 0.19916316866874695,
      "num_tokens": 72802927.0,
      "step": 35930
    },
    {
      "entropy": 5.942129182815552,
      "epoch": 3.140347810888753,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004026322287744277,
      "loss": 5.0894,
      "mean_token_accuracy": 0.22457202672958373,
      "num_tokens": 72813387.0,
      "step": 35935
    },
    {
      "entropy": 6.013668441772461,
      "epoch": 3.1407847592414577,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004026065650571466,
      "loss": 5.1703,
      "mean_token_accuracy": 0.21614699810743332,
      "num_tokens": 72824109.0,
      "step": 35940
    },
    {
      "entropy": 5.999838590621948,
      "epoch": 3.141221707594162,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004025808988923047,
      "loss": 5.1755,
      "mean_token_accuracy": 0.20839475244283676,
      "num_tokens": 72834107.0,
      "step": 35945
    },
    {
      "entropy": 6.029338836669922,
      "epoch": 3.141658655946867,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040255523028039433,
      "loss": 5.2056,
      "mean_token_accuracy": 0.21301597058773042,
      "num_tokens": 72843726.0,
      "step": 35950
    },
    {
      "entropy": 6.029319143295288,
      "epoch": 3.1420956042995716,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00040252955922190776,
      "loss": 5.298,
      "mean_token_accuracy": 0.20504157096147538,
      "num_tokens": 72854203.0,
      "step": 35955
    },
    {
      "entropy": 6.009540987014771,
      "epoch": 3.1425325526522765,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040250388571733744,
      "loss": 5.2213,
      "mean_token_accuracy": 0.20537275224924087,
      "num_tokens": 72864689.0,
      "step": 35960
    },
    {
      "entropy": 5.969363832473755,
      "epoch": 3.1429695010049814,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004024782097671758,
      "loss": 5.2646,
      "mean_token_accuracy": 0.2106610879302025,
      "num_tokens": 72875040.0,
      "step": 35965
    },
    {
      "entropy": 5.97716212272644,
      "epoch": 3.143406449357686,
      "grad_norm": 1.25,
      "learning_rate": 0.0004024525313719153,
      "loss": 5.232,
      "mean_token_accuracy": 0.2089971721172333,
      "num_tokens": 72885174.0,
      "step": 35970
    },
    {
      "entropy": 5.9887073040008545,
      "epoch": 3.143843397710391,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040242685053204844,
      "loss": 5.2179,
      "mean_token_accuracy": 0.21116133779287338,
      "num_tokens": 72895344.0,
      "step": 35975
    },
    {
      "entropy": 6.074653005599975,
      "epoch": 3.1442803460630953,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004024011672480678,
      "loss": 5.2766,
      "mean_token_accuracy": 0.20658514648675919,
      "num_tokens": 72905950.0,
      "step": 35980
    },
    {
      "entropy": 6.035929489135742,
      "epoch": 3.1447172944158,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040237548152046603,
      "loss": 5.2361,
      "mean_token_accuracy": 0.2102895423769951,
      "num_tokens": 72916828.0,
      "step": 35985
    },
    {
      "entropy": 6.011080551147461,
      "epoch": 3.1451542427685046,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040234979334973577,
      "loss": 5.2875,
      "mean_token_accuracy": 0.2113719254732132,
      "num_tokens": 72926951.0,
      "step": 35990
    },
    {
      "entropy": 5.9285430908203125,
      "epoch": 3.1455911911212096,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004023241027363698,
      "loss": 5.1573,
      "mean_token_accuracy": 0.2136702448129654,
      "num_tokens": 72937909.0,
      "step": 35995
    },
    {
      "entropy": 6.012809562683105,
      "epoch": 3.146028139473914,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004022984096808608,
      "loss": 5.2231,
      "mean_token_accuracy": 0.20738113075494766,
      "num_tokens": 72949898.0,
      "step": 36000
    },
    {
      "epoch": 3.146028139473914,
      "eval_entropy": 5.723772223577505,
      "eval_loss": 5.297130107879639,
      "eval_mean_token_accuracy": 0.21454855861025132,
      "eval_num_tokens": 72949898.0,
      "eval_runtime": 21.6478,
      "eval_samples_per_second": 1709.83,
      "eval_steps_per_second": 213.74,
      "step": 36000
    }
  ],
  "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": 1.619442135272448e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}