{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.170028818443804,
  "eval_steps": 3000,
  "global_step": 33000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 4.952998208999634,
      "epoch": 0.0004803073967339097,
      "grad_norm": 14.5625,
      "learning_rate": 2e-06,
      "loss": 14.2582,
      "mean_token_accuracy": 0.0,
      "num_tokens": 10821.0,
      "step": 5
    },
    {
      "entropy": 4.955690383911133,
      "epoch": 0.0009606147934678194,
      "grad_norm": 14.0,
      "learning_rate": 4.5e-06,
      "loss": 14.2611,
      "mean_token_accuracy": 0.0,
      "num_tokens": 21838.0,
      "step": 10
    },
    {
      "entropy": 4.961869859695435,
      "epoch": 0.001440922190201729,
      "grad_norm": 15.125,
      "learning_rate": 7e-06,
      "loss": 14.2406,
      "mean_token_accuracy": 0.0001630714046768844,
      "num_tokens": 33483.0,
      "step": 15
    },
    {
      "entropy": 5.000402975082397,
      "epoch": 0.0019212295869356388,
      "grad_norm": 17.625,
      "learning_rate": 9.5e-06,
      "loss": 14.0342,
      "mean_token_accuracy": 0.0,
      "num_tokens": 45150.0,
      "step": 20
    },
    {
      "entropy": 5.144548416137695,
      "epoch": 0.0024015369836695487,
      "grad_norm": 21.625,
      "learning_rate": 1.2e-05,
      "loss": 13.6497,
      "mean_token_accuracy": 0.0,
      "num_tokens": 55862.0,
      "step": 25
    },
    {
      "entropy": 5.467367696762085,
      "epoch": 0.002881844380403458,
      "grad_norm": 29.25,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 13.0786,
      "mean_token_accuracy": 0.0,
      "num_tokens": 66766.0,
      "step": 30
    },
    {
      "entropy": 6.960640001296997,
      "epoch": 0.0033621517771373678,
      "grad_norm": 24.5,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 11.9791,
      "mean_token_accuracy": 0.0,
      "num_tokens": 77012.0,
      "step": 35
    },
    {
      "entropy": 10.176104927062989,
      "epoch": 0.0038424591738712775,
      "grad_norm": 3.328125,
      "learning_rate": 1.95e-05,
      "loss": 10.8581,
      "mean_token_accuracy": 0.0017824553069658577,
      "num_tokens": 89034.0,
      "step": 40
    },
    {
      "entropy": 10.718319034576416,
      "epoch": 0.004322766570605188,
      "grad_norm": 2.9375,
      "learning_rate": 2.2e-05,
      "loss": 10.6429,
      "mean_token_accuracy": 0.01355557688511908,
      "num_tokens": 100341.0,
      "step": 45
    },
    {
      "entropy": 10.739382457733154,
      "epoch": 0.004803073967339097,
      "grad_norm": 2.453125,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 10.5214,
      "mean_token_accuracy": 0.014157400839030743,
      "num_tokens": 111997.0,
      "step": 50
    },
    {
      "entropy": 10.734613037109375,
      "epoch": 0.005283381364073006,
      "grad_norm": 2.0625,
      "learning_rate": 2.7e-05,
      "loss": 10.3541,
      "mean_token_accuracy": 0.014309123158454895,
      "num_tokens": 123705.0,
      "step": 55
    },
    {
      "entropy": 10.708789348602295,
      "epoch": 0.005763688760806916,
      "grad_norm": 2.0,
      "learning_rate": 2.95e-05,
      "loss": 10.1416,
      "mean_token_accuracy": 0.01604629335924983,
      "num_tokens": 134091.0,
      "step": 60
    },
    {
      "entropy": 10.720145320892334,
      "epoch": 0.006243996157540826,
      "grad_norm": 1.78125,
      "learning_rate": 3.2e-05,
      "loss": 10.0631,
      "mean_token_accuracy": 0.01453780010342598,
      "num_tokens": 145731.0,
      "step": 65
    },
    {
      "entropy": 10.739076805114745,
      "epoch": 0.0067243035542747355,
      "grad_norm": 1.8984375,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 9.9517,
      "mean_token_accuracy": 0.017807454522699117,
      "num_tokens": 156979.0,
      "step": 70
    },
    {
      "entropy": 10.738404560089112,
      "epoch": 0.007204610951008645,
      "grad_norm": 1.8203125,
      "learning_rate": 3.7e-05,
      "loss": 9.8629,
      "mean_token_accuracy": 0.024540318921208382,
      "num_tokens": 167021.0,
      "step": 75
    },
    {
      "entropy": 10.731132507324219,
      "epoch": 0.007684918347742555,
      "grad_norm": 1.8359375,
      "learning_rate": 3.95e-05,
      "loss": 9.819,
      "mean_token_accuracy": 0.020528202410787344,
      "num_tokens": 179269.0,
      "step": 80
    },
    {
      "entropy": 10.73000726699829,
      "epoch": 0.008165225744476465,
      "grad_norm": 1.703125,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 9.7216,
      "mean_token_accuracy": 0.020909248571842908,
      "num_tokens": 190328.0,
      "step": 85
    },
    {
      "entropy": 10.728069496154784,
      "epoch": 0.008645533141210375,
      "grad_norm": 1.859375,
      "learning_rate": 4.45e-05,
      "loss": 9.5738,
      "mean_token_accuracy": 0.021778783388435842,
      "num_tokens": 201522.0,
      "step": 90
    },
    {
      "entropy": 10.719436931610108,
      "epoch": 0.009125840537944284,
      "grad_norm": 1.7578125,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 9.4983,
      "mean_token_accuracy": 0.022924750670790674,
      "num_tokens": 212963.0,
      "step": 95
    },
    {
      "entropy": 10.704556751251221,
      "epoch": 0.009606147934678195,
      "grad_norm": 1.5625,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 9.4627,
      "mean_token_accuracy": 0.026258311606943606,
      "num_tokens": 224446.0,
      "step": 100
    },
    {
      "entropy": 10.686229038238526,
      "epoch": 0.010086455331412104,
      "grad_norm": 1.75,
      "learning_rate": 5.2e-05,
      "loss": 9.2916,
      "mean_token_accuracy": 0.03442668188363314,
      "num_tokens": 235334.0,
      "step": 105
    },
    {
      "entropy": 10.660537052154542,
      "epoch": 0.010566762728146013,
      "grad_norm": 1.6484375,
      "learning_rate": 5.45e-05,
      "loss": 9.1974,
      "mean_token_accuracy": 0.038018427044153216,
      "num_tokens": 247024.0,
      "step": 110
    },
    {
      "entropy": 10.629736328125,
      "epoch": 0.011047070124879923,
      "grad_norm": 1.7265625,
      "learning_rate": 5.7e-05,
      "loss": 9.044,
      "mean_token_accuracy": 0.03711554128676653,
      "num_tokens": 258354.0,
      "step": 115
    },
    {
      "entropy": 10.596047592163085,
      "epoch": 0.011527377521613832,
      "grad_norm": 1.6171875,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 8.954,
      "mean_token_accuracy": 0.03982700537890196,
      "num_tokens": 271202.0,
      "step": 120
    },
    {
      "entropy": 10.551871967315673,
      "epoch": 0.012007684918347743,
      "grad_norm": 1.578125,
      "learning_rate": 6.2e-05,
      "loss": 8.8293,
      "mean_token_accuracy": 0.03978835418820381,
      "num_tokens": 282405.0,
      "step": 125
    },
    {
      "entropy": 10.49206190109253,
      "epoch": 0.012487992315081652,
      "grad_norm": 1.609375,
      "learning_rate": 6.450000000000001e-05,
      "loss": 8.6826,
      "mean_token_accuracy": 0.039380787685513494,
      "num_tokens": 294211.0,
      "step": 130
    },
    {
      "entropy": 10.426816463470459,
      "epoch": 0.012968299711815562,
      "grad_norm": 1.5546875,
      "learning_rate": 6.7e-05,
      "loss": 8.6088,
      "mean_token_accuracy": 0.03625727668404579,
      "num_tokens": 305420.0,
      "step": 135
    },
    {
      "entropy": 10.343353271484375,
      "epoch": 0.013448607108549471,
      "grad_norm": 1.46875,
      "learning_rate": 6.950000000000001e-05,
      "loss": 8.4538,
      "mean_token_accuracy": 0.04094459749758243,
      "num_tokens": 316430.0,
      "step": 140
    },
    {
      "entropy": 10.245965957641602,
      "epoch": 0.013928914505283382,
      "grad_norm": 1.4296875,
      "learning_rate": 7.2e-05,
      "loss": 8.3471,
      "mean_token_accuracy": 0.03853845577687025,
      "num_tokens": 328406.0,
      "step": 145
    },
    {
      "entropy": 10.129299736022949,
      "epoch": 0.01440922190201729,
      "grad_norm": 1.3984375,
      "learning_rate": 7.45e-05,
      "loss": 8.2251,
      "mean_token_accuracy": 0.0405678428709507,
      "num_tokens": 339051.0,
      "step": 150
    },
    {
      "entropy": 9.980317115783691,
      "epoch": 0.014889529298751201,
      "grad_norm": 1.3125,
      "learning_rate": 7.7e-05,
      "loss": 8.1161,
      "mean_token_accuracy": 0.04249956700950861,
      "num_tokens": 350308.0,
      "step": 155
    },
    {
      "entropy": 9.8000244140625,
      "epoch": 0.01536983669548511,
      "grad_norm": 1.2109375,
      "learning_rate": 7.950000000000001e-05,
      "loss": 8.0193,
      "mean_token_accuracy": 0.04042474515736103,
      "num_tokens": 362159.0,
      "step": 160
    },
    {
      "entropy": 9.598854160308838,
      "epoch": 0.01585014409221902,
      "grad_norm": 1.1015625,
      "learning_rate": 8.2e-05,
      "loss": 7.951,
      "mean_token_accuracy": 0.0442409684881568,
      "num_tokens": 375157.0,
      "step": 165
    },
    {
      "entropy": 9.361251831054688,
      "epoch": 0.01633045148895293,
      "grad_norm": 1.0,
      "learning_rate": 8.450000000000001e-05,
      "loss": 7.7913,
      "mean_token_accuracy": 0.05417908914387226,
      "num_tokens": 387318.0,
      "step": 170
    },
    {
      "entropy": 9.094889831542968,
      "epoch": 0.01681075888568684,
      "grad_norm": 0.84765625,
      "learning_rate": 8.7e-05,
      "loss": 7.7466,
      "mean_token_accuracy": 0.06120046563446522,
      "num_tokens": 398933.0,
      "step": 175
    },
    {
      "entropy": 8.835952472686767,
      "epoch": 0.01729106628242075,
      "grad_norm": 0.9453125,
      "learning_rate": 8.95e-05,
      "loss": 7.6633,
      "mean_token_accuracy": 0.05573962703347206,
      "num_tokens": 410460.0,
      "step": 180
    },
    {
      "entropy": 8.607842922210693,
      "epoch": 0.01777137367915466,
      "grad_norm": 0.890625,
      "learning_rate": 9.2e-05,
      "loss": 7.6204,
      "mean_token_accuracy": 0.056572451069951055,
      "num_tokens": 421894.0,
      "step": 185
    },
    {
      "entropy": 8.41255807876587,
      "epoch": 0.01825168107588857,
      "grad_norm": 0.69140625,
      "learning_rate": 9.45e-05,
      "loss": 7.6005,
      "mean_token_accuracy": 0.05312460921704769,
      "num_tokens": 432976.0,
      "step": 190
    },
    {
      "entropy": 8.314041900634766,
      "epoch": 0.018731988472622477,
      "grad_norm": 0.86328125,
      "learning_rate": 9.7e-05,
      "loss": 7.6613,
      "mean_token_accuracy": 0.05617779567837715,
      "num_tokens": 444859.0,
      "step": 195
    },
    {
      "entropy": 8.262031841278077,
      "epoch": 0.01921229586935639,
      "grad_norm": 0.859375,
      "learning_rate": 9.95e-05,
      "loss": 7.5566,
      "mean_token_accuracy": 0.06568880900740623,
      "num_tokens": 456508.0,
      "step": 200
    },
    {
      "entropy": 8.133868789672851,
      "epoch": 0.0196926032660903,
      "grad_norm": 0.7578125,
      "learning_rate": 0.000102,
      "loss": 7.5667,
      "mean_token_accuracy": 0.06184492409229279,
      "num_tokens": 469064.0,
      "step": 205
    },
    {
      "entropy": 8.096502780914307,
      "epoch": 0.020172910662824207,
      "grad_norm": 1.125,
      "learning_rate": 0.00010449999999999999,
      "loss": 7.4995,
      "mean_token_accuracy": 0.06528568714857101,
      "num_tokens": 480743.0,
      "step": 210
    },
    {
      "entropy": 8.07320432662964,
      "epoch": 0.020653218059558116,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000107,
      "loss": 7.5222,
      "mean_token_accuracy": 0.07013107351958751,
      "num_tokens": 493378.0,
      "step": 215
    },
    {
      "entropy": 8.062201929092407,
      "epoch": 0.021133525456292025,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0001095,
      "loss": 7.4632,
      "mean_token_accuracy": 0.07239910811185837,
      "num_tokens": 505641.0,
      "step": 220
    },
    {
      "entropy": 8.092798614501953,
      "epoch": 0.021613832853025938,
      "grad_norm": 0.828125,
      "learning_rate": 0.000112,
      "loss": 7.5153,
      "mean_token_accuracy": 0.0712558664381504,
      "num_tokens": 517783.0,
      "step": 225
    },
    {
      "entropy": 8.056060886383056,
      "epoch": 0.022094140249759846,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0001145,
      "loss": 7.3885,
      "mean_token_accuracy": 0.0776093952357769,
      "num_tokens": 528407.0,
      "step": 230
    },
    {
      "entropy": 8.07971124649048,
      "epoch": 0.022574447646493755,
      "grad_norm": 1.421875,
      "learning_rate": 0.00011700000000000001,
      "loss": 7.499,
      "mean_token_accuracy": 0.07200518772006034,
      "num_tokens": 539974.0,
      "step": 235
    },
    {
      "entropy": 8.014474582672118,
      "epoch": 0.023054755043227664,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00011949999999999999,
      "loss": 7.3531,
      "mean_token_accuracy": 0.07475511692464351,
      "num_tokens": 551127.0,
      "step": 240
    },
    {
      "entropy": 8.012955379486083,
      "epoch": 0.023535062439961577,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000122,
      "loss": 7.3664,
      "mean_token_accuracy": 0.08100151270627975,
      "num_tokens": 561980.0,
      "step": 245
    },
    {
      "entropy": 8.09322385787964,
      "epoch": 0.024015369836695485,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0001245,
      "loss": 7.3988,
      "mean_token_accuracy": 0.07981968075037002,
      "num_tokens": 573840.0,
      "step": 250
    },
    {
      "entropy": 7.97099084854126,
      "epoch": 0.024495677233429394,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000127,
      "loss": 7.2948,
      "mean_token_accuracy": 0.0818309374153614,
      "num_tokens": 585110.0,
      "step": 255
    },
    {
      "entropy": 7.940869426727295,
      "epoch": 0.024975984630163303,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0001295,
      "loss": 7.2052,
      "mean_token_accuracy": 0.08267272785305976,
      "num_tokens": 596902.0,
      "step": 260
    },
    {
      "entropy": 7.915973138809204,
      "epoch": 0.025456292026897216,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000132,
      "loss": 7.3346,
      "mean_token_accuracy": 0.08441563807427883,
      "num_tokens": 608738.0,
      "step": 265
    },
    {
      "entropy": 7.930963659286499,
      "epoch": 0.025936599423631124,
      "grad_norm": 0.9375,
      "learning_rate": 0.00013450000000000002,
      "loss": 7.2662,
      "mean_token_accuracy": 0.07967347651720047,
      "num_tokens": 621078.0,
      "step": 270
    },
    {
      "entropy": 7.923765468597412,
      "epoch": 0.026416906820365033,
      "grad_norm": 1.375,
      "learning_rate": 0.00013700000000000002,
      "loss": 7.3281,
      "mean_token_accuracy": 0.08380152806639671,
      "num_tokens": 632333.0,
      "step": 275
    },
    {
      "entropy": 7.8553708553314205,
      "epoch": 0.026897214217098942,
      "grad_norm": 1.125,
      "learning_rate": 0.0001395,
      "loss": 7.1392,
      "mean_token_accuracy": 0.08382196128368377,
      "num_tokens": 644474.0,
      "step": 280
    },
    {
      "entropy": 7.874654054641724,
      "epoch": 0.027377521613832854,
      "grad_norm": 1.265625,
      "learning_rate": 0.00014199999999999998,
      "loss": 7.2046,
      "mean_token_accuracy": 0.08822997659444809,
      "num_tokens": 655608.0,
      "step": 285
    },
    {
      "entropy": 7.8405436992645265,
      "epoch": 0.027857829010566763,
      "grad_norm": 1.59375,
      "learning_rate": 0.0001445,
      "loss": 7.2322,
      "mean_token_accuracy": 0.08056687787175179,
      "num_tokens": 667371.0,
      "step": 290
    },
    {
      "entropy": 7.788603258132935,
      "epoch": 0.028338136407300672,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000147,
      "loss": 7.2274,
      "mean_token_accuracy": 0.08374285846948623,
      "num_tokens": 679152.0,
      "step": 295
    },
    {
      "entropy": 7.757482385635376,
      "epoch": 0.02881844380403458,
      "grad_norm": 1.296875,
      "learning_rate": 0.0001495,
      "loss": 7.1415,
      "mean_token_accuracy": 0.09327690675854683,
      "num_tokens": 690359.0,
      "step": 300
    },
    {
      "entropy": 7.725837469100952,
      "epoch": 0.029298751200768493,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000152,
      "loss": 7.1113,
      "mean_token_accuracy": 0.0908672846853733,
      "num_tokens": 701819.0,
      "step": 305
    },
    {
      "entropy": 7.791640710830689,
      "epoch": 0.029779058597502402,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00015450000000000001,
      "loss": 7.1544,
      "mean_token_accuracy": 0.08916135728359223,
      "num_tokens": 711048.0,
      "step": 310
    },
    {
      "entropy": 7.7330770015716555,
      "epoch": 0.03025936599423631,
      "grad_norm": 1.21875,
      "learning_rate": 0.000157,
      "loss": 7.0486,
      "mean_token_accuracy": 0.0980133630335331,
      "num_tokens": 721399.0,
      "step": 315
    },
    {
      "entropy": 7.682194662094116,
      "epoch": 0.03073967339097022,
      "grad_norm": 2.078125,
      "learning_rate": 0.0001595,
      "loss": 7.0748,
      "mean_token_accuracy": 0.09791787564754487,
      "num_tokens": 732144.0,
      "step": 320
    },
    {
      "entropy": 7.704891967773437,
      "epoch": 0.03121998078770413,
      "grad_norm": 1.5546875,
      "learning_rate": 0.000162,
      "loss": 7.0133,
      "mean_token_accuracy": 0.09394430741667747,
      "num_tokens": 743962.0,
      "step": 325
    },
    {
      "entropy": 7.67185001373291,
      "epoch": 0.03170028818443804,
      "grad_norm": 1.375,
      "learning_rate": 0.00016450000000000001,
      "loss": 7.1606,
      "mean_token_accuracy": 0.09519302621483802,
      "num_tokens": 755641.0,
      "step": 330
    },
    {
      "entropy": 7.65534348487854,
      "epoch": 0.03218059558117195,
      "grad_norm": 1.8203125,
      "learning_rate": 0.00016700000000000002,
      "loss": 6.9888,
      "mean_token_accuracy": 0.1040790244936943,
      "num_tokens": 766798.0,
      "step": 335
    },
    {
      "entropy": 7.634436845779419,
      "epoch": 0.03266090297790586,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00016950000000000003,
      "loss": 7.0797,
      "mean_token_accuracy": 0.10063858479261398,
      "num_tokens": 778097.0,
      "step": 340
    },
    {
      "entropy": 7.547883176803589,
      "epoch": 0.03314121037463977,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00017199999999999998,
      "loss": 6.9963,
      "mean_token_accuracy": 0.10573035553097725,
      "num_tokens": 789775.0,
      "step": 345
    },
    {
      "entropy": 7.640296983718872,
      "epoch": 0.03362151777137368,
      "grad_norm": 1.21875,
      "learning_rate": 0.00017449999999999999,
      "loss": 7.0568,
      "mean_token_accuracy": 0.09917445480823517,
      "num_tokens": 800155.0,
      "step": 350
    },
    {
      "entropy": 7.624283695220948,
      "epoch": 0.034101825168107586,
      "grad_norm": 1.5234375,
      "learning_rate": 0.000177,
      "loss": 7.0508,
      "mean_token_accuracy": 0.10235279574990272,
      "num_tokens": 811213.0,
      "step": 355
    },
    {
      "entropy": 7.644110107421875,
      "epoch": 0.0345821325648415,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0001795,
      "loss": 7.0427,
      "mean_token_accuracy": 0.09794286787509918,
      "num_tokens": 822832.0,
      "step": 360
    },
    {
      "entropy": 7.549008846282959,
      "epoch": 0.03506243996157541,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000182,
      "loss": 7.0118,
      "mean_token_accuracy": 0.10539761856198311,
      "num_tokens": 834824.0,
      "step": 365
    },
    {
      "entropy": 7.598927068710327,
      "epoch": 0.03554274735830932,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0001845,
      "loss": 7.046,
      "mean_token_accuracy": 0.10215208157896996,
      "num_tokens": 846969.0,
      "step": 370
    },
    {
      "entropy": 7.629372644424438,
      "epoch": 0.03602305475504323,
      "grad_norm": 1.9921875,
      "learning_rate": 0.000187,
      "loss": 7.1144,
      "mean_token_accuracy": 0.09570988789200782,
      "num_tokens": 859271.0,
      "step": 375
    },
    {
      "entropy": 7.489568710327148,
      "epoch": 0.03650336215177714,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0001895,
      "loss": 7.0168,
      "mean_token_accuracy": 0.1066791370511055,
      "num_tokens": 869754.0,
      "step": 380
    },
    {
      "entropy": 7.518455076217651,
      "epoch": 0.036983669548511046,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000192,
      "loss": 6.9428,
      "mean_token_accuracy": 0.10733397230505944,
      "num_tokens": 880519.0,
      "step": 385
    },
    {
      "entropy": 7.433148956298828,
      "epoch": 0.037463976945244955,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0001945,
      "loss": 6.9057,
      "mean_token_accuracy": 0.10587815120816231,
      "num_tokens": 890765.0,
      "step": 390
    },
    {
      "entropy": 7.51275486946106,
      "epoch": 0.037944284341978864,
      "grad_norm": 1.359375,
      "learning_rate": 0.00019700000000000002,
      "loss": 6.9144,
      "mean_token_accuracy": 0.10545952916145325,
      "num_tokens": 901717.0,
      "step": 395
    },
    {
      "entropy": 7.56442084312439,
      "epoch": 0.03842459173871278,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00019950000000000002,
      "loss": 6.9775,
      "mean_token_accuracy": 0.10485790073871612,
      "num_tokens": 913334.0,
      "step": 400
    },
    {
      "entropy": 7.5254968166351315,
      "epoch": 0.03890489913544669,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000202,
      "loss": 6.9865,
      "mean_token_accuracy": 0.10525585263967514,
      "num_tokens": 924753.0,
      "step": 405
    },
    {
      "entropy": 7.441157817840576,
      "epoch": 0.0393852065321806,
      "grad_norm": 1.296875,
      "learning_rate": 0.00020449999999999998,
      "loss": 6.921,
      "mean_token_accuracy": 0.10738600119948387,
      "num_tokens": 935692.0,
      "step": 410
    },
    {
      "entropy": 7.468551778793335,
      "epoch": 0.039865513928914506,
      "grad_norm": 0.96875,
      "learning_rate": 0.000207,
      "loss": 7.0066,
      "mean_token_accuracy": 0.10407103821635247,
      "num_tokens": 947902.0,
      "step": 415
    },
    {
      "entropy": 7.422157144546508,
      "epoch": 0.040345821325648415,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002095,
      "loss": 6.9215,
      "mean_token_accuracy": 0.10983402356505394,
      "num_tokens": 959437.0,
      "step": 420
    },
    {
      "entropy": 7.329869747161865,
      "epoch": 0.040826128722382324,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000212,
      "loss": 6.7483,
      "mean_token_accuracy": 0.11806870326399803,
      "num_tokens": 970191.0,
      "step": 425
    },
    {
      "entropy": 7.379027938842773,
      "epoch": 0.04130643611911623,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002145,
      "loss": 6.8638,
      "mean_token_accuracy": 0.10865884721279144,
      "num_tokens": 981811.0,
      "step": 430
    },
    {
      "entropy": 7.4306111335754395,
      "epoch": 0.04178674351585014,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00021700000000000002,
      "loss": 6.9707,
      "mean_token_accuracy": 0.10563067272305489,
      "num_tokens": 994034.0,
      "step": 435
    },
    {
      "entropy": 7.317675256729126,
      "epoch": 0.04226705091258405,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0002195,
      "loss": 6.9255,
      "mean_token_accuracy": 0.10667062103748322,
      "num_tokens": 1005506.0,
      "step": 440
    },
    {
      "entropy": 7.382990884780884,
      "epoch": 0.042747358309317966,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000222,
      "loss": 6.8792,
      "mean_token_accuracy": 0.10067129284143447,
      "num_tokens": 1017103.0,
      "step": 445
    },
    {
      "entropy": 7.375922918319702,
      "epoch": 0.043227665706051875,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002245,
      "loss": 6.9606,
      "mean_token_accuracy": 0.10542604103684425,
      "num_tokens": 1029606.0,
      "step": 450
    },
    {
      "entropy": 7.423660087585449,
      "epoch": 0.043707973102785784,
      "grad_norm": 1.40625,
      "learning_rate": 0.00022700000000000002,
      "loss": 6.9113,
      "mean_token_accuracy": 0.10665275976061821,
      "num_tokens": 1041899.0,
      "step": 455
    },
    {
      "entropy": 7.303192949295044,
      "epoch": 0.04418828049951969,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00022950000000000002,
      "loss": 6.8294,
      "mean_token_accuracy": 0.11543489247560501,
      "num_tokens": 1053577.0,
      "step": 460
    },
    {
      "entropy": 7.357319164276123,
      "epoch": 0.0446685878962536,
      "grad_norm": 1.0625,
      "learning_rate": 0.00023200000000000003,
      "loss": 6.8798,
      "mean_token_accuracy": 0.11370891258120537,
      "num_tokens": 1065541.0,
      "step": 465
    },
    {
      "entropy": 7.381957864761352,
      "epoch": 0.04514889529298751,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00023449999999999998,
      "loss": 6.8192,
      "mean_token_accuracy": 0.1103691153228283,
      "num_tokens": 1076138.0,
      "step": 470
    },
    {
      "entropy": 7.3531769752502445,
      "epoch": 0.04562920268972142,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000237,
      "loss": 6.8676,
      "mean_token_accuracy": 0.11432998552918434,
      "num_tokens": 1087620.0,
      "step": 475
    },
    {
      "entropy": 7.271972751617431,
      "epoch": 0.04610951008645533,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002395,
      "loss": 6.765,
      "mean_token_accuracy": 0.11456627622246743,
      "num_tokens": 1099493.0,
      "step": 480
    },
    {
      "entropy": 7.294817638397217,
      "epoch": 0.046589817483189244,
      "grad_norm": 1.140625,
      "learning_rate": 0.000242,
      "loss": 6.8052,
      "mean_token_accuracy": 0.11549188196659088,
      "num_tokens": 1110559.0,
      "step": 485
    },
    {
      "entropy": 7.216236162185669,
      "epoch": 0.04707012487992315,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002445,
      "loss": 6.7756,
      "mean_token_accuracy": 0.10924254432320595,
      "num_tokens": 1121051.0,
      "step": 490
    },
    {
      "entropy": 7.257420635223388,
      "epoch": 0.04755043227665706,
      "grad_norm": 1.15625,
      "learning_rate": 0.000247,
      "loss": 6.7499,
      "mean_token_accuracy": 0.11539025753736495,
      "num_tokens": 1132305.0,
      "step": 495
    },
    {
      "entropy": 7.321677541732788,
      "epoch": 0.04803073967339097,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002495,
      "loss": 6.7571,
      "mean_token_accuracy": 0.11674077361822129,
      "num_tokens": 1142979.0,
      "step": 500
    },
    {
      "entropy": 7.188573884963989,
      "epoch": 0.04851104707012488,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000252,
      "loss": 6.7549,
      "mean_token_accuracy": 0.11924750506877899,
      "num_tokens": 1153889.0,
      "step": 505
    },
    {
      "entropy": 7.244559097290039,
      "epoch": 0.04899135446685879,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002545,
      "loss": 6.7565,
      "mean_token_accuracy": 0.11460114493966103,
      "num_tokens": 1166030.0,
      "step": 510
    },
    {
      "entropy": 7.383937501907349,
      "epoch": 0.0494716618635927,
      "grad_norm": 1.375,
      "learning_rate": 0.000257,
      "loss": 6.9289,
      "mean_token_accuracy": 0.10644723549485206,
      "num_tokens": 1177545.0,
      "step": 515
    },
    {
      "entropy": 7.166497278213501,
      "epoch": 0.049951969260326606,
      "grad_norm": 1.71875,
      "learning_rate": 0.0002595,
      "loss": 6.7357,
      "mean_token_accuracy": 0.12091782912611962,
      "num_tokens": 1188944.0,
      "step": 520
    },
    {
      "entropy": 7.2184765338897705,
      "epoch": 0.05043227665706052,
      "grad_norm": 1.375,
      "learning_rate": 0.000262,
      "loss": 6.8006,
      "mean_token_accuracy": 0.11136688962578774,
      "num_tokens": 1199517.0,
      "step": 525
    },
    {
      "entropy": 7.175792932510376,
      "epoch": 0.05091258405379443,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00026450000000000003,
      "loss": 6.7376,
      "mean_token_accuracy": 0.11152568534016609,
      "num_tokens": 1211536.0,
      "step": 530
    },
    {
      "entropy": 7.124487686157226,
      "epoch": 0.05139289145052834,
      "grad_norm": 1.21875,
      "learning_rate": 0.00026700000000000004,
      "loss": 6.6428,
      "mean_token_accuracy": 0.12353976517915725,
      "num_tokens": 1224159.0,
      "step": 535
    },
    {
      "entropy": 7.199137878417969,
      "epoch": 0.05187319884726225,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00026950000000000005,
      "loss": 6.7006,
      "mean_token_accuracy": 0.12614266201853752,
      "num_tokens": 1235226.0,
      "step": 540
    },
    {
      "entropy": 7.1002240657806395,
      "epoch": 0.05235350624399616,
      "grad_norm": 1.203125,
      "learning_rate": 0.00027200000000000005,
      "loss": 6.6409,
      "mean_token_accuracy": 0.12885337471961975,
      "num_tokens": 1245862.0,
      "step": 545
    },
    {
      "entropy": 7.0684572696685795,
      "epoch": 0.052833813640730067,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002745,
      "loss": 6.6448,
      "mean_token_accuracy": 0.12506688982248307,
      "num_tokens": 1257512.0,
      "step": 550
    },
    {
      "entropy": 7.15536675453186,
      "epoch": 0.053314121037463975,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000277,
      "loss": 6.7532,
      "mean_token_accuracy": 0.11791636049747467,
      "num_tokens": 1269317.0,
      "step": 555
    },
    {
      "entropy": 7.05263729095459,
      "epoch": 0.053794428434197884,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002795,
      "loss": 6.6387,
      "mean_token_accuracy": 0.11247576996684075,
      "num_tokens": 1281878.0,
      "step": 560
    },
    {
      "entropy": 7.032025909423828,
      "epoch": 0.05427473583093179,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00028199999999999997,
      "loss": 6.6567,
      "mean_token_accuracy": 0.12334890589118004,
      "num_tokens": 1293178.0,
      "step": 565
    },
    {
      "entropy": 7.021555423736572,
      "epoch": 0.05475504322766571,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002845,
      "loss": 6.641,
      "mean_token_accuracy": 0.12092255204916,
      "num_tokens": 1304969.0,
      "step": 570
    },
    {
      "entropy": 7.173486518859863,
      "epoch": 0.05523535062439962,
      "grad_norm": 1.25,
      "learning_rate": 0.000287,
      "loss": 6.6615,
      "mean_token_accuracy": 0.12557224780321122,
      "num_tokens": 1317301.0,
      "step": 575
    },
    {
      "entropy": 7.086919069290161,
      "epoch": 0.05571565802113353,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002895,
      "loss": 6.6481,
      "mean_token_accuracy": 0.12229160591959953,
      "num_tokens": 1327688.0,
      "step": 580
    },
    {
      "entropy": 7.105197381973267,
      "epoch": 0.056195965417867436,
      "grad_norm": 1.40625,
      "learning_rate": 0.000292,
      "loss": 6.5803,
      "mean_token_accuracy": 0.1251593180000782,
      "num_tokens": 1338750.0,
      "step": 585
    },
    {
      "entropy": 7.099721193313599,
      "epoch": 0.056676272814601344,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002945,
      "loss": 6.7331,
      "mean_token_accuracy": 0.11501363664865494,
      "num_tokens": 1350899.0,
      "step": 590
    },
    {
      "entropy": 7.106310939788818,
      "epoch": 0.05715658021133525,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000297,
      "loss": 6.6048,
      "mean_token_accuracy": 0.12520924732089042,
      "num_tokens": 1363184.0,
      "step": 595
    },
    {
      "entropy": 7.0411724090576175,
      "epoch": 0.05763688760806916,
      "grad_norm": 1.25,
      "learning_rate": 0.0002995,
      "loss": 6.582,
      "mean_token_accuracy": 0.1271129161119461,
      "num_tokens": 1374192.0,
      "step": 600
    },
    {
      "entropy": 6.922922134399414,
      "epoch": 0.05811719500480307,
      "grad_norm": 1.28125,
      "learning_rate": 0.000302,
      "loss": 6.594,
      "mean_token_accuracy": 0.12445875108242035,
      "num_tokens": 1384888.0,
      "step": 605
    },
    {
      "entropy": 7.115780639648437,
      "epoch": 0.05859750240153699,
      "grad_norm": 1.375,
      "learning_rate": 0.0003045,
      "loss": 6.6765,
      "mean_token_accuracy": 0.1298594705760479,
      "num_tokens": 1395139.0,
      "step": 610
    },
    {
      "entropy": 6.956879568099976,
      "epoch": 0.059077809798270896,
      "grad_norm": 1.4375,
      "learning_rate": 0.000307,
      "loss": 6.487,
      "mean_token_accuracy": 0.12400545477867127,
      "num_tokens": 1407219.0,
      "step": 615
    },
    {
      "entropy": 7.013321304321289,
      "epoch": 0.059558117195004805,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003095,
      "loss": 6.6015,
      "mean_token_accuracy": 0.12258810997009277,
      "num_tokens": 1418917.0,
      "step": 620
    },
    {
      "entropy": 7.036109256744385,
      "epoch": 0.060038424591738714,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000312,
      "loss": 6.5567,
      "mean_token_accuracy": 0.12197562456130981,
      "num_tokens": 1430792.0,
      "step": 625
    },
    {
      "entropy": 6.967579412460327,
      "epoch": 0.06051873198847262,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003145,
      "loss": 6.5753,
      "mean_token_accuracy": 0.12886715084314346,
      "num_tokens": 1442825.0,
      "step": 630
    },
    {
      "entropy": 7.002191114425659,
      "epoch": 0.06099903938520653,
      "grad_norm": 1.078125,
      "learning_rate": 0.000317,
      "loss": 6.6694,
      "mean_token_accuracy": 0.1262420743703842,
      "num_tokens": 1454078.0,
      "step": 635
    },
    {
      "entropy": 6.952101945877075,
      "epoch": 0.06147934678194044,
      "grad_norm": 1.8125,
      "learning_rate": 0.0003195,
      "loss": 6.6083,
      "mean_token_accuracy": 0.12252153679728509,
      "num_tokens": 1465164.0,
      "step": 640
    },
    {
      "entropy": 7.042235279083252,
      "epoch": 0.06195965417867435,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000322,
      "loss": 6.5864,
      "mean_token_accuracy": 0.12691230326890945,
      "num_tokens": 1476813.0,
      "step": 645
    },
    {
      "entropy": 6.95318078994751,
      "epoch": 0.06243996157540826,
      "grad_norm": 1.4375,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.6147,
      "mean_token_accuracy": 0.12464898824691772,
      "num_tokens": 1488547.0,
      "step": 650
    },
    {
      "entropy": 7.05318603515625,
      "epoch": 0.06292026897214217,
      "grad_norm": 1.3125,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.6125,
      "mean_token_accuracy": 0.12129621878266335,
      "num_tokens": 1500714.0,
      "step": 655
    },
    {
      "entropy": 6.901967525482178,
      "epoch": 0.06340057636887608,
      "grad_norm": 1.09375,
      "learning_rate": 0.00032950000000000004,
      "loss": 6.5812,
      "mean_token_accuracy": 0.13022611066699027,
      "num_tokens": 1512212.0,
      "step": 660
    },
    {
      "entropy": 6.97374701499939,
      "epoch": 0.06388088376560999,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.5642,
      "mean_token_accuracy": 0.1291198454797268,
      "num_tokens": 1523610.0,
      "step": 665
    },
    {
      "entropy": 6.8925048351287845,
      "epoch": 0.0643611911623439,
      "grad_norm": 1.25,
      "learning_rate": 0.00033450000000000005,
      "loss": 6.5201,
      "mean_token_accuracy": 0.12605366334319115,
      "num_tokens": 1535543.0,
      "step": 670
    },
    {
      "entropy": 7.089626169204712,
      "epoch": 0.06484149855907781,
      "grad_norm": 1.3125,
      "learning_rate": 0.000337,
      "loss": 6.74,
      "mean_token_accuracy": 0.12375304996967315,
      "num_tokens": 1547183.0,
      "step": 675
    },
    {
      "entropy": 6.861551904678345,
      "epoch": 0.06532180595581172,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0003395,
      "loss": 6.4903,
      "mean_token_accuracy": 0.13013082817196847,
      "num_tokens": 1559265.0,
      "step": 680
    },
    {
      "entropy": 6.936911153793335,
      "epoch": 0.06580211335254563,
      "grad_norm": 1.375,
      "learning_rate": 0.000342,
      "loss": 6.5209,
      "mean_token_accuracy": 0.1260334573686123,
      "num_tokens": 1569774.0,
      "step": 685
    },
    {
      "entropy": 6.816721200942993,
      "epoch": 0.06628242074927954,
      "grad_norm": 1.171875,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.4525,
      "mean_token_accuracy": 0.13374078422784805,
      "num_tokens": 1579408.0,
      "step": 690
    },
    {
      "entropy": 6.881241273880005,
      "epoch": 0.06676272814601344,
      "grad_norm": 1.40625,
      "learning_rate": 0.000347,
      "loss": 6.4841,
      "mean_token_accuracy": 0.13311079293489456,
      "num_tokens": 1591849.0,
      "step": 695
    },
    {
      "entropy": 6.893535804748535,
      "epoch": 0.06724303554274735,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003495,
      "loss": 6.5399,
      "mean_token_accuracy": 0.1287438191473484,
      "num_tokens": 1603269.0,
      "step": 700
    },
    {
      "entropy": 6.999371290206909,
      "epoch": 0.06772334293948126,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000352,
      "loss": 6.579,
      "mean_token_accuracy": 0.12688603401184081,
      "num_tokens": 1613639.0,
      "step": 705
    },
    {
      "entropy": 6.96681170463562,
      "epoch": 0.06820365033621517,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003545,
      "loss": 6.6065,
      "mean_token_accuracy": 0.1259676620364189,
      "num_tokens": 1625393.0,
      "step": 710
    },
    {
      "entropy": 6.8337006092071535,
      "epoch": 0.0686839577329491,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000357,
      "loss": 6.4307,
      "mean_token_accuracy": 0.1348931148648262,
      "num_tokens": 1636355.0,
      "step": 715
    },
    {
      "entropy": 6.902727127075195,
      "epoch": 0.069164265129683,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003595,
      "loss": 6.4784,
      "mean_token_accuracy": 0.13106829673051834,
      "num_tokens": 1646157.0,
      "step": 720
    },
    {
      "entropy": 6.7788458347320555,
      "epoch": 0.06964457252641691,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000362,
      "loss": 6.4699,
      "mean_token_accuracy": 0.12550933212041854,
      "num_tokens": 1657827.0,
      "step": 725
    },
    {
      "entropy": 6.847135925292969,
      "epoch": 0.07012487992315082,
      "grad_norm": 1.5,
      "learning_rate": 0.0003645,
      "loss": 6.5532,
      "mean_token_accuracy": 0.13552703633904456,
      "num_tokens": 1669023.0,
      "step": 730
    },
    {
      "entropy": 6.893656158447266,
      "epoch": 0.07060518731988473,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000367,
      "loss": 6.4955,
      "mean_token_accuracy": 0.12944890484213828,
      "num_tokens": 1680048.0,
      "step": 735
    },
    {
      "entropy": 6.9871117115020756,
      "epoch": 0.07108549471661864,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003695,
      "loss": 6.6982,
      "mean_token_accuracy": 0.11849347949028015,
      "num_tokens": 1690907.0,
      "step": 740
    },
    {
      "entropy": 6.907092571258545,
      "epoch": 0.07156580211335255,
      "grad_norm": 1.4609375,
      "learning_rate": 0.000372,
      "loss": 6.525,
      "mean_token_accuracy": 0.13299438506364822,
      "num_tokens": 1702337.0,
      "step": 745
    },
    {
      "entropy": 6.883594512939453,
      "epoch": 0.07204610951008646,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003745,
      "loss": 6.4486,
      "mean_token_accuracy": 0.1341588594019413,
      "num_tokens": 1713242.0,
      "step": 750
    },
    {
      "entropy": 6.749587869644165,
      "epoch": 0.07252641690682037,
      "grad_norm": 1.484375,
      "learning_rate": 0.000377,
      "loss": 6.4226,
      "mean_token_accuracy": 0.13001055493950844,
      "num_tokens": 1724939.0,
      "step": 755
    },
    {
      "entropy": 6.8215819835662845,
      "epoch": 0.07300672430355427,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003795,
      "loss": 6.3946,
      "mean_token_accuracy": 0.1426876053214073,
      "num_tokens": 1737311.0,
      "step": 760
    },
    {
      "entropy": 6.9048539161682125,
      "epoch": 0.07348703170028818,
      "grad_norm": 1.28125,
      "learning_rate": 0.000382,
      "loss": 6.5289,
      "mean_token_accuracy": 0.13060924485325814,
      "num_tokens": 1749528.0,
      "step": 765
    },
    {
      "entropy": 6.729816436767578,
      "epoch": 0.07396733909702209,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003845,
      "loss": 6.3903,
      "mean_token_accuracy": 0.13687298968434333,
      "num_tokens": 1760304.0,
      "step": 770
    },
    {
      "entropy": 6.770515632629395,
      "epoch": 0.074447646493756,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.3784,
      "mean_token_accuracy": 0.13586725369095803,
      "num_tokens": 1772014.0,
      "step": 775
    },
    {
      "entropy": 6.735719966888428,
      "epoch": 0.07492795389048991,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.4403,
      "mean_token_accuracy": 0.13041759729385377,
      "num_tokens": 1784585.0,
      "step": 780
    },
    {
      "entropy": 6.807892465591431,
      "epoch": 0.07540826128722382,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.3763,
      "mean_token_accuracy": 0.1385021425783634,
      "num_tokens": 1795246.0,
      "step": 785
    },
    {
      "entropy": 6.708891773223877,
      "epoch": 0.07588856868395773,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.3674,
      "mean_token_accuracy": 0.13637370616197586,
      "num_tokens": 1806826.0,
      "step": 790
    },
    {
      "entropy": 6.739760875701904,
      "epoch": 0.07636887608069164,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.4892,
      "mean_token_accuracy": 0.13308436870574952,
      "num_tokens": 1818766.0,
      "step": 795
    },
    {
      "entropy": 6.768296575546264,
      "epoch": 0.07684918347742556,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003995,
      "loss": 6.3435,
      "mean_token_accuracy": 0.141699942946434,
      "num_tokens": 1830824.0,
      "step": 800
    },
    {
      "entropy": 6.714781379699707,
      "epoch": 0.07732949087415947,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000402,
      "loss": 6.4306,
      "mean_token_accuracy": 0.13212079629302026,
      "num_tokens": 1841792.0,
      "step": 805
    },
    {
      "entropy": 6.706526756286621,
      "epoch": 0.07780979827089338,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004045,
      "loss": 6.3676,
      "mean_token_accuracy": 0.13363081067800522,
      "num_tokens": 1854683.0,
      "step": 810
    },
    {
      "entropy": 6.775065326690674,
      "epoch": 0.07829010566762729,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.4657,
      "mean_token_accuracy": 0.13291899263858795,
      "num_tokens": 1865826.0,
      "step": 815
    },
    {
      "entropy": 6.768963766098023,
      "epoch": 0.0787704130643612,
      "grad_norm": 1.25,
      "learning_rate": 0.0004095,
      "loss": 6.4149,
      "mean_token_accuracy": 0.13940379321575164,
      "num_tokens": 1876759.0,
      "step": 820
    },
    {
      "entropy": 6.710623407363892,
      "epoch": 0.0792507204610951,
      "grad_norm": 1.09375,
      "learning_rate": 0.000412,
      "loss": 6.4535,
      "mean_token_accuracy": 0.1341131992638111,
      "num_tokens": 1888013.0,
      "step": 825
    },
    {
      "entropy": 6.827510213851928,
      "epoch": 0.07973102785782901,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004145,
      "loss": 6.4581,
      "mean_token_accuracy": 0.1362846978008747,
      "num_tokens": 1898964.0,
      "step": 830
    },
    {
      "entropy": 6.655255270004273,
      "epoch": 0.08021133525456292,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000417,
      "loss": 6.3854,
      "mean_token_accuracy": 0.13690235167741777,
      "num_tokens": 1909728.0,
      "step": 835
    },
    {
      "entropy": 6.653443908691406,
      "epoch": 0.08069164265129683,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004195,
      "loss": 6.3046,
      "mean_token_accuracy": 0.14418013766407967,
      "num_tokens": 1919895.0,
      "step": 840
    },
    {
      "entropy": 6.61823296546936,
      "epoch": 0.08117195004803074,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000422,
      "loss": 6.2275,
      "mean_token_accuracy": 0.14545366540551186,
      "num_tokens": 1930839.0,
      "step": 845
    },
    {
      "entropy": 6.723414564132691,
      "epoch": 0.08165225744476465,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004245,
      "loss": 6.4191,
      "mean_token_accuracy": 0.14012102857232095,
      "num_tokens": 1942605.0,
      "step": 850
    },
    {
      "entropy": 6.665984439849853,
      "epoch": 0.08213256484149856,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000427,
      "loss": 6.4117,
      "mean_token_accuracy": 0.13332209810614587,
      "num_tokens": 1954863.0,
      "step": 855
    },
    {
      "entropy": 6.725205516815185,
      "epoch": 0.08261287223823247,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004295,
      "loss": 6.4879,
      "mean_token_accuracy": 0.1358294278383255,
      "num_tokens": 1966088.0,
      "step": 860
    },
    {
      "entropy": 6.663434267044067,
      "epoch": 0.08309317963496637,
      "grad_norm": 1.421875,
      "learning_rate": 0.000432,
      "loss": 6.3286,
      "mean_token_accuracy": 0.13870726078748702,
      "num_tokens": 1978081.0,
      "step": 865
    },
    {
      "entropy": 6.728060007095337,
      "epoch": 0.08357348703170028,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004345,
      "loss": 6.3965,
      "mean_token_accuracy": 0.12974184826016427,
      "num_tokens": 1991212.0,
      "step": 870
    },
    {
      "entropy": 6.6904360294342045,
      "epoch": 0.08405379442843419,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000437,
      "loss": 6.3822,
      "mean_token_accuracy": 0.13286416828632355,
      "num_tokens": 2002074.0,
      "step": 875
    },
    {
      "entropy": 6.6402044773101805,
      "epoch": 0.0845341018251681,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004395,
      "loss": 6.2916,
      "mean_token_accuracy": 0.14489293247461318,
      "num_tokens": 2013039.0,
      "step": 880
    },
    {
      "entropy": 6.650302171707153,
      "epoch": 0.08501440922190202,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000442,
      "loss": 6.4218,
      "mean_token_accuracy": 0.13508904427289964,
      "num_tokens": 2024744.0,
      "step": 885
    },
    {
      "entropy": 6.657005834579468,
      "epoch": 0.08549471661863593,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004445,
      "loss": 6.2951,
      "mean_token_accuracy": 0.144847309589386,
      "num_tokens": 2035587.0,
      "step": 890
    },
    {
      "entropy": 6.607360124588013,
      "epoch": 0.08597502401536984,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000447,
      "loss": 6.3681,
      "mean_token_accuracy": 0.13551491349935532,
      "num_tokens": 2047924.0,
      "step": 895
    },
    {
      "entropy": 6.6262932300567625,
      "epoch": 0.08645533141210375,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.3086,
      "mean_token_accuracy": 0.14029427096247674,
      "num_tokens": 2060161.0,
      "step": 900
    },
    {
      "entropy": 6.626390981674194,
      "epoch": 0.08693563880883766,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.3231,
      "mean_token_accuracy": 0.14045732319355012,
      "num_tokens": 2071674.0,
      "step": 905
    },
    {
      "entropy": 6.669071340560913,
      "epoch": 0.08741594620557157,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.4579,
      "mean_token_accuracy": 0.13296820148825644,
      "num_tokens": 2084025.0,
      "step": 910
    },
    {
      "entropy": 6.6538848876953125,
      "epoch": 0.08789625360230548,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.3756,
      "mean_token_accuracy": 0.14029246047139168,
      "num_tokens": 2094501.0,
      "step": 915
    },
    {
      "entropy": 6.681252670288086,
      "epoch": 0.08837656099903939,
      "grad_norm": 1.3125,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.3749,
      "mean_token_accuracy": 0.14268773496150972,
      "num_tokens": 2106207.0,
      "step": 920
    },
    {
      "entropy": 6.6069988250732425,
      "epoch": 0.0888568683957733,
      "grad_norm": 1.28125,
      "learning_rate": 0.000462,
      "loss": 6.2643,
      "mean_token_accuracy": 0.14611646607518197,
      "num_tokens": 2116879.0,
      "step": 925
    },
    {
      "entropy": 6.578690099716186,
      "epoch": 0.0893371757925072,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004645,
      "loss": 6.3318,
      "mean_token_accuracy": 0.13842730224132538,
      "num_tokens": 2128884.0,
      "step": 930
    },
    {
      "entropy": 6.4773077964782715,
      "epoch": 0.08981748318924111,
      "grad_norm": 1.109375,
      "learning_rate": 0.000467,
      "loss": 6.2839,
      "mean_token_accuracy": 0.13629047125577926,
      "num_tokens": 2140810.0,
      "step": 935
    },
    {
      "entropy": 6.670573139190674,
      "epoch": 0.09029779058597502,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004695,
      "loss": 6.3424,
      "mean_token_accuracy": 0.13401698023080827,
      "num_tokens": 2152312.0,
      "step": 940
    },
    {
      "entropy": 6.51163682937622,
      "epoch": 0.09077809798270893,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000472,
      "loss": 6.2788,
      "mean_token_accuracy": 0.136185372620821,
      "num_tokens": 2164408.0,
      "step": 945
    },
    {
      "entropy": 6.558056449890136,
      "epoch": 0.09125840537944284,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004745,
      "loss": 6.2898,
      "mean_token_accuracy": 0.1440206691622734,
      "num_tokens": 2176050.0,
      "step": 950
    },
    {
      "entropy": 6.531415700912476,
      "epoch": 0.09173871277617675,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000477,
      "loss": 6.1747,
      "mean_token_accuracy": 0.14442746862769126,
      "num_tokens": 2187354.0,
      "step": 955
    },
    {
      "entropy": 6.446061897277832,
      "epoch": 0.09221902017291066,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004795,
      "loss": 6.1956,
      "mean_token_accuracy": 0.14385264068841935,
      "num_tokens": 2198686.0,
      "step": 960
    },
    {
      "entropy": 6.423680925369263,
      "epoch": 0.09269932756964457,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000482,
      "loss": 6.2228,
      "mean_token_accuracy": 0.14219090715050697,
      "num_tokens": 2209780.0,
      "step": 965
    },
    {
      "entropy": 6.576705598831177,
      "epoch": 0.09317963496637849,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004845,
      "loss": 6.3136,
      "mean_token_accuracy": 0.14841534793376923,
      "num_tokens": 2221760.0,
      "step": 970
    },
    {
      "entropy": 6.538150453567505,
      "epoch": 0.0936599423631124,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000487,
      "loss": 6.3321,
      "mean_token_accuracy": 0.14067101255059242,
      "num_tokens": 2233767.0,
      "step": 975
    },
    {
      "entropy": 6.526997327804565,
      "epoch": 0.0941402497598463,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004895,
      "loss": 6.2696,
      "mean_token_accuracy": 0.13998018503189086,
      "num_tokens": 2244841.0,
      "step": 980
    },
    {
      "entropy": 6.529129076004028,
      "epoch": 0.09462055715658022,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000492,
      "loss": 6.1252,
      "mean_token_accuracy": 0.14948576986789702,
      "num_tokens": 2255620.0,
      "step": 985
    },
    {
      "entropy": 6.469129943847657,
      "epoch": 0.09510086455331412,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004945,
      "loss": 6.3184,
      "mean_token_accuracy": 0.14613131657242776,
      "num_tokens": 2266345.0,
      "step": 990
    },
    {
      "entropy": 6.5079692840576175,
      "epoch": 0.09558117195004803,
      "grad_norm": 1.296875,
      "learning_rate": 0.000497,
      "loss": 6.2527,
      "mean_token_accuracy": 0.1458457499742508,
      "num_tokens": 2277190.0,
      "step": 995
    },
    {
      "entropy": 6.46534948348999,
      "epoch": 0.09606147934678194,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004995,
      "loss": 6.1264,
      "mean_token_accuracy": 0.15072371661663056,
      "num_tokens": 2288241.0,
      "step": 1000
    },
    {
      "entropy": 6.495367813110351,
      "epoch": 0.09654178674351585,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999999983283737,
      "loss": 6.2644,
      "mean_token_accuracy": 0.1425079435110092,
      "num_tokens": 2301092.0,
      "step": 1005
    },
    {
      "entropy": 6.48754186630249,
      "epoch": 0.09702209414024976,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999999915373924,
      "loss": 6.1828,
      "mean_token_accuracy": 0.14111206606030463,
      "num_tokens": 2313622.0,
      "step": 1010
    },
    {
      "entropy": 6.343118476867676,
      "epoch": 0.09750240153698367,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999999795225793,
      "loss": 6.1425,
      "mean_token_accuracy": 0.14880686104297638,
      "num_tokens": 2325699.0,
      "step": 1015
    },
    {
      "entropy": 6.502762937545777,
      "epoch": 0.09798270893371758,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999999622839347,
      "loss": 6.2125,
      "mean_token_accuracy": 0.1470236748456955,
      "num_tokens": 2337816.0,
      "step": 1020
    },
    {
      "entropy": 6.620756721496582,
      "epoch": 0.09846301633045149,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999999398214593,
      "loss": 6.3494,
      "mean_token_accuracy": 0.1372337594628334,
      "num_tokens": 2348793.0,
      "step": 1025
    },
    {
      "entropy": 6.463247442245484,
      "epoch": 0.0989433237271854,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999999121351532,
      "loss": 6.2452,
      "mean_token_accuracy": 0.14675871282815933,
      "num_tokens": 2360209.0,
      "step": 1030
    },
    {
      "entropy": 6.566287660598755,
      "epoch": 0.0994236311239193,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004999998792250173,
      "loss": 6.2558,
      "mean_token_accuracy": 0.14142224341630935,
      "num_tokens": 2370597.0,
      "step": 1035
    },
    {
      "entropy": 6.506043529510498,
      "epoch": 0.09990393852065321,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004999998410910524,
      "loss": 6.2907,
      "mean_token_accuracy": 0.14004258513450624,
      "num_tokens": 2381466.0,
      "step": 1040
    },
    {
      "entropy": 6.63688383102417,
      "epoch": 0.10038424591738712,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004999997977332592,
      "loss": 6.3923,
      "mean_token_accuracy": 0.1357167035341263,
      "num_tokens": 2394507.0,
      "step": 1045
    },
    {
      "entropy": 6.429609298706055,
      "epoch": 0.10086455331412104,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004999997491516389,
      "loss": 6.2382,
      "mean_token_accuracy": 0.14680789411067963,
      "num_tokens": 2406367.0,
      "step": 1050
    },
    {
      "entropy": 6.439948606491089,
      "epoch": 0.10134486071085495,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004999996953461925,
      "loss": 6.2089,
      "mean_token_accuracy": 0.1446738436818123,
      "num_tokens": 2416962.0,
      "step": 1055
    },
    {
      "entropy": 6.582912969589233,
      "epoch": 0.10182516810758886,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004999996363169212,
      "loss": 6.2643,
      "mean_token_accuracy": 0.13576384484767914,
      "num_tokens": 2429441.0,
      "step": 1060
    },
    {
      "entropy": 6.374668884277344,
      "epoch": 0.10230547550432277,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999995720638266,
      "loss": 6.1497,
      "mean_token_accuracy": 0.1466536670923233,
      "num_tokens": 2441360.0,
      "step": 1065
    },
    {
      "entropy": 6.437990379333496,
      "epoch": 0.10278578290105668,
      "grad_norm": 1.234375,
      "learning_rate": 0.00049999950258691,
      "loss": 6.1671,
      "mean_token_accuracy": 0.14428199231624603,
      "num_tokens": 2452928.0,
      "step": 1070
    },
    {
      "entropy": 6.362707710266113,
      "epoch": 0.10326609029779059,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999994278861731,
      "loss": 6.0991,
      "mean_token_accuracy": 0.14837993010878564,
      "num_tokens": 2464218.0,
      "step": 1075
    },
    {
      "entropy": 6.4544370651245115,
      "epoch": 0.1037463976945245,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004999993479616175,
      "loss": 6.183,
      "mean_token_accuracy": 0.14867693558335304,
      "num_tokens": 2475084.0,
      "step": 1080
    },
    {
      "entropy": 6.419835948944092,
      "epoch": 0.1042267050912584,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004999992628132451,
      "loss": 6.1854,
      "mean_token_accuracy": 0.14430805891752244,
      "num_tokens": 2486845.0,
      "step": 1085
    },
    {
      "entropy": 6.383818006515503,
      "epoch": 0.10470701248799232,
      "grad_norm": 1.375,
      "learning_rate": 0.0004999991724410582,
      "loss": 6.2687,
      "mean_token_accuracy": 0.14187756553292274,
      "num_tokens": 2498049.0,
      "step": 1090
    },
    {
      "entropy": 6.444679689407349,
      "epoch": 0.10518731988472622,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999990768450583,
      "loss": 6.1959,
      "mean_token_accuracy": 0.14557206630706787,
      "num_tokens": 2509511.0,
      "step": 1095
    },
    {
      "entropy": 6.473533487319946,
      "epoch": 0.10566762728146013,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999989760252482,
      "loss": 6.1493,
      "mean_token_accuracy": 0.15422460734844207,
      "num_tokens": 2521518.0,
      "step": 1100
    },
    {
      "entropy": 6.244095516204834,
      "epoch": 0.10614793467819404,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999988699816299,
      "loss": 5.9871,
      "mean_token_accuracy": 0.1582183226943016,
      "num_tokens": 2533141.0,
      "step": 1105
    },
    {
      "entropy": 6.313006591796875,
      "epoch": 0.10662824207492795,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999987587142058,
      "loss": 5.987,
      "mean_token_accuracy": 0.1536654606461525,
      "num_tokens": 2542907.0,
      "step": 1110
    },
    {
      "entropy": 6.3394622802734375,
      "epoch": 0.10710854947166186,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999986422229789,
      "loss": 6.1417,
      "mean_token_accuracy": 0.14201202169060706,
      "num_tokens": 2554497.0,
      "step": 1115
    },
    {
      "entropy": 6.543412351608277,
      "epoch": 0.10758885686839577,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999985205079514,
      "loss": 6.2688,
      "mean_token_accuracy": 0.14553593546152116,
      "num_tokens": 2567035.0,
      "step": 1120
    },
    {
      "entropy": 6.282679748535156,
      "epoch": 0.10806916426512968,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004999983935691265,
      "loss": 6.0562,
      "mean_token_accuracy": 0.15392716675996782,
      "num_tokens": 2578003.0,
      "step": 1125
    },
    {
      "entropy": 6.480337524414063,
      "epoch": 0.10854947166186359,
      "grad_norm": 1.109375,
      "learning_rate": 0.000499998261406507,
      "loss": 6.3434,
      "mean_token_accuracy": 0.1433580659329891,
      "num_tokens": 2590272.0,
      "step": 1130
    },
    {
      "entropy": 6.362969541549683,
      "epoch": 0.10902977905859751,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999981240200958,
      "loss": 6.1541,
      "mean_token_accuracy": 0.145244736969471,
      "num_tokens": 2602874.0,
      "step": 1135
    },
    {
      "entropy": 6.355509281158447,
      "epoch": 0.10951008645533142,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004999979814098966,
      "loss": 6.1054,
      "mean_token_accuracy": 0.1554892733693123,
      "num_tokens": 2613835.0,
      "step": 1140
    },
    {
      "entropy": 6.463612365722656,
      "epoch": 0.10999039385206533,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999978335759121,
      "loss": 6.238,
      "mean_token_accuracy": 0.1424919068813324,
      "num_tokens": 2624986.0,
      "step": 1145
    },
    {
      "entropy": 6.391502523422242,
      "epoch": 0.11047070124879924,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999976805181461,
      "loss": 6.2362,
      "mean_token_accuracy": 0.14005767554044724,
      "num_tokens": 2637654.0,
      "step": 1150
    },
    {
      "entropy": 6.405415058135986,
      "epoch": 0.11095100864553314,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000499997522236602,
      "loss": 6.1399,
      "mean_token_accuracy": 0.15252109467983246,
      "num_tokens": 2648947.0,
      "step": 1155
    },
    {
      "entropy": 6.3985276222229,
      "epoch": 0.11143131604226705,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999973587312837,
      "loss": 6.1239,
      "mean_token_accuracy": 0.14815355390310286,
      "num_tokens": 2659573.0,
      "step": 1160
    },
    {
      "entropy": 6.249864339828491,
      "epoch": 0.11191162343900096,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999971900021947,
      "loss": 6.0978,
      "mean_token_accuracy": 0.1548936977982521,
      "num_tokens": 2671491.0,
      "step": 1165
    },
    {
      "entropy": 6.381734895706177,
      "epoch": 0.11239193083573487,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999970160493391,
      "loss": 6.0828,
      "mean_token_accuracy": 0.1447075366973877,
      "num_tokens": 2682705.0,
      "step": 1170
    },
    {
      "entropy": 6.272091722488403,
      "epoch": 0.11287223823246878,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999968368727209,
      "loss": 6.0287,
      "mean_token_accuracy": 0.14921507537364959,
      "num_tokens": 2693539.0,
      "step": 1175
    },
    {
      "entropy": 6.254890108108521,
      "epoch": 0.11335254562920269,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999966524723442,
      "loss": 6.075,
      "mean_token_accuracy": 0.15476031005382537,
      "num_tokens": 2705403.0,
      "step": 1180
    },
    {
      "entropy": 6.456190967559815,
      "epoch": 0.1138328530259366,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004999964628482135,
      "loss": 6.1955,
      "mean_token_accuracy": 0.1451671041548252,
      "num_tokens": 2717320.0,
      "step": 1185
    },
    {
      "entropy": 6.350939464569092,
      "epoch": 0.1143131604226705,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999962680003328,
      "loss": 6.1526,
      "mean_token_accuracy": 0.1435933716595173,
      "num_tokens": 2730056.0,
      "step": 1190
    },
    {
      "entropy": 6.3369707584381105,
      "epoch": 0.11479346781940442,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000499996067928707,
      "loss": 6.0253,
      "mean_token_accuracy": 0.15263447314500808,
      "num_tokens": 2741399.0,
      "step": 1195
    },
    {
      "entropy": 6.4518022537231445,
      "epoch": 0.11527377521613832,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004999958626333406,
      "loss": 6.2564,
      "mean_token_accuracy": 0.1389774426817894,
      "num_tokens": 2752277.0,
      "step": 1200
    },
    {
      "entropy": 6.364963531494141,
      "epoch": 0.11575408261287223,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999956521142383,
      "loss": 6.134,
      "mean_token_accuracy": 0.14522297084331512,
      "num_tokens": 2763172.0,
      "step": 1205
    },
    {
      "entropy": 6.406414270401001,
      "epoch": 0.11623439000960614,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999954363714051,
      "loss": 6.1445,
      "mean_token_accuracy": 0.15052591264247894,
      "num_tokens": 2775096.0,
      "step": 1210
    },
    {
      "entropy": 6.3844269752502445,
      "epoch": 0.11671469740634005,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999952154048459,
      "loss": 6.221,
      "mean_token_accuracy": 0.1505927786231041,
      "num_tokens": 2786007.0,
      "step": 1215
    },
    {
      "entropy": 6.287612104415894,
      "epoch": 0.11719500480307397,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000499994989214566,
      "loss": 6.0292,
      "mean_token_accuracy": 0.15772099494934083,
      "num_tokens": 2797265.0,
      "step": 1220
    },
    {
      "entropy": 6.365857458114624,
      "epoch": 0.11767531219980788,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999947578005705,
      "loss": 6.1966,
      "mean_token_accuracy": 0.14764621183276178,
      "num_tokens": 2809427.0,
      "step": 1225
    },
    {
      "entropy": 6.3595850467681885,
      "epoch": 0.11815561959654179,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999945211628648,
      "loss": 6.1154,
      "mean_token_accuracy": 0.15281978696584703,
      "num_tokens": 2822880.0,
      "step": 1230
    },
    {
      "entropy": 6.394454145431519,
      "epoch": 0.1186359269932757,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999942793014544,
      "loss": 6.201,
      "mean_token_accuracy": 0.14849669709801674,
      "num_tokens": 2834760.0,
      "step": 1235
    },
    {
      "entropy": 6.325704431533813,
      "epoch": 0.11911623439000961,
      "grad_norm": 1.109375,
      "learning_rate": 0.000499994032216345,
      "loss": 6.1413,
      "mean_token_accuracy": 0.14735834896564484,
      "num_tokens": 2846517.0,
      "step": 1240
    },
    {
      "entropy": 6.302383375167847,
      "epoch": 0.11959654178674352,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999937799075422,
      "loss": 6.0306,
      "mean_token_accuracy": 0.15673189759254455,
      "num_tokens": 2857280.0,
      "step": 1245
    },
    {
      "entropy": 6.341720819473267,
      "epoch": 0.12007684918347743,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499993522375052,
      "loss": 6.0833,
      "mean_token_accuracy": 0.15285103321075438,
      "num_tokens": 2869735.0,
      "step": 1250
    },
    {
      "entropy": 6.2822376728057865,
      "epoch": 0.12055715658021134,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999932596188802,
      "loss": 6.0779,
      "mean_token_accuracy": 0.15265348553657532,
      "num_tokens": 2881681.0,
      "step": 1255
    },
    {
      "entropy": 6.21174373626709,
      "epoch": 0.12103746397694524,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999929916390331,
      "loss": 6.0949,
      "mean_token_accuracy": 0.154794842004776,
      "num_tokens": 2893828.0,
      "step": 1260
    },
    {
      "entropy": 6.129497814178467,
      "epoch": 0.12151777137367915,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999927184355169,
      "loss": 6.036,
      "mean_token_accuracy": 0.1517324849963188,
      "num_tokens": 2905804.0,
      "step": 1265
    },
    {
      "entropy": 6.52538423538208,
      "epoch": 0.12199807877041306,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999924400083377,
      "loss": 6.1836,
      "mean_token_accuracy": 0.14400093555450438,
      "num_tokens": 2918709.0,
      "step": 1270
    },
    {
      "entropy": 6.26390323638916,
      "epoch": 0.12247838616714697,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999921563575022,
      "loss": 6.0853,
      "mean_token_accuracy": 0.1494814559817314,
      "num_tokens": 2930589.0,
      "step": 1275
    },
    {
      "entropy": 6.3428949356079105,
      "epoch": 0.12295869356388088,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999918674830169,
      "loss": 6.15,
      "mean_token_accuracy": 0.15355577766895295,
      "num_tokens": 2940635.0,
      "step": 1280
    },
    {
      "entropy": 6.185381269454956,
      "epoch": 0.12343900096061479,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999915733848886,
      "loss": 6.0204,
      "mean_token_accuracy": 0.15577883869409562,
      "num_tokens": 2952731.0,
      "step": 1285
    },
    {
      "entropy": 6.257724475860596,
      "epoch": 0.1239193083573487,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000499991274063124,
      "loss": 6.0307,
      "mean_token_accuracy": 0.15278594493865966,
      "num_tokens": 2964544.0,
      "step": 1290
    },
    {
      "entropy": 6.300468444824219,
      "epoch": 0.12439961575408261,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004999909695177301,
      "loss": 5.9802,
      "mean_token_accuracy": 0.15509692430496216,
      "num_tokens": 2974996.0,
      "step": 1295
    },
    {
      "entropy": 6.149973678588867,
      "epoch": 0.12487992315081652,
      "grad_norm": 1.109375,
      "learning_rate": 0.000499990659748714,
      "loss": 5.9729,
      "mean_token_accuracy": 0.15493070483207702,
      "num_tokens": 2986496.0,
      "step": 1300
    },
    {
      "entropy": 6.232842063903808,
      "epoch": 0.12536023054755044,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999903447560828,
      "loss": 6.0612,
      "mean_token_accuracy": 0.15265824794769287,
      "num_tokens": 2997693.0,
      "step": 1305
    },
    {
      "entropy": 6.226562786102295,
      "epoch": 0.12584053794428435,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999900245398439,
      "loss": 6.0157,
      "mean_token_accuracy": 0.15674570202827454,
      "num_tokens": 3009155.0,
      "step": 1310
    },
    {
      "entropy": 6.221279764175415,
      "epoch": 0.12632084534101826,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999896991000047,
      "loss": 5.9339,
      "mean_token_accuracy": 0.16307595893740653,
      "num_tokens": 3020944.0,
      "step": 1315
    },
    {
      "entropy": 6.228196477890014,
      "epoch": 0.12680115273775217,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999893684365729,
      "loss": 5.9926,
      "mean_token_accuracy": 0.15848468393087387,
      "num_tokens": 3032893.0,
      "step": 1320
    },
    {
      "entropy": 6.234899044036865,
      "epoch": 0.12728146013448607,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999890325495559,
      "loss": 6.0734,
      "mean_token_accuracy": 0.14982036575675012,
      "num_tokens": 3044937.0,
      "step": 1325
    },
    {
      "entropy": 6.239958381652832,
      "epoch": 0.12776176753121998,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999886914389617,
      "loss": 5.9763,
      "mean_token_accuracy": 0.16002579778432846,
      "num_tokens": 3056071.0,
      "step": 1330
    },
    {
      "entropy": 6.2782426357269285,
      "epoch": 0.1282420749279539,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004999883451047981,
      "loss": 6.1258,
      "mean_token_accuracy": 0.15075602605938912,
      "num_tokens": 3067724.0,
      "step": 1335
    },
    {
      "entropy": 6.279479598999023,
      "epoch": 0.1287223823246878,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999879935470733,
      "loss": 6.029,
      "mean_token_accuracy": 0.159932242333889,
      "num_tokens": 3080779.0,
      "step": 1340
    },
    {
      "entropy": 6.198295736312867,
      "epoch": 0.1292026897214217,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999876367657954,
      "loss": 6.0253,
      "mean_token_accuracy": 0.15193908512592316,
      "num_tokens": 3091991.0,
      "step": 1345
    },
    {
      "entropy": 6.165930318832397,
      "epoch": 0.12968299711815562,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999872747609725,
      "loss": 6.0053,
      "mean_token_accuracy": 0.15580039322376252,
      "num_tokens": 3103713.0,
      "step": 1350
    },
    {
      "entropy": 6.256294059753418,
      "epoch": 0.13016330451488953,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999869075326132,
      "loss": 6.0243,
      "mean_token_accuracy": 0.15310140401124955,
      "num_tokens": 3115891.0,
      "step": 1355
    },
    {
      "entropy": 6.100836944580078,
      "epoch": 0.13064361191162344,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499986535080726,
      "loss": 5.8932,
      "mean_token_accuracy": 0.16618788689374925,
      "num_tokens": 3126223.0,
      "step": 1360
    },
    {
      "entropy": 6.277415800094604,
      "epoch": 0.13112391930835735,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999861574053196,
      "loss": 6.0803,
      "mean_token_accuracy": 0.1464972734451294,
      "num_tokens": 3137537.0,
      "step": 1365
    },
    {
      "entropy": 6.311663389205933,
      "epoch": 0.13160422670509125,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999857745064027,
      "loss": 6.0675,
      "mean_token_accuracy": 0.15843178257346152,
      "num_tokens": 3150126.0,
      "step": 1370
    },
    {
      "entropy": 6.187855768203735,
      "epoch": 0.13208453410182516,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499985386383984,
      "loss": 6.0576,
      "mean_token_accuracy": 0.15574488639831544,
      "num_tokens": 3161684.0,
      "step": 1375
    },
    {
      "entropy": 6.19967212677002,
      "epoch": 0.13256484149855907,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999849930380729,
      "loss": 5.9453,
      "mean_token_accuracy": 0.1588939517736435,
      "num_tokens": 3172342.0,
      "step": 1380
    },
    {
      "entropy": 6.310555124282837,
      "epoch": 0.13304514889529298,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999845944686781,
      "loss": 6.0841,
      "mean_token_accuracy": 0.15674352198839187,
      "num_tokens": 3183681.0,
      "step": 1385
    },
    {
      "entropy": 6.053848028182983,
      "epoch": 0.1335254562920269,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999841906758093,
      "loss": 5.8837,
      "mean_token_accuracy": 0.16934000104665756,
      "num_tokens": 3194122.0,
      "step": 1390
    },
    {
      "entropy": 6.18535680770874,
      "epoch": 0.1340057636887608,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999837816594757,
      "loss": 5.9358,
      "mean_token_accuracy": 0.15829690545797348,
      "num_tokens": 3205719.0,
      "step": 1395
    },
    {
      "entropy": 6.125392961502075,
      "epoch": 0.1344860710854947,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999833674196865,
      "loss": 5.9729,
      "mean_token_accuracy": 0.15530545860528946,
      "num_tokens": 3216262.0,
      "step": 1400
    },
    {
      "entropy": 6.297972679138184,
      "epoch": 0.13496637848222862,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999829479564518,
      "loss": 6.0252,
      "mean_token_accuracy": 0.1493932545185089,
      "num_tokens": 3228035.0,
      "step": 1405
    },
    {
      "entropy": 6.070080995559692,
      "epoch": 0.13544668587896252,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499982523269781,
      "loss": 5.8804,
      "mean_token_accuracy": 0.16375838965177536,
      "num_tokens": 3239643.0,
      "step": 1410
    },
    {
      "entropy": 6.191074371337891,
      "epoch": 0.13592699327569643,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999820933596842,
      "loss": 5.9715,
      "mean_token_accuracy": 0.16024953722953797,
      "num_tokens": 3250293.0,
      "step": 1415
    },
    {
      "entropy": 6.235940551757812,
      "epoch": 0.13640730067243034,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499981658226171,
      "loss": 6.0849,
      "mean_token_accuracy": 0.14876422584056853,
      "num_tokens": 3262448.0,
      "step": 1420
    },
    {
      "entropy": 6.09799861907959,
      "epoch": 0.13688760806916425,
      "grad_norm": 1.078125,
      "learning_rate": 0.000499981217869252,
      "loss": 5.9479,
      "mean_token_accuracy": 0.16196773797273636,
      "num_tokens": 3274041.0,
      "step": 1425
    },
    {
      "entropy": 6.258628892898559,
      "epoch": 0.1373679154658982,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499980772288937,
      "loss": 5.9781,
      "mean_token_accuracy": 0.15339865237474443,
      "num_tokens": 3285063.0,
      "step": 1430
    },
    {
      "entropy": 6.194421911239624,
      "epoch": 0.1378482228626321,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999803214852367,
      "loss": 5.9904,
      "mean_token_accuracy": 0.15928526520729064,
      "num_tokens": 3297031.0,
      "step": 1435
    },
    {
      "entropy": 6.193781995773316,
      "epoch": 0.138328530259366,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999798654581613,
      "loss": 5.9831,
      "mean_token_accuracy": 0.15303055197000504,
      "num_tokens": 3308187.0,
      "step": 1440
    },
    {
      "entropy": 6.149940299987793,
      "epoch": 0.13880883765609991,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999794042077214,
      "loss": 5.9776,
      "mean_token_accuracy": 0.15377987325191497,
      "num_tokens": 3319439.0,
      "step": 1445
    },
    {
      "entropy": 6.193714046478272,
      "epoch": 0.13928914505283382,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999789377339279,
      "loss": 5.9825,
      "mean_token_accuracy": 0.15619124472141266,
      "num_tokens": 3331715.0,
      "step": 1450
    },
    {
      "entropy": 6.1586973667144775,
      "epoch": 0.13976945244956773,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999784660367915,
      "loss": 5.9445,
      "mean_token_accuracy": 0.16621134728193282,
      "num_tokens": 3342281.0,
      "step": 1455
    },
    {
      "entropy": 6.333666086196899,
      "epoch": 0.14024975984630164,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999779891163231,
      "loss": 6.0767,
      "mean_token_accuracy": 0.1537630721926689,
      "num_tokens": 3353530.0,
      "step": 1460
    },
    {
      "entropy": 6.150418090820312,
      "epoch": 0.14073006724303555,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999775069725339,
      "loss": 6.0413,
      "mean_token_accuracy": 0.14952244013547897,
      "num_tokens": 3365083.0,
      "step": 1465
    },
    {
      "entropy": 6.237746095657348,
      "epoch": 0.14121037463976946,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499977019605435,
      "loss": 5.994,
      "mean_token_accuracy": 0.15313961654901503,
      "num_tokens": 3378143.0,
      "step": 1470
    },
    {
      "entropy": 6.189295482635498,
      "epoch": 0.14169068203650337,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999765270150378,
      "loss": 5.9751,
      "mean_token_accuracy": 0.15560294240713118,
      "num_tokens": 3388438.0,
      "step": 1475
    },
    {
      "entropy": 6.083598518371582,
      "epoch": 0.14217098943323728,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999760292013536,
      "loss": 5.9712,
      "mean_token_accuracy": 0.1537217065691948,
      "num_tokens": 3400150.0,
      "step": 1480
    },
    {
      "entropy": 6.116258144378662,
      "epoch": 0.14265129682997119,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999755261643941,
      "loss": 5.842,
      "mean_token_accuracy": 0.1626240774989128,
      "num_tokens": 3411073.0,
      "step": 1485
    },
    {
      "entropy": 6.100852298736572,
      "epoch": 0.1431316042267051,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999750179041709,
      "loss": 5.8908,
      "mean_token_accuracy": 0.17015900164842607,
      "num_tokens": 3422196.0,
      "step": 1490
    },
    {
      "entropy": 6.193173599243164,
      "epoch": 0.143611911623439,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999745044206959,
      "loss": 5.9324,
      "mean_token_accuracy": 0.15437426567077636,
      "num_tokens": 3432568.0,
      "step": 1495
    },
    {
      "entropy": 6.133666181564331,
      "epoch": 0.1440922190201729,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999739857139809,
      "loss": 5.9461,
      "mean_token_accuracy": 0.15741463452577592,
      "num_tokens": 3444096.0,
      "step": 1500
    },
    {
      "entropy": 6.035371446609497,
      "epoch": 0.14457252641690682,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499973461784038,
      "loss": 5.8437,
      "mean_token_accuracy": 0.164261032640934,
      "num_tokens": 3455502.0,
      "step": 1505
    },
    {
      "entropy": 6.17965931892395,
      "epoch": 0.14505283381364073,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999729326308792,
      "loss": 5.9825,
      "mean_token_accuracy": 0.152863709628582,
      "num_tokens": 3467753.0,
      "step": 1510
    },
    {
      "entropy": 6.079440689086914,
      "epoch": 0.14553314121037464,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000499972398254517,
      "loss": 5.8548,
      "mean_token_accuracy": 0.16266588419675826,
      "num_tokens": 3479206.0,
      "step": 1515
    },
    {
      "entropy": 6.087345123291016,
      "epoch": 0.14601344860710855,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499971858654964,
      "loss": 5.8727,
      "mean_token_accuracy": 0.16008733659982682,
      "num_tokens": 3490914.0,
      "step": 1520
    },
    {
      "entropy": 6.117412471771241,
      "epoch": 0.14649375600384246,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999713138322321,
      "loss": 5.9381,
      "mean_token_accuracy": 0.15474505573511124,
      "num_tokens": 3502932.0,
      "step": 1525
    },
    {
      "entropy": 6.052666330337525,
      "epoch": 0.14697406340057637,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999707637863346,
      "loss": 5.8568,
      "mean_token_accuracy": 0.16760103702545165,
      "num_tokens": 3515071.0,
      "step": 1530
    },
    {
      "entropy": 6.197118139266967,
      "epoch": 0.14745437079731027,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999702085172838,
      "loss": 5.9731,
      "mean_token_accuracy": 0.15752421617507933,
      "num_tokens": 3526845.0,
      "step": 1535
    },
    {
      "entropy": 6.107962656021118,
      "epoch": 0.14793467819404418,
      "grad_norm": 1.25,
      "learning_rate": 0.0004999696480250929,
      "loss": 5.97,
      "mean_token_accuracy": 0.15250069200992583,
      "num_tokens": 3538810.0,
      "step": 1540
    },
    {
      "entropy": 6.2200414657592775,
      "epoch": 0.1484149855907781,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999690823097747,
      "loss": 5.9264,
      "mean_token_accuracy": 0.16580826342105864,
      "num_tokens": 3549570.0,
      "step": 1545
    },
    {
      "entropy": 5.904806661605835,
      "epoch": 0.148895292987512,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999685113713426,
      "loss": 5.8501,
      "mean_token_accuracy": 0.16919772773981095,
      "num_tokens": 3560937.0,
      "step": 1550
    },
    {
      "entropy": 6.123501825332641,
      "epoch": 0.1493756003842459,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999679352098096,
      "loss": 5.8894,
      "mean_token_accuracy": 0.15619551986455918,
      "num_tokens": 3573603.0,
      "step": 1555
    },
    {
      "entropy": 6.082352209091186,
      "epoch": 0.14985590778097982,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999673538251891,
      "loss": 5.9088,
      "mean_token_accuracy": 0.1532599925994873,
      "num_tokens": 3586034.0,
      "step": 1560
    },
    {
      "entropy": 6.018591833114624,
      "epoch": 0.15033621517771373,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999667672174947,
      "loss": 5.8289,
      "mean_token_accuracy": 0.1625208377838135,
      "num_tokens": 3597305.0,
      "step": 1565
    },
    {
      "entropy": 6.1929418563842775,
      "epoch": 0.15081652257444764,
      "grad_norm": 1.046875,
      "learning_rate": 0.00049996617538674,
      "loss": 5.9833,
      "mean_token_accuracy": 0.14946810826659201,
      "num_tokens": 3608758.0,
      "step": 1570
    },
    {
      "entropy": 6.1501671314239506,
      "epoch": 0.15129682997118155,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999655783329386,
      "loss": 5.9919,
      "mean_token_accuracy": 0.15281759947538376,
      "num_tokens": 3621508.0,
      "step": 1575
    },
    {
      "entropy": 6.159276962280273,
      "epoch": 0.15177713736791545,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999649760561046,
      "loss": 5.9382,
      "mean_token_accuracy": 0.16324293464422227,
      "num_tokens": 3634826.0,
      "step": 1580
    },
    {
      "entropy": 5.917337799072266,
      "epoch": 0.15225744476464936,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999643685562519,
      "loss": 5.7775,
      "mean_token_accuracy": 0.17960015386343003,
      "num_tokens": 3646427.0,
      "step": 1585
    },
    {
      "entropy": 6.045775270462036,
      "epoch": 0.15273775216138327,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999637558333945,
      "loss": 5.9014,
      "mean_token_accuracy": 0.16223143190145492,
      "num_tokens": 3658157.0,
      "step": 1590
    },
    {
      "entropy": 6.132830429077148,
      "epoch": 0.15321805955811718,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999631378875467,
      "loss": 5.9486,
      "mean_token_accuracy": 0.15801500529050827,
      "num_tokens": 3670695.0,
      "step": 1595
    },
    {
      "entropy": 6.065447902679443,
      "epoch": 0.15369836695485112,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999625147187228,
      "loss": 5.8667,
      "mean_token_accuracy": 0.1583435446023941,
      "num_tokens": 3683379.0,
      "step": 1600
    },
    {
      "entropy": 6.047661972045899,
      "epoch": 0.15417867435158503,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999618863269373,
      "loss": 5.7623,
      "mean_token_accuracy": 0.17143424898386,
      "num_tokens": 3694411.0,
      "step": 1605
    },
    {
      "entropy": 5.992631816864014,
      "epoch": 0.15465898174831894,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999612527122049,
      "loss": 5.8231,
      "mean_token_accuracy": 0.1685330703854561,
      "num_tokens": 3703833.0,
      "step": 1610
    },
    {
      "entropy": 6.119769525527954,
      "epoch": 0.15513928914505284,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999606138745402,
      "loss": 5.8962,
      "mean_token_accuracy": 0.16868149489164352,
      "num_tokens": 3714702.0,
      "step": 1615
    },
    {
      "entropy": 6.150318717956543,
      "epoch": 0.15561959654178675,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999599698139581,
      "loss": 5.9627,
      "mean_token_accuracy": 0.16385385543107986,
      "num_tokens": 3726000.0,
      "step": 1620
    },
    {
      "entropy": 6.108389234542846,
      "epoch": 0.15609990393852066,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999593205304734,
      "loss": 5.859,
      "mean_token_accuracy": 0.15902963280677795,
      "num_tokens": 3737681.0,
      "step": 1625
    },
    {
      "entropy": 6.1753983974456785,
      "epoch": 0.15658021133525457,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999586660241012,
      "loss": 5.9649,
      "mean_token_accuracy": 0.15404864847660066,
      "num_tokens": 3749477.0,
      "step": 1630
    },
    {
      "entropy": 6.151984453201294,
      "epoch": 0.15706051873198848,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999580062948569,
      "loss": 5.9426,
      "mean_token_accuracy": 0.15559921264648438,
      "num_tokens": 3759946.0,
      "step": 1635
    },
    {
      "entropy": 6.033643436431885,
      "epoch": 0.1575408261287224,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999573413427556,
      "loss": 5.8405,
      "mean_token_accuracy": 0.160524982213974,
      "num_tokens": 3770147.0,
      "step": 1640
    },
    {
      "entropy": 6.0757527351379395,
      "epoch": 0.1580211335254563,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999566711678128,
      "loss": 5.794,
      "mean_token_accuracy": 0.16993406414985657,
      "num_tokens": 3781028.0,
      "step": 1645
    },
    {
      "entropy": 6.029065990447998,
      "epoch": 0.1585014409221902,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999559957700442,
      "loss": 5.8655,
      "mean_token_accuracy": 0.15539842694997788,
      "num_tokens": 3792529.0,
      "step": 1650
    },
    {
      "entropy": 6.171193170547485,
      "epoch": 0.15898174831892412,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999553151494653,
      "loss": 5.9723,
      "mean_token_accuracy": 0.16454491764307022,
      "num_tokens": 3804630.0,
      "step": 1655
    },
    {
      "entropy": 5.867986536026001,
      "epoch": 0.15946205571565802,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999546293060919,
      "loss": 5.7366,
      "mean_token_accuracy": 0.17225825041532516,
      "num_tokens": 3815019.0,
      "step": 1660
    },
    {
      "entropy": 6.107065629959107,
      "epoch": 0.15994236311239193,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00049995393823994,
      "loss": 5.8429,
      "mean_token_accuracy": 0.16143251210451126,
      "num_tokens": 3825648.0,
      "step": 1665
    },
    {
      "entropy": 6.041478061676026,
      "epoch": 0.16042267050912584,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999532419510255,
      "loss": 5.8323,
      "mean_token_accuracy": 0.16772198379039766,
      "num_tokens": 3836779.0,
      "step": 1670
    },
    {
      "entropy": 5.989100790023803,
      "epoch": 0.16090297790585975,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000499952540439365,
      "loss": 5.8595,
      "mean_token_accuracy": 0.1596663162112236,
      "num_tokens": 3848350.0,
      "step": 1675
    },
    {
      "entropy": 6.105287694931031,
      "epoch": 0.16138328530259366,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999518337049743,
      "loss": 5.8766,
      "mean_token_accuracy": 0.16466120183467864,
      "num_tokens": 3859792.0,
      "step": 1680
    },
    {
      "entropy": 6.038152456283569,
      "epoch": 0.16186359269932757,
      "grad_norm": 0.984375,
      "learning_rate": 0.00049995112174787,
      "loss": 5.8118,
      "mean_token_accuracy": 0.1610649883747101,
      "num_tokens": 3871539.0,
      "step": 1685
    },
    {
      "entropy": 5.94040699005127,
      "epoch": 0.16234390009606148,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999504045680687,
      "loss": 5.8553,
      "mean_token_accuracy": 0.165919728577137,
      "num_tokens": 3885025.0,
      "step": 1690
    },
    {
      "entropy": 6.108747243881226,
      "epoch": 0.1628242074927954,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999496821655869,
      "loss": 5.9671,
      "mean_token_accuracy": 0.15593098998069763,
      "num_tokens": 3896519.0,
      "step": 1695
    },
    {
      "entropy": 6.010381031036377,
      "epoch": 0.1633045148895293,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999489545404414,
      "loss": 5.7259,
      "mean_token_accuracy": 0.16787517666816712,
      "num_tokens": 3907636.0,
      "step": 1700
    },
    {
      "entropy": 5.983977651596069,
      "epoch": 0.1637848222862632,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999482216926493,
      "loss": 5.8537,
      "mean_token_accuracy": 0.16455249339342118,
      "num_tokens": 3917746.0,
      "step": 1705
    },
    {
      "entropy": 6.00126838684082,
      "epoch": 0.1642651296829971,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999474836222273,
      "loss": 5.7735,
      "mean_token_accuracy": 0.16910521686077118,
      "num_tokens": 3928901.0,
      "step": 1710
    },
    {
      "entropy": 5.979851198196411,
      "epoch": 0.16474543707973102,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999467403291928,
      "loss": 5.8926,
      "mean_token_accuracy": 0.16778817474842073,
      "num_tokens": 3940094.0,
      "step": 1715
    },
    {
      "entropy": 5.993012142181397,
      "epoch": 0.16522574447646493,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999459918135628,
      "loss": 5.8116,
      "mean_token_accuracy": 0.1678248405456543,
      "num_tokens": 3952714.0,
      "step": 1720
    },
    {
      "entropy": 5.990395832061767,
      "epoch": 0.16570605187319884,
      "grad_norm": 1.09375,
      "learning_rate": 0.000499945238075355,
      "loss": 5.8016,
      "mean_token_accuracy": 0.16812108904123307,
      "num_tokens": 3963107.0,
      "step": 1725
    },
    {
      "entropy": 5.992589044570923,
      "epoch": 0.16618635926993275,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999444791145865,
      "loss": 5.7408,
      "mean_token_accuracy": 0.1692016214132309,
      "num_tokens": 3973454.0,
      "step": 1730
    },
    {
      "entropy": 5.983733463287353,
      "epoch": 0.16666666666666666,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999437149312754,
      "loss": 5.8393,
      "mean_token_accuracy": 0.16407619565725326,
      "num_tokens": 3985352.0,
      "step": 1735
    },
    {
      "entropy": 6.094290447235108,
      "epoch": 0.16714697406340057,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499942945525439,
      "loss": 5.8161,
      "mean_token_accuracy": 0.1678325891494751,
      "num_tokens": 3997088.0,
      "step": 1740
    },
    {
      "entropy": 5.9968156814575195,
      "epoch": 0.16762728146013448,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999421708970954,
      "loss": 5.8345,
      "mean_token_accuracy": 0.1668623298406601,
      "num_tokens": 4008551.0,
      "step": 1745
    },
    {
      "entropy": 5.992422819137573,
      "epoch": 0.16810758885686838,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999413910462625,
      "loss": 5.8707,
      "mean_token_accuracy": 0.1622787207365036,
      "num_tokens": 4020886.0,
      "step": 1750
    },
    {
      "entropy": 6.026627683639527,
      "epoch": 0.1685878962536023,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999406059729586,
      "loss": 5.8259,
      "mean_token_accuracy": 0.15881873518228531,
      "num_tokens": 4032811.0,
      "step": 1755
    },
    {
      "entropy": 6.032187414169312,
      "epoch": 0.1690682036503362,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999398156772016,
      "loss": 5.8069,
      "mean_token_accuracy": 0.1716371476650238,
      "num_tokens": 4044102.0,
      "step": 1760
    },
    {
      "entropy": 5.965451192855835,
      "epoch": 0.16954851104707014,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00049993902015901,
      "loss": 5.8225,
      "mean_token_accuracy": 0.16970249265432358,
      "num_tokens": 4056849.0,
      "step": 1765
    },
    {
      "entropy": 5.969043207168579,
      "epoch": 0.17002881844380405,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999382194184023,
      "loss": 5.8634,
      "mean_token_accuracy": 0.16509611159563065,
      "num_tokens": 4068392.0,
      "step": 1770
    },
    {
      "entropy": 6.03976616859436,
      "epoch": 0.17050912584053796,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999374134553972,
      "loss": 5.6964,
      "mean_token_accuracy": 0.16902782618999482,
      "num_tokens": 4078949.0,
      "step": 1775
    },
    {
      "entropy": 5.9731302738189695,
      "epoch": 0.17098943323727187,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999366022700131,
      "loss": 5.8517,
      "mean_token_accuracy": 0.15924588441848755,
      "num_tokens": 4091432.0,
      "step": 1780
    },
    {
      "entropy": 5.927183818817139,
      "epoch": 0.17146974063400577,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999357858622691,
      "loss": 5.7154,
      "mean_token_accuracy": 0.17314967513084412,
      "num_tokens": 4103001.0,
      "step": 1785
    },
    {
      "entropy": 6.011290836334228,
      "epoch": 0.17195004803073968,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999349642321842,
      "loss": 5.8191,
      "mean_token_accuracy": 0.17158177942037584,
      "num_tokens": 4113570.0,
      "step": 1790
    },
    {
      "entropy": 5.992266416549683,
      "epoch": 0.1724303554274736,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999341373797772,
      "loss": 5.8541,
      "mean_token_accuracy": 0.16368370950222016,
      "num_tokens": 4125658.0,
      "step": 1795
    },
    {
      "entropy": 5.947376489639282,
      "epoch": 0.1729106628242075,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999333053050675,
      "loss": 5.7418,
      "mean_token_accuracy": 0.17171041816473007,
      "num_tokens": 4137424.0,
      "step": 1800
    },
    {
      "entropy": 5.968177795410156,
      "epoch": 0.1733909702209414,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999324680080744,
      "loss": 5.7765,
      "mean_token_accuracy": 0.16682788729667664,
      "num_tokens": 4149772.0,
      "step": 1805
    },
    {
      "entropy": 5.920679140090942,
      "epoch": 0.17387127761767532,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999316254888172,
      "loss": 5.7741,
      "mean_token_accuracy": 0.1707597255706787,
      "num_tokens": 4162246.0,
      "step": 1810
    },
    {
      "entropy": 5.959510278701782,
      "epoch": 0.17435158501440923,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999307777473157,
      "loss": 5.7535,
      "mean_token_accuracy": 0.1732211798429489,
      "num_tokens": 4173449.0,
      "step": 1815
    },
    {
      "entropy": 5.920207357406616,
      "epoch": 0.17483189241114314,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999299247835893,
      "loss": 5.7832,
      "mean_token_accuracy": 0.16665025502443315,
      "num_tokens": 4185560.0,
      "step": 1820
    },
    {
      "entropy": 5.9777604103088375,
      "epoch": 0.17531219980787704,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000499929066597658,
      "loss": 5.8396,
      "mean_token_accuracy": 0.15737783163785934,
      "num_tokens": 4197843.0,
      "step": 1825
    },
    {
      "entropy": 6.042659330368042,
      "epoch": 0.17579250720461095,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999282031895418,
      "loss": 5.697,
      "mean_token_accuracy": 0.16807238161563873,
      "num_tokens": 4210009.0,
      "step": 1830
    },
    {
      "entropy": 5.951578092575073,
      "epoch": 0.17627281460134486,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999273345592604,
      "loss": 5.9491,
      "mean_token_accuracy": 0.1559513933956623,
      "num_tokens": 4222213.0,
      "step": 1835
    },
    {
      "entropy": 6.02875337600708,
      "epoch": 0.17675312199807877,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999264607068343,
      "loss": 5.7886,
      "mean_token_accuracy": 0.1635829284787178,
      "num_tokens": 4234159.0,
      "step": 1840
    },
    {
      "entropy": 5.958674907684326,
      "epoch": 0.17723342939481268,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999255816322837,
      "loss": 5.8528,
      "mean_token_accuracy": 0.15672947987914085,
      "num_tokens": 4246358.0,
      "step": 1845
    },
    {
      "entropy": 6.076961183547974,
      "epoch": 0.1777137367915466,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499924697335629,
      "loss": 5.9279,
      "mean_token_accuracy": 0.16347040683031083,
      "num_tokens": 4256698.0,
      "step": 1850
    },
    {
      "entropy": 5.922059392929077,
      "epoch": 0.1781940441882805,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999238078168906,
      "loss": 5.7106,
      "mean_token_accuracy": 0.17682354152202606,
      "num_tokens": 4268007.0,
      "step": 1855
    },
    {
      "entropy": 6.002375745773316,
      "epoch": 0.1786743515850144,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999229130760894,
      "loss": 5.8196,
      "mean_token_accuracy": 0.16765979379415513,
      "num_tokens": 4278583.0,
      "step": 1860
    },
    {
      "entropy": 6.071978282928467,
      "epoch": 0.17915465898174832,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499922013113246,
      "loss": 5.8557,
      "mean_token_accuracy": 0.15993767529726027,
      "num_tokens": 4290267.0,
      "step": 1865
    },
    {
      "entropy": 5.926740837097168,
      "epoch": 0.17963496637848222,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999211079283814,
      "loss": 5.7515,
      "mean_token_accuracy": 0.17423082739114762,
      "num_tokens": 4302177.0,
      "step": 1870
    },
    {
      "entropy": 5.883774328231811,
      "epoch": 0.18011527377521613,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999201975215164,
      "loss": 5.7041,
      "mean_token_accuracy": 0.17202669382095337,
      "num_tokens": 4312420.0,
      "step": 1875
    },
    {
      "entropy": 6.049909019470215,
      "epoch": 0.18059558117195004,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999192818926725,
      "loss": 5.8324,
      "mean_token_accuracy": 0.16264985203742982,
      "num_tokens": 4324050.0,
      "step": 1880
    },
    {
      "entropy": 5.910491991043091,
      "epoch": 0.18107588856868395,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999183610418706,
      "loss": 5.6986,
      "mean_token_accuracy": 0.1735323429107666,
      "num_tokens": 4335336.0,
      "step": 1885
    },
    {
      "entropy": 5.850298452377319,
      "epoch": 0.18155619596541786,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999174349691322,
      "loss": 5.7101,
      "mean_token_accuracy": 0.1698858320713043,
      "num_tokens": 4346239.0,
      "step": 1890
    },
    {
      "entropy": 5.930282306671143,
      "epoch": 0.18203650336215177,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999165036744788,
      "loss": 5.7899,
      "mean_token_accuracy": 0.1672823816537857,
      "num_tokens": 4357195.0,
      "step": 1895
    },
    {
      "entropy": 5.934293127059936,
      "epoch": 0.18251681075888568,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999155671579322,
      "loss": 5.713,
      "mean_token_accuracy": 0.17279640585184097,
      "num_tokens": 4368905.0,
      "step": 1900
    },
    {
      "entropy": 5.985248327255249,
      "epoch": 0.1829971181556196,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499914625419514,
      "loss": 5.7924,
      "mean_token_accuracy": 0.16554614007472992,
      "num_tokens": 4381708.0,
      "step": 1905
    },
    {
      "entropy": 5.808998727798462,
      "epoch": 0.1834774255523535,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999136784592459,
      "loss": 5.6416,
      "mean_token_accuracy": 0.18022481054067613,
      "num_tokens": 4393199.0,
      "step": 1910
    },
    {
      "entropy": 5.969259786605835,
      "epoch": 0.1839577329490874,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999127262771502,
      "loss": 5.8635,
      "mean_token_accuracy": 0.1600896716117859,
      "num_tokens": 4403986.0,
      "step": 1915
    },
    {
      "entropy": 5.962497043609619,
      "epoch": 0.1844380403458213,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999117688732487,
      "loss": 5.7064,
      "mean_token_accuracy": 0.1686962679028511,
      "num_tokens": 4415574.0,
      "step": 1920
    },
    {
      "entropy": 5.917124795913696,
      "epoch": 0.18491834774255522,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999108062475638,
      "loss": 5.7456,
      "mean_token_accuracy": 0.1732297196984291,
      "num_tokens": 4427579.0,
      "step": 1925
    },
    {
      "entropy": 5.935105228424073,
      "epoch": 0.18539865513928913,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000499909838400118,
      "loss": 5.7297,
      "mean_token_accuracy": 0.1685871258378029,
      "num_tokens": 4440045.0,
      "step": 1930
    },
    {
      "entropy": 5.947299671173096,
      "epoch": 0.18587896253602307,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999088653309334,
      "loss": 5.8274,
      "mean_token_accuracy": 0.16587052941322328,
      "num_tokens": 4450129.0,
      "step": 1935
    },
    {
      "entropy": 5.921767616271973,
      "epoch": 0.18635926993275698,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999078870400329,
      "loss": 5.6916,
      "mean_token_accuracy": 0.17326790392398833,
      "num_tokens": 4461378.0,
      "step": 1940
    },
    {
      "entropy": 5.938140058517456,
      "epoch": 0.18683957732949089,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999069035274391,
      "loss": 5.7165,
      "mean_token_accuracy": 0.1744956612586975,
      "num_tokens": 4472249.0,
      "step": 1945
    },
    {
      "entropy": 5.883719635009766,
      "epoch": 0.1873198847262248,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999059147931747,
      "loss": 5.7713,
      "mean_token_accuracy": 0.16725743114948272,
      "num_tokens": 4484243.0,
      "step": 1950
    },
    {
      "entropy": 5.955920314788818,
      "epoch": 0.1878001921229587,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999049208372629,
      "loss": 5.6979,
      "mean_token_accuracy": 0.1676156923174858,
      "num_tokens": 4495776.0,
      "step": 1955
    },
    {
      "entropy": 5.874173736572265,
      "epoch": 0.1882804995196926,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999039216597267,
      "loss": 5.7708,
      "mean_token_accuracy": 0.17022369354963302,
      "num_tokens": 4508581.0,
      "step": 1960
    },
    {
      "entropy": 5.967559289932251,
      "epoch": 0.18876080691642652,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999029172605892,
      "loss": 5.7088,
      "mean_token_accuracy": 0.16853721737861632,
      "num_tokens": 4519753.0,
      "step": 1965
    },
    {
      "entropy": 5.8548189163208,
      "epoch": 0.18924111431316043,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999019076398738,
      "loss": 5.7709,
      "mean_token_accuracy": 0.17433851957321167,
      "num_tokens": 4529957.0,
      "step": 1970
    },
    {
      "entropy": 5.984222507476806,
      "epoch": 0.18972142170989434,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499900892797604,
      "loss": 5.7479,
      "mean_token_accuracy": 0.16784974783658982,
      "num_tokens": 4541181.0,
      "step": 1975
    },
    {
      "entropy": 5.887285852432251,
      "epoch": 0.19020172910662825,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998998727338031,
      "loss": 5.7067,
      "mean_token_accuracy": 0.17156032621860504,
      "num_tokens": 4552235.0,
      "step": 1980
    },
    {
      "entropy": 5.951331043243409,
      "epoch": 0.19068203650336216,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998988474484952,
      "loss": 5.7343,
      "mean_token_accuracy": 0.17553815692663194,
      "num_tokens": 4564045.0,
      "step": 1985
    },
    {
      "entropy": 5.845712804794312,
      "epoch": 0.19116234390009607,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998978169417038,
      "loss": 5.6063,
      "mean_token_accuracy": 0.1784340426325798,
      "num_tokens": 4575141.0,
      "step": 1990
    },
    {
      "entropy": 5.845089387893677,
      "epoch": 0.19164265129682997,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998967812134529,
      "loss": 5.7324,
      "mean_token_accuracy": 0.1685302123427391,
      "num_tokens": 4586491.0,
      "step": 1995
    },
    {
      "entropy": 5.947291421890259,
      "epoch": 0.19212295869356388,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998957402637664,
      "loss": 5.722,
      "mean_token_accuracy": 0.17145114988088608,
      "num_tokens": 4598421.0,
      "step": 2000
    },
    {
      "entropy": 5.85858998298645,
      "epoch": 0.1926032660902978,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998946940926687,
      "loss": 5.6278,
      "mean_token_accuracy": 0.18301957547664643,
      "num_tokens": 4609688.0,
      "step": 2005
    },
    {
      "entropy": 5.852959537506104,
      "epoch": 0.1930835734870317,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499893642700184,
      "loss": 5.7817,
      "mean_token_accuracy": 0.16823463290929794,
      "num_tokens": 4621169.0,
      "step": 2010
    },
    {
      "entropy": 5.936551713943482,
      "epoch": 0.1935638808837656,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004998925860863368,
      "loss": 5.7287,
      "mean_token_accuracy": 0.1727944165468216,
      "num_tokens": 4632512.0,
      "step": 2015
    },
    {
      "entropy": 5.921540403366089,
      "epoch": 0.19404418828049952,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004998915242511516,
      "loss": 5.8023,
      "mean_token_accuracy": 0.16758645325899124,
      "num_tokens": 4644356.0,
      "step": 2020
    },
    {
      "entropy": 5.950337505340576,
      "epoch": 0.19452449567723343,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998904571946528,
      "loss": 5.7704,
      "mean_token_accuracy": 0.16971734017133713,
      "num_tokens": 4655601.0,
      "step": 2025
    },
    {
      "entropy": 5.899324655532837,
      "epoch": 0.19500480307396734,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004998893849168655,
      "loss": 5.7704,
      "mean_token_accuracy": 0.16545508950948715,
      "num_tokens": 4666813.0,
      "step": 2030
    },
    {
      "entropy": 5.944556856155396,
      "epoch": 0.19548511047070125,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998883074178144,
      "loss": 5.764,
      "mean_token_accuracy": 0.16672957092523574,
      "num_tokens": 4679123.0,
      "step": 2035
    },
    {
      "entropy": 5.964180326461792,
      "epoch": 0.19596541786743515,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998872246975247,
      "loss": 5.658,
      "mean_token_accuracy": 0.175301992893219,
      "num_tokens": 4690485.0,
      "step": 2040
    },
    {
      "entropy": 5.712597417831421,
      "epoch": 0.19644572526416906,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998861367560213,
      "loss": 5.685,
      "mean_token_accuracy": 0.17356650531291962,
      "num_tokens": 4702573.0,
      "step": 2045
    },
    {
      "entropy": 6.018131875991822,
      "epoch": 0.19692603266090297,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998850435933296,
      "loss": 5.7212,
      "mean_token_accuracy": 0.16638207733631133,
      "num_tokens": 4715653.0,
      "step": 2050
    },
    {
      "entropy": 5.8512365341186525,
      "epoch": 0.19740634005763688,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998839452094749,
      "loss": 5.7248,
      "mean_token_accuracy": 0.17304063439369202,
      "num_tokens": 4726051.0,
      "step": 2055
    },
    {
      "entropy": 5.898526000976562,
      "epoch": 0.1978866474543708,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998828416044829,
      "loss": 5.8526,
      "mean_token_accuracy": 0.16188929975032806,
      "num_tokens": 4738176.0,
      "step": 2060
    },
    {
      "entropy": 5.963538885116577,
      "epoch": 0.1983669548511047,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499881732778379,
      "loss": 5.7637,
      "mean_token_accuracy": 0.16928643435239793,
      "num_tokens": 4749547.0,
      "step": 2065
    },
    {
      "entropy": 5.823513078689575,
      "epoch": 0.1988472622478386,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499880618731189,
      "loss": 5.6765,
      "mean_token_accuracy": 0.17271803468465804,
      "num_tokens": 4761967.0,
      "step": 2070
    },
    {
      "entropy": 5.874024534225464,
      "epoch": 0.19932756964457252,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998794994629388,
      "loss": 5.6645,
      "mean_token_accuracy": 0.17876208424568177,
      "num_tokens": 4772847.0,
      "step": 2075
    },
    {
      "entropy": 5.877955722808838,
      "epoch": 0.19980787704130643,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998783749736545,
      "loss": 5.7292,
      "mean_token_accuracy": 0.16857509166002274,
      "num_tokens": 4785049.0,
      "step": 2080
    },
    {
      "entropy": 5.901623296737671,
      "epoch": 0.20028818443804033,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998772452633619,
      "loss": 5.7387,
      "mean_token_accuracy": 0.16586733162403106,
      "num_tokens": 4797484.0,
      "step": 2085
    },
    {
      "entropy": 5.869886159896851,
      "epoch": 0.20076849183477424,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998761103320876,
      "loss": 5.6839,
      "mean_token_accuracy": 0.16756847649812698,
      "num_tokens": 4809422.0,
      "step": 2090
    },
    {
      "entropy": 5.904060888290405,
      "epoch": 0.20124879923150815,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998749701798577,
      "loss": 5.8555,
      "mean_token_accuracy": 0.16425696462392808,
      "num_tokens": 4822340.0,
      "step": 2095
    },
    {
      "entropy": 5.91972541809082,
      "epoch": 0.2017291066282421,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998738248066986,
      "loss": 5.6478,
      "mean_token_accuracy": 0.1721834734082222,
      "num_tokens": 4833382.0,
      "step": 2100
    },
    {
      "entropy": 5.785834503173828,
      "epoch": 0.202209414024976,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998726742126372,
      "loss": 5.6559,
      "mean_token_accuracy": 0.17670872658491135,
      "num_tokens": 4843916.0,
      "step": 2105
    },
    {
      "entropy": 5.9129715919494625,
      "epoch": 0.2026897214217099,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998715183976999,
      "loss": 5.6465,
      "mean_token_accuracy": 0.175136299431324,
      "num_tokens": 4855420.0,
      "step": 2110
    },
    {
      "entropy": 5.899812126159668,
      "epoch": 0.20317002881844382,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998703573619137,
      "loss": 5.7348,
      "mean_token_accuracy": 0.16458186358213425,
      "num_tokens": 4866540.0,
      "step": 2115
    },
    {
      "entropy": 5.9016939163208,
      "epoch": 0.20365033621517772,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004998691911053056,
      "loss": 5.6732,
      "mean_token_accuracy": 0.17268897444009781,
      "num_tokens": 4878860.0,
      "step": 2120
    },
    {
      "entropy": 5.974357414245605,
      "epoch": 0.20413064361191163,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998680196279026,
      "loss": 5.7979,
      "mean_token_accuracy": 0.16729146391153335,
      "num_tokens": 4889692.0,
      "step": 2125
    },
    {
      "entropy": 5.968846464157105,
      "epoch": 0.20461095100864554,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998668429297319,
      "loss": 5.7415,
      "mean_token_accuracy": 0.1737045720219612,
      "num_tokens": 4901337.0,
      "step": 2130
    },
    {
      "entropy": 5.777680540084839,
      "epoch": 0.20509125840537945,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998656610108208,
      "loss": 5.6488,
      "mean_token_accuracy": 0.1813000425696373,
      "num_tokens": 4912340.0,
      "step": 2135
    },
    {
      "entropy": 5.86177020072937,
      "epoch": 0.20557156580211336,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998644738711969,
      "loss": 5.6257,
      "mean_token_accuracy": 0.1789335146546364,
      "num_tokens": 4923429.0,
      "step": 2140
    },
    {
      "entropy": 5.904550504684448,
      "epoch": 0.20605187319884727,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004998632815108874,
      "loss": 5.6893,
      "mean_token_accuracy": 0.1724928915500641,
      "num_tokens": 4934826.0,
      "step": 2145
    },
    {
      "entropy": 5.867998170852661,
      "epoch": 0.20653218059558118,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998620839299203,
      "loss": 5.766,
      "mean_token_accuracy": 0.17130101472139359,
      "num_tokens": 4945189.0,
      "step": 2150
    },
    {
      "entropy": 5.955338764190674,
      "epoch": 0.2070124879923151,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998608811283233,
      "loss": 5.6609,
      "mean_token_accuracy": 0.17488691955804825,
      "num_tokens": 4956060.0,
      "step": 2155
    },
    {
      "entropy": 5.850173854827881,
      "epoch": 0.207492795389049,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998596731061244,
      "loss": 5.6778,
      "mean_token_accuracy": 0.1728389099240303,
      "num_tokens": 4967710.0,
      "step": 2160
    },
    {
      "entropy": 5.8504749774932865,
      "epoch": 0.2079731027857829,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998584598633516,
      "loss": 5.6944,
      "mean_token_accuracy": 0.17247060984373092,
      "num_tokens": 4978716.0,
      "step": 2165
    },
    {
      "entropy": 5.9108704090118405,
      "epoch": 0.2084534101825168,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998572414000329,
      "loss": 5.7613,
      "mean_token_accuracy": 0.16938658952713012,
      "num_tokens": 4990742.0,
      "step": 2170
    },
    {
      "entropy": 5.872229480743409,
      "epoch": 0.20893371757925072,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998560177161969,
      "loss": 5.7111,
      "mean_token_accuracy": 0.170537830889225,
      "num_tokens": 5001217.0,
      "step": 2175
    },
    {
      "entropy": 5.884965562820435,
      "epoch": 0.20941402497598463,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004998547888118718,
      "loss": 5.6905,
      "mean_token_accuracy": 0.17088903486728668,
      "num_tokens": 5012750.0,
      "step": 2180
    },
    {
      "entropy": 5.863210535049438,
      "epoch": 0.20989433237271854,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998535546870862,
      "loss": 5.7103,
      "mean_token_accuracy": 0.1782528981566429,
      "num_tokens": 5023818.0,
      "step": 2185
    },
    {
      "entropy": 5.938375854492188,
      "epoch": 0.21037463976945245,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998523153418687,
      "loss": 5.7242,
      "mean_token_accuracy": 0.1727166622877121,
      "num_tokens": 5035060.0,
      "step": 2190
    },
    {
      "entropy": 5.894730424880981,
      "epoch": 0.21085494716618636,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998510707762481,
      "loss": 5.7598,
      "mean_token_accuracy": 0.16584042757749556,
      "num_tokens": 5046780.0,
      "step": 2195
    },
    {
      "entropy": 5.820294713973999,
      "epoch": 0.21133525456292027,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998498209902533,
      "loss": 5.7333,
      "mean_token_accuracy": 0.16727135479450225,
      "num_tokens": 5059152.0,
      "step": 2200
    },
    {
      "entropy": 5.911803770065307,
      "epoch": 0.21181556195965417,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998485659839134,
      "loss": 5.6956,
      "mean_token_accuracy": 0.17464353293180465,
      "num_tokens": 5069971.0,
      "step": 2205
    },
    {
      "entropy": 5.827769708633423,
      "epoch": 0.21229586935638808,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998473057572575,
      "loss": 5.644,
      "mean_token_accuracy": 0.17635224610567093,
      "num_tokens": 5080640.0,
      "step": 2210
    },
    {
      "entropy": 5.841212940216065,
      "epoch": 0.212776176753122,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998460403103146,
      "loss": 5.6737,
      "mean_token_accuracy": 0.1733733147382736,
      "num_tokens": 5091240.0,
      "step": 2215
    },
    {
      "entropy": 5.77046570777893,
      "epoch": 0.2132564841498559,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998447696431146,
      "loss": 5.6845,
      "mean_token_accuracy": 0.16993408799171447,
      "num_tokens": 5102920.0,
      "step": 2220
    },
    {
      "entropy": 5.9066606044769285,
      "epoch": 0.2137367915465898,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998434937556865,
      "loss": 5.7005,
      "mean_token_accuracy": 0.1688303142786026,
      "num_tokens": 5114240.0,
      "step": 2225
    },
    {
      "entropy": 5.993351554870605,
      "epoch": 0.21421709894332372,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998422126480602,
      "loss": 5.9259,
      "mean_token_accuracy": 0.15784672945737838,
      "num_tokens": 5126653.0,
      "step": 2230
    },
    {
      "entropy": 5.971493339538574,
      "epoch": 0.21469740634005763,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998409263202653,
      "loss": 5.7667,
      "mean_token_accuracy": 0.1656375914812088,
      "num_tokens": 5138778.0,
      "step": 2235
    },
    {
      "entropy": 5.849447393417359,
      "epoch": 0.21517771373679154,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998396347723318,
      "loss": 5.624,
      "mean_token_accuracy": 0.17630810886621476,
      "num_tokens": 5148670.0,
      "step": 2240
    },
    {
      "entropy": 5.685166835784912,
      "epoch": 0.21565802113352545,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998383380042895,
      "loss": 5.6681,
      "mean_token_accuracy": 0.17174769341945648,
      "num_tokens": 5159617.0,
      "step": 2245
    },
    {
      "entropy": 5.921658134460449,
      "epoch": 0.21613832853025935,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998370360161688,
      "loss": 5.7246,
      "mean_token_accuracy": 0.17249593138694763,
      "num_tokens": 5171225.0,
      "step": 2250
    },
    {
      "entropy": 5.881531429290772,
      "epoch": 0.21661863592699326,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998357288079996,
      "loss": 5.7225,
      "mean_token_accuracy": 0.17121425718069078,
      "num_tokens": 5184586.0,
      "step": 2255
    },
    {
      "entropy": 5.809099578857422,
      "epoch": 0.21709894332372717,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998344163798125,
      "loss": 5.6839,
      "mean_token_accuracy": 0.16990743279457093,
      "num_tokens": 5196089.0,
      "step": 2260
    },
    {
      "entropy": 5.88064432144165,
      "epoch": 0.21757925072046108,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004998330987316379,
      "loss": 5.7237,
      "mean_token_accuracy": 0.17189674377441405,
      "num_tokens": 5207581.0,
      "step": 2265
    },
    {
      "entropy": 5.819460391998291,
      "epoch": 0.21805955811719502,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998317758635062,
      "loss": 5.6785,
      "mean_token_accuracy": 0.1701069101691246,
      "num_tokens": 5219108.0,
      "step": 2270
    },
    {
      "entropy": 5.844608736038208,
      "epoch": 0.21853986551392893,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998304477754484,
      "loss": 5.6322,
      "mean_token_accuracy": 0.17310244739055633,
      "num_tokens": 5231137.0,
      "step": 2275
    },
    {
      "entropy": 5.944477605819702,
      "epoch": 0.21902017291066284,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998291144674952,
      "loss": 5.7005,
      "mean_token_accuracy": 0.1749888077378273,
      "num_tokens": 5243528.0,
      "step": 2280
    },
    {
      "entropy": 5.702851104736328,
      "epoch": 0.21950048030739674,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998277759396776,
      "loss": 5.6449,
      "mean_token_accuracy": 0.1756882846355438,
      "num_tokens": 5254869.0,
      "step": 2285
    },
    {
      "entropy": 5.825752067565918,
      "epoch": 0.21998078770413065,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998264321920265,
      "loss": 5.6384,
      "mean_token_accuracy": 0.1748126581311226,
      "num_tokens": 5265370.0,
      "step": 2290
    },
    {
      "entropy": 5.841421604156494,
      "epoch": 0.22046109510086456,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998250832245734,
      "loss": 5.5943,
      "mean_token_accuracy": 0.1766410768032074,
      "num_tokens": 5277812.0,
      "step": 2295
    },
    {
      "entropy": 5.787303256988525,
      "epoch": 0.22094140249759847,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998237290373494,
      "loss": 5.5874,
      "mean_token_accuracy": 0.17861277461051941,
      "num_tokens": 5289107.0,
      "step": 2300
    },
    {
      "entropy": 5.789830303192138,
      "epoch": 0.22142170989433238,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499822369630386,
      "loss": 5.6356,
      "mean_token_accuracy": 0.16736368983983993,
      "num_tokens": 5301489.0,
      "step": 2305
    },
    {
      "entropy": 5.7769101619720455,
      "epoch": 0.2219020172910663,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998210050037148,
      "loss": 5.618,
      "mean_token_accuracy": 0.1812180057168007,
      "num_tokens": 5313273.0,
      "step": 2310
    },
    {
      "entropy": 5.85909104347229,
      "epoch": 0.2223823246878002,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998196351573674,
      "loss": 5.6456,
      "mean_token_accuracy": 0.17510980367660522,
      "num_tokens": 5324227.0,
      "step": 2315
    },
    {
      "entropy": 5.776947546005249,
      "epoch": 0.2228626320845341,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998182600913757,
      "loss": 5.5856,
      "mean_token_accuracy": 0.18297051787376403,
      "num_tokens": 5335138.0,
      "step": 2320
    },
    {
      "entropy": 5.796558809280396,
      "epoch": 0.22334293948126802,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998168798057715,
      "loss": 5.6683,
      "mean_token_accuracy": 0.17540759295225145,
      "num_tokens": 5345652.0,
      "step": 2325
    },
    {
      "entropy": 5.9193799018859865,
      "epoch": 0.22382324687800192,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000499815494300587,
      "loss": 5.7596,
      "mean_token_accuracy": 0.17200602740049362,
      "num_tokens": 5357999.0,
      "step": 2330
    },
    {
      "entropy": 5.852154779434204,
      "epoch": 0.22430355427473583,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998141035758542,
      "loss": 5.669,
      "mean_token_accuracy": 0.17199552208185195,
      "num_tokens": 5369032.0,
      "step": 2335
    },
    {
      "entropy": 5.799731969833374,
      "epoch": 0.22478386167146974,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998127076316054,
      "loss": 5.5855,
      "mean_token_accuracy": 0.17641636580228806,
      "num_tokens": 5380692.0,
      "step": 2340
    },
    {
      "entropy": 5.864359760284424,
      "epoch": 0.22526416906820365,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998113064678734,
      "loss": 5.7228,
      "mean_token_accuracy": 0.16467591673135756,
      "num_tokens": 5392320.0,
      "step": 2345
    },
    {
      "entropy": 5.877440690994263,
      "epoch": 0.22574447646493756,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004998099000846901,
      "loss": 5.7007,
      "mean_token_accuracy": 0.1754738360643387,
      "num_tokens": 5404374.0,
      "step": 2350
    },
    {
      "entropy": 5.855792570114136,
      "epoch": 0.22622478386167147,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998084884820887,
      "loss": 5.6965,
      "mean_token_accuracy": 0.17115845978260041,
      "num_tokens": 5415471.0,
      "step": 2355
    },
    {
      "entropy": 5.710199642181396,
      "epoch": 0.22670509125840538,
      "grad_norm": 1.125,
      "learning_rate": 0.0004998070716601016,
      "loss": 5.5231,
      "mean_token_accuracy": 0.17625806480646133,
      "num_tokens": 5425971.0,
      "step": 2360
    },
    {
      "entropy": 5.852556991577148,
      "epoch": 0.2271853986551393,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998056496187618,
      "loss": 5.6506,
      "mean_token_accuracy": 0.17488612681627275,
      "num_tokens": 5437638.0,
      "step": 2365
    },
    {
      "entropy": 5.849278402328491,
      "epoch": 0.2276657060518732,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998042223581025,
      "loss": 5.6783,
      "mean_token_accuracy": 0.1768277943134308,
      "num_tokens": 5449463.0,
      "step": 2370
    },
    {
      "entropy": 5.841348314285279,
      "epoch": 0.2281460134486071,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998027898781565,
      "loss": 5.7432,
      "mean_token_accuracy": 0.1705165147781372,
      "num_tokens": 5460972.0,
      "step": 2375
    },
    {
      "entropy": 5.983591842651367,
      "epoch": 0.228626320845341,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998013521789574,
      "loss": 5.7514,
      "mean_token_accuracy": 0.16724159568548203,
      "num_tokens": 5471575.0,
      "step": 2380
    },
    {
      "entropy": 5.754042243957519,
      "epoch": 0.22910662824207492,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997999092605384,
      "loss": 5.6009,
      "mean_token_accuracy": 0.178335303068161,
      "num_tokens": 5483041.0,
      "step": 2385
    },
    {
      "entropy": 5.795910120010376,
      "epoch": 0.22958693563880883,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499798461122933,
      "loss": 5.6288,
      "mean_token_accuracy": 0.1755184829235077,
      "num_tokens": 5494040.0,
      "step": 2390
    },
    {
      "entropy": 5.774956464767456,
      "epoch": 0.23006724303554274,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997970077661748,
      "loss": 5.5311,
      "mean_token_accuracy": 0.18456185460090638,
      "num_tokens": 5504217.0,
      "step": 2395
    },
    {
      "entropy": 5.877355241775513,
      "epoch": 0.23054755043227665,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997955491902977,
      "loss": 5.6954,
      "mean_token_accuracy": 0.1742586001753807,
      "num_tokens": 5517180.0,
      "step": 2400
    },
    {
      "entropy": 5.792226076126099,
      "epoch": 0.23102785782901056,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997940853953354,
      "loss": 5.6086,
      "mean_token_accuracy": 0.17229509651660918,
      "num_tokens": 5529626.0,
      "step": 2405
    },
    {
      "entropy": 5.843723583221435,
      "epoch": 0.23150816522574447,
      "grad_norm": 0.890625,
      "learning_rate": 0.000499792616381322,
      "loss": 5.6466,
      "mean_token_accuracy": 0.18131813108921052,
      "num_tokens": 5540844.0,
      "step": 2410
    },
    {
      "entropy": 5.8429769515991214,
      "epoch": 0.23198847262247838,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997911421482914,
      "loss": 5.57,
      "mean_token_accuracy": 0.1793536514043808,
      "num_tokens": 5551732.0,
      "step": 2415
    },
    {
      "entropy": 5.815704107284546,
      "epoch": 0.23246878001921228,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499789662696278,
      "loss": 5.7051,
      "mean_token_accuracy": 0.17402584850788116,
      "num_tokens": 5562692.0,
      "step": 2420
    },
    {
      "entropy": 5.91572937965393,
      "epoch": 0.2329490874159462,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997881780253162,
      "loss": 5.7007,
      "mean_token_accuracy": 0.17055609226226806,
      "num_tokens": 5574884.0,
      "step": 2425
    },
    {
      "entropy": 5.709230899810791,
      "epoch": 0.2334293948126801,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997866881354403,
      "loss": 5.5462,
      "mean_token_accuracy": 0.18206045031547546,
      "num_tokens": 5585335.0,
      "step": 2430
    },
    {
      "entropy": 5.862868022918701,
      "epoch": 0.23390970220941404,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499785193026685,
      "loss": 5.6923,
      "mean_token_accuracy": 0.17130919694900512,
      "num_tokens": 5597621.0,
      "step": 2435
    },
    {
      "entropy": 5.702938747406006,
      "epoch": 0.23439000960614795,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997836926990851,
      "loss": 5.5569,
      "mean_token_accuracy": 0.17519160956144333,
      "num_tokens": 5609795.0,
      "step": 2440
    },
    {
      "entropy": 5.82079005241394,
      "epoch": 0.23487031700288186,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997821871526752,
      "loss": 5.6514,
      "mean_token_accuracy": 0.17482785731554032,
      "num_tokens": 5621473.0,
      "step": 2445
    },
    {
      "entropy": 5.843278980255127,
      "epoch": 0.23535062439961577,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997806763874905,
      "loss": 5.7307,
      "mean_token_accuracy": 0.174130442738533,
      "num_tokens": 5635064.0,
      "step": 2450
    },
    {
      "entropy": 5.9370584964752195,
      "epoch": 0.23583093179634967,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997791604035659,
      "loss": 5.7352,
      "mean_token_accuracy": 0.17039123475551604,
      "num_tokens": 5647382.0,
      "step": 2455
    },
    {
      "entropy": 5.830444431304931,
      "epoch": 0.23631123919308358,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997776392009366,
      "loss": 5.6191,
      "mean_token_accuracy": 0.17757430523633957,
      "num_tokens": 5658417.0,
      "step": 2460
    },
    {
      "entropy": 5.7811634063720705,
      "epoch": 0.2367915465898175,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004997761127796381,
      "loss": 5.5757,
      "mean_token_accuracy": 0.1834951251745224,
      "num_tokens": 5672190.0,
      "step": 2465
    },
    {
      "entropy": 5.717312669754028,
      "epoch": 0.2372718539865514,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004997745811397056,
      "loss": 5.5926,
      "mean_token_accuracy": 0.1851950004696846,
      "num_tokens": 5684259.0,
      "step": 2470
    },
    {
      "entropy": 5.784187078475952,
      "epoch": 0.2377521613832853,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997730442811748,
      "loss": 5.6604,
      "mean_token_accuracy": 0.17508698701858522,
      "num_tokens": 5696505.0,
      "step": 2475
    },
    {
      "entropy": 5.720936918258667,
      "epoch": 0.23823246878001922,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997715022040814,
      "loss": 5.5441,
      "mean_token_accuracy": 0.1788373351097107,
      "num_tokens": 5707638.0,
      "step": 2480
    },
    {
      "entropy": 5.753598546981811,
      "epoch": 0.23871277617675313,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000499769954908461,
      "loss": 5.5293,
      "mean_token_accuracy": 0.1850265458226204,
      "num_tokens": 5719098.0,
      "step": 2485
    },
    {
      "entropy": 5.9342193603515625,
      "epoch": 0.23919308357348704,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997684023943498,
      "loss": 5.7511,
      "mean_token_accuracy": 0.16247327476739884,
      "num_tokens": 5731345.0,
      "step": 2490
    },
    {
      "entropy": 5.649840927124023,
      "epoch": 0.23967339097022095,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997668446617837,
      "loss": 5.525,
      "mean_token_accuracy": 0.17844921350479126,
      "num_tokens": 5742604.0,
      "step": 2495
    },
    {
      "entropy": 5.754658555984497,
      "epoch": 0.24015369836695485,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997652817107989,
      "loss": 5.5469,
      "mean_token_accuracy": 0.18343215733766555,
      "num_tokens": 5754620.0,
      "step": 2500
    },
    {
      "entropy": 5.781588220596314,
      "epoch": 0.24063400576368876,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997637135414315,
      "loss": 5.6402,
      "mean_token_accuracy": 0.17647155076265336,
      "num_tokens": 5764764.0,
      "step": 2505
    },
    {
      "entropy": 5.660699033737183,
      "epoch": 0.24111431316042267,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997621401537183,
      "loss": 5.5191,
      "mean_token_accuracy": 0.18696373254060744,
      "num_tokens": 5776341.0,
      "step": 2510
    },
    {
      "entropy": 5.818323850631714,
      "epoch": 0.24159462055715658,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004997605615476955,
      "loss": 5.739,
      "mean_token_accuracy": 0.16692858338356018,
      "num_tokens": 5788604.0,
      "step": 2515
    },
    {
      "entropy": 5.825610828399658,
      "epoch": 0.2420749279538905,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997589777234,
      "loss": 5.6251,
      "mean_token_accuracy": 0.17850779592990876,
      "num_tokens": 5800194.0,
      "step": 2520
    },
    {
      "entropy": 5.751928377151489,
      "epoch": 0.2425552353506244,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004997573886808684,
      "loss": 5.6256,
      "mean_token_accuracy": 0.18556526601314544,
      "num_tokens": 5812978.0,
      "step": 2525
    },
    {
      "entropy": 5.761942911148071,
      "epoch": 0.2430355427473583,
      "grad_norm": 0.875,
      "learning_rate": 0.0004997557944201375,
      "loss": 5.535,
      "mean_token_accuracy": 0.18203383684158325,
      "num_tokens": 5824591.0,
      "step": 2530
    },
    {
      "entropy": 5.713061332702637,
      "epoch": 0.24351585014409222,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997541949412445,
      "loss": 5.6289,
      "mean_token_accuracy": 0.1750789538025856,
      "num_tokens": 5835970.0,
      "step": 2535
    },
    {
      "entropy": 5.851242256164551,
      "epoch": 0.24399615754082613,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997525902442266,
      "loss": 5.6674,
      "mean_token_accuracy": 0.17141913771629333,
      "num_tokens": 5848687.0,
      "step": 2540
    },
    {
      "entropy": 5.764578819274902,
      "epoch": 0.24447646493756003,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997509803291207,
      "loss": 5.641,
      "mean_token_accuracy": 0.1750764176249504,
      "num_tokens": 5860820.0,
      "step": 2545
    },
    {
      "entropy": 5.759086322784424,
      "epoch": 0.24495677233429394,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997493651959647,
      "loss": 5.5807,
      "mean_token_accuracy": 0.1781640961766243,
      "num_tokens": 5872172.0,
      "step": 2550
    },
    {
      "entropy": 5.821686220169068,
      "epoch": 0.24543707973102785,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997477448447955,
      "loss": 5.6243,
      "mean_token_accuracy": 0.17710455805063247,
      "num_tokens": 5883574.0,
      "step": 2555
    },
    {
      "entropy": 5.838350677490235,
      "epoch": 0.24591738712776176,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997461192756512,
      "loss": 5.7091,
      "mean_token_accuracy": 0.17892282754182814,
      "num_tokens": 5895223.0,
      "step": 2560
    },
    {
      "entropy": 5.739510202407837,
      "epoch": 0.24639769452449567,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997444884885694,
      "loss": 5.5408,
      "mean_token_accuracy": 0.1767376109957695,
      "num_tokens": 5907448.0,
      "step": 2565
    },
    {
      "entropy": 5.812138986587525,
      "epoch": 0.24687800192122958,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997428524835879,
      "loss": 5.6334,
      "mean_token_accuracy": 0.1804532065987587,
      "num_tokens": 5918906.0,
      "step": 2570
    },
    {
      "entropy": 5.752865028381348,
      "epoch": 0.2473583093179635,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004997412112607446,
      "loss": 5.501,
      "mean_token_accuracy": 0.18547787815332412,
      "num_tokens": 5930619.0,
      "step": 2575
    },
    {
      "entropy": 5.849464845657349,
      "epoch": 0.2478386167146974,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997395648200778,
      "loss": 5.7079,
      "mean_token_accuracy": 0.17087607085704803,
      "num_tokens": 5941554.0,
      "step": 2580
    },
    {
      "entropy": 5.8415814399719235,
      "epoch": 0.2483189241114313,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004997379131616257,
      "loss": 5.6732,
      "mean_token_accuracy": 0.17775924354791642,
      "num_tokens": 5953309.0,
      "step": 2585
    },
    {
      "entropy": 5.769577217102051,
      "epoch": 0.24879923150816521,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004997362562854266,
      "loss": 5.6135,
      "mean_token_accuracy": 0.17712407559156418,
      "num_tokens": 5965286.0,
      "step": 2590
    },
    {
      "entropy": 5.778793144226074,
      "epoch": 0.24927953890489912,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997345941915187,
      "loss": 5.5918,
      "mean_token_accuracy": 0.17737313807010652,
      "num_tokens": 5976558.0,
      "step": 2595
    },
    {
      "entropy": 5.780767154693604,
      "epoch": 0.24975984630163303,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997329268799412,
      "loss": 5.6362,
      "mean_token_accuracy": 0.17200108617544174,
      "num_tokens": 5988793.0,
      "step": 2600
    },
    {
      "entropy": 5.7742997169494625,
      "epoch": 0.25024015369836694,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997312543507322,
      "loss": 5.6641,
      "mean_token_accuracy": 0.16963352560997008,
      "num_tokens": 5999825.0,
      "step": 2605
    },
    {
      "entropy": 5.83264274597168,
      "epoch": 0.2507204610951009,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997295766039309,
      "loss": 5.6299,
      "mean_token_accuracy": 0.1738866686820984,
      "num_tokens": 6010408.0,
      "step": 2610
    },
    {
      "entropy": 5.799222660064697,
      "epoch": 0.25120076849183476,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997278936395761,
      "loss": 5.6418,
      "mean_token_accuracy": 0.1749333217740059,
      "num_tokens": 6021161.0,
      "step": 2615
    },
    {
      "entropy": 5.686932611465454,
      "epoch": 0.2516810758885687,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997262054577071,
      "loss": 5.4878,
      "mean_token_accuracy": 0.18859331905841828,
      "num_tokens": 6032646.0,
      "step": 2620
    },
    {
      "entropy": 5.804806661605835,
      "epoch": 0.2521613832853026,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997245120583627,
      "loss": 5.6657,
      "mean_token_accuracy": 0.1764261692762375,
      "num_tokens": 6045019.0,
      "step": 2625
    },
    {
      "entropy": 5.651595687866211,
      "epoch": 0.2526416906820365,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004997228134415825,
      "loss": 5.5362,
      "mean_token_accuracy": 0.18478751182556152,
      "num_tokens": 6055644.0,
      "step": 2630
    },
    {
      "entropy": 5.809038543701172,
      "epoch": 0.2531219980787704,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997211096074059,
      "loss": 5.5779,
      "mean_token_accuracy": 0.18167155236005783,
      "num_tokens": 6066488.0,
      "step": 2635
    },
    {
      "entropy": 5.771279287338257,
      "epoch": 0.25360230547550433,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997194005558722,
      "loss": 5.6018,
      "mean_token_accuracy": 0.17499032467603684,
      "num_tokens": 6078300.0,
      "step": 2640
    },
    {
      "entropy": 5.731289052963257,
      "epoch": 0.2540826128722382,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997176862870216,
      "loss": 5.5867,
      "mean_token_accuracy": 0.1714366629719734,
      "num_tokens": 6089180.0,
      "step": 2645
    },
    {
      "entropy": 5.78031816482544,
      "epoch": 0.25456292026897215,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997159668008933,
      "loss": 5.6185,
      "mean_token_accuracy": 0.18340758681297303,
      "num_tokens": 6099093.0,
      "step": 2650
    },
    {
      "entropy": 5.818167781829834,
      "epoch": 0.25504322766570603,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997142420975277,
      "loss": 5.6335,
      "mean_token_accuracy": 0.178047214448452,
      "num_tokens": 6109161.0,
      "step": 2655
    },
    {
      "entropy": 5.863713979721069,
      "epoch": 0.25552353506243997,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997125121769647,
      "loss": 5.7377,
      "mean_token_accuracy": 0.16746834814548492,
      "num_tokens": 6120031.0,
      "step": 2660
    },
    {
      "entropy": 5.716607522964478,
      "epoch": 0.25600384245917385,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997107770392444,
      "loss": 5.5956,
      "mean_token_accuracy": 0.17772964537143707,
      "num_tokens": 6132398.0,
      "step": 2665
    },
    {
      "entropy": 5.752996158599854,
      "epoch": 0.2564841498559078,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499709036684407,
      "loss": 5.5746,
      "mean_token_accuracy": 0.17592063248157502,
      "num_tokens": 6143099.0,
      "step": 2670
    },
    {
      "entropy": 5.736369705200195,
      "epoch": 0.25696445725264166,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004997072911124932,
      "loss": 5.5385,
      "mean_token_accuracy": 0.18297111839056016,
      "num_tokens": 6154815.0,
      "step": 2675
    },
    {
      "entropy": 5.770055437088013,
      "epoch": 0.2574447646493756,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997055403235432,
      "loss": 5.6053,
      "mean_token_accuracy": 0.18265556544065475,
      "num_tokens": 6165696.0,
      "step": 2680
    },
    {
      "entropy": 5.7239914894104,
      "epoch": 0.2579250720461095,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997037843175978,
      "loss": 5.6237,
      "mean_token_accuracy": 0.17451792806386948,
      "num_tokens": 6177042.0,
      "step": 2685
    },
    {
      "entropy": 5.856169700622559,
      "epoch": 0.2584053794428434,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997020230946978,
      "loss": 5.6197,
      "mean_token_accuracy": 0.1756477549672127,
      "num_tokens": 6188691.0,
      "step": 2690
    },
    {
      "entropy": 5.612815475463867,
      "epoch": 0.25888568683957736,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997002566548841,
      "loss": 5.4458,
      "mean_token_accuracy": 0.184897281229496,
      "num_tokens": 6199931.0,
      "step": 2695
    },
    {
      "entropy": 5.721167039871216,
      "epoch": 0.25936599423631124,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004996984849981976,
      "loss": 5.5864,
      "mean_token_accuracy": 0.18731192648410797,
      "num_tokens": 6211493.0,
      "step": 2700
    },
    {
      "entropy": 5.792260313034058,
      "epoch": 0.2598463016330452,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996967081246794,
      "loss": 5.5167,
      "mean_token_accuracy": 0.17840367555618286,
      "num_tokens": 6221505.0,
      "step": 2705
    },
    {
      "entropy": 5.643460273742676,
      "epoch": 0.26032660902977905,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004996949260343711,
      "loss": 5.4986,
      "mean_token_accuracy": 0.18759121745824814,
      "num_tokens": 6233058.0,
      "step": 2710
    },
    {
      "entropy": 5.726798248291016,
      "epoch": 0.260806916426513,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996931387273137,
      "loss": 5.6449,
      "mean_token_accuracy": 0.17207851707935334,
      "num_tokens": 6244713.0,
      "step": 2715
    },
    {
      "entropy": 5.75096960067749,
      "epoch": 0.2612872238232469,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004996913462035487,
      "loss": 5.59,
      "mean_token_accuracy": 0.17319334894418717,
      "num_tokens": 6256624.0,
      "step": 2720
    },
    {
      "entropy": 5.7101236343383786,
      "epoch": 0.2617675312199808,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000499689548463118,
      "loss": 5.5794,
      "mean_token_accuracy": 0.17421182990074158,
      "num_tokens": 6271134.0,
      "step": 2725
    },
    {
      "entropy": 5.844261217117309,
      "epoch": 0.2622478386167147,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996877455060631,
      "loss": 5.6879,
      "mean_token_accuracy": 0.17699914574623107,
      "num_tokens": 6282082.0,
      "step": 2730
    },
    {
      "entropy": 5.764342355728149,
      "epoch": 0.2627281460134486,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004996859373324259,
      "loss": 5.6248,
      "mean_token_accuracy": 0.17453763633966446,
      "num_tokens": 6293869.0,
      "step": 2735
    },
    {
      "entropy": 5.779357290267944,
      "epoch": 0.2632084534101825,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004996841239422485,
      "loss": 5.6448,
      "mean_token_accuracy": 0.1693587228655815,
      "num_tokens": 6304858.0,
      "step": 2740
    },
    {
      "entropy": 5.76701545715332,
      "epoch": 0.26368876080691644,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004996823053355729,
      "loss": 5.5765,
      "mean_token_accuracy": 0.17856617122888566,
      "num_tokens": 6316738.0,
      "step": 2745
    },
    {
      "entropy": 5.723379611968994,
      "epoch": 0.2641690682036503,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996804815124413,
      "loss": 5.548,
      "mean_token_accuracy": 0.17553849667310714,
      "num_tokens": 6327691.0,
      "step": 2750
    },
    {
      "entropy": 5.798751974105835,
      "epoch": 0.26464937560038426,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996786524728962,
      "loss": 5.5599,
      "mean_token_accuracy": 0.176359224319458,
      "num_tokens": 6337964.0,
      "step": 2755
    },
    {
      "entropy": 5.669320964813233,
      "epoch": 0.26512968299711814,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004996768182169797,
      "loss": 5.6018,
      "mean_token_accuracy": 0.177447210252285,
      "num_tokens": 6349764.0,
      "step": 2760
    },
    {
      "entropy": 5.802992105484009,
      "epoch": 0.2656099903938521,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996749787447349,
      "loss": 5.5699,
      "mean_token_accuracy": 0.17743425965309143,
      "num_tokens": 6360538.0,
      "step": 2765
    },
    {
      "entropy": 5.745979881286621,
      "epoch": 0.26609029779058596,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499673134056204,
      "loss": 5.5983,
      "mean_token_accuracy": 0.1776156932115555,
      "num_tokens": 6372489.0,
      "step": 2770
    },
    {
      "entropy": 5.608844423294068,
      "epoch": 0.2665706051873199,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996712841514303,
      "loss": 5.4713,
      "mean_token_accuracy": 0.18549432456493378,
      "num_tokens": 6383080.0,
      "step": 2775
    },
    {
      "entropy": 5.823545598983765,
      "epoch": 0.2670509125840538,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004996694290304563,
      "loss": 5.5946,
      "mean_token_accuracy": 0.1786493942141533,
      "num_tokens": 6394033.0,
      "step": 2780
    },
    {
      "entropy": 5.703853988647461,
      "epoch": 0.2675312199807877,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004996675686933255,
      "loss": 5.597,
      "mean_token_accuracy": 0.17830613106489182,
      "num_tokens": 6404990.0,
      "step": 2785
    },
    {
      "entropy": 5.722035598754883,
      "epoch": 0.2680115273775216,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004996657031400807,
      "loss": 5.4683,
      "mean_token_accuracy": 0.186190427839756,
      "num_tokens": 6416337.0,
      "step": 2790
    },
    {
      "entropy": 5.710982465744019,
      "epoch": 0.26849183477425553,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996638323707655,
      "loss": 5.6723,
      "mean_token_accuracy": 0.1744049683213234,
      "num_tokens": 6427892.0,
      "step": 2795
    },
    {
      "entropy": 5.754525852203369,
      "epoch": 0.2689721421709894,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004996619563854232,
      "loss": 5.5776,
      "mean_token_accuracy": 0.18131734281778336,
      "num_tokens": 6440300.0,
      "step": 2800
    },
    {
      "entropy": 5.676147937774658,
      "epoch": 0.26945244956772335,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004996600751840974,
      "loss": 5.4817,
      "mean_token_accuracy": 0.18177040964365004,
      "num_tokens": 6451499.0,
      "step": 2805
    },
    {
      "entropy": 5.7406394481658936,
      "epoch": 0.26993275696445723,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004996581887668317,
      "loss": 5.5633,
      "mean_token_accuracy": 0.18365933150053024,
      "num_tokens": 6462882.0,
      "step": 2810
    },
    {
      "entropy": 5.697629499435425,
      "epoch": 0.27041306436119117,
      "grad_norm": 0.953125,
      "learning_rate": 0.00049965629713367,
      "loss": 5.5503,
      "mean_token_accuracy": 0.17984720915555955,
      "num_tokens": 6475042.0,
      "step": 2815
    },
    {
      "entropy": 5.7321325778961185,
      "epoch": 0.27089337175792505,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004996544002846561,
      "loss": 5.5598,
      "mean_token_accuracy": 0.17862652242183685,
      "num_tokens": 6486245.0,
      "step": 2820
    },
    {
      "entropy": 5.704649543762207,
      "epoch": 0.271373679154659,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996524982198343,
      "loss": 5.4526,
      "mean_token_accuracy": 0.1825141653418541,
      "num_tokens": 6497705.0,
      "step": 2825
    },
    {
      "entropy": 5.639297246932983,
      "epoch": 0.27185398655139287,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996505909392485,
      "loss": 5.4936,
      "mean_token_accuracy": 0.1846439689397812,
      "num_tokens": 6508193.0,
      "step": 2830
    },
    {
      "entropy": 5.765617036819458,
      "epoch": 0.2723342939481268,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996486784429429,
      "loss": 5.612,
      "mean_token_accuracy": 0.18255474865436555,
      "num_tokens": 6519228.0,
      "step": 2835
    },
    {
      "entropy": 5.747947263717651,
      "epoch": 0.2728146013448607,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996467607309622,
      "loss": 5.5679,
      "mean_token_accuracy": 0.17834291160106658,
      "num_tokens": 6530743.0,
      "step": 2840
    },
    {
      "entropy": 5.690560340881348,
      "epoch": 0.2732949087415946,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004996448378033507,
      "loss": 5.4551,
      "mean_token_accuracy": 0.17746061682701111,
      "num_tokens": 6542588.0,
      "step": 2845
    },
    {
      "entropy": 5.786451196670532,
      "epoch": 0.2737752161383285,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996429096601532,
      "loss": 5.6517,
      "mean_token_accuracy": 0.16629489734768868,
      "num_tokens": 6554002.0,
      "step": 2850
    },
    {
      "entropy": 5.8046431064605715,
      "epoch": 0.27425552353506244,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996409763014144,
      "loss": 5.6878,
      "mean_token_accuracy": 0.17579563558101655,
      "num_tokens": 6563358.0,
      "step": 2855
    },
    {
      "entropy": 5.70266695022583,
      "epoch": 0.2747358309317964,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004996390377271791,
      "loss": 5.597,
      "mean_token_accuracy": 0.17393501996994018,
      "num_tokens": 6574986.0,
      "step": 2860
    },
    {
      "entropy": 5.694790029525757,
      "epoch": 0.27521613832853026,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004996370939374924,
      "loss": 5.5337,
      "mean_token_accuracy": 0.17862565964460372,
      "num_tokens": 6587407.0,
      "step": 2865
    },
    {
      "entropy": 5.679986953735352,
      "epoch": 0.2756964457252642,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996351449323994,
      "loss": 5.4914,
      "mean_token_accuracy": 0.18411503434181214,
      "num_tokens": 6598483.0,
      "step": 2870
    },
    {
      "entropy": 5.684548234939575,
      "epoch": 0.2761767531219981,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004996331907119455,
      "loss": 5.5874,
      "mean_token_accuracy": 0.18107069432735443,
      "num_tokens": 6610749.0,
      "step": 2875
    },
    {
      "entropy": 5.7515064716339115,
      "epoch": 0.276657060518732,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004996312312761758,
      "loss": 5.5523,
      "mean_token_accuracy": 0.1702647790312767,
      "num_tokens": 6621762.0,
      "step": 2880
    },
    {
      "entropy": 5.740266036987305,
      "epoch": 0.2771373679154659,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499629266625136,
      "loss": 5.6195,
      "mean_token_accuracy": 0.1732449159026146,
      "num_tokens": 6632296.0,
      "step": 2885
    },
    {
      "entropy": 5.732256841659546,
      "epoch": 0.27761767531219983,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004996272967588715,
      "loss": 5.5649,
      "mean_token_accuracy": 0.18136372566223144,
      "num_tokens": 6644413.0,
      "step": 2890
    },
    {
      "entropy": 5.668825292587281,
      "epoch": 0.2780979827089337,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996253216774283,
      "loss": 5.5148,
      "mean_token_accuracy": 0.18241942077875137,
      "num_tokens": 6655634.0,
      "step": 2895
    },
    {
      "entropy": 5.68368968963623,
      "epoch": 0.27857829010566765,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996233413808521,
      "loss": 5.4722,
      "mean_token_accuracy": 0.18518155813217163,
      "num_tokens": 6665977.0,
      "step": 2900
    },
    {
      "entropy": 5.742200183868408,
      "epoch": 0.27905859750240153,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004996213558691889,
      "loss": 5.5236,
      "mean_token_accuracy": 0.17702546864748,
      "num_tokens": 6677165.0,
      "step": 2905
    },
    {
      "entropy": 5.718938302993775,
      "epoch": 0.27953890489913547,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004996193651424848,
      "loss": 5.6368,
      "mean_token_accuracy": 0.1747002512216568,
      "num_tokens": 6688182.0,
      "step": 2910
    },
    {
      "entropy": 5.719934892654419,
      "epoch": 0.28001921229586935,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499617369200786,
      "loss": 5.5332,
      "mean_token_accuracy": 0.1799617737531662,
      "num_tokens": 6699259.0,
      "step": 2915
    },
    {
      "entropy": 5.722778844833374,
      "epoch": 0.2804995196926033,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004996153680441389,
      "loss": 5.5587,
      "mean_token_accuracy": 0.1818137377500534,
      "num_tokens": 6710071.0,
      "step": 2920
    },
    {
      "entropy": 5.74660005569458,
      "epoch": 0.28097982708933716,
      "grad_norm": 1.015625,
      "learning_rate": 0.00049961336167259,
      "loss": 5.5901,
      "mean_token_accuracy": 0.18049627989530564,
      "num_tokens": 6720267.0,
      "step": 2925
    },
    {
      "entropy": 5.77274603843689,
      "epoch": 0.2814601344860711,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996113500861857,
      "loss": 5.5781,
      "mean_token_accuracy": 0.17617071717977523,
      "num_tokens": 6730919.0,
      "step": 2930
    },
    {
      "entropy": 5.648800373077393,
      "epoch": 0.281940441882805,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996093332849729,
      "loss": 5.5655,
      "mean_token_accuracy": 0.18327539116144181,
      "num_tokens": 6741480.0,
      "step": 2935
    },
    {
      "entropy": 5.748388051986694,
      "epoch": 0.2824207492795389,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996073112689983,
      "loss": 5.5065,
      "mean_token_accuracy": 0.18213073313236236,
      "num_tokens": 6751816.0,
      "step": 2940
    },
    {
      "entropy": 5.665603494644165,
      "epoch": 0.2829010566762728,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004996052840383088,
      "loss": 5.5378,
      "mean_token_accuracy": 0.18160058706998825,
      "num_tokens": 6763645.0,
      "step": 2945
    },
    {
      "entropy": 5.678023004531861,
      "epoch": 0.28338136407300674,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996032515929516,
      "loss": 5.5128,
      "mean_token_accuracy": 0.18091680556535722,
      "num_tokens": 6774795.0,
      "step": 2950
    },
    {
      "entropy": 5.664436340332031,
      "epoch": 0.2838616714697406,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004996012139329738,
      "loss": 5.4312,
      "mean_token_accuracy": 0.18459193110466005,
      "num_tokens": 6786833.0,
      "step": 2955
    },
    {
      "entropy": 5.591051626205444,
      "epoch": 0.28434197886647455,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004995991710584228,
      "loss": 5.4431,
      "mean_token_accuracy": 0.18741845786571504,
      "num_tokens": 6798134.0,
      "step": 2960
    },
    {
      "entropy": 5.7143707275390625,
      "epoch": 0.28482228626320844,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004995971229693459,
      "loss": 5.564,
      "mean_token_accuracy": 0.18205250948667526,
      "num_tokens": 6807918.0,
      "step": 2965
    },
    {
      "entropy": 5.700773429870606,
      "epoch": 0.28530259365994237,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004995950696657909,
      "loss": 5.6352,
      "mean_token_accuracy": 0.17493324726819992,
      "num_tokens": 6818689.0,
      "step": 2970
    },
    {
      "entropy": 5.689798498153687,
      "epoch": 0.28578290105667625,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004995930111478051,
      "loss": 5.4959,
      "mean_token_accuracy": 0.18560752868652344,
      "num_tokens": 6830871.0,
      "step": 2975
    },
    {
      "entropy": 5.646875667572021,
      "epoch": 0.2862632084534102,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004995909474154365,
      "loss": 5.486,
      "mean_token_accuracy": 0.18816983848810195,
      "num_tokens": 6842857.0,
      "step": 2980
    },
    {
      "entropy": 5.6857727527618405,
      "epoch": 0.28674351585014407,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004995888784687331,
      "loss": 5.5149,
      "mean_token_accuracy": 0.1798010364174843,
      "num_tokens": 6853727.0,
      "step": 2985
    },
    {
      "entropy": 5.747433519363403,
      "epoch": 0.287223823246878,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995868043077428,
      "loss": 5.5493,
      "mean_token_accuracy": 0.18155408054590225,
      "num_tokens": 6864728.0,
      "step": 2990
    },
    {
      "entropy": 5.74623293876648,
      "epoch": 0.2877041306436119,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995847249325137,
      "loss": 5.6145,
      "mean_token_accuracy": 0.17592370361089707,
      "num_tokens": 6875001.0,
      "step": 2995
    },
    {
      "entropy": 5.656736660003662,
      "epoch": 0.2881844380403458,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004995826403430942,
      "loss": 5.4857,
      "mean_token_accuracy": 0.1892928287386894,
      "num_tokens": 6887026.0,
      "step": 3000
    },
    {
      "epoch": 0.2881844380403458,
      "eval_entropy": 5.517254452656561,
      "eval_loss": 5.54893159866333,
      "eval_mean_token_accuracy": 0.18836901158363223,
      "eval_num_tokens": 6887026.0,
      "eval_runtime": 20.6637,
      "eval_samples_per_second": 1588.05,
      "eval_steps_per_second": 198.512,
      "step": 3000
    },
    {
      "entropy": 5.8717550277709964,
      "epoch": 0.2886647454370797,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004995805505395328,
      "loss": 5.6218,
      "mean_token_accuracy": 0.17854404374957084,
      "num_tokens": 6900118.0,
      "step": 3005
    },
    {
      "entropy": 5.688295984268189,
      "epoch": 0.28914505283381364,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004995784555218778,
      "loss": 5.5199,
      "mean_token_accuracy": 0.18320534080266954,
      "num_tokens": 6912135.0,
      "step": 3010
    },
    {
      "entropy": 5.600100708007813,
      "epoch": 0.2896253602305475,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995763552901779,
      "loss": 5.434,
      "mean_token_accuracy": 0.18747649043798448,
      "num_tokens": 6923781.0,
      "step": 3015
    },
    {
      "entropy": 5.716433191299439,
      "epoch": 0.29010566762728146,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004995742498444818,
      "loss": 5.509,
      "mean_token_accuracy": 0.17203439027071,
      "num_tokens": 6935804.0,
      "step": 3020
    },
    {
      "entropy": 5.672420835494995,
      "epoch": 0.2905859750240154,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004995721391848387,
      "loss": 5.545,
      "mean_token_accuracy": 0.18459217697381974,
      "num_tokens": 6948289.0,
      "step": 3025
    },
    {
      "entropy": 5.705364942550659,
      "epoch": 0.2910662824207493,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995700233112972,
      "loss": 5.4888,
      "mean_token_accuracy": 0.18250523060560225,
      "num_tokens": 6958719.0,
      "step": 3030
    },
    {
      "entropy": 5.657338047027588,
      "epoch": 0.2915465898174832,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004995679022239066,
      "loss": 5.5208,
      "mean_token_accuracy": 0.1882460296154022,
      "num_tokens": 6969041.0,
      "step": 3035
    },
    {
      "entropy": 5.713998460769654,
      "epoch": 0.2920268972142171,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004995657759227162,
      "loss": 5.6195,
      "mean_token_accuracy": 0.18382472097873687,
      "num_tokens": 6981206.0,
      "step": 3040
    },
    {
      "entropy": 5.733813333511352,
      "epoch": 0.29250720461095103,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004995636444077751,
      "loss": 5.5,
      "mean_token_accuracy": 0.18588536828756333,
      "num_tokens": 6993169.0,
      "step": 3045
    },
    {
      "entropy": 5.687197494506836,
      "epoch": 0.2929875120076849,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004995615076791333,
      "loss": 5.5687,
      "mean_token_accuracy": 0.1816737398505211,
      "num_tokens": 7005175.0,
      "step": 3050
    },
    {
      "entropy": 5.711205911636353,
      "epoch": 0.29346781940441885,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004995593657368399,
      "loss": 5.5769,
      "mean_token_accuracy": 0.18063196241855622,
      "num_tokens": 7016738.0,
      "step": 3055
    },
    {
      "entropy": 5.722720336914063,
      "epoch": 0.29394812680115273,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499557218580945,
      "loss": 5.4698,
      "mean_token_accuracy": 0.1917486384510994,
      "num_tokens": 7028399.0,
      "step": 3060
    },
    {
      "entropy": 5.712218952178955,
      "epoch": 0.29442843419788667,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004995550662114981,
      "loss": 5.616,
      "mean_token_accuracy": 0.17428845763206482,
      "num_tokens": 7039528.0,
      "step": 3065
    },
    {
      "entropy": 5.572362947463989,
      "epoch": 0.29490874159462055,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004995529086285495,
      "loss": 5.3916,
      "mean_token_accuracy": 0.19135287553071975,
      "num_tokens": 7051399.0,
      "step": 3070
    },
    {
      "entropy": 5.5535026550292965,
      "epoch": 0.2953890489913545,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499550745832149,
      "loss": 5.4031,
      "mean_token_accuracy": 0.19149063378572465,
      "num_tokens": 7062452.0,
      "step": 3075
    },
    {
      "entropy": 5.677850723266602,
      "epoch": 0.29586935638808837,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004995485778223471,
      "loss": 5.5474,
      "mean_token_accuracy": 0.1808340921998024,
      "num_tokens": 7074052.0,
      "step": 3080
    },
    {
      "entropy": 5.749167728424072,
      "epoch": 0.2963496637848223,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004995464045991939,
      "loss": 5.5685,
      "mean_token_accuracy": 0.17625710219144822,
      "num_tokens": 7086692.0,
      "step": 3085
    },
    {
      "entropy": 5.664271974563599,
      "epoch": 0.2968299711815562,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00049954422616274,
      "loss": 5.5488,
      "mean_token_accuracy": 0.18172660619020461,
      "num_tokens": 7098221.0,
      "step": 3090
    },
    {
      "entropy": 5.622652196884156,
      "epoch": 0.2973102785782901,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004995420425130359,
      "loss": 5.5068,
      "mean_token_accuracy": 0.1877342715859413,
      "num_tokens": 7110129.0,
      "step": 3095
    },
    {
      "entropy": 5.6984845161437985,
      "epoch": 0.297790585975024,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004995398536501324,
      "loss": 5.5847,
      "mean_token_accuracy": 0.1803930550813675,
      "num_tokens": 7121523.0,
      "step": 3100
    },
    {
      "entropy": 5.709534502029419,
      "epoch": 0.29827089337175794,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004995376595740801,
      "loss": 5.5071,
      "mean_token_accuracy": 0.1876872569322586,
      "num_tokens": 7133482.0,
      "step": 3105
    },
    {
      "entropy": 5.600775051116943,
      "epoch": 0.2987512007684918,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995354602849302,
      "loss": 5.5006,
      "mean_token_accuracy": 0.18258904665708542,
      "num_tokens": 7144031.0,
      "step": 3110
    },
    {
      "entropy": 5.745647048950195,
      "epoch": 0.29923150816522576,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004995332557827337,
      "loss": 5.5654,
      "mean_token_accuracy": 0.17680131942033767,
      "num_tokens": 7155469.0,
      "step": 3115
    },
    {
      "entropy": 5.7442200660705565,
      "epoch": 0.29971181556195964,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995310460675416,
      "loss": 5.5132,
      "mean_token_accuracy": 0.17851166278123856,
      "num_tokens": 7167077.0,
      "step": 3120
    },
    {
      "entropy": 5.592399930953979,
      "epoch": 0.3001921229586936,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004995288311394053,
      "loss": 5.4402,
      "mean_token_accuracy": 0.18590566366910935,
      "num_tokens": 7178917.0,
      "step": 3125
    },
    {
      "entropy": 5.747562742233276,
      "epoch": 0.30067243035542746,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004995266109983764,
      "loss": 5.5461,
      "mean_token_accuracy": 0.17973315119743347,
      "num_tokens": 7190284.0,
      "step": 3130
    },
    {
      "entropy": 5.689589881896973,
      "epoch": 0.3011527377521614,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004995243856445062,
      "loss": 5.5424,
      "mean_token_accuracy": 0.18706399947404861,
      "num_tokens": 7202391.0,
      "step": 3135
    },
    {
      "entropy": 5.6660867691040036,
      "epoch": 0.3016330451488953,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004995221550778466,
      "loss": 5.4595,
      "mean_token_accuracy": 0.18576400578022004,
      "num_tokens": 7212882.0,
      "step": 3140
    },
    {
      "entropy": 5.583697080612183,
      "epoch": 0.3021133525456292,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004995199192984491,
      "loss": 5.4489,
      "mean_token_accuracy": 0.1874859169125557,
      "num_tokens": 7224102.0,
      "step": 3145
    },
    {
      "entropy": 5.662063217163086,
      "epoch": 0.3025936599423631,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004995176783063657,
      "loss": 5.4004,
      "mean_token_accuracy": 0.1879839137196541,
      "num_tokens": 7234579.0,
      "step": 3150
    },
    {
      "entropy": 5.706206369400024,
      "epoch": 0.30307396733909703,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995154321016487,
      "loss": 5.5484,
      "mean_token_accuracy": 0.18038555681705476,
      "num_tokens": 7246065.0,
      "step": 3155
    },
    {
      "entropy": 5.697294759750366,
      "epoch": 0.3035542747358309,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004995131806843499,
      "loss": 5.4995,
      "mean_token_accuracy": 0.17942927032709122,
      "num_tokens": 7258899.0,
      "step": 3160
    },
    {
      "entropy": 5.712396049499512,
      "epoch": 0.30403458213256485,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004995109240545218,
      "loss": 5.5854,
      "mean_token_accuracy": 0.16906598955392838,
      "num_tokens": 7271887.0,
      "step": 3165
    },
    {
      "entropy": 5.709551382064819,
      "epoch": 0.3045148895292987,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004995086622122167,
      "loss": 5.5355,
      "mean_token_accuracy": 0.17753610461950303,
      "num_tokens": 7282773.0,
      "step": 3170
    },
    {
      "entropy": 5.605632638931274,
      "epoch": 0.30499519692603266,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004995063951574871,
      "loss": 5.3773,
      "mean_token_accuracy": 0.19257392436265947,
      "num_tokens": 7294460.0,
      "step": 3175
    },
    {
      "entropy": 5.5793797969818115,
      "epoch": 0.30547550432276654,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004995041228903856,
      "loss": 5.4241,
      "mean_token_accuracy": 0.18916317224502563,
      "num_tokens": 7305125.0,
      "step": 3180
    },
    {
      "entropy": 5.525802755355835,
      "epoch": 0.3059558117195005,
      "grad_norm": 0.78515625,
      "learning_rate": 0.000499501845410965,
      "loss": 5.4308,
      "mean_token_accuracy": 0.18405420184135438,
      "num_tokens": 7317738.0,
      "step": 3185
    },
    {
      "entropy": 5.682470989227295,
      "epoch": 0.30643611911623436,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004994995627192781,
      "loss": 5.4505,
      "mean_token_accuracy": 0.19196930825710296,
      "num_tokens": 7329820.0,
      "step": 3190
    },
    {
      "entropy": 5.6539403915405275,
      "epoch": 0.3069164265129683,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004994972748153781,
      "loss": 5.4967,
      "mean_token_accuracy": 0.17736558765172958,
      "num_tokens": 7341471.0,
      "step": 3195
    },
    {
      "entropy": 5.701456069946289,
      "epoch": 0.30739673390970224,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499494981699318,
      "loss": 5.628,
      "mean_token_accuracy": 0.18697974234819412,
      "num_tokens": 7352459.0,
      "step": 3200
    },
    {
      "entropy": 5.652633571624756,
      "epoch": 0.3078770413064361,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499492683371151,
      "loss": 5.4152,
      "mean_token_accuracy": 0.1909807428717613,
      "num_tokens": 7363129.0,
      "step": 3205
    },
    {
      "entropy": 5.656173229217529,
      "epoch": 0.30835734870317005,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004994903798309306,
      "loss": 5.4975,
      "mean_token_accuracy": 0.1836207091808319,
      "num_tokens": 7373726.0,
      "step": 3210
    },
    {
      "entropy": 5.67913851737976,
      "epoch": 0.30883765609990393,
      "grad_norm": 1.0,
      "learning_rate": 0.0004994880710787102,
      "loss": 5.5312,
      "mean_token_accuracy": 0.18742814809083938,
      "num_tokens": 7383847.0,
      "step": 3215
    },
    {
      "entropy": 5.678493785858154,
      "epoch": 0.30931796349663787,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004994857571145432,
      "loss": 5.528,
      "mean_token_accuracy": 0.17938138991594316,
      "num_tokens": 7396425.0,
      "step": 3220
    },
    {
      "entropy": 5.720796203613281,
      "epoch": 0.30979827089337175,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004994834379384837,
      "loss": 5.5958,
      "mean_token_accuracy": 0.17902967631816863,
      "num_tokens": 7407591.0,
      "step": 3225
    },
    {
      "entropy": 5.674501752853393,
      "epoch": 0.3102785782901057,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004994811135505851,
      "loss": 5.459,
      "mean_token_accuracy": 0.18573354184627533,
      "num_tokens": 7419668.0,
      "step": 3230
    },
    {
      "entropy": 5.704877758026123,
      "epoch": 0.31075888568683957,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994787839509018,
      "loss": 5.5598,
      "mean_token_accuracy": 0.17747900784015655,
      "num_tokens": 7431969.0,
      "step": 3235
    },
    {
      "entropy": 5.772231531143189,
      "epoch": 0.3112391930835735,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004994764491394876,
      "loss": 5.5638,
      "mean_token_accuracy": 0.1785635992884636,
      "num_tokens": 7443467.0,
      "step": 3240
    },
    {
      "entropy": 5.702940320968628,
      "epoch": 0.3117195004803074,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004994741091163969,
      "loss": 5.5465,
      "mean_token_accuracy": 0.18591915667057038,
      "num_tokens": 7455220.0,
      "step": 3245
    },
    {
      "entropy": 5.640719938278198,
      "epoch": 0.3121998078770413,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499471763881684,
      "loss": 5.4527,
      "mean_token_accuracy": 0.18479665219783784,
      "num_tokens": 7466452.0,
      "step": 3250
    },
    {
      "entropy": 5.6805953025817875,
      "epoch": 0.3126801152737752,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004994694134354031,
      "loss": 5.4975,
      "mean_token_accuracy": 0.19328097999095917,
      "num_tokens": 7478798.0,
      "step": 3255
    },
    {
      "entropy": 5.71582818031311,
      "epoch": 0.31316042267050914,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000499467057777609,
      "loss": 5.5839,
      "mean_token_accuracy": 0.1815870925784111,
      "num_tokens": 7491039.0,
      "step": 3260
    },
    {
      "entropy": 5.6178038120269775,
      "epoch": 0.313640730067243,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004994646969083565,
      "loss": 5.4549,
      "mean_token_accuracy": 0.1828945755958557,
      "num_tokens": 7501455.0,
      "step": 3265
    },
    {
      "entropy": 5.664418077468872,
      "epoch": 0.31412103746397696,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004994623308277002,
      "loss": 5.5032,
      "mean_token_accuracy": 0.18493229746818543,
      "num_tokens": 7513484.0,
      "step": 3270
    },
    {
      "entropy": 5.721431922912598,
      "epoch": 0.31460134486071084,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000499459959535695,
      "loss": 5.4825,
      "mean_token_accuracy": 0.18559184819459915,
      "num_tokens": 7524868.0,
      "step": 3275
    },
    {
      "entropy": 5.519072532653809,
      "epoch": 0.3150816522574448,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004994575830323962,
      "loss": 5.3589,
      "mean_token_accuracy": 0.19548385143280028,
      "num_tokens": 7536290.0,
      "step": 3280
    },
    {
      "entropy": 5.604799842834472,
      "epoch": 0.31556195965417866,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004994552013178586,
      "loss": 5.48,
      "mean_token_accuracy": 0.18394231498241426,
      "num_tokens": 7546994.0,
      "step": 3285
    },
    {
      "entropy": 5.665694236755371,
      "epoch": 0.3160422670509126,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499452814392138,
      "loss": 5.4675,
      "mean_token_accuracy": 0.18123726844787597,
      "num_tokens": 7559231.0,
      "step": 3290
    },
    {
      "entropy": 5.536679315567016,
      "epoch": 0.3165225744476465,
      "grad_norm": 0.875,
      "learning_rate": 0.0004994504222552894,
      "loss": 5.4434,
      "mean_token_accuracy": 0.18347110748291015,
      "num_tokens": 7570894.0,
      "step": 3295
    },
    {
      "entropy": 5.71035327911377,
      "epoch": 0.3170028818443804,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004994480249073684,
      "loss": 5.5494,
      "mean_token_accuracy": 0.17979539036750794,
      "num_tokens": 7581765.0,
      "step": 3300
    },
    {
      "entropy": 5.612960243225098,
      "epoch": 0.3174831892411143,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004994456223484308,
      "loss": 5.3981,
      "mean_token_accuracy": 0.19041623175144196,
      "num_tokens": 7591828.0,
      "step": 3305
    },
    {
      "entropy": 5.583401203155518,
      "epoch": 0.31796349663784823,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004994432145785323,
      "loss": 5.4716,
      "mean_token_accuracy": 0.18562580198049544,
      "num_tokens": 7604230.0,
      "step": 3310
    },
    {
      "entropy": 5.616955995559692,
      "epoch": 0.3184438040345821,
      "grad_norm": 1.0,
      "learning_rate": 0.0004994408015977288,
      "loss": 5.4265,
      "mean_token_accuracy": 0.1882665678858757,
      "num_tokens": 7615780.0,
      "step": 3315
    },
    {
      "entropy": 5.640593147277832,
      "epoch": 0.31892411143131605,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004994383834060764,
      "loss": 5.3919,
      "mean_token_accuracy": 0.18447416573762893,
      "num_tokens": 7627633.0,
      "step": 3320
    },
    {
      "entropy": 5.608803081512451,
      "epoch": 0.31940441882804993,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004994359600036311,
      "loss": 5.5195,
      "mean_token_accuracy": 0.18261253833770752,
      "num_tokens": 7639487.0,
      "step": 3325
    },
    {
      "entropy": 5.651657962799073,
      "epoch": 0.31988472622478387,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004994335313904493,
      "loss": 5.4392,
      "mean_token_accuracy": 0.18792017847299575,
      "num_tokens": 7649976.0,
      "step": 3330
    },
    {
      "entropy": 5.766524124145508,
      "epoch": 0.32036503362151775,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004994310975665873,
      "loss": 5.6262,
      "mean_token_accuracy": 0.17568368166685105,
      "num_tokens": 7662555.0,
      "step": 3335
    },
    {
      "entropy": 5.5539319038391115,
      "epoch": 0.3208453410182517,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004994286585321017,
      "loss": 5.3943,
      "mean_token_accuracy": 0.19039005935192108,
      "num_tokens": 7673869.0,
      "step": 3340
    },
    {
      "entropy": 5.526929521560669,
      "epoch": 0.32132564841498557,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000499426214287049,
      "loss": 5.4915,
      "mean_token_accuracy": 0.1830260917544365,
      "num_tokens": 7687149.0,
      "step": 3345
    },
    {
      "entropy": 5.745597219467163,
      "epoch": 0.3218059558117195,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004994237648314862,
      "loss": 5.54,
      "mean_token_accuracy": 0.17828639447689057,
      "num_tokens": 7698352.0,
      "step": 3350
    },
    {
      "entropy": 5.590162038803101,
      "epoch": 0.3222862632084534,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004994213101654697,
      "loss": 5.3734,
      "mean_token_accuracy": 0.19427183121442795,
      "num_tokens": 7709117.0,
      "step": 3355
    },
    {
      "entropy": 5.597160816192627,
      "epoch": 0.3227665706051873,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000499418850289057,
      "loss": 5.4077,
      "mean_token_accuracy": 0.19042223393917085,
      "num_tokens": 7721731.0,
      "step": 3360
    },
    {
      "entropy": 5.554579448699951,
      "epoch": 0.32324687800192126,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004994163852023048,
      "loss": 5.3841,
      "mean_token_accuracy": 0.1844659060239792,
      "num_tokens": 7733010.0,
      "step": 3365
    },
    {
      "entropy": 5.59989914894104,
      "epoch": 0.32372718539865514,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004994139149052706,
      "loss": 5.3662,
      "mean_token_accuracy": 0.19753526598215104,
      "num_tokens": 7743905.0,
      "step": 3370
    },
    {
      "entropy": 5.625758695602417,
      "epoch": 0.3242074927953891,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004994114393980117,
      "loss": 5.3983,
      "mean_token_accuracy": 0.1913401886820793,
      "num_tokens": 7754891.0,
      "step": 3375
    },
    {
      "entropy": 5.632540130615235,
      "epoch": 0.32468780019212296,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004994089586805856,
      "loss": 5.521,
      "mean_token_accuracy": 0.17907267212867736,
      "num_tokens": 7766831.0,
      "step": 3380
    },
    {
      "entropy": 5.6776283264160154,
      "epoch": 0.3251681075888569,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004994064727530496,
      "loss": 5.5922,
      "mean_token_accuracy": 0.17850053757429124,
      "num_tokens": 7778239.0,
      "step": 3385
    },
    {
      "entropy": 5.658124542236328,
      "epoch": 0.3256484149855908,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004994039816154618,
      "loss": 5.3979,
      "mean_token_accuracy": 0.19412729442119597,
      "num_tokens": 7789449.0,
      "step": 3390
    },
    {
      "entropy": 5.679704189300537,
      "epoch": 0.3261287223823247,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00049940148526788,
      "loss": 5.4911,
      "mean_token_accuracy": 0.18398293256759643,
      "num_tokens": 7799507.0,
      "step": 3395
    },
    {
      "entropy": 5.6712682247161865,
      "epoch": 0.3266090297790586,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004993989837103618,
      "loss": 5.4766,
      "mean_token_accuracy": 0.19145456999540328,
      "num_tokens": 7811730.0,
      "step": 3400
    },
    {
      "entropy": 5.541997814178467,
      "epoch": 0.3270893371757925,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004993964769429657,
      "loss": 5.3581,
      "mean_token_accuracy": 0.19294817596673966,
      "num_tokens": 7823108.0,
      "step": 3405
    },
    {
      "entropy": 5.617771482467651,
      "epoch": 0.3275696445725264,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004993939649657498,
      "loss": 5.4307,
      "mean_token_accuracy": 0.18940595984458924,
      "num_tokens": 7835384.0,
      "step": 3410
    },
    {
      "entropy": 5.658884811401367,
      "epoch": 0.32804995196926034,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004993914477787721,
      "loss": 5.4704,
      "mean_token_accuracy": 0.19038086384534836,
      "num_tokens": 7845860.0,
      "step": 3415
    },
    {
      "entropy": 5.630373620986939,
      "epoch": 0.3285302593659942,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004993889253820915,
      "loss": 5.5263,
      "mean_token_accuracy": 0.1801797404885292,
      "num_tokens": 7856861.0,
      "step": 3420
    },
    {
      "entropy": 5.68350625038147,
      "epoch": 0.32901056676272816,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004993863977757663,
      "loss": 5.4844,
      "mean_token_accuracy": 0.18732366859912872,
      "num_tokens": 7868902.0,
      "step": 3425
    },
    {
      "entropy": 5.662414789199829,
      "epoch": 0.32949087415946204,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004993838649598552,
      "loss": 5.4886,
      "mean_token_accuracy": 0.18540164977312087,
      "num_tokens": 7880369.0,
      "step": 3430
    },
    {
      "entropy": 5.6702476978302006,
      "epoch": 0.329971181556196,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004993813269344171,
      "loss": 5.4879,
      "mean_token_accuracy": 0.18206911385059357,
      "num_tokens": 7892041.0,
      "step": 3435
    },
    {
      "entropy": 5.576903867721557,
      "epoch": 0.33045148895292986,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004993787836995108,
      "loss": 5.4226,
      "mean_token_accuracy": 0.1905149057507515,
      "num_tokens": 7903644.0,
      "step": 3440
    },
    {
      "entropy": 5.60099048614502,
      "epoch": 0.3309317963496638,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004993762352551954,
      "loss": 5.5044,
      "mean_token_accuracy": 0.18278414756059647,
      "num_tokens": 7915679.0,
      "step": 3445
    },
    {
      "entropy": 5.685850667953491,
      "epoch": 0.3314121037463977,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004993736816015301,
      "loss": 5.5227,
      "mean_token_accuracy": 0.18584686517715454,
      "num_tokens": 7928020.0,
      "step": 3450
    },
    {
      "entropy": 5.706366539001465,
      "epoch": 0.3318924111431316,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004993711227385742,
      "loss": 5.5743,
      "mean_token_accuracy": 0.18078709840774537,
      "num_tokens": 7940521.0,
      "step": 3455
    },
    {
      "entropy": 5.6799705028533936,
      "epoch": 0.3323727185398655,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004993685586663871,
      "loss": 5.5121,
      "mean_token_accuracy": 0.17839492708444596,
      "num_tokens": 7952234.0,
      "step": 3460
    },
    {
      "entropy": 5.655110263824463,
      "epoch": 0.33285302593659943,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004993659893850281,
      "loss": 5.5683,
      "mean_token_accuracy": 0.18211730420589448,
      "num_tokens": 7964900.0,
      "step": 3465
    },
    {
      "entropy": 5.7237406253814695,
      "epoch": 0.3333333333333333,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004993634148945573,
      "loss": 5.5663,
      "mean_token_accuracy": 0.182293364405632,
      "num_tokens": 7976661.0,
      "step": 3470
    },
    {
      "entropy": 5.630974531173706,
      "epoch": 0.33381364073006725,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004993608351950341,
      "loss": 5.4047,
      "mean_token_accuracy": 0.18842126727104186,
      "num_tokens": 7987605.0,
      "step": 3475
    },
    {
      "entropy": 5.526041412353516,
      "epoch": 0.33429394812680113,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004993582502865185,
      "loss": 5.4611,
      "mean_token_accuracy": 0.1937727600336075,
      "num_tokens": 7998879.0,
      "step": 3480
    },
    {
      "entropy": 5.628568124771118,
      "epoch": 0.33477425552353507,
      "grad_norm": 0.875,
      "learning_rate": 0.0004993556601690706,
      "loss": 5.4519,
      "mean_token_accuracy": 0.18745657056570053,
      "num_tokens": 8010634.0,
      "step": 3485
    },
    {
      "entropy": 5.671873331069946,
      "epoch": 0.33525456292026895,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004993530648427505,
      "loss": 5.4563,
      "mean_token_accuracy": 0.181246218085289,
      "num_tokens": 8022134.0,
      "step": 3490
    },
    {
      "entropy": 5.589345169067383,
      "epoch": 0.3357348703170029,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004993504643076184,
      "loss": 5.4126,
      "mean_token_accuracy": 0.18356649279594422,
      "num_tokens": 8034454.0,
      "step": 3495
    },
    {
      "entropy": 5.656652736663818,
      "epoch": 0.33621517771373677,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004993478585637347,
      "loss": 5.5004,
      "mean_token_accuracy": 0.17905702888965608,
      "num_tokens": 8046175.0,
      "step": 3500
    },
    {
      "entropy": 5.747422742843628,
      "epoch": 0.3366954851104707,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004993452476111599,
      "loss": 5.5546,
      "mean_token_accuracy": 0.18119859397411348,
      "num_tokens": 8057335.0,
      "step": 3505
    },
    {
      "entropy": 5.555219221115112,
      "epoch": 0.3371757925072046,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004993426314499546,
      "loss": 5.3693,
      "mean_token_accuracy": 0.19091724455356598,
      "num_tokens": 8068860.0,
      "step": 3510
    },
    {
      "entropy": 5.595325565338134,
      "epoch": 0.3376560999039385,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993400100801796,
      "loss": 5.4425,
      "mean_token_accuracy": 0.18414415121078492,
      "num_tokens": 8078957.0,
      "step": 3515
    },
    {
      "entropy": 5.710655450820923,
      "epoch": 0.3381364073006724,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004993373835018956,
      "loss": 5.5512,
      "mean_token_accuracy": 0.17972831279039383,
      "num_tokens": 8090143.0,
      "step": 3520
    },
    {
      "entropy": 5.562273979187012,
      "epoch": 0.33861671469740634,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004993347517151638,
      "loss": 5.3538,
      "mean_token_accuracy": 0.1909133017063141,
      "num_tokens": 8101961.0,
      "step": 3525
    },
    {
      "entropy": 5.440454292297363,
      "epoch": 0.3390970220941403,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004993321147200452,
      "loss": 5.305,
      "mean_token_accuracy": 0.19344836473464966,
      "num_tokens": 8112635.0,
      "step": 3530
    },
    {
      "entropy": 5.658025026321411,
      "epoch": 0.33957732949087416,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499329472516601,
      "loss": 5.4841,
      "mean_token_accuracy": 0.18330524116754532,
      "num_tokens": 8123730.0,
      "step": 3535
    },
    {
      "entropy": 5.680354881286621,
      "epoch": 0.3400576368876081,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004993268251048925,
      "loss": 5.491,
      "mean_token_accuracy": 0.18179696649312974,
      "num_tokens": 8135258.0,
      "step": 3540
    },
    {
      "entropy": 5.5405772686004635,
      "epoch": 0.340537944284342,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004993241724849814,
      "loss": 5.3467,
      "mean_token_accuracy": 0.19629553258419036,
      "num_tokens": 8145946.0,
      "step": 3545
    },
    {
      "entropy": 5.65671672821045,
      "epoch": 0.3410182516810759,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000499321514656929,
      "loss": 5.5585,
      "mean_token_accuracy": 0.18194225430488586,
      "num_tokens": 8158422.0,
      "step": 3550
    },
    {
      "entropy": 5.634524154663086,
      "epoch": 0.3414985590778098,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004993188516207972,
      "loss": 5.4852,
      "mean_token_accuracy": 0.1771097257733345,
      "num_tokens": 8170849.0,
      "step": 3555
    },
    {
      "entropy": 5.557367610931396,
      "epoch": 0.34197886647454373,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004993161833766478,
      "loss": 5.4224,
      "mean_token_accuracy": 0.18828796595335007,
      "num_tokens": 8180907.0,
      "step": 3560
    },
    {
      "entropy": 5.694223213195801,
      "epoch": 0.3424591738712776,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004993135099245426,
      "loss": 5.4888,
      "mean_token_accuracy": 0.18229319155216217,
      "num_tokens": 8193418.0,
      "step": 3565
    },
    {
      "entropy": 5.634408569335937,
      "epoch": 0.34293948126801155,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004993108312645438,
      "loss": 5.5723,
      "mean_token_accuracy": 0.18664569556713104,
      "num_tokens": 8203873.0,
      "step": 3570
    },
    {
      "entropy": 5.762505292892456,
      "epoch": 0.34341978866474543,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004993081473967135,
      "loss": 5.5715,
      "mean_token_accuracy": 0.17954090237617493,
      "num_tokens": 8215150.0,
      "step": 3575
    },
    {
      "entropy": 5.665427827835083,
      "epoch": 0.34390009606147937,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004993054583211143,
      "loss": 5.5281,
      "mean_token_accuracy": 0.18246499598026275,
      "num_tokens": 8225732.0,
      "step": 3580
    },
    {
      "entropy": 5.664745283126831,
      "epoch": 0.34438040345821325,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004993027640378081,
      "loss": 5.4783,
      "mean_token_accuracy": 0.18502768874168396,
      "num_tokens": 8237523.0,
      "step": 3585
    },
    {
      "entropy": 5.658119249343872,
      "epoch": 0.3448607108549472,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499300064546858,
      "loss": 5.4662,
      "mean_token_accuracy": 0.18917794674634933,
      "num_tokens": 8249660.0,
      "step": 3590
    },
    {
      "entropy": 5.528011798858643,
      "epoch": 0.34534101825168106,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004992973598483264,
      "loss": 5.3091,
      "mean_token_accuracy": 0.195291231572628,
      "num_tokens": 8261224.0,
      "step": 3595
    },
    {
      "entropy": 5.541525983810425,
      "epoch": 0.345821325648415,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499294649942276,
      "loss": 5.5357,
      "mean_token_accuracy": 0.18459745049476622,
      "num_tokens": 8272177.0,
      "step": 3600
    },
    {
      "entropy": 5.636389636993409,
      "epoch": 0.3463016330451489,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004992919348287699,
      "loss": 5.4267,
      "mean_token_accuracy": 0.1912529721856117,
      "num_tokens": 8284449.0,
      "step": 3605
    },
    {
      "entropy": 5.64267406463623,
      "epoch": 0.3467819404418828,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004992892145078711,
      "loss": 5.493,
      "mean_token_accuracy": 0.19091300517320633,
      "num_tokens": 8297028.0,
      "step": 3610
    },
    {
      "entropy": 5.592890691757202,
      "epoch": 0.3472622478386167,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004992864889796427,
      "loss": 5.3719,
      "mean_token_accuracy": 0.1985611706972122,
      "num_tokens": 8308348.0,
      "step": 3615
    },
    {
      "entropy": 5.544464874267578,
      "epoch": 0.34774255523535064,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004992837582441481,
      "loss": 5.3816,
      "mean_token_accuracy": 0.19663999527692794,
      "num_tokens": 8318951.0,
      "step": 3620
    },
    {
      "entropy": 5.636328315734863,
      "epoch": 0.3482228626320845,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004992810223014506,
      "loss": 5.4969,
      "mean_token_accuracy": 0.18055207431316375,
      "num_tokens": 8329979.0,
      "step": 3625
    },
    {
      "entropy": 5.643764209747315,
      "epoch": 0.34870317002881845,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004992782811516137,
      "loss": 5.4394,
      "mean_token_accuracy": 0.19049171060323716,
      "num_tokens": 8340332.0,
      "step": 3630
    },
    {
      "entropy": 5.624525833129883,
      "epoch": 0.34918347742555234,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004992755347947011,
      "loss": 5.4455,
      "mean_token_accuracy": 0.1862572282552719,
      "num_tokens": 8350376.0,
      "step": 3635
    },
    {
      "entropy": 5.608530282974243,
      "epoch": 0.34966378482228627,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004992727832307766,
      "loss": 5.3723,
      "mean_token_accuracy": 0.1924934431910515,
      "num_tokens": 8361884.0,
      "step": 3640
    },
    {
      "entropy": 5.444787120819091,
      "epoch": 0.35014409221902015,
      "grad_norm": 0.875,
      "learning_rate": 0.0004992700264599039,
      "loss": 5.313,
      "mean_token_accuracy": 0.19729706197977065,
      "num_tokens": 8372914.0,
      "step": 3645
    },
    {
      "entropy": 5.6330359935760494,
      "epoch": 0.3506243996157541,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004992672644821473,
      "loss": 5.4823,
      "mean_token_accuracy": 0.18294698148965835,
      "num_tokens": 8385800.0,
      "step": 3650
    },
    {
      "entropy": 5.585728883743286,
      "epoch": 0.35110470701248797,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004992644972975707,
      "loss": 5.3707,
      "mean_token_accuracy": 0.19337564557790757,
      "num_tokens": 8397783.0,
      "step": 3655
    },
    {
      "entropy": 5.579562187194824,
      "epoch": 0.3515850144092219,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004992617249062383,
      "loss": 5.43,
      "mean_token_accuracy": 0.18734387904405594,
      "num_tokens": 8409096.0,
      "step": 3660
    },
    {
      "entropy": 5.537727117538452,
      "epoch": 0.3520653218059558,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004992589473082147,
      "loss": 5.3413,
      "mean_token_accuracy": 0.1902912124991417,
      "num_tokens": 8420833.0,
      "step": 3665
    },
    {
      "entropy": 5.604993677139282,
      "epoch": 0.3525456292026897,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004992561645035641,
      "loss": 5.4175,
      "mean_token_accuracy": 0.18516699671745301,
      "num_tokens": 8431848.0,
      "step": 3670
    },
    {
      "entropy": 5.700228500366211,
      "epoch": 0.3530259365994236,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004992533764923515,
      "loss": 5.5086,
      "mean_token_accuracy": 0.18160357177257538,
      "num_tokens": 8443620.0,
      "step": 3675
    },
    {
      "entropy": 5.629539251327515,
      "epoch": 0.35350624399615754,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004992505832746412,
      "loss": 5.502,
      "mean_token_accuracy": 0.1890736609697342,
      "num_tokens": 8455339.0,
      "step": 3680
    },
    {
      "entropy": 5.532653284072876,
      "epoch": 0.3539865513928914,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004992477848504983,
      "loss": 5.3373,
      "mean_token_accuracy": 0.20125855654478073,
      "num_tokens": 8466342.0,
      "step": 3685
    },
    {
      "entropy": 5.606335020065307,
      "epoch": 0.35446685878962536,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004992449812199877,
      "loss": 5.3829,
      "mean_token_accuracy": 0.19097561985254288,
      "num_tokens": 8476955.0,
      "step": 3690
    },
    {
      "entropy": 5.483207845687867,
      "epoch": 0.3549471661863593,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004992421723831745,
      "loss": 5.3572,
      "mean_token_accuracy": 0.18940153419971467,
      "num_tokens": 8488148.0,
      "step": 3695
    },
    {
      "entropy": 5.599952650070191,
      "epoch": 0.3554274735830932,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004992393583401239,
      "loss": 5.3687,
      "mean_token_accuracy": 0.1926783189177513,
      "num_tokens": 8499014.0,
      "step": 3700
    },
    {
      "entropy": 5.6040562152862545,
      "epoch": 0.3559077809798271,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004992365390909014,
      "loss": 5.4318,
      "mean_token_accuracy": 0.1858607843518257,
      "num_tokens": 8510191.0,
      "step": 3705
    },
    {
      "entropy": 5.583250379562378,
      "epoch": 0.356388088376561,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004992337146355721,
      "loss": 5.3295,
      "mean_token_accuracy": 0.19488962441682817,
      "num_tokens": 8522368.0,
      "step": 3710
    },
    {
      "entropy": 5.519905471801758,
      "epoch": 0.35686839577329493,
      "grad_norm": 0.875,
      "learning_rate": 0.0004992308849742019,
      "loss": 5.3455,
      "mean_token_accuracy": 0.19564225971698762,
      "num_tokens": 8533692.0,
      "step": 3715
    },
    {
      "entropy": 5.549171066284179,
      "epoch": 0.3573487031700288,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004992280501068563,
      "loss": 5.4301,
      "mean_token_accuracy": 0.18890246897935867,
      "num_tokens": 8545283.0,
      "step": 3720
    },
    {
      "entropy": 5.658993053436279,
      "epoch": 0.35782901056676275,
      "grad_norm": 1.0,
      "learning_rate": 0.0004992252100336012,
      "loss": 5.4097,
      "mean_token_accuracy": 0.18976463079452516,
      "num_tokens": 8555094.0,
      "step": 3725
    },
    {
      "entropy": 5.564365720748901,
      "epoch": 0.35830931796349663,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004992223647545027,
      "loss": 5.4475,
      "mean_token_accuracy": 0.18259308189153672,
      "num_tokens": 8565179.0,
      "step": 3730
    },
    {
      "entropy": 5.529795169830322,
      "epoch": 0.35878962536023057,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004992195142696266,
      "loss": 5.3479,
      "mean_token_accuracy": 0.18502709716558458,
      "num_tokens": 8575764.0,
      "step": 3735
    },
    {
      "entropy": 5.451181125640869,
      "epoch": 0.35926993275696445,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004992166585790391,
      "loss": 5.3068,
      "mean_token_accuracy": 0.1993624046444893,
      "num_tokens": 8586052.0,
      "step": 3740
    },
    {
      "entropy": 5.620107841491699,
      "epoch": 0.3597502401536984,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004992137976828067,
      "loss": 5.4476,
      "mean_token_accuracy": 0.18854943215847014,
      "num_tokens": 8597832.0,
      "step": 3745
    },
    {
      "entropy": 5.603310203552246,
      "epoch": 0.36023054755043227,
      "grad_norm": 0.875,
      "learning_rate": 0.0004992109315809955,
      "loss": 5.4426,
      "mean_token_accuracy": 0.1903956025838852,
      "num_tokens": 8609917.0,
      "step": 3750
    },
    {
      "entropy": 5.71903510093689,
      "epoch": 0.3607108549471662,
      "grad_norm": 1.0,
      "learning_rate": 0.0004992080602736725,
      "loss": 5.4623,
      "mean_token_accuracy": 0.18415833562612532,
      "num_tokens": 8620137.0,
      "step": 3755
    },
    {
      "entropy": 5.559347629547119,
      "epoch": 0.3611911623439001,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004992051837609039,
      "loss": 5.4785,
      "mean_token_accuracy": 0.19106023013591766,
      "num_tokens": 8630620.0,
      "step": 3760
    },
    {
      "entropy": 5.550971841812133,
      "epoch": 0.361671469740634,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004992023020427568,
      "loss": 5.3295,
      "mean_token_accuracy": 0.19601178914308548,
      "num_tokens": 8642820.0,
      "step": 3765
    },
    {
      "entropy": 5.565942287445068,
      "epoch": 0.3621517771373679,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004991994151192979,
      "loss": 5.341,
      "mean_token_accuracy": 0.19805176854133605,
      "num_tokens": 8654081.0,
      "step": 3770
    },
    {
      "entropy": 5.577949523925781,
      "epoch": 0.36263208453410184,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004991965229905943,
      "loss": 5.3503,
      "mean_token_accuracy": 0.19264333993196486,
      "num_tokens": 8665097.0,
      "step": 3775
    },
    {
      "entropy": 5.552213668823242,
      "epoch": 0.3631123919308357,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004991936256567133,
      "loss": 5.3474,
      "mean_token_accuracy": 0.19428305327892303,
      "num_tokens": 8676576.0,
      "step": 3780
    },
    {
      "entropy": 5.539322233200073,
      "epoch": 0.36359269932756966,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000499190723117722,
      "loss": 5.3619,
      "mean_token_accuracy": 0.18652667999267578,
      "num_tokens": 8689217.0,
      "step": 3785
    },
    {
      "entropy": 5.65005145072937,
      "epoch": 0.36407300672430354,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004991878153736877,
      "loss": 5.4994,
      "mean_token_accuracy": 0.18317543119192123,
      "num_tokens": 8701101.0,
      "step": 3790
    },
    {
      "entropy": 5.604945135116577,
      "epoch": 0.3645533141210375,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004991849024246781,
      "loss": 5.4069,
      "mean_token_accuracy": 0.19224559813737868,
      "num_tokens": 8712769.0,
      "step": 3795
    },
    {
      "entropy": 5.467254781723023,
      "epoch": 0.36503362151777136,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004991819842707608,
      "loss": 5.3379,
      "mean_token_accuracy": 0.1974531129002571,
      "num_tokens": 8724691.0,
      "step": 3800
    },
    {
      "entropy": 5.520613098144532,
      "epoch": 0.3655139289145053,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004991790609120035,
      "loss": 5.3102,
      "mean_token_accuracy": 0.1991004079580307,
      "num_tokens": 8738008.0,
      "step": 3805
    },
    {
      "entropy": 5.639196872711182,
      "epoch": 0.3659942363112392,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499176132348474,
      "loss": 5.5183,
      "mean_token_accuracy": 0.18328616619110108,
      "num_tokens": 8750353.0,
      "step": 3810
    },
    {
      "entropy": 5.554964828491211,
      "epoch": 0.3664745437079731,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004991731985802405,
      "loss": 5.3201,
      "mean_token_accuracy": 0.1959972083568573,
      "num_tokens": 8761319.0,
      "step": 3815
    },
    {
      "entropy": 5.565137004852295,
      "epoch": 0.366954851104707,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004991702596073708,
      "loss": 5.382,
      "mean_token_accuracy": 0.18771857768297195,
      "num_tokens": 8772722.0,
      "step": 3820
    },
    {
      "entropy": 5.666484832763672,
      "epoch": 0.36743515850144093,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004991673154299335,
      "loss": 5.4685,
      "mean_token_accuracy": 0.18808210641145706,
      "num_tokens": 8783886.0,
      "step": 3825
    },
    {
      "entropy": 5.564854669570923,
      "epoch": 0.3679154658981748,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004991643660479967,
      "loss": 5.3901,
      "mean_token_accuracy": 0.1928197354078293,
      "num_tokens": 8795161.0,
      "step": 3830
    },
    {
      "entropy": 5.494964551925659,
      "epoch": 0.36839577329490875,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004991614114616289,
      "loss": 5.3555,
      "mean_token_accuracy": 0.19581058472394944,
      "num_tokens": 8805417.0,
      "step": 3835
    },
    {
      "entropy": 5.540738725662232,
      "epoch": 0.3688760806916426,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004991584516708988,
      "loss": 5.4235,
      "mean_token_accuracy": 0.18542098104953766,
      "num_tokens": 8817980.0,
      "step": 3840
    },
    {
      "entropy": 5.4776569366455075,
      "epoch": 0.36935638808837656,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004991554866758751,
      "loss": 5.2202,
      "mean_token_accuracy": 0.20287279039621353,
      "num_tokens": 8829814.0,
      "step": 3845
    },
    {
      "entropy": 5.465519428253174,
      "epoch": 0.36983669548511044,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004991525164766265,
      "loss": 5.3321,
      "mean_token_accuracy": 0.1917806312441826,
      "num_tokens": 8840220.0,
      "step": 3850
    },
    {
      "entropy": 5.576660633087158,
      "epoch": 0.3703170028818444,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004991495410732222,
      "loss": 5.4567,
      "mean_token_accuracy": 0.18965231329202653,
      "num_tokens": 8852156.0,
      "step": 3855
    },
    {
      "entropy": 5.629491853713989,
      "epoch": 0.37079731027857826,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004991465604657311,
      "loss": 5.381,
      "mean_token_accuracy": 0.18899564146995546,
      "num_tokens": 8863362.0,
      "step": 3860
    },
    {
      "entropy": 5.54644079208374,
      "epoch": 0.3712776176753122,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004991435746542224,
      "loss": 5.3972,
      "mean_token_accuracy": 0.1930238977074623,
      "num_tokens": 8874957.0,
      "step": 3865
    },
    {
      "entropy": 5.508120584487915,
      "epoch": 0.37175792507204614,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004991405836387655,
      "loss": 5.3321,
      "mean_token_accuracy": 0.19720477610826492,
      "num_tokens": 8886171.0,
      "step": 3870
    },
    {
      "entropy": 5.570309400558472,
      "epoch": 0.37223823246878,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004991375874194298,
      "loss": 5.4571,
      "mean_token_accuracy": 0.18851193636655808,
      "num_tokens": 8898103.0,
      "step": 3875
    },
    {
      "entropy": 5.5911322116851805,
      "epoch": 0.37271853986551395,
      "grad_norm": 0.859375,
      "learning_rate": 0.000499134585996285,
      "loss": 5.4296,
      "mean_token_accuracy": 0.1882340207695961,
      "num_tokens": 8910195.0,
      "step": 3880
    },
    {
      "entropy": 5.512739896774292,
      "epoch": 0.37319884726224783,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004991315793694004,
      "loss": 5.3757,
      "mean_token_accuracy": 0.19175583571195604,
      "num_tokens": 8922865.0,
      "step": 3885
    },
    {
      "entropy": 5.536996984481812,
      "epoch": 0.37367915465898177,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004991285675388463,
      "loss": 5.4142,
      "mean_token_accuracy": 0.18459559977054596,
      "num_tokens": 8933913.0,
      "step": 3890
    },
    {
      "entropy": 5.528164482116699,
      "epoch": 0.37415946205571565,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004991255505046922,
      "loss": 5.3323,
      "mean_token_accuracy": 0.19299012273550034,
      "num_tokens": 8945749.0,
      "step": 3895
    },
    {
      "entropy": 5.574621534347534,
      "epoch": 0.3746397694524496,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004991225282670083,
      "loss": 5.3368,
      "mean_token_accuracy": 0.18833704441785812,
      "num_tokens": 8957454.0,
      "step": 3900
    },
    {
      "entropy": 5.576537847518921,
      "epoch": 0.37512007684918347,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499119500825865,
      "loss": 5.3979,
      "mean_token_accuracy": 0.18703885674476622,
      "num_tokens": 8968173.0,
      "step": 3905
    },
    {
      "entropy": 5.520160102844239,
      "epoch": 0.3756003842459174,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004991164681813323,
      "loss": 5.4052,
      "mean_token_accuracy": 0.18889042139053344,
      "num_tokens": 8979187.0,
      "step": 3910
    },
    {
      "entropy": 5.56157865524292,
      "epoch": 0.3760806916426513,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004991134303334807,
      "loss": 5.3242,
      "mean_token_accuracy": 0.20045467913150788,
      "num_tokens": 8990007.0,
      "step": 3915
    },
    {
      "entropy": 5.564577341079712,
      "epoch": 0.3765609990393852,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004991103872823807,
      "loss": 5.459,
      "mean_token_accuracy": 0.18554384261369705,
      "num_tokens": 9001108.0,
      "step": 3920
    },
    {
      "entropy": 5.639509105682373,
      "epoch": 0.3770413064361191,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499107339028103,
      "loss": 5.4649,
      "mean_token_accuracy": 0.18744481950998307,
      "num_tokens": 9012704.0,
      "step": 3925
    },
    {
      "entropy": 5.5753583908081055,
      "epoch": 0.37752161383285304,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004991042855707184,
      "loss": 5.3891,
      "mean_token_accuracy": 0.1935709610581398,
      "num_tokens": 9024351.0,
      "step": 3930
    },
    {
      "entropy": 5.500352001190185,
      "epoch": 0.3780019212295869,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004991012269102977,
      "loss": 5.3106,
      "mean_token_accuracy": 0.19251321405172347,
      "num_tokens": 9035099.0,
      "step": 3935
    },
    {
      "entropy": 5.559963321685791,
      "epoch": 0.37848222862632086,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004990981630469119,
      "loss": 5.3502,
      "mean_token_accuracy": 0.19000618606805803,
      "num_tokens": 9045675.0,
      "step": 3940
    },
    {
      "entropy": 5.567147064208984,
      "epoch": 0.37896253602305474,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004990950939806323,
      "loss": 5.3301,
      "mean_token_accuracy": 0.1966022342443466,
      "num_tokens": 9057133.0,
      "step": 3945
    },
    {
      "entropy": 5.543869400024414,
      "epoch": 0.3794428434197887,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00049909201971153,
      "loss": 5.3182,
      "mean_token_accuracy": 0.19737419933080674,
      "num_tokens": 9067806.0,
      "step": 3950
    },
    {
      "entropy": 5.42619218826294,
      "epoch": 0.37992315081652256,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004990889402396763,
      "loss": 5.3451,
      "mean_token_accuracy": 0.18544983863830566,
      "num_tokens": 9079893.0,
      "step": 3955
    },
    {
      "entropy": 5.554620456695557,
      "epoch": 0.3804034582132565,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004990858555651431,
      "loss": 5.3677,
      "mean_token_accuracy": 0.1892174646258354,
      "num_tokens": 9091583.0,
      "step": 3960
    },
    {
      "entropy": 5.580472564697265,
      "epoch": 0.3808837656099904,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004990827656880015,
      "loss": 5.3692,
      "mean_token_accuracy": 0.19650019258260726,
      "num_tokens": 9102919.0,
      "step": 3965
    },
    {
      "entropy": 5.542777109146118,
      "epoch": 0.3813640730067243,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004990796706083235,
      "loss": 5.3396,
      "mean_token_accuracy": 0.1959129184484482,
      "num_tokens": 9113876.0,
      "step": 3970
    },
    {
      "entropy": 5.593253469467163,
      "epoch": 0.3818443804034582,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004990765703261809,
      "loss": 5.4932,
      "mean_token_accuracy": 0.17902480214834213,
      "num_tokens": 9125630.0,
      "step": 3975
    },
    {
      "entropy": 5.5642547607421875,
      "epoch": 0.38232468780019213,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004990734648416458,
      "loss": 5.4176,
      "mean_token_accuracy": 0.19356685727834702,
      "num_tokens": 9137302.0,
      "step": 3980
    },
    {
      "entropy": 5.564790391921997,
      "epoch": 0.382804995196926,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990703541547901,
      "loss": 5.3705,
      "mean_token_accuracy": 0.19759535640478135,
      "num_tokens": 9148048.0,
      "step": 3985
    },
    {
      "entropy": 5.598634147644043,
      "epoch": 0.38328530259365995,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004990672382656863,
      "loss": 5.4333,
      "mean_token_accuracy": 0.18877681642770766,
      "num_tokens": 9158656.0,
      "step": 3990
    },
    {
      "entropy": 5.583253717422485,
      "epoch": 0.38376560999039383,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004990641171744064,
      "loss": 5.3099,
      "mean_token_accuracy": 0.19103344529867172,
      "num_tokens": 9169324.0,
      "step": 3995
    },
    {
      "entropy": 5.45250563621521,
      "epoch": 0.38424591738712777,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004990609908810231,
      "loss": 5.2751,
      "mean_token_accuracy": 0.19656312167644502,
      "num_tokens": 9180238.0,
      "step": 4000
    },
    {
      "entropy": 5.407513284683228,
      "epoch": 0.38472622478386165,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004990578593856089,
      "loss": 5.1374,
      "mean_token_accuracy": 0.20177774280309677,
      "num_tokens": 9190766.0,
      "step": 4005
    },
    {
      "entropy": 5.489398097991943,
      "epoch": 0.3852065321805956,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004990547226882366,
      "loss": 5.307,
      "mean_token_accuracy": 0.19777344465255736,
      "num_tokens": 9202445.0,
      "step": 4010
    },
    {
      "entropy": 5.535893154144287,
      "epoch": 0.38568683957732947,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004990515807889788,
      "loss": 5.3869,
      "mean_token_accuracy": 0.1935746118426323,
      "num_tokens": 9212383.0,
      "step": 4015
    },
    {
      "entropy": 5.494548654556274,
      "epoch": 0.3861671469740634,
      "grad_norm": 1.0,
      "learning_rate": 0.0004990484336879087,
      "loss": 5.2896,
      "mean_token_accuracy": 0.19888617098331451,
      "num_tokens": 9223619.0,
      "step": 4020
    },
    {
      "entropy": 5.55720591545105,
      "epoch": 0.3866474543707973,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004990452813850992,
      "loss": 5.4291,
      "mean_token_accuracy": 0.1925532728433609,
      "num_tokens": 9234137.0,
      "step": 4025
    },
    {
      "entropy": 5.609319067001342,
      "epoch": 0.3871277617675312,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004990421238806236,
      "loss": 5.3548,
      "mean_token_accuracy": 0.19390372633934022,
      "num_tokens": 9246093.0,
      "step": 4030
    },
    {
      "entropy": 5.585018253326416,
      "epoch": 0.38760806916426516,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004990389611745551,
      "loss": 5.4478,
      "mean_token_accuracy": 0.18219424933195114,
      "num_tokens": 9258021.0,
      "step": 4035
    },
    {
      "entropy": 5.569950771331787,
      "epoch": 0.38808837656099904,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004990357932669672,
      "loss": 5.4922,
      "mean_token_accuracy": 0.191328227519989,
      "num_tokens": 9267668.0,
      "step": 4040
    },
    {
      "entropy": 5.5912493705749515,
      "epoch": 0.388568683957733,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004990326201579335,
      "loss": 5.3887,
      "mean_token_accuracy": 0.18971785604953767,
      "num_tokens": 9279381.0,
      "step": 4045
    },
    {
      "entropy": 5.55603551864624,
      "epoch": 0.38904899135446686,
      "grad_norm": 0.875,
      "learning_rate": 0.0004990294418475274,
      "loss": 5.4724,
      "mean_token_accuracy": 0.1800238832831383,
      "num_tokens": 9292578.0,
      "step": 4050
    },
    {
      "entropy": 5.563618135452271,
      "epoch": 0.3895292987512008,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004990262583358231,
      "loss": 5.251,
      "mean_token_accuracy": 0.19234683662652968,
      "num_tokens": 9304116.0,
      "step": 4055
    },
    {
      "entropy": 5.583739519119263,
      "epoch": 0.3900096061479347,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004990230696228943,
      "loss": 5.4749,
      "mean_token_accuracy": 0.18630178272724152,
      "num_tokens": 9315877.0,
      "step": 4060
    },
    {
      "entropy": 5.529163599014282,
      "epoch": 0.3904899135446686,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004990198757088149,
      "loss": 5.3569,
      "mean_token_accuracy": 0.195740707218647,
      "num_tokens": 9326859.0,
      "step": 4065
    },
    {
      "entropy": 5.593496370315552,
      "epoch": 0.3909702209414025,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004990166765936593,
      "loss": 5.4358,
      "mean_token_accuracy": 0.1861550435423851,
      "num_tokens": 9338067.0,
      "step": 4070
    },
    {
      "entropy": 5.554104518890381,
      "epoch": 0.3914505283381364,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004990134722775016,
      "loss": 5.4741,
      "mean_token_accuracy": 0.18901266455650328,
      "num_tokens": 9349564.0,
      "step": 4075
    },
    {
      "entropy": 5.485367202758789,
      "epoch": 0.3919308357348703,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004990102627604162,
      "loss": 5.2697,
      "mean_token_accuracy": 0.1950869917869568,
      "num_tokens": 9359962.0,
      "step": 4080
    },
    {
      "entropy": 5.491148948669434,
      "epoch": 0.39241114313160425,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004990070480424778,
      "loss": 5.2919,
      "mean_token_accuracy": 0.1947808250784874,
      "num_tokens": 9370251.0,
      "step": 4085
    },
    {
      "entropy": 5.539888000488281,
      "epoch": 0.3928914505283381,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004990038281237608,
      "loss": 5.407,
      "mean_token_accuracy": 0.18602183014154433,
      "num_tokens": 9381848.0,
      "step": 4090
    },
    {
      "entropy": 5.485216188430786,
      "epoch": 0.39337175792507206,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004990006030043401,
      "loss": 5.3329,
      "mean_token_accuracy": 0.1948874056339264,
      "num_tokens": 9392475.0,
      "step": 4095
    },
    {
      "entropy": 5.599017238616943,
      "epoch": 0.39385206532180594,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004989973726842906,
      "loss": 5.4511,
      "mean_token_accuracy": 0.18498369604349135,
      "num_tokens": 9404014.0,
      "step": 4100
    },
    {
      "entropy": 5.545511054992676,
      "epoch": 0.3943323727185399,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004989941371636872,
      "loss": 5.3304,
      "mean_token_accuracy": 0.191384394466877,
      "num_tokens": 9415590.0,
      "step": 4105
    },
    {
      "entropy": 5.587719488143921,
      "epoch": 0.39481268011527376,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004989908964426051,
      "loss": 5.4889,
      "mean_token_accuracy": 0.1834881380200386,
      "num_tokens": 9427486.0,
      "step": 4110
    },
    {
      "entropy": 5.5759295463562015,
      "epoch": 0.3952929875120077,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004989876505211194,
      "loss": 5.364,
      "mean_token_accuracy": 0.19686854779720306,
      "num_tokens": 9438212.0,
      "step": 4115
    },
    {
      "entropy": 5.517956018447876,
      "epoch": 0.3957732949087416,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004989843993993056,
      "loss": 5.3204,
      "mean_token_accuracy": 0.20115900039672852,
      "num_tokens": 9450922.0,
      "step": 4120
    },
    {
      "entropy": 5.466002607345581,
      "epoch": 0.3962536023054755,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004989811430772392,
      "loss": 5.2866,
      "mean_token_accuracy": 0.1922501415014267,
      "num_tokens": 9461676.0,
      "step": 4125
    },
    {
      "entropy": 5.38816385269165,
      "epoch": 0.3967339097022094,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004989778815549957,
      "loss": 5.2534,
      "mean_token_accuracy": 0.2010972335934639,
      "num_tokens": 9473517.0,
      "step": 4130
    },
    {
      "entropy": 5.575114965438843,
      "epoch": 0.39721421709894333,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004989746148326508,
      "loss": 5.387,
      "mean_token_accuracy": 0.18667260110378264,
      "num_tokens": 9485660.0,
      "step": 4135
    },
    {
      "entropy": 5.585870838165283,
      "epoch": 0.3976945244956772,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004989713429102805,
      "loss": 5.4057,
      "mean_token_accuracy": 0.19433852434158325,
      "num_tokens": 9495983.0,
      "step": 4140
    },
    {
      "entropy": 5.535733604431153,
      "epoch": 0.39817483189241115,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004989680657879607,
      "loss": 5.3528,
      "mean_token_accuracy": 0.18286803513765335,
      "num_tokens": 9507438.0,
      "step": 4145
    },
    {
      "entropy": 5.441587018966675,
      "epoch": 0.39865513928914503,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004989647834657675,
      "loss": 5.2582,
      "mean_token_accuracy": 0.1978207230567932,
      "num_tokens": 9517508.0,
      "step": 4150
    },
    {
      "entropy": 5.486960077285767,
      "epoch": 0.39913544668587897,
      "grad_norm": 0.890625,
      "learning_rate": 0.000498961495943777,
      "loss": 5.2811,
      "mean_token_accuracy": 0.1986250624060631,
      "num_tokens": 9527641.0,
      "step": 4155
    },
    {
      "entropy": 5.5311973094940186,
      "epoch": 0.39961575408261285,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004989582032220656,
      "loss": 5.3698,
      "mean_token_accuracy": 0.19058652222156525,
      "num_tokens": 9538762.0,
      "step": 4160
    },
    {
      "entropy": 5.5386857986450195,
      "epoch": 0.4000960614793468,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004989549053007096,
      "loss": 5.3918,
      "mean_token_accuracy": 0.18760392665863038,
      "num_tokens": 9550246.0,
      "step": 4165
    },
    {
      "entropy": 5.560341119766235,
      "epoch": 0.40057636887608067,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004989516021797858,
      "loss": 5.3257,
      "mean_token_accuracy": 0.19101699441671371,
      "num_tokens": 9562341.0,
      "step": 4170
    },
    {
      "entropy": 5.551169538497925,
      "epoch": 0.4010566762728146,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000498948293859371,
      "loss": 5.4136,
      "mean_token_accuracy": 0.18860781490802764,
      "num_tokens": 9574484.0,
      "step": 4175
    },
    {
      "entropy": 5.568818998336792,
      "epoch": 0.4015369836695485,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004989449803395415,
      "loss": 5.3669,
      "mean_token_accuracy": 0.19462387561798095,
      "num_tokens": 9585241.0,
      "step": 4180
    },
    {
      "entropy": 5.497219467163086,
      "epoch": 0.4020172910662824,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004989416616203747,
      "loss": 5.3726,
      "mean_token_accuracy": 0.18923058211803437,
      "num_tokens": 9598683.0,
      "step": 4185
    },
    {
      "entropy": 5.539427614212036,
      "epoch": 0.4024975984630163,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004989383377019476,
      "loss": 5.3849,
      "mean_token_accuracy": 0.19240100532770157,
      "num_tokens": 9610127.0,
      "step": 4190
    },
    {
      "entropy": 5.5731137752532955,
      "epoch": 0.40297790585975024,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004989350085843371,
      "loss": 5.2965,
      "mean_token_accuracy": 0.19976160675287247,
      "num_tokens": 9620009.0,
      "step": 4195
    },
    {
      "entropy": 5.477526235580444,
      "epoch": 0.4034582132564842,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004989316742676207,
      "loss": 5.3905,
      "mean_token_accuracy": 0.19826774597167968,
      "num_tokens": 9632479.0,
      "step": 4200
    },
    {
      "entropy": 5.595583438873291,
      "epoch": 0.40393852065321806,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004989283347518757,
      "loss": 5.4474,
      "mean_token_accuracy": 0.19625402688980104,
      "num_tokens": 9644447.0,
      "step": 4205
    },
    {
      "entropy": 5.559397220611572,
      "epoch": 0.404418828049952,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004989249900371797,
      "loss": 5.3427,
      "mean_token_accuracy": 0.1875220090150833,
      "num_tokens": 9656886.0,
      "step": 4210
    },
    {
      "entropy": 5.5763678550720215,
      "epoch": 0.4048991354466859,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004989216401236103,
      "loss": 5.392,
      "mean_token_accuracy": 0.19164305329322814,
      "num_tokens": 9667079.0,
      "step": 4215
    },
    {
      "entropy": 5.600238513946533,
      "epoch": 0.4053794428434198,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004989182850112455,
      "loss": 5.4316,
      "mean_token_accuracy": 0.19006281048059465,
      "num_tokens": 9678237.0,
      "step": 4220
    },
    {
      "entropy": 5.487116909027099,
      "epoch": 0.4058597502401537,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004989149247001629,
      "loss": 5.2881,
      "mean_token_accuracy": 0.1941922962665558,
      "num_tokens": 9688809.0,
      "step": 4225
    },
    {
      "entropy": 5.612413787841797,
      "epoch": 0.40634005763688763,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004989115591904407,
      "loss": 5.3921,
      "mean_token_accuracy": 0.19743535816669464,
      "num_tokens": 9700460.0,
      "step": 4230
    },
    {
      "entropy": 5.4953470706939695,
      "epoch": 0.4068203650336215,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004989081884821569,
      "loss": 5.3272,
      "mean_token_accuracy": 0.19113785326480864,
      "num_tokens": 9711379.0,
      "step": 4235
    },
    {
      "entropy": 5.471403121948242,
      "epoch": 0.40730067243035545,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004989048125753899,
      "loss": 5.2663,
      "mean_token_accuracy": 0.19734414517879487,
      "num_tokens": 9722792.0,
      "step": 4240
    },
    {
      "entropy": 5.5470785140991214,
      "epoch": 0.40778097982708933,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000498901431470218,
      "loss": 5.3118,
      "mean_token_accuracy": 0.19861934781074525,
      "num_tokens": 9733524.0,
      "step": 4245
    },
    {
      "entropy": 5.586074495315552,
      "epoch": 0.40826128722382327,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004988980451667198,
      "loss": 5.3721,
      "mean_token_accuracy": 0.1926126554608345,
      "num_tokens": 9745430.0,
      "step": 4250
    },
    {
      "entropy": 5.494037771224976,
      "epoch": 0.40874159462055715,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004988946536649737,
      "loss": 5.2897,
      "mean_token_accuracy": 0.19434840828180314,
      "num_tokens": 9756927.0,
      "step": 4255
    },
    {
      "entropy": 5.53960952758789,
      "epoch": 0.4092219020172911,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004988912569650585,
      "loss": 5.321,
      "mean_token_accuracy": 0.19842132329940795,
      "num_tokens": 9768350.0,
      "step": 4260
    },
    {
      "entropy": 5.471000623703003,
      "epoch": 0.40970220941402496,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004988878550670533,
      "loss": 5.2705,
      "mean_token_accuracy": 0.20159802734851837,
      "num_tokens": 9779238.0,
      "step": 4265
    },
    {
      "entropy": 5.5897942066192625,
      "epoch": 0.4101825168107589,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004988844479710369,
      "loss": 5.4434,
      "mean_token_accuracy": 0.1883348345756531,
      "num_tokens": 9789262.0,
      "step": 4270
    },
    {
      "entropy": 5.611835670471192,
      "epoch": 0.4106628242074928,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004988810356770884,
      "loss": 5.4081,
      "mean_token_accuracy": 0.19093504697084426,
      "num_tokens": 9800909.0,
      "step": 4275
    },
    {
      "entropy": 5.531945133209229,
      "epoch": 0.4111431316042267,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000498877618185287,
      "loss": 5.3718,
      "mean_token_accuracy": 0.19317576587200164,
      "num_tokens": 9812538.0,
      "step": 4280
    },
    {
      "entropy": 5.577809858322143,
      "epoch": 0.4116234390009606,
      "grad_norm": 1.0,
      "learning_rate": 0.0004988741954957121,
      "loss": 5.357,
      "mean_token_accuracy": 0.19297734797000884,
      "num_tokens": 9824071.0,
      "step": 4285
    },
    {
      "entropy": 5.627749490737915,
      "epoch": 0.41210374639769454,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004988707676084432,
      "loss": 5.4586,
      "mean_token_accuracy": 0.18716205060482025,
      "num_tokens": 9836713.0,
      "step": 4290
    },
    {
      "entropy": 5.476572704315186,
      "epoch": 0.4125840537944284,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004988673345235597,
      "loss": 5.2915,
      "mean_token_accuracy": 0.19582398235797882,
      "num_tokens": 9847336.0,
      "step": 4295
    },
    {
      "entropy": 5.457479047775268,
      "epoch": 0.41306436119116235,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004988638962411416,
      "loss": 5.2992,
      "mean_token_accuracy": 0.19223696291446685,
      "num_tokens": 9858567.0,
      "step": 4300
    },
    {
      "entropy": 5.599784708023071,
      "epoch": 0.41354466858789624,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004988604527612685,
      "loss": 5.3744,
      "mean_token_accuracy": 0.1840699717402458,
      "num_tokens": 9869254.0,
      "step": 4305
    },
    {
      "entropy": 5.5117292404174805,
      "epoch": 0.4140249759846302,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004988570040840205,
      "loss": 5.313,
      "mean_token_accuracy": 0.19747684746980668,
      "num_tokens": 9879635.0,
      "step": 4310
    },
    {
      "entropy": 5.527637577056884,
      "epoch": 0.41450528338136405,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004988535502094774,
      "loss": 5.3703,
      "mean_token_accuracy": 0.19628743827342987,
      "num_tokens": 9891728.0,
      "step": 4315
    },
    {
      "entropy": 5.458841991424561,
      "epoch": 0.414985590778098,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004988500911377198,
      "loss": 5.1866,
      "mean_token_accuracy": 0.19835380762815474,
      "num_tokens": 9903140.0,
      "step": 4320
    },
    {
      "entropy": 5.491971206665039,
      "epoch": 0.41546589817483187,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004988466268688276,
      "loss": 5.3201,
      "mean_token_accuracy": 0.19649144262075424,
      "num_tokens": 9915562.0,
      "step": 4325
    },
    {
      "entropy": 5.515983200073242,
      "epoch": 0.4159462055715658,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004988431574028814,
      "loss": 5.2756,
      "mean_token_accuracy": 0.18987508863210678,
      "num_tokens": 9925720.0,
      "step": 4330
    },
    {
      "entropy": 5.537591171264649,
      "epoch": 0.4164265129682997,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004988396827399618,
      "loss": 5.3474,
      "mean_token_accuracy": 0.1914285972714424,
      "num_tokens": 9936492.0,
      "step": 4335
    },
    {
      "entropy": 5.46312985420227,
      "epoch": 0.4169068203650336,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004988362028801495,
      "loss": 5.2899,
      "mean_token_accuracy": 0.19118731766939162,
      "num_tokens": 9948658.0,
      "step": 4340
    },
    {
      "entropy": 5.383918571472168,
      "epoch": 0.4173871277617675,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004988327178235253,
      "loss": 5.2456,
      "mean_token_accuracy": 0.20390928983688356,
      "num_tokens": 9959710.0,
      "step": 4345
    },
    {
      "entropy": 5.509265470504761,
      "epoch": 0.41786743515850144,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004988292275701699,
      "loss": 5.2379,
      "mean_token_accuracy": 0.19945309311151505,
      "num_tokens": 9970302.0,
      "step": 4350
    },
    {
      "entropy": 5.521309900283813,
      "epoch": 0.4183477425552353,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004988257321201646,
      "loss": 5.3504,
      "mean_token_accuracy": 0.19332970678806305,
      "num_tokens": 9981776.0,
      "step": 4355
    },
    {
      "entropy": 5.524739694595337,
      "epoch": 0.41882804995196926,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004988222314735902,
      "loss": 5.3316,
      "mean_token_accuracy": 0.19160499274730683,
      "num_tokens": 9993471.0,
      "step": 4360
    },
    {
      "entropy": 5.5496828079223635,
      "epoch": 0.41930835734870314,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004988187256305284,
      "loss": 5.3535,
      "mean_token_accuracy": 0.19131845980882645,
      "num_tokens": 10003818.0,
      "step": 4365
    },
    {
      "entropy": 5.520091152191162,
      "epoch": 0.4197886647454371,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004988152145910603,
      "loss": 5.419,
      "mean_token_accuracy": 0.18176987618207932,
      "num_tokens": 10015507.0,
      "step": 4370
    },
    {
      "entropy": 5.471072483062744,
      "epoch": 0.420268972142171,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004988116983552675,
      "loss": 5.2562,
      "mean_token_accuracy": 0.19782975167036057,
      "num_tokens": 10026994.0,
      "step": 4375
    },
    {
      "entropy": 5.565588188171387,
      "epoch": 0.4207492795389049,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004988081769232317,
      "loss": 5.4051,
      "mean_token_accuracy": 0.1849276453256607,
      "num_tokens": 10039547.0,
      "step": 4380
    },
    {
      "entropy": 5.545499515533447,
      "epoch": 0.42122958693563883,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004988046502950346,
      "loss": 5.4261,
      "mean_token_accuracy": 0.18821370005607604,
      "num_tokens": 10050568.0,
      "step": 4385
    },
    {
      "entropy": 5.482304859161377,
      "epoch": 0.4217098943323727,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000498801118470758,
      "loss": 5.3288,
      "mean_token_accuracy": 0.19275769144296645,
      "num_tokens": 10061780.0,
      "step": 4390
    },
    {
      "entropy": 5.450534868240356,
      "epoch": 0.42219020172910665,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000498797581450484,
      "loss": 5.2912,
      "mean_token_accuracy": 0.19367838501930237,
      "num_tokens": 10073101.0,
      "step": 4395
    },
    {
      "entropy": 5.451673316955566,
      "epoch": 0.42267050912584053,
      "grad_norm": 1.0,
      "learning_rate": 0.0004987940392342948,
      "loss": 5.2355,
      "mean_token_accuracy": 0.20872799009084703,
      "num_tokens": 10084397.0,
      "step": 4400
    },
    {
      "entropy": 5.516866254806518,
      "epoch": 0.42315081652257447,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004987904918222726,
      "loss": 5.3131,
      "mean_token_accuracy": 0.19882151931524278,
      "num_tokens": 10095635.0,
      "step": 4405
    },
    {
      "entropy": 5.4873948097229,
      "epoch": 0.42363112391930835,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004987869392144996,
      "loss": 5.3735,
      "mean_token_accuracy": 0.19766258597373962,
      "num_tokens": 10107744.0,
      "step": 4410
    },
    {
      "entropy": 5.523633241653442,
      "epoch": 0.4241114313160423,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004987833814110584,
      "loss": 5.3019,
      "mean_token_accuracy": 0.19736873358488083,
      "num_tokens": 10118552.0,
      "step": 4415
    },
    {
      "entropy": 5.584984254837036,
      "epoch": 0.42459173871277617,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004987798184120316,
      "loss": 5.4329,
      "mean_token_accuracy": 0.18282995671033858,
      "num_tokens": 10130813.0,
      "step": 4420
    },
    {
      "entropy": 5.5322612762451175,
      "epoch": 0.4250720461095101,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004987762502175018,
      "loss": 5.4422,
      "mean_token_accuracy": 0.19148289561271667,
      "num_tokens": 10142196.0,
      "step": 4425
    },
    {
      "entropy": 5.487265157699585,
      "epoch": 0.425552353506244,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000498772676827552,
      "loss": 5.3357,
      "mean_token_accuracy": 0.19926347434520722,
      "num_tokens": 10155012.0,
      "step": 4430
    },
    {
      "entropy": 5.574202394485473,
      "epoch": 0.4260326609029779,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004987690982422652,
      "loss": 5.4298,
      "mean_token_accuracy": 0.1860943153500557,
      "num_tokens": 10166160.0,
      "step": 4435
    },
    {
      "entropy": 5.524494934082031,
      "epoch": 0.4265129682997118,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004987655144617243,
      "loss": 5.2143,
      "mean_token_accuracy": 0.20166747272014618,
      "num_tokens": 10177663.0,
      "step": 4440
    },
    {
      "entropy": 5.41952166557312,
      "epoch": 0.42699327569644574,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004987619254860126,
      "loss": 5.2868,
      "mean_token_accuracy": 0.20027084350585939,
      "num_tokens": 10189127.0,
      "step": 4445
    },
    {
      "entropy": 5.62891731262207,
      "epoch": 0.4274735830931796,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004987583313152134,
      "loss": 5.4641,
      "mean_token_accuracy": 0.18564036190509797,
      "num_tokens": 10200765.0,
      "step": 4450
    },
    {
      "entropy": 5.561132860183716,
      "epoch": 0.42795389048991356,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004987547319494104,
      "loss": 5.3107,
      "mean_token_accuracy": 0.20098303705453874,
      "num_tokens": 10213047.0,
      "step": 4455
    },
    {
      "entropy": 5.453295373916626,
      "epoch": 0.42843419788664744,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004987511273886867,
      "loss": 5.2818,
      "mean_token_accuracy": 0.19822690933942794,
      "num_tokens": 10223790.0,
      "step": 4460
    },
    {
      "entropy": 5.498892402648925,
      "epoch": 0.4289145052833814,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004987475176331263,
      "loss": 5.2759,
      "mean_token_accuracy": 0.19361269623041152,
      "num_tokens": 10235077.0,
      "step": 4465
    },
    {
      "entropy": 5.511125183105468,
      "epoch": 0.42939481268011526,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004987439026828129,
      "loss": 5.3443,
      "mean_token_accuracy": 0.18790692389011382,
      "num_tokens": 10248844.0,
      "step": 4470
    },
    {
      "entropy": 5.498596572875977,
      "epoch": 0.4298751200768492,
      "grad_norm": 1.0,
      "learning_rate": 0.0004987402825378305,
      "loss": 5.2707,
      "mean_token_accuracy": 0.20468346476554872,
      "num_tokens": 10259013.0,
      "step": 4475
    },
    {
      "entropy": 5.401968908309937,
      "epoch": 0.4303554274735831,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004987366571982631,
      "loss": 5.2942,
      "mean_token_accuracy": 0.1975047007203102,
      "num_tokens": 10271415.0,
      "step": 4480
    },
    {
      "entropy": 5.456958818435669,
      "epoch": 0.430835734870317,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004987330266641948,
      "loss": 5.3057,
      "mean_token_accuracy": 0.19396223574876786,
      "num_tokens": 10283204.0,
      "step": 4485
    },
    {
      "entropy": 5.642587327957154,
      "epoch": 0.4313160422670509,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004987293909357101,
      "loss": 5.4656,
      "mean_token_accuracy": 0.18648719489574433,
      "num_tokens": 10294111.0,
      "step": 4490
    },
    {
      "entropy": 5.6016778469085695,
      "epoch": 0.43179634966378483,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004987257500128933,
      "loss": 5.3711,
      "mean_token_accuracy": 0.18458185642957686,
      "num_tokens": 10305417.0,
      "step": 4495
    },
    {
      "entropy": 5.391876840591431,
      "epoch": 0.4322766570605187,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004987221038958288,
      "loss": 5.2406,
      "mean_token_accuracy": 0.1981041520833969,
      "num_tokens": 10316416.0,
      "step": 4500
    },
    {
      "entropy": 5.54496808052063,
      "epoch": 0.43275696445725265,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004987184525846015,
      "loss": 5.4128,
      "mean_token_accuracy": 0.19055870175361633,
      "num_tokens": 10328251.0,
      "step": 4505
    },
    {
      "entropy": 5.593237972259521,
      "epoch": 0.4332372718539865,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004987147960792958,
      "loss": 5.375,
      "mean_token_accuracy": 0.19966753274202348,
      "num_tokens": 10340189.0,
      "step": 4510
    },
    {
      "entropy": 5.508662605285645,
      "epoch": 0.43371757925072046,
      "grad_norm": 0.875,
      "learning_rate": 0.0004987111343799971,
      "loss": 5.3016,
      "mean_token_accuracy": 0.19588980674743653,
      "num_tokens": 10351660.0,
      "step": 4515
    },
    {
      "entropy": 5.487722015380859,
      "epoch": 0.43419788664745435,
      "grad_norm": 0.984375,
      "learning_rate": 0.00049870746748679,
      "loss": 5.3456,
      "mean_token_accuracy": 0.19546182155609132,
      "num_tokens": 10362240.0,
      "step": 4520
    },
    {
      "entropy": 5.629311752319336,
      "epoch": 0.4346781940441883,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004987037953997598,
      "loss": 5.3776,
      "mean_token_accuracy": 0.198500619828701,
      "num_tokens": 10373646.0,
      "step": 4525
    },
    {
      "entropy": 5.49867582321167,
      "epoch": 0.43515850144092216,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004987001181189918,
      "loss": 5.4002,
      "mean_token_accuracy": 0.190858456492424,
      "num_tokens": 10385540.0,
      "step": 4530
    },
    {
      "entropy": 5.516322469711303,
      "epoch": 0.4356388088376561,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004986964356445713,
      "loss": 5.3947,
      "mean_token_accuracy": 0.18782368004322053,
      "num_tokens": 10396641.0,
      "step": 4535
    },
    {
      "entropy": 5.5667399883270265,
      "epoch": 0.43611911623439004,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004986927479765837,
      "loss": 5.3527,
      "mean_token_accuracy": 0.18772210627794267,
      "num_tokens": 10408165.0,
      "step": 4540
    },
    {
      "entropy": 5.503515148162842,
      "epoch": 0.4365994236311239,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004986890551151148,
      "loss": 5.3226,
      "mean_token_accuracy": 0.1969214618206024,
      "num_tokens": 10419646.0,
      "step": 4545
    },
    {
      "entropy": 5.521216773986817,
      "epoch": 0.43707973102785785,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004986853570602503,
      "loss": 5.3317,
      "mean_token_accuracy": 0.19431255608797074,
      "num_tokens": 10431051.0,
      "step": 4550
    },
    {
      "entropy": 5.59027943611145,
      "epoch": 0.43756003842459174,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004986816538120758,
      "loss": 5.3985,
      "mean_token_accuracy": 0.1879848912358284,
      "num_tokens": 10441833.0,
      "step": 4555
    },
    {
      "entropy": 5.534223651885986,
      "epoch": 0.43804034582132567,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004986779453706778,
      "loss": 5.3622,
      "mean_token_accuracy": 0.1911931961774826,
      "num_tokens": 10453544.0,
      "step": 4560
    },
    {
      "entropy": 5.4847815990447994,
      "epoch": 0.43852065321805955,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004986742317361419,
      "loss": 5.2157,
      "mean_token_accuracy": 0.20285268127918243,
      "num_tokens": 10464140.0,
      "step": 4565
    },
    {
      "entropy": 5.375190258026123,
      "epoch": 0.4390009606147935,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004986705129085546,
      "loss": 5.2747,
      "mean_token_accuracy": 0.2015322670340538,
      "num_tokens": 10475739.0,
      "step": 4570
    },
    {
      "entropy": 5.4098118305206295,
      "epoch": 0.43948126801152737,
      "grad_norm": 0.875,
      "learning_rate": 0.0004986667888880021,
      "loss": 5.3202,
      "mean_token_accuracy": 0.20002770870923997,
      "num_tokens": 10487451.0,
      "step": 4575
    },
    {
      "entropy": 5.557822608947754,
      "epoch": 0.4399615754082613,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004986630596745709,
      "loss": 5.3687,
      "mean_token_accuracy": 0.18746123164892198,
      "num_tokens": 10499367.0,
      "step": 4580
    },
    {
      "entropy": 5.498035621643067,
      "epoch": 0.4404418828049952,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004986593252683477,
      "loss": 5.181,
      "mean_token_accuracy": 0.2056640088558197,
      "num_tokens": 10511192.0,
      "step": 4585
    },
    {
      "entropy": 5.362164258956909,
      "epoch": 0.4409221902017291,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004986555856694191,
      "loss": 5.2248,
      "mean_token_accuracy": 0.19351800680160522,
      "num_tokens": 10522754.0,
      "step": 4590
    },
    {
      "entropy": 5.533795595169067,
      "epoch": 0.441402497598463,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004986518408778718,
      "loss": 5.3597,
      "mean_token_accuracy": 0.18763887286186218,
      "num_tokens": 10533439.0,
      "step": 4595
    },
    {
      "entropy": 5.594131183624268,
      "epoch": 0.44188280499519694,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004986480908937929,
      "loss": 5.3193,
      "mean_token_accuracy": 0.19278666824102403,
      "num_tokens": 10543654.0,
      "step": 4600
    },
    {
      "entropy": 5.408310222625732,
      "epoch": 0.4423631123919308,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004986443357172695,
      "loss": 5.2346,
      "mean_token_accuracy": 0.19730304181575775,
      "num_tokens": 10554936.0,
      "step": 4605
    },
    {
      "entropy": 5.482907390594482,
      "epoch": 0.44284341978866476,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004986405753483887,
      "loss": 5.2963,
      "mean_token_accuracy": 0.1938473716378212,
      "num_tokens": 10566410.0,
      "step": 4610
    },
    {
      "entropy": 5.4643957138061525,
      "epoch": 0.44332372718539864,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004986368097872377,
      "loss": 5.2889,
      "mean_token_accuracy": 0.2042735531926155,
      "num_tokens": 10577258.0,
      "step": 4615
    },
    {
      "entropy": 5.512962102890015,
      "epoch": 0.4438040345821326,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004986330390339042,
      "loss": 5.3965,
      "mean_token_accuracy": 0.1925565704703331,
      "num_tokens": 10588783.0,
      "step": 4620
    },
    {
      "entropy": 5.4776952266693115,
      "epoch": 0.44428434197886646,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004986292630884755,
      "loss": 5.2651,
      "mean_token_accuracy": 0.19924047887325286,
      "num_tokens": 10599778.0,
      "step": 4625
    },
    {
      "entropy": 5.492894506454467,
      "epoch": 0.4447646493756004,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004986254819510393,
      "loss": 5.2996,
      "mean_token_accuracy": 0.19947718679904938,
      "num_tokens": 10611211.0,
      "step": 4630
    },
    {
      "entropy": 5.411989831924439,
      "epoch": 0.4452449567723343,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004986216956216835,
      "loss": 5.2261,
      "mean_token_accuracy": 0.19829935133457183,
      "num_tokens": 10624535.0,
      "step": 4635
    },
    {
      "entropy": 5.436934947967529,
      "epoch": 0.4457252641690682,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000498617904100496,
      "loss": 5.2762,
      "mean_token_accuracy": 0.20355055034160613,
      "num_tokens": 10635206.0,
      "step": 4640
    },
    {
      "entropy": 5.509118270874024,
      "epoch": 0.4462055715658021,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004986141073875646,
      "loss": 5.3176,
      "mean_token_accuracy": 0.20377421677112578,
      "num_tokens": 10646993.0,
      "step": 4645
    },
    {
      "entropy": 5.483718061447144,
      "epoch": 0.44668587896253603,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004986103054829779,
      "loss": 5.255,
      "mean_token_accuracy": 0.20013512223958968,
      "num_tokens": 10657262.0,
      "step": 4650
    },
    {
      "entropy": 5.433026313781738,
      "epoch": 0.4471661863592699,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004986064983868237,
      "loss": 5.2956,
      "mean_token_accuracy": 0.19394896775484086,
      "num_tokens": 10668962.0,
      "step": 4655
    },
    {
      "entropy": 5.498749923706055,
      "epoch": 0.44764649375600385,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004986026860991906,
      "loss": 5.3024,
      "mean_token_accuracy": 0.2030349165201187,
      "num_tokens": 10680183.0,
      "step": 4660
    },
    {
      "entropy": 5.444003534317017,
      "epoch": 0.44812680115273773,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004985988686201672,
      "loss": 5.248,
      "mean_token_accuracy": 0.2050817146897316,
      "num_tokens": 10691221.0,
      "step": 4665
    },
    {
      "entropy": 5.445614910125732,
      "epoch": 0.44860710854947167,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004985950459498419,
      "loss": 5.2737,
      "mean_token_accuracy": 0.19548940360546113,
      "num_tokens": 10703462.0,
      "step": 4670
    },
    {
      "entropy": 5.472289371490478,
      "epoch": 0.44908741594620555,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004985912180883037,
      "loss": 5.2579,
      "mean_token_accuracy": 0.1996539682149887,
      "num_tokens": 10713774.0,
      "step": 4675
    },
    {
      "entropy": 5.542710828781128,
      "epoch": 0.4495677233429395,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004985873850356411,
      "loss": 5.4542,
      "mean_token_accuracy": 0.18273407071828843,
      "num_tokens": 10725109.0,
      "step": 4680
    },
    {
      "entropy": 5.487839603424073,
      "epoch": 0.45004803073967337,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004985835467919436,
      "loss": 5.276,
      "mean_token_accuracy": 0.2000733196735382,
      "num_tokens": 10735695.0,
      "step": 4685
    },
    {
      "entropy": 5.489776182174682,
      "epoch": 0.4505283381364073,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004985797033572999,
      "loss": 5.3495,
      "mean_token_accuracy": 0.1979081153869629,
      "num_tokens": 10747554.0,
      "step": 4690
    },
    {
      "entropy": 5.552037048339844,
      "epoch": 0.4510086455331412,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004985758547317994,
      "loss": 5.3594,
      "mean_token_accuracy": 0.1919667512178421,
      "num_tokens": 10759175.0,
      "step": 4695
    },
    {
      "entropy": 5.449206447601318,
      "epoch": 0.4514889529298751,
      "grad_norm": 1.0,
      "learning_rate": 0.0004985720009155315,
      "loss": 5.2289,
      "mean_token_accuracy": 0.20362470149993897,
      "num_tokens": 10770365.0,
      "step": 4700
    },
    {
      "entropy": 5.4378249168396,
      "epoch": 0.45196926032660906,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004985681419085856,
      "loss": 5.3269,
      "mean_token_accuracy": 0.19245791286230088,
      "num_tokens": 10783096.0,
      "step": 4705
    },
    {
      "entropy": 5.597962427139282,
      "epoch": 0.45244956772334294,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004985642777110513,
      "loss": 5.4144,
      "mean_token_accuracy": 0.19019332826137542,
      "num_tokens": 10795312.0,
      "step": 4710
    },
    {
      "entropy": 5.46242504119873,
      "epoch": 0.4529298751200769,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004985604083230183,
      "loss": 5.2484,
      "mean_token_accuracy": 0.198370298743248,
      "num_tokens": 10806451.0,
      "step": 4715
    },
    {
      "entropy": 5.525980758666992,
      "epoch": 0.45341018251681076,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004985565337445765,
      "loss": 5.39,
      "mean_token_accuracy": 0.19138407260179519,
      "num_tokens": 10818185.0,
      "step": 4720
    },
    {
      "entropy": 5.5021544933319095,
      "epoch": 0.4538904899135447,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004985526539758158,
      "loss": 5.256,
      "mean_token_accuracy": 0.20659695863723754,
      "num_tokens": 10830203.0,
      "step": 4725
    },
    {
      "entropy": 5.45419716835022,
      "epoch": 0.4543707973102786,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004985487690168263,
      "loss": 5.2978,
      "mean_token_accuracy": 0.1879442274570465,
      "num_tokens": 10842953.0,
      "step": 4730
    },
    {
      "entropy": 5.614833831787109,
      "epoch": 0.4548511047070125,
      "grad_norm": 1.03125,
      "learning_rate": 0.000498544878867698,
      "loss": 5.4491,
      "mean_token_accuracy": 0.1875968888401985,
      "num_tokens": 10853322.0,
      "step": 4735
    },
    {
      "entropy": 5.57782301902771,
      "epoch": 0.4553314121037464,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004985409835285215,
      "loss": 5.3421,
      "mean_token_accuracy": 0.19194249510765077,
      "num_tokens": 10865480.0,
      "step": 4740
    },
    {
      "entropy": 5.419948720932007,
      "epoch": 0.45581171950048033,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004985370829993873,
      "loss": 5.2021,
      "mean_token_accuracy": 0.20466244518756865,
      "num_tokens": 10875700.0,
      "step": 4745
    },
    {
      "entropy": 5.412119626998901,
      "epoch": 0.4562920268972142,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004985331772803857,
      "loss": 5.2999,
      "mean_token_accuracy": 0.19970083683729173,
      "num_tokens": 10887350.0,
      "step": 4750
    },
    {
      "entropy": 5.500506734848022,
      "epoch": 0.45677233429394815,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004985292663716074,
      "loss": 5.2391,
      "mean_token_accuracy": 0.19918713420629502,
      "num_tokens": 10899355.0,
      "step": 4755
    },
    {
      "entropy": 5.450326538085937,
      "epoch": 0.457252641690682,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004985253502731435,
      "loss": 5.2538,
      "mean_token_accuracy": 0.20215147137641906,
      "num_tokens": 10911521.0,
      "step": 4760
    },
    {
      "entropy": 5.425291919708252,
      "epoch": 0.45773294908741596,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004985214289850845,
      "loss": 5.32,
      "mean_token_accuracy": 0.1956801801919937,
      "num_tokens": 10923363.0,
      "step": 4765
    },
    {
      "entropy": 5.500451278686524,
      "epoch": 0.45821325648414984,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004985175025075217,
      "loss": 5.3109,
      "mean_token_accuracy": 0.19380827397108077,
      "num_tokens": 10934965.0,
      "step": 4770
    },
    {
      "entropy": 5.484968948364258,
      "epoch": 0.4586935638808838,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004985135708405462,
      "loss": 5.3108,
      "mean_token_accuracy": 0.19206822514533997,
      "num_tokens": 10946126.0,
      "step": 4775
    },
    {
      "entropy": 5.506540966033936,
      "epoch": 0.45917387127761766,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004985096339842493,
      "loss": 5.379,
      "mean_token_accuracy": 0.19345731139183045,
      "num_tokens": 10957667.0,
      "step": 4780
    },
    {
      "entropy": 5.626012229919434,
      "epoch": 0.4596541786743516,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004985056919387224,
      "loss": 5.4613,
      "mean_token_accuracy": 0.18610572963953018,
      "num_tokens": 10969999.0,
      "step": 4785
    },
    {
      "entropy": 5.454364395141601,
      "epoch": 0.4601344860710855,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004985017447040569,
      "loss": 5.1668,
      "mean_token_accuracy": 0.20562463253736496,
      "num_tokens": 10980672.0,
      "step": 4790
    },
    {
      "entropy": 5.4818586826324465,
      "epoch": 0.4606147934678194,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004984977922803447,
      "loss": 5.2782,
      "mean_token_accuracy": 0.1968653455376625,
      "num_tokens": 10991471.0,
      "step": 4795
    },
    {
      "entropy": 5.480503225326538,
      "epoch": 0.4610951008645533,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004984938346676772,
      "loss": 5.3803,
      "mean_token_accuracy": 0.19473806768655777,
      "num_tokens": 11001960.0,
      "step": 4800
    },
    {
      "entropy": 5.521226167678833,
      "epoch": 0.46157540826128723,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004984898718661468,
      "loss": 5.2378,
      "mean_token_accuracy": 0.19595222771167756,
      "num_tokens": 11013274.0,
      "step": 4805
    },
    {
      "entropy": 5.480836486816406,
      "epoch": 0.4620557156580211,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004984859038758451,
      "loss": 5.244,
      "mean_token_accuracy": 0.20800837874412537,
      "num_tokens": 11024635.0,
      "step": 4810
    },
    {
      "entropy": 5.458645820617676,
      "epoch": 0.46253602305475505,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004984819306968642,
      "loss": 5.28,
      "mean_token_accuracy": 0.19557492882013322,
      "num_tokens": 11036317.0,
      "step": 4815
    },
    {
      "entropy": 5.4449347972869875,
      "epoch": 0.46301633045148893,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004984779523292966,
      "loss": 5.2632,
      "mean_token_accuracy": 0.20042200684547423,
      "num_tokens": 11047776.0,
      "step": 4820
    },
    {
      "entropy": 5.497167348861694,
      "epoch": 0.46349663784822287,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004984739687732345,
      "loss": 5.2918,
      "mean_token_accuracy": 0.19992978274822235,
      "num_tokens": 11058746.0,
      "step": 4825
    },
    {
      "entropy": 5.5446727752685545,
      "epoch": 0.46397694524495675,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004984699800287705,
      "loss": 5.3311,
      "mean_token_accuracy": 0.19627001583576204,
      "num_tokens": 11068892.0,
      "step": 4830
    },
    {
      "entropy": 5.458117532730102,
      "epoch": 0.4644572526416907,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000498465986095997,
      "loss": 5.313,
      "mean_token_accuracy": 0.19664016664028167,
      "num_tokens": 11080038.0,
      "step": 4835
    },
    {
      "entropy": 5.495731544494629,
      "epoch": 0.46493756003842457,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004984619869750069,
      "loss": 5.2649,
      "mean_token_accuracy": 0.19860966503620148,
      "num_tokens": 11091261.0,
      "step": 4840
    },
    {
      "entropy": 5.403757572174072,
      "epoch": 0.4654178674351585,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000498457982665893,
      "loss": 5.2161,
      "mean_token_accuracy": 0.20508018732070923,
      "num_tokens": 11103015.0,
      "step": 4845
    },
    {
      "entropy": 5.456387567520141,
      "epoch": 0.4658981748318924,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004984539731687483,
      "loss": 5.2,
      "mean_token_accuracy": 0.20083018988370896,
      "num_tokens": 11115075.0,
      "step": 4850
    },
    {
      "entropy": 5.49595046043396,
      "epoch": 0.4663784822286263,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004984499584836659,
      "loss": 5.1976,
      "mean_token_accuracy": 0.19300755709409714,
      "num_tokens": 11125937.0,
      "step": 4855
    },
    {
      "entropy": 5.4198315143585205,
      "epoch": 0.4668587896253602,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000498445938610739,
      "loss": 5.2865,
      "mean_token_accuracy": 0.19478206783533097,
      "num_tokens": 11136863.0,
      "step": 4860
    },
    {
      "entropy": 5.53124361038208,
      "epoch": 0.46733909702209414,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004984419135500608,
      "loss": 5.3513,
      "mean_token_accuracy": 0.18467848598957062,
      "num_tokens": 11149934.0,
      "step": 4865
    },
    {
      "entropy": 5.453778982162476,
      "epoch": 0.4678194044188281,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004984378833017249,
      "loss": 5.2836,
      "mean_token_accuracy": 0.19541754573583603,
      "num_tokens": 11162208.0,
      "step": 4870
    },
    {
      "entropy": 5.462684917449951,
      "epoch": 0.46829971181556196,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004984338478658248,
      "loss": 5.3332,
      "mean_token_accuracy": 0.1997538447380066,
      "num_tokens": 11172892.0,
      "step": 4875
    },
    {
      "entropy": 5.461723470687867,
      "epoch": 0.4687800192122959,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004984298072424542,
      "loss": 5.2795,
      "mean_token_accuracy": 0.20340183824300767,
      "num_tokens": 11184339.0,
      "step": 4880
    },
    {
      "entropy": 5.540046453475952,
      "epoch": 0.4692603266090298,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000498425761431707,
      "loss": 5.3493,
      "mean_token_accuracy": 0.19144022464752197,
      "num_tokens": 11195681.0,
      "step": 4885
    },
    {
      "entropy": 5.493838596343994,
      "epoch": 0.4697406340057637,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000498421710433677,
      "loss": 5.2603,
      "mean_token_accuracy": 0.18896204233169556,
      "num_tokens": 11207982.0,
      "step": 4890
    },
    {
      "entropy": 5.511179971694946,
      "epoch": 0.4702209414024976,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004984176542484584,
      "loss": 5.3315,
      "mean_token_accuracy": 0.19578453302383422,
      "num_tokens": 11218901.0,
      "step": 4895
    },
    {
      "entropy": 5.440368032455444,
      "epoch": 0.47070124879923153,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004984135928761452,
      "loss": 5.2989,
      "mean_token_accuracy": 0.19516911953687668,
      "num_tokens": 11230764.0,
      "step": 4900
    },
    {
      "entropy": 5.404439163208008,
      "epoch": 0.4711815561959654,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004984095263168317,
      "loss": 5.2196,
      "mean_token_accuracy": 0.2027796745300293,
      "num_tokens": 11241489.0,
      "step": 4905
    },
    {
      "entropy": 5.34543023109436,
      "epoch": 0.47166186359269935,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004984054545706124,
      "loss": 5.196,
      "mean_token_accuracy": 0.20225575864315032,
      "num_tokens": 11254500.0,
      "step": 4910
    },
    {
      "entropy": 5.533623266220093,
      "epoch": 0.47214217098943323,
      "grad_norm": 0.84375,
      "learning_rate": 0.000498401377637582,
      "loss": 5.3351,
      "mean_token_accuracy": 0.19645069986581803,
      "num_tokens": 11266235.0,
      "step": 4915
    },
    {
      "entropy": 5.520373964309693,
      "epoch": 0.47262247838616717,
      "grad_norm": 0.875,
      "learning_rate": 0.000498397295517835,
      "loss": 5.2801,
      "mean_token_accuracy": 0.20444937348365783,
      "num_tokens": 11277471.0,
      "step": 4920
    },
    {
      "entropy": 5.349462604522705,
      "epoch": 0.47310278578290105,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004983932082114659,
      "loss": 5.2014,
      "mean_token_accuracy": 0.19666503518819808,
      "num_tokens": 11290501.0,
      "step": 4925
    },
    {
      "entropy": 5.404211330413818,
      "epoch": 0.473583093179635,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004983891157185699,
      "loss": 5.2444,
      "mean_token_accuracy": 0.2027844339609146,
      "num_tokens": 11301933.0,
      "step": 4930
    },
    {
      "entropy": 5.5084710121154785,
      "epoch": 0.47406340057636887,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004983850180392421,
      "loss": 5.3085,
      "mean_token_accuracy": 0.19369672536849974,
      "num_tokens": 11314539.0,
      "step": 4935
    },
    {
      "entropy": 5.441911029815674,
      "epoch": 0.4745437079731028,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004983809151735775,
      "loss": 5.2238,
      "mean_token_accuracy": 0.20218825191259385,
      "num_tokens": 11326641.0,
      "step": 4940
    },
    {
      "entropy": 5.475595951080322,
      "epoch": 0.4750240153698367,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004983768071216713,
      "loss": 5.312,
      "mean_token_accuracy": 0.19576069414615632,
      "num_tokens": 11337214.0,
      "step": 4945
    },
    {
      "entropy": 5.502204561233521,
      "epoch": 0.4755043227665706,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004983726938836189,
      "loss": 5.2682,
      "mean_token_accuracy": 0.20043425112962723,
      "num_tokens": 11349540.0,
      "step": 4950
    },
    {
      "entropy": 5.4716967105865475,
      "epoch": 0.4759846301633045,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004983685754595159,
      "loss": 5.2739,
      "mean_token_accuracy": 0.19362666010856627,
      "num_tokens": 11359817.0,
      "step": 4955
    },
    {
      "entropy": 5.498916053771973,
      "epoch": 0.47646493756003844,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004983644518494578,
      "loss": 5.3405,
      "mean_token_accuracy": 0.19233174622058868,
      "num_tokens": 11371120.0,
      "step": 4960
    },
    {
      "entropy": 5.424308347702026,
      "epoch": 0.4769452449567723,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004983603230535403,
      "loss": 5.2994,
      "mean_token_accuracy": 0.19341708719730377,
      "num_tokens": 11382071.0,
      "step": 4965
    },
    {
      "entropy": 5.556464004516601,
      "epoch": 0.47742555235350626,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004983561890718594,
      "loss": 5.3125,
      "mean_token_accuracy": 0.19371788948774338,
      "num_tokens": 11392938.0,
      "step": 4970
    },
    {
      "entropy": 5.482561731338501,
      "epoch": 0.47790585975024014,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000498352049904511,
      "loss": 5.3115,
      "mean_token_accuracy": 0.2003222167491913,
      "num_tokens": 11405416.0,
      "step": 4975
    },
    {
      "entropy": 5.496841764450073,
      "epoch": 0.4783861671469741,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004983479055515914,
      "loss": 5.3024,
      "mean_token_accuracy": 0.20455323159694672,
      "num_tokens": 11417008.0,
      "step": 4980
    },
    {
      "entropy": 5.519680595397949,
      "epoch": 0.47886647454370795,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004983437560131964,
      "loss": 5.2917,
      "mean_token_accuracy": 0.19942499548196793,
      "num_tokens": 11428154.0,
      "step": 4985
    },
    {
      "entropy": 5.523438835144043,
      "epoch": 0.4793467819404419,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004983396012894228,
      "loss": 5.2817,
      "mean_token_accuracy": 0.19546004086732865,
      "num_tokens": 11438559.0,
      "step": 4990
    },
    {
      "entropy": 5.551545333862305,
      "epoch": 0.47982708933717577,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004983354413803666,
      "loss": 5.3645,
      "mean_token_accuracy": 0.1964424192905426,
      "num_tokens": 11449501.0,
      "step": 4995
    },
    {
      "entropy": 5.479586744308472,
      "epoch": 0.4803073967339097,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004983312762861248,
      "loss": 5.3492,
      "mean_token_accuracy": 0.19353883862495422,
      "num_tokens": 11461841.0,
      "step": 5000
    },
    {
      "entropy": 5.480322265625,
      "epoch": 0.4807877041306436,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004983271060067939,
      "loss": 5.2627,
      "mean_token_accuracy": 0.1952250838279724,
      "num_tokens": 11472876.0,
      "step": 5005
    },
    {
      "entropy": 5.4854522228240965,
      "epoch": 0.4812680115273775,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004983229305424707,
      "loss": 5.3259,
      "mean_token_accuracy": 0.19420148581266403,
      "num_tokens": 11484541.0,
      "step": 5010
    },
    {
      "entropy": 5.34986834526062,
      "epoch": 0.4817483189241114,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004983187498932522,
      "loss": 5.1671,
      "mean_token_accuracy": 0.20383797883987426,
      "num_tokens": 11496071.0,
      "step": 5015
    },
    {
      "entropy": 5.458111763000488,
      "epoch": 0.48222862632084534,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004983145640592354,
      "loss": 5.2409,
      "mean_token_accuracy": 0.19875885546207428,
      "num_tokens": 11507574.0,
      "step": 5020
    },
    {
      "entropy": 5.464467763900757,
      "epoch": 0.4827089337175792,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004983103730405176,
      "loss": 5.2829,
      "mean_token_accuracy": 0.20098519772291185,
      "num_tokens": 11519738.0,
      "step": 5025
    },
    {
      "entropy": 5.479814291000366,
      "epoch": 0.48318924111431316,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000498306176837196,
      "loss": 5.2848,
      "mean_token_accuracy": 0.2017580211162567,
      "num_tokens": 11530669.0,
      "step": 5030
    },
    {
      "entropy": 5.352247524261474,
      "epoch": 0.48366954851104704,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004983019754493681,
      "loss": 5.198,
      "mean_token_accuracy": 0.20411175340414048,
      "num_tokens": 11541680.0,
      "step": 5035
    },
    {
      "entropy": 5.408116579055786,
      "epoch": 0.484149855907781,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004982977688771314,
      "loss": 5.24,
      "mean_token_accuracy": 0.19973273873329161,
      "num_tokens": 11553204.0,
      "step": 5040
    },
    {
      "entropy": 5.486913728713989,
      "epoch": 0.4846301633045149,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004982935571205835,
      "loss": 5.2753,
      "mean_token_accuracy": 0.1930134505033493,
      "num_tokens": 11564841.0,
      "step": 5045
    },
    {
      "entropy": 5.420222187042237,
      "epoch": 0.4851104707012488,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004982893401798223,
      "loss": 5.2814,
      "mean_token_accuracy": 0.19732938408851625,
      "num_tokens": 11576723.0,
      "step": 5050
    },
    {
      "entropy": 5.540726900100708,
      "epoch": 0.48559077809798273,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004982851180549456,
      "loss": 5.2597,
      "mean_token_accuracy": 0.20198208540678025,
      "num_tokens": 11587727.0,
      "step": 5055
    },
    {
      "entropy": 5.3975653648376465,
      "epoch": 0.4860710854947166,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004982808907460515,
      "loss": 5.2503,
      "mean_token_accuracy": 0.1943666383624077,
      "num_tokens": 11598508.0,
      "step": 5060
    },
    {
      "entropy": 5.430011320114136,
      "epoch": 0.48655139289145055,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004982766582532382,
      "loss": 5.1803,
      "mean_token_accuracy": 0.20839327275753022,
      "num_tokens": 11609684.0,
      "step": 5065
    },
    {
      "entropy": 5.479270267486572,
      "epoch": 0.48703170028818443,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004982724205766038,
      "loss": 5.3047,
      "mean_token_accuracy": 0.18830664306879044,
      "num_tokens": 11621642.0,
      "step": 5070
    },
    {
      "entropy": 5.42671914100647,
      "epoch": 0.48751200768491837,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004982681777162468,
      "loss": 5.2387,
      "mean_token_accuracy": 0.19864881187677383,
      "num_tokens": 11632778.0,
      "step": 5075
    },
    {
      "entropy": 5.405061674118042,
      "epoch": 0.48799231508165225,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004982639296722657,
      "loss": 5.2241,
      "mean_token_accuracy": 0.2059781774878502,
      "num_tokens": 11643581.0,
      "step": 5080
    },
    {
      "entropy": 5.575912761688232,
      "epoch": 0.4884726224783862,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004982596764447591,
      "loss": 5.3957,
      "mean_token_accuracy": 0.18960180282592773,
      "num_tokens": 11656126.0,
      "step": 5085
    },
    {
      "entropy": 5.5779776096344,
      "epoch": 0.48895292987512007,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004982554180338258,
      "loss": 5.3102,
      "mean_token_accuracy": 0.20083604604005814,
      "num_tokens": 11667765.0,
      "step": 5090
    },
    {
      "entropy": 5.442415237426758,
      "epoch": 0.489433237271854,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004982511544395646,
      "loss": 5.3126,
      "mean_token_accuracy": 0.1927064538002014,
      "num_tokens": 11679268.0,
      "step": 5095
    },
    {
      "entropy": 5.463594770431518,
      "epoch": 0.4899135446685879,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004982468856620745,
      "loss": 5.2396,
      "mean_token_accuracy": 0.20180205553770064,
      "num_tokens": 11690049.0,
      "step": 5100
    },
    {
      "entropy": 5.472123432159424,
      "epoch": 0.4903938520653218,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004982426117014545,
      "loss": 5.2225,
      "mean_token_accuracy": 0.20619832575321198,
      "num_tokens": 11701782.0,
      "step": 5105
    },
    {
      "entropy": 5.3492760181427,
      "epoch": 0.4908741594620557,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004982383325578041,
      "loss": 5.2511,
      "mean_token_accuracy": 0.20259998142719268,
      "num_tokens": 11712488.0,
      "step": 5110
    },
    {
      "entropy": 5.471000146865845,
      "epoch": 0.49135446685878964,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004982340482312226,
      "loss": 5.287,
      "mean_token_accuracy": 0.19829708933830262,
      "num_tokens": 11724036.0,
      "step": 5115
    },
    {
      "entropy": 5.430007696151733,
      "epoch": 0.4918347742555235,
      "grad_norm": 0.875,
      "learning_rate": 0.0004982297587218092,
      "loss": 5.257,
      "mean_token_accuracy": 0.2030409649014473,
      "num_tokens": 11736461.0,
      "step": 5120
    },
    {
      "entropy": 5.333529472351074,
      "epoch": 0.49231508165225746,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004982254640296637,
      "loss": 5.2018,
      "mean_token_accuracy": 0.2050809696316719,
      "num_tokens": 11747549.0,
      "step": 5125
    },
    {
      "entropy": 5.553578901290893,
      "epoch": 0.49279538904899134,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004982211641548857,
      "loss": 5.3789,
      "mean_token_accuracy": 0.18653028309345246,
      "num_tokens": 11759012.0,
      "step": 5130
    },
    {
      "entropy": 5.5170821189880375,
      "epoch": 0.4932756964457253,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004982168590975752,
      "loss": 5.2675,
      "mean_token_accuracy": 0.20033161640167235,
      "num_tokens": 11770535.0,
      "step": 5135
    },
    {
      "entropy": 5.425612306594848,
      "epoch": 0.49375600384245916,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004982125488578321,
      "loss": 5.3002,
      "mean_token_accuracy": 0.19846944659948348,
      "num_tokens": 11783708.0,
      "step": 5140
    },
    {
      "entropy": 5.48554162979126,
      "epoch": 0.4942363112391931,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004982082334357563,
      "loss": 5.2675,
      "mean_token_accuracy": 0.1977352738380432,
      "num_tokens": 11794565.0,
      "step": 5145
    },
    {
      "entropy": 5.513044452667236,
      "epoch": 0.494716618635927,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004982039128314481,
      "loss": 5.2835,
      "mean_token_accuracy": 0.2005320444703102,
      "num_tokens": 11805631.0,
      "step": 5150
    },
    {
      "entropy": 5.485285472869873,
      "epoch": 0.4951969260326609,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004981995870450079,
      "loss": 5.315,
      "mean_token_accuracy": 0.19396546483039856,
      "num_tokens": 11817041.0,
      "step": 5155
    },
    {
      "entropy": 5.498223400115966,
      "epoch": 0.4956772334293948,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004981952560765361,
      "loss": 5.2456,
      "mean_token_accuracy": 0.20346460044384002,
      "num_tokens": 11829090.0,
      "step": 5160
    },
    {
      "entropy": 5.569083547592163,
      "epoch": 0.49615754082612873,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004981909199261331,
      "loss": 5.3767,
      "mean_token_accuracy": 0.19005566388368605,
      "num_tokens": 11839980.0,
      "step": 5165
    },
    {
      "entropy": 5.495912361145019,
      "epoch": 0.4966378482228626,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004981865785938998,
      "loss": 5.2764,
      "mean_token_accuracy": 0.19872369319200517,
      "num_tokens": 11851217.0,
      "step": 5170
    },
    {
      "entropy": 5.475414037704468,
      "epoch": 0.49711815561959655,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004981822320799367,
      "loss": 5.2839,
      "mean_token_accuracy": 0.19672577530145646,
      "num_tokens": 11862506.0,
      "step": 5175
    },
    {
      "entropy": 5.492313194274902,
      "epoch": 0.49759846301633043,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004981778803843449,
      "loss": 5.3101,
      "mean_token_accuracy": 0.19360736459493638,
      "num_tokens": 11874844.0,
      "step": 5180
    },
    {
      "entropy": 5.419992923736572,
      "epoch": 0.49807877041306436,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004981735235072256,
      "loss": 5.2242,
      "mean_token_accuracy": 0.20608001351356506,
      "num_tokens": 11885231.0,
      "step": 5185
    },
    {
      "entropy": 5.4680455207824705,
      "epoch": 0.49855907780979825,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004981691614486796,
      "loss": 5.276,
      "mean_token_accuracy": 0.19828298836946487,
      "num_tokens": 11897222.0,
      "step": 5190
    },
    {
      "entropy": 5.424116420745849,
      "epoch": 0.4990393852065322,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004981647942088084,
      "loss": 5.2368,
      "mean_token_accuracy": 0.19893894493579864,
      "num_tokens": 11909493.0,
      "step": 5195
    },
    {
      "entropy": 5.493898153305054,
      "epoch": 0.49951969260326606,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004981604217877135,
      "loss": 5.2405,
      "mean_token_accuracy": 0.20231085419654846,
      "num_tokens": 11920833.0,
      "step": 5200
    },
    {
      "entropy": 5.393985605239868,
      "epoch": 0.5,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000498156044185496,
      "loss": 5.1388,
      "mean_token_accuracy": 0.2119492843747139,
      "num_tokens": 11932910.0,
      "step": 5205
    },
    {
      "entropy": 5.494914722442627,
      "epoch": 0.5004803073967339,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004981516614022579,
      "loss": 5.2717,
      "mean_token_accuracy": 0.20780968368053437,
      "num_tokens": 11944536.0,
      "step": 5210
    },
    {
      "entropy": 5.42256875038147,
      "epoch": 0.5009606147934679,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004981472734381008,
      "loss": 5.2623,
      "mean_token_accuracy": 0.2059893861413002,
      "num_tokens": 11955724.0,
      "step": 5215
    },
    {
      "entropy": 5.404328107833862,
      "epoch": 0.5014409221902018,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004981428802931267,
      "loss": 5.2396,
      "mean_token_accuracy": 0.2013075277209282,
      "num_tokens": 11965921.0,
      "step": 5220
    },
    {
      "entropy": 5.413595867156983,
      "epoch": 0.5019212295869356,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004981384819674375,
      "loss": 5.2042,
      "mean_token_accuracy": 0.20479653477668763,
      "num_tokens": 11976356.0,
      "step": 5225
    },
    {
      "entropy": 5.429383611679077,
      "epoch": 0.5024015369836695,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004981340784611354,
      "loss": 5.2421,
      "mean_token_accuracy": 0.19580158293247224,
      "num_tokens": 11987959.0,
      "step": 5230
    },
    {
      "entropy": 5.420312976837158,
      "epoch": 0.5028818443804035,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004981296697743224,
      "loss": 5.2761,
      "mean_token_accuracy": 0.20255880653858185,
      "num_tokens": 12000089.0,
      "step": 5235
    },
    {
      "entropy": 5.467278861999512,
      "epoch": 0.5033621517771374,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004981252559071012,
      "loss": 5.3348,
      "mean_token_accuracy": 0.19850112795829772,
      "num_tokens": 12011286.0,
      "step": 5240
    },
    {
      "entropy": 5.44477424621582,
      "epoch": 0.5038424591738713,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004981208368595739,
      "loss": 5.2156,
      "mean_token_accuracy": 0.20099994987249375,
      "num_tokens": 12022809.0,
      "step": 5245
    },
    {
      "entropy": 5.398239994049073,
      "epoch": 0.5043227665706052,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004981164126318435,
      "loss": 5.1948,
      "mean_token_accuracy": 0.20346730351448059,
      "num_tokens": 12034311.0,
      "step": 5250
    },
    {
      "entropy": 5.382494497299194,
      "epoch": 0.5048030739673391,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004981119832240124,
      "loss": 5.1927,
      "mean_token_accuracy": 0.21001314222812653,
      "num_tokens": 12046405.0,
      "step": 5255
    },
    {
      "entropy": 5.432193851470947,
      "epoch": 0.505283381364073,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004981075486361837,
      "loss": 5.2393,
      "mean_token_accuracy": 0.20229213833808898,
      "num_tokens": 12059022.0,
      "step": 5260
    },
    {
      "entropy": 5.478096342086792,
      "epoch": 0.5057636887608069,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004981031088684601,
      "loss": 5.2709,
      "mean_token_accuracy": 0.19771818220615386,
      "num_tokens": 12070666.0,
      "step": 5265
    },
    {
      "entropy": 5.411506795883179,
      "epoch": 0.5062439961575408,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004980986639209448,
      "loss": 5.2532,
      "mean_token_accuracy": 0.20054394900798797,
      "num_tokens": 12082702.0,
      "step": 5270
    },
    {
      "entropy": 5.432961368560791,
      "epoch": 0.5067243035542748,
      "grad_norm": 0.921875,
      "learning_rate": 0.000498094213793741,
      "loss": 5.2666,
      "mean_token_accuracy": 0.20299312323331833,
      "num_tokens": 12094034.0,
      "step": 5275
    },
    {
      "entropy": 5.520953607559204,
      "epoch": 0.5072046109510087,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000498089758486952,
      "loss": 5.2545,
      "mean_token_accuracy": 0.18852686434984206,
      "num_tokens": 12104597.0,
      "step": 5280
    },
    {
      "entropy": 5.435879707336426,
      "epoch": 0.5076849183477425,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004980852980006812,
      "loss": 5.2468,
      "mean_token_accuracy": 0.19348674565553664,
      "num_tokens": 12116577.0,
      "step": 5285
    },
    {
      "entropy": 5.432578134536743,
      "epoch": 0.5081652257444764,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004980808323350323,
      "loss": 5.2093,
      "mean_token_accuracy": 0.19678463637828827,
      "num_tokens": 12128005.0,
      "step": 5290
    },
    {
      "entropy": 5.472235870361328,
      "epoch": 0.5086455331412104,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004980763614901089,
      "loss": 5.3248,
      "mean_token_accuracy": 0.1937783271074295,
      "num_tokens": 12138591.0,
      "step": 5295
    },
    {
      "entropy": 5.4084892749786375,
      "epoch": 0.5091258405379443,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004980718854660146,
      "loss": 5.3009,
      "mean_token_accuracy": 0.1967466205358505,
      "num_tokens": 12150010.0,
      "step": 5300
    },
    {
      "entropy": 5.48543152809143,
      "epoch": 0.5096061479346782,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004980674042628537,
      "loss": 5.2466,
      "mean_token_accuracy": 0.19615691304206848,
      "num_tokens": 12161268.0,
      "step": 5305
    },
    {
      "entropy": 5.41184024810791,
      "epoch": 0.5100864553314121,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00049806291788073,
      "loss": 5.208,
      "mean_token_accuracy": 0.20249421298503875,
      "num_tokens": 12172647.0,
      "step": 5310
    },
    {
      "entropy": 5.416033935546875,
      "epoch": 0.510566762728146,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004980584263197477,
      "loss": 5.2858,
      "mean_token_accuracy": 0.20228659212589264,
      "num_tokens": 12183521.0,
      "step": 5315
    },
    {
      "entropy": 5.583924198150635,
      "epoch": 0.5110470701248799,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004980539295800111,
      "loss": 5.338,
      "mean_token_accuracy": 0.19121068716049194,
      "num_tokens": 12193616.0,
      "step": 5320
    },
    {
      "entropy": 5.420936727523804,
      "epoch": 0.5115273775216138,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004980494276616246,
      "loss": 5.2638,
      "mean_token_accuracy": 0.19847189038991928,
      "num_tokens": 12204817.0,
      "step": 5325
    },
    {
      "entropy": 5.486912584304809,
      "epoch": 0.5120076849183477,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004980449205646926,
      "loss": 5.3294,
      "mean_token_accuracy": 0.19133158326148986,
      "num_tokens": 12217281.0,
      "step": 5330
    },
    {
      "entropy": 5.506886053085327,
      "epoch": 0.5124879923150817,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00049804040828932,
      "loss": 5.2328,
      "mean_token_accuracy": 0.2016789048910141,
      "num_tokens": 12228705.0,
      "step": 5335
    },
    {
      "entropy": 5.466443920135498,
      "epoch": 0.5129682997118156,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004980358908356113,
      "loss": 5.3238,
      "mean_token_accuracy": 0.19411759823560715,
      "num_tokens": 12240010.0,
      "step": 5340
    },
    {
      "entropy": 5.502632284164429,
      "epoch": 0.5134486071085494,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004980313682036717,
      "loss": 5.3632,
      "mean_token_accuracy": 0.19064028859138488,
      "num_tokens": 12251384.0,
      "step": 5345
    },
    {
      "entropy": 5.483712148666382,
      "epoch": 0.5139289145052833,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004980268403936058,
      "loss": 5.2208,
      "mean_token_accuracy": 0.19753179997205733,
      "num_tokens": 12262548.0,
      "step": 5350
    },
    {
      "entropy": 5.408590459823609,
      "epoch": 0.5144092219020173,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004980223074055189,
      "loss": 5.1751,
      "mean_token_accuracy": 0.20852846503257752,
      "num_tokens": 12274022.0,
      "step": 5355
    },
    {
      "entropy": 5.410207748413086,
      "epoch": 0.5148895292987512,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004980177692395164,
      "loss": 5.1468,
      "mean_token_accuracy": 0.20906689018011093,
      "num_tokens": 12285000.0,
      "step": 5360
    },
    {
      "entropy": 5.378578042984008,
      "epoch": 0.5153698366954851,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004980132258957035,
      "loss": 5.2078,
      "mean_token_accuracy": 0.20281624048948288,
      "num_tokens": 12297472.0,
      "step": 5365
    },
    {
      "entropy": 5.472492647171021,
      "epoch": 0.515850144092219,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004980086773741856,
      "loss": 5.2594,
      "mean_token_accuracy": 0.20173135846853257,
      "num_tokens": 12309310.0,
      "step": 5370
    },
    {
      "entropy": 5.511911869049072,
      "epoch": 0.516330451488953,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004980041236750685,
      "loss": 5.333,
      "mean_token_accuracy": 0.19360153675079345,
      "num_tokens": 12321897.0,
      "step": 5375
    },
    {
      "entropy": 5.3833719253540036,
      "epoch": 0.5168107588856868,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004979995647984577,
      "loss": 5.1859,
      "mean_token_accuracy": 0.21272823363542556,
      "num_tokens": 12333484.0,
      "step": 5380
    },
    {
      "entropy": 5.409833669662476,
      "epoch": 0.5172910662824207,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004979950007444593,
      "loss": 5.2267,
      "mean_token_accuracy": 0.2010849192738533,
      "num_tokens": 12345154.0,
      "step": 5385
    },
    {
      "entropy": 5.5202230453491214,
      "epoch": 0.5177713736791547,
      "grad_norm": 1.0,
      "learning_rate": 0.0004979904315131792,
      "loss": 5.2784,
      "mean_token_accuracy": 0.20154620558023453,
      "num_tokens": 12354934.0,
      "step": 5390
    },
    {
      "entropy": 5.576596736907959,
      "epoch": 0.5182516810758886,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004979858571047233,
      "loss": 5.4274,
      "mean_token_accuracy": 0.18995014876127242,
      "num_tokens": 12367729.0,
      "step": 5395
    },
    {
      "entropy": 5.442936277389526,
      "epoch": 0.5187319884726225,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004979812775191979,
      "loss": 5.2204,
      "mean_token_accuracy": 0.20359728038311004,
      "num_tokens": 12380366.0,
      "step": 5400
    },
    {
      "entropy": 5.388678073883057,
      "epoch": 0.5192122958693564,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004979766927567094,
      "loss": 5.2636,
      "mean_token_accuracy": 0.19992765635252,
      "num_tokens": 12392338.0,
      "step": 5405
    },
    {
      "entropy": 5.467579412460327,
      "epoch": 0.5196926032660903,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004979721028173643,
      "loss": 5.2255,
      "mean_token_accuracy": 0.20543574094772338,
      "num_tokens": 12403280.0,
      "step": 5410
    },
    {
      "entropy": 5.42668137550354,
      "epoch": 0.5201729106628242,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000497967507701269,
      "loss": 5.214,
      "mean_token_accuracy": 0.20364041924476622,
      "num_tokens": 12414279.0,
      "step": 5415
    },
    {
      "entropy": 5.348680734634399,
      "epoch": 0.5206532180595581,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004979629074085303,
      "loss": 5.1304,
      "mean_token_accuracy": 0.20140196681022643,
      "num_tokens": 12425486.0,
      "step": 5420
    },
    {
      "entropy": 5.379352807998657,
      "epoch": 0.521133525456292,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004979583019392548,
      "loss": 5.2447,
      "mean_token_accuracy": 0.193988399207592,
      "num_tokens": 12437295.0,
      "step": 5425
    },
    {
      "entropy": 5.478912115097046,
      "epoch": 0.521613832853026,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004979536912935497,
      "loss": 5.2053,
      "mean_token_accuracy": 0.2021231472492218,
      "num_tokens": 12448684.0,
      "step": 5430
    },
    {
      "entropy": 5.362641334533691,
      "epoch": 0.5220941402497599,
      "grad_norm": 0.796875,
      "learning_rate": 0.000497949075471522,
      "loss": 5.1613,
      "mean_token_accuracy": 0.20797635167837142,
      "num_tokens": 12461209.0,
      "step": 5435
    },
    {
      "entropy": 5.442259883880615,
      "epoch": 0.5225744476464937,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004979444544732786,
      "loss": 5.2431,
      "mean_token_accuracy": 0.20280103534460067,
      "num_tokens": 12472863.0,
      "step": 5440
    },
    {
      "entropy": 5.501968431472778,
      "epoch": 0.5230547550432276,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000497939828298927,
      "loss": 5.2664,
      "mean_token_accuracy": 0.1969785735011101,
      "num_tokens": 12483510.0,
      "step": 5445
    },
    {
      "entropy": 5.493886137008667,
      "epoch": 0.5235350624399616,
      "grad_norm": 0.875,
      "learning_rate": 0.0004979351969485747,
      "loss": 5.2796,
      "mean_token_accuracy": 0.2020692527294159,
      "num_tokens": 12495532.0,
      "step": 5450
    },
    {
      "entropy": 5.342804622650147,
      "epoch": 0.5240153698366955,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004979305604223291,
      "loss": 5.2013,
      "mean_token_accuracy": 0.203008995950222,
      "num_tokens": 12508413.0,
      "step": 5455
    },
    {
      "entropy": 5.460912895202637,
      "epoch": 0.5244956772334294,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004979259187202978,
      "loss": 5.2896,
      "mean_token_accuracy": 0.20008356869220734,
      "num_tokens": 12520705.0,
      "step": 5460
    },
    {
      "entropy": 5.441412305831909,
      "epoch": 0.5249759846301633,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004979212718425887,
      "loss": 5.2486,
      "mean_token_accuracy": 0.20349239110946654,
      "num_tokens": 12531560.0,
      "step": 5465
    },
    {
      "entropy": 5.435775470733643,
      "epoch": 0.5254562920268973,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004979166197893096,
      "loss": 5.2571,
      "mean_token_accuracy": 0.197725285589695,
      "num_tokens": 12542539.0,
      "step": 5470
    },
    {
      "entropy": 5.496302080154419,
      "epoch": 0.5259365994236311,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004979119625605683,
      "loss": 5.2994,
      "mean_token_accuracy": 0.19296068400144578,
      "num_tokens": 12555228.0,
      "step": 5475
    },
    {
      "entropy": 5.517956209182739,
      "epoch": 0.526416906820365,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004979073001564734,
      "loss": 5.2158,
      "mean_token_accuracy": 0.20317000895738602,
      "num_tokens": 12565715.0,
      "step": 5480
    },
    {
      "entropy": 5.400572633743286,
      "epoch": 0.5268972142170989,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004979026325771328,
      "loss": 5.1812,
      "mean_token_accuracy": 0.2036986157298088,
      "num_tokens": 12576896.0,
      "step": 5485
    },
    {
      "entropy": 5.454914903640747,
      "epoch": 0.5273775216138329,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004978979598226549,
      "loss": 5.2149,
      "mean_token_accuracy": 0.2030993804335594,
      "num_tokens": 12588117.0,
      "step": 5490
    },
    {
      "entropy": 5.428133201599121,
      "epoch": 0.5278578290105668,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004978932818931483,
      "loss": 5.2237,
      "mean_token_accuracy": 0.20230944007635115,
      "num_tokens": 12599987.0,
      "step": 5495
    },
    {
      "entropy": 5.452146720886231,
      "epoch": 0.5283381364073007,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004978885987887216,
      "loss": 5.3156,
      "mean_token_accuracy": 0.19870949983596803,
      "num_tokens": 12611453.0,
      "step": 5500
    },
    {
      "entropy": 5.521382427215576,
      "epoch": 0.5288184438040345,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004978839105094833,
      "loss": 5.2644,
      "mean_token_accuracy": 0.19538675844669343,
      "num_tokens": 12623068.0,
      "step": 5505
    },
    {
      "entropy": 5.503361558914184,
      "epoch": 0.5292987512007685,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004978792170555426,
      "loss": 5.2974,
      "mean_token_accuracy": 0.2002948820590973,
      "num_tokens": 12634952.0,
      "step": 5510
    },
    {
      "entropy": 5.418151998519898,
      "epoch": 0.5297790585975024,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004978745184270083,
      "loss": 5.2396,
      "mean_token_accuracy": 0.19544745832681656,
      "num_tokens": 12647027.0,
      "step": 5515
    },
    {
      "entropy": 5.385825204849243,
      "epoch": 0.5302593659942363,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004978698146239893,
      "loss": 5.1162,
      "mean_token_accuracy": 0.212474824488163,
      "num_tokens": 12657249.0,
      "step": 5520
    },
    {
      "entropy": 5.399412775039673,
      "epoch": 0.5307396733909702,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004978651056465952,
      "loss": 5.1464,
      "mean_token_accuracy": 0.2152785211801529,
      "num_tokens": 12666714.0,
      "step": 5525
    },
    {
      "entropy": 5.414678716659546,
      "epoch": 0.5312199807877042,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000497860391494935,
      "loss": 5.2172,
      "mean_token_accuracy": 0.2012980103492737,
      "num_tokens": 12678627.0,
      "step": 5530
    },
    {
      "entropy": 5.455015230178833,
      "epoch": 0.531700288184438,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004978556721691183,
      "loss": 5.2711,
      "mean_token_accuracy": 0.1999930053949356,
      "num_tokens": 12691009.0,
      "step": 5535
    },
    {
      "entropy": 5.50864577293396,
      "epoch": 0.5321805955811719,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004978509476692547,
      "loss": 5.2335,
      "mean_token_accuracy": 0.1997248113155365,
      "num_tokens": 12702503.0,
      "step": 5540
    },
    {
      "entropy": 5.448041296005249,
      "epoch": 0.5326609029779059,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004978462179954538,
      "loss": 5.2778,
      "mean_token_accuracy": 0.19693051427602767,
      "num_tokens": 12714018.0,
      "step": 5545
    },
    {
      "entropy": 5.514851140975952,
      "epoch": 0.5331412103746398,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004978414831478253,
      "loss": 5.3707,
      "mean_token_accuracy": 0.1978973224759102,
      "num_tokens": 12726726.0,
      "step": 5550
    },
    {
      "entropy": 5.4406757831573485,
      "epoch": 0.5336215177713737,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004978367431264794,
      "loss": 5.2215,
      "mean_token_accuracy": 0.20387665927410126,
      "num_tokens": 12737790.0,
      "step": 5555
    },
    {
      "entropy": 5.364364814758301,
      "epoch": 0.5341018251681076,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004978319979315261,
      "loss": 5.1201,
      "mean_token_accuracy": 0.20641730576753617,
      "num_tokens": 12749473.0,
      "step": 5560
    },
    {
      "entropy": 5.395363712310791,
      "epoch": 0.5345821325648416,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004978272475630752,
      "loss": 5.2616,
      "mean_token_accuracy": 0.2007603868842125,
      "num_tokens": 12762220.0,
      "step": 5565
    },
    {
      "entropy": 5.3281409740448,
      "epoch": 0.5350624399615754,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004978224920212374,
      "loss": 5.0664,
      "mean_token_accuracy": 0.211602383852005,
      "num_tokens": 12772788.0,
      "step": 5570
    },
    {
      "entropy": 5.3264015197753904,
      "epoch": 0.5355427473583093,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004978177313061232,
      "loss": 5.177,
      "mean_token_accuracy": 0.20179829895496368,
      "num_tokens": 12784552.0,
      "step": 5575
    },
    {
      "entropy": 5.434843587875366,
      "epoch": 0.5360230547550432,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004978129654178426,
      "loss": 5.2034,
      "mean_token_accuracy": 0.20716970264911652,
      "num_tokens": 12797618.0,
      "step": 5580
    },
    {
      "entropy": 5.382349634170533,
      "epoch": 0.5365033621517772,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004978081943565067,
      "loss": 5.1428,
      "mean_token_accuracy": 0.20363860428333283,
      "num_tokens": 12807789.0,
      "step": 5585
    },
    {
      "entropy": 5.3887444019317625,
      "epoch": 0.5369836695485111,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004978034181222261,
      "loss": 5.2736,
      "mean_token_accuracy": 0.20326304137706758,
      "num_tokens": 12819020.0,
      "step": 5590
    },
    {
      "entropy": 5.455112600326538,
      "epoch": 0.537463976945245,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004977986367151119,
      "loss": 5.2458,
      "mean_token_accuracy": 0.20026920437812806,
      "num_tokens": 12829527.0,
      "step": 5595
    },
    {
      "entropy": 5.501059579849243,
      "epoch": 0.5379442843419788,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004977938501352747,
      "loss": 5.2268,
      "mean_token_accuracy": 0.19757758378982543,
      "num_tokens": 12840821.0,
      "step": 5600
    },
    {
      "entropy": 5.454826545715332,
      "epoch": 0.5384245917387128,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004977890583828259,
      "loss": 5.2224,
      "mean_token_accuracy": 0.2033350020647049,
      "num_tokens": 12853314.0,
      "step": 5605
    },
    {
      "entropy": 5.3799371242523195,
      "epoch": 0.5389048991354467,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004977842614578768,
      "loss": 5.2325,
      "mean_token_accuracy": 0.1946890637278557,
      "num_tokens": 12863507.0,
      "step": 5610
    },
    {
      "entropy": 5.479557514190674,
      "epoch": 0.5393852065321806,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004977794593605386,
      "loss": 5.2356,
      "mean_token_accuracy": 0.20206407010555266,
      "num_tokens": 12874142.0,
      "step": 5615
    },
    {
      "entropy": 5.506076335906982,
      "epoch": 0.5398655139289145,
      "grad_norm": 0.84375,
      "learning_rate": 0.000497774652090923,
      "loss": 5.2411,
      "mean_token_accuracy": 0.20961515605449677,
      "num_tokens": 12884548.0,
      "step": 5620
    },
    {
      "entropy": 5.466309928894043,
      "epoch": 0.5403458213256485,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004977698396491414,
      "loss": 5.2406,
      "mean_token_accuracy": 0.20028359442949295,
      "num_tokens": 12895562.0,
      "step": 5625
    },
    {
      "entropy": 5.367431688308716,
      "epoch": 0.5408261287223823,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004977650220353055,
      "loss": 5.231,
      "mean_token_accuracy": 0.20302457213401795,
      "num_tokens": 12905308.0,
      "step": 5630
    },
    {
      "entropy": 5.445009374618531,
      "epoch": 0.5413064361191162,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004977601992495274,
      "loss": 5.2573,
      "mean_token_accuracy": 0.19644587337970734,
      "num_tokens": 12916731.0,
      "step": 5635
    },
    {
      "entropy": 5.547921276092529,
      "epoch": 0.5417867435158501,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004977553712919189,
      "loss": 5.3119,
      "mean_token_accuracy": 0.19527269154787064,
      "num_tokens": 12927609.0,
      "step": 5640
    },
    {
      "entropy": 5.500287246704102,
      "epoch": 0.5422670509125841,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004977505381625921,
      "loss": 5.2443,
      "mean_token_accuracy": 0.19932116568088531,
      "num_tokens": 12939445.0,
      "step": 5645
    },
    {
      "entropy": 5.368523740768433,
      "epoch": 0.542747358309318,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004977456998616593,
      "loss": 5.2327,
      "mean_token_accuracy": 0.1991736352443695,
      "num_tokens": 12951638.0,
      "step": 5650
    },
    {
      "entropy": 5.435686540603638,
      "epoch": 0.5432276657060519,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004977408563892327,
      "loss": 5.2123,
      "mean_token_accuracy": 0.2036779284477234,
      "num_tokens": 12963157.0,
      "step": 5655
    },
    {
      "entropy": 5.427400159835815,
      "epoch": 0.5437079731027857,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004977360077454249,
      "loss": 5.1963,
      "mean_token_accuracy": 0.19695059359073638,
      "num_tokens": 12974187.0,
      "step": 5660
    },
    {
      "entropy": 5.444734525680542,
      "epoch": 0.5441882804995197,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004977311539303483,
      "loss": 5.2678,
      "mean_token_accuracy": 0.2042669177055359,
      "num_tokens": 12986612.0,
      "step": 5665
    },
    {
      "entropy": 5.465189361572266,
      "epoch": 0.5446685878962536,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004977262949441158,
      "loss": 5.2887,
      "mean_token_accuracy": 0.19179251343011855,
      "num_tokens": 12998343.0,
      "step": 5670
    },
    {
      "entropy": 5.4359818458557125,
      "epoch": 0.5451488952929875,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004977214307868399,
      "loss": 5.2614,
      "mean_token_accuracy": 0.19960355758666992,
      "num_tokens": 13009272.0,
      "step": 5675
    },
    {
      "entropy": 5.475467252731323,
      "epoch": 0.5456292026897214,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000497716561458634,
      "loss": 5.2353,
      "mean_token_accuracy": 0.20250756442546844,
      "num_tokens": 13019961.0,
      "step": 5680
    },
    {
      "entropy": 5.357641792297363,
      "epoch": 0.5461095100864554,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004977116869596107,
      "loss": 5.1961,
      "mean_token_accuracy": 0.20876758247613908,
      "num_tokens": 13030827.0,
      "step": 5685
    },
    {
      "entropy": 5.530929040908814,
      "epoch": 0.5465898174831892,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004977068072898834,
      "loss": 5.3909,
      "mean_token_accuracy": 0.18940035700798036,
      "num_tokens": 13043702.0,
      "step": 5690
    },
    {
      "entropy": 5.344489479064942,
      "epoch": 0.5470701248799231,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004977019224495652,
      "loss": 5.0776,
      "mean_token_accuracy": 0.2140758141875267,
      "num_tokens": 13054474.0,
      "step": 5695
    },
    {
      "entropy": 5.437290668487549,
      "epoch": 0.547550432276657,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004976970324387698,
      "loss": 5.2275,
      "mean_token_accuracy": 0.1997435286641121,
      "num_tokens": 13064920.0,
      "step": 5700
    },
    {
      "entropy": 5.485542058944702,
      "epoch": 0.548030739673391,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004976921372576104,
      "loss": 5.3156,
      "mean_token_accuracy": 0.20201597213745118,
      "num_tokens": 13077214.0,
      "step": 5705
    },
    {
      "entropy": 5.442243814468384,
      "epoch": 0.5485110470701249,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004976872369062011,
      "loss": 5.2065,
      "mean_token_accuracy": 0.19895603507757187,
      "num_tokens": 13089472.0,
      "step": 5710
    },
    {
      "entropy": 5.432494497299194,
      "epoch": 0.5489913544668588,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004976823313846552,
      "loss": 5.2958,
      "mean_token_accuracy": 0.20202500224113465,
      "num_tokens": 13100369.0,
      "step": 5715
    },
    {
      "entropy": 5.456142520904541,
      "epoch": 0.5494716618635928,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004976774206930869,
      "loss": 5.2343,
      "mean_token_accuracy": 0.1933125302195549,
      "num_tokens": 13110813.0,
      "step": 5720
    },
    {
      "entropy": 5.4524623394012455,
      "epoch": 0.5499519692603266,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004976725048316101,
      "loss": 5.2522,
      "mean_token_accuracy": 0.2034139961004257,
      "num_tokens": 13121946.0,
      "step": 5725
    },
    {
      "entropy": 5.397434043884277,
      "epoch": 0.5504322766570605,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004976675838003388,
      "loss": 5.1711,
      "mean_token_accuracy": 0.20791017562150954,
      "num_tokens": 13133286.0,
      "step": 5730
    },
    {
      "entropy": 5.359141635894775,
      "epoch": 0.5509125840537944,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004976626575993877,
      "loss": 5.1656,
      "mean_token_accuracy": 0.20593208819627762,
      "num_tokens": 13145023.0,
      "step": 5735
    },
    {
      "entropy": 5.3973226070404055,
      "epoch": 0.5513928914505284,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004976577262288705,
      "loss": 5.1303,
      "mean_token_accuracy": 0.20884603410959243,
      "num_tokens": 13156379.0,
      "step": 5740
    },
    {
      "entropy": 5.363160419464111,
      "epoch": 0.5518731988472623,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004976527896889023,
      "loss": 5.1758,
      "mean_token_accuracy": 0.21286931484937668,
      "num_tokens": 13166313.0,
      "step": 5745
    },
    {
      "entropy": 5.463663005828858,
      "epoch": 0.5523535062439962,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004976478479795974,
      "loss": 5.2528,
      "mean_token_accuracy": 0.19728380292654038,
      "num_tokens": 13177968.0,
      "step": 5750
    },
    {
      "entropy": 5.490283870697022,
      "epoch": 0.55283381364073,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004976429011010706,
      "loss": 5.2535,
      "mean_token_accuracy": 0.20040113031864165,
      "num_tokens": 13189464.0,
      "step": 5755
    },
    {
      "entropy": 5.446268033981323,
      "epoch": 0.553314121037464,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004976379490534366,
      "loss": 5.2521,
      "mean_token_accuracy": 0.19554601907730101,
      "num_tokens": 13201716.0,
      "step": 5760
    },
    {
      "entropy": 5.372710418701172,
      "epoch": 0.5537944284341979,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004976329918368107,
      "loss": 5.0973,
      "mean_token_accuracy": 0.20496036112308502,
      "num_tokens": 13212976.0,
      "step": 5765
    },
    {
      "entropy": 5.306728839874268,
      "epoch": 0.5542747358309318,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004976280294513079,
      "loss": 5.1177,
      "mean_token_accuracy": 0.2160654291510582,
      "num_tokens": 13225265.0,
      "step": 5770
    },
    {
      "entropy": 5.551449632644653,
      "epoch": 0.5547550432276657,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004976230618970431,
      "loss": 5.4104,
      "mean_token_accuracy": 0.18847486823797227,
      "num_tokens": 13236443.0,
      "step": 5775
    },
    {
      "entropy": 5.431248521804809,
      "epoch": 0.5552353506243997,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000497618089174132,
      "loss": 5.2782,
      "mean_token_accuracy": 0.20320553034543992,
      "num_tokens": 13248681.0,
      "step": 5780
    },
    {
      "entropy": 5.440784502029419,
      "epoch": 0.5557156580211335,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004976131112826898,
      "loss": 5.2393,
      "mean_token_accuracy": 0.2048149138689041,
      "num_tokens": 13258719.0,
      "step": 5785
    },
    {
      "entropy": 5.4210944175720215,
      "epoch": 0.5561959654178674,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004976081282228323,
      "loss": 5.1994,
      "mean_token_accuracy": 0.20102282464504242,
      "num_tokens": 13270544.0,
      "step": 5790
    },
    {
      "entropy": 5.456362628936768,
      "epoch": 0.5566762728146013,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000497603139994675,
      "loss": 5.1655,
      "mean_token_accuracy": 0.20941447019577025,
      "num_tokens": 13280759.0,
      "step": 5795
    },
    {
      "entropy": 5.349844598770142,
      "epoch": 0.5571565802113353,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004975981465983338,
      "loss": 5.2553,
      "mean_token_accuracy": 0.20415948033332826,
      "num_tokens": 13292496.0,
      "step": 5800
    },
    {
      "entropy": 5.431897115707398,
      "epoch": 0.5576368876080692,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004975931480339246,
      "loss": 5.2483,
      "mean_token_accuracy": 0.19945974200963973,
      "num_tokens": 13303003.0,
      "step": 5805
    },
    {
      "entropy": 5.459396457672119,
      "epoch": 0.5581171950048031,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004975881443015635,
      "loss": 5.2273,
      "mean_token_accuracy": 0.20237836241722107,
      "num_tokens": 13315036.0,
      "step": 5810
    },
    {
      "entropy": 5.405623149871826,
      "epoch": 0.5585975024015369,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004975831354013667,
      "loss": 5.2042,
      "mean_token_accuracy": 0.20477402657270433,
      "num_tokens": 13327529.0,
      "step": 5815
    },
    {
      "entropy": 5.396123552322388,
      "epoch": 0.5590778097982709,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004975781213334503,
      "loss": 5.1979,
      "mean_token_accuracy": 0.2049066975712776,
      "num_tokens": 13339494.0,
      "step": 5820
    },
    {
      "entropy": 5.434308481216431,
      "epoch": 0.5595581171950048,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004975731020979309,
      "loss": 5.2058,
      "mean_token_accuracy": 0.19406634867191314,
      "num_tokens": 13351184.0,
      "step": 5825
    },
    {
      "entropy": 5.376484489440918,
      "epoch": 0.5600384245917387,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004975680776949249,
      "loss": 5.1326,
      "mean_token_accuracy": 0.2064900889992714,
      "num_tokens": 13362402.0,
      "step": 5830
    },
    {
      "entropy": 5.383229541778564,
      "epoch": 0.5605187319884726,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004975630481245492,
      "loss": 5.3042,
      "mean_token_accuracy": 0.1998065561056137,
      "num_tokens": 13375257.0,
      "step": 5835
    },
    {
      "entropy": 5.4490478515625,
      "epoch": 0.5609990393852066,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004975580133869202,
      "loss": 5.2356,
      "mean_token_accuracy": 0.20361828207969665,
      "num_tokens": 13385917.0,
      "step": 5840
    },
    {
      "entropy": 5.498636722564697,
      "epoch": 0.5614793467819404,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004975529734821552,
      "loss": 5.2459,
      "mean_token_accuracy": 0.19649287015199662,
      "num_tokens": 13397962.0,
      "step": 5845
    },
    {
      "entropy": 5.38640866279602,
      "epoch": 0.5619596541786743,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004975479284103708,
      "loss": 5.1398,
      "mean_token_accuracy": 0.20798306167125702,
      "num_tokens": 13408654.0,
      "step": 5850
    },
    {
      "entropy": 5.400376796722412,
      "epoch": 0.5624399615754082,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004975428781716845,
      "loss": 5.2573,
      "mean_token_accuracy": 0.19754728525877,
      "num_tokens": 13419790.0,
      "step": 5855
    },
    {
      "entropy": 5.411325407028198,
      "epoch": 0.5629202689721422,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004975378227662134,
      "loss": 5.176,
      "mean_token_accuracy": 0.20228849053382875,
      "num_tokens": 13430305.0,
      "step": 5860
    },
    {
      "entropy": 5.444153499603272,
      "epoch": 0.5634005763688761,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004975327621940746,
      "loss": 5.2726,
      "mean_token_accuracy": 0.2002544119954109,
      "num_tokens": 13441730.0,
      "step": 5865
    },
    {
      "entropy": 5.4841084480285645,
      "epoch": 0.56388088376561,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004975276964553861,
      "loss": 5.2665,
      "mean_token_accuracy": 0.19990339279174804,
      "num_tokens": 13452003.0,
      "step": 5870
    },
    {
      "entropy": 5.412625741958618,
      "epoch": 0.5643611911623438,
      "grad_norm": 0.875,
      "learning_rate": 0.0004975226255502651,
      "loss": 5.1386,
      "mean_token_accuracy": 0.2083074390888214,
      "num_tokens": 13463953.0,
      "step": 5875
    },
    {
      "entropy": 5.459894847869873,
      "epoch": 0.5648414985590778,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004975175494788297,
      "loss": 5.273,
      "mean_token_accuracy": 0.19823551774024964,
      "num_tokens": 13473820.0,
      "step": 5880
    },
    {
      "entropy": 5.502310132980346,
      "epoch": 0.5653218059558117,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004975124682411974,
      "loss": 5.2922,
      "mean_token_accuracy": 0.20164339393377304,
      "num_tokens": 13484181.0,
      "step": 5885
    },
    {
      "entropy": 5.432629108428955,
      "epoch": 0.5658021133525456,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004975073818374863,
      "loss": 5.2505,
      "mean_token_accuracy": 0.19621144384145736,
      "num_tokens": 13494480.0,
      "step": 5890
    },
    {
      "entropy": 5.37396206855774,
      "epoch": 0.5662824207492796,
      "grad_norm": 0.875,
      "learning_rate": 0.0004975022902678145,
      "loss": 5.1846,
      "mean_token_accuracy": 0.2060466930270195,
      "num_tokens": 13506026.0,
      "step": 5895
    },
    {
      "entropy": 5.48086199760437,
      "epoch": 0.5667627281460135,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004974971935323003,
      "loss": 5.2901,
      "mean_token_accuracy": 0.20040195435285568,
      "num_tokens": 13516849.0,
      "step": 5900
    },
    {
      "entropy": 5.431895637512207,
      "epoch": 0.5672430355427474,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004974920916310619,
      "loss": 5.166,
      "mean_token_accuracy": 0.2056526079773903,
      "num_tokens": 13527386.0,
      "step": 5905
    },
    {
      "entropy": 5.389743185043335,
      "epoch": 0.5677233429394812,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004974869845642178,
      "loss": 5.3018,
      "mean_token_accuracy": 0.19635573774576187,
      "num_tokens": 13540238.0,
      "step": 5910
    },
    {
      "entropy": 5.3790994644165036,
      "epoch": 0.5682036503362152,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004974818723318866,
      "loss": 5.1674,
      "mean_token_accuracy": 0.2107623741030693,
      "num_tokens": 13551808.0,
      "step": 5915
    },
    {
      "entropy": 5.400789070129394,
      "epoch": 0.5686839577329491,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004974767549341868,
      "loss": 5.2114,
      "mean_token_accuracy": 0.20286811590194703,
      "num_tokens": 13563795.0,
      "step": 5920
    },
    {
      "entropy": 5.450816202163696,
      "epoch": 0.569164265129683,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004974716323712376,
      "loss": 5.1799,
      "mean_token_accuracy": 0.20528349727392198,
      "num_tokens": 13575168.0,
      "step": 5925
    },
    {
      "entropy": 5.4283411502838135,
      "epoch": 0.5696445725264169,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004974665046431576,
      "loss": 5.2716,
      "mean_token_accuracy": 0.203519506752491,
      "num_tokens": 13587013.0,
      "step": 5930
    },
    {
      "entropy": 5.348874235153199,
      "epoch": 0.5701248799231509,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004974613717500659,
      "loss": 5.1476,
      "mean_token_accuracy": 0.20087188482284546,
      "num_tokens": 13598138.0,
      "step": 5935
    },
    {
      "entropy": 5.446438884735107,
      "epoch": 0.5706051873198847,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004974562336920818,
      "loss": 5.2518,
      "mean_token_accuracy": 0.19929346293210984,
      "num_tokens": 13609174.0,
      "step": 5940
    },
    {
      "entropy": 5.370011234283448,
      "epoch": 0.5710854947166186,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004974510904693245,
      "loss": 5.1519,
      "mean_token_accuracy": 0.21060054004192352,
      "num_tokens": 13619724.0,
      "step": 5945
    },
    {
      "entropy": 5.397034406661987,
      "epoch": 0.5715658021133525,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004974459420819134,
      "loss": 5.2007,
      "mean_token_accuracy": 0.21060845404863357,
      "num_tokens": 13631114.0,
      "step": 5950
    },
    {
      "entropy": 5.399286508560181,
      "epoch": 0.5720461095100865,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000497440788529968,
      "loss": 5.1351,
      "mean_token_accuracy": 0.20767421871423722,
      "num_tokens": 13643035.0,
      "step": 5955
    },
    {
      "entropy": 5.427866983413696,
      "epoch": 0.5725264169068204,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004974356298136081,
      "loss": 5.2063,
      "mean_token_accuracy": 0.20366256386041642,
      "num_tokens": 13656603.0,
      "step": 5960
    },
    {
      "entropy": 5.359571886062622,
      "epoch": 0.5730067243035543,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004974304659329533,
      "loss": 5.2258,
      "mean_token_accuracy": 0.19997829794883729,
      "num_tokens": 13668420.0,
      "step": 5965
    },
    {
      "entropy": 5.4472008228302,
      "epoch": 0.5734870317002881,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004974252968881236,
      "loss": 5.2387,
      "mean_token_accuracy": 0.19976507872343063,
      "num_tokens": 13680521.0,
      "step": 5970
    },
    {
      "entropy": 5.367796516418457,
      "epoch": 0.5739673390970221,
      "grad_norm": 0.953125,
      "learning_rate": 0.000497420122679239,
      "loss": 5.1409,
      "mean_token_accuracy": 0.2116909921169281,
      "num_tokens": 13692503.0,
      "step": 5975
    },
    {
      "entropy": 5.314179277420044,
      "epoch": 0.574447646493756,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004974149433064196,
      "loss": 5.0185,
      "mean_token_accuracy": 0.21637237817049026,
      "num_tokens": 13703318.0,
      "step": 5980
    },
    {
      "entropy": 5.398615550994873,
      "epoch": 0.5749279538904899,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004974097587697856,
      "loss": 5.2097,
      "mean_token_accuracy": 0.20367472916841506,
      "num_tokens": 13713957.0,
      "step": 5985
    },
    {
      "entropy": 5.43631501197815,
      "epoch": 0.5754082612872238,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004974045690694575,
      "loss": 5.3037,
      "mean_token_accuracy": 0.19581447094678878,
      "num_tokens": 13725300.0,
      "step": 5990
    },
    {
      "entropy": 5.464006423950195,
      "epoch": 0.5758885686839578,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004973993742055557,
      "loss": 5.275,
      "mean_token_accuracy": 0.19415323585271835,
      "num_tokens": 13735463.0,
      "step": 5995
    },
    {
      "entropy": 5.498620653152466,
      "epoch": 0.5763688760806917,
      "grad_norm": 0.875,
      "learning_rate": 0.0004973941741782007,
      "loss": 5.3493,
      "mean_token_accuracy": 0.18962887972593306,
      "num_tokens": 13749062.0,
      "step": 6000
    },
    {
      "epoch": 0.5763688760806917,
      "eval_entropy": 5.239894051716886,
      "eval_loss": 5.227808475494385,
      "eval_mean_token_accuracy": 0.2094871479080675,
      "eval_num_tokens": 13749062.0,
      "eval_runtime": 20.4148,
      "eval_samples_per_second": 1607.41,
      "eval_steps_per_second": 200.932,
      "step": 6000
    },
    {
      "entropy": 5.490328025817871,
      "epoch": 0.5768491834774255,
      "grad_norm": 0.875,
      "learning_rate": 0.0004973889689875135,
      "loss": 5.2928,
      "mean_token_accuracy": 0.1911982014775276,
      "num_tokens": 13761766.0,
      "step": 6005
    },
    {
      "entropy": 5.462340307235718,
      "epoch": 0.5773294908741594,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004973837586336147,
      "loss": 5.2599,
      "mean_token_accuracy": 0.21026235818862915,
      "num_tokens": 13774702.0,
      "step": 6010
    },
    {
      "entropy": 5.416801166534424,
      "epoch": 0.5778097982708934,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004973785431166254,
      "loss": 5.1846,
      "mean_token_accuracy": 0.19890785366296768,
      "num_tokens": 13786215.0,
      "step": 6015
    },
    {
      "entropy": 5.359220361709594,
      "epoch": 0.5782901056676273,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004973733224366666,
      "loss": 5.1956,
      "mean_token_accuracy": 0.2026144728064537,
      "num_tokens": 13797697.0,
      "step": 6020
    },
    {
      "entropy": 5.442294073104859,
      "epoch": 0.5787704130643612,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004973680965938597,
      "loss": 5.272,
      "mean_token_accuracy": 0.20416500866413118,
      "num_tokens": 13810085.0,
      "step": 6025
    },
    {
      "entropy": 5.450731611251831,
      "epoch": 0.579250720461095,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004973628655883258,
      "loss": 5.1663,
      "mean_token_accuracy": 0.20892454832792282,
      "num_tokens": 13822054.0,
      "step": 6030
    },
    {
      "entropy": 5.316114234924316,
      "epoch": 0.579731027857829,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004973576294201865,
      "loss": 5.1506,
      "mean_token_accuracy": 0.20882944464683534,
      "num_tokens": 13834342.0,
      "step": 6035
    },
    {
      "entropy": 5.410109376907348,
      "epoch": 0.5802113352545629,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004973523880895633,
      "loss": 5.2413,
      "mean_token_accuracy": 0.19906787872314452,
      "num_tokens": 13846598.0,
      "step": 6040
    },
    {
      "entropy": 5.44398078918457,
      "epoch": 0.5806916426512968,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004973471415965779,
      "loss": 5.162,
      "mean_token_accuracy": 0.20871315598487855,
      "num_tokens": 13857515.0,
      "step": 6045
    },
    {
      "entropy": 5.304894924163818,
      "epoch": 0.5811719500480308,
      "grad_norm": 0.921875,
      "learning_rate": 0.000497341889941352,
      "loss": 5.1122,
      "mean_token_accuracy": 0.2111505761742592,
      "num_tokens": 13868333.0,
      "step": 6050
    },
    {
      "entropy": 5.409986352920532,
      "epoch": 0.5816522574447647,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004973366331240078,
      "loss": 5.2342,
      "mean_token_accuracy": 0.20056295543909072,
      "num_tokens": 13879048.0,
      "step": 6055
    },
    {
      "entropy": 5.478386354446411,
      "epoch": 0.5821325648414986,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000497331371144667,
      "loss": 5.3195,
      "mean_token_accuracy": 0.19381048232316972,
      "num_tokens": 13891358.0,
      "step": 6060
    },
    {
      "entropy": 5.570667839050293,
      "epoch": 0.5826128722382324,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004973261040034521,
      "loss": 5.3298,
      "mean_token_accuracy": 0.1916772097349167,
      "num_tokens": 13902680.0,
      "step": 6065
    },
    {
      "entropy": 5.4681798934936525,
      "epoch": 0.5830931796349664,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004973208317004852,
      "loss": 5.2401,
      "mean_token_accuracy": 0.19889348596334458,
      "num_tokens": 13913863.0,
      "step": 6070
    },
    {
      "entropy": 5.415434122085571,
      "epoch": 0.5835734870317003,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004973155542358889,
      "loss": 5.1958,
      "mean_token_accuracy": 0.2036360502243042,
      "num_tokens": 13924917.0,
      "step": 6075
    },
    {
      "entropy": 5.298160362243652,
      "epoch": 0.5840537944284342,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004973102716097853,
      "loss": 5.0056,
      "mean_token_accuracy": 0.21209852248430253,
      "num_tokens": 13936022.0,
      "step": 6080
    },
    {
      "entropy": 5.43290753364563,
      "epoch": 0.5845341018251681,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004973049838222973,
      "loss": 5.2329,
      "mean_token_accuracy": 0.19903909265995026,
      "num_tokens": 13947594.0,
      "step": 6085
    },
    {
      "entropy": 5.4542155265808105,
      "epoch": 0.5850144092219021,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004972996908735479,
      "loss": 5.2002,
      "mean_token_accuracy": 0.20452118515968323,
      "num_tokens": 13960698.0,
      "step": 6090
    },
    {
      "entropy": 5.424703931808471,
      "epoch": 0.585494716618636,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004972943927636597,
      "loss": 5.222,
      "mean_token_accuracy": 0.2001119390130043,
      "num_tokens": 13971004.0,
      "step": 6095
    },
    {
      "entropy": 5.334628009796143,
      "epoch": 0.5859750240153698,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004972890894927558,
      "loss": 5.2059,
      "mean_token_accuracy": 0.2064224198460579,
      "num_tokens": 13982500.0,
      "step": 6100
    },
    {
      "entropy": 5.515211486816407,
      "epoch": 0.5864553314121037,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004972837810609592,
      "loss": 5.3519,
      "mean_token_accuracy": 0.19525067508220673,
      "num_tokens": 13994129.0,
      "step": 6105
    },
    {
      "entropy": 5.408864545822143,
      "epoch": 0.5869356388088377,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004972784674683933,
      "loss": 5.1285,
      "mean_token_accuracy": 0.20696143358945845,
      "num_tokens": 14005581.0,
      "step": 6110
    },
    {
      "entropy": 5.4026776313781735,
      "epoch": 0.5874159462055716,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004972731487151815,
      "loss": 5.2099,
      "mean_token_accuracy": 0.1985423147678375,
      "num_tokens": 14015866.0,
      "step": 6115
    },
    {
      "entropy": 5.364549398422241,
      "epoch": 0.5878962536023055,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004972678248014471,
      "loss": 5.2089,
      "mean_token_accuracy": 0.20720898658037185,
      "num_tokens": 14027002.0,
      "step": 6120
    },
    {
      "entropy": 5.375023221969604,
      "epoch": 0.5883765609990393,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004972624957273139,
      "loss": 5.2273,
      "mean_token_accuracy": 0.20587487816810607,
      "num_tokens": 14038962.0,
      "step": 6125
    },
    {
      "entropy": 5.42085280418396,
      "epoch": 0.5888568683957733,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004972571614929055,
      "loss": 5.2307,
      "mean_token_accuracy": 0.20569755285978317,
      "num_tokens": 14049956.0,
      "step": 6130
    },
    {
      "entropy": 5.3963987827301025,
      "epoch": 0.5893371757925072,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004972518220983457,
      "loss": 5.2396,
      "mean_token_accuracy": 0.19862050861120223,
      "num_tokens": 14062413.0,
      "step": 6135
    },
    {
      "entropy": 5.466190767288208,
      "epoch": 0.5898174831892411,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004972464775437586,
      "loss": 5.2071,
      "mean_token_accuracy": 0.20270342975854874,
      "num_tokens": 14075166.0,
      "step": 6140
    },
    {
      "entropy": 5.368501901626587,
      "epoch": 0.590297790585975,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004972411278292683,
      "loss": 5.1505,
      "mean_token_accuracy": 0.20578243136405944,
      "num_tokens": 14085208.0,
      "step": 6145
    },
    {
      "entropy": 5.508627986907959,
      "epoch": 0.590778097982709,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004972357729549988,
      "loss": 5.2982,
      "mean_token_accuracy": 0.1953480437397957,
      "num_tokens": 14096688.0,
      "step": 6150
    },
    {
      "entropy": 5.300438976287841,
      "epoch": 0.5912584053794429,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004972304129210746,
      "loss": 5.1082,
      "mean_token_accuracy": 0.2173052564263344,
      "num_tokens": 14108737.0,
      "step": 6155
    },
    {
      "entropy": 5.398712205886841,
      "epoch": 0.5917387127761767,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004972250477276202,
      "loss": 5.1843,
      "mean_token_accuracy": 0.20542621612548828,
      "num_tokens": 14120121.0,
      "step": 6160
    },
    {
      "entropy": 5.4251556396484375,
      "epoch": 0.5922190201729106,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004972196773747599,
      "loss": 5.225,
      "mean_token_accuracy": 0.2091844528913498,
      "num_tokens": 14132157.0,
      "step": 6165
    },
    {
      "entropy": 5.438767004013061,
      "epoch": 0.5926993275696446,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004972143018626186,
      "loss": 5.1531,
      "mean_token_accuracy": 0.20691134333610534,
      "num_tokens": 14143092.0,
      "step": 6170
    },
    {
      "entropy": 5.375551033020019,
      "epoch": 0.5931796349663785,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004972089211913211,
      "loss": 5.1642,
      "mean_token_accuracy": 0.20463316291570663,
      "num_tokens": 14153973.0,
      "step": 6175
    },
    {
      "entropy": 5.484472274780273,
      "epoch": 0.5936599423631124,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004972035353609923,
      "loss": 5.219,
      "mean_token_accuracy": 0.19699754267930986,
      "num_tokens": 14164919.0,
      "step": 6180
    },
    {
      "entropy": 5.374893856048584,
      "epoch": 0.5941402497598463,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004971981443717572,
      "loss": 5.2306,
      "mean_token_accuracy": 0.20131659060716628,
      "num_tokens": 14176803.0,
      "step": 6185
    },
    {
      "entropy": 5.40919132232666,
      "epoch": 0.5946205571565802,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004971927482237409,
      "loss": 5.1317,
      "mean_token_accuracy": 0.20775262117385865,
      "num_tokens": 14190087.0,
      "step": 6190
    },
    {
      "entropy": 5.445462894439697,
      "epoch": 0.5951008645533141,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004971873469170689,
      "loss": 5.3704,
      "mean_token_accuracy": 0.19574974328279496,
      "num_tokens": 14202242.0,
      "step": 6195
    },
    {
      "entropy": 5.4123053550720215,
      "epoch": 0.595581171950048,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004971819404518664,
      "loss": 5.2765,
      "mean_token_accuracy": 0.20171388536691665,
      "num_tokens": 14214020.0,
      "step": 6200
    },
    {
      "entropy": 5.429128551483155,
      "epoch": 0.5960614793467819,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000497176528828259,
      "loss": 5.1507,
      "mean_token_accuracy": 0.2125359520316124,
      "num_tokens": 14225455.0,
      "step": 6205
    },
    {
      "entropy": 5.381184864044189,
      "epoch": 0.5965417867435159,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004971711120463722,
      "loss": 5.1567,
      "mean_token_accuracy": 0.2035444676876068,
      "num_tokens": 14235955.0,
      "step": 6210
    },
    {
      "entropy": 5.385305023193359,
      "epoch": 0.5970220941402498,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000497165690106332,
      "loss": 5.229,
      "mean_token_accuracy": 0.19900637567043306,
      "num_tokens": 14246107.0,
      "step": 6215
    },
    {
      "entropy": 5.390568113327026,
      "epoch": 0.5975024015369836,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004971602630082642,
      "loss": 5.1787,
      "mean_token_accuracy": 0.205305877327919,
      "num_tokens": 14257042.0,
      "step": 6220
    },
    {
      "entropy": 5.523938179016113,
      "epoch": 0.5979827089337176,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004971548307522947,
      "loss": 5.3611,
      "mean_token_accuracy": 0.19819364100694656,
      "num_tokens": 14268670.0,
      "step": 6225
    },
    {
      "entropy": 5.3982919216156,
      "epoch": 0.5984630163304515,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004971493933385498,
      "loss": 5.1406,
      "mean_token_accuracy": 0.2070077434182167,
      "num_tokens": 14279538.0,
      "step": 6230
    },
    {
      "entropy": 5.418197059631348,
      "epoch": 0.5989433237271854,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004971439507671556,
      "loss": 5.2354,
      "mean_token_accuracy": 0.20025917142629623,
      "num_tokens": 14291928.0,
      "step": 6235
    },
    {
      "entropy": 5.396153593063355,
      "epoch": 0.5994236311239193,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004971385030382384,
      "loss": 5.144,
      "mean_token_accuracy": 0.21402665823698044,
      "num_tokens": 14302958.0,
      "step": 6240
    },
    {
      "entropy": 5.3904797554016115,
      "epoch": 0.5999039385206533,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004971330501519248,
      "loss": 5.2064,
      "mean_token_accuracy": 0.2017538532614708,
      "num_tokens": 14315585.0,
      "step": 6245
    },
    {
      "entropy": 5.387982606887817,
      "epoch": 0.6003842459173871,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004971275921083414,
      "loss": 5.2681,
      "mean_token_accuracy": 0.19753913581371307,
      "num_tokens": 14327228.0,
      "step": 6250
    },
    {
      "entropy": 5.49901180267334,
      "epoch": 0.600864553314121,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000497122128907615,
      "loss": 5.2828,
      "mean_token_accuracy": 0.19430439919233322,
      "num_tokens": 14338079.0,
      "step": 6255
    },
    {
      "entropy": 5.425865125656128,
      "epoch": 0.6013448607108549,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004971166605498722,
      "loss": 5.1383,
      "mean_token_accuracy": 0.2087297558784485,
      "num_tokens": 14350220.0,
      "step": 6260
    },
    {
      "entropy": 5.42844820022583,
      "epoch": 0.6018251681075889,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004971111870352401,
      "loss": 5.2454,
      "mean_token_accuracy": 0.19418667554855346,
      "num_tokens": 14360761.0,
      "step": 6265
    },
    {
      "entropy": 5.417655992507934,
      "epoch": 0.6023054755043228,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004971057083638458,
      "loss": 5.233,
      "mean_token_accuracy": 0.20236316621303557,
      "num_tokens": 14372056.0,
      "step": 6270
    },
    {
      "entropy": 5.484567451477051,
      "epoch": 0.6027857829010567,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004971002245358166,
      "loss": 5.2361,
      "mean_token_accuracy": 0.20370341837406158,
      "num_tokens": 14382601.0,
      "step": 6275
    },
    {
      "entropy": 5.539703369140625,
      "epoch": 0.6032660902977905,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004970947355512795,
      "loss": 5.3307,
      "mean_token_accuracy": 0.18877172619104385,
      "num_tokens": 14393749.0,
      "step": 6280
    },
    {
      "entropy": 5.423619985580444,
      "epoch": 0.6037463976945245,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004970892414103622,
      "loss": 5.2212,
      "mean_token_accuracy": 0.19914059340953827,
      "num_tokens": 14405001.0,
      "step": 6285
    },
    {
      "entropy": 5.310844278335571,
      "epoch": 0.6042267050912584,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004970837421131921,
      "loss": 5.0742,
      "mean_token_accuracy": 0.20705017894506456,
      "num_tokens": 14415875.0,
      "step": 6290
    },
    {
      "entropy": 5.320971298217773,
      "epoch": 0.6047070124879923,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004970782376598972,
      "loss": 5.1301,
      "mean_token_accuracy": 0.20543057918548585,
      "num_tokens": 14426896.0,
      "step": 6295
    },
    {
      "entropy": 5.403817462921142,
      "epoch": 0.6051873198847262,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004970727280506048,
      "loss": 5.1983,
      "mean_token_accuracy": 0.1990131065249443,
      "num_tokens": 14438712.0,
      "step": 6300
    },
    {
      "entropy": 5.4132297992706295,
      "epoch": 0.6056676272814602,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004970672132854431,
      "loss": 5.1585,
      "mean_token_accuracy": 0.2055175244808197,
      "num_tokens": 14450952.0,
      "step": 6305
    },
    {
      "entropy": 5.420947313308716,
      "epoch": 0.6061479346781941,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004970616933645403,
      "loss": 5.185,
      "mean_token_accuracy": 0.20710327923297883,
      "num_tokens": 14462139.0,
      "step": 6310
    },
    {
      "entropy": 5.369873285293579,
      "epoch": 0.6066282420749279,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004970561682880242,
      "loss": 5.2439,
      "mean_token_accuracy": 0.20087671726942063,
      "num_tokens": 14472817.0,
      "step": 6315
    },
    {
      "entropy": 5.432730388641358,
      "epoch": 0.6071085494716618,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004970506380560233,
      "loss": 5.2618,
      "mean_token_accuracy": 0.19673240184783936,
      "num_tokens": 14484241.0,
      "step": 6320
    },
    {
      "entropy": 5.35672287940979,
      "epoch": 0.6075888568683958,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004970451026686659,
      "loss": 5.1535,
      "mean_token_accuracy": 0.2068097934126854,
      "num_tokens": 14496327.0,
      "step": 6325
    },
    {
      "entropy": 5.459692859649659,
      "epoch": 0.6080691642651297,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004970395621260806,
      "loss": 5.1918,
      "mean_token_accuracy": 0.19893223643302918,
      "num_tokens": 14508041.0,
      "step": 6330
    },
    {
      "entropy": 5.4205903053283695,
      "epoch": 0.6085494716618636,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000497034016428396,
      "loss": 5.1801,
      "mean_token_accuracy": 0.21205435395240785,
      "num_tokens": 14519023.0,
      "step": 6335
    },
    {
      "entropy": 5.382953214645386,
      "epoch": 0.6090297790585975,
      "grad_norm": 1.0,
      "learning_rate": 0.0004970284655757409,
      "loss": 5.1772,
      "mean_token_accuracy": 0.20403306037187577,
      "num_tokens": 14530783.0,
      "step": 6340
    },
    {
      "entropy": 5.311904191970825,
      "epoch": 0.6095100864553314,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004970229095682439,
      "loss": 5.1423,
      "mean_token_accuracy": 0.21318317204713821,
      "num_tokens": 14541680.0,
      "step": 6345
    },
    {
      "entropy": 5.447034883499145,
      "epoch": 0.6099903938520653,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004970173484060344,
      "loss": 5.1805,
      "mean_token_accuracy": 0.20109282284975052,
      "num_tokens": 14553792.0,
      "step": 6350
    },
    {
      "entropy": 5.4284569263458256,
      "epoch": 0.6104707012487992,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004970117820892414,
      "loss": 5.2452,
      "mean_token_accuracy": 0.1991787552833557,
      "num_tokens": 14565410.0,
      "step": 6355
    },
    {
      "entropy": 5.372927808761597,
      "epoch": 0.6109510086455331,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004970062106179939,
      "loss": 5.0929,
      "mean_token_accuracy": 0.2088688388466835,
      "num_tokens": 14576722.0,
      "step": 6360
    },
    {
      "entropy": 5.334823942184448,
      "epoch": 0.6114313160422671,
      "grad_norm": 0.875,
      "learning_rate": 0.0004970006339924214,
      "loss": 5.1112,
      "mean_token_accuracy": 0.2104211300611496,
      "num_tokens": 14589443.0,
      "step": 6365
    },
    {
      "entropy": 5.43429388999939,
      "epoch": 0.611911623439001,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004969950522126534,
      "loss": 5.2553,
      "mean_token_accuracy": 0.19828147441148758,
      "num_tokens": 14600731.0,
      "step": 6370
    },
    {
      "entropy": 5.386390352249146,
      "epoch": 0.6123919308357348,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004969894652788196,
      "loss": 5.1268,
      "mean_token_accuracy": 0.2068344920873642,
      "num_tokens": 14611152.0,
      "step": 6375
    },
    {
      "entropy": 5.414159822463989,
      "epoch": 0.6128722382324687,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004969838731910494,
      "loss": 5.2421,
      "mean_token_accuracy": 0.19756061732769012,
      "num_tokens": 14624526.0,
      "step": 6380
    },
    {
      "entropy": 5.393636798858642,
      "epoch": 0.6133525456292027,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004969782759494729,
      "loss": 5.1859,
      "mean_token_accuracy": 0.2060685098171234,
      "num_tokens": 14635527.0,
      "step": 6385
    },
    {
      "entropy": 5.447384214401245,
      "epoch": 0.6138328530259366,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00049697267355422,
      "loss": 5.2091,
      "mean_token_accuracy": 0.20596279948949814,
      "num_tokens": 14647372.0,
      "step": 6390
    },
    {
      "entropy": 5.461427116394043,
      "epoch": 0.6143131604226705,
      "grad_norm": 0.875,
      "learning_rate": 0.0004969670660054208,
      "loss": 5.2379,
      "mean_token_accuracy": 0.197710782289505,
      "num_tokens": 14659216.0,
      "step": 6395
    },
    {
      "entropy": 5.394044208526611,
      "epoch": 0.6147934678194045,
      "grad_norm": 1.0,
      "learning_rate": 0.0004969614533032054,
      "loss": 5.1964,
      "mean_token_accuracy": 0.20306333601474763,
      "num_tokens": 14670336.0,
      "step": 6400
    },
    {
      "entropy": 5.32331395149231,
      "epoch": 0.6152737752161384,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004969558354477041,
      "loss": 5.0572,
      "mean_token_accuracy": 0.21537792533636094,
      "num_tokens": 14680668.0,
      "step": 6405
    },
    {
      "entropy": 5.4224403381347654,
      "epoch": 0.6157540826128722,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004969502124390474,
      "loss": 5.2893,
      "mean_token_accuracy": 0.19843865633010865,
      "num_tokens": 14691521.0,
      "step": 6410
    },
    {
      "entropy": 5.42945704460144,
      "epoch": 0.6162343900096061,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004969445842773658,
      "loss": 5.1466,
      "mean_token_accuracy": 0.19957467168569565,
      "num_tokens": 14702473.0,
      "step": 6415
    },
    {
      "entropy": 5.348221063613892,
      "epoch": 0.6167146974063401,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00049693895096279,
      "loss": 5.111,
      "mean_token_accuracy": 0.20814729779958724,
      "num_tokens": 14713057.0,
      "step": 6420
    },
    {
      "entropy": 5.464430809020996,
      "epoch": 0.617195004803074,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004969333124954508,
      "loss": 5.2257,
      "mean_token_accuracy": 0.19840198755264282,
      "num_tokens": 14725203.0,
      "step": 6425
    },
    {
      "entropy": 5.362083530426025,
      "epoch": 0.6176753121998079,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004969276688754791,
      "loss": 5.1878,
      "mean_token_accuracy": 0.2072448968887329,
      "num_tokens": 14735851.0,
      "step": 6430
    },
    {
      "entropy": 5.4686243534088135,
      "epoch": 0.6181556195965417,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004969220201030059,
      "loss": 5.26,
      "mean_token_accuracy": 0.20460502952337264,
      "num_tokens": 14747876.0,
      "step": 6435
    },
    {
      "entropy": 5.392566585540772,
      "epoch": 0.6186359269932757,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004969163661781624,
      "loss": 5.1921,
      "mean_token_accuracy": 0.20352910161018373,
      "num_tokens": 14758673.0,
      "step": 6440
    },
    {
      "entropy": 5.34576530456543,
      "epoch": 0.6191162343900096,
      "grad_norm": 0.875,
      "learning_rate": 0.0004969107071010798,
      "loss": 5.1227,
      "mean_token_accuracy": 0.2122005432844162,
      "num_tokens": 14770505.0,
      "step": 6445
    },
    {
      "entropy": 5.457253408432007,
      "epoch": 0.6195965417867435,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004969050428718895,
      "loss": 5.2003,
      "mean_token_accuracy": 0.20007197111845015,
      "num_tokens": 14782518.0,
      "step": 6450
    },
    {
      "entropy": 5.37495059967041,
      "epoch": 0.6200768491834774,
      "grad_norm": 1.0,
      "learning_rate": 0.000496899373490723,
      "loss": 5.1333,
      "mean_token_accuracy": 0.20667885690927507,
      "num_tokens": 14794608.0,
      "step": 6455
    },
    {
      "entropy": 5.289442682266236,
      "epoch": 0.6205571565802114,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000496893698957712,
      "loss": 5.0724,
      "mean_token_accuracy": 0.21167365312576295,
      "num_tokens": 14804725.0,
      "step": 6460
    },
    {
      "entropy": 5.409210443496704,
      "epoch": 0.6210374639769453,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004968880192729882,
      "loss": 5.1939,
      "mean_token_accuracy": 0.20232508033514024,
      "num_tokens": 14816178.0,
      "step": 6465
    },
    {
      "entropy": 5.437175846099853,
      "epoch": 0.6215177713736791,
      "grad_norm": 0.875,
      "learning_rate": 0.0004968823344366835,
      "loss": 5.1706,
      "mean_token_accuracy": 0.20008497983217238,
      "num_tokens": 14827808.0,
      "step": 6470
    },
    {
      "entropy": 5.342192268371582,
      "epoch": 0.621998078770413,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004968766444489298,
      "loss": 5.1263,
      "mean_token_accuracy": 0.20254679918289184,
      "num_tokens": 14839423.0,
      "step": 6475
    },
    {
      "entropy": 5.4316998481750485,
      "epoch": 0.622478386167147,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004968709493098593,
      "loss": 5.2501,
      "mean_token_accuracy": 0.20475541949272155,
      "num_tokens": 14849678.0,
      "step": 6480
    },
    {
      "entropy": 5.30169644355774,
      "epoch": 0.6229586935638809,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004968652490196041,
      "loss": 5.1484,
      "mean_token_accuracy": 0.20848240852355956,
      "num_tokens": 14860912.0,
      "step": 6485
    },
    {
      "entropy": 5.382401466369629,
      "epoch": 0.6234390009606148,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004968595435782967,
      "loss": 5.148,
      "mean_token_accuracy": 0.20803963094949723,
      "num_tokens": 14873028.0,
      "step": 6490
    },
    {
      "entropy": 5.483411121368408,
      "epoch": 0.6239193083573487,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004968538329860695,
      "loss": 5.1408,
      "mean_token_accuracy": 0.2057709202170372,
      "num_tokens": 14883985.0,
      "step": 6495
    },
    {
      "entropy": 5.396556329727173,
      "epoch": 0.6243996157540826,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004968481172430549,
      "loss": 5.2073,
      "mean_token_accuracy": 0.1978598415851593,
      "num_tokens": 14896569.0,
      "step": 6500
    },
    {
      "entropy": 5.389759349822998,
      "epoch": 0.6248799231508165,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000496842396349386,
      "loss": 5.1869,
      "mean_token_accuracy": 0.2137570858001709,
      "num_tokens": 14907874.0,
      "step": 6505
    },
    {
      "entropy": 5.395158100128174,
      "epoch": 0.6253602305475504,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004968366703051952,
      "loss": 5.1817,
      "mean_token_accuracy": 0.20938907861709594,
      "num_tokens": 14919501.0,
      "step": 6510
    },
    {
      "entropy": 5.392323684692383,
      "epoch": 0.6258405379442843,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004968309391106157,
      "loss": 5.1592,
      "mean_token_accuracy": 0.20563398003578187,
      "num_tokens": 14929876.0,
      "step": 6515
    },
    {
      "entropy": 5.343015813827515,
      "epoch": 0.6263208453410183,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004968252027657806,
      "loss": 5.2392,
      "mean_token_accuracy": 0.1985512539744377,
      "num_tokens": 14941317.0,
      "step": 6520
    },
    {
      "entropy": 5.463213968276977,
      "epoch": 0.6268011527377522,
      "grad_norm": 0.875,
      "learning_rate": 0.0004968194612708229,
      "loss": 5.2216,
      "mean_token_accuracy": 0.20489775538444518,
      "num_tokens": 14953038.0,
      "step": 6525
    },
    {
      "entropy": 5.376852226257324,
      "epoch": 0.627281460134486,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004968137146258759,
      "loss": 5.1452,
      "mean_token_accuracy": 0.20700180977582933,
      "num_tokens": 14964664.0,
      "step": 6530
    },
    {
      "entropy": 5.295059680938721,
      "epoch": 0.6277617675312199,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004968079628310732,
      "loss": 5.1326,
      "mean_token_accuracy": 0.2132351815700531,
      "num_tokens": 14975761.0,
      "step": 6535
    },
    {
      "entropy": 5.37347149848938,
      "epoch": 0.6282420749279539,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004968022058865482,
      "loss": 5.0625,
      "mean_token_accuracy": 0.20843595117330552,
      "num_tokens": 14987670.0,
      "step": 6540
    },
    {
      "entropy": 5.328548908233643,
      "epoch": 0.6287223823246878,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004967964437924346,
      "loss": 5.0409,
      "mean_token_accuracy": 0.21331107169389724,
      "num_tokens": 14999390.0,
      "step": 6545
    },
    {
      "entropy": 5.3107908248901365,
      "epoch": 0.6292026897214217,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004967906765488662,
      "loss": 5.1506,
      "mean_token_accuracy": 0.20612156242132187,
      "num_tokens": 15011691.0,
      "step": 6550
    },
    {
      "entropy": 5.506126642227173,
      "epoch": 0.6296829971181557,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004967849041559769,
      "loss": 5.2845,
      "mean_token_accuracy": 0.19802508056163787,
      "num_tokens": 15023539.0,
      "step": 6555
    },
    {
      "entropy": 5.437392234802246,
      "epoch": 0.6301633045148896,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004967791266139006,
      "loss": 5.2519,
      "mean_token_accuracy": 0.2090195745229721,
      "num_tokens": 15035529.0,
      "step": 6560
    },
    {
      "entropy": 5.380367279052734,
      "epoch": 0.6306436119116234,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004967733439227716,
      "loss": 5.1923,
      "mean_token_accuracy": 0.20294786095619202,
      "num_tokens": 15047072.0,
      "step": 6565
    },
    {
      "entropy": 5.457985305786133,
      "epoch": 0.6311239193083573,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000496767556082724,
      "loss": 5.2363,
      "mean_token_accuracy": 0.20444143414497376,
      "num_tokens": 15058951.0,
      "step": 6570
    },
    {
      "entropy": 5.346799659729004,
      "epoch": 0.6316042267050913,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004967617630938924,
      "loss": 5.2285,
      "mean_token_accuracy": 0.20584283769130707,
      "num_tokens": 15070647.0,
      "step": 6575
    },
    {
      "entropy": 5.403177738189697,
      "epoch": 0.6320845341018252,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000496755964956411,
      "loss": 5.1488,
      "mean_token_accuracy": 0.21485621929168702,
      "num_tokens": 15082324.0,
      "step": 6580
    },
    {
      "entropy": 5.46810827255249,
      "epoch": 0.6325648414985591,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004967501616704147,
      "loss": 5.2867,
      "mean_token_accuracy": 0.19468965828418733,
      "num_tokens": 15094641.0,
      "step": 6585
    },
    {
      "entropy": 5.5217138767242435,
      "epoch": 0.633045148895293,
      "grad_norm": 0.921875,
      "learning_rate": 0.000496744353236038,
      "loss": 5.2162,
      "mean_token_accuracy": 0.2036300480365753,
      "num_tokens": 15106610.0,
      "step": 6590
    },
    {
      "entropy": 5.346227836608887,
      "epoch": 0.633525456292027,
      "grad_norm": 0.921875,
      "learning_rate": 0.000496738539653416,
      "loss": 5.1768,
      "mean_token_accuracy": 0.20563569962978362,
      "num_tokens": 15118463.0,
      "step": 6595
    },
    {
      "entropy": 5.375806379318237,
      "epoch": 0.6340057636887608,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004967327209226835,
      "loss": 5.1699,
      "mean_token_accuracy": 0.20686524510383605,
      "num_tokens": 15130154.0,
      "step": 6600
    },
    {
      "entropy": 5.334732484817505,
      "epoch": 0.6344860710854947,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004967268970439755,
      "loss": 5.1426,
      "mean_token_accuracy": 0.20653853863477706,
      "num_tokens": 15140854.0,
      "step": 6605
    },
    {
      "entropy": 5.365774440765381,
      "epoch": 0.6349663784822286,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004967210680174274,
      "loss": 5.1444,
      "mean_token_accuracy": 0.21392789632081985,
      "num_tokens": 15151075.0,
      "step": 6610
    },
    {
      "entropy": 5.323400211334229,
      "epoch": 0.6354466858789626,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004967152338431745,
      "loss": 5.1109,
      "mean_token_accuracy": 0.20944196581840516,
      "num_tokens": 15162091.0,
      "step": 6615
    },
    {
      "entropy": 5.469527292251587,
      "epoch": 0.6359269932756965,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004967093945213523,
      "loss": 5.2488,
      "mean_token_accuracy": 0.19773634374141694,
      "num_tokens": 15175321.0,
      "step": 6620
    },
    {
      "entropy": 5.362401342391967,
      "epoch": 0.6364073006724303,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004967035500520962,
      "loss": 5.1801,
      "mean_token_accuracy": 0.20843495875597,
      "num_tokens": 15187058.0,
      "step": 6625
    },
    {
      "entropy": 5.343773031234742,
      "epoch": 0.6368876080691642,
      "grad_norm": 0.875,
      "learning_rate": 0.0004966977004355421,
      "loss": 5.164,
      "mean_token_accuracy": 0.20927274525165557,
      "num_tokens": 15198886.0,
      "step": 6630
    },
    {
      "entropy": 5.475527667999268,
      "epoch": 0.6373679154658982,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004966918456718256,
      "loss": 5.2076,
      "mean_token_accuracy": 0.2070307806134224,
      "num_tokens": 15211026.0,
      "step": 6635
    },
    {
      "entropy": 5.3034186363220215,
      "epoch": 0.6378482228626321,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004966859857610828,
      "loss": 5.1127,
      "mean_token_accuracy": 0.21772426515817642,
      "num_tokens": 15222094.0,
      "step": 6640
    },
    {
      "entropy": 5.339668703079224,
      "epoch": 0.638328530259366,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004966801207034497,
      "loss": 5.2197,
      "mean_token_accuracy": 0.20501981526613236,
      "num_tokens": 15232464.0,
      "step": 6645
    },
    {
      "entropy": 5.419715166091919,
      "epoch": 0.6388088376560999,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004966742504990624,
      "loss": 5.1229,
      "mean_token_accuracy": 0.2082277402281761,
      "num_tokens": 15244564.0,
      "step": 6650
    },
    {
      "entropy": 5.435512828826904,
      "epoch": 0.6392891450528339,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004966683751480572,
      "loss": 5.2635,
      "mean_token_accuracy": 0.20538551956415177,
      "num_tokens": 15255822.0,
      "step": 6655
    },
    {
      "entropy": 5.412751960754394,
      "epoch": 0.6397694524495677,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004966624946505706,
      "loss": 5.1786,
      "mean_token_accuracy": 0.19883393496274948,
      "num_tokens": 15267336.0,
      "step": 6660
    },
    {
      "entropy": 5.331346321105957,
      "epoch": 0.6402497598463016,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004966566090067391,
      "loss": 5.0723,
      "mean_token_accuracy": 0.21310587376356124,
      "num_tokens": 15278073.0,
      "step": 6665
    },
    {
      "entropy": 5.403138160705566,
      "epoch": 0.6407300672430355,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004966507182166993,
      "loss": 5.2625,
      "mean_token_accuracy": 0.19759885519742965,
      "num_tokens": 15290119.0,
      "step": 6670
    },
    {
      "entropy": 5.322434282302856,
      "epoch": 0.6412103746397695,
      "grad_norm": 0.875,
      "learning_rate": 0.000496644822280588,
      "loss": 5.0446,
      "mean_token_accuracy": 0.21892160922288895,
      "num_tokens": 15302235.0,
      "step": 6675
    },
    {
      "entropy": 5.427629661560059,
      "epoch": 0.6416906820365034,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000496638921198542,
      "loss": 5.2677,
      "mean_token_accuracy": 0.19753460288047792,
      "num_tokens": 15312924.0,
      "step": 6680
    },
    {
      "entropy": 5.290898180007934,
      "epoch": 0.6421709894332372,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004966330149706984,
      "loss": 5.0718,
      "mean_token_accuracy": 0.22088615149259566,
      "num_tokens": 15323453.0,
      "step": 6685
    },
    {
      "entropy": 5.351494979858399,
      "epoch": 0.6426512968299711,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004966271035971944,
      "loss": 5.1248,
      "mean_token_accuracy": 0.21146300137043,
      "num_tokens": 15334907.0,
      "step": 6690
    },
    {
      "entropy": 5.326338529586792,
      "epoch": 0.6431316042267051,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004966211870781671,
      "loss": 5.1016,
      "mean_token_accuracy": 0.20986914038658142,
      "num_tokens": 15347211.0,
      "step": 6695
    },
    {
      "entropy": 5.408086109161377,
      "epoch": 0.643611911623439,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000496615265413754,
      "loss": 5.2898,
      "mean_token_accuracy": 0.19876152276992798,
      "num_tokens": 15360171.0,
      "step": 6700
    },
    {
      "entropy": 5.511696147918701,
      "epoch": 0.6440922190201729,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004966093386040923,
      "loss": 5.3405,
      "mean_token_accuracy": 0.20140828639268876,
      "num_tokens": 15372209.0,
      "step": 6705
    },
    {
      "entropy": 5.465041446685791,
      "epoch": 0.6445725264169068,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00049660340664932,
      "loss": 5.2282,
      "mean_token_accuracy": 0.20112976729869841,
      "num_tokens": 15383753.0,
      "step": 6710
    },
    {
      "entropy": 5.449148797988892,
      "epoch": 0.6450528338136408,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004965974695495746,
      "loss": 5.1971,
      "mean_token_accuracy": 0.20426873713731766,
      "num_tokens": 15395202.0,
      "step": 6715
    },
    {
      "entropy": 5.4275689125061035,
      "epoch": 0.6455331412103746,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004965915273049938,
      "loss": 5.1692,
      "mean_token_accuracy": 0.20671294182538985,
      "num_tokens": 15407121.0,
      "step": 6720
    },
    {
      "entropy": 5.349760961532593,
      "epoch": 0.6460134486071085,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004965855799157158,
      "loss": 5.1539,
      "mean_token_accuracy": 0.20702461898326874,
      "num_tokens": 15418063.0,
      "step": 6725
    },
    {
      "entropy": 5.337334775924683,
      "epoch": 0.6464937560038425,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004965796273818787,
      "loss": 5.169,
      "mean_token_accuracy": 0.20369377583265305,
      "num_tokens": 15428830.0,
      "step": 6730
    },
    {
      "entropy": 5.368178176879883,
      "epoch": 0.6469740634005764,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004965736697036206,
      "loss": 5.0987,
      "mean_token_accuracy": 0.20680338740348816,
      "num_tokens": 15441009.0,
      "step": 6735
    },
    {
      "entropy": 5.376477241516113,
      "epoch": 0.6474543707973103,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004965677068810798,
      "loss": 5.1617,
      "mean_token_accuracy": 0.21096395552158356,
      "num_tokens": 15453008.0,
      "step": 6740
    },
    {
      "entropy": 5.437696170806885,
      "epoch": 0.6479346781940442,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004965617389143947,
      "loss": 5.1847,
      "mean_token_accuracy": 0.2094212830066681,
      "num_tokens": 15463907.0,
      "step": 6745
    },
    {
      "entropy": 5.413573455810547,
      "epoch": 0.6484149855907781,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000496555765803704,
      "loss": 5.0966,
      "mean_token_accuracy": 0.20675279200077057,
      "num_tokens": 15474706.0,
      "step": 6750
    },
    {
      "entropy": 5.356460475921631,
      "epoch": 0.648895292987512,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004965497875491462,
      "loss": 5.1685,
      "mean_token_accuracy": 0.21038753688335418,
      "num_tokens": 15486355.0,
      "step": 6755
    },
    {
      "entropy": 5.381618499755859,
      "epoch": 0.6493756003842459,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004965438041508604,
      "loss": 5.175,
      "mean_token_accuracy": 0.20263763815164565,
      "num_tokens": 15498184.0,
      "step": 6760
    },
    {
      "entropy": 5.467784833908081,
      "epoch": 0.6498559077809798,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004965378156089851,
      "loss": 5.2274,
      "mean_token_accuracy": 0.20794383883476258,
      "num_tokens": 15509461.0,
      "step": 6765
    },
    {
      "entropy": 5.371381330490112,
      "epoch": 0.6503362151777138,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004965318219236597,
      "loss": 5.1576,
      "mean_token_accuracy": 0.20625602900981904,
      "num_tokens": 15520169.0,
      "step": 6770
    },
    {
      "entropy": 5.392260408401489,
      "epoch": 0.6508165225744477,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004965258230950232,
      "loss": 5.2383,
      "mean_token_accuracy": 0.2018094003200531,
      "num_tokens": 15532783.0,
      "step": 6775
    },
    {
      "entropy": 5.422503900527954,
      "epoch": 0.6512968299711815,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004965198191232148,
      "loss": 5.1539,
      "mean_token_accuracy": 0.1978270396590233,
      "num_tokens": 15543639.0,
      "step": 6780
    },
    {
      "entropy": 5.336484575271607,
      "epoch": 0.6517771373679154,
      "grad_norm": 0.828125,
      "learning_rate": 0.000496513810008374,
      "loss": 5.1398,
      "mean_token_accuracy": 0.21311411261558533,
      "num_tokens": 15555149.0,
      "step": 6785
    },
    {
      "entropy": 5.421627998352051,
      "epoch": 0.6522574447646494,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004965077957506403,
      "loss": 5.2016,
      "mean_token_accuracy": 0.20595904141664506,
      "num_tokens": 15566922.0,
      "step": 6790
    },
    {
      "entropy": 5.423614406585694,
      "epoch": 0.6527377521613833,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004965017763501533,
      "loss": 5.2482,
      "mean_token_accuracy": 0.20370101928710938,
      "num_tokens": 15578047.0,
      "step": 6795
    },
    {
      "entropy": 5.456806087493897,
      "epoch": 0.6532180595581172,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004964957518070529,
      "loss": 5.2422,
      "mean_token_accuracy": 0.19982969462871553,
      "num_tokens": 15590378.0,
      "step": 6800
    },
    {
      "entropy": 5.470078182220459,
      "epoch": 0.6536983669548511,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004964897221214788,
      "loss": 5.2029,
      "mean_token_accuracy": 0.2055142343044281,
      "num_tokens": 15600866.0,
      "step": 6805
    },
    {
      "entropy": 5.358665132522583,
      "epoch": 0.654178674351585,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000496483687293571,
      "loss": 5.0604,
      "mean_token_accuracy": 0.20851106941699982,
      "num_tokens": 15612617.0,
      "step": 6810
    },
    {
      "entropy": 5.32964596748352,
      "epoch": 0.6546589817483189,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004964776473234696,
      "loss": 5.1538,
      "mean_token_accuracy": 0.210930635035038,
      "num_tokens": 15624446.0,
      "step": 6815
    },
    {
      "entropy": 5.386249351501465,
      "epoch": 0.6551392891450528,
      "grad_norm": 0.875,
      "learning_rate": 0.000496471602211315,
      "loss": 5.1334,
      "mean_token_accuracy": 0.21521304100751876,
      "num_tokens": 15635300.0,
      "step": 6820
    },
    {
      "entropy": 5.310616159439087,
      "epoch": 0.6556195965417867,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004964655519572475,
      "loss": 5.0548,
      "mean_token_accuracy": 0.2114316463470459,
      "num_tokens": 15646452.0,
      "step": 6825
    },
    {
      "entropy": 5.372830820083618,
      "epoch": 0.6560999039385207,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004964594965614072,
      "loss": 5.23,
      "mean_token_accuracy": 0.2003457024693489,
      "num_tokens": 15658514.0,
      "step": 6830
    },
    {
      "entropy": 5.457715845108032,
      "epoch": 0.6565802113352546,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004964534360239353,
      "loss": 5.1516,
      "mean_token_accuracy": 0.20427311211824417,
      "num_tokens": 15670222.0,
      "step": 6835
    },
    {
      "entropy": 5.360155248641968,
      "epoch": 0.6570605187319885,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000496447370344972,
      "loss": 5.1706,
      "mean_token_accuracy": 0.2081298425793648,
      "num_tokens": 15681627.0,
      "step": 6840
    },
    {
      "entropy": 5.386286354064941,
      "epoch": 0.6575408261287223,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004964412995246584,
      "loss": 5.1324,
      "mean_token_accuracy": 0.2057268053293228,
      "num_tokens": 15692632.0,
      "step": 6845
    },
    {
      "entropy": 5.3280738353729244,
      "epoch": 0.6580211335254563,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004964352235631354,
      "loss": 5.1627,
      "mean_token_accuracy": 0.2094754010438919,
      "num_tokens": 15704736.0,
      "step": 6850
    },
    {
      "entropy": 5.414780855178833,
      "epoch": 0.6585014409221902,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000496429142460544,
      "loss": 5.2386,
      "mean_token_accuracy": 0.20740071833133697,
      "num_tokens": 15717057.0,
      "step": 6855
    },
    {
      "entropy": 5.486054849624634,
      "epoch": 0.6589817483189241,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004964230562170254,
      "loss": 5.3012,
      "mean_token_accuracy": 0.19647495746612548,
      "num_tokens": 15728577.0,
      "step": 6860
    },
    {
      "entropy": 5.431281471252442,
      "epoch": 0.659462055715658,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004964169648327209,
      "loss": 5.146,
      "mean_token_accuracy": 0.20458454340696336,
      "num_tokens": 15739714.0,
      "step": 6865
    },
    {
      "entropy": 5.361445617675781,
      "epoch": 0.659942363112392,
      "grad_norm": 0.84375,
      "learning_rate": 0.000496410868307772,
      "loss": 5.0668,
      "mean_token_accuracy": 0.2168011710047722,
      "num_tokens": 15751339.0,
      "step": 6870
    },
    {
      "entropy": 5.493907260894775,
      "epoch": 0.6604226705091258,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004964047666423203,
      "loss": 5.3399,
      "mean_token_accuracy": 0.1936228781938553,
      "num_tokens": 15763177.0,
      "step": 6875
    },
    {
      "entropy": 5.418540954589844,
      "epoch": 0.6609029779058597,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004963986598365072,
      "loss": 5.2063,
      "mean_token_accuracy": 0.20496341586112976,
      "num_tokens": 15774691.0,
      "step": 6880
    },
    {
      "entropy": 5.457896614074707,
      "epoch": 0.6613832853025937,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004963925478904746,
      "loss": 5.2129,
      "mean_token_accuracy": 0.20555105805397034,
      "num_tokens": 15785561.0,
      "step": 6885
    },
    {
      "entropy": 5.407823801040649,
      "epoch": 0.6618635926993276,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004963864308043645,
      "loss": 5.2425,
      "mean_token_accuracy": 0.1990278869867325,
      "num_tokens": 15795687.0,
      "step": 6890
    },
    {
      "entropy": 5.383008575439453,
      "epoch": 0.6623439000960615,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004963803085783189,
      "loss": 5.1898,
      "mean_token_accuracy": 0.20886379331350327,
      "num_tokens": 15807191.0,
      "step": 6895
    },
    {
      "entropy": 5.384150743484497,
      "epoch": 0.6628242074927954,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004963741812124799,
      "loss": 5.1586,
      "mean_token_accuracy": 0.2077916979789734,
      "num_tokens": 15819585.0,
      "step": 6900
    },
    {
      "entropy": 5.437775945663452,
      "epoch": 0.6633045148895294,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004963680487069898,
      "loss": 5.2261,
      "mean_token_accuracy": 0.2039772167801857,
      "num_tokens": 15831580.0,
      "step": 6905
    },
    {
      "entropy": 5.3639538288116455,
      "epoch": 0.6637848222862632,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004963619110619908,
      "loss": 5.1765,
      "mean_token_accuracy": 0.20615523904561997,
      "num_tokens": 15843848.0,
      "step": 6910
    },
    {
      "entropy": 5.454598808288575,
      "epoch": 0.6642651296829971,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004963557682776256,
      "loss": 5.2622,
      "mean_token_accuracy": 0.1992006316781044,
      "num_tokens": 15855091.0,
      "step": 6915
    },
    {
      "entropy": 5.424618053436279,
      "epoch": 0.664745437079731,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004963496203540368,
      "loss": 5.1895,
      "mean_token_accuracy": 0.2092268005013466,
      "num_tokens": 15866676.0,
      "step": 6920
    },
    {
      "entropy": 5.351919221878052,
      "epoch": 0.665225744476465,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004963434672913671,
      "loss": 5.0442,
      "mean_token_accuracy": 0.21184577643871308,
      "num_tokens": 15877901.0,
      "step": 6925
    },
    {
      "entropy": 5.357399892807007,
      "epoch": 0.6657060518731989,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004963373090897592,
      "loss": 5.1378,
      "mean_token_accuracy": 0.19635393619537353,
      "num_tokens": 15889284.0,
      "step": 6930
    },
    {
      "entropy": 5.361107349395752,
      "epoch": 0.6661863592699327,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004963311457493563,
      "loss": 5.093,
      "mean_token_accuracy": 0.21453468650579452,
      "num_tokens": 15901218.0,
      "step": 6935
    },
    {
      "entropy": 5.304901123046875,
      "epoch": 0.6666666666666666,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004963249772703015,
      "loss": 5.0339,
      "mean_token_accuracy": 0.2130548357963562,
      "num_tokens": 15912452.0,
      "step": 6940
    },
    {
      "entropy": 5.377503347396851,
      "epoch": 0.6671469740634006,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004963188036527378,
      "loss": 5.1485,
      "mean_token_accuracy": 0.2080046772956848,
      "num_tokens": 15924353.0,
      "step": 6945
    },
    {
      "entropy": 5.435291051864624,
      "epoch": 0.6676272814601345,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004963126248968087,
      "loss": 5.1745,
      "mean_token_accuracy": 0.21261215806007386,
      "num_tokens": 15936295.0,
      "step": 6950
    },
    {
      "entropy": 5.375908088684082,
      "epoch": 0.6681075888568684,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004963064410026577,
      "loss": 5.1476,
      "mean_token_accuracy": 0.1997729331254959,
      "num_tokens": 15948407.0,
      "step": 6955
    },
    {
      "entropy": 5.357903242111206,
      "epoch": 0.6685878962536023,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004963002519704281,
      "loss": 5.129,
      "mean_token_accuracy": 0.21047675460577012,
      "num_tokens": 15959975.0,
      "step": 6960
    },
    {
      "entropy": 5.4241190433502195,
      "epoch": 0.6690682036503363,
      "grad_norm": 0.77734375,
      "learning_rate": 0.000496294057800264,
      "loss": 5.1173,
      "mean_token_accuracy": 0.2131599172949791,
      "num_tokens": 15973084.0,
      "step": 6965
    },
    {
      "entropy": 5.282241725921631,
      "epoch": 0.6695485110470701,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004962878584923089,
      "loss": 5.0553,
      "mean_token_accuracy": 0.21309628933668137,
      "num_tokens": 15984444.0,
      "step": 6970
    },
    {
      "entropy": 5.317758846282959,
      "epoch": 0.670028818443804,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004962816540467068,
      "loss": 5.1528,
      "mean_token_accuracy": 0.20982538908720016,
      "num_tokens": 15996611.0,
      "step": 6975
    },
    {
      "entropy": 5.406557416915893,
      "epoch": 0.6705091258405379,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004962754444636017,
      "loss": 5.1415,
      "mean_token_accuracy": 0.20628152340650557,
      "num_tokens": 16007144.0,
      "step": 6980
    },
    {
      "entropy": 5.369329595565796,
      "epoch": 0.6709894332372719,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000496269229743138,
      "loss": 5.16,
      "mean_token_accuracy": 0.21100229024887085,
      "num_tokens": 16018878.0,
      "step": 6985
    },
    {
      "entropy": 5.394302606582642,
      "epoch": 0.6714697406340058,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004962630098854597,
      "loss": 5.1551,
      "mean_token_accuracy": 0.20226527452468873,
      "num_tokens": 16031273.0,
      "step": 6990
    },
    {
      "entropy": 5.430879783630371,
      "epoch": 0.6719500480307397,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004962567848907113,
      "loss": 5.2264,
      "mean_token_accuracy": 0.20150285661220552,
      "num_tokens": 16043101.0,
      "step": 6995
    },
    {
      "entropy": 5.345595502853394,
      "epoch": 0.6724303554274735,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004962505547590374,
      "loss": 5.1096,
      "mean_token_accuracy": 0.2055200770497322,
      "num_tokens": 16054922.0,
      "step": 7000
    },
    {
      "entropy": 5.409153890609741,
      "epoch": 0.6729106628242075,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004962443194905826,
      "loss": 5.1877,
      "mean_token_accuracy": 0.20647785663604737,
      "num_tokens": 16065461.0,
      "step": 7005
    },
    {
      "entropy": 5.447549057006836,
      "epoch": 0.6733909702209414,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004962380790854916,
      "loss": 5.2467,
      "mean_token_accuracy": 0.20228500813245773,
      "num_tokens": 16076715.0,
      "step": 7010
    },
    {
      "entropy": 5.37047324180603,
      "epoch": 0.6738712776176753,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004962318335439094,
      "loss": 5.1571,
      "mean_token_accuracy": 0.20498238503932953,
      "num_tokens": 16088937.0,
      "step": 7015
    },
    {
      "entropy": 5.338178634643555,
      "epoch": 0.6743515850144092,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004962255828659809,
      "loss": 5.102,
      "mean_token_accuracy": 0.21312863379716873,
      "num_tokens": 16100004.0,
      "step": 7020
    },
    {
      "entropy": 5.384662294387818,
      "epoch": 0.6748318924111432,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004962193270518513,
      "loss": 5.1815,
      "mean_token_accuracy": 0.20741551667451857,
      "num_tokens": 16112435.0,
      "step": 7025
    },
    {
      "entropy": 5.430048894882202,
      "epoch": 0.675312199807877,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004962130661016659,
      "loss": 5.2739,
      "mean_token_accuracy": 0.20215365290641785,
      "num_tokens": 16124246.0,
      "step": 7030
    },
    {
      "entropy": 5.504215669631958,
      "epoch": 0.6757925072046109,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004962068000155699,
      "loss": 5.2541,
      "mean_token_accuracy": 0.19731413722038268,
      "num_tokens": 16136291.0,
      "step": 7035
    },
    {
      "entropy": 5.332998466491699,
      "epoch": 0.6762728146013448,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004962005287937088,
      "loss": 5.0966,
      "mean_token_accuracy": 0.21123160123825074,
      "num_tokens": 16148641.0,
      "step": 7040
    },
    {
      "entropy": 5.327459621429443,
      "epoch": 0.6767531219980788,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004961942524362283,
      "loss": 5.0949,
      "mean_token_accuracy": 0.2182485729455948,
      "num_tokens": 16159152.0,
      "step": 7045
    },
    {
      "entropy": 5.451176023483276,
      "epoch": 0.6772334293948127,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004961879709432741,
      "loss": 5.1951,
      "mean_token_accuracy": 0.2038820743560791,
      "num_tokens": 16170019.0,
      "step": 7050
    },
    {
      "entropy": 5.448411893844605,
      "epoch": 0.6777137367915466,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000496181684314992,
      "loss": 5.2101,
      "mean_token_accuracy": 0.2010012224316597,
      "num_tokens": 16181714.0,
      "step": 7055
    },
    {
      "entropy": 5.380151891708374,
      "epoch": 0.6781940441882806,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004961753925515279,
      "loss": 5.239,
      "mean_token_accuracy": 0.19977569878101348,
      "num_tokens": 16192822.0,
      "step": 7060
    },
    {
      "entropy": 5.30504355430603,
      "epoch": 0.6786743515850144,
      "grad_norm": 0.953125,
      "learning_rate": 0.000496169095653028,
      "loss": 5.0455,
      "mean_token_accuracy": 0.21294282972812653,
      "num_tokens": 16203660.0,
      "step": 7065
    },
    {
      "entropy": 5.333891201019287,
      "epoch": 0.6791546589817483,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004961627936196384,
      "loss": 5.1329,
      "mean_token_accuracy": 0.21205862313508989,
      "num_tokens": 16215495.0,
      "step": 7070
    },
    {
      "entropy": 5.3648933410644535,
      "epoch": 0.6796349663784822,
      "grad_norm": 0.875,
      "learning_rate": 0.0004961564864515055,
      "loss": 5.0898,
      "mean_token_accuracy": 0.2154749244451523,
      "num_tokens": 16227532.0,
      "step": 7075
    },
    {
      "entropy": 5.2359169006347654,
      "epoch": 0.6801152737752162,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004961501741487757,
      "loss": 4.9647,
      "mean_token_accuracy": 0.22248569875955582,
      "num_tokens": 16238979.0,
      "step": 7080
    },
    {
      "entropy": 5.23373851776123,
      "epoch": 0.6805955811719501,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004961438567115955,
      "loss": 5.0217,
      "mean_token_accuracy": 0.21414166688919067,
      "num_tokens": 16249146.0,
      "step": 7085
    },
    {
      "entropy": 5.333969593048096,
      "epoch": 0.681075888568684,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004961375341401116,
      "loss": 5.1266,
      "mean_token_accuracy": 0.20568785965442657,
      "num_tokens": 16262269.0,
      "step": 7090
    },
    {
      "entropy": 5.355562210083008,
      "epoch": 0.6815561959654178,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004961312064344708,
      "loss": 5.0935,
      "mean_token_accuracy": 0.21303499937057496,
      "num_tokens": 16273877.0,
      "step": 7095
    },
    {
      "entropy": 5.448276615142822,
      "epoch": 0.6820365033621518,
      "grad_norm": 0.90625,
      "learning_rate": 0.00049612487359482,
      "loss": 5.3509,
      "mean_token_accuracy": 0.20162592679262162,
      "num_tokens": 16285623.0,
      "step": 7100
    },
    {
      "entropy": 5.330206871032715,
      "epoch": 0.6825168107588857,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004961185356213062,
      "loss": 5.0716,
      "mean_token_accuracy": 0.2116587206721306,
      "num_tokens": 16296242.0,
      "step": 7105
    },
    {
      "entropy": 5.4196404933929445,
      "epoch": 0.6829971181556196,
      "grad_norm": 0.875,
      "learning_rate": 0.0004961121925140767,
      "loss": 5.2492,
      "mean_token_accuracy": 0.20306337624788284,
      "num_tokens": 16308400.0,
      "step": 7110
    },
    {
      "entropy": 5.3860764503479,
      "epoch": 0.6834774255523535,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004961058442732786,
      "loss": 5.1371,
      "mean_token_accuracy": 0.2072478339076042,
      "num_tokens": 16319201.0,
      "step": 7115
    },
    {
      "entropy": 5.400764799118042,
      "epoch": 0.6839577329490875,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004960994908990594,
      "loss": 5.1894,
      "mean_token_accuracy": 0.20453314781188964,
      "num_tokens": 16331462.0,
      "step": 7120
    },
    {
      "entropy": 5.402371740341186,
      "epoch": 0.6844380403458213,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004960931323915665,
      "loss": 5.1832,
      "mean_token_accuracy": 0.2105350375175476,
      "num_tokens": 16343428.0,
      "step": 7125
    },
    {
      "entropy": 5.371869373321533,
      "epoch": 0.6849183477425552,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004960867687509475,
      "loss": 5.0703,
      "mean_token_accuracy": 0.21095869690179825,
      "num_tokens": 16356270.0,
      "step": 7130
    },
    {
      "entropy": 5.346034240722656,
      "epoch": 0.6853986551392891,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004960803999773504,
      "loss": 5.1204,
      "mean_token_accuracy": 0.20673516988754273,
      "num_tokens": 16367039.0,
      "step": 7135
    },
    {
      "entropy": 5.319623565673828,
      "epoch": 0.6858789625360231,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004960740260709228,
      "loss": 5.166,
      "mean_token_accuracy": 0.20859277248382568,
      "num_tokens": 16378601.0,
      "step": 7140
    },
    {
      "entropy": 5.423642492294311,
      "epoch": 0.686359269932757,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004960676470318128,
      "loss": 5.1914,
      "mean_token_accuracy": 0.21026677638292313,
      "num_tokens": 16391096.0,
      "step": 7145
    },
    {
      "entropy": 5.3713376998901365,
      "epoch": 0.6868395773294909,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004960612628601683,
      "loss": 5.032,
      "mean_token_accuracy": 0.21509682089090348,
      "num_tokens": 16401527.0,
      "step": 7150
    },
    {
      "entropy": 5.3664943218231205,
      "epoch": 0.6873198847262247,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004960548735561379,
      "loss": 5.1856,
      "mean_token_accuracy": 0.19482938796281815,
      "num_tokens": 16413948.0,
      "step": 7155
    },
    {
      "entropy": 5.377300453186035,
      "epoch": 0.6878001921229587,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004960484791198697,
      "loss": 5.1005,
      "mean_token_accuracy": 0.21152703911066056,
      "num_tokens": 16425718.0,
      "step": 7160
    },
    {
      "entropy": 5.355294132232666,
      "epoch": 0.6882804995196926,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004960420795515121,
      "loss": 5.1452,
      "mean_token_accuracy": 0.20685127824544908,
      "num_tokens": 16437606.0,
      "step": 7165
    },
    {
      "entropy": 5.368129777908325,
      "epoch": 0.6887608069164265,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004960356748512138,
      "loss": 5.1632,
      "mean_token_accuracy": 0.21085579842329025,
      "num_tokens": 16448964.0,
      "step": 7170
    },
    {
      "entropy": 5.367068433761597,
      "epoch": 0.6892411143131604,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004960292650191236,
      "loss": 5.1762,
      "mean_token_accuracy": 0.2089980036020279,
      "num_tokens": 16459974.0,
      "step": 7175
    },
    {
      "entropy": 5.383929824829101,
      "epoch": 0.6897214217098944,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004960228500553899,
      "loss": 5.1035,
      "mean_token_accuracy": 0.2089856892824173,
      "num_tokens": 16471143.0,
      "step": 7180
    },
    {
      "entropy": 5.3491700172424315,
      "epoch": 0.6902017291066282,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004960164299601623,
      "loss": 5.1183,
      "mean_token_accuracy": 0.21433530747890472,
      "num_tokens": 16482090.0,
      "step": 7185
    },
    {
      "entropy": 5.33709831237793,
      "epoch": 0.6906820365033621,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004960100047335892,
      "loss": 5.1283,
      "mean_token_accuracy": 0.21117574721574783,
      "num_tokens": 16493695.0,
      "step": 7190
    },
    {
      "entropy": 5.381958770751953,
      "epoch": 0.691162343900096,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004960035743758202,
      "loss": 5.2039,
      "mean_token_accuracy": 0.2063015878200531,
      "num_tokens": 16505317.0,
      "step": 7195
    },
    {
      "entropy": 5.369424200057983,
      "epoch": 0.69164265129683,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004959971388870044,
      "loss": 5.1328,
      "mean_token_accuracy": 0.20625419169664383,
      "num_tokens": 16516719.0,
      "step": 7200
    },
    {
      "entropy": 5.4026594161987305,
      "epoch": 0.6921229586935639,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004959906982672912,
      "loss": 5.0865,
      "mean_token_accuracy": 0.21335774064064025,
      "num_tokens": 16528302.0,
      "step": 7205
    },
    {
      "entropy": 5.338581562042236,
      "epoch": 0.6926032660902978,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004959842525168302,
      "loss": 5.1174,
      "mean_token_accuracy": 0.20836355090141295,
      "num_tokens": 16538471.0,
      "step": 7210
    },
    {
      "entropy": 5.398002958297729,
      "epoch": 0.6930835734870316,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004959778016357712,
      "loss": 5.1196,
      "mean_token_accuracy": 0.21086780428886415,
      "num_tokens": 16549283.0,
      "step": 7215
    },
    {
      "entropy": 5.350003290176391,
      "epoch": 0.6935638808837656,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004959713456242637,
      "loss": 5.0938,
      "mean_token_accuracy": 0.20237651765346526,
      "num_tokens": 16560895.0,
      "step": 7220
    },
    {
      "entropy": 5.3539876461029055,
      "epoch": 0.6940441882804995,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004959648844824576,
      "loss": 5.1838,
      "mean_token_accuracy": 0.20399041920900346,
      "num_tokens": 16571306.0,
      "step": 7225
    },
    {
      "entropy": 5.338198566436768,
      "epoch": 0.6945244956772334,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004959584182105032,
      "loss": 5.0979,
      "mean_token_accuracy": 0.21113806068897248,
      "num_tokens": 16582399.0,
      "step": 7230
    },
    {
      "entropy": 5.340732908248901,
      "epoch": 0.6950048030739674,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004959519468085504,
      "loss": 5.0686,
      "mean_token_accuracy": 0.21296932846307753,
      "num_tokens": 16593617.0,
      "step": 7235
    },
    {
      "entropy": 5.304125547409058,
      "epoch": 0.6954851104707013,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004959454702767494,
      "loss": 5.1402,
      "mean_token_accuracy": 0.20957745015621185,
      "num_tokens": 16604109.0,
      "step": 7240
    },
    {
      "entropy": 5.461039209365845,
      "epoch": 0.6959654178674352,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004959389886152507,
      "loss": 5.2396,
      "mean_token_accuracy": 0.20265789479017257,
      "num_tokens": 16616007.0,
      "step": 7245
    },
    {
      "entropy": 5.360119104385376,
      "epoch": 0.696445725264169,
      "grad_norm": 1.0,
      "learning_rate": 0.0004959325018242048,
      "loss": 5.0849,
      "mean_token_accuracy": 0.21437074840068818,
      "num_tokens": 16627609.0,
      "step": 7250
    },
    {
      "entropy": 5.374688911437988,
      "epoch": 0.696926032660903,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004959260099037622,
      "loss": 5.0763,
      "mean_token_accuracy": 0.20924076139926912,
      "num_tokens": 16639842.0,
      "step": 7255
    },
    {
      "entropy": 5.406261014938354,
      "epoch": 0.6974063400576369,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004959195128540737,
      "loss": 5.2363,
      "mean_token_accuracy": 0.19595388025045396,
      "num_tokens": 16651620.0,
      "step": 7260
    },
    {
      "entropy": 5.38639497756958,
      "epoch": 0.6978866474543708,
      "grad_norm": 0.8203125,
      "learning_rate": 0.00049591301067529,
      "loss": 5.1552,
      "mean_token_accuracy": 0.2072630152106285,
      "num_tokens": 16663637.0,
      "step": 7265
    },
    {
      "entropy": 5.420674228668213,
      "epoch": 0.6983669548511047,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004959065033675623,
      "loss": 5.199,
      "mean_token_accuracy": 0.20475565791130065,
      "num_tokens": 16675685.0,
      "step": 7270
    },
    {
      "entropy": 5.426479625701904,
      "epoch": 0.6988472622478387,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004958999909310414,
      "loss": 5.1288,
      "mean_token_accuracy": 0.20318289548158647,
      "num_tokens": 16686268.0,
      "step": 7275
    },
    {
      "entropy": 5.332116174697876,
      "epoch": 0.6993275696445725,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004958934733658788,
      "loss": 5.1455,
      "mean_token_accuracy": 0.2084924265742302,
      "num_tokens": 16698382.0,
      "step": 7280
    },
    {
      "entropy": 5.41964373588562,
      "epoch": 0.6998078770413064,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004958869506722256,
      "loss": 5.2303,
      "mean_token_accuracy": 0.20180756896734237,
      "num_tokens": 16710421.0,
      "step": 7285
    },
    {
      "entropy": 5.354325628280639,
      "epoch": 0.7002881844380403,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004958804228502332,
      "loss": 5.0595,
      "mean_token_accuracy": 0.2088662087917328,
      "num_tokens": 16721658.0,
      "step": 7290
    },
    {
      "entropy": 5.317934846878051,
      "epoch": 0.7007684918347743,
      "grad_norm": 0.875,
      "learning_rate": 0.0004958738899000534,
      "loss": 5.0619,
      "mean_token_accuracy": 0.21445441693067552,
      "num_tokens": 16734100.0,
      "step": 7295
    },
    {
      "entropy": 5.315471744537353,
      "epoch": 0.7012487992315082,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004958673518218377,
      "loss": 5.1139,
      "mean_token_accuracy": 0.21227584332227706,
      "num_tokens": 16746386.0,
      "step": 7300
    },
    {
      "entropy": 5.411015748977661,
      "epoch": 0.7017291066282421,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004958608086157379,
      "loss": 5.1549,
      "mean_token_accuracy": 0.20739419758319855,
      "num_tokens": 16756967.0,
      "step": 7305
    },
    {
      "entropy": 5.274027681350708,
      "epoch": 0.7022094140249759,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000495854260281906,
      "loss": 5.0615,
      "mean_token_accuracy": 0.2082788646221161,
      "num_tokens": 16767454.0,
      "step": 7310
    },
    {
      "entropy": 5.30084376335144,
      "epoch": 0.7026897214217099,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004958477068204941,
      "loss": 5.1297,
      "mean_token_accuracy": 0.21405288577079773,
      "num_tokens": 16779533.0,
      "step": 7315
    },
    {
      "entropy": 5.395409297943115,
      "epoch": 0.7031700288184438,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000495841148231654,
      "loss": 5.1364,
      "mean_token_accuracy": 0.20829372406005858,
      "num_tokens": 16792019.0,
      "step": 7320
    },
    {
      "entropy": 5.339911985397339,
      "epoch": 0.7036503362151777,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004958345845155383,
      "loss": 5.1056,
      "mean_token_accuracy": 0.20919940769672393,
      "num_tokens": 16803089.0,
      "step": 7325
    },
    {
      "entropy": 5.353569412231446,
      "epoch": 0.7041306436119116,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004958280156722992,
      "loss": 5.0619,
      "mean_token_accuracy": 0.21677895337343217,
      "num_tokens": 16813651.0,
      "step": 7330
    },
    {
      "entropy": 5.306046819686889,
      "epoch": 0.7046109510086456,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004958214417020894,
      "loss": 5.0886,
      "mean_token_accuracy": 0.21640269607305526,
      "num_tokens": 16825460.0,
      "step": 7335
    },
    {
      "entropy": 5.304646444320679,
      "epoch": 0.7050912584053795,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004958148626050614,
      "loss": 5.095,
      "mean_token_accuracy": 0.20732040405273439,
      "num_tokens": 16836344.0,
      "step": 7340
    },
    {
      "entropy": 5.3684392929077145,
      "epoch": 0.7055715658021133,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000495808278381368,
      "loss": 5.1172,
      "mean_token_accuracy": 0.2069009706377983,
      "num_tokens": 16848274.0,
      "step": 7345
    },
    {
      "entropy": 5.340195751190185,
      "epoch": 0.7060518731988472,
      "grad_norm": 1.03125,
      "learning_rate": 0.000495801689031162,
      "loss": 5.1264,
      "mean_token_accuracy": 0.20316795110702515,
      "num_tokens": 16860363.0,
      "step": 7350
    },
    {
      "entropy": 5.360002136230468,
      "epoch": 0.7065321805955812,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004957950945545965,
      "loss": 5.0991,
      "mean_token_accuracy": 0.21089154481887817,
      "num_tokens": 16872037.0,
      "step": 7355
    },
    {
      "entropy": 5.439253330230713,
      "epoch": 0.7070124879923151,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004957884949518246,
      "loss": 5.2171,
      "mean_token_accuracy": 0.20246604532003404,
      "num_tokens": 16883845.0,
      "step": 7360
    },
    {
      "entropy": 5.392954397201538,
      "epoch": 0.707492795389049,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004957818902229992,
      "loss": 5.1702,
      "mean_token_accuracy": 0.20164575427770615,
      "num_tokens": 16894965.0,
      "step": 7365
    },
    {
      "entropy": 5.362032699584961,
      "epoch": 0.7079731027857828,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004957752803682741,
      "loss": 5.1452,
      "mean_token_accuracy": 0.20596324205398558,
      "num_tokens": 16907475.0,
      "step": 7370
    },
    {
      "entropy": 5.323917484283447,
      "epoch": 0.7084534101825168,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004957686653878024,
      "loss": 5.0587,
      "mean_token_accuracy": 0.21193834841251374,
      "num_tokens": 16918825.0,
      "step": 7375
    },
    {
      "entropy": 5.328707218170166,
      "epoch": 0.7089337175792507,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000495762045281738,
      "loss": 5.0931,
      "mean_token_accuracy": 0.21276040077209474,
      "num_tokens": 16930663.0,
      "step": 7380
    },
    {
      "entropy": 5.3733714580535885,
      "epoch": 0.7094140249759846,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004957554200502344,
      "loss": 5.1269,
      "mean_token_accuracy": 0.21101990938186646,
      "num_tokens": 16942089.0,
      "step": 7385
    },
    {
      "entropy": 5.337329864501953,
      "epoch": 0.7098943323727186,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004957487896934454,
      "loss": 5.1525,
      "mean_token_accuracy": 0.20968486964702607,
      "num_tokens": 16954203.0,
      "step": 7390
    },
    {
      "entropy": 5.337719869613648,
      "epoch": 0.7103746397694525,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000495742154211525,
      "loss": 5.0989,
      "mean_token_accuracy": 0.21202256083488463,
      "num_tokens": 16966047.0,
      "step": 7395
    },
    {
      "entropy": 5.261015748977661,
      "epoch": 0.7108549471661864,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004957355136046272,
      "loss": 5.003,
      "mean_token_accuracy": 0.2191260740160942,
      "num_tokens": 16977726.0,
      "step": 7400
    },
    {
      "entropy": 5.312539958953858,
      "epoch": 0.7113352545629202,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004957288678729064,
      "loss": 5.0752,
      "mean_token_accuracy": 0.21404675543308258,
      "num_tokens": 16990917.0,
      "step": 7405
    },
    {
      "entropy": 5.346804237365722,
      "epoch": 0.7118155619596542,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004957222170165166,
      "loss": 5.1222,
      "mean_token_accuracy": 0.21714337170124054,
      "num_tokens": 17001703.0,
      "step": 7410
    },
    {
      "entropy": 5.333375310897827,
      "epoch": 0.7122958693563881,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004957155610356124,
      "loss": 5.0,
      "mean_token_accuracy": 0.21575921773910522,
      "num_tokens": 17011918.0,
      "step": 7415
    },
    {
      "entropy": 5.40339732170105,
      "epoch": 0.712776176753122,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004957088999303481,
      "loss": 5.1655,
      "mean_token_accuracy": 0.20513920485973358,
      "num_tokens": 17022653.0,
      "step": 7420
    },
    {
      "entropy": 5.299003314971924,
      "epoch": 0.7132564841498559,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004957022337008787,
      "loss": 5.0172,
      "mean_token_accuracy": 0.21925644874572753,
      "num_tokens": 17034775.0,
      "step": 7425
    },
    {
      "entropy": 5.269684600830078,
      "epoch": 0.7137367915465899,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004956955623473587,
      "loss": 5.0771,
      "mean_token_accuracy": 0.203242127597332,
      "num_tokens": 17046835.0,
      "step": 7430
    },
    {
      "entropy": 5.40220456123352,
      "epoch": 0.7142170989433237,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000495688885869943,
      "loss": 5.1831,
      "mean_token_accuracy": 0.20669764280319214,
      "num_tokens": 17059163.0,
      "step": 7435
    },
    {
      "entropy": 5.4141336441040036,
      "epoch": 0.7146974063400576,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004956822042687868,
      "loss": 5.1726,
      "mean_token_accuracy": 0.20387499183416366,
      "num_tokens": 17070777.0,
      "step": 7440
    },
    {
      "entropy": 5.3857988834381105,
      "epoch": 0.7151777137367915,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004956755175440451,
      "loss": 5.1435,
      "mean_token_accuracy": 0.21015833616256713,
      "num_tokens": 17083770.0,
      "step": 7445
    },
    {
      "entropy": 5.319206094741821,
      "epoch": 0.7156580211335255,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004956688256958731,
      "loss": 5.1347,
      "mean_token_accuracy": 0.20641033202409745,
      "num_tokens": 17096049.0,
      "step": 7450
    },
    {
      "entropy": 5.306346368789673,
      "epoch": 0.7161383285302594,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004956621287244262,
      "loss": 5.0188,
      "mean_token_accuracy": 0.21796371936798095,
      "num_tokens": 17107030.0,
      "step": 7455
    },
    {
      "entropy": 5.307922983169556,
      "epoch": 0.7166186359269933,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004956554266298599,
      "loss": 5.0832,
      "mean_token_accuracy": 0.21178277879953383,
      "num_tokens": 17118638.0,
      "step": 7460
    },
    {
      "entropy": 5.372864627838135,
      "epoch": 0.7170989433237271,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004956487194123298,
      "loss": 5.157,
      "mean_token_accuracy": 0.2121974930167198,
      "num_tokens": 17129783.0,
      "step": 7465
    },
    {
      "entropy": 5.394932889938355,
      "epoch": 0.7175792507204611,
      "grad_norm": 0.875,
      "learning_rate": 0.0004956420070719918,
      "loss": 5.1537,
      "mean_token_accuracy": 0.2046264410018921,
      "num_tokens": 17141939.0,
      "step": 7470
    },
    {
      "entropy": 5.339280891418457,
      "epoch": 0.718059558117195,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004956352896090014,
      "loss": 5.1021,
      "mean_token_accuracy": 0.212686225771904,
      "num_tokens": 17153471.0,
      "step": 7475
    },
    {
      "entropy": 5.379578018188477,
      "epoch": 0.7185398655139289,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004956285670235147,
      "loss": 5.1728,
      "mean_token_accuracy": 0.2014916643500328,
      "num_tokens": 17164659.0,
      "step": 7480
    },
    {
      "entropy": 5.389385032653808,
      "epoch": 0.7190201729106628,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004956218393156879,
      "loss": 5.0919,
      "mean_token_accuracy": 0.2149190679192543,
      "num_tokens": 17175749.0,
      "step": 7485
    },
    {
      "entropy": 5.35507173538208,
      "epoch": 0.7195004803073968,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004956151064856772,
      "loss": 5.1763,
      "mean_token_accuracy": 0.20613805949687958,
      "num_tokens": 17187482.0,
      "step": 7490
    },
    {
      "entropy": 5.337770128250122,
      "epoch": 0.7199807877041307,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004956083685336386,
      "loss": 5.1123,
      "mean_token_accuracy": 0.21079347282648087,
      "num_tokens": 17199904.0,
      "step": 7495
    },
    {
      "entropy": 5.324865436553955,
      "epoch": 0.7204610951008645,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004956016254597289,
      "loss": 5.0734,
      "mean_token_accuracy": 0.20917451232671738,
      "num_tokens": 17210649.0,
      "step": 7500
    },
    {
      "entropy": 5.357712697982788,
      "epoch": 0.7209414024975984,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004955948772641044,
      "loss": 5.1331,
      "mean_token_accuracy": 0.20300116389989853,
      "num_tokens": 17223069.0,
      "step": 7505
    },
    {
      "entropy": 5.38078122138977,
      "epoch": 0.7214217098943324,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000495588123946922,
      "loss": 5.0862,
      "mean_token_accuracy": 0.2128705322742462,
      "num_tokens": 17235144.0,
      "step": 7510
    },
    {
      "entropy": 5.344494152069092,
      "epoch": 0.7219020172910663,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004955813655083384,
      "loss": 5.0434,
      "mean_token_accuracy": 0.21243091821670532,
      "num_tokens": 17246816.0,
      "step": 7515
    },
    {
      "entropy": 5.301909255981445,
      "epoch": 0.7223823246878002,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004955746019485104,
      "loss": 5.1115,
      "mean_token_accuracy": 0.2140478804707527,
      "num_tokens": 17259056.0,
      "step": 7520
    },
    {
      "entropy": 5.389876890182495,
      "epoch": 0.722862632084534,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000495567833267595,
      "loss": 5.1393,
      "mean_token_accuracy": 0.20456192791461944,
      "num_tokens": 17270391.0,
      "step": 7525
    },
    {
      "entropy": 5.391180419921875,
      "epoch": 0.723342939481268,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004955610594657496,
      "loss": 5.0518,
      "mean_token_accuracy": 0.2089456558227539,
      "num_tokens": 17281608.0,
      "step": 7530
    },
    {
      "entropy": 5.3281005859375,
      "epoch": 0.7238232468780019,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004955542805431313,
      "loss": 5.1971,
      "mean_token_accuracy": 0.2003318876028061,
      "num_tokens": 17293649.0,
      "step": 7535
    },
    {
      "entropy": 5.341959142684937,
      "epoch": 0.7243035542747358,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004955474964998976,
      "loss": 5.1564,
      "mean_token_accuracy": 0.2083909347653389,
      "num_tokens": 17305548.0,
      "step": 7540
    },
    {
      "entropy": 5.452064657211304,
      "epoch": 0.7247838616714697,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004955407073362058,
      "loss": 5.1121,
      "mean_token_accuracy": 0.21089767813682556,
      "num_tokens": 17317380.0,
      "step": 7545
    },
    {
      "entropy": 5.336724042892456,
      "epoch": 0.7252641690682037,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004955339130522136,
      "loss": 5.096,
      "mean_token_accuracy": 0.21307118684053422,
      "num_tokens": 17329263.0,
      "step": 7550
    },
    {
      "entropy": 5.366559267044067,
      "epoch": 0.7257444764649376,
      "grad_norm": 0.90625,
      "learning_rate": 0.000495527113648079,
      "loss": 5.234,
      "mean_token_accuracy": 0.20036546885967255,
      "num_tokens": 17343268.0,
      "step": 7555
    },
    {
      "entropy": 5.3912629127502445,
      "epoch": 0.7262247838616714,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004955203091239596,
      "loss": 5.0812,
      "mean_token_accuracy": 0.21764165312051773,
      "num_tokens": 17353979.0,
      "step": 7560
    },
    {
      "entropy": 5.36877212524414,
      "epoch": 0.7267050912584054,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004955134994800134,
      "loss": 5.14,
      "mean_token_accuracy": 0.20684333592653276,
      "num_tokens": 17367587.0,
      "step": 7565
    },
    {
      "entropy": 5.362344741821289,
      "epoch": 0.7271853986551393,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004955066847163984,
      "loss": 5.1468,
      "mean_token_accuracy": 0.20489041209220887,
      "num_tokens": 17379093.0,
      "step": 7570
    },
    {
      "entropy": 5.416521453857422,
      "epoch": 0.7276657060518732,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004954998648332731,
      "loss": 5.1152,
      "mean_token_accuracy": 0.21321907341480256,
      "num_tokens": 17391254.0,
      "step": 7575
    },
    {
      "entropy": 5.321313953399658,
      "epoch": 0.7281460134486071,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004954930398307956,
      "loss": 5.0969,
      "mean_token_accuracy": 0.2080830603837967,
      "num_tokens": 17402532.0,
      "step": 7580
    },
    {
      "entropy": 5.375644779205322,
      "epoch": 0.7286263208453411,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004954862097091245,
      "loss": 5.1453,
      "mean_token_accuracy": 0.21211414486169816,
      "num_tokens": 17413849.0,
      "step": 7585
    },
    {
      "entropy": 5.289227819442749,
      "epoch": 0.729106628242075,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004954793744684184,
      "loss": 5.0486,
      "mean_token_accuracy": 0.2181496202945709,
      "num_tokens": 17424982.0,
      "step": 7590
    },
    {
      "entropy": 5.428251171112061,
      "epoch": 0.7295869356388088,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004954725341088358,
      "loss": 5.2645,
      "mean_token_accuracy": 0.19453805983066558,
      "num_tokens": 17437774.0,
      "step": 7595
    },
    {
      "entropy": 5.400572490692139,
      "epoch": 0.7300672430355427,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004954656886305356,
      "loss": 5.1133,
      "mean_token_accuracy": 0.21708402186632156,
      "num_tokens": 17448676.0,
      "step": 7600
    },
    {
      "entropy": 5.393203687667847,
      "epoch": 0.7305475504322767,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004954588380336768,
      "loss": 5.1016,
      "mean_token_accuracy": 0.20877535939216613,
      "num_tokens": 17459576.0,
      "step": 7605
    },
    {
      "entropy": 5.266789722442627,
      "epoch": 0.7310278578290106,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004954519823184184,
      "loss": 5.0357,
      "mean_token_accuracy": 0.21944561898708342,
      "num_tokens": 17471220.0,
      "step": 7610
    },
    {
      "entropy": 5.3855186939239506,
      "epoch": 0.7315081652257445,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004954451214849196,
      "loss": 5.1609,
      "mean_token_accuracy": 0.20333958715200423,
      "num_tokens": 17483939.0,
      "step": 7615
    },
    {
      "entropy": 5.376885652542114,
      "epoch": 0.7319884726224783,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004954382555333397,
      "loss": 5.1324,
      "mean_token_accuracy": 0.20945583283901215,
      "num_tokens": 17496058.0,
      "step": 7620
    },
    {
      "entropy": 5.354236650466919,
      "epoch": 0.7324687800192123,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000495431384463838,
      "loss": 5.1543,
      "mean_token_accuracy": 0.20988546013832093,
      "num_tokens": 17507750.0,
      "step": 7625
    },
    {
      "entropy": 5.365172433853149,
      "epoch": 0.7329490874159462,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004954245082765741,
      "loss": 5.093,
      "mean_token_accuracy": 0.216850945353508,
      "num_tokens": 17520208.0,
      "step": 7630
    },
    {
      "entropy": 5.314209222793579,
      "epoch": 0.7334293948126801,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004954176269717077,
      "loss": 5.0218,
      "mean_token_accuracy": 0.2171405389904976,
      "num_tokens": 17531655.0,
      "step": 7635
    },
    {
      "entropy": 5.265064668655396,
      "epoch": 0.733909702209414,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004954107405493984,
      "loss": 5.0904,
      "mean_token_accuracy": 0.20970546305179597,
      "num_tokens": 17543766.0,
      "step": 7640
    },
    {
      "entropy": 5.423938131332397,
      "epoch": 0.734390009606148,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004954038490098064,
      "loss": 5.1778,
      "mean_token_accuracy": 0.20048845261335374,
      "num_tokens": 17555710.0,
      "step": 7645
    },
    {
      "entropy": 5.272390079498291,
      "epoch": 0.7348703170028819,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004953969523530914,
      "loss": 5.0195,
      "mean_token_accuracy": 0.21508183926343918,
      "num_tokens": 17566674.0,
      "step": 7650
    },
    {
      "entropy": 5.379325485229492,
      "epoch": 0.7353506243996157,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004953900505794136,
      "loss": 5.1231,
      "mean_token_accuracy": 0.2039305627346039,
      "num_tokens": 17579320.0,
      "step": 7655
    },
    {
      "entropy": 5.393657970428467,
      "epoch": 0.7358309317963496,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004953831436889332,
      "loss": 5.0835,
      "mean_token_accuracy": 0.2030976355075836,
      "num_tokens": 17590679.0,
      "step": 7660
    },
    {
      "entropy": 5.3323766708374025,
      "epoch": 0.7363112391930836,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004953762316818106,
      "loss": 5.0441,
      "mean_token_accuracy": 0.22344406545162201,
      "num_tokens": 17600928.0,
      "step": 7665
    },
    {
      "entropy": 5.220042419433594,
      "epoch": 0.7367915465898175,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004953693145582064,
      "loss": 5.0109,
      "mean_token_accuracy": 0.22107046842575073,
      "num_tokens": 17611906.0,
      "step": 7670
    },
    {
      "entropy": 5.303012037277222,
      "epoch": 0.7372718539865514,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000495362392318281,
      "loss": 5.2031,
      "mean_token_accuracy": 0.20802225172519684,
      "num_tokens": 17623165.0,
      "step": 7675
    },
    {
      "entropy": 5.404568815231324,
      "epoch": 0.7377521613832853,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004953554649621951,
      "loss": 5.0766,
      "mean_token_accuracy": 0.21476980447769164,
      "num_tokens": 17633632.0,
      "step": 7680
    },
    {
      "entropy": 5.340730381011963,
      "epoch": 0.7382324687800192,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004953485324901098,
      "loss": 5.0968,
      "mean_token_accuracy": 0.21471381634473802,
      "num_tokens": 17644801.0,
      "step": 7685
    },
    {
      "entropy": 5.295978355407715,
      "epoch": 0.7387127761767531,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004953415949021857,
      "loss": 5.1068,
      "mean_token_accuracy": 0.20715835839509963,
      "num_tokens": 17656852.0,
      "step": 7690
    },
    {
      "entropy": 5.396936464309692,
      "epoch": 0.739193083573487,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004953346521985843,
      "loss": 5.1144,
      "mean_token_accuracy": 0.20758841633796693,
      "num_tokens": 17667538.0,
      "step": 7695
    },
    {
      "entropy": 5.310084486007691,
      "epoch": 0.7396733909702209,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004953277043794663,
      "loss": 5.0416,
      "mean_token_accuracy": 0.21788775324821472,
      "num_tokens": 17680049.0,
      "step": 7700
    },
    {
      "entropy": 5.265052604675293,
      "epoch": 0.7401536983669549,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004953207514449933,
      "loss": 5.0863,
      "mean_token_accuracy": 0.21244745403528215,
      "num_tokens": 17692611.0,
      "step": 7705
    },
    {
      "entropy": 5.324451303482055,
      "epoch": 0.7406340057636888,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004953137933953267,
      "loss": 4.9862,
      "mean_token_accuracy": 0.21537172198295593,
      "num_tokens": 17704145.0,
      "step": 7710
    },
    {
      "entropy": 5.399234437942505,
      "epoch": 0.7411143131604226,
      "grad_norm": 1.015625,
      "learning_rate": 0.000495306830230628,
      "loss": 5.118,
      "mean_token_accuracy": 0.2077773094177246,
      "num_tokens": 17715467.0,
      "step": 7715
    },
    {
      "entropy": 5.339452409744263,
      "epoch": 0.7415946205571565,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004952998619510589,
      "loss": 5.1061,
      "mean_token_accuracy": 0.2151417464017868,
      "num_tokens": 17726709.0,
      "step": 7720
    },
    {
      "entropy": 5.300859642028809,
      "epoch": 0.7420749279538905,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004952928885567811,
      "loss": 5.0504,
      "mean_token_accuracy": 0.2152445212006569,
      "num_tokens": 17736896.0,
      "step": 7725
    },
    {
      "entropy": 5.371372413635254,
      "epoch": 0.7425552353506244,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004952859100479566,
      "loss": 5.108,
      "mean_token_accuracy": 0.2078382968902588,
      "num_tokens": 17747658.0,
      "step": 7730
    },
    {
      "entropy": 5.329500293731689,
      "epoch": 0.7430355427473583,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004952789264247474,
      "loss": 5.0856,
      "mean_token_accuracy": 0.2106537029147148,
      "num_tokens": 17757833.0,
      "step": 7735
    },
    {
      "entropy": 5.322234773635865,
      "epoch": 0.7435158501440923,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004952719376873156,
      "loss": 5.0538,
      "mean_token_accuracy": 0.21105607002973556,
      "num_tokens": 17769144.0,
      "step": 7740
    },
    {
      "entropy": 5.363447046279907,
      "epoch": 0.7439961575408262,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004952649438358234,
      "loss": 5.1582,
      "mean_token_accuracy": 0.20478810071945192,
      "num_tokens": 17781357.0,
      "step": 7745
    },
    {
      "entropy": 5.4434644222259525,
      "epoch": 0.74447646493756,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004952579448704334,
      "loss": 5.1057,
      "mean_token_accuracy": 0.20709287226200104,
      "num_tokens": 17793640.0,
      "step": 7750
    },
    {
      "entropy": 5.315048599243164,
      "epoch": 0.7449567723342939,
      "grad_norm": 0.890625,
      "learning_rate": 0.000495250940791308,
      "loss": 5.0862,
      "mean_token_accuracy": 0.21254309564828872,
      "num_tokens": 17804228.0,
      "step": 7755
    },
    {
      "entropy": 5.3066788673400875,
      "epoch": 0.7454370797310279,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004952439315986096,
      "loss": 5.1009,
      "mean_token_accuracy": 0.20862143337726594,
      "num_tokens": 17816153.0,
      "step": 7760
    },
    {
      "entropy": 5.326050329208374,
      "epoch": 0.7459173871277618,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004952369172925012,
      "loss": 5.125,
      "mean_token_accuracy": 0.21110926270484925,
      "num_tokens": 17828464.0,
      "step": 7765
    },
    {
      "entropy": 5.23726453781128,
      "epoch": 0.7463976945244957,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004952298978731454,
      "loss": 5.0119,
      "mean_token_accuracy": 0.21157591193914413,
      "num_tokens": 17840126.0,
      "step": 7770
    },
    {
      "entropy": 5.290395927429199,
      "epoch": 0.7468780019212296,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004952228733407055,
      "loss": 5.0634,
      "mean_token_accuracy": 0.21206205636262893,
      "num_tokens": 17851735.0,
      "step": 7775
    },
    {
      "entropy": 5.395587587356568,
      "epoch": 0.7473583093179635,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004952158436953444,
      "loss": 5.0383,
      "mean_token_accuracy": 0.2085823029279709,
      "num_tokens": 17862897.0,
      "step": 7780
    },
    {
      "entropy": 5.339494132995606,
      "epoch": 0.7478386167146974,
      "grad_norm": 1.0,
      "learning_rate": 0.0004952088089372252,
      "loss": 5.1081,
      "mean_token_accuracy": 0.2109610691666603,
      "num_tokens": 17874235.0,
      "step": 7785
    },
    {
      "entropy": 5.303083419799805,
      "epoch": 0.7483189241114313,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004952017690665114,
      "loss": 5.1201,
      "mean_token_accuracy": 0.21084200590848923,
      "num_tokens": 17885677.0,
      "step": 7790
    },
    {
      "entropy": 5.335223150253296,
      "epoch": 0.7487992315081652,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004951947240833664,
      "loss": 5.0293,
      "mean_token_accuracy": 0.21606981754302979,
      "num_tokens": 17896204.0,
      "step": 7795
    },
    {
      "entropy": 5.394116544723511,
      "epoch": 0.7492795389048992,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004951876739879537,
      "loss": 5.1022,
      "mean_token_accuracy": 0.21018677353858947,
      "num_tokens": 17908514.0,
      "step": 7800
    },
    {
      "entropy": 5.322999954223633,
      "epoch": 0.7497598463016331,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004951806187804371,
      "loss": 5.0163,
      "mean_token_accuracy": 0.21938319355249405,
      "num_tokens": 17918493.0,
      "step": 7805
    },
    {
      "entropy": 5.3222287654876705,
      "epoch": 0.7502401536983669,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004951735584609804,
      "loss": 5.0314,
      "mean_token_accuracy": 0.2209548458456993,
      "num_tokens": 17929016.0,
      "step": 7810
    },
    {
      "entropy": 5.274287223815918,
      "epoch": 0.7507204610951008,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004951664930297474,
      "loss": 5.0732,
      "mean_token_accuracy": 0.21650074273347855,
      "num_tokens": 17940674.0,
      "step": 7815
    },
    {
      "entropy": 5.375551986694336,
      "epoch": 0.7512007684918348,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000495159422486902,
      "loss": 5.1082,
      "mean_token_accuracy": 0.2095817357301712,
      "num_tokens": 17951431.0,
      "step": 7820
    },
    {
      "entropy": 5.468128776550293,
      "epoch": 0.7516810758885687,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004951523468326088,
      "loss": 5.1503,
      "mean_token_accuracy": 0.20269854366779327,
      "num_tokens": 17961434.0,
      "step": 7825
    },
    {
      "entropy": 5.2070191383361815,
      "epoch": 0.7521613832853026,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004951452660670317,
      "loss": 4.9883,
      "mean_token_accuracy": 0.22168319970369338,
      "num_tokens": 17971917.0,
      "step": 7830
    },
    {
      "entropy": 5.326111650466919,
      "epoch": 0.7526416906820365,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004951381801903352,
      "loss": 5.0667,
      "mean_token_accuracy": 0.21750225126743317,
      "num_tokens": 17983754.0,
      "step": 7835
    },
    {
      "entropy": 5.3291443347930905,
      "epoch": 0.7531219980787704,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004951310892026839,
      "loss": 5.0518,
      "mean_token_accuracy": 0.21318816542625427,
      "num_tokens": 17994088.0,
      "step": 7840
    },
    {
      "entropy": 5.294413757324219,
      "epoch": 0.7536023054755043,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004951239931042424,
      "loss": 5.0353,
      "mean_token_accuracy": 0.21851316541433335,
      "num_tokens": 18004847.0,
      "step": 7845
    },
    {
      "entropy": 5.322088098526001,
      "epoch": 0.7540826128722382,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004951168918951753,
      "loss": 5.0616,
      "mean_token_accuracy": 0.2216856747865677,
      "num_tokens": 18015756.0,
      "step": 7850
    },
    {
      "entropy": 5.3369475364685055,
      "epoch": 0.7545629202689721,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004951097855756476,
      "loss": 5.1642,
      "mean_token_accuracy": 0.21026596128940583,
      "num_tokens": 18026417.0,
      "step": 7855
    },
    {
      "entropy": 5.361872243881225,
      "epoch": 0.7550432276657061,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004951026741458243,
      "loss": 5.0833,
      "mean_token_accuracy": 0.2136737063527107,
      "num_tokens": 18038205.0,
      "step": 7860
    },
    {
      "entropy": 5.354436206817627,
      "epoch": 0.75552353506244,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004950955576058705,
      "loss": 5.1272,
      "mean_token_accuracy": 0.21078554093837737,
      "num_tokens": 18049223.0,
      "step": 7865
    },
    {
      "entropy": 5.32838134765625,
      "epoch": 0.7560038424591738,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004950884359559513,
      "loss": 5.0758,
      "mean_token_accuracy": 0.2092729315161705,
      "num_tokens": 18061855.0,
      "step": 7870
    },
    {
      "entropy": 5.297578048706055,
      "epoch": 0.7564841498559077,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004950813091962324,
      "loss": 5.0706,
      "mean_token_accuracy": 0.2168395772576332,
      "num_tokens": 18073969.0,
      "step": 7875
    },
    {
      "entropy": 5.322752189636231,
      "epoch": 0.7569644572526417,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004950741773268788,
      "loss": 5.0274,
      "mean_token_accuracy": 0.21702370494604112,
      "num_tokens": 18085786.0,
      "step": 7880
    },
    {
      "entropy": 5.3562328815460205,
      "epoch": 0.7574447646493756,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004950670403480562,
      "loss": 5.0405,
      "mean_token_accuracy": 0.21527558267116548,
      "num_tokens": 18096263.0,
      "step": 7885
    },
    {
      "entropy": 5.351226139068603,
      "epoch": 0.7579250720461095,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004950598982599306,
      "loss": 5.1495,
      "mean_token_accuracy": 0.20090456008911134,
      "num_tokens": 18108769.0,
      "step": 7890
    },
    {
      "entropy": 5.335327768325806,
      "epoch": 0.7584053794428435,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004950527510626675,
      "loss": 5.1385,
      "mean_token_accuracy": 0.20942235589027405,
      "num_tokens": 18120255.0,
      "step": 7895
    },
    {
      "entropy": 5.3293149948120115,
      "epoch": 0.7588856868395774,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004950455987564329,
      "loss": 5.1093,
      "mean_token_accuracy": 0.2142955929040909,
      "num_tokens": 18132016.0,
      "step": 7900
    },
    {
      "entropy": 5.296963500976562,
      "epoch": 0.7593659942363112,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004950384413413931,
      "loss": 5.0254,
      "mean_token_accuracy": 0.21653357595205308,
      "num_tokens": 18144101.0,
      "step": 7905
    },
    {
      "entropy": 5.355410957336426,
      "epoch": 0.7598463016330451,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004950312788177139,
      "loss": 5.0279,
      "mean_token_accuracy": 0.2169403076171875,
      "num_tokens": 18155086.0,
      "step": 7910
    },
    {
      "entropy": 5.31473331451416,
      "epoch": 0.7603266090297791,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004950241111855618,
      "loss": 5.0841,
      "mean_token_accuracy": 0.21038519442081452,
      "num_tokens": 18166902.0,
      "step": 7915
    },
    {
      "entropy": 5.317033672332764,
      "epoch": 0.760806916426513,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004950169384451031,
      "loss": 5.0497,
      "mean_token_accuracy": 0.21716347485780715,
      "num_tokens": 18178258.0,
      "step": 7920
    },
    {
      "entropy": 5.38117036819458,
      "epoch": 0.7612872238232469,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004950097605965045,
      "loss": 5.1052,
      "mean_token_accuracy": 0.2088675618171692,
      "num_tokens": 18190580.0,
      "step": 7925
    },
    {
      "entropy": 5.3288915157318115,
      "epoch": 0.7617675312199808,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004950025776399326,
      "loss": 5.1631,
      "mean_token_accuracy": 0.208085136115551,
      "num_tokens": 18204912.0,
      "step": 7930
    },
    {
      "entropy": 5.334179067611695,
      "epoch": 0.7622478386167147,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000494995389575554,
      "loss": 5.0743,
      "mean_token_accuracy": 0.2075268045067787,
      "num_tokens": 18215129.0,
      "step": 7935
    },
    {
      "entropy": 5.371470260620117,
      "epoch": 0.7627281460134486,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004949881964035357,
      "loss": 5.1858,
      "mean_token_accuracy": 0.20329192280769348,
      "num_tokens": 18225640.0,
      "step": 7940
    },
    {
      "entropy": 5.311095714569092,
      "epoch": 0.7632084534101825,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004949809981240448,
      "loss": 5.0558,
      "mean_token_accuracy": 0.21754969507455826,
      "num_tokens": 18236641.0,
      "step": 7945
    },
    {
      "entropy": 5.255709409713745,
      "epoch": 0.7636887608069164,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004949737947372483,
      "loss": 5.0513,
      "mean_token_accuracy": 0.2097135454416275,
      "num_tokens": 18249244.0,
      "step": 7950
    },
    {
      "entropy": 5.345564603805542,
      "epoch": 0.7641690682036504,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004949665862433134,
      "loss": 5.1059,
      "mean_token_accuracy": 0.20937389880418777,
      "num_tokens": 18261595.0,
      "step": 7955
    },
    {
      "entropy": 5.313157224655152,
      "epoch": 0.7646493756003843,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004949593726424077,
      "loss": 5.0899,
      "mean_token_accuracy": 0.2140410140156746,
      "num_tokens": 18272609.0,
      "step": 7960
    },
    {
      "entropy": 5.390051078796387,
      "epoch": 0.7651296829971181,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004949521539346985,
      "loss": 5.2332,
      "mean_token_accuracy": 0.2105963110923767,
      "num_tokens": 18283886.0,
      "step": 7965
    },
    {
      "entropy": 5.3576515197753904,
      "epoch": 0.765609990393852,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004949449301203533,
      "loss": 5.1474,
      "mean_token_accuracy": 0.21307448148727418,
      "num_tokens": 18294794.0,
      "step": 7970
    },
    {
      "entropy": 5.429322099685669,
      "epoch": 0.766090297790586,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004949377011995399,
      "loss": 5.1347,
      "mean_token_accuracy": 0.2075539469718933,
      "num_tokens": 18306250.0,
      "step": 7975
    },
    {
      "entropy": 5.373021411895752,
      "epoch": 0.7665706051873199,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004949304671724263,
      "loss": 5.108,
      "mean_token_accuracy": 0.21300649046897888,
      "num_tokens": 18317743.0,
      "step": 7980
    },
    {
      "entropy": 5.273615455627441,
      "epoch": 0.7670509125840538,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004949232280391802,
      "loss": 4.9306,
      "mean_token_accuracy": 0.21797778606414794,
      "num_tokens": 18328287.0,
      "step": 7985
    },
    {
      "entropy": 5.350247526168824,
      "epoch": 0.7675312199807877,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004949159837999698,
      "loss": 5.1513,
      "mean_token_accuracy": 0.2110004484653473,
      "num_tokens": 18341354.0,
      "step": 7990
    },
    {
      "entropy": 5.406481552124023,
      "epoch": 0.7680115273775217,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004949087344549633,
      "loss": 5.1524,
      "mean_token_accuracy": 0.21035605370998384,
      "num_tokens": 18353822.0,
      "step": 7995
    },
    {
      "entropy": 5.39883861541748,
      "epoch": 0.7684918347742555,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000494901480004329,
      "loss": 5.0851,
      "mean_token_accuracy": 0.20871742367744445,
      "num_tokens": 18365491.0,
      "step": 8000
    },
    {
      "entropy": 5.236415958404541,
      "epoch": 0.7689721421709894,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004948942204482351,
      "loss": 5.0515,
      "mean_token_accuracy": 0.2154657155275345,
      "num_tokens": 18376329.0,
      "step": 8005
    },
    {
      "entropy": 5.38602843284607,
      "epoch": 0.7694524495677233,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004948869557868506,
      "loss": 5.1465,
      "mean_token_accuracy": 0.2062172085046768,
      "num_tokens": 18387105.0,
      "step": 8010
    },
    {
      "entropy": 5.319233179092407,
      "epoch": 0.7699327569644573,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004948796860203439,
      "loss": 5.0812,
      "mean_token_accuracy": 0.2077920764684677,
      "num_tokens": 18398113.0,
      "step": 8015
    },
    {
      "entropy": 5.320543098449707,
      "epoch": 0.7704130643611912,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004948724111488838,
      "loss": 5.058,
      "mean_token_accuracy": 0.21251536756753922,
      "num_tokens": 18409315.0,
      "step": 8020
    },
    {
      "entropy": 5.274054956436157,
      "epoch": 0.770893371757925,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004948651311726391,
      "loss": 5.0931,
      "mean_token_accuracy": 0.2064313530921936,
      "num_tokens": 18421110.0,
      "step": 8025
    },
    {
      "entropy": 5.385641431808471,
      "epoch": 0.7713736791546589,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004948578460917789,
      "loss": 5.1648,
      "mean_token_accuracy": 0.2025717467069626,
      "num_tokens": 18433145.0,
      "step": 8030
    },
    {
      "entropy": 5.3862261295318605,
      "epoch": 0.7718539865513929,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004948505559064725,
      "loss": 5.105,
      "mean_token_accuracy": 0.21000165343284607,
      "num_tokens": 18444689.0,
      "step": 8035
    },
    {
      "entropy": 5.302114629745484,
      "epoch": 0.7723342939481268,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004948432606168889,
      "loss": 5.0482,
      "mean_token_accuracy": 0.21657320708036423,
      "num_tokens": 18455943.0,
      "step": 8040
    },
    {
      "entropy": 5.251695394515991,
      "epoch": 0.7728146013448607,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004948359602231976,
      "loss": 5.0193,
      "mean_token_accuracy": 0.21835987120866776,
      "num_tokens": 18466763.0,
      "step": 8045
    },
    {
      "entropy": 5.3209226608276365,
      "epoch": 0.7732949087415946,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004948286547255681,
      "loss": 5.1497,
      "mean_token_accuracy": 0.2014711394906044,
      "num_tokens": 18477647.0,
      "step": 8050
    },
    {
      "entropy": 5.350843143463135,
      "epoch": 0.7737752161383286,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00049482134412417,
      "loss": 5.0313,
      "mean_token_accuracy": 0.22013407796621323,
      "num_tokens": 18488465.0,
      "step": 8055
    },
    {
      "entropy": 5.40940113067627,
      "epoch": 0.7742555235350624,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000494814028419173,
      "loss": 5.1593,
      "mean_token_accuracy": 0.21280452907085418,
      "num_tokens": 18500705.0,
      "step": 8060
    },
    {
      "entropy": 5.307103443145752,
      "epoch": 0.7747358309317963,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000494806707610747,
      "loss": 5.0691,
      "mean_token_accuracy": 0.21186197698116302,
      "num_tokens": 18512253.0,
      "step": 8065
    },
    {
      "entropy": 5.310587644577026,
      "epoch": 0.7752161383285303,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000494799381699062,
      "loss": 5.0671,
      "mean_token_accuracy": 0.21496582329273223,
      "num_tokens": 18524231.0,
      "step": 8070
    },
    {
      "entropy": 5.360444974899292,
      "epoch": 0.7756964457252642,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004947920506842879,
      "loss": 5.1154,
      "mean_token_accuracy": 0.213245165348053,
      "num_tokens": 18535824.0,
      "step": 8075
    },
    {
      "entropy": 5.388378047943116,
      "epoch": 0.7761767531219981,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004947847145665951,
      "loss": 5.1664,
      "mean_token_accuracy": 0.2116045042872429,
      "num_tokens": 18547957.0,
      "step": 8080
    },
    {
      "entropy": 5.3663976192474365,
      "epoch": 0.776657060518732,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004947773733461539,
      "loss": 5.1064,
      "mean_token_accuracy": 0.21368202865123748,
      "num_tokens": 18559297.0,
      "step": 8085
    },
    {
      "entropy": 5.364014577865601,
      "epoch": 0.777137367915466,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004947700270231347,
      "loss": 5.1714,
      "mean_token_accuracy": 0.20959067940711976,
      "num_tokens": 18570264.0,
      "step": 8090
    },
    {
      "entropy": 5.347544145584107,
      "epoch": 0.7776176753121998,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004947626755977079,
      "loss": 5.0572,
      "mean_token_accuracy": 0.21751748025417328,
      "num_tokens": 18582417.0,
      "step": 8095
    },
    {
      "entropy": 5.3305253982543945,
      "epoch": 0.7780979827089337,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004947553190700444,
      "loss": 5.0748,
      "mean_token_accuracy": 0.20772644877433777,
      "num_tokens": 18593703.0,
      "step": 8100
    },
    {
      "entropy": 5.285561513900757,
      "epoch": 0.7785782901056676,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000494747957440315,
      "loss": 5.0576,
      "mean_token_accuracy": 0.21156463921070098,
      "num_tokens": 18605050.0,
      "step": 8105
    },
    {
      "entropy": 5.365184020996094,
      "epoch": 0.7790585975024016,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004947405907086905,
      "loss": 5.1108,
      "mean_token_accuracy": 0.20796026289463043,
      "num_tokens": 18616962.0,
      "step": 8110
    },
    {
      "entropy": 5.360433864593506,
      "epoch": 0.7795389048991355,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004947332188753419,
      "loss": 5.0873,
      "mean_token_accuracy": 0.2115780830383301,
      "num_tokens": 18628640.0,
      "step": 8115
    },
    {
      "entropy": 5.315173006057739,
      "epoch": 0.7800192122958693,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004947258419404405,
      "loss": 5.1007,
      "mean_token_accuracy": 0.2134520173072815,
      "num_tokens": 18640488.0,
      "step": 8120
    },
    {
      "entropy": 5.311030435562134,
      "epoch": 0.7804995196926032,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004947184599041576,
      "loss": 5.0696,
      "mean_token_accuracy": 0.2179467037320137,
      "num_tokens": 18652095.0,
      "step": 8125
    },
    {
      "entropy": 5.328585624694824,
      "epoch": 0.7809798270893372,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004947110727666644,
      "loss": 5.0619,
      "mean_token_accuracy": 0.21874432563781737,
      "num_tokens": 18662767.0,
      "step": 8130
    },
    {
      "entropy": 5.305939245223999,
      "epoch": 0.7814601344860711,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004947036805281325,
      "loss": 5.1146,
      "mean_token_accuracy": 0.2128242924809456,
      "num_tokens": 18673568.0,
      "step": 8135
    },
    {
      "entropy": 5.347345685958862,
      "epoch": 0.781940441882805,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004946962831887336,
      "loss": 5.0637,
      "mean_token_accuracy": 0.21972844749689102,
      "num_tokens": 18684456.0,
      "step": 8140
    },
    {
      "entropy": 5.34295015335083,
      "epoch": 0.7824207492795389,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004946888807486393,
      "loss": 5.1296,
      "mean_token_accuracy": 0.2074110984802246,
      "num_tokens": 18695942.0,
      "step": 8145
    },
    {
      "entropy": 5.397829580307007,
      "epoch": 0.7829010566762729,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004946814732080214,
      "loss": 5.1574,
      "mean_token_accuracy": 0.21366949677467345,
      "num_tokens": 18708110.0,
      "step": 8150
    },
    {
      "entropy": 5.388080692291259,
      "epoch": 0.7833813640730067,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004946740605670523,
      "loss": 5.1447,
      "mean_token_accuracy": 0.21066084057092666,
      "num_tokens": 18718864.0,
      "step": 8155
    },
    {
      "entropy": 5.281895637512207,
      "epoch": 0.7838616714697406,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004946666428259037,
      "loss": 5.012,
      "mean_token_accuracy": 0.21510991901159288,
      "num_tokens": 18730338.0,
      "step": 8160
    },
    {
      "entropy": 5.372270345687866,
      "epoch": 0.7843419788664745,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004946592199847478,
      "loss": 5.1802,
      "mean_token_accuracy": 0.206123685836792,
      "num_tokens": 18743433.0,
      "step": 8165
    },
    {
      "entropy": 5.380597448348999,
      "epoch": 0.7848222862632085,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004946517920437571,
      "loss": 5.0675,
      "mean_token_accuracy": 0.21284492760896684,
      "num_tokens": 18754631.0,
      "step": 8170
    },
    {
      "entropy": 5.294105005264282,
      "epoch": 0.7853025936599424,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004946443590031041,
      "loss": 5.0741,
      "mean_token_accuracy": 0.21417883336544036,
      "num_tokens": 18766118.0,
      "step": 8175
    },
    {
      "entropy": 5.352212285995483,
      "epoch": 0.7857829010566763,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004946369208629613,
      "loss": 5.1001,
      "mean_token_accuracy": 0.21100191920995712,
      "num_tokens": 18777743.0,
      "step": 8180
    },
    {
      "entropy": 5.308093643188476,
      "epoch": 0.7862632084534101,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004946294776235013,
      "loss": 5.0461,
      "mean_token_accuracy": 0.22079116553068162,
      "num_tokens": 18789684.0,
      "step": 8185
    },
    {
      "entropy": 5.2396527290344235,
      "epoch": 0.7867435158501441,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004946220292848971,
      "loss": 5.0244,
      "mean_token_accuracy": 0.21681814789772033,
      "num_tokens": 18801088.0,
      "step": 8190
    },
    {
      "entropy": 5.369229555130005,
      "epoch": 0.787223823246878,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004946145758473214,
      "loss": 5.1275,
      "mean_token_accuracy": 0.20462090522050858,
      "num_tokens": 18812478.0,
      "step": 8195
    },
    {
      "entropy": 5.516317081451416,
      "epoch": 0.7877041306436119,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004946071173109475,
      "loss": 5.1289,
      "mean_token_accuracy": 0.20881302058696746,
      "num_tokens": 18823573.0,
      "step": 8200
    },
    {
      "entropy": 5.305065679550171,
      "epoch": 0.7881844380403458,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004945996536759484,
      "loss": 5.0398,
      "mean_token_accuracy": 0.2163756161928177,
      "num_tokens": 18833880.0,
      "step": 8205
    },
    {
      "entropy": 5.273091793060303,
      "epoch": 0.7886647454370798,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004945921849424974,
      "loss": 5.0074,
      "mean_token_accuracy": 0.2144491642713547,
      "num_tokens": 18844367.0,
      "step": 8210
    },
    {
      "entropy": 5.300557708740234,
      "epoch": 0.7891450528338136,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004945847111107679,
      "loss": 5.0638,
      "mean_token_accuracy": 0.21261599957942962,
      "num_tokens": 18856309.0,
      "step": 8215
    },
    {
      "entropy": 5.315127229690551,
      "epoch": 0.7896253602305475,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004945772321809334,
      "loss": 5.0261,
      "mean_token_accuracy": 0.21350652426481248,
      "num_tokens": 18868873.0,
      "step": 8220
    },
    {
      "entropy": 5.25799298286438,
      "epoch": 0.7901056676272814,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004945697481531677,
      "loss": 5.0016,
      "mean_token_accuracy": 0.21754464954137803,
      "num_tokens": 18879953.0,
      "step": 8225
    },
    {
      "entropy": 5.383896112442017,
      "epoch": 0.7905859750240154,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004945622590276443,
      "loss": 5.1434,
      "mean_token_accuracy": 0.2055958777666092,
      "num_tokens": 18892220.0,
      "step": 8230
    },
    {
      "entropy": 5.326837396621704,
      "epoch": 0.7910662824207493,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004945547648045373,
      "loss": 5.0626,
      "mean_token_accuracy": 0.21681371182203293,
      "num_tokens": 18903261.0,
      "step": 8235
    },
    {
      "entropy": 5.32203426361084,
      "epoch": 0.7915465898174832,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004945472654840206,
      "loss": 5.1122,
      "mean_token_accuracy": 0.2147361472249031,
      "num_tokens": 18913702.0,
      "step": 8240
    },
    {
      "entropy": 5.3890540599823,
      "epoch": 0.7920268972142172,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004945397610662683,
      "loss": 5.1455,
      "mean_token_accuracy": 0.20847590863704682,
      "num_tokens": 18925024.0,
      "step": 8245
    },
    {
      "entropy": 5.3334733009338375,
      "epoch": 0.792507204610951,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004945322515514547,
      "loss": 5.0534,
      "mean_token_accuracy": 0.22270526587963105,
      "num_tokens": 18935939.0,
      "step": 8250
    },
    {
      "entropy": 5.319299554824829,
      "epoch": 0.7929875120076849,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000494524736939754,
      "loss": 5.0327,
      "mean_token_accuracy": 0.21452128887176514,
      "num_tokens": 18946091.0,
      "step": 8255
    },
    {
      "entropy": 5.421309566497802,
      "epoch": 0.7934678194044188,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004945172172313408,
      "loss": 5.1064,
      "mean_token_accuracy": 0.2051161289215088,
      "num_tokens": 18957471.0,
      "step": 8260
    },
    {
      "entropy": 5.43027548789978,
      "epoch": 0.7939481268011528,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004945096924263896,
      "loss": 5.2325,
      "mean_token_accuracy": 0.20539658516645432,
      "num_tokens": 18969462.0,
      "step": 8265
    },
    {
      "entropy": 5.317670583724976,
      "epoch": 0.7944284341978867,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004945021625250753,
      "loss": 5.0804,
      "mean_token_accuracy": 0.21442054808139802,
      "num_tokens": 18980228.0,
      "step": 8270
    },
    {
      "entropy": 5.445771312713623,
      "epoch": 0.7949087415946205,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004944946275275724,
      "loss": 5.2291,
      "mean_token_accuracy": 0.20927242785692216,
      "num_tokens": 18991661.0,
      "step": 8275
    },
    {
      "entropy": 5.3949380874633786,
      "epoch": 0.7953890489913544,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004944870874340561,
      "loss": 5.1204,
      "mean_token_accuracy": 0.21041174679994584,
      "num_tokens": 19003082.0,
      "step": 8280
    },
    {
      "entropy": 5.321384048461914,
      "epoch": 0.7958693563880884,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004944795422447013,
      "loss": 4.9948,
      "mean_token_accuracy": 0.21227362304925917,
      "num_tokens": 19015408.0,
      "step": 8285
    },
    {
      "entropy": 5.326467084884643,
      "epoch": 0.7963496637848223,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004944719919596835,
      "loss": 5.0648,
      "mean_token_accuracy": 0.21030440479516982,
      "num_tokens": 19027246.0,
      "step": 8290
    },
    {
      "entropy": 5.340162897109986,
      "epoch": 0.7968299711815562,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004944644365791776,
      "loss": 5.0795,
      "mean_token_accuracy": 0.21516409665346145,
      "num_tokens": 19038126.0,
      "step": 8295
    },
    {
      "entropy": 5.280876731872558,
      "epoch": 0.7973102785782901,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000494456876103359,
      "loss": 5.0353,
      "mean_token_accuracy": 0.21743886917829514,
      "num_tokens": 19047998.0,
      "step": 8300
    },
    {
      "entropy": 5.3070981979370115,
      "epoch": 0.7977905859750241,
      "grad_norm": 1.0,
      "learning_rate": 0.0004944493105324035,
      "loss": 5.1044,
      "mean_token_accuracy": 0.21068391352891921,
      "num_tokens": 19058969.0,
      "step": 8305
    },
    {
      "entropy": 5.3324806690216064,
      "epoch": 0.7982708933717579,
      "grad_norm": 0.875,
      "learning_rate": 0.0004944417398664866,
      "loss": 5.0357,
      "mean_token_accuracy": 0.21973257660865783,
      "num_tokens": 19070314.0,
      "step": 8310
    },
    {
      "entropy": 5.366440010070801,
      "epoch": 0.7987512007684918,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004944341641057843,
      "loss": 5.0683,
      "mean_token_accuracy": 0.2162510111927986,
      "num_tokens": 19082575.0,
      "step": 8315
    },
    {
      "entropy": 5.225700712203979,
      "epoch": 0.7992315081652257,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004944265832504721,
      "loss": 5.0348,
      "mean_token_accuracy": 0.21975623071193695,
      "num_tokens": 19095719.0,
      "step": 8320
    },
    {
      "entropy": 5.343756818771363,
      "epoch": 0.7997118155619597,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004944189973007262,
      "loss": 5.0369,
      "mean_token_accuracy": 0.2150997281074524,
      "num_tokens": 19106894.0,
      "step": 8325
    },
    {
      "entropy": 5.297736692428589,
      "epoch": 0.8001921229586936,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004944114062567229,
      "loss": 5.051,
      "mean_token_accuracy": 0.20648773461580278,
      "num_tokens": 19118206.0,
      "step": 8330
    },
    {
      "entropy": 5.372063493728637,
      "epoch": 0.8006724303554275,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004944038101186381,
      "loss": 5.1117,
      "mean_token_accuracy": 0.20490544736385347,
      "num_tokens": 19129613.0,
      "step": 8335
    },
    {
      "entropy": 5.26626672744751,
      "epoch": 0.8011527377521613,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004943962088866483,
      "loss": 5.0392,
      "mean_token_accuracy": 0.21679096966981887,
      "num_tokens": 19142214.0,
      "step": 8340
    },
    {
      "entropy": 5.345350170135498,
      "epoch": 0.8016330451488953,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004943886025609301,
      "loss": 5.1491,
      "mean_token_accuracy": 0.2101259633898735,
      "num_tokens": 19153906.0,
      "step": 8345
    },
    {
      "entropy": 5.359352874755859,
      "epoch": 0.8021133525456292,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00049438099114166,
      "loss": 5.0315,
      "mean_token_accuracy": 0.20551080554723739,
      "num_tokens": 19163677.0,
      "step": 8350
    },
    {
      "entropy": 5.357611560821534,
      "epoch": 0.8025936599423631,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004943733746290147,
      "loss": 5.1017,
      "mean_token_accuracy": 0.21362351924180983,
      "num_tokens": 19175240.0,
      "step": 8355
    },
    {
      "entropy": 5.287661933898926,
      "epoch": 0.803073967339097,
      "grad_norm": 0.984375,
      "learning_rate": 0.000494365753023171,
      "loss": 4.9527,
      "mean_token_accuracy": 0.21951040923595427,
      "num_tokens": 19185529.0,
      "step": 8360
    },
    {
      "entropy": 5.259682273864746,
      "epoch": 0.803554274735831,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004943581263243059,
      "loss": 5.022,
      "mean_token_accuracy": 0.21926667541265488,
      "num_tokens": 19197153.0,
      "step": 8365
    },
    {
      "entropy": 5.367073440551758,
      "epoch": 0.8040345821325648,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004943504945325965,
      "loss": 5.1829,
      "mean_token_accuracy": 0.20731729716062547,
      "num_tokens": 19208439.0,
      "step": 8370
    },
    {
      "entropy": 5.293757677078247,
      "epoch": 0.8045148895292987,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004943428576482198,
      "loss": 5.0241,
      "mean_token_accuracy": 0.2193980723619461,
      "num_tokens": 19218570.0,
      "step": 8375
    },
    {
      "entropy": 5.3497954368591305,
      "epoch": 0.8049951969260326,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004943352156713535,
      "loss": 5.0371,
      "mean_token_accuracy": 0.21767054349184037,
      "num_tokens": 19229080.0,
      "step": 8380
    },
    {
      "entropy": 5.340333366394043,
      "epoch": 0.8054755043227666,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004943275686021747,
      "loss": 5.1181,
      "mean_token_accuracy": 0.21056045442819596,
      "num_tokens": 19239299.0,
      "step": 8385
    },
    {
      "entropy": 5.373806667327881,
      "epoch": 0.8059558117195005,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000494319916440861,
      "loss": 5.0632,
      "mean_token_accuracy": 0.21684606969356537,
      "num_tokens": 19250691.0,
      "step": 8390
    },
    {
      "entropy": 5.350543403625489,
      "epoch": 0.8064361191162344,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004943122591875901,
      "loss": 5.041,
      "mean_token_accuracy": 0.21793127954006195,
      "num_tokens": 19260868.0,
      "step": 8395
    },
    {
      "entropy": 5.397047519683838,
      "epoch": 0.8069164265129684,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004943045968425398,
      "loss": 5.1504,
      "mean_token_accuracy": 0.20332364588975907,
      "num_tokens": 19271924.0,
      "step": 8400
    },
    {
      "entropy": 5.302755308151245,
      "epoch": 0.8073967339097022,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004942969294058878,
      "loss": 5.0851,
      "mean_token_accuracy": 0.21210103631019592,
      "num_tokens": 19283340.0,
      "step": 8405
    },
    {
      "entropy": 5.391398572921753,
      "epoch": 0.8078770413064361,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004942892568778125,
      "loss": 5.1527,
      "mean_token_accuracy": 0.20925139933824538,
      "num_tokens": 19294903.0,
      "step": 8410
    },
    {
      "entropy": 5.326468420028687,
      "epoch": 0.80835734870317,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004942815792584917,
      "loss": 5.0379,
      "mean_token_accuracy": 0.21071590185165406,
      "num_tokens": 19305670.0,
      "step": 8415
    },
    {
      "entropy": 5.340024852752686,
      "epoch": 0.808837656099904,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004942738965481038,
      "loss": 5.0487,
      "mean_token_accuracy": 0.21396310329437257,
      "num_tokens": 19317043.0,
      "step": 8420
    },
    {
      "entropy": 5.35851526260376,
      "epoch": 0.8093179634966379,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004942662087468272,
      "loss": 5.1161,
      "mean_token_accuracy": 0.21393977403640746,
      "num_tokens": 19329161.0,
      "step": 8425
    },
    {
      "entropy": 5.365776824951172,
      "epoch": 0.8097982708933718,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004942585158548402,
      "loss": 5.0951,
      "mean_token_accuracy": 0.21492688953876496,
      "num_tokens": 19340706.0,
      "step": 8430
    },
    {
      "entropy": 5.351538419723511,
      "epoch": 0.8102785782901056,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004942508178723214,
      "loss": 5.0695,
      "mean_token_accuracy": 0.21012449413537979,
      "num_tokens": 19352422.0,
      "step": 8435
    },
    {
      "entropy": 5.34750919342041,
      "epoch": 0.8107588856868396,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004942431147994499,
      "loss": 5.1138,
      "mean_token_accuracy": 0.21097689121961594,
      "num_tokens": 19364109.0,
      "step": 8440
    },
    {
      "entropy": 5.36461877822876,
      "epoch": 0.8112391930835735,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004942354066364042,
      "loss": 5.0983,
      "mean_token_accuracy": 0.20818905234336854,
      "num_tokens": 19377386.0,
      "step": 8445
    },
    {
      "entropy": 5.352046632766724,
      "epoch": 0.8117195004803074,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004942276933833634,
      "loss": 5.0407,
      "mean_token_accuracy": 0.21737380027770997,
      "num_tokens": 19388453.0,
      "step": 8450
    },
    {
      "entropy": 5.341142082214356,
      "epoch": 0.8121998078770413,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004942199750405064,
      "loss": 5.149,
      "mean_token_accuracy": 0.20875073075294495,
      "num_tokens": 19402240.0,
      "step": 8455
    },
    {
      "entropy": 5.292131137847901,
      "epoch": 0.8126801152737753,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004942122516080127,
      "loss": 5.0706,
      "mean_token_accuracy": 0.21504217237234116,
      "num_tokens": 19414260.0,
      "step": 8460
    },
    {
      "entropy": 5.281808710098266,
      "epoch": 0.8131604226705091,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004942045230860614,
      "loss": 4.9903,
      "mean_token_accuracy": 0.22041517049074172,
      "num_tokens": 19426356.0,
      "step": 8465
    },
    {
      "entropy": 5.3594505310058596,
      "epoch": 0.813640730067243,
      "grad_norm": 0.875,
      "learning_rate": 0.0004941967894748319,
      "loss": 5.0845,
      "mean_token_accuracy": 0.21201793402433394,
      "num_tokens": 19438881.0,
      "step": 8470
    },
    {
      "entropy": 5.292220973968506,
      "epoch": 0.8141210374639769,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004941890507745039,
      "loss": 5.0323,
      "mean_token_accuracy": 0.21854334622621535,
      "num_tokens": 19450109.0,
      "step": 8475
    },
    {
      "entropy": 5.3408753871917725,
      "epoch": 0.8146013448607109,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004941813069852569,
      "loss": 5.0583,
      "mean_token_accuracy": 0.21378877460956575,
      "num_tokens": 19461155.0,
      "step": 8480
    },
    {
      "entropy": 5.429159784317017,
      "epoch": 0.8150816522574448,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004941735581072708,
      "loss": 5.2237,
      "mean_token_accuracy": 0.2097587302327156,
      "num_tokens": 19474364.0,
      "step": 8485
    },
    {
      "entropy": 5.29606261253357,
      "epoch": 0.8155619596541787,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004941658041407255,
      "loss": 5.0262,
      "mean_token_accuracy": 0.22532742768526076,
      "num_tokens": 19485271.0,
      "step": 8490
    },
    {
      "entropy": 5.382001733779907,
      "epoch": 0.8160422670509125,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000494158045085801,
      "loss": 5.2056,
      "mean_token_accuracy": 0.20260287970304489,
      "num_tokens": 19497117.0,
      "step": 8495
    },
    {
      "entropy": 5.349798107147217,
      "epoch": 0.8165225744476465,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004941502809426776,
      "loss": 4.9768,
      "mean_token_accuracy": 0.2214510425925255,
      "num_tokens": 19507035.0,
      "step": 8500
    },
    {
      "entropy": 5.278189277648925,
      "epoch": 0.8170028818443804,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004941425117115354,
      "loss": 4.9856,
      "mean_token_accuracy": 0.21761000901460648,
      "num_tokens": 19517791.0,
      "step": 8505
    },
    {
      "entropy": 5.323719310760498,
      "epoch": 0.8174831892411143,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004941347373925547,
      "loss": 5.061,
      "mean_token_accuracy": 0.21803533285856247,
      "num_tokens": 19529577.0,
      "step": 8510
    },
    {
      "entropy": 5.38933539390564,
      "epoch": 0.8179634966378482,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004941269579859161,
      "loss": 5.165,
      "mean_token_accuracy": 0.21070607900619506,
      "num_tokens": 19542355.0,
      "step": 8515
    },
    {
      "entropy": 5.370808076858521,
      "epoch": 0.8184438040345822,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004941191734918002,
      "loss": 5.108,
      "mean_token_accuracy": 0.20824917107820512,
      "num_tokens": 19553819.0,
      "step": 8520
    },
    {
      "entropy": 5.4003965854644775,
      "epoch": 0.818924111431316,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000494111383910388,
      "loss": 5.1297,
      "mean_token_accuracy": 0.20304519236087798,
      "num_tokens": 19565816.0,
      "step": 8525
    },
    {
      "entropy": 5.325316286087036,
      "epoch": 0.8194044188280499,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004941035892418597,
      "loss": 5.0478,
      "mean_token_accuracy": 0.21650688648223876,
      "num_tokens": 19577991.0,
      "step": 8530
    },
    {
      "entropy": 5.290628480911255,
      "epoch": 0.8198847262247838,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004940957894863968,
      "loss": 5.0424,
      "mean_token_accuracy": 0.2112143099308014,
      "num_tokens": 19590225.0,
      "step": 8535
    },
    {
      "entropy": 5.289161825180054,
      "epoch": 0.8203650336215178,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004940879846441804,
      "loss": 5.0784,
      "mean_token_accuracy": 0.21836695820093155,
      "num_tokens": 19602966.0,
      "step": 8540
    },
    {
      "entropy": 5.383546352386475,
      "epoch": 0.8208453410182517,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004940801747153914,
      "loss": 5.0633,
      "mean_token_accuracy": 0.21092184334993364,
      "num_tokens": 19613148.0,
      "step": 8545
    },
    {
      "entropy": 5.220239496231079,
      "epoch": 0.8213256484149856,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004940723597002113,
      "loss": 4.9199,
      "mean_token_accuracy": 0.22161339223384857,
      "num_tokens": 19624361.0,
      "step": 8550
    },
    {
      "entropy": 5.244068050384522,
      "epoch": 0.8218059558117194,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004940645395988216,
      "loss": 5.0319,
      "mean_token_accuracy": 0.21759173423051834,
      "num_tokens": 19635915.0,
      "step": 8555
    },
    {
      "entropy": 5.35276837348938,
      "epoch": 0.8222862632084534,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004940567144114036,
      "loss": 5.0604,
      "mean_token_accuracy": 0.21141322255134581,
      "num_tokens": 19646701.0,
      "step": 8560
    },
    {
      "entropy": 5.42860860824585,
      "epoch": 0.8227665706051873,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004940488841381393,
      "loss": 5.169,
      "mean_token_accuracy": 0.20708441883325576,
      "num_tokens": 19657804.0,
      "step": 8565
    },
    {
      "entropy": 5.310262870788574,
      "epoch": 0.8232468780019212,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004940410487792103,
      "loss": 5.1565,
      "mean_token_accuracy": 0.21374066174030304,
      "num_tokens": 19670202.0,
      "step": 8570
    },
    {
      "entropy": 5.219448041915894,
      "epoch": 0.8237271853986552,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004940332083347986,
      "loss": 4.9423,
      "mean_token_accuracy": 0.22693694531917571,
      "num_tokens": 19681362.0,
      "step": 8575
    },
    {
      "entropy": 5.3117447853088375,
      "epoch": 0.8242074927953891,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004940253628050861,
      "loss": 5.0848,
      "mean_token_accuracy": 0.20922519862651826,
      "num_tokens": 19693429.0,
      "step": 8580
    },
    {
      "entropy": 5.372284746170044,
      "epoch": 0.824687800192123,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004940175121902552,
      "loss": 5.1206,
      "mean_token_accuracy": 0.21577929854393005,
      "num_tokens": 19704667.0,
      "step": 8585
    },
    {
      "entropy": 5.338471603393555,
      "epoch": 0.8251681075888568,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000494009656490488,
      "loss": 5.0589,
      "mean_token_accuracy": 0.21465186774730682,
      "num_tokens": 19716002.0,
      "step": 8590
    },
    {
      "entropy": 5.337630128860473,
      "epoch": 0.8256484149855908,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004940017957059668,
      "loss": 5.0928,
      "mean_token_accuracy": 0.21340962499380112,
      "num_tokens": 19726626.0,
      "step": 8595
    },
    {
      "entropy": 5.307701444625854,
      "epoch": 0.8261287223823247,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004939939298368742,
      "loss": 5.0509,
      "mean_token_accuracy": 0.21765271723270416,
      "num_tokens": 19738250.0,
      "step": 8600
    },
    {
      "entropy": 5.413993501663208,
      "epoch": 0.8266090297790586,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004939860588833929,
      "loss": 5.1388,
      "mean_token_accuracy": 0.21273907274007797,
      "num_tokens": 19750119.0,
      "step": 8605
    },
    {
      "entropy": 5.352220487594605,
      "epoch": 0.8270893371757925,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004939781828457055,
      "loss": 5.1136,
      "mean_token_accuracy": 0.208088319003582,
      "num_tokens": 19761709.0,
      "step": 8610
    },
    {
      "entropy": 5.325716400146485,
      "epoch": 0.8275696445725265,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000493970301723995,
      "loss": 5.0936,
      "mean_token_accuracy": 0.20774221569299697,
      "num_tokens": 19773562.0,
      "step": 8615
    },
    {
      "entropy": 5.463512372970581,
      "epoch": 0.8280499519692603,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004939624155184443,
      "loss": 5.1067,
      "mean_token_accuracy": 0.21097196340560914,
      "num_tokens": 19785327.0,
      "step": 8620
    },
    {
      "entropy": 5.377057981491089,
      "epoch": 0.8285302593659942,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004939545242292365,
      "loss": 5.13,
      "mean_token_accuracy": 0.20602378398180007,
      "num_tokens": 19796760.0,
      "step": 8625
    },
    {
      "entropy": 5.3201452732086185,
      "epoch": 0.8290105667627281,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004939466278565547,
      "loss": 5.0565,
      "mean_token_accuracy": 0.21484012007713318,
      "num_tokens": 19807271.0,
      "step": 8630
    },
    {
      "entropy": 5.329145622253418,
      "epoch": 0.8294908741594621,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004939387264005825,
      "loss": 5.0729,
      "mean_token_accuracy": 0.21570000946521758,
      "num_tokens": 19818098.0,
      "step": 8635
    },
    {
      "entropy": 5.31826114654541,
      "epoch": 0.829971181556196,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004939308198615031,
      "loss": 5.0639,
      "mean_token_accuracy": 0.21507036983966826,
      "num_tokens": 19829482.0,
      "step": 8640
    },
    {
      "entropy": 5.362896871566773,
      "epoch": 0.8304514889529299,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004939229082395001,
      "loss": 5.096,
      "mean_token_accuracy": 0.21485564708709717,
      "num_tokens": 19840883.0,
      "step": 8645
    },
    {
      "entropy": 5.263069200515747,
      "epoch": 0.8309317963496637,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004939149915347573,
      "loss": 5.0032,
      "mean_token_accuracy": 0.21648780405521392,
      "num_tokens": 19852619.0,
      "step": 8650
    },
    {
      "entropy": 5.247555780410766,
      "epoch": 0.8314121037463977,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004939070697474585,
      "loss": 4.9613,
      "mean_token_accuracy": 0.215274079144001,
      "num_tokens": 19863316.0,
      "step": 8655
    },
    {
      "entropy": 5.372004413604737,
      "epoch": 0.8318924111431316,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004938991428777875,
      "loss": 5.1523,
      "mean_token_accuracy": 0.20794903934001924,
      "num_tokens": 19873965.0,
      "step": 8660
    },
    {
      "entropy": 5.366307163238526,
      "epoch": 0.8323727185398655,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004938912109259284,
      "loss": 5.0694,
      "mean_token_accuracy": 0.21538155078887938,
      "num_tokens": 19884685.0,
      "step": 8665
    },
    {
      "entropy": 5.293383407592773,
      "epoch": 0.8328530259365994,
      "grad_norm": 0.875,
      "learning_rate": 0.0004938832738920654,
      "loss": 5.0317,
      "mean_token_accuracy": 0.21668114364147187,
      "num_tokens": 19897594.0,
      "step": 8670
    },
    {
      "entropy": 5.323529148101807,
      "epoch": 0.8333333333333334,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004938753317763826,
      "loss": 5.1108,
      "mean_token_accuracy": 0.21189523488283157,
      "num_tokens": 19908093.0,
      "step": 8675
    },
    {
      "entropy": 5.393456840515137,
      "epoch": 0.8338136407300673,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004938673845790646,
      "loss": 5.1359,
      "mean_token_accuracy": 0.20514165461063386,
      "num_tokens": 19920502.0,
      "step": 8680
    },
    {
      "entropy": 5.289888000488281,
      "epoch": 0.8342939481268011,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004938594323002957,
      "loss": 4.9834,
      "mean_token_accuracy": 0.21902598738670348,
      "num_tokens": 19931437.0,
      "step": 8685
    },
    {
      "entropy": 5.280273818969727,
      "epoch": 0.834774255523535,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004938514749402609,
      "loss": 5.0536,
      "mean_token_accuracy": 0.21799136102199554,
      "num_tokens": 19942592.0,
      "step": 8690
    },
    {
      "entropy": 5.3899688720703125,
      "epoch": 0.835254562920269,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004938435124991447,
      "loss": 5.2182,
      "mean_token_accuracy": 0.20100459903478624,
      "num_tokens": 19953802.0,
      "step": 8695
    },
    {
      "entropy": 5.356327676773072,
      "epoch": 0.8357348703170029,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004938355449771318,
      "loss": 5.0738,
      "mean_token_accuracy": 0.21255610138177872,
      "num_tokens": 19964686.0,
      "step": 8700
    },
    {
      "entropy": 5.3338714122772215,
      "epoch": 0.8362151777137368,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004938275723744075,
      "loss": 5.1166,
      "mean_token_accuracy": 0.211995492875576,
      "num_tokens": 19975603.0,
      "step": 8705
    },
    {
      "entropy": 5.376679801940918,
      "epoch": 0.8366954851104706,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004938195946911567,
      "loss": 5.1591,
      "mean_token_accuracy": 0.20652659833431244,
      "num_tokens": 19986205.0,
      "step": 8710
    },
    {
      "entropy": 5.339930486679077,
      "epoch": 0.8371757925072046,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004938116119275645,
      "loss": 4.9791,
      "mean_token_accuracy": 0.21873142421245576,
      "num_tokens": 19997323.0,
      "step": 8715
    },
    {
      "entropy": 5.310747480392456,
      "epoch": 0.8376560999039385,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004938036240838166,
      "loss": 5.0398,
      "mean_token_accuracy": 0.21743457466363908,
      "num_tokens": 20007334.0,
      "step": 8720
    },
    {
      "entropy": 5.385940027236939,
      "epoch": 0.8381364073006724,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004937956311600983,
      "loss": 5.0503,
      "mean_token_accuracy": 0.21304183900356294,
      "num_tokens": 20017476.0,
      "step": 8725
    },
    {
      "entropy": 5.2531270503997805,
      "epoch": 0.8386167146974063,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004937876331565951,
      "loss": 5.0108,
      "mean_token_accuracy": 0.21013785004615784,
      "num_tokens": 20029323.0,
      "step": 8730
    },
    {
      "entropy": 5.329405832290649,
      "epoch": 0.8390970220941403,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004937796300734926,
      "loss": 5.02,
      "mean_token_accuracy": 0.21011769622564316,
      "num_tokens": 20040317.0,
      "step": 8735
    },
    {
      "entropy": 5.3954352855682375,
      "epoch": 0.8395773294908742,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004937716219109769,
      "loss": 5.0857,
      "mean_token_accuracy": 0.21819895803928374,
      "num_tokens": 20050495.0,
      "step": 8740
    },
    {
      "entropy": 5.4144049167633055,
      "epoch": 0.840057636887608,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004937636086692336,
      "loss": 5.1234,
      "mean_token_accuracy": 0.20799389630556106,
      "num_tokens": 20061185.0,
      "step": 8745
    },
    {
      "entropy": 5.314114189147949,
      "epoch": 0.840537944284342,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000493755590348449,
      "loss": 5.0837,
      "mean_token_accuracy": 0.2115689516067505,
      "num_tokens": 20073746.0,
      "step": 8750
    },
    {
      "entropy": 5.373018074035644,
      "epoch": 0.8410182516810759,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004937475669488091,
      "loss": 5.1187,
      "mean_token_accuracy": 0.21122402548789979,
      "num_tokens": 20085667.0,
      "step": 8755
    },
    {
      "entropy": 5.362546777725219,
      "epoch": 0.8414985590778098,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004937395384705004,
      "loss": 4.9992,
      "mean_token_accuracy": 0.2196834459900856,
      "num_tokens": 20096017.0,
      "step": 8760
    },
    {
      "entropy": 5.307758808135986,
      "epoch": 0.8419788664745437,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004937315049137089,
      "loss": 5.0744,
      "mean_token_accuracy": 0.21469683945178986,
      "num_tokens": 20106928.0,
      "step": 8765
    },
    {
      "entropy": 5.298285961151123,
      "epoch": 0.8424591738712777,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004937234662786216,
      "loss": 5.0566,
      "mean_token_accuracy": 0.21256544589996337,
      "num_tokens": 20118768.0,
      "step": 8770
    },
    {
      "entropy": 5.463348627090454,
      "epoch": 0.8429394812680115,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004937154225654246,
      "loss": 5.1956,
      "mean_token_accuracy": 0.20620521157979965,
      "num_tokens": 20129009.0,
      "step": 8775
    },
    {
      "entropy": 5.325447130203247,
      "epoch": 0.8434197886647454,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004937073737743051,
      "loss": 5.077,
      "mean_token_accuracy": 0.21056914031505586,
      "num_tokens": 20141577.0,
      "step": 8780
    },
    {
      "entropy": 5.380077314376831,
      "epoch": 0.8439000960614793,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004936993199054499,
      "loss": 5.1363,
      "mean_token_accuracy": 0.2108604907989502,
      "num_tokens": 20153636.0,
      "step": 8785
    },
    {
      "entropy": 5.303895092010498,
      "epoch": 0.8443804034582133,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004936912609590458,
      "loss": 5.0255,
      "mean_token_accuracy": 0.2143542945384979,
      "num_tokens": 20165719.0,
      "step": 8790
    },
    {
      "entropy": 5.362807846069336,
      "epoch": 0.8448607108549472,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00049368319693528,
      "loss": 5.0929,
      "mean_token_accuracy": 0.2065473571419716,
      "num_tokens": 20178260.0,
      "step": 8795
    },
    {
      "entropy": 5.354299736022949,
      "epoch": 0.8453410182516811,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004936751278343397,
      "loss": 5.1141,
      "mean_token_accuracy": 0.20961180776357652,
      "num_tokens": 20189667.0,
      "step": 8800
    },
    {
      "entropy": 5.346553802490234,
      "epoch": 0.845821325648415,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004936670536564123,
      "loss": 5.061,
      "mean_token_accuracy": 0.2096709728240967,
      "num_tokens": 20201451.0,
      "step": 8805
    },
    {
      "entropy": 5.351399612426758,
      "epoch": 0.8463016330451489,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004936589744016853,
      "loss": 5.1172,
      "mean_token_accuracy": 0.21129563748836516,
      "num_tokens": 20213257.0,
      "step": 8810
    },
    {
      "entropy": 5.354643297195435,
      "epoch": 0.8467819404418828,
      "grad_norm": 0.859375,
      "learning_rate": 0.000493650890070346,
      "loss": 5.0692,
      "mean_token_accuracy": 0.21912156343460082,
      "num_tokens": 20225301.0,
      "step": 8815
    },
    {
      "entropy": 5.35613169670105,
      "epoch": 0.8472622478386167,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004936428006625824,
      "loss": 5.0443,
      "mean_token_accuracy": 0.21862908601760864,
      "num_tokens": 20236183.0,
      "step": 8820
    },
    {
      "entropy": 5.385243892669678,
      "epoch": 0.8477425552353506,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004936347061785823,
      "loss": 5.1429,
      "mean_token_accuracy": 0.20484864860773086,
      "num_tokens": 20248663.0,
      "step": 8825
    },
    {
      "entropy": 5.311046695709228,
      "epoch": 0.8482228626320846,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004936266066185334,
      "loss": 5.0269,
      "mean_token_accuracy": 0.22330467402935028,
      "num_tokens": 20259293.0,
      "step": 8830
    },
    {
      "entropy": 5.3645570278167725,
      "epoch": 0.8487031700288185,
      "grad_norm": 0.875,
      "learning_rate": 0.0004936185019826239,
      "loss": 5.1616,
      "mean_token_accuracy": 0.2026146560907364,
      "num_tokens": 20270601.0,
      "step": 8835
    },
    {
      "entropy": 5.3683113098144535,
      "epoch": 0.8491834774255523,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004936103922710419,
      "loss": 5.1372,
      "mean_token_accuracy": 0.20879282802343369,
      "num_tokens": 20281471.0,
      "step": 8840
    },
    {
      "entropy": 5.311202621459961,
      "epoch": 0.8496637848222862,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004936022774839759,
      "loss": 4.9854,
      "mean_token_accuracy": 0.21537231802940368,
      "num_tokens": 20291941.0,
      "step": 8845
    },
    {
      "entropy": 5.349550771713257,
      "epoch": 0.8501440922190202,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004935941576216141,
      "loss": 5.1171,
      "mean_token_accuracy": 0.2099014401435852,
      "num_tokens": 20302631.0,
      "step": 8850
    },
    {
      "entropy": 5.377071571350098,
      "epoch": 0.8506243996157541,
      "grad_norm": 0.84375,
      "learning_rate": 0.000493586032684145,
      "loss": 5.1725,
      "mean_token_accuracy": 0.2099987491965294,
      "num_tokens": 20314955.0,
      "step": 8855
    },
    {
      "entropy": 5.3678529262542725,
      "epoch": 0.851104707012488,
      "grad_norm": 0.875,
      "learning_rate": 0.0004935779026717573,
      "loss": 5.1001,
      "mean_token_accuracy": 0.21613814532756806,
      "num_tokens": 20326791.0,
      "step": 8860
    },
    {
      "entropy": 5.355668687820435,
      "epoch": 0.8515850144092219,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004935697675846396,
      "loss": 5.0827,
      "mean_token_accuracy": 0.2179255962371826,
      "num_tokens": 20338912.0,
      "step": 8865
    },
    {
      "entropy": 5.334821033477783,
      "epoch": 0.8520653218059558,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004935616274229811,
      "loss": 5.0462,
      "mean_token_accuracy": 0.22144256085157393,
      "num_tokens": 20349414.0,
      "step": 8870
    },
    {
      "entropy": 5.380909490585327,
      "epoch": 0.8525456292026897,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004935534821869705,
      "loss": 5.0837,
      "mean_token_accuracy": 0.215879425406456,
      "num_tokens": 20360346.0,
      "step": 8875
    },
    {
      "entropy": 5.3450206279754635,
      "epoch": 0.8530259365994236,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004935453318767971,
      "loss": 5.0615,
      "mean_token_accuracy": 0.21613271981477739,
      "num_tokens": 20371965.0,
      "step": 8880
    },
    {
      "entropy": 5.297636318206787,
      "epoch": 0.8535062439961575,
      "grad_norm": 0.96875,
      "learning_rate": 0.00049353717649265,
      "loss": 5.0896,
      "mean_token_accuracy": 0.2207036092877388,
      "num_tokens": 20382632.0,
      "step": 8885
    },
    {
      "entropy": 5.329258728027344,
      "epoch": 0.8539865513928915,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004935290160347185,
      "loss": 5.0682,
      "mean_token_accuracy": 0.22044046372175216,
      "num_tokens": 20395796.0,
      "step": 8890
    },
    {
      "entropy": 5.375508737564087,
      "epoch": 0.8544668587896254,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004935208505031922,
      "loss": 5.1047,
      "mean_token_accuracy": 0.21535910665988922,
      "num_tokens": 20407113.0,
      "step": 8895
    },
    {
      "entropy": 5.347590017318725,
      "epoch": 0.8549471661863592,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004935126798982606,
      "loss": 5.1123,
      "mean_token_accuracy": 0.21519984304904938,
      "num_tokens": 20419012.0,
      "step": 8900
    },
    {
      "entropy": 5.356325006484985,
      "epoch": 0.8554274735830932,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004935045042201135,
      "loss": 5.0763,
      "mean_token_accuracy": 0.21092951595783233,
      "num_tokens": 20430337.0,
      "step": 8905
    },
    {
      "entropy": 5.316872453689575,
      "epoch": 0.8559077809798271,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004934963234689407,
      "loss": 5.0824,
      "mean_token_accuracy": 0.20934270471334457,
      "num_tokens": 20441614.0,
      "step": 8910
    },
    {
      "entropy": 5.306456708908081,
      "epoch": 0.856388088376561,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000493488137644932,
      "loss": 4.9574,
      "mean_token_accuracy": 0.22647832781076432,
      "num_tokens": 20453170.0,
      "step": 8915
    },
    {
      "entropy": 5.224303817749023,
      "epoch": 0.8568683957732949,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004934799467482774,
      "loss": 4.9211,
      "mean_token_accuracy": 0.22477201521396636,
      "num_tokens": 20464585.0,
      "step": 8920
    },
    {
      "entropy": 5.303629446029663,
      "epoch": 0.8573487031700289,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004934717507791673,
      "loss": 4.9243,
      "mean_token_accuracy": 0.22721005827188492,
      "num_tokens": 20473903.0,
      "step": 8925
    },
    {
      "entropy": 5.345573472976684,
      "epoch": 0.8578290105667628,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004934635497377919,
      "loss": 5.0853,
      "mean_token_accuracy": 0.21045285016298293,
      "num_tokens": 20484778.0,
      "step": 8930
    },
    {
      "entropy": 5.333237791061402,
      "epoch": 0.8583093179634966,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004934553436243415,
      "loss": 5.1197,
      "mean_token_accuracy": 0.20598161071538926,
      "num_tokens": 20495534.0,
      "step": 8935
    },
    {
      "entropy": 5.438832569122314,
      "epoch": 0.8587896253602305,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004934471324390067,
      "loss": 5.1604,
      "mean_token_accuracy": 0.21332629770040512,
      "num_tokens": 20506853.0,
      "step": 8940
    },
    {
      "entropy": 5.353634452819824,
      "epoch": 0.8592699327569645,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004934389161819783,
      "loss": 5.0921,
      "mean_token_accuracy": 0.21390892565250397,
      "num_tokens": 20519112.0,
      "step": 8945
    },
    {
      "entropy": 5.362328672409058,
      "epoch": 0.8597502401536984,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004934306948534467,
      "loss": 5.0868,
      "mean_token_accuracy": 0.21352950781583785,
      "num_tokens": 20530060.0,
      "step": 8950
    },
    {
      "entropy": 5.339789342880249,
      "epoch": 0.8602305475504323,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004934224684536031,
      "loss": 5.0169,
      "mean_token_accuracy": 0.2173382371664047,
      "num_tokens": 20541802.0,
      "step": 8955
    },
    {
      "entropy": 5.231397294998169,
      "epoch": 0.8607108549471661,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004934142369826382,
      "loss": 4.9326,
      "mean_token_accuracy": 0.22756012678146362,
      "num_tokens": 20552494.0,
      "step": 8960
    },
    {
      "entropy": 5.290261268615723,
      "epoch": 0.8611911623439001,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004934060004407434,
      "loss": 5.0762,
      "mean_token_accuracy": 0.21892891377210616,
      "num_tokens": 20564467.0,
      "step": 8965
    },
    {
      "entropy": 5.341933012008667,
      "epoch": 0.861671469740634,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004933977588281099,
      "loss": 5.1309,
      "mean_token_accuracy": 0.21144873648881912,
      "num_tokens": 20575399.0,
      "step": 8970
    },
    {
      "entropy": 5.396449327468872,
      "epoch": 0.8621517771373679,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004933895121449288,
      "loss": 5.1008,
      "mean_token_accuracy": 0.21717502921819687,
      "num_tokens": 20586212.0,
      "step": 8975
    },
    {
      "entropy": 5.420312595367432,
      "epoch": 0.8626320845341018,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004933812603913917,
      "loss": 5.1118,
      "mean_token_accuracy": 0.21174852550029755,
      "num_tokens": 20597020.0,
      "step": 8980
    },
    {
      "entropy": 5.326293087005615,
      "epoch": 0.8631123919308358,
      "grad_norm": 1.0,
      "learning_rate": 0.0004933730035676903,
      "loss": 5.0552,
      "mean_token_accuracy": 0.21294718086719513,
      "num_tokens": 20609447.0,
      "step": 8985
    },
    {
      "entropy": 5.299590349197388,
      "epoch": 0.8635926993275697,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004933647416740161,
      "loss": 5.037,
      "mean_token_accuracy": 0.214912748336792,
      "num_tokens": 20621780.0,
      "step": 8990
    },
    {
      "entropy": 5.256945419311523,
      "epoch": 0.8640730067243035,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000493356474710561,
      "loss": 4.9902,
      "mean_token_accuracy": 0.21920427978038787,
      "num_tokens": 20632331.0,
      "step": 8995
    },
    {
      "entropy": 5.3006034851074215,
      "epoch": 0.8645533141210374,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000493348202677517,
      "loss": 4.9487,
      "mean_token_accuracy": 0.22219757735729218,
      "num_tokens": 20643357.0,
      "step": 9000
    },
    {
      "epoch": 0.8645533141210374,
      "eval_entropy": 5.13524659280949,
      "eval_loss": 5.074194431304932,
      "eval_mean_token_accuracy": 0.22209948177532501,
      "eval_num_tokens": 20643357.0,
      "eval_runtime": 20.5605,
      "eval_samples_per_second": 1596.02,
      "eval_steps_per_second": 199.509,
      "step": 9000
    },
    {
      "entropy": 5.259588813781738,
      "epoch": 0.8650336215177714,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004933399255750761,
      "loss": 5.0191,
      "mean_token_accuracy": 0.22431397587060928,
      "num_tokens": 20655212.0,
      "step": 9005
    },
    {
      "entropy": 5.192686128616333,
      "epoch": 0.8655139289145053,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004933316434034304,
      "loss": 4.8997,
      "mean_token_accuracy": 0.22778640389442445,
      "num_tokens": 20666467.0,
      "step": 9010
    },
    {
      "entropy": 5.295734882354736,
      "epoch": 0.8659942363112392,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004933233561627723,
      "loss": 5.0117,
      "mean_token_accuracy": 0.221454980969429,
      "num_tokens": 20677709.0,
      "step": 9015
    },
    {
      "entropy": 5.351622247695923,
      "epoch": 0.866474543707973,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004933150638532942,
      "loss": 5.0533,
      "mean_token_accuracy": 0.21002226918935776,
      "num_tokens": 20688880.0,
      "step": 9020
    },
    {
      "entropy": 5.3735559463500975,
      "epoch": 0.866954851104707,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004933067664751885,
      "loss": 5.1915,
      "mean_token_accuracy": 0.20039801746606828,
      "num_tokens": 20701384.0,
      "step": 9025
    },
    {
      "entropy": 5.425477361679077,
      "epoch": 0.8674351585014409,
      "grad_norm": 0.890625,
      "learning_rate": 0.000493298464028648,
      "loss": 5.137,
      "mean_token_accuracy": 0.20833015143871308,
      "num_tokens": 20713736.0,
      "step": 9030
    },
    {
      "entropy": 5.236796140670776,
      "epoch": 0.8679154658981748,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004932901565138653,
      "loss": 4.9564,
      "mean_token_accuracy": 0.22253143638372422,
      "num_tokens": 20725424.0,
      "step": 9035
    },
    {
      "entropy": 5.293302392959594,
      "epoch": 0.8683957732949087,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004932818439310334,
      "loss": 4.9889,
      "mean_token_accuracy": 0.21736721843481063,
      "num_tokens": 20737454.0,
      "step": 9040
    },
    {
      "entropy": 5.397310256958008,
      "epoch": 0.8688760806916427,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004932735262803452,
      "loss": 5.1192,
      "mean_token_accuracy": 0.21272976994514464,
      "num_tokens": 20748882.0,
      "step": 9045
    },
    {
      "entropy": 5.309205389022827,
      "epoch": 0.8693563880883766,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004932652035619939,
      "loss": 4.9973,
      "mean_token_accuracy": 0.22420158237218857,
      "num_tokens": 20759622.0,
      "step": 9050
    },
    {
      "entropy": 5.269955396652222,
      "epoch": 0.8698366954851104,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004932568757761727,
      "loss": 5.0058,
      "mean_token_accuracy": 0.2151384621858597,
      "num_tokens": 20770658.0,
      "step": 9055
    },
    {
      "entropy": 5.369260215759278,
      "epoch": 0.8703170028818443,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004932485429230748,
      "loss": 5.0489,
      "mean_token_accuracy": 0.2149381086230278,
      "num_tokens": 20782176.0,
      "step": 9060
    },
    {
      "entropy": 5.388922500610351,
      "epoch": 0.8707973102785783,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000493240205002894,
      "loss": 5.1333,
      "mean_token_accuracy": 0.20277243703603745,
      "num_tokens": 20793592.0,
      "step": 9065
    },
    {
      "entropy": 5.304251670837402,
      "epoch": 0.8712776176753122,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004932318620158235,
      "loss": 5.0073,
      "mean_token_accuracy": 0.21001942604780197,
      "num_tokens": 20804226.0,
      "step": 9070
    },
    {
      "entropy": 5.3437543392181395,
      "epoch": 0.8717579250720461,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004932235139620574,
      "loss": 5.1428,
      "mean_token_accuracy": 0.20608436763286592,
      "num_tokens": 20815336.0,
      "step": 9075
    },
    {
      "entropy": 5.290644311904908,
      "epoch": 0.8722382324687801,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004932151608417892,
      "loss": 5.0449,
      "mean_token_accuracy": 0.22099872678518295,
      "num_tokens": 20826936.0,
      "step": 9080
    },
    {
      "entropy": 5.329099512100219,
      "epoch": 0.872718539865514,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004932068026552127,
      "loss": 5.0998,
      "mean_token_accuracy": 0.21392923295497895,
      "num_tokens": 20838159.0,
      "step": 9085
    },
    {
      "entropy": 5.324334383010864,
      "epoch": 0.8731988472622478,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004931984394025224,
      "loss": 4.9956,
      "mean_token_accuracy": 0.21418420374393463,
      "num_tokens": 20848605.0,
      "step": 9090
    },
    {
      "entropy": 5.340549945831299,
      "epoch": 0.8736791546589817,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004931900710839123,
      "loss": 5.1031,
      "mean_token_accuracy": 0.2132624238729477,
      "num_tokens": 20860494.0,
      "step": 9095
    },
    {
      "entropy": 5.368777465820313,
      "epoch": 0.8741594620557157,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004931816976995766,
      "loss": 5.102,
      "mean_token_accuracy": 0.21883988678455352,
      "num_tokens": 20871327.0,
      "step": 9100
    },
    {
      "entropy": 5.3373048305511475,
      "epoch": 0.8746397694524496,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004931733192497097,
      "loss": 5.0645,
      "mean_token_accuracy": 0.21759789437055588,
      "num_tokens": 20883032.0,
      "step": 9105
    },
    {
      "entropy": 5.321063995361328,
      "epoch": 0.8751200768491835,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004931649357345062,
      "loss": 5.0924,
      "mean_token_accuracy": 0.20907170623540877,
      "num_tokens": 20894327.0,
      "step": 9110
    },
    {
      "entropy": 5.272947072982788,
      "epoch": 0.8756003842459174,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004931565471541606,
      "loss": 4.968,
      "mean_token_accuracy": 0.22293614894151687,
      "num_tokens": 20905489.0,
      "step": 9115
    },
    {
      "entropy": 5.305966377258301,
      "epoch": 0.8760806916426513,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004931481535088679,
      "loss": 5.0449,
      "mean_token_accuracy": 0.2159699261188507,
      "num_tokens": 20917044.0,
      "step": 9120
    },
    {
      "entropy": 5.3416472434997555,
      "epoch": 0.8765609990393852,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004931397547988229,
      "loss": 5.0285,
      "mean_token_accuracy": 0.22210747748613358,
      "num_tokens": 20928065.0,
      "step": 9125
    },
    {
      "entropy": 5.333930349349975,
      "epoch": 0.8770413064361191,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004931313510242204,
      "loss": 5.0392,
      "mean_token_accuracy": 0.21704798936843872,
      "num_tokens": 20939396.0,
      "step": 9130
    },
    {
      "entropy": 5.413697528839111,
      "epoch": 0.877521613832853,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004931229421852557,
      "loss": 5.1301,
      "mean_token_accuracy": 0.21396100372076035,
      "num_tokens": 20950844.0,
      "step": 9135
    },
    {
      "entropy": 5.355669736862183,
      "epoch": 0.878001921229587,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000493114528282124,
      "loss": 5.0349,
      "mean_token_accuracy": 0.21624603271484374,
      "num_tokens": 20961567.0,
      "step": 9140
    },
    {
      "entropy": 5.394284963607788,
      "epoch": 0.8784822286263209,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004931061093150206,
      "loss": 5.1728,
      "mean_token_accuracy": 0.21613683551549911,
      "num_tokens": 20972762.0,
      "step": 9145
    },
    {
      "entropy": 5.299278783798218,
      "epoch": 0.8789625360230547,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004930976852841409,
      "loss": 5.0397,
      "mean_token_accuracy": 0.2137425348162651,
      "num_tokens": 20983848.0,
      "step": 9150
    },
    {
      "entropy": 5.333186626434326,
      "epoch": 0.8794428434197886,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004930892561896806,
      "loss": 5.0699,
      "mean_token_accuracy": 0.21684213429689408,
      "num_tokens": 20996418.0,
      "step": 9155
    },
    {
      "entropy": 5.315537166595459,
      "epoch": 0.8799231508165226,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004930808220318354,
      "loss": 5.0569,
      "mean_token_accuracy": 0.2223145470023155,
      "num_tokens": 21008714.0,
      "step": 9160
    },
    {
      "entropy": 5.3496985912323,
      "epoch": 0.8804034582132565,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004930723828108012,
      "loss": 5.0208,
      "mean_token_accuracy": 0.21808206737041474,
      "num_tokens": 21020453.0,
      "step": 9165
    },
    {
      "entropy": 5.202108526229859,
      "epoch": 0.8808837656099904,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004930639385267736,
      "loss": 5.0082,
      "mean_token_accuracy": 0.21429841518402098,
      "num_tokens": 21032211.0,
      "step": 9170
    },
    {
      "entropy": 5.348698711395263,
      "epoch": 0.8813640730067243,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000493055489179949,
      "loss": 5.0793,
      "mean_token_accuracy": 0.2088883936405182,
      "num_tokens": 21042995.0,
      "step": 9175
    },
    {
      "entropy": 5.264824342727661,
      "epoch": 0.8818443804034583,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004930470347705234,
      "loss": 4.9505,
      "mean_token_accuracy": 0.22010566145181656,
      "num_tokens": 21054306.0,
      "step": 9180
    },
    {
      "entropy": 5.293511486053466,
      "epoch": 0.8823246878001921,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000493038575298693,
      "loss": 5.117,
      "mean_token_accuracy": 0.2064583718776703,
      "num_tokens": 21067144.0,
      "step": 9185
    },
    {
      "entropy": 5.45893669128418,
      "epoch": 0.882804995196926,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004930301107646545,
      "loss": 5.0934,
      "mean_token_accuracy": 0.20680933445692062,
      "num_tokens": 21078885.0,
      "step": 9190
    },
    {
      "entropy": 5.413471841812134,
      "epoch": 0.8832853025936599,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004930216411686042,
      "loss": 5.1471,
      "mean_token_accuracy": 0.20421212166547775,
      "num_tokens": 21089482.0,
      "step": 9195
    },
    {
      "entropy": 5.397814989089966,
      "epoch": 0.8837656099903939,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004930131665107387,
      "loss": 5.1435,
      "mean_token_accuracy": 0.20462200343608855,
      "num_tokens": 21102046.0,
      "step": 9200
    },
    {
      "entropy": 5.292316246032715,
      "epoch": 0.8842459173871278,
      "grad_norm": 1.03125,
      "learning_rate": 0.000493004686791255,
      "loss": 4.9802,
      "mean_token_accuracy": 0.21848117411136628,
      "num_tokens": 21112827.0,
      "step": 9205
    },
    {
      "entropy": 5.364362525939941,
      "epoch": 0.8847262247838616,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004929962020103496,
      "loss": 5.0978,
      "mean_token_accuracy": 0.2128182590007782,
      "num_tokens": 21123108.0,
      "step": 9210
    },
    {
      "entropy": 5.36696949005127,
      "epoch": 0.8852065321805955,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004929877121682198,
      "loss": 5.0974,
      "mean_token_accuracy": 0.21727408468723297,
      "num_tokens": 21134955.0,
      "step": 9215
    },
    {
      "entropy": 5.30641016960144,
      "epoch": 0.8856868395773295,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004929792172650627,
      "loss": 5.0056,
      "mean_token_accuracy": 0.22059776932001113,
      "num_tokens": 21146415.0,
      "step": 9220
    },
    {
      "entropy": 5.29180006980896,
      "epoch": 0.8861671469740634,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004929707173010753,
      "loss": 5.0263,
      "mean_token_accuracy": 0.2198776051402092,
      "num_tokens": 21158691.0,
      "step": 9225
    },
    {
      "entropy": 5.307462978363037,
      "epoch": 0.8866474543707973,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004929622122764552,
      "loss": 5.0497,
      "mean_token_accuracy": 0.21131929755210876,
      "num_tokens": 21170063.0,
      "step": 9230
    },
    {
      "entropy": 5.37916374206543,
      "epoch": 0.8871277617675313,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004929537021913997,
      "loss": 5.1352,
      "mean_token_accuracy": 0.21041790544986724,
      "num_tokens": 21182427.0,
      "step": 9235
    },
    {
      "entropy": 5.291451072692871,
      "epoch": 0.8876080691642652,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004929451870461064,
      "loss": 4.9917,
      "mean_token_accuracy": 0.22073093056678772,
      "num_tokens": 21195011.0,
      "step": 9240
    },
    {
      "entropy": 5.28820390701294,
      "epoch": 0.888088376560999,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004929366668407731,
      "loss": 5.0244,
      "mean_token_accuracy": 0.22147382646799088,
      "num_tokens": 21207886.0,
      "step": 9245
    },
    {
      "entropy": 5.318432474136353,
      "epoch": 0.8885686839577329,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004929281415755974,
      "loss": 5.0088,
      "mean_token_accuracy": 0.22068893015384675,
      "num_tokens": 21219285.0,
      "step": 9250
    },
    {
      "entropy": 5.348536634445191,
      "epoch": 0.8890489913544669,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004929196112507775,
      "loss": 5.0596,
      "mean_token_accuracy": 0.21293590515851973,
      "num_tokens": 21231208.0,
      "step": 9255
    },
    {
      "entropy": 5.28056492805481,
      "epoch": 0.8895292987512008,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004929110758665112,
      "loss": 4.9529,
      "mean_token_accuracy": 0.21895646303892136,
      "num_tokens": 21242573.0,
      "step": 9260
    },
    {
      "entropy": 5.218964385986328,
      "epoch": 0.8900096061479347,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004929025354229969,
      "loss": 4.9165,
      "mean_token_accuracy": 0.23360105454921723,
      "num_tokens": 21254966.0,
      "step": 9265
    },
    {
      "entropy": 5.292690467834473,
      "epoch": 0.8904899135446686,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004928939899204326,
      "loss": 5.0508,
      "mean_token_accuracy": 0.21411998867988585,
      "num_tokens": 21266167.0,
      "step": 9270
    },
    {
      "entropy": 5.333870697021484,
      "epoch": 0.8909702209414025,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000492885439359017,
      "loss": 5.0908,
      "mean_token_accuracy": 0.21503318548202516,
      "num_tokens": 21277036.0,
      "step": 9275
    },
    {
      "entropy": 5.2855884552001955,
      "epoch": 0.8914505283381364,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004928768837389485,
      "loss": 5.0252,
      "mean_token_accuracy": 0.21801000088453293,
      "num_tokens": 21288887.0,
      "step": 9280
    },
    {
      "entropy": 5.353742885589599,
      "epoch": 0.8919308357348703,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004928683230604257,
      "loss": 5.1251,
      "mean_token_accuracy": 0.21071304827928544,
      "num_tokens": 21300664.0,
      "step": 9285
    },
    {
      "entropy": 5.3726386547088625,
      "epoch": 0.8924111431316042,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004928597573236474,
      "loss": 4.9962,
      "mean_token_accuracy": 0.22287717461585999,
      "num_tokens": 21311535.0,
      "step": 9290
    },
    {
      "entropy": 5.222602033615113,
      "epoch": 0.8928914505283382,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004928511865288123,
      "loss": 4.9734,
      "mean_token_accuracy": 0.2204432010650635,
      "num_tokens": 21323248.0,
      "step": 9295
    },
    {
      "entropy": 5.308815908432007,
      "epoch": 0.8933717579250721,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004928426106761197,
      "loss": 5.0148,
      "mean_token_accuracy": 0.21792270094156266,
      "num_tokens": 21334901.0,
      "step": 9300
    },
    {
      "entropy": 5.455125379562378,
      "epoch": 0.8938520653218059,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004928340297657685,
      "loss": 5.1052,
      "mean_token_accuracy": 0.2138903796672821,
      "num_tokens": 21345902.0,
      "step": 9305
    },
    {
      "entropy": 5.402592992782592,
      "epoch": 0.8943323727185398,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004928254437979578,
      "loss": 5.133,
      "mean_token_accuracy": 0.20897181034088136,
      "num_tokens": 21358240.0,
      "step": 9310
    },
    {
      "entropy": 5.230735206604004,
      "epoch": 0.8948126801152738,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004928168527728873,
      "loss": 4.8925,
      "mean_token_accuracy": 0.21925444304943084,
      "num_tokens": 21368209.0,
      "step": 9315
    },
    {
      "entropy": 5.25173773765564,
      "epoch": 0.8952929875120077,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004928082566907562,
      "loss": 5.0057,
      "mean_token_accuracy": 0.2107519507408142,
      "num_tokens": 21379484.0,
      "step": 9320
    },
    {
      "entropy": 5.312851762771606,
      "epoch": 0.8957732949087416,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004927996555517642,
      "loss": 5.0216,
      "mean_token_accuracy": 0.21481863558292388,
      "num_tokens": 21391750.0,
      "step": 9325
    },
    {
      "entropy": 5.382816219329834,
      "epoch": 0.8962536023054755,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004927910493561109,
      "loss": 5.0895,
      "mean_token_accuracy": 0.21238654255867004,
      "num_tokens": 21402583.0,
      "step": 9330
    },
    {
      "entropy": 5.321469926834107,
      "epoch": 0.8967339097022095,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000492782438103996,
      "loss": 5.0098,
      "mean_token_accuracy": 0.208832223713398,
      "num_tokens": 21412892.0,
      "step": 9335
    },
    {
      "entropy": 5.344041061401367,
      "epoch": 0.8972142170989433,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004927738217956197,
      "loss": 5.0278,
      "mean_token_accuracy": 0.21638358533382415,
      "num_tokens": 21422828.0,
      "step": 9340
    },
    {
      "entropy": 5.2740538120269775,
      "epoch": 0.8976945244956772,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004927652004311819,
      "loss": 5.0361,
      "mean_token_accuracy": 0.21591609567403794,
      "num_tokens": 21434441.0,
      "step": 9345
    },
    {
      "entropy": 5.3726622581481935,
      "epoch": 0.8981748318924111,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004927565740108828,
      "loss": 5.1708,
      "mean_token_accuracy": 0.20641844868659973,
      "num_tokens": 21445945.0,
      "step": 9350
    },
    {
      "entropy": 5.382807874679566,
      "epoch": 0.8986551392891451,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004927479425349226,
      "loss": 5.0867,
      "mean_token_accuracy": 0.2191966876387596,
      "num_tokens": 21457265.0,
      "step": 9355
    },
    {
      "entropy": 5.318716812133789,
      "epoch": 0.899135446685879,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004927393060035018,
      "loss": 5.0417,
      "mean_token_accuracy": 0.2137404516339302,
      "num_tokens": 21469090.0,
      "step": 9360
    },
    {
      "entropy": 5.31899471282959,
      "epoch": 0.8996157540826129,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004927306644168207,
      "loss": 5.0912,
      "mean_token_accuracy": 0.2055736556649208,
      "num_tokens": 21480280.0,
      "step": 9365
    },
    {
      "entropy": 5.3223069190979,
      "epoch": 0.9000960614793467,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004927220177750803,
      "loss": 5.0088,
      "mean_token_accuracy": 0.22321386635303497,
      "num_tokens": 21491856.0,
      "step": 9370
    },
    {
      "entropy": 5.381522178649902,
      "epoch": 0.9005763688760807,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004927133660784811,
      "loss": 5.0641,
      "mean_token_accuracy": 0.21337833106517792,
      "num_tokens": 21501814.0,
      "step": 9375
    },
    {
      "entropy": 5.309741592407226,
      "epoch": 0.9010566762728146,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004927047093272241,
      "loss": 5.0745,
      "mean_token_accuracy": 0.21261882334947585,
      "num_tokens": 21514725.0,
      "step": 9380
    },
    {
      "entropy": 5.272862339019776,
      "epoch": 0.9015369836695485,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00049269604752151,
      "loss": 5.0208,
      "mean_token_accuracy": 0.2207929238677025,
      "num_tokens": 21526523.0,
      "step": 9385
    },
    {
      "entropy": 5.261856508255005,
      "epoch": 0.9020172910662824,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004926873806615403,
      "loss": 4.9921,
      "mean_token_accuracy": 0.2205223873257637,
      "num_tokens": 21537947.0,
      "step": 9390
    },
    {
      "entropy": 5.3909515857696535,
      "epoch": 0.9024975984630164,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004926787087475158,
      "loss": 5.1007,
      "mean_token_accuracy": 0.20380255877971648,
      "num_tokens": 21550238.0,
      "step": 9395
    },
    {
      "entropy": 5.319186401367188,
      "epoch": 0.9029779058597502,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004926700317796382,
      "loss": 5.1061,
      "mean_token_accuracy": 0.2101788729429245,
      "num_tokens": 21561546.0,
      "step": 9400
    },
    {
      "entropy": 5.346573257446289,
      "epoch": 0.9034582132564841,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004926613497581088,
      "loss": 5.0737,
      "mean_token_accuracy": 0.2156229257583618,
      "num_tokens": 21572034.0,
      "step": 9405
    },
    {
      "entropy": 5.335413599014283,
      "epoch": 0.9039385206532181,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004926526626831292,
      "loss": 5.0298,
      "mean_token_accuracy": 0.21373840570449829,
      "num_tokens": 21583427.0,
      "step": 9410
    },
    {
      "entropy": 5.324167680740357,
      "epoch": 0.904418828049952,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004926439705549011,
      "loss": 5.0336,
      "mean_token_accuracy": 0.22302501350641252,
      "num_tokens": 21595071.0,
      "step": 9415
    },
    {
      "entropy": 5.37883677482605,
      "epoch": 0.9048991354466859,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004926352733736262,
      "loss": 5.0842,
      "mean_token_accuracy": 0.2141408309340477,
      "num_tokens": 21606007.0,
      "step": 9420
    },
    {
      "entropy": 5.328155660629273,
      "epoch": 0.9053794428434198,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004926265711395065,
      "loss": 5.0336,
      "mean_token_accuracy": 0.21804341822862625,
      "num_tokens": 21617988.0,
      "step": 9425
    },
    {
      "entropy": 5.317275142669677,
      "epoch": 0.9058597502401537,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000492617863852744,
      "loss": 5.0456,
      "mean_token_accuracy": 0.215871000289917,
      "num_tokens": 21629832.0,
      "step": 9430
    },
    {
      "entropy": 5.315813064575195,
      "epoch": 0.9063400576368876,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004926091515135409,
      "loss": 5.0544,
      "mean_token_accuracy": 0.20848238170146943,
      "num_tokens": 21641295.0,
      "step": 9435
    },
    {
      "entropy": 5.229184865951538,
      "epoch": 0.9068203650336215,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004926004341220995,
      "loss": 4.9162,
      "mean_token_accuracy": 0.22651044577360152,
      "num_tokens": 21652829.0,
      "step": 9440
    },
    {
      "entropy": 5.407023668289185,
      "epoch": 0.9073006724303554,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004925917116786222,
      "loss": 5.1288,
      "mean_token_accuracy": 0.21182576864957808,
      "num_tokens": 21665032.0,
      "step": 9445
    },
    {
      "entropy": 5.191027212142944,
      "epoch": 0.9077809798270894,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004925829841833114,
      "loss": 4.9647,
      "mean_token_accuracy": 0.22772566825151444,
      "num_tokens": 21675814.0,
      "step": 9450
    },
    {
      "entropy": 5.297474145889282,
      "epoch": 0.9082612872238233,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004925742516363699,
      "loss": 5.0279,
      "mean_token_accuracy": 0.21724895834922792,
      "num_tokens": 21687581.0,
      "step": 9455
    },
    {
      "entropy": 5.3963446617126465,
      "epoch": 0.9087415946205571,
      "grad_norm": 0.875,
      "learning_rate": 0.0004925655140380002,
      "loss": 5.0483,
      "mean_token_accuracy": 0.20929873883724212,
      "num_tokens": 21700167.0,
      "step": 9460
    },
    {
      "entropy": 5.296067810058593,
      "epoch": 0.909221902017291,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004925567713884054,
      "loss": 5.0978,
      "mean_token_accuracy": 0.2146335795521736,
      "num_tokens": 21712309.0,
      "step": 9465
    },
    {
      "entropy": 5.317354917526245,
      "epoch": 0.909702209414025,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004925480236877884,
      "loss": 5.0229,
      "mean_token_accuracy": 0.22230723798274993,
      "num_tokens": 21723391.0,
      "step": 9470
    },
    {
      "entropy": 5.311816740036011,
      "epoch": 0.9101825168107589,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004925392709363522,
      "loss": 4.9585,
      "mean_token_accuracy": 0.2186511978507042,
      "num_tokens": 21735052.0,
      "step": 9475
    },
    {
      "entropy": 5.340672969818115,
      "epoch": 0.9106628242074928,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004925305131343001,
      "loss": 5.0947,
      "mean_token_accuracy": 0.2141306772828102,
      "num_tokens": 21746424.0,
      "step": 9480
    },
    {
      "entropy": 5.310224294662476,
      "epoch": 0.9111431316042267,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004925217502818355,
      "loss": 5.0836,
      "mean_token_accuracy": 0.21054065227508545,
      "num_tokens": 21756960.0,
      "step": 9485
    },
    {
      "entropy": 5.34330472946167,
      "epoch": 0.9116234390009607,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004925129823791616,
      "loss": 5.0266,
      "mean_token_accuracy": 0.22672560065984726,
      "num_tokens": 21767621.0,
      "step": 9490
    },
    {
      "entropy": 5.305933809280395,
      "epoch": 0.9121037463976945,
      "grad_norm": 0.875,
      "learning_rate": 0.0004925042094264822,
      "loss": 5.0655,
      "mean_token_accuracy": 0.21312022507190703,
      "num_tokens": 21779403.0,
      "step": 9495
    },
    {
      "entropy": 5.431873178482055,
      "epoch": 0.9125840537944284,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000492495431424001,
      "loss": 5.1585,
      "mean_token_accuracy": 0.20595968663692474,
      "num_tokens": 21791039.0,
      "step": 9500
    },
    {
      "entropy": 5.425524520874023,
      "epoch": 0.9130643611911623,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004924866483719216,
      "loss": 5.1394,
      "mean_token_accuracy": 0.20783050507307052,
      "num_tokens": 21803731.0,
      "step": 9505
    },
    {
      "entropy": 5.362973403930664,
      "epoch": 0.9135446685878963,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004924778602704481,
      "loss": 5.0388,
      "mean_token_accuracy": 0.21072967499494552,
      "num_tokens": 21816139.0,
      "step": 9510
    },
    {
      "entropy": 5.208991003036499,
      "epoch": 0.9140249759846302,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004924690671197845,
      "loss": 4.9065,
      "mean_token_accuracy": 0.22255628556013107,
      "num_tokens": 21826861.0,
      "step": 9515
    },
    {
      "entropy": 5.284266662597656,
      "epoch": 0.914505283381364,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004924602689201348,
      "loss": 5.0729,
      "mean_token_accuracy": 0.21916309893131256,
      "num_tokens": 21838083.0,
      "step": 9520
    },
    {
      "entropy": 5.385208225250244,
      "epoch": 0.9149855907780979,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004924514656717034,
      "loss": 5.0718,
      "mean_token_accuracy": 0.21638004779815673,
      "num_tokens": 21849232.0,
      "step": 9525
    },
    {
      "entropy": 5.371644973754883,
      "epoch": 0.9154658981748319,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004924426573746948,
      "loss": 5.1085,
      "mean_token_accuracy": 0.20857543200254441,
      "num_tokens": 21861650.0,
      "step": 9530
    },
    {
      "entropy": 5.432660531997681,
      "epoch": 0.9159462055715658,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004924338440293131,
      "loss": 5.1595,
      "mean_token_accuracy": 0.21015776693820953,
      "num_tokens": 21873369.0,
      "step": 9535
    },
    {
      "entropy": 5.3409665584564205,
      "epoch": 0.9164265129682997,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004924250256357635,
      "loss": 5.0407,
      "mean_token_accuracy": 0.21708695590496063,
      "num_tokens": 21884563.0,
      "step": 9540
    },
    {
      "entropy": 5.351705121994018,
      "epoch": 0.9169068203650336,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004924162021942502,
      "loss": 5.0763,
      "mean_token_accuracy": 0.21649834960699083,
      "num_tokens": 21896114.0,
      "step": 9545
    },
    {
      "entropy": 5.335648393630981,
      "epoch": 0.9173871277617676,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004924073737049784,
      "loss": 5.0818,
      "mean_token_accuracy": 0.22015207558870314,
      "num_tokens": 21907386.0,
      "step": 9550
    },
    {
      "entropy": 5.372521257400512,
      "epoch": 0.9178674351585014,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004923985401681528,
      "loss": 5.1223,
      "mean_token_accuracy": 0.21338898986577987,
      "num_tokens": 21918312.0,
      "step": 9555
    },
    {
      "entropy": 5.309265899658203,
      "epoch": 0.9183477425552353,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004923897015839788,
      "loss": 4.9851,
      "mean_token_accuracy": 0.22309899777173997,
      "num_tokens": 21930101.0,
      "step": 9560
    },
    {
      "entropy": 5.302239894866943,
      "epoch": 0.9188280499519692,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004923808579526613,
      "loss": 5.1125,
      "mean_token_accuracy": 0.21264244019985198,
      "num_tokens": 21941739.0,
      "step": 9565
    },
    {
      "entropy": 5.359229612350464,
      "epoch": 0.9193083573487032,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004923720092744059,
      "loss": 5.0857,
      "mean_token_accuracy": 0.20372410416603087,
      "num_tokens": 21953342.0,
      "step": 9570
    },
    {
      "entropy": 5.415268039703369,
      "epoch": 0.9197886647454371,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004923631555494179,
      "loss": 5.1212,
      "mean_token_accuracy": 0.2114642396569252,
      "num_tokens": 21963506.0,
      "step": 9575
    },
    {
      "entropy": 5.39159083366394,
      "epoch": 0.920268972142171,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004923542967779028,
      "loss": 5.1491,
      "mean_token_accuracy": 0.20872991681098937,
      "num_tokens": 21975078.0,
      "step": 9580
    },
    {
      "entropy": 5.327833795547486,
      "epoch": 0.920749279538905,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004923454329600664,
      "loss": 4.9846,
      "mean_token_accuracy": 0.21937451362609864,
      "num_tokens": 21987128.0,
      "step": 9585
    },
    {
      "entropy": 5.329600715637207,
      "epoch": 0.9212295869356388,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004923365640961143,
      "loss": 5.0953,
      "mean_token_accuracy": 0.21669147461652755,
      "num_tokens": 22000106.0,
      "step": 9590
    },
    {
      "entropy": 5.319876861572266,
      "epoch": 0.9217098943323727,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004923276901862526,
      "loss": 4.9948,
      "mean_token_accuracy": 0.21855123490095138,
      "num_tokens": 22010571.0,
      "step": 9595
    },
    {
      "entropy": 5.268307495117187,
      "epoch": 0.9221902017291066,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004923188112306874,
      "loss": 4.9662,
      "mean_token_accuracy": 0.22862545698881148,
      "num_tokens": 22021399.0,
      "step": 9600
    },
    {
      "entropy": 5.2910998344421385,
      "epoch": 0.9226705091258406,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004923099272296246,
      "loss": 4.9366,
      "mean_token_accuracy": 0.2241847425699234,
      "num_tokens": 22032302.0,
      "step": 9605
    },
    {
      "entropy": 5.309083080291748,
      "epoch": 0.9231508165225745,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004923010381832706,
      "loss": 4.9872,
      "mean_token_accuracy": 0.223245470225811,
      "num_tokens": 22043762.0,
      "step": 9610
    },
    {
      "entropy": 5.330354022979736,
      "epoch": 0.9236311239193083,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004922921440918318,
      "loss": 5.0386,
      "mean_token_accuracy": 0.21934472769498825,
      "num_tokens": 22054885.0,
      "step": 9615
    },
    {
      "entropy": 5.400898694992065,
      "epoch": 0.9241114313160422,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004922832449555144,
      "loss": 5.1097,
      "mean_token_accuracy": 0.21035009026527404,
      "num_tokens": 22066627.0,
      "step": 9620
    },
    {
      "entropy": 5.226740217208862,
      "epoch": 0.9245917387127762,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004922743407745255,
      "loss": 4.9597,
      "mean_token_accuracy": 0.22060729414224625,
      "num_tokens": 22078062.0,
      "step": 9625
    },
    {
      "entropy": 5.284878492355347,
      "epoch": 0.9250720461095101,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004922654315490714,
      "loss": 4.9665,
      "mean_token_accuracy": 0.21778267174959182,
      "num_tokens": 22089465.0,
      "step": 9630
    },
    {
      "entropy": 5.3933202743530275,
      "epoch": 0.925552353506244,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004922565172793593,
      "loss": 5.1056,
      "mean_token_accuracy": 0.21052461862564087,
      "num_tokens": 22101688.0,
      "step": 9635
    },
    {
      "entropy": 5.304271554946899,
      "epoch": 0.9260326609029779,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004922475979655958,
      "loss": 5.0835,
      "mean_token_accuracy": 0.2142885997891426,
      "num_tokens": 22112912.0,
      "step": 9640
    },
    {
      "entropy": 5.391285848617554,
      "epoch": 0.9265129682997119,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004922386736079883,
      "loss": 5.1887,
      "mean_token_accuracy": 0.20508348643779756,
      "num_tokens": 22125079.0,
      "step": 9645
    },
    {
      "entropy": 5.422160005569458,
      "epoch": 0.9269932756964457,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004922297442067438,
      "loss": 5.0708,
      "mean_token_accuracy": 0.22008590549230575,
      "num_tokens": 22137146.0,
      "step": 9650
    },
    {
      "entropy": 5.288139915466308,
      "epoch": 0.9274735830931796,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004922208097620697,
      "loss": 5.0044,
      "mean_token_accuracy": 0.21349945068359374,
      "num_tokens": 22148330.0,
      "step": 9655
    },
    {
      "entropy": 5.202523612976075,
      "epoch": 0.9279538904899135,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004922118702741735,
      "loss": 5.0023,
      "mean_token_accuracy": 0.2154933825135231,
      "num_tokens": 22160511.0,
      "step": 9660
    },
    {
      "entropy": 5.309700870513916,
      "epoch": 0.9284341978866475,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004922029257432625,
      "loss": 5.0009,
      "mean_token_accuracy": 0.22090534567832948,
      "num_tokens": 22171298.0,
      "step": 9665
    },
    {
      "entropy": 5.303619813919068,
      "epoch": 0.9289145052833814,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004921939761695446,
      "loss": 5.0509,
      "mean_token_accuracy": 0.21446681767702103,
      "num_tokens": 22181897.0,
      "step": 9670
    },
    {
      "entropy": 5.389228487014771,
      "epoch": 0.9293948126801153,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004921850215532275,
      "loss": 5.1403,
      "mean_token_accuracy": 0.21048984825611114,
      "num_tokens": 22194630.0,
      "step": 9675
    },
    {
      "entropy": 5.345093727111816,
      "epoch": 0.9298751200768491,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004921760618945192,
      "loss": 5.0501,
      "mean_token_accuracy": 0.22013561874628068,
      "num_tokens": 22206041.0,
      "step": 9680
    },
    {
      "entropy": 5.278069305419922,
      "epoch": 0.9303554274735831,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004921670971936276,
      "loss": 5.0502,
      "mean_token_accuracy": 0.2145330488681793,
      "num_tokens": 22218434.0,
      "step": 9685
    },
    {
      "entropy": 5.379455709457398,
      "epoch": 0.930835734870317,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004921581274507607,
      "loss": 5.0325,
      "mean_token_accuracy": 0.2174965262413025,
      "num_tokens": 22229182.0,
      "step": 9690
    },
    {
      "entropy": 5.3874797344207765,
      "epoch": 0.9313160422670509,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000492149152666127,
      "loss": 5.1369,
      "mean_token_accuracy": 0.21050627529621124,
      "num_tokens": 22241142.0,
      "step": 9695
    },
    {
      "entropy": 5.344835901260376,
      "epoch": 0.9317963496637848,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004921401728399348,
      "loss": 5.0582,
      "mean_token_accuracy": 0.20936338156461715,
      "num_tokens": 22252695.0,
      "step": 9700
    },
    {
      "entropy": 5.258717060089111,
      "epoch": 0.9322766570605188,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004921311879723926,
      "loss": 4.9922,
      "mean_token_accuracy": 0.22483751475811004,
      "num_tokens": 22264159.0,
      "step": 9705
    },
    {
      "entropy": 5.367508602142334,
      "epoch": 0.9327569644572526,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004921221980637088,
      "loss": 5.0606,
      "mean_token_accuracy": 0.21813715398311614,
      "num_tokens": 22276154.0,
      "step": 9710
    },
    {
      "entropy": 5.257150888442993,
      "epoch": 0.9332372718539865,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004921132031140925,
      "loss": 4.9315,
      "mean_token_accuracy": 0.2206779435276985,
      "num_tokens": 22288328.0,
      "step": 9715
    },
    {
      "entropy": 5.30983715057373,
      "epoch": 0.9337175792507204,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004921042031237521,
      "loss": 5.0306,
      "mean_token_accuracy": 0.21737836301326752,
      "num_tokens": 22300154.0,
      "step": 9720
    },
    {
      "entropy": 5.210334587097168,
      "epoch": 0.9341978866474544,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004920951980928969,
      "loss": 4.8798,
      "mean_token_accuracy": 0.23314393311738968,
      "num_tokens": 22311176.0,
      "step": 9725
    },
    {
      "entropy": 5.340295124053955,
      "epoch": 0.9346781940441883,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004920861880217359,
      "loss": 5.0707,
      "mean_token_accuracy": 0.20963584035634994,
      "num_tokens": 22321953.0,
      "step": 9730
    },
    {
      "entropy": 5.320473861694336,
      "epoch": 0.9351585014409222,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004920771729104781,
      "loss": 5.0496,
      "mean_token_accuracy": 0.21505106389522552,
      "num_tokens": 22333152.0,
      "step": 9735
    },
    {
      "entropy": 5.336848020553589,
      "epoch": 0.9356388088376562,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004920681527593329,
      "loss": 5.057,
      "mean_token_accuracy": 0.217726993560791,
      "num_tokens": 22344848.0,
      "step": 9740
    },
    {
      "entropy": 5.334695672988891,
      "epoch": 0.93611911623439,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004920591275685098,
      "loss": 4.9916,
      "mean_token_accuracy": 0.21745637208223342,
      "num_tokens": 22355100.0,
      "step": 9745
    },
    {
      "entropy": 5.294270944595337,
      "epoch": 0.9365994236311239,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004920500973382184,
      "loss": 5.031,
      "mean_token_accuracy": 0.21864738315343857,
      "num_tokens": 22366045.0,
      "step": 9750
    },
    {
      "entropy": 5.314795780181885,
      "epoch": 0.9370797310278578,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004920410620686682,
      "loss": 5.0096,
      "mean_token_accuracy": 0.21304722577333451,
      "num_tokens": 22378781.0,
      "step": 9755
    },
    {
      "entropy": 5.419394302368164,
      "epoch": 0.9375600384245918,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004920320217600689,
      "loss": 5.1104,
      "mean_token_accuracy": 0.2104163095355034,
      "num_tokens": 22390121.0,
      "step": 9760
    },
    {
      "entropy": 5.288579845428467,
      "epoch": 0.9380403458213257,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004920229764126306,
      "loss": 4.9911,
      "mean_token_accuracy": 0.22346494644880294,
      "num_tokens": 22402261.0,
      "step": 9765
    },
    {
      "entropy": 5.315666627883911,
      "epoch": 0.9385206532180596,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004920139260265632,
      "loss": 5.0567,
      "mean_token_accuracy": 0.22018129825592042,
      "num_tokens": 22413939.0,
      "step": 9770
    },
    {
      "entropy": 5.325642251968384,
      "epoch": 0.9390009606147934,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004920048706020769,
      "loss": 4.9847,
      "mean_token_accuracy": 0.21607000827789308,
      "num_tokens": 22425439.0,
      "step": 9775
    },
    {
      "entropy": 5.366525030136108,
      "epoch": 0.9394812680115274,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004919958101393817,
      "loss": 5.0767,
      "mean_token_accuracy": 0.20951039493083953,
      "num_tokens": 22436989.0,
      "step": 9780
    },
    {
      "entropy": 5.356968688964844,
      "epoch": 0.9399615754082613,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004919867446386883,
      "loss": 5.0992,
      "mean_token_accuracy": 0.2107743039727211,
      "num_tokens": 22449181.0,
      "step": 9785
    },
    {
      "entropy": 5.251317405700684,
      "epoch": 0.9404418828049952,
      "grad_norm": 0.90625,
      "learning_rate": 0.000491977674100207,
      "loss": 4.9334,
      "mean_token_accuracy": 0.2237143859267235,
      "num_tokens": 22460592.0,
      "step": 9790
    },
    {
      "entropy": 5.239273595809936,
      "epoch": 0.9409221902017291,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004919685985241483,
      "loss": 4.9876,
      "mean_token_accuracy": 0.22301929891109468,
      "num_tokens": 22471718.0,
      "step": 9795
    },
    {
      "entropy": 5.351552820205688,
      "epoch": 0.9414024975984631,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000491959517910723,
      "loss": 5.0441,
      "mean_token_accuracy": 0.22287043780088425,
      "num_tokens": 22482413.0,
      "step": 9800
    },
    {
      "entropy": 5.301695489883423,
      "epoch": 0.9418828049951969,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004919504322601421,
      "loss": 4.9707,
      "mean_token_accuracy": 0.21922060400247573,
      "num_tokens": 22494016.0,
      "step": 9805
    },
    {
      "entropy": 5.328707456588745,
      "epoch": 0.9423631123919308,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004919413415726162,
      "loss": 5.0609,
      "mean_token_accuracy": 0.21316564530134202,
      "num_tokens": 22505015.0,
      "step": 9810
    },
    {
      "entropy": 5.3101362705230715,
      "epoch": 0.9428434197886647,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004919322458483566,
      "loss": 5.0754,
      "mean_token_accuracy": 0.21946054250001906,
      "num_tokens": 22516691.0,
      "step": 9815
    },
    {
      "entropy": 5.326116847991943,
      "epoch": 0.9433237271853987,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004919231450875745,
      "loss": 5.0232,
      "mean_token_accuracy": 0.21881574988365174,
      "num_tokens": 22527740.0,
      "step": 9820
    },
    {
      "entropy": 5.304749298095703,
      "epoch": 0.9438040345821326,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004919140392904809,
      "loss": 5.0285,
      "mean_token_accuracy": 0.21940327137708665,
      "num_tokens": 22537685.0,
      "step": 9825
    },
    {
      "entropy": 5.297449159622192,
      "epoch": 0.9442843419788665,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004919049284572875,
      "loss": 4.945,
      "mean_token_accuracy": 0.22557729929685594,
      "num_tokens": 22548282.0,
      "step": 9830
    },
    {
      "entropy": 5.250690174102783,
      "epoch": 0.9447646493756003,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004918958125882058,
      "loss": 4.896,
      "mean_token_accuracy": 0.22713269740343095,
      "num_tokens": 22559277.0,
      "step": 9835
    },
    {
      "entropy": 5.268884229660034,
      "epoch": 0.9452449567723343,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004918866916834474,
      "loss": 5.0408,
      "mean_token_accuracy": 0.2137300565838814,
      "num_tokens": 22570664.0,
      "step": 9840
    },
    {
      "entropy": 5.286205005645752,
      "epoch": 0.9457252641690682,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004918775657432239,
      "loss": 4.9828,
      "mean_token_accuracy": 0.22495179921388625,
      "num_tokens": 22580410.0,
      "step": 9845
    },
    {
      "entropy": 5.279875659942627,
      "epoch": 0.9462055715658021,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004918684347677474,
      "loss": 5.0412,
      "mean_token_accuracy": 0.21835660934448242,
      "num_tokens": 22591630.0,
      "step": 9850
    },
    {
      "entropy": 5.317875051498413,
      "epoch": 0.946685878962536,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004918592987572298,
      "loss": 5.0356,
      "mean_token_accuracy": 0.2163267105817795,
      "num_tokens": 22603658.0,
      "step": 9855
    },
    {
      "entropy": 5.367918539047241,
      "epoch": 0.94716618635927,
      "grad_norm": 1.0,
      "learning_rate": 0.0004918501577118832,
      "loss": 5.0759,
      "mean_token_accuracy": 0.20692822486162185,
      "num_tokens": 22614812.0,
      "step": 9860
    },
    {
      "entropy": 5.319218540191651,
      "epoch": 0.9476464937560038,
      "grad_norm": 0.921875,
      "learning_rate": 0.00049184101163192,
      "loss": 4.9939,
      "mean_token_accuracy": 0.21730560064315796,
      "num_tokens": 22626924.0,
      "step": 9865
    },
    {
      "entropy": 5.364023113250733,
      "epoch": 0.9481268011527377,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004918318605175522,
      "loss": 5.036,
      "mean_token_accuracy": 0.2125631868839264,
      "num_tokens": 22638715.0,
      "step": 9870
    },
    {
      "entropy": 5.375600051879883,
      "epoch": 0.9486071085494716,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004918227043689924,
      "loss": 5.1665,
      "mean_token_accuracy": 0.2041703850030899,
      "num_tokens": 22650257.0,
      "step": 9875
    },
    {
      "entropy": 5.291876983642578,
      "epoch": 0.9490874159462056,
      "grad_norm": 0.875,
      "learning_rate": 0.0004918135431864534,
      "loss": 5.0381,
      "mean_token_accuracy": 0.21445752680301666,
      "num_tokens": 22662480.0,
      "step": 9880
    },
    {
      "entropy": 5.352047157287598,
      "epoch": 0.9495677233429395,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004918043769701478,
      "loss": 5.0241,
      "mean_token_accuracy": 0.21248966306447983,
      "num_tokens": 22673927.0,
      "step": 9885
    },
    {
      "entropy": 5.404024362564087,
      "epoch": 0.9500480307396734,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004917952057202882,
      "loss": 5.0819,
      "mean_token_accuracy": 0.22141211926937104,
      "num_tokens": 22684670.0,
      "step": 9890
    },
    {
      "entropy": 5.372463893890381,
      "epoch": 0.9505283381364072,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004917860294370877,
      "loss": 5.1396,
      "mean_token_accuracy": 0.21322518438100815,
      "num_tokens": 22696571.0,
      "step": 9895
    },
    {
      "entropy": 5.238849401473999,
      "epoch": 0.9510086455331412,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004917768481207593,
      "loss": 5.0013,
      "mean_token_accuracy": 0.22458896785974503,
      "num_tokens": 22708987.0,
      "step": 9900
    },
    {
      "entropy": 5.3238318920135494,
      "epoch": 0.9514889529298751,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004917676617715162,
      "loss": 5.0685,
      "mean_token_accuracy": 0.22402574717998505,
      "num_tokens": 22720769.0,
      "step": 9905
    },
    {
      "entropy": 5.32821888923645,
      "epoch": 0.951969260326609,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004917584703895717,
      "loss": 4.9951,
      "mean_token_accuracy": 0.22140030115842818,
      "num_tokens": 22732592.0,
      "step": 9910
    },
    {
      "entropy": 5.345959424972534,
      "epoch": 0.952449567723343,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004917492739751391,
      "loss": 5.0372,
      "mean_token_accuracy": 0.2117053672671318,
      "num_tokens": 22744440.0,
      "step": 9915
    },
    {
      "entropy": 5.2869163990020756,
      "epoch": 0.9529298751200769,
      "grad_norm": 1.0,
      "learning_rate": 0.000491740072528432,
      "loss": 4.9956,
      "mean_token_accuracy": 0.21757068186998368,
      "num_tokens": 22756336.0,
      "step": 9920
    },
    {
      "entropy": 5.288923549652099,
      "epoch": 0.9534101825168108,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000491730866049664,
      "loss": 5.0218,
      "mean_token_accuracy": 0.21755135655403138,
      "num_tokens": 22768256.0,
      "step": 9925
    },
    {
      "entropy": 5.375495052337646,
      "epoch": 0.9538904899135446,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004917216545390489,
      "loss": 5.0703,
      "mean_token_accuracy": 0.21447910219430924,
      "num_tokens": 22779693.0,
      "step": 9930
    },
    {
      "entropy": 5.246587753295898,
      "epoch": 0.9543707973102786,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004917124379968004,
      "loss": 4.914,
      "mean_token_accuracy": 0.22556027323007583,
      "num_tokens": 22791417.0,
      "step": 9935
    },
    {
      "entropy": 5.340655946731568,
      "epoch": 0.9548511047070125,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004917032164231327,
      "loss": 5.0041,
      "mean_token_accuracy": 0.2189626470208168,
      "num_tokens": 22803541.0,
      "step": 9940
    },
    {
      "entropy": 5.359480381011963,
      "epoch": 0.9553314121037464,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004916939898182598,
      "loss": 5.0476,
      "mean_token_accuracy": 0.21102361679077147,
      "num_tokens": 22814648.0,
      "step": 9945
    },
    {
      "entropy": 5.330089473724366,
      "epoch": 0.9558117195004803,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004916847581823958,
      "loss": 5.0343,
      "mean_token_accuracy": 0.21618885397911072,
      "num_tokens": 22825672.0,
      "step": 9950
    },
    {
      "entropy": 5.299649858474732,
      "epoch": 0.9562920268972143,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004916755215157552,
      "loss": 4.9918,
      "mean_token_accuracy": 0.21889250576496125,
      "num_tokens": 22837663.0,
      "step": 9955
    },
    {
      "entropy": 5.202730131149292,
      "epoch": 0.9567723342939481,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004916662798185524,
      "loss": 4.8927,
      "mean_token_accuracy": 0.22119835168123245,
      "num_tokens": 22849531.0,
      "step": 9960
    },
    {
      "entropy": 5.308682632446289,
      "epoch": 0.957252641690682,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004916570330910019,
      "loss": 5.0499,
      "mean_token_accuracy": 0.21198931634426116,
      "num_tokens": 22861784.0,
      "step": 9965
    },
    {
      "entropy": 5.375360631942749,
      "epoch": 0.9577329490874159,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004916477813333185,
      "loss": 5.0782,
      "mean_token_accuracy": 0.21123120486736296,
      "num_tokens": 22873482.0,
      "step": 9970
    },
    {
      "entropy": 5.2821496486663815,
      "epoch": 0.9582132564841499,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004916385245457168,
      "loss": 4.9789,
      "mean_token_accuracy": 0.22451970130205154,
      "num_tokens": 22883917.0,
      "step": 9975
    },
    {
      "entropy": 5.306679582595825,
      "epoch": 0.9586935638808838,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000491629262728412,
      "loss": 5.0172,
      "mean_token_accuracy": 0.22300494760274886,
      "num_tokens": 22896167.0,
      "step": 9980
    },
    {
      "entropy": 5.315229749679565,
      "epoch": 0.9591738712776177,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004916199958816188,
      "loss": 5.0589,
      "mean_token_accuracy": 0.218895848095417,
      "num_tokens": 22908350.0,
      "step": 9985
    },
    {
      "entropy": 5.434739112854004,
      "epoch": 0.9596541786743515,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004916107240055527,
      "loss": 5.1529,
      "mean_token_accuracy": 0.21266272515058518,
      "num_tokens": 22918913.0,
      "step": 9990
    },
    {
      "entropy": 5.211382722854614,
      "epoch": 0.9601344860710855,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004916014471004287,
      "loss": 4.9787,
      "mean_token_accuracy": 0.220284166932106,
      "num_tokens": 22929336.0,
      "step": 9995
    },
    {
      "entropy": 5.267848968505859,
      "epoch": 0.9606147934678194,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004915921651664622,
      "loss": 4.9827,
      "mean_token_accuracy": 0.21854242980480193,
      "num_tokens": 22940406.0,
      "step": 10000
    },
    {
      "entropy": 5.2923791885375975,
      "epoch": 0.9610951008645533,
      "grad_norm": 0.90625,
      "learning_rate": 0.000491582878203869,
      "loss": 4.9785,
      "mean_token_accuracy": 0.21876122653484345,
      "num_tokens": 22951518.0,
      "step": 10005
    },
    {
      "entropy": 5.243970537185669,
      "epoch": 0.9615754082612872,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004915735862128643,
      "loss": 4.9629,
      "mean_token_accuracy": 0.22157049626111985,
      "num_tokens": 22962309.0,
      "step": 10010
    },
    {
      "entropy": 5.279350900650025,
      "epoch": 0.9620557156580212,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004915642891936641,
      "loss": 4.9899,
      "mean_token_accuracy": 0.2215436428785324,
      "num_tokens": 22973996.0,
      "step": 10015
    },
    {
      "entropy": 5.357512903213501,
      "epoch": 0.962536023054755,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004915549871464841,
      "loss": 5.0984,
      "mean_token_accuracy": 0.21900363266468048,
      "num_tokens": 22984964.0,
      "step": 10020
    },
    {
      "entropy": 5.366422033309936,
      "epoch": 0.9630163304514889,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004915456800715403,
      "loss": 5.061,
      "mean_token_accuracy": 0.2139080911874771,
      "num_tokens": 22996058.0,
      "step": 10025
    },
    {
      "entropy": 5.239693832397461,
      "epoch": 0.9634966378482228,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000491536367969049,
      "loss": 4.916,
      "mean_token_accuracy": 0.2231506049633026,
      "num_tokens": 23006140.0,
      "step": 10030
    },
    {
      "entropy": 5.310165977478027,
      "epoch": 0.9639769452449568,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004915270508392261,
      "loss": 5.1054,
      "mean_token_accuracy": 0.21028921157121658,
      "num_tokens": 23018129.0,
      "step": 10035
    },
    {
      "entropy": 5.338807582855225,
      "epoch": 0.9644572526416907,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000491517728682288,
      "loss": 4.9287,
      "mean_token_accuracy": 0.22368063777685165,
      "num_tokens": 23028673.0,
      "step": 10040
    },
    {
      "entropy": 5.293785572052002,
      "epoch": 0.9649375600384246,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004915084014984512,
      "loss": 4.9903,
      "mean_token_accuracy": 0.22126883864402772,
      "num_tokens": 23040778.0,
      "step": 10045
    },
    {
      "entropy": 5.236273241043091,
      "epoch": 0.9654178674351584,
      "grad_norm": 1.0,
      "learning_rate": 0.0004914990692879322,
      "loss": 5.0067,
      "mean_token_accuracy": 0.2269112303853035,
      "num_tokens": 23052170.0,
      "step": 10050
    },
    {
      "entropy": 5.263945579528809,
      "epoch": 0.9658981748318924,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004914897320509478,
      "loss": 4.9778,
      "mean_token_accuracy": 0.22065386772155762,
      "num_tokens": 23063455.0,
      "step": 10055
    },
    {
      "entropy": 5.241347122192383,
      "epoch": 0.9663784822286263,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004914803897877146,
      "loss": 4.8932,
      "mean_token_accuracy": 0.22524694800376893,
      "num_tokens": 23075009.0,
      "step": 10060
    },
    {
      "entropy": 5.32644624710083,
      "epoch": 0.9668587896253602,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004914710424984495,
      "loss": 5.0256,
      "mean_token_accuracy": 0.2151941180229187,
      "num_tokens": 23085921.0,
      "step": 10065
    },
    {
      "entropy": 5.387622833251953,
      "epoch": 0.9673390970220941,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004914616901833696,
      "loss": 5.1417,
      "mean_token_accuracy": 0.21165000051259994,
      "num_tokens": 23098181.0,
      "step": 10070
    },
    {
      "entropy": 5.241356039047242,
      "epoch": 0.9678194044188281,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000491452332842692,
      "loss": 4.9911,
      "mean_token_accuracy": 0.2154693901538849,
      "num_tokens": 23109904.0,
      "step": 10075
    },
    {
      "entropy": 5.315796661376953,
      "epoch": 0.968299711815562,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000491442970476634,
      "loss": 5.0702,
      "mean_token_accuracy": 0.21607670038938523,
      "num_tokens": 23122356.0,
      "step": 10080
    },
    {
      "entropy": 5.324264669418335,
      "epoch": 0.9687800192122958,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004914336030854129,
      "loss": 4.9754,
      "mean_token_accuracy": 0.22599320262670516,
      "num_tokens": 23134055.0,
      "step": 10085
    },
    {
      "entropy": 5.281080913543701,
      "epoch": 0.9692603266090298,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004914242306692461,
      "loss": 4.9376,
      "mean_token_accuracy": 0.22522149085998536,
      "num_tokens": 23145388.0,
      "step": 10090
    },
    {
      "entropy": 5.332269954681396,
      "epoch": 0.9697406340057637,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004914148532283516,
      "loss": 5.049,
      "mean_token_accuracy": 0.20636244416236876,
      "num_tokens": 23157083.0,
      "step": 10095
    },
    {
      "entropy": 5.267354917526245,
      "epoch": 0.9702209414024976,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004914054707629466,
      "loss": 4.9883,
      "mean_token_accuracy": 0.2220076471567154,
      "num_tokens": 23169154.0,
      "step": 10100
    },
    {
      "entropy": 5.242982196807861,
      "epoch": 0.9707012487992315,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004913960832732493,
      "loss": 4.9125,
      "mean_token_accuracy": 0.22122575491666793,
      "num_tokens": 23179879.0,
      "step": 10105
    },
    {
      "entropy": 5.307461404800415,
      "epoch": 0.9711815561959655,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004913866907594774,
      "loss": 5.0259,
      "mean_token_accuracy": 0.22066918313503264,
      "num_tokens": 23190632.0,
      "step": 10110
    },
    {
      "entropy": 5.337830257415772,
      "epoch": 0.9716618635926993,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004913772932218491,
      "loss": 5.0701,
      "mean_token_accuracy": 0.20966452211141587,
      "num_tokens": 23202060.0,
      "step": 10115
    },
    {
      "entropy": 5.313678169250489,
      "epoch": 0.9721421709894332,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004913678906605825,
      "loss": 5.0385,
      "mean_token_accuracy": 0.21130431443452835,
      "num_tokens": 23214103.0,
      "step": 10120
    },
    {
      "entropy": 5.33049578666687,
      "epoch": 0.9726224783861671,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004913584830758961,
      "loss": 5.0015,
      "mean_token_accuracy": 0.22173945903778075,
      "num_tokens": 23224676.0,
      "step": 10125
    },
    {
      "entropy": 5.337496519088745,
      "epoch": 0.9731027857829011,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004913490704680081,
      "loss": 5.0126,
      "mean_token_accuracy": 0.22231239080429077,
      "num_tokens": 23235478.0,
      "step": 10130
    },
    {
      "entropy": 5.303532552719116,
      "epoch": 0.973583093179635,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004913396528371371,
      "loss": 5.0681,
      "mean_token_accuracy": 0.2190092220902443,
      "num_tokens": 23246913.0,
      "step": 10135
    },
    {
      "entropy": 5.303474283218383,
      "epoch": 0.9740634005763689,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004913302301835018,
      "loss": 5.0341,
      "mean_token_accuracy": 0.21854409277439119,
      "num_tokens": 23258472.0,
      "step": 10140
    },
    {
      "entropy": 5.419639730453492,
      "epoch": 0.9745437079731027,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000491320802507321,
      "loss": 5.1482,
      "mean_token_accuracy": 0.20678024590015412,
      "num_tokens": 23270443.0,
      "step": 10145
    },
    {
      "entropy": 5.365138149261474,
      "epoch": 0.9750240153698367,
      "grad_norm": 1.0,
      "learning_rate": 0.0004913113698088133,
      "loss": 5.0038,
      "mean_token_accuracy": 0.22145238518714905,
      "num_tokens": 23281199.0,
      "step": 10150
    },
    {
      "entropy": 5.253880500793457,
      "epoch": 0.9755043227665706,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000491301932088198,
      "loss": 4.9885,
      "mean_token_accuracy": 0.22177451699972153,
      "num_tokens": 23291980.0,
      "step": 10155
    },
    {
      "entropy": 5.340717077255249,
      "epoch": 0.9759846301633045,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004912924893456942,
      "loss": 5.0916,
      "mean_token_accuracy": 0.2173662155866623,
      "num_tokens": 23303673.0,
      "step": 10160
    },
    {
      "entropy": 5.392231607437134,
      "epoch": 0.9764649375600384,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000491283041581521,
      "loss": 5.033,
      "mean_token_accuracy": 0.21498239934444427,
      "num_tokens": 23315830.0,
      "step": 10165
    },
    {
      "entropy": 5.3358473777771,
      "epoch": 0.9769452449567724,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004912735887958978,
      "loss": 5.0656,
      "mean_token_accuracy": 0.2116302505135536,
      "num_tokens": 23329860.0,
      "step": 10170
    },
    {
      "entropy": 5.290106391906738,
      "epoch": 0.9774255523535063,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004912641309890441,
      "loss": 4.9791,
      "mean_token_accuracy": 0.2175542324781418,
      "num_tokens": 23340610.0,
      "step": 10175
    },
    {
      "entropy": 5.31144347190857,
      "epoch": 0.9779058597502401,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004912546681611794,
      "loss": 5.0111,
      "mean_token_accuracy": 0.22013431489467622,
      "num_tokens": 23352153.0,
      "step": 10180
    },
    {
      "entropy": 5.246566200256348,
      "epoch": 0.978386167146974,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004912452003125234,
      "loss": 4.9606,
      "mean_token_accuracy": 0.22157482504844667,
      "num_tokens": 23363630.0,
      "step": 10185
    },
    {
      "entropy": 5.245562362670898,
      "epoch": 0.978866474543708,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000491235727443296,
      "loss": 4.9908,
      "mean_token_accuracy": 0.21523314863443374,
      "num_tokens": 23375922.0,
      "step": 10190
    },
    {
      "entropy": 5.2912712574005125,
      "epoch": 0.9793467819404419,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004912262495537171,
      "loss": 4.9344,
      "mean_token_accuracy": 0.22206820249557496,
      "num_tokens": 23386446.0,
      "step": 10195
    },
    {
      "entropy": 5.244411945343018,
      "epoch": 0.9798270893371758,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004912167666440068,
      "loss": 4.8949,
      "mean_token_accuracy": 0.2304428279399872,
      "num_tokens": 23398458.0,
      "step": 10200
    },
    {
      "entropy": 5.2837268829345705,
      "epoch": 0.9803073967339097,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004912072787143852,
      "loss": 4.9815,
      "mean_token_accuracy": 0.21959258764982223,
      "num_tokens": 23409725.0,
      "step": 10205
    },
    {
      "entropy": 5.265960645675659,
      "epoch": 0.9807877041306436,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004911977857650725,
      "loss": 5.0294,
      "mean_token_accuracy": 0.22082363963127136,
      "num_tokens": 23420468.0,
      "step": 10210
    },
    {
      "entropy": 5.324489736557007,
      "epoch": 0.9812680115273775,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004911882877962893,
      "loss": 5.0192,
      "mean_token_accuracy": 0.21556002050638198,
      "num_tokens": 23431386.0,
      "step": 10215
    },
    {
      "entropy": 5.363737440109253,
      "epoch": 0.9817483189241114,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004911787848082559,
      "loss": 5.038,
      "mean_token_accuracy": 0.21810971796512604,
      "num_tokens": 23442759.0,
      "step": 10220
    },
    {
      "entropy": 5.274034070968628,
      "epoch": 0.9822286263208453,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004911692768011931,
      "loss": 4.9727,
      "mean_token_accuracy": 0.22235131710767747,
      "num_tokens": 23454372.0,
      "step": 10225
    },
    {
      "entropy": 5.350617361068726,
      "epoch": 0.9827089337175793,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004911597637753217,
      "loss": 5.0507,
      "mean_token_accuracy": 0.2179918110370636,
      "num_tokens": 23465557.0,
      "step": 10230
    },
    {
      "entropy": 5.241461133956909,
      "epoch": 0.9831892411143132,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004911502457308623,
      "loss": 4.8703,
      "mean_token_accuracy": 0.22498429119586943,
      "num_tokens": 23475850.0,
      "step": 10235
    },
    {
      "entropy": 5.257108354568482,
      "epoch": 0.983669548511047,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000491140722668036,
      "loss": 4.9284,
      "mean_token_accuracy": 0.22309089899063111,
      "num_tokens": 23487145.0,
      "step": 10240
    },
    {
      "entropy": 5.3209474086761475,
      "epoch": 0.984149855907781,
      "grad_norm": 0.921875,
      "learning_rate": 0.000491131194587064,
      "loss": 5.0354,
      "mean_token_accuracy": 0.21114782392978668,
      "num_tokens": 23498112.0,
      "step": 10245
    },
    {
      "entropy": 5.272340297698975,
      "epoch": 0.9846301633045149,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004911216614881675,
      "loss": 4.9326,
      "mean_token_accuracy": 0.22073310613632202,
      "num_tokens": 23509865.0,
      "step": 10250
    },
    {
      "entropy": 5.385131072998047,
      "epoch": 0.9851104707012488,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004911121233715677,
      "loss": 5.105,
      "mean_token_accuracy": 0.20461009740829467,
      "num_tokens": 23521153.0,
      "step": 10255
    },
    {
      "entropy": 5.350685167312622,
      "epoch": 0.9855907780979827,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004911025802374861,
      "loss": 5.1313,
      "mean_token_accuracy": 0.21253945380449296,
      "num_tokens": 23532372.0,
      "step": 10260
    },
    {
      "entropy": 5.333655261993409,
      "epoch": 0.9860710854947167,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004910930320861442,
      "loss": 5.0668,
      "mean_token_accuracy": 0.21637446135282518,
      "num_tokens": 23543494.0,
      "step": 10265
    },
    {
      "entropy": 5.34989104270935,
      "epoch": 0.9865513928914506,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004910834789177639,
      "loss": 5.0186,
      "mean_token_accuracy": 0.2135024771094322,
      "num_tokens": 23554020.0,
      "step": 10270
    },
    {
      "entropy": 5.376517200469971,
      "epoch": 0.9870317002881844,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004910739207325668,
      "loss": 5.1637,
      "mean_token_accuracy": 0.20898036062717437,
      "num_tokens": 23565032.0,
      "step": 10275
    },
    {
      "entropy": 5.271209239959717,
      "epoch": 0.9875120076849183,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004910643575307749,
      "loss": 5.0332,
      "mean_token_accuracy": 0.21593343913555146,
      "num_tokens": 23576243.0,
      "step": 10280
    },
    {
      "entropy": 5.328129768371582,
      "epoch": 0.9879923150816523,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004910547893126102,
      "loss": 5.0797,
      "mean_token_accuracy": 0.2132748618721962,
      "num_tokens": 23586111.0,
      "step": 10285
    },
    {
      "entropy": 5.379953145980835,
      "epoch": 0.9884726224783862,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004910452160782948,
      "loss": 5.0832,
      "mean_token_accuracy": 0.21226355135440828,
      "num_tokens": 23597718.0,
      "step": 10290
    },
    {
      "entropy": 5.4288835525512695,
      "epoch": 0.9889529298751201,
      "grad_norm": 0.90625,
      "learning_rate": 0.000491035637828051,
      "loss": 5.0874,
      "mean_token_accuracy": 0.21133963912725448,
      "num_tokens": 23608241.0,
      "step": 10295
    },
    {
      "entropy": 5.376296186447144,
      "epoch": 0.989433237271854,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004910260545621012,
      "loss": 5.0582,
      "mean_token_accuracy": 0.2145102798938751,
      "num_tokens": 23619120.0,
      "step": 10300
    },
    {
      "entropy": 5.247903728485108,
      "epoch": 0.9899135446685879,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004910164662806679,
      "loss": 4.9228,
      "mean_token_accuracy": 0.22468816936016084,
      "num_tokens": 23630642.0,
      "step": 10305
    },
    {
      "entropy": 5.283850002288818,
      "epoch": 0.9903938520653218,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004910068729839736,
      "loss": 4.9569,
      "mean_token_accuracy": 0.2230352997779846,
      "num_tokens": 23641987.0,
      "step": 10310
    },
    {
      "entropy": 5.367715311050415,
      "epoch": 0.9908741594620557,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004909972746722413,
      "loss": 5.1198,
      "mean_token_accuracy": 0.21470424979925157,
      "num_tokens": 23653144.0,
      "step": 10315
    },
    {
      "entropy": 5.317087030410766,
      "epoch": 0.9913544668587896,
      "grad_norm": 1.0,
      "learning_rate": 0.0004909876713456935,
      "loss": 5.0427,
      "mean_token_accuracy": 0.21591320037841796,
      "num_tokens": 23663807.0,
      "step": 10320
    },
    {
      "entropy": 5.302252626419067,
      "epoch": 0.9918347742555236,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004909780630045534,
      "loss": 4.9678,
      "mean_token_accuracy": 0.22029399871826172,
      "num_tokens": 23674341.0,
      "step": 10325
    },
    {
      "entropy": 5.311558389663697,
      "epoch": 0.9923150816522575,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000490968449649044,
      "loss": 5.0836,
      "mean_token_accuracy": 0.21235835254192353,
      "num_tokens": 23685415.0,
      "step": 10330
    },
    {
      "entropy": 5.357914781570434,
      "epoch": 0.9927953890489913,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004909588312793884,
      "loss": 5.0611,
      "mean_token_accuracy": 0.22123935222625732,
      "num_tokens": 23696652.0,
      "step": 10335
    },
    {
      "entropy": 5.331394386291504,
      "epoch": 0.9932756964457252,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004909492078958101,
      "loss": 4.9951,
      "mean_token_accuracy": 0.22037477791309357,
      "num_tokens": 23708939.0,
      "step": 10340
    },
    {
      "entropy": 5.342646169662475,
      "epoch": 0.9937560038424592,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004909395794985324,
      "loss": 4.991,
      "mean_token_accuracy": 0.21956510096788406,
      "num_tokens": 23720587.0,
      "step": 10345
    },
    {
      "entropy": 5.212669944763183,
      "epoch": 0.9942363112391931,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004909299460877788,
      "loss": 4.9533,
      "mean_token_accuracy": 0.21953347623348235,
      "num_tokens": 23732345.0,
      "step": 10350
    },
    {
      "entropy": 5.322557497024536,
      "epoch": 0.994716618635927,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004909203076637732,
      "loss": 5.032,
      "mean_token_accuracy": 0.21387986242771148,
      "num_tokens": 23743149.0,
      "step": 10355
    },
    {
      "entropy": 5.417452669143676,
      "epoch": 0.9951969260326609,
      "grad_norm": 1.125,
      "learning_rate": 0.0004909106642267392,
      "loss": 5.0155,
      "mean_token_accuracy": 0.22344211488962173,
      "num_tokens": 23755508.0,
      "step": 10360
    },
    {
      "entropy": 5.2970113277435305,
      "epoch": 0.9956772334293948,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004909010157769006,
      "loss": 4.921,
      "mean_token_accuracy": 0.217673259973526,
      "num_tokens": 23766313.0,
      "step": 10365
    },
    {
      "entropy": 5.201802682876587,
      "epoch": 0.9961575408261287,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004908913623144814,
      "loss": 4.9244,
      "mean_token_accuracy": 0.2295621618628502,
      "num_tokens": 23777776.0,
      "step": 10370
    },
    {
      "entropy": 5.279236364364624,
      "epoch": 0.9966378482228626,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000490881703839706,
      "loss": 5.0813,
      "mean_token_accuracy": 0.22236378639936447,
      "num_tokens": 23791143.0,
      "step": 10375
    },
    {
      "entropy": 5.3937296867370605,
      "epoch": 0.9971181556195965,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004908720403527984,
      "loss": 5.0599,
      "mean_token_accuracy": 0.21299023628234864,
      "num_tokens": 23802446.0,
      "step": 10380
    },
    {
      "entropy": 5.227496767044068,
      "epoch": 0.9975984630163305,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000490862371853983,
      "loss": 4.9152,
      "mean_token_accuracy": 0.22675246596336365,
      "num_tokens": 23813757.0,
      "step": 10385
    },
    {
      "entropy": 5.280193233489991,
      "epoch": 0.9980787704130644,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004908526983434844,
      "loss": 4.9924,
      "mean_token_accuracy": 0.21932236850261688,
      "num_tokens": 23824934.0,
      "step": 10390
    },
    {
      "entropy": 5.386311340332031,
      "epoch": 0.9985590778097982,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000490843019821527,
      "loss": 5.105,
      "mean_token_accuracy": 0.21267533898353577,
      "num_tokens": 23836872.0,
      "step": 10395
    },
    {
      "entropy": 5.406396055221558,
      "epoch": 0.9990393852065321,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004908333362883358,
      "loss": 5.0109,
      "mean_token_accuracy": 0.21347676217556,
      "num_tokens": 23847853.0,
      "step": 10400
    },
    {
      "entropy": 5.226173830032349,
      "epoch": 0.9995196926032661,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004908236477441353,
      "loss": 4.9517,
      "mean_token_accuracy": 0.22802724689245224,
      "num_tokens": 23859208.0,
      "step": 10405
    },
    {
      "entropy": 5.218823528289795,
      "epoch": 1.0,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004908139541891505,
      "loss": 4.9163,
      "mean_token_accuracy": 0.2201109617948532,
      "num_tokens": 23868536.0,
      "step": 10410
    },
    {
      "entropy": 5.315934467315674,
      "epoch": 1.0004803073967339,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004908042556236066,
      "loss": 4.9022,
      "mean_token_accuracy": 0.22351286858320235,
      "num_tokens": 23881335.0,
      "step": 10415
    },
    {
      "entropy": 5.269560194015503,
      "epoch": 1.0009606147934678,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004907945520477286,
      "loss": 4.848,
      "mean_token_accuracy": 0.2277863398194313,
      "num_tokens": 23891694.0,
      "step": 10420
    },
    {
      "entropy": 5.242224931716919,
      "epoch": 1.0014409221902016,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004907848434617419,
      "loss": 4.9044,
      "mean_token_accuracy": 0.22318483144044876,
      "num_tokens": 23903560.0,
      "step": 10425
    },
    {
      "entropy": 5.381755304336548,
      "epoch": 1.0019212295869357,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000490775129865872,
      "loss": 5.0927,
      "mean_token_accuracy": 0.20759996324777602,
      "num_tokens": 23914525.0,
      "step": 10430
    },
    {
      "entropy": 5.349568033218384,
      "epoch": 1.0024015369836696,
      "grad_norm": 0.875,
      "learning_rate": 0.0004907654112603442,
      "loss": 4.9658,
      "mean_token_accuracy": 0.22538032680749892,
      "num_tokens": 23926492.0,
      "step": 10435
    },
    {
      "entropy": 5.3642548561096195,
      "epoch": 1.0028818443804035,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004907556876453843,
      "loss": 4.9333,
      "mean_token_accuracy": 0.22279513478279114,
      "num_tokens": 23936326.0,
      "step": 10440
    },
    {
      "entropy": 5.146279811859131,
      "epoch": 1.0033621517771374,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000490745959021218,
      "loss": 4.735,
      "mean_token_accuracy": 0.23274027854204177,
      "num_tokens": 23947173.0,
      "step": 10445
    },
    {
      "entropy": 5.2692399501800535,
      "epoch": 1.0038424591738713,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004907362253880711,
      "loss": 4.9026,
      "mean_token_accuracy": 0.22597576379776002,
      "num_tokens": 23959771.0,
      "step": 10450
    },
    {
      "entropy": 5.293594217300415,
      "epoch": 1.0043227665706052,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004907264867461697,
      "loss": 4.9885,
      "mean_token_accuracy": 0.21834689676761626,
      "num_tokens": 23970908.0,
      "step": 10455
    },
    {
      "entropy": 5.288987302780152,
      "epoch": 1.004803073967339,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004907167430957399,
      "loss": 4.9252,
      "mean_token_accuracy": 0.2204930528998375,
      "num_tokens": 23981640.0,
      "step": 10460
    },
    {
      "entropy": 5.355242347717285,
      "epoch": 1.005283381364073,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004907069944370077,
      "loss": 4.9426,
      "mean_token_accuracy": 0.22529313266277312,
      "num_tokens": 23992913.0,
      "step": 10465
    },
    {
      "entropy": 5.2970287799835205,
      "epoch": 1.005763688760807,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004906972407701998,
      "loss": 4.9308,
      "mean_token_accuracy": 0.21502486914396285,
      "num_tokens": 24003819.0,
      "step": 10470
    },
    {
      "entropy": 5.328417634963989,
      "epoch": 1.006243996157541,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004906874820955423,
      "loss": 4.9532,
      "mean_token_accuracy": 0.22127001434564592,
      "num_tokens": 24013997.0,
      "step": 10475
    },
    {
      "entropy": 5.268332529067993,
      "epoch": 1.0067243035542748,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004906777184132621,
      "loss": 4.9161,
      "mean_token_accuracy": 0.22133830487728118,
      "num_tokens": 24025244.0,
      "step": 10480
    },
    {
      "entropy": 5.234721517562866,
      "epoch": 1.0072046109510087,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004906679497235856,
      "loss": 4.8832,
      "mean_token_accuracy": 0.2266906589269638,
      "num_tokens": 24037701.0,
      "step": 10485
    },
    {
      "entropy": 5.355649375915528,
      "epoch": 1.0076849183477425,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004906581760267397,
      "loss": 4.9593,
      "mean_token_accuracy": 0.21943172961473464,
      "num_tokens": 24049369.0,
      "step": 10490
    },
    {
      "entropy": 5.276905107498169,
      "epoch": 1.0081652257444764,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004906483973229513,
      "loss": 4.9339,
      "mean_token_accuracy": 0.22334463894367218,
      "num_tokens": 24060625.0,
      "step": 10495
    },
    {
      "entropy": 5.365232515335083,
      "epoch": 1.0086455331412103,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004906386136124476,
      "loss": 4.9588,
      "mean_token_accuracy": 0.2181503489613533,
      "num_tokens": 24072527.0,
      "step": 10500
    },
    {
      "entropy": 5.363365411758423,
      "epoch": 1.0091258405379442,
      "grad_norm": 1.0,
      "learning_rate": 0.0004906288248954554,
      "loss": 4.9889,
      "mean_token_accuracy": 0.21111027151346207,
      "num_tokens": 24083979.0,
      "step": 10505
    },
    {
      "entropy": 5.265048456192017,
      "epoch": 1.0096061479346783,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004906190311722023,
      "loss": 4.9505,
      "mean_token_accuracy": 0.22555046230554582,
      "num_tokens": 24095926.0,
      "step": 10510
    },
    {
      "entropy": 5.2602174282073975,
      "epoch": 1.0100864553314122,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004906092324429155,
      "loss": 4.8504,
      "mean_token_accuracy": 0.22824413031339646,
      "num_tokens": 24107898.0,
      "step": 10515
    },
    {
      "entropy": 5.293238401412964,
      "epoch": 1.010566762728146,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004905994287078227,
      "loss": 4.8929,
      "mean_token_accuracy": 0.2249892011284828,
      "num_tokens": 24118713.0,
      "step": 10520
    },
    {
      "entropy": 5.2938319683074955,
      "epoch": 1.01104707012488,
      "grad_norm": 0.875,
      "learning_rate": 0.0004905896199671512,
      "loss": 4.9363,
      "mean_token_accuracy": 0.2259400486946106,
      "num_tokens": 24130410.0,
      "step": 10525
    },
    {
      "entropy": 5.352580642700195,
      "epoch": 1.0115273775216138,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000490579806221129,
      "loss": 4.9537,
      "mean_token_accuracy": 0.2219889521598816,
      "num_tokens": 24140961.0,
      "step": 10530
    },
    {
      "entropy": 5.345658969879151,
      "epoch": 1.0120076849183477,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004905699874699838,
      "loss": 5.0119,
      "mean_token_accuracy": 0.21565217077732085,
      "num_tokens": 24153467.0,
      "step": 10535
    },
    {
      "entropy": 5.200050210952758,
      "epoch": 1.0124879923150816,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004905601637139436,
      "loss": 4.8219,
      "mean_token_accuracy": 0.23107449859380721,
      "num_tokens": 24163873.0,
      "step": 10540
    },
    {
      "entropy": 5.203955078125,
      "epoch": 1.0129682997118155,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004905503349532365,
      "loss": 4.9302,
      "mean_token_accuracy": 0.2254146933555603,
      "num_tokens": 24175852.0,
      "step": 10545
    },
    {
      "entropy": 5.309887933731079,
      "epoch": 1.0134486071085496,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004905405011880906,
      "loss": 4.9605,
      "mean_token_accuracy": 0.2194860890507698,
      "num_tokens": 24187759.0,
      "step": 10550
    },
    {
      "entropy": 5.317674684524536,
      "epoch": 1.0139289145052834,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004905306624187343,
      "loss": 4.862,
      "mean_token_accuracy": 0.2264263555407524,
      "num_tokens": 24199725.0,
      "step": 10555
    },
    {
      "entropy": 5.354394626617432,
      "epoch": 1.0144092219020173,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004905208186453961,
      "loss": 4.9868,
      "mean_token_accuracy": 0.21699429601430892,
      "num_tokens": 24210739.0,
      "step": 10560
    },
    {
      "entropy": 5.3906103610992435,
      "epoch": 1.0148895292987512,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004905109698683044,
      "loss": 5.0439,
      "mean_token_accuracy": 0.2114240989089012,
      "num_tokens": 24221873.0,
      "step": 10565
    },
    {
      "entropy": 5.304267358779907,
      "epoch": 1.015369836695485,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004905011160876878,
      "loss": 4.9171,
      "mean_token_accuracy": 0.22567193657159806,
      "num_tokens": 24233503.0,
      "step": 10570
    },
    {
      "entropy": 5.278196001052857,
      "epoch": 1.015850144092219,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004904912573037753,
      "loss": 4.9246,
      "mean_token_accuracy": 0.219121253490448,
      "num_tokens": 24244465.0,
      "step": 10575
    },
    {
      "entropy": 5.202612066268921,
      "epoch": 1.0163304514889528,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004904813935167957,
      "loss": 4.8166,
      "mean_token_accuracy": 0.23441030383110045,
      "num_tokens": 24255807.0,
      "step": 10580
    },
    {
      "entropy": 5.372105503082276,
      "epoch": 1.0168107588856867,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004904715247269779,
      "loss": 5.0244,
      "mean_token_accuracy": 0.21195382624864578,
      "num_tokens": 24268195.0,
      "step": 10585
    },
    {
      "entropy": 5.233973789215088,
      "epoch": 1.0172910662824208,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004904616509345514,
      "loss": 4.8019,
      "mean_token_accuracy": 0.22695298939943315,
      "num_tokens": 24278915.0,
      "step": 10590
    },
    {
      "entropy": 5.292721700668335,
      "epoch": 1.0177713736791547,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004904517721397449,
      "loss": 4.9691,
      "mean_token_accuracy": 0.2240957200527191,
      "num_tokens": 24289115.0,
      "step": 10595
    },
    {
      "entropy": 5.258921098709107,
      "epoch": 1.0182516810758886,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004904418883427881,
      "loss": 4.9377,
      "mean_token_accuracy": 0.2179606392979622,
      "num_tokens": 24300646.0,
      "step": 10600
    },
    {
      "entropy": 5.42149510383606,
      "epoch": 1.0187319884726225,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004904319995439104,
      "loss": 5.0791,
      "mean_token_accuracy": 0.21247263848781586,
      "num_tokens": 24312344.0,
      "step": 10605
    },
    {
      "entropy": 5.275315761566162,
      "epoch": 1.0192122958693564,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004904221057433412,
      "loss": 4.8725,
      "mean_token_accuracy": 0.22786226868629456,
      "num_tokens": 24323382.0,
      "step": 10610
    },
    {
      "entropy": 5.219202947616577,
      "epoch": 1.0196926032660902,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004904122069413105,
      "loss": 4.8517,
      "mean_token_accuracy": 0.22670889943838118,
      "num_tokens": 24336820.0,
      "step": 10615
    },
    {
      "entropy": 5.310491466522217,
      "epoch": 1.0201729106628241,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000490402303138048,
      "loss": 4.8971,
      "mean_token_accuracy": 0.22554201632738113,
      "num_tokens": 24348547.0,
      "step": 10620
    },
    {
      "entropy": 5.425422525405883,
      "epoch": 1.0206532180595582,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004903923943337836,
      "loss": 5.0558,
      "mean_token_accuracy": 0.21923378854990005,
      "num_tokens": 24359586.0,
      "step": 10625
    },
    {
      "entropy": 5.347232103347778,
      "epoch": 1.021133525456292,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004903824805287475,
      "loss": 4.9764,
      "mean_token_accuracy": 0.21540000438690185,
      "num_tokens": 24371647.0,
      "step": 10630
    },
    {
      "entropy": 5.304556465148925,
      "epoch": 1.021613832853026,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004903725617231696,
      "loss": 4.8535,
      "mean_token_accuracy": 0.22002808898687362,
      "num_tokens": 24382308.0,
      "step": 10635
    },
    {
      "entropy": 5.290077829360962,
      "epoch": 1.0220941402497599,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004903626379172805,
      "loss": 4.9265,
      "mean_token_accuracy": 0.2266160249710083,
      "num_tokens": 24393224.0,
      "step": 10640
    },
    {
      "entropy": 5.243235301971436,
      "epoch": 1.0225744476464937,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004903527091113102,
      "loss": 4.9657,
      "mean_token_accuracy": 0.21775728613138198,
      "num_tokens": 24405220.0,
      "step": 10645
    },
    {
      "entropy": 5.3952381134033205,
      "epoch": 1.0230547550432276,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004903427753054897,
      "loss": 4.988,
      "mean_token_accuracy": 0.21733782887458802,
      "num_tokens": 24416406.0,
      "step": 10650
    },
    {
      "entropy": 5.290863847732544,
      "epoch": 1.0235350624399615,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004903328365000492,
      "loss": 5.0012,
      "mean_token_accuracy": 0.2190077230334282,
      "num_tokens": 24428595.0,
      "step": 10655
    },
    {
      "entropy": 5.298348760604858,
      "epoch": 1.0240153698366954,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004903228926952199,
      "loss": 4.8844,
      "mean_token_accuracy": 0.22837687879800797,
      "num_tokens": 24439568.0,
      "step": 10660
    },
    {
      "entropy": 5.331699371337891,
      "epoch": 1.0244956772334295,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004903129438912322,
      "loss": 4.9465,
      "mean_token_accuracy": 0.21995493173599243,
      "num_tokens": 24451736.0,
      "step": 10665
    },
    {
      "entropy": 5.272909116744995,
      "epoch": 1.0249759846301634,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004903029900883174,
      "loss": 4.9079,
      "mean_token_accuracy": 0.2253137558698654,
      "num_tokens": 24463463.0,
      "step": 10670
    },
    {
      "entropy": 5.291452550888062,
      "epoch": 1.0254562920268973,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004902930312867063,
      "loss": 4.9318,
      "mean_token_accuracy": 0.21667072623968125,
      "num_tokens": 24474425.0,
      "step": 10675
    },
    {
      "entropy": 5.308586931228637,
      "epoch": 1.0259365994236311,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004902830674866306,
      "loss": 4.852,
      "mean_token_accuracy": 0.22820862382650375,
      "num_tokens": 24486482.0,
      "step": 10680
    },
    {
      "entropy": 5.239372491836548,
      "epoch": 1.026416906820365,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004902730986883211,
      "loss": 4.8957,
      "mean_token_accuracy": 0.21790147721767425,
      "num_tokens": 24497709.0,
      "step": 10685
    },
    {
      "entropy": 5.243060779571533,
      "epoch": 1.026897214217099,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004902631248920096,
      "loss": 4.8724,
      "mean_token_accuracy": 0.23006686568260193,
      "num_tokens": 24508406.0,
      "step": 10690
    },
    {
      "entropy": 5.200401067733765,
      "epoch": 1.0273775216138328,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004902531460979274,
      "loss": 4.9302,
      "mean_token_accuracy": 0.22070737928152084,
      "num_tokens": 24520688.0,
      "step": 10695
    },
    {
      "entropy": 5.237102365493774,
      "epoch": 1.0278578290105667,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004902431623063065,
      "loss": 4.8462,
      "mean_token_accuracy": 0.22616141587495803,
      "num_tokens": 24531473.0,
      "step": 10700
    },
    {
      "entropy": 5.283131647109985,
      "epoch": 1.0283381364073008,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004902331735173785,
      "loss": 4.9308,
      "mean_token_accuracy": 0.2180193617939949,
      "num_tokens": 24543263.0,
      "step": 10705
    },
    {
      "entropy": 5.2635664463043215,
      "epoch": 1.0288184438040346,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004902231797313752,
      "loss": 4.8625,
      "mean_token_accuracy": 0.22937071770429612,
      "num_tokens": 24554067.0,
      "step": 10710
    },
    {
      "entropy": 5.199433708190918,
      "epoch": 1.0292987512007685,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004902131809485288,
      "loss": 4.8522,
      "mean_token_accuracy": 0.23227118849754333,
      "num_tokens": 24566291.0,
      "step": 10715
    },
    {
      "entropy": 5.33905177116394,
      "epoch": 1.0297790585975024,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004902031771690713,
      "loss": 4.9766,
      "mean_token_accuracy": 0.2186967045068741,
      "num_tokens": 24579248.0,
      "step": 10720
    },
    {
      "entropy": 5.268744850158692,
      "epoch": 1.0302593659942363,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004901931683932352,
      "loss": 4.8691,
      "mean_token_accuracy": 0.23284034729003905,
      "num_tokens": 24591047.0,
      "step": 10725
    },
    {
      "entropy": 5.299459505081177,
      "epoch": 1.0307396733909702,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004901831546212526,
      "loss": 4.9452,
      "mean_token_accuracy": 0.22014039307832717,
      "num_tokens": 24602454.0,
      "step": 10730
    },
    {
      "entropy": 5.284708547592163,
      "epoch": 1.031219980787704,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004901731358533562,
      "loss": 4.9538,
      "mean_token_accuracy": 0.2242993086576462,
      "num_tokens": 24614168.0,
      "step": 10735
    },
    {
      "entropy": 5.285967206954956,
      "epoch": 1.031700288184438,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004901631120897785,
      "loss": 4.9385,
      "mean_token_accuracy": 0.22476571798324585,
      "num_tokens": 24625588.0,
      "step": 10740
    },
    {
      "entropy": 5.229019451141357,
      "epoch": 1.032180595581172,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004901530833307522,
      "loss": 4.9043,
      "mean_token_accuracy": 0.22213018536567689,
      "num_tokens": 24637744.0,
      "step": 10745
    },
    {
      "entropy": 5.214313268661499,
      "epoch": 1.032660902977906,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004901430495765103,
      "loss": 4.8555,
      "mean_token_accuracy": 0.22907426953315735,
      "num_tokens": 24649631.0,
      "step": 10750
    },
    {
      "entropy": 5.272592163085937,
      "epoch": 1.0331412103746398,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004901330108272855,
      "loss": 4.8942,
      "mean_token_accuracy": 0.2248514175415039,
      "num_tokens": 24660622.0,
      "step": 10755
    },
    {
      "entropy": 5.3516308784484865,
      "epoch": 1.0336215177713737,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004901229670833111,
      "loss": 4.9139,
      "mean_token_accuracy": 0.22330948263406752,
      "num_tokens": 24671072.0,
      "step": 10760
    },
    {
      "entropy": 5.275476169586182,
      "epoch": 1.0341018251681076,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004901129183448201,
      "loss": 4.9628,
      "mean_token_accuracy": 0.21949354708194732,
      "num_tokens": 24683160.0,
      "step": 10765
    },
    {
      "entropy": 5.328217363357544,
      "epoch": 1.0345821325648414,
      "grad_norm": 0.875,
      "learning_rate": 0.0004901028646120459,
      "loss": 4.9684,
      "mean_token_accuracy": 0.22561192214488984,
      "num_tokens": 24694703.0,
      "step": 10770
    },
    {
      "entropy": 5.29872875213623,
      "epoch": 1.0350624399615753,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000490092805885222,
      "loss": 4.897,
      "mean_token_accuracy": 0.22266589254140853,
      "num_tokens": 24706429.0,
      "step": 10775
    },
    {
      "entropy": 5.225975561141968,
      "epoch": 1.0355427473583094,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004900827421645816,
      "loss": 4.9274,
      "mean_token_accuracy": 0.2213777184486389,
      "num_tokens": 24717682.0,
      "step": 10780
    },
    {
      "entropy": 5.363853311538696,
      "epoch": 1.0360230547550433,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004900726734503589,
      "loss": 4.9896,
      "mean_token_accuracy": 0.22104839384555816,
      "num_tokens": 24728789.0,
      "step": 10785
    },
    {
      "entropy": 5.243290376663208,
      "epoch": 1.0365033621517772,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004900625997427872,
      "loss": 4.8095,
      "mean_token_accuracy": 0.23038740009069442,
      "num_tokens": 24739788.0,
      "step": 10790
    },
    {
      "entropy": 5.266545343399048,
      "epoch": 1.036983669548511,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004900525210421006,
      "loss": 4.9494,
      "mean_token_accuracy": 0.2233280599117279,
      "num_tokens": 24751828.0,
      "step": 10795
    },
    {
      "entropy": 5.307193326950073,
      "epoch": 1.037463976945245,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004900424373485329,
      "loss": 4.8922,
      "mean_token_accuracy": 0.2292135775089264,
      "num_tokens": 24762061.0,
      "step": 10800
    },
    {
      "entropy": 5.243637323379517,
      "epoch": 1.0379442843419788,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004900323486623185,
      "loss": 4.9002,
      "mean_token_accuracy": 0.22845154404640197,
      "num_tokens": 24774018.0,
      "step": 10805
    },
    {
      "entropy": 5.270729827880859,
      "epoch": 1.0384245917387127,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004900222549836914,
      "loss": 4.8817,
      "mean_token_accuracy": 0.22876522392034532,
      "num_tokens": 24785206.0,
      "step": 10810
    },
    {
      "entropy": 5.289447450637818,
      "epoch": 1.0389048991354466,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000490012156312886,
      "loss": 4.9464,
      "mean_token_accuracy": 0.2215624064207077,
      "num_tokens": 24797730.0,
      "step": 10815
    },
    {
      "entropy": 5.3395623683929445,
      "epoch": 1.0393852065321807,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004900020526501369,
      "loss": 5.0364,
      "mean_token_accuracy": 0.2170681908726692,
      "num_tokens": 24809267.0,
      "step": 10820
    },
    {
      "entropy": 5.322444725036621,
      "epoch": 1.0398655139289146,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004899919439956785,
      "loss": 4.8967,
      "mean_token_accuracy": 0.2186542496085167,
      "num_tokens": 24820461.0,
      "step": 10825
    },
    {
      "entropy": 5.234636449813843,
      "epoch": 1.0403458213256485,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004899818303497455,
      "loss": 4.9393,
      "mean_token_accuracy": 0.22122032344341278,
      "num_tokens": 24831959.0,
      "step": 10830
    },
    {
      "entropy": 5.269577407836914,
      "epoch": 1.0408261287223823,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004899717117125728,
      "loss": 4.9114,
      "mean_token_accuracy": 0.22643207162618637,
      "num_tokens": 24843655.0,
      "step": 10835
    },
    {
      "entropy": 5.248018360137939,
      "epoch": 1.0413064361191162,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004899615880843953,
      "loss": 4.8739,
      "mean_token_accuracy": 0.22124601602554322,
      "num_tokens": 24855643.0,
      "step": 10840
    },
    {
      "entropy": 5.284094429016113,
      "epoch": 1.04178674351585,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004899514594654481,
      "loss": 4.8458,
      "mean_token_accuracy": 0.23002680540084838,
      "num_tokens": 24865745.0,
      "step": 10845
    },
    {
      "entropy": 5.234852123260498,
      "epoch": 1.042267050912584,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004899413258559662,
      "loss": 4.8455,
      "mean_token_accuracy": 0.228420490026474,
      "num_tokens": 24876567.0,
      "step": 10850
    },
    {
      "entropy": 5.207185363769531,
      "epoch": 1.0427473583093179,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004899311872561849,
      "loss": 4.8429,
      "mean_token_accuracy": 0.22723314464092254,
      "num_tokens": 24888294.0,
      "step": 10855
    },
    {
      "entropy": 5.308998060226441,
      "epoch": 1.043227665706052,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004899210436663398,
      "loss": 5.04,
      "mean_token_accuracy": 0.21451251059770585,
      "num_tokens": 24899559.0,
      "step": 10860
    },
    {
      "entropy": 5.309617757797241,
      "epoch": 1.0437079731027858,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004899108950866661,
      "loss": 4.9039,
      "mean_token_accuracy": 0.22729567885398866,
      "num_tokens": 24911205.0,
      "step": 10865
    },
    {
      "entropy": 5.307304525375367,
      "epoch": 1.0441882804995197,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004899007415173997,
      "loss": 4.895,
      "mean_token_accuracy": 0.22373060137033463,
      "num_tokens": 24922635.0,
      "step": 10870
    },
    {
      "entropy": 5.248117589950562,
      "epoch": 1.0446685878962536,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004898905829587762,
      "loss": 4.8904,
      "mean_token_accuracy": 0.22630150914192199,
      "num_tokens": 24932941.0,
      "step": 10875
    },
    {
      "entropy": 5.333377552032471,
      "epoch": 1.0451488952929875,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004898804194110313,
      "loss": 4.9715,
      "mean_token_accuracy": 0.22128599137067795,
      "num_tokens": 24943980.0,
      "step": 10880
    },
    {
      "entropy": 5.296769142150879,
      "epoch": 1.0456292026897214,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004898702508744012,
      "loss": 4.9242,
      "mean_token_accuracy": 0.2180767312645912,
      "num_tokens": 24955233.0,
      "step": 10885
    },
    {
      "entropy": 5.300024557113647,
      "epoch": 1.0461095100864553,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004898600773491221,
      "loss": 4.9697,
      "mean_token_accuracy": 0.2209122970700264,
      "num_tokens": 24966971.0,
      "step": 10890
    },
    {
      "entropy": 5.337382221221924,
      "epoch": 1.0465898174831891,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004898498988354297,
      "loss": 4.9464,
      "mean_token_accuracy": 0.21950223296880722,
      "num_tokens": 24977515.0,
      "step": 10895
    },
    {
      "entropy": 5.265654802322388,
      "epoch": 1.0470701248799232,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004898397153335608,
      "loss": 4.8928,
      "mean_token_accuracy": 0.22675986140966414,
      "num_tokens": 24988131.0,
      "step": 10900
    },
    {
      "entropy": 5.304479598999023,
      "epoch": 1.0475504322766571,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004898295268437517,
      "loss": 4.9763,
      "mean_token_accuracy": 0.22106654793024064,
      "num_tokens": 24999238.0,
      "step": 10905
    },
    {
      "entropy": 5.339041090011596,
      "epoch": 1.048030739673391,
      "grad_norm": 1.0,
      "learning_rate": 0.0004898193333662388,
      "loss": 5.0013,
      "mean_token_accuracy": 0.21588689237833023,
      "num_tokens": 25011443.0,
      "step": 10910
    },
    {
      "entropy": 5.304559659957886,
      "epoch": 1.0485110470701249,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004898091349012588,
      "loss": 4.8461,
      "mean_token_accuracy": 0.22937976717948913,
      "num_tokens": 25022151.0,
      "step": 10915
    },
    {
      "entropy": 5.310842800140381,
      "epoch": 1.0489913544668588,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004897989314490486,
      "loss": 4.9393,
      "mean_token_accuracy": 0.22364698499441146,
      "num_tokens": 25032730.0,
      "step": 10920
    },
    {
      "entropy": 5.1821037292480465,
      "epoch": 1.0494716618635926,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004897887230098451,
      "loss": 4.8784,
      "mean_token_accuracy": 0.22659228295087813,
      "num_tokens": 25044455.0,
      "step": 10925
    },
    {
      "entropy": 5.221290874481201,
      "epoch": 1.0499519692603265,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004897785095838852,
      "loss": 4.8413,
      "mean_token_accuracy": 0.23516307175159454,
      "num_tokens": 25056647.0,
      "step": 10930
    },
    {
      "entropy": 5.26479868888855,
      "epoch": 1.0504322766570606,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004897682911714061,
      "loss": 4.8504,
      "mean_token_accuracy": 0.23068436980247498,
      "num_tokens": 25067618.0,
      "step": 10935
    },
    {
      "entropy": 5.2493171215057375,
      "epoch": 1.0509125840537945,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000489758067772645,
      "loss": 4.8582,
      "mean_token_accuracy": 0.22356654554605485,
      "num_tokens": 25079753.0,
      "step": 10940
    },
    {
      "entropy": 5.250501203536987,
      "epoch": 1.0513928914505284,
      "grad_norm": 1.0,
      "learning_rate": 0.0004897478393878392,
      "loss": 4.9325,
      "mean_token_accuracy": 0.22210078090429305,
      "num_tokens": 25091023.0,
      "step": 10945
    },
    {
      "entropy": 5.3351030349731445,
      "epoch": 1.0518731988472623,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004897376060172264,
      "loss": 4.9243,
      "mean_token_accuracy": 0.2184640273451805,
      "num_tokens": 25101396.0,
      "step": 10950
    },
    {
      "entropy": 5.283926677703858,
      "epoch": 1.0523535062439962,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004897273676610438,
      "loss": 4.9573,
      "mean_token_accuracy": 0.22059548944234847,
      "num_tokens": 25112562.0,
      "step": 10955
    },
    {
      "entropy": 5.288591718673706,
      "epoch": 1.05283381364073,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004897171243195295,
      "loss": 4.9483,
      "mean_token_accuracy": 0.22751814275979995,
      "num_tokens": 25124401.0,
      "step": 10960
    },
    {
      "entropy": 5.291896915435791,
      "epoch": 1.053314121037464,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000489706875992921,
      "loss": 4.9155,
      "mean_token_accuracy": 0.22412931323051452,
      "num_tokens": 25137345.0,
      "step": 10965
    },
    {
      "entropy": 5.281894207000732,
      "epoch": 1.0537944284341978,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004896966226814565,
      "loss": 4.9528,
      "mean_token_accuracy": 0.22082947045564652,
      "num_tokens": 25149300.0,
      "step": 10970
    },
    {
      "entropy": 5.276123809814453,
      "epoch": 1.054274735830932,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004896863643853739,
      "loss": 4.8883,
      "mean_token_accuracy": 0.2278333455324173,
      "num_tokens": 25159510.0,
      "step": 10975
    },
    {
      "entropy": 5.285507297515869,
      "epoch": 1.0547550432276658,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004896761011049114,
      "loss": 4.8414,
      "mean_token_accuracy": 0.23181968331336975,
      "num_tokens": 25172346.0,
      "step": 10980
    },
    {
      "entropy": 5.327755784988403,
      "epoch": 1.0552353506243997,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004896658328403074,
      "loss": 5.0211,
      "mean_token_accuracy": 0.22096251398324968,
      "num_tokens": 25183471.0,
      "step": 10985
    },
    {
      "entropy": 5.34070611000061,
      "epoch": 1.0557156580211335,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004896555595918001,
      "loss": 4.9515,
      "mean_token_accuracy": 0.22527804374694824,
      "num_tokens": 25195067.0,
      "step": 10990
    },
    {
      "entropy": 5.3092926979064945,
      "epoch": 1.0561959654178674,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004896452813596281,
      "loss": 4.9381,
      "mean_token_accuracy": 0.22057298123836516,
      "num_tokens": 25206165.0,
      "step": 10995
    },
    {
      "entropy": 5.251141166687011,
      "epoch": 1.0566762728146013,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004896349981440301,
      "loss": 4.8879,
      "mean_token_accuracy": 0.23374197334051133,
      "num_tokens": 25217229.0,
      "step": 11000
    },
    {
      "entropy": 5.343147706985474,
      "epoch": 1.0571565802113352,
      "grad_norm": 1.0,
      "learning_rate": 0.0004896247099452447,
      "loss": 5.0186,
      "mean_token_accuracy": 0.21516940295696257,
      "num_tokens": 25229349.0,
      "step": 11005
    },
    {
      "entropy": 5.304940462112427,
      "epoch": 1.057636887608069,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004896144167635108,
      "loss": 4.9068,
      "mean_token_accuracy": 0.22352226972579955,
      "num_tokens": 25240367.0,
      "step": 11010
    },
    {
      "entropy": 5.303400135040283,
      "epoch": 1.0581171950048032,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004896041185990675,
      "loss": 4.9871,
      "mean_token_accuracy": 0.2135757774114609,
      "num_tokens": 25252025.0,
      "step": 11015
    },
    {
      "entropy": 5.272900867462158,
      "epoch": 1.058597502401537,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004895938154521538,
      "loss": 4.9017,
      "mean_token_accuracy": 0.224807009100914,
      "num_tokens": 25264846.0,
      "step": 11020
    },
    {
      "entropy": 5.24267053604126,
      "epoch": 1.059077809798271,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004895835073230089,
      "loss": 4.8933,
      "mean_token_accuracy": 0.224464151263237,
      "num_tokens": 25276150.0,
      "step": 11025
    },
    {
      "entropy": 5.289191722869873,
      "epoch": 1.0595581171950048,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004895731942118722,
      "loss": 4.905,
      "mean_token_accuracy": 0.22091283649206161,
      "num_tokens": 25287819.0,
      "step": 11030
    },
    {
      "entropy": 5.326931190490723,
      "epoch": 1.0600384245917387,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004895628761189829,
      "loss": 4.9296,
      "mean_token_accuracy": 0.22610394060611724,
      "num_tokens": 25300467.0,
      "step": 11035
    },
    {
      "entropy": 5.203870248794556,
      "epoch": 1.0605187319884726,
      "grad_norm": 0.875,
      "learning_rate": 0.0004895525530445809,
      "loss": 4.8957,
      "mean_token_accuracy": 0.22015904039144515,
      "num_tokens": 25312447.0,
      "step": 11040
    },
    {
      "entropy": 5.3005616664886475,
      "epoch": 1.0609990393852065,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004895422249889057,
      "loss": 4.9815,
      "mean_token_accuracy": 0.218388332426548,
      "num_tokens": 25324160.0,
      "step": 11045
    },
    {
      "entropy": 5.278023433685303,
      "epoch": 1.0614793467819403,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004895318919521971,
      "loss": 4.8267,
      "mean_token_accuracy": 0.23409009575843812,
      "num_tokens": 25335106.0,
      "step": 11050
    },
    {
      "entropy": 5.207167720794677,
      "epoch": 1.0619596541786744,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004895215539346949,
      "loss": 4.8096,
      "mean_token_accuracy": 0.2274819865822792,
      "num_tokens": 25345246.0,
      "step": 11055
    },
    {
      "entropy": 5.277943706512451,
      "epoch": 1.0624399615754083,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004895112109366393,
      "loss": 4.9383,
      "mean_token_accuracy": 0.218681001663208,
      "num_tokens": 25356665.0,
      "step": 11060
    },
    {
      "entropy": 5.268387699127198,
      "epoch": 1.0629202689721422,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004895008629582703,
      "loss": 4.8828,
      "mean_token_accuracy": 0.22639892548322677,
      "num_tokens": 25368128.0,
      "step": 11065
    },
    {
      "entropy": 5.239960241317749,
      "epoch": 1.063400576368876,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004894905099998283,
      "loss": 4.8883,
      "mean_token_accuracy": 0.225995209813118,
      "num_tokens": 25379061.0,
      "step": 11070
    },
    {
      "entropy": 5.291909074783325,
      "epoch": 1.06388088376561,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004894801520615535,
      "loss": 4.8939,
      "mean_token_accuracy": 0.2299271821975708,
      "num_tokens": 25390157.0,
      "step": 11075
    },
    {
      "entropy": 5.2465818405151365,
      "epoch": 1.0643611911623438,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004894697891436863,
      "loss": 4.8906,
      "mean_token_accuracy": 0.2228396341204643,
      "num_tokens": 25400446.0,
      "step": 11080
    },
    {
      "entropy": 5.287621021270752,
      "epoch": 1.0648414985590777,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004894594212464676,
      "loss": 4.9352,
      "mean_token_accuracy": 0.22121475785970687,
      "num_tokens": 25412064.0,
      "step": 11085
    },
    {
      "entropy": 5.259828042984009,
      "epoch": 1.0653218059558118,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004894490483701381,
      "loss": 4.9297,
      "mean_token_accuracy": 0.2274885281920433,
      "num_tokens": 25422898.0,
      "step": 11090
    },
    {
      "entropy": 5.2978808879852295,
      "epoch": 1.0658021133525457,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004894386705149382,
      "loss": 4.9047,
      "mean_token_accuracy": 0.21848641186952591,
      "num_tokens": 25434318.0,
      "step": 11095
    },
    {
      "entropy": 5.2812341213226315,
      "epoch": 1.0662824207492796,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004894282876811093,
      "loss": 4.8607,
      "mean_token_accuracy": 0.22751497328281403,
      "num_tokens": 25444941.0,
      "step": 11100
    },
    {
      "entropy": 5.262346172332764,
      "epoch": 1.0667627281460135,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004894178998688921,
      "loss": 4.8988,
      "mean_token_accuracy": 0.2232814759016037,
      "num_tokens": 25455971.0,
      "step": 11105
    },
    {
      "entropy": 5.2257513999938965,
      "epoch": 1.0672430355427474,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004894075070785281,
      "loss": 4.8581,
      "mean_token_accuracy": 0.22697691470384598,
      "num_tokens": 25467587.0,
      "step": 11110
    },
    {
      "entropy": 5.235648345947266,
      "epoch": 1.0677233429394812,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004893971093102585,
      "loss": 4.8792,
      "mean_token_accuracy": 0.23085685074329376,
      "num_tokens": 25479137.0,
      "step": 11115
    },
    {
      "entropy": 5.378875732421875,
      "epoch": 1.0682036503362151,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004893867065643245,
      "loss": 5.0194,
      "mean_token_accuracy": 0.2175419107079506,
      "num_tokens": 25491755.0,
      "step": 11120
    },
    {
      "entropy": 5.208220291137695,
      "epoch": 1.068683957732949,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004893762988409678,
      "loss": 4.8188,
      "mean_token_accuracy": 0.23100871443748475,
      "num_tokens": 25503581.0,
      "step": 11125
    },
    {
      "entropy": 5.2728108882904055,
      "epoch": 1.069164265129683,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004893658861404301,
      "loss": 4.9399,
      "mean_token_accuracy": 0.22313530147075653,
      "num_tokens": 25515032.0,
      "step": 11130
    },
    {
      "entropy": 5.269526481628418,
      "epoch": 1.069644572526417,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004893554684629529,
      "loss": 4.8908,
      "mean_token_accuracy": 0.22198642939329147,
      "num_tokens": 25526347.0,
      "step": 11135
    },
    {
      "entropy": 5.33236517906189,
      "epoch": 1.0701248799231509,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004893450458087784,
      "loss": 4.9308,
      "mean_token_accuracy": 0.22438198775053025,
      "num_tokens": 25538287.0,
      "step": 11140
    },
    {
      "entropy": 5.168224573135376,
      "epoch": 1.0706051873198847,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004893346181781483,
      "loss": 4.8839,
      "mean_token_accuracy": 0.2358560785651207,
      "num_tokens": 25548431.0,
      "step": 11145
    },
    {
      "entropy": 5.2847521781921385,
      "epoch": 1.0710854947166186,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004893241855713048,
      "loss": 5.0198,
      "mean_token_accuracy": 0.21750132590532303,
      "num_tokens": 25558706.0,
      "step": 11150
    },
    {
      "entropy": 5.347308874130249,
      "epoch": 1.0715658021133525,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004893137479884903,
      "loss": 4.9097,
      "mean_token_accuracy": 0.22480886429548264,
      "num_tokens": 25569920.0,
      "step": 11155
    },
    {
      "entropy": 5.253406953811646,
      "epoch": 1.0720461095100864,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004893033054299468,
      "loss": 4.8449,
      "mean_token_accuracy": 0.2349695399403572,
      "num_tokens": 25581340.0,
      "step": 11160
    },
    {
      "entropy": 5.264286136627197,
      "epoch": 1.0725264169068203,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000489292857895917,
      "loss": 4.94,
      "mean_token_accuracy": 0.21906747072935104,
      "num_tokens": 25592477.0,
      "step": 11165
    },
    {
      "entropy": 5.2814020156860355,
      "epoch": 1.0730067243035544,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004892824053866432,
      "loss": 4.882,
      "mean_token_accuracy": 0.23097696602344514,
      "num_tokens": 25604583.0,
      "step": 11170
    },
    {
      "entropy": 5.386000156402588,
      "epoch": 1.0734870317002883,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004892719479023683,
      "loss": 5.0279,
      "mean_token_accuracy": 0.2186363309621811,
      "num_tokens": 25614743.0,
      "step": 11175
    },
    {
      "entropy": 5.161961412429809,
      "epoch": 1.0739673390970221,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000489261485443335,
      "loss": 4.7523,
      "mean_token_accuracy": 0.23406897485256195,
      "num_tokens": 25626796.0,
      "step": 11180
    },
    {
      "entropy": 5.281435346603393,
      "epoch": 1.074447646493756,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004892510180097863,
      "loss": 4.8903,
      "mean_token_accuracy": 0.2238363355398178,
      "num_tokens": 25638014.0,
      "step": 11185
    },
    {
      "entropy": 5.232827520370483,
      "epoch": 1.07492795389049,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004892405456019651,
      "loss": 4.9006,
      "mean_token_accuracy": 0.23067580610513688,
      "num_tokens": 25649555.0,
      "step": 11190
    },
    {
      "entropy": 5.299401807785034,
      "epoch": 1.0754082612872238,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004892300682201147,
      "loss": 5.0072,
      "mean_token_accuracy": 0.22073896676301957,
      "num_tokens": 25662886.0,
      "step": 11195
    },
    {
      "entropy": 5.319880485534668,
      "epoch": 1.0758885686839577,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004892195858644782,
      "loss": 4.8675,
      "mean_token_accuracy": 0.2300463005900383,
      "num_tokens": 25673802.0,
      "step": 11200
    },
    {
      "entropy": 5.239248657226563,
      "epoch": 1.0763688760806915,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000489209098535299,
      "loss": 4.9001,
      "mean_token_accuracy": 0.23211382031440736,
      "num_tokens": 25685545.0,
      "step": 11205
    },
    {
      "entropy": 5.225132942199707,
      "epoch": 1.0768491834774256,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004891986062328205,
      "loss": 4.8381,
      "mean_token_accuracy": 0.22820284813642502,
      "num_tokens": 25697955.0,
      "step": 11210
    },
    {
      "entropy": 5.278775978088379,
      "epoch": 1.0773294908741595,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004891881089572865,
      "loss": 4.9199,
      "mean_token_accuracy": 0.22155346870422363,
      "num_tokens": 25709707.0,
      "step": 11215
    },
    {
      "entropy": 5.302041721343994,
      "epoch": 1.0778097982708934,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004891776067089406,
      "loss": 4.9675,
      "mean_token_accuracy": 0.22354414016008378,
      "num_tokens": 25720576.0,
      "step": 11220
    },
    {
      "entropy": 5.271542644500732,
      "epoch": 1.0782901056676273,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004891670994880266,
      "loss": 4.8471,
      "mean_token_accuracy": 0.2279942363500595,
      "num_tokens": 25731387.0,
      "step": 11225
    },
    {
      "entropy": 5.211033296585083,
      "epoch": 1.0787704130643612,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004891565872947888,
      "loss": 4.8348,
      "mean_token_accuracy": 0.23124773502349855,
      "num_tokens": 25741813.0,
      "step": 11230
    },
    {
      "entropy": 5.229121494293213,
      "epoch": 1.079250720461095,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004891460701294706,
      "loss": 4.9097,
      "mean_token_accuracy": 0.22490486949682237,
      "num_tokens": 25753763.0,
      "step": 11235
    },
    {
      "entropy": 5.249711036682129,
      "epoch": 1.079731027857829,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004891355479923167,
      "loss": 4.8694,
      "mean_token_accuracy": 0.2245772212743759,
      "num_tokens": 25765269.0,
      "step": 11240
    },
    {
      "entropy": 5.267370986938476,
      "epoch": 1.080211335254563,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004891250208835712,
      "loss": 4.915,
      "mean_token_accuracy": 0.2294594466686249,
      "num_tokens": 25776841.0,
      "step": 11245
    },
    {
      "entropy": 5.354419898986817,
      "epoch": 1.080691642651297,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004891144888034784,
      "loss": 4.9501,
      "mean_token_accuracy": 0.21193725615739822,
      "num_tokens": 25788151.0,
      "step": 11250
    },
    {
      "entropy": 5.284690523147583,
      "epoch": 1.0811719500480308,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004891039517522832,
      "loss": 4.8536,
      "mean_token_accuracy": 0.22946816980838775,
      "num_tokens": 25799194.0,
      "step": 11255
    },
    {
      "entropy": 5.2536064147949215,
      "epoch": 1.0816522574447647,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004890934097302299,
      "loss": 4.8953,
      "mean_token_accuracy": 0.22067134231328964,
      "num_tokens": 25809483.0,
      "step": 11260
    },
    {
      "entropy": 5.20205659866333,
      "epoch": 1.0821325648414986,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004890828627375632,
      "loss": 4.8791,
      "mean_token_accuracy": 0.21775562167167664,
      "num_tokens": 25820781.0,
      "step": 11265
    },
    {
      "entropy": 5.3336633205413815,
      "epoch": 1.0826128722382324,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004890723107745283,
      "loss": 4.9553,
      "mean_token_accuracy": 0.21996574401855468,
      "num_tokens": 25832111.0,
      "step": 11270
    },
    {
      "entropy": 5.330847072601318,
      "epoch": 1.0830931796349663,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004890617538413699,
      "loss": 4.9711,
      "mean_token_accuracy": 0.22579692006111146,
      "num_tokens": 25843300.0,
      "step": 11275
    },
    {
      "entropy": 5.279792261123657,
      "epoch": 1.0835734870317002,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004890511919383333,
      "loss": 4.8949,
      "mean_token_accuracy": 0.2203872263431549,
      "num_tokens": 25855337.0,
      "step": 11280
    },
    {
      "entropy": 5.379551601409912,
      "epoch": 1.084053794428434,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004890406250656636,
      "loss": 5.0001,
      "mean_token_accuracy": 0.21658748686313628,
      "num_tokens": 25866683.0,
      "step": 11285
    },
    {
      "entropy": 5.2679853439331055,
      "epoch": 1.0845341018251682,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004890300532236062,
      "loss": 4.8772,
      "mean_token_accuracy": 0.22585312724113465,
      "num_tokens": 25878041.0,
      "step": 11290
    },
    {
      "entropy": 5.306599712371826,
      "epoch": 1.085014409221902,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004890194764124064,
      "loss": 4.8871,
      "mean_token_accuracy": 0.22083671241998673,
      "num_tokens": 25889059.0,
      "step": 11295
    },
    {
      "entropy": 5.203682804107666,
      "epoch": 1.085494716618636,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004890088946323099,
      "loss": 4.8062,
      "mean_token_accuracy": 0.23782912343740464,
      "num_tokens": 25900062.0,
      "step": 11300
    },
    {
      "entropy": 5.292300462722778,
      "epoch": 1.0859750240153698,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004889983078835623,
      "loss": 5.0089,
      "mean_token_accuracy": 0.2220240965485573,
      "num_tokens": 25911813.0,
      "step": 11305
    },
    {
      "entropy": 5.313268041610717,
      "epoch": 1.0864553314121037,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004889877161664096,
      "loss": 4.8269,
      "mean_token_accuracy": 0.23053346276283265,
      "num_tokens": 25922321.0,
      "step": 11310
    },
    {
      "entropy": 5.281772327423096,
      "epoch": 1.0869356388088376,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004889771194810974,
      "loss": 4.9015,
      "mean_token_accuracy": 0.2300559923052788,
      "num_tokens": 25933631.0,
      "step": 11315
    },
    {
      "entropy": 5.261170864105225,
      "epoch": 1.0874159462055715,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004889665178278719,
      "loss": 4.9627,
      "mean_token_accuracy": 0.2196210131049156,
      "num_tokens": 25945190.0,
      "step": 11320
    },
    {
      "entropy": 5.3156407356262205,
      "epoch": 1.0878962536023056,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004889559112069792,
      "loss": 4.8942,
      "mean_token_accuracy": 0.23000511825084685,
      "num_tokens": 25956817.0,
      "step": 11325
    },
    {
      "entropy": 5.382984828948975,
      "epoch": 1.0883765609990395,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004889452996186657,
      "loss": 4.9565,
      "mean_token_accuracy": 0.22990044057369233,
      "num_tokens": 25968134.0,
      "step": 11330
    },
    {
      "entropy": 5.195108699798584,
      "epoch": 1.0888568683957733,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004889346830631774,
      "loss": 4.8764,
      "mean_token_accuracy": 0.22957141250371932,
      "num_tokens": 25981825.0,
      "step": 11335
    },
    {
      "entropy": 5.354644441604615,
      "epoch": 1.0893371757925072,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000488924061540761,
      "loss": 4.989,
      "mean_token_accuracy": 0.21383907198905944,
      "num_tokens": 25993962.0,
      "step": 11340
    },
    {
      "entropy": 5.251808309555054,
      "epoch": 1.089817483189241,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004889134350516633,
      "loss": 4.869,
      "mean_token_accuracy": 0.22989823073148727,
      "num_tokens": 26005705.0,
      "step": 11345
    },
    {
      "entropy": 5.247726535797119,
      "epoch": 1.090297790585975,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004889028035961308,
      "loss": 4.9054,
      "mean_token_accuracy": 0.22980657815933228,
      "num_tokens": 26017111.0,
      "step": 11350
    },
    {
      "entropy": 5.3020837783813475,
      "epoch": 1.0907780979827089,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004888921671744103,
      "loss": 4.9531,
      "mean_token_accuracy": 0.2205898270010948,
      "num_tokens": 26028097.0,
      "step": 11355
    },
    {
      "entropy": 5.282825708389282,
      "epoch": 1.0912584053794427,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004888815257867488,
      "loss": 4.8944,
      "mean_token_accuracy": 0.22508398592472076,
      "num_tokens": 26039281.0,
      "step": 11360
    },
    {
      "entropy": 5.28433895111084,
      "epoch": 1.0917387127761768,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004888708794333934,
      "loss": 4.9677,
      "mean_token_accuracy": 0.2168301224708557,
      "num_tokens": 26051133.0,
      "step": 11365
    },
    {
      "entropy": 5.2489758968353275,
      "epoch": 1.0922190201729107,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004888602281145913,
      "loss": 4.9493,
      "mean_token_accuracy": 0.22295374274253846,
      "num_tokens": 26062220.0,
      "step": 11370
    },
    {
      "entropy": 5.262448072433472,
      "epoch": 1.0926993275696446,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004888495718305897,
      "loss": 4.8614,
      "mean_token_accuracy": 0.23666325956583023,
      "num_tokens": 26072556.0,
      "step": 11375
    },
    {
      "entropy": 5.326375198364258,
      "epoch": 1.0931796349663785,
      "grad_norm": 1.015625,
      "learning_rate": 0.000488838910581636,
      "loss": 4.9927,
      "mean_token_accuracy": 0.21592565327882768,
      "num_tokens": 26084218.0,
      "step": 11380
    },
    {
      "entropy": 5.37132420539856,
      "epoch": 1.0936599423631124,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004888282443679777,
      "loss": 5.0797,
      "mean_token_accuracy": 0.21330417394638063,
      "num_tokens": 26096272.0,
      "step": 11385
    },
    {
      "entropy": 5.264055013656616,
      "epoch": 1.0941402497598463,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004888175731898627,
      "loss": 4.9507,
      "mean_token_accuracy": 0.2227497547864914,
      "num_tokens": 26109247.0,
      "step": 11390
    },
    {
      "entropy": 5.270467901229859,
      "epoch": 1.0946205571565801,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004888068970475384,
      "loss": 4.8863,
      "mean_token_accuracy": 0.22501642853021622,
      "num_tokens": 26121340.0,
      "step": 11395
    },
    {
      "entropy": 5.277775669097901,
      "epoch": 1.0951008645533142,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004887962159412529,
      "loss": 4.9074,
      "mean_token_accuracy": 0.22068292051553726,
      "num_tokens": 26133135.0,
      "step": 11400
    },
    {
      "entropy": 5.243629837036133,
      "epoch": 1.0955811719500481,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004887855298712541,
      "loss": 4.8236,
      "mean_token_accuracy": 0.2285383552312851,
      "num_tokens": 26145324.0,
      "step": 11405
    },
    {
      "entropy": 5.174122524261475,
      "epoch": 1.096061479346782,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048877483883779,
      "loss": 4.8322,
      "mean_token_accuracy": 0.22797962427139282,
      "num_tokens": 26156739.0,
      "step": 11410
    },
    {
      "entropy": 5.263250827789307,
      "epoch": 1.0965417867435159,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004887641428411091,
      "loss": 4.9209,
      "mean_token_accuracy": 0.22310953140258788,
      "num_tokens": 26169543.0,
      "step": 11415
    },
    {
      "entropy": 5.290619421005249,
      "epoch": 1.0970220941402498,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004887534418814595,
      "loss": 4.915,
      "mean_token_accuracy": 0.22453030049800873,
      "num_tokens": 26181317.0,
      "step": 11420
    },
    {
      "entropy": 5.327699470520019,
      "epoch": 1.0975024015369836,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004887427359590897,
      "loss": 4.993,
      "mean_token_accuracy": 0.2185678228735924,
      "num_tokens": 26193402.0,
      "step": 11425
    },
    {
      "entropy": 5.230723142623901,
      "epoch": 1.0979827089337175,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004887320250742482,
      "loss": 4.8685,
      "mean_token_accuracy": 0.22495629042387008,
      "num_tokens": 26204660.0,
      "step": 11430
    },
    {
      "entropy": 5.357244634628296,
      "epoch": 1.0984630163304514,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004887213092271838,
      "loss": 4.9163,
      "mean_token_accuracy": 0.22498188465833663,
      "num_tokens": 26215593.0,
      "step": 11435
    },
    {
      "entropy": 5.212852239608765,
      "epoch": 1.0989433237271853,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004887105884181451,
      "loss": 4.8371,
      "mean_token_accuracy": 0.23485222309827805,
      "num_tokens": 26227481.0,
      "step": 11440
    },
    {
      "entropy": 5.281425428390503,
      "epoch": 1.0994236311239194,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004886998626473813,
      "loss": 4.9909,
      "mean_token_accuracy": 0.21913453191518784,
      "num_tokens": 26238260.0,
      "step": 11445
    },
    {
      "entropy": 5.3118151187896725,
      "epoch": 1.0999039385206533,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004886891319151411,
      "loss": 4.911,
      "mean_token_accuracy": 0.22022621035575868,
      "num_tokens": 26249749.0,
      "step": 11450
    },
    {
      "entropy": 5.331310987472534,
      "epoch": 1.1003842459173871,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004886783962216738,
      "loss": 4.9035,
      "mean_token_accuracy": 0.227567683160305,
      "num_tokens": 26260639.0,
      "step": 11455
    },
    {
      "entropy": 5.284071731567383,
      "epoch": 1.100864553314121,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004886676555672287,
      "loss": 4.9204,
      "mean_token_accuracy": 0.22389831095933915,
      "num_tokens": 26271797.0,
      "step": 11460
    },
    {
      "entropy": 5.312815093994141,
      "epoch": 1.101344860710855,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004886569099520551,
      "loss": 4.91,
      "mean_token_accuracy": 0.21784194856882094,
      "num_tokens": 26283298.0,
      "step": 11465
    },
    {
      "entropy": 5.2375401020050045,
      "epoch": 1.1018251681075888,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004886461593764024,
      "loss": 4.7421,
      "mean_token_accuracy": 0.23445857614278792,
      "num_tokens": 26293632.0,
      "step": 11470
    },
    {
      "entropy": 5.278605937957764,
      "epoch": 1.1023054755043227,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004886354038405204,
      "loss": 4.9877,
      "mean_token_accuracy": 0.21642430275678634,
      "num_tokens": 26307053.0,
      "step": 11475
    },
    {
      "entropy": 5.259592199325562,
      "epoch": 1.1027857829010568,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004886246433446586,
      "loss": 4.8312,
      "mean_token_accuracy": 0.23544746190309523,
      "num_tokens": 26317136.0,
      "step": 11480
    },
    {
      "entropy": 5.271986627578736,
      "epoch": 1.1032660902977907,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004886138778890669,
      "loss": 4.9259,
      "mean_token_accuracy": 0.22516705244779586,
      "num_tokens": 26328865.0,
      "step": 11485
    },
    {
      "entropy": 5.282785034179687,
      "epoch": 1.1037463976945245,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004886031074739953,
      "loss": 4.9385,
      "mean_token_accuracy": 0.22470292448997498,
      "num_tokens": 26341010.0,
      "step": 11490
    },
    {
      "entropy": 5.3507547855377195,
      "epoch": 1.1042267050912584,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004885923320996938,
      "loss": 4.9829,
      "mean_token_accuracy": 0.21977316737174987,
      "num_tokens": 26352203.0,
      "step": 11495
    },
    {
      "entropy": 5.337172031402588,
      "epoch": 1.1047070124879923,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004885815517664127,
      "loss": 4.9422,
      "mean_token_accuracy": 0.221173956990242,
      "num_tokens": 26363096.0,
      "step": 11500
    },
    {
      "entropy": 5.20210018157959,
      "epoch": 1.1051873198847262,
      "grad_norm": 1.046875,
      "learning_rate": 0.000488570766474402,
      "loss": 4.8479,
      "mean_token_accuracy": 0.2333227202296257,
      "num_tokens": 26374776.0,
      "step": 11505
    },
    {
      "entropy": 5.2437022686004635,
      "epoch": 1.10566762728146,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004885599762239124,
      "loss": 4.9484,
      "mean_token_accuracy": 0.2269499570131302,
      "num_tokens": 26386654.0,
      "step": 11510
    },
    {
      "entropy": 5.245615434646607,
      "epoch": 1.106147934678194,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004885491810151943,
      "loss": 4.9159,
      "mean_token_accuracy": 0.2227490559220314,
      "num_tokens": 26400022.0,
      "step": 11515
    },
    {
      "entropy": 5.231359958648682,
      "epoch": 1.106628242074928,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004885383808484982,
      "loss": 4.9026,
      "mean_token_accuracy": 0.2260092243552208,
      "num_tokens": 26412308.0,
      "step": 11520
    },
    {
      "entropy": 5.33850359916687,
      "epoch": 1.107108549471662,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004885275757240751,
      "loss": 4.9826,
      "mean_token_accuracy": 0.21035897433757783,
      "num_tokens": 26424506.0,
      "step": 11525
    },
    {
      "entropy": 5.320164537429809,
      "epoch": 1.1075888568683958,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004885167656421757,
      "loss": 4.8823,
      "mean_token_accuracy": 0.2294747546315193,
      "num_tokens": 26435020.0,
      "step": 11530
    },
    {
      "entropy": 5.159334993362426,
      "epoch": 1.1080691642651297,
      "grad_norm": 0.859375,
      "learning_rate": 0.000488505950603051,
      "loss": 4.798,
      "mean_token_accuracy": 0.23228764533996582,
      "num_tokens": 26447952.0,
      "step": 11535
    },
    {
      "entropy": 5.261581945419311,
      "epoch": 1.1085494716618636,
      "grad_norm": 1.0,
      "learning_rate": 0.000488495130606952,
      "loss": 4.8955,
      "mean_token_accuracy": 0.22537608742713927,
      "num_tokens": 26460001.0,
      "step": 11540
    },
    {
      "entropy": 5.254501295089722,
      "epoch": 1.1090297790585975,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004884843056541302,
      "loss": 4.8668,
      "mean_token_accuracy": 0.22915585786104203,
      "num_tokens": 26472139.0,
      "step": 11545
    },
    {
      "entropy": 5.249441432952881,
      "epoch": 1.1095100864553313,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004884734757448367,
      "loss": 4.9092,
      "mean_token_accuracy": 0.22575765997171401,
      "num_tokens": 26483855.0,
      "step": 11550
    },
    {
      "entropy": 5.281817245483398,
      "epoch": 1.1099903938520654,
      "grad_norm": 0.84375,
      "learning_rate": 0.000488462640879323,
      "loss": 4.9918,
      "mean_token_accuracy": 0.2159971699118614,
      "num_tokens": 26496529.0,
      "step": 11555
    },
    {
      "entropy": 5.2981163501739506,
      "epoch": 1.1104707012487993,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004884518010578405,
      "loss": 4.8651,
      "mean_token_accuracy": 0.23040514141321183,
      "num_tokens": 26506663.0,
      "step": 11560
    },
    {
      "entropy": 5.299975156784058,
      "epoch": 1.1109510086455332,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004884409562806411,
      "loss": 4.9925,
      "mean_token_accuracy": 0.21918700784444808,
      "num_tokens": 26519091.0,
      "step": 11565
    },
    {
      "entropy": 5.2664471626281735,
      "epoch": 1.111431316042267,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004884301065479765,
      "loss": 4.9137,
      "mean_token_accuracy": 0.22284814715385437,
      "num_tokens": 26530145.0,
      "step": 11570
    },
    {
      "entropy": 5.296102142333984,
      "epoch": 1.111911623439001,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004884192518600986,
      "loss": 4.8589,
      "mean_token_accuracy": 0.2247842952609062,
      "num_tokens": 26541228.0,
      "step": 11575
    },
    {
      "entropy": 5.304616689682007,
      "epoch": 1.1123919308357348,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004884083922172593,
      "loss": 4.897,
      "mean_token_accuracy": 0.2260119065642357,
      "num_tokens": 26551359.0,
      "step": 11580
    },
    {
      "entropy": 5.190745878219604,
      "epoch": 1.1128722382324687,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004883975276197108,
      "loss": 4.9123,
      "mean_token_accuracy": 0.22713827937841416,
      "num_tokens": 26562911.0,
      "step": 11585
    },
    {
      "entropy": 5.355407571792602,
      "epoch": 1.1133525456292026,
      "grad_norm": 1.0,
      "learning_rate": 0.0004883866580677055,
      "loss": 4.8915,
      "mean_token_accuracy": 0.2313615635037422,
      "num_tokens": 26573198.0,
      "step": 11590
    },
    {
      "entropy": 5.287772130966187,
      "epoch": 1.1138328530259365,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004883757835614956,
      "loss": 4.9053,
      "mean_token_accuracy": 0.22553828805685044,
      "num_tokens": 26584932.0,
      "step": 11595
    },
    {
      "entropy": 5.373600339889526,
      "epoch": 1.1143131604226706,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004883649041013335,
      "loss": 5.0445,
      "mean_token_accuracy": 0.21245040595531464,
      "num_tokens": 26596042.0,
      "step": 11600
    },
    {
      "entropy": 5.314949369430542,
      "epoch": 1.1147934678194045,
      "grad_norm": 1.03125,
      "learning_rate": 0.000488354019687472,
      "loss": 4.9498,
      "mean_token_accuracy": 0.22052074819803238,
      "num_tokens": 26607359.0,
      "step": 11605
    },
    {
      "entropy": 5.339853763580322,
      "epoch": 1.1152737752161384,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004883431303201636,
      "loss": 4.8921,
      "mean_token_accuracy": 0.225559201836586,
      "num_tokens": 26617095.0,
      "step": 11610
    },
    {
      "entropy": 5.23981556892395,
      "epoch": 1.1157540826128722,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004883322359996613,
      "loss": 4.9566,
      "mean_token_accuracy": 0.21801867634057998,
      "num_tokens": 26628450.0,
      "step": 11615
    },
    {
      "entropy": 5.243373823165894,
      "epoch": 1.1162343900096061,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004883213367262179,
      "loss": 4.8514,
      "mean_token_accuracy": 0.23036189675331115,
      "num_tokens": 26639428.0,
      "step": 11620
    },
    {
      "entropy": 5.1971923351287845,
      "epoch": 1.11671469740634,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004883104325000866,
      "loss": 4.8572,
      "mean_token_accuracy": 0.23037500083446502,
      "num_tokens": 26650730.0,
      "step": 11625
    },
    {
      "entropy": 5.2549420356750485,
      "epoch": 1.1171950048030739,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004882995233215203,
      "loss": 4.8545,
      "mean_token_accuracy": 0.22746994346380234,
      "num_tokens": 26661745.0,
      "step": 11630
    },
    {
      "entropy": 5.208850860595703,
      "epoch": 1.117675312199808,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004882886091907726,
      "loss": 4.8037,
      "mean_token_accuracy": 0.23410269767045974,
      "num_tokens": 26672976.0,
      "step": 11635
    },
    {
      "entropy": 5.323585462570191,
      "epoch": 1.1181556195965419,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048827769010809666,
      "loss": 4.9536,
      "mean_token_accuracy": 0.22333840131759644,
      "num_tokens": 26683503.0,
      "step": 11640
    },
    {
      "entropy": 5.259787130355835,
      "epoch": 1.1186359269932757,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048826676607374606,
      "loss": 4.8949,
      "mean_token_accuracy": 0.22300700098276138,
      "num_tokens": 26695160.0,
      "step": 11645
    },
    {
      "entropy": 5.200079345703125,
      "epoch": 1.1191162343900096,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048825583708797434,
      "loss": 4.8995,
      "mean_token_accuracy": 0.23131009191274643,
      "num_tokens": 26706514.0,
      "step": 11650
    },
    {
      "entropy": 5.1858802318573,
      "epoch": 1.1195965417867435,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004882449031510354,
      "loss": 4.7958,
      "mean_token_accuracy": 0.23193188458681108,
      "num_tokens": 26717268.0,
      "step": 11655
    },
    {
      "entropy": 5.3842504024505615,
      "epoch": 1.1200768491834774,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000488233964263183,
      "loss": 5.0491,
      "mean_token_accuracy": 0.21794600486755372,
      "num_tokens": 26728821.0,
      "step": 11660
    },
    {
      "entropy": 5.346639776229859,
      "epoch": 1.1205571565802113,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048822302042467115,
      "loss": 4.9793,
      "mean_token_accuracy": 0.21585893183946608,
      "num_tokens": 26740476.0,
      "step": 11665
    },
    {
      "entropy": 5.28800892829895,
      "epoch": 1.1210374639769451,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004882120716357539,
      "loss": 4.9864,
      "mean_token_accuracy": 0.21786582767963408,
      "num_tokens": 26751289.0,
      "step": 11670
    },
    {
      "entropy": 5.266548347473145,
      "epoch": 1.1215177713736793,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004882011178966854,
      "loss": 4.8995,
      "mean_token_accuracy": 0.22814032882452012,
      "num_tokens": 26762115.0,
      "step": 11675
    },
    {
      "entropy": 5.238596248626709,
      "epoch": 1.1219980787704131,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004881901592077201,
      "loss": 4.8303,
      "mean_token_accuracy": 0.2285684823989868,
      "num_tokens": 26772837.0,
      "step": 11680
    },
    {
      "entropy": 5.333533906936646,
      "epoch": 1.122478386167147,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004881791955691123,
      "loss": 4.9283,
      "mean_token_accuracy": 0.22646736055612565,
      "num_tokens": 26783721.0,
      "step": 11685
    },
    {
      "entropy": 5.262446594238281,
      "epoch": 1.122958693563881,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048816822698111655,
      "loss": 4.8833,
      "mean_token_accuracy": 0.2303398221731186,
      "num_tokens": 26794320.0,
      "step": 11690
    },
    {
      "entropy": 5.317741775512696,
      "epoch": 1.1234390009606148,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048815725344398766,
      "loss": 5.013,
      "mean_token_accuracy": 0.21603443771600722,
      "num_tokens": 26807053.0,
      "step": 11695
    },
    {
      "entropy": 5.395951795578003,
      "epoch": 1.1239193083573487,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048814627495798017,
      "loss": 4.9285,
      "mean_token_accuracy": 0.21572482138872145,
      "num_tokens": 26817245.0,
      "step": 11700
    },
    {
      "entropy": 5.262052917480469,
      "epoch": 1.1243996157540825,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004881352915233492,
      "loss": 4.9683,
      "mean_token_accuracy": 0.22109626829624177,
      "num_tokens": 26829498.0,
      "step": 11705
    },
    {
      "entropy": 5.189062404632568,
      "epoch": 1.1248799231508164,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048812430314034956,
      "loss": 4.8518,
      "mean_token_accuracy": 0.22912632375955583,
      "num_tokens": 26840534.0,
      "step": 11710
    },
    {
      "entropy": 5.297791576385498,
      "epoch": 1.1253602305475505,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004881133098092365,
      "loss": 4.9034,
      "mean_token_accuracy": 0.21584645211696624,
      "num_tokens": 26851305.0,
      "step": 11715
    },
    {
      "entropy": 5.250143003463745,
      "epoch": 1.1258405379442844,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004881023115302652,
      "loss": 4.8996,
      "mean_token_accuracy": 0.22503663450479508,
      "num_tokens": 26862620.0,
      "step": 11720
    },
    {
      "entropy": 5.283990335464478,
      "epoch": 1.1263208453410183,
      "grad_norm": 0.953125,
      "learning_rate": 0.000488091308303691,
      "loss": 4.933,
      "mean_token_accuracy": 0.2213476002216339,
      "num_tokens": 26874108.0,
      "step": 11725
    },
    {
      "entropy": 5.29652624130249,
      "epoch": 1.1268011527377522,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004880803001297694,
      "loss": 4.9086,
      "mean_token_accuracy": 0.21890902072191237,
      "num_tokens": 26886523.0,
      "step": 11730
    },
    {
      "entropy": 5.248139810562134,
      "epoch": 1.127281460134486,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004880692870087559,
      "loss": 4.848,
      "mean_token_accuracy": 0.23364873081445695,
      "num_tokens": 26897316.0,
      "step": 11735
    },
    {
      "entropy": 5.2217775821685795,
      "epoch": 1.12776176753122,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048805826894090626,
      "loss": 4.8927,
      "mean_token_accuracy": 0.22073718011379242,
      "num_tokens": 26909807.0,
      "step": 11740
    },
    {
      "entropy": 5.331627559661865,
      "epoch": 1.1282420749279538,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048804724592647626,
      "loss": 4.9844,
      "mean_token_accuracy": 0.220663358271122,
      "num_tokens": 26922054.0,
      "step": 11745
    },
    {
      "entropy": 5.279937076568603,
      "epoch": 1.1287223823246877,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004880362179657218,
      "loss": 4.8503,
      "mean_token_accuracy": 0.22239796966314315,
      "num_tokens": 26933838.0,
      "step": 11750
    },
    {
      "entropy": 5.243291568756104,
      "epoch": 1.1292026897214218,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048802518505889904,
      "loss": 4.9253,
      "mean_token_accuracy": 0.2240079253911972,
      "num_tokens": 26945396.0,
      "step": 11755
    },
    {
      "entropy": 5.3203565120697025,
      "epoch": 1.1296829971181557,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048801414720626404,
      "loss": 5.0047,
      "mean_token_accuracy": 0.21910335570573808,
      "num_tokens": 26957809.0,
      "step": 11760
    },
    {
      "entropy": 5.327291440963745,
      "epoch": 1.1301633045148896,
      "grad_norm": 1.0,
      "learning_rate": 0.00048800310440807294,
      "loss": 4.861,
      "mean_token_accuracy": 0.2299187645316124,
      "num_tokens": 26969024.0,
      "step": 11765
    },
    {
      "entropy": 5.2379968643188475,
      "epoch": 1.1306436119116234,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004879920566645823,
      "loss": 4.9097,
      "mean_token_accuracy": 0.22367744743824006,
      "num_tokens": 26980689.0,
      "step": 11770
    },
    {
      "entropy": 5.3291302680969235,
      "epoch": 1.1311239193083573,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004879810039760486,
      "loss": 4.9892,
      "mean_token_accuracy": 0.2206874117255211,
      "num_tokens": 26992115.0,
      "step": 11775
    },
    {
      "entropy": 5.3287688255310055,
      "epoch": 1.1316042267050912,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004879699463427284,
      "loss": 4.9026,
      "mean_token_accuracy": 0.22223807126283646,
      "num_tokens": 27003042.0,
      "step": 11780
    },
    {
      "entropy": 5.273615312576294,
      "epoch": 1.132084534101825,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004879588837648785,
      "loss": 4.8593,
      "mean_token_accuracy": 0.2304479405283928,
      "num_tokens": 27013863.0,
      "step": 11785
    },
    {
      "entropy": 5.159205102920533,
      "epoch": 1.1325648414985592,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048794781624275554,
      "loss": 4.809,
      "mean_token_accuracy": 0.23135840594768525,
      "num_tokens": 27024423.0,
      "step": 11790
    },
    {
      "entropy": 5.212631797790527,
      "epoch": 1.133045148895293,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048793674377661664,
      "loss": 4.9123,
      "mean_token_accuracy": 0.22103788554668427,
      "num_tokens": 27035460.0,
      "step": 11795
    },
    {
      "entropy": 5.294098997116089,
      "epoch": 1.133525456292027,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048792566636671886,
      "loss": 4.9018,
      "mean_token_accuracy": 0.2241029679775238,
      "num_tokens": 27046731.0,
      "step": 11800
    },
    {
      "entropy": 5.298119926452637,
      "epoch": 1.1340057636887608,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004879145840133194,
      "loss": 4.9088,
      "mean_token_accuracy": 0.22704579830169677,
      "num_tokens": 27057758.0,
      "step": 11805
    },
    {
      "entropy": 5.358719682693481,
      "epoch": 1.1344860710854947,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004879034967166755,
      "loss": 4.969,
      "mean_token_accuracy": 0.22055717259645463,
      "num_tokens": 27068055.0,
      "step": 11810
    },
    {
      "entropy": 5.23629584312439,
      "epoch": 1.1349663784822286,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004878924044770446,
      "loss": 4.8486,
      "mean_token_accuracy": 0.2240639716386795,
      "num_tokens": 27078602.0,
      "step": 11815
    },
    {
      "entropy": 5.236640596389771,
      "epoch": 1.1354466858789625,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004878813072946843,
      "loss": 4.9232,
      "mean_token_accuracy": 0.22116781920194625,
      "num_tokens": 27089944.0,
      "step": 11820
    },
    {
      "entropy": 5.300722885131836,
      "epoch": 1.1359269932756964,
      "grad_norm": 1.0,
      "learning_rate": 0.00048787020516985203,
      "loss": 4.8854,
      "mean_token_accuracy": 0.2296626850962639,
      "num_tokens": 27100486.0,
      "step": 11825
    },
    {
      "entropy": 5.285901308059692,
      "epoch": 1.1364073006724302,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048785909810280576,
      "loss": 4.9287,
      "mean_token_accuracy": 0.2241004601120949,
      "num_tokens": 27111435.0,
      "step": 11830
    },
    {
      "entropy": 5.299823522567749,
      "epoch": 1.1368876080691643,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004878479860938033,
      "loss": 4.9376,
      "mean_token_accuracy": 0.22263829857110978,
      "num_tokens": 27123482.0,
      "step": 11835
    },
    {
      "entropy": 5.276788187026978,
      "epoch": 1.1373679154658982,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048783686914310266,
      "loss": 4.867,
      "mean_token_accuracy": 0.22502729296684265,
      "num_tokens": 27135009.0,
      "step": 11840
    },
    {
      "entropy": 5.246808338165283,
      "epoch": 1.137848222862632,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004878257472509619,
      "loss": 4.8454,
      "mean_token_accuracy": 0.23011981695890427,
      "num_tokens": 27145230.0,
      "step": 11845
    },
    {
      "entropy": 5.241552686691284,
      "epoch": 1.138328530259366,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004878146204176392,
      "loss": 4.8779,
      "mean_token_accuracy": 0.22578407526016236,
      "num_tokens": 27156490.0,
      "step": 11850
    },
    {
      "entropy": 5.296277284622192,
      "epoch": 1.1388088376560999,
      "grad_norm": 1.046875,
      "learning_rate": 0.000487803488643393,
      "loss": 4.8614,
      "mean_token_accuracy": 0.23098314851522445,
      "num_tokens": 27166727.0,
      "step": 11855
    },
    {
      "entropy": 5.2258258819580075,
      "epoch": 1.1392891450528337,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048779235192848166,
      "loss": 4.8342,
      "mean_token_accuracy": 0.22840515673160552,
      "num_tokens": 27177969.0,
      "step": 11860
    },
    {
      "entropy": 5.288859939575195,
      "epoch": 1.1397694524495678,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004877812102731638,
      "loss": 4.9769,
      "mean_token_accuracy": 0.22422119081020356,
      "num_tokens": 27189637.0,
      "step": 11865
    },
    {
      "entropy": 5.407187366485596,
      "epoch": 1.1402497598463017,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048777006367769804,
      "loss": 5.093,
      "mean_token_accuracy": 0.2188020259141922,
      "num_tokens": 27201124.0,
      "step": 11870
    },
    {
      "entropy": 5.323542213439941,
      "epoch": 1.1407300672430356,
      "grad_norm": 1.0,
      "learning_rate": 0.0004877589121423432,
      "loss": 4.9595,
      "mean_token_accuracy": 0.21857482194900513,
      "num_tokens": 27212399.0,
      "step": 11875
    },
    {
      "entropy": 5.238043165206909,
      "epoch": 1.1412103746397695,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004877477556673582,
      "loss": 4.8847,
      "mean_token_accuracy": 0.2281758114695549,
      "num_tokens": 27224469.0,
      "step": 11880
    },
    {
      "entropy": 5.284638595581055,
      "epoch": 1.1416906820365034,
      "grad_norm": 0.9375,
      "learning_rate": 0.000487736594253002,
      "loss": 4.8581,
      "mean_token_accuracy": 0.2314703106880188,
      "num_tokens": 27236026.0,
      "step": 11885
    },
    {
      "entropy": 5.252954483032227,
      "epoch": 1.1421709894332372,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048772542789953384,
      "loss": 4.8667,
      "mean_token_accuracy": 0.22041328996419907,
      "num_tokens": 27248332.0,
      "step": 11890
    },
    {
      "entropy": 5.244831132888794,
      "epoch": 1.1426512968299711,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048771425660721284,
      "loss": 4.8485,
      "mean_token_accuracy": 0.2237658306956291,
      "num_tokens": 27258701.0,
      "step": 11895
    },
    {
      "entropy": 5.284875106811524,
      "epoch": 1.143131604226705,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048770308037629853,
      "loss": 4.9841,
      "mean_token_accuracy": 0.22013316452503204,
      "num_tokens": 27269808.0,
      "step": 11900
    },
    {
      "entropy": 5.299513053894043,
      "epoch": 1.143611911623439,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004876918992070502,
      "loss": 5.009,
      "mean_token_accuracy": 0.22142900973558427,
      "num_tokens": 27282440.0,
      "step": 11905
    },
    {
      "entropy": 5.38394718170166,
      "epoch": 1.144092219020173,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004876807130997276,
      "loss": 4.9992,
      "mean_token_accuracy": 0.22663030177354812,
      "num_tokens": 27294430.0,
      "step": 11910
    },
    {
      "entropy": 5.349733448028564,
      "epoch": 1.1445725264169069,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004876695220545903,
      "loss": 4.9761,
      "mean_token_accuracy": 0.22637368589639664,
      "num_tokens": 27305896.0,
      "step": 11915
    },
    {
      "entropy": 5.375237369537354,
      "epoch": 1.1450528338136408,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048765832607189824,
      "loss": 4.9568,
      "mean_token_accuracy": 0.22072764486074448,
      "num_tokens": 27318195.0,
      "step": 11920
    },
    {
      "entropy": 5.2663938999176025,
      "epoch": 1.1455331412103746,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048764712515191136,
      "loss": 4.8727,
      "mean_token_accuracy": 0.220981402695179,
      "num_tokens": 27329055.0,
      "step": 11925
    },
    {
      "entropy": 5.248434591293335,
      "epoch": 1.1460134486071085,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048763591929488966,
      "loss": 4.8774,
      "mean_token_accuracy": 0.22833212465047836,
      "num_tokens": 27340296.0,
      "step": 11930
    },
    {
      "entropy": 5.279546737670898,
      "epoch": 1.1464937560038424,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004876247085010933,
      "loss": 4.9195,
      "mean_token_accuracy": 0.22433834671974182,
      "num_tokens": 27350995.0,
      "step": 11935
    },
    {
      "entropy": 5.294404077529907,
      "epoch": 1.1469740634005763,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048761349277078253,
      "loss": 4.9571,
      "mean_token_accuracy": 0.2165197491645813,
      "num_tokens": 27362463.0,
      "step": 11940
    },
    {
      "entropy": 5.270102071762085,
      "epoch": 1.1474543707973104,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048760227210421775,
      "loss": 4.8319,
      "mean_token_accuracy": 0.22566030025482178,
      "num_tokens": 27373579.0,
      "step": 11945
    },
    {
      "entropy": 5.2840043067932125,
      "epoch": 1.1479346781940443,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004875910465016596,
      "loss": 4.8311,
      "mean_token_accuracy": 0.23404351025819778,
      "num_tokens": 27383813.0,
      "step": 11950
    },
    {
      "entropy": 5.276977300643921,
      "epoch": 1.1484149855907781,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004875798159633686,
      "loss": 4.9453,
      "mean_token_accuracy": 0.21644957214593888,
      "num_tokens": 27395651.0,
      "step": 11955
    },
    {
      "entropy": 5.270737648010254,
      "epoch": 1.148895292987512,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004875685804896055,
      "loss": 4.9254,
      "mean_token_accuracy": 0.22063556611537932,
      "num_tokens": 27407827.0,
      "step": 11960
    },
    {
      "entropy": 5.282165622711181,
      "epoch": 1.149375600384246,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004875573400806312,
      "loss": 4.9652,
      "mean_token_accuracy": 0.22058391124010085,
      "num_tokens": 27418777.0,
      "step": 11965
    },
    {
      "entropy": 5.3356610298156735,
      "epoch": 1.1498559077809798,
      "grad_norm": 1.0,
      "learning_rate": 0.00048754609473670654,
      "loss": 4.9542,
      "mean_token_accuracy": 0.21569475829601287,
      "num_tokens": 27429626.0,
      "step": 11970
    },
    {
      "entropy": 5.3574834823608395,
      "epoch": 1.1503362151777137,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004875348444580927,
      "loss": 5.1518,
      "mean_token_accuracy": 0.21480620354413987,
      "num_tokens": 27442304.0,
      "step": 11975
    },
    {
      "entropy": 5.287232208251953,
      "epoch": 1.1508165225744476,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000487523589245051,
      "loss": 4.8915,
      "mean_token_accuracy": 0.22925875335931778,
      "num_tokens": 27454421.0,
      "step": 11980
    },
    {
      "entropy": 5.245364618301392,
      "epoch": 1.1512968299711814,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004875123290978425,
      "loss": 4.7923,
      "mean_token_accuracy": 0.23335683941841126,
      "num_tokens": 27465220.0,
      "step": 11985
    },
    {
      "entropy": 5.23538818359375,
      "epoch": 1.1517771373679155,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048750106401672876,
      "loss": 4.9102,
      "mean_token_accuracy": 0.2190852403640747,
      "num_tokens": 27476542.0,
      "step": 11990
    },
    {
      "entropy": 5.2412971496582035,
      "epoch": 1.1522574447646494,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048748979400197134,
      "loss": 4.7948,
      "mean_token_accuracy": 0.2340750738978386,
      "num_tokens": 27487721.0,
      "step": 11995
    },
    {
      "entropy": 5.306460046768189,
      "epoch": 1.1527377521613833,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048747851905383183,
      "loss": 4.896,
      "mean_token_accuracy": 0.2306316614151001,
      "num_tokens": 27498982.0,
      "step": 12000
    },
    {
      "epoch": 1.1527377521613833,
      "eval_entropy": 5.074991735117079,
      "eval_loss": 4.985158443450928,
      "eval_mean_token_accuracy": 0.23014877696972483,
      "eval_num_tokens": 27498982.0,
      "eval_runtime": 20.3173,
      "eval_samples_per_second": 1615.129,
      "eval_steps_per_second": 201.897,
      "step": 12000
    },
    {
      "entropy": 5.1887836933135985,
      "epoch": 1.1532180595581172,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004874672391725721,
      "loss": 4.794,
      "mean_token_accuracy": 0.22834218740463258,
      "num_tokens": 27509863.0,
      "step": 12005
    },
    {
      "entropy": 5.22808871269226,
      "epoch": 1.153698366954851,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004874559543584539,
      "loss": 4.9074,
      "mean_token_accuracy": 0.22429670989513398,
      "num_tokens": 27521061.0,
      "step": 12010
    },
    {
      "entropy": 5.264443731307983,
      "epoch": 1.154178674351585,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004874446646117394,
      "loss": 4.9217,
      "mean_token_accuracy": 0.22197792530059815,
      "num_tokens": 27532187.0,
      "step": 12015
    },
    {
      "entropy": 5.283362960815429,
      "epoch": 1.154658981748319,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004874333699326906,
      "loss": 4.8993,
      "mean_token_accuracy": 0.22761330753564835,
      "num_tokens": 27543371.0,
      "step": 12020
    },
    {
      "entropy": 5.2648295879364015,
      "epoch": 1.155139289145053,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004874220703215697,
      "loss": 4.8454,
      "mean_token_accuracy": 0.22452566176652908,
      "num_tokens": 27554423.0,
      "step": 12025
    },
    {
      "entropy": 5.250241422653199,
      "epoch": 1.1556195965417868,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004874107657786391,
      "loss": 4.9062,
      "mean_token_accuracy": 0.22544174641370773,
      "num_tokens": 27566397.0,
      "step": 12030
    },
    {
      "entropy": 5.360298824310303,
      "epoch": 1.1560999039385207,
      "grad_norm": 1.0,
      "learning_rate": 0.00048739945630416124,
      "loss": 4.9368,
      "mean_token_accuracy": 0.22632402926683426,
      "num_tokens": 27578668.0,
      "step": 12035
    },
    {
      "entropy": 5.24913215637207,
      "epoch": 1.1565802113352546,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004873881418983987,
      "loss": 4.9262,
      "mean_token_accuracy": 0.21739766150712966,
      "num_tokens": 27589345.0,
      "step": 12040
    },
    {
      "entropy": 5.199416971206665,
      "epoch": 1.1570605187319885,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004873768225616141,
      "loss": 4.8128,
      "mean_token_accuracy": 0.23532032817602158,
      "num_tokens": 27600482.0,
      "step": 12045
    },
    {
      "entropy": 5.331480598449707,
      "epoch": 1.1575408261287223,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048736549829407047,
      "loss": 4.9548,
      "mean_token_accuracy": 0.22439651489257811,
      "num_tokens": 27611543.0,
      "step": 12050
    },
    {
      "entropy": 5.316651296615601,
      "epoch": 1.1580211335254562,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004873541690960305,
      "loss": 4.8943,
      "mean_token_accuracy": 0.2311217501759529,
      "num_tokens": 27623503.0,
      "step": 12055
    },
    {
      "entropy": 5.210362482070923,
      "epoch": 1.15850144092219,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004873428349677573,
      "loss": 4.7949,
      "mean_token_accuracy": 0.23657965958118438,
      "num_tokens": 27633593.0,
      "step": 12060
    },
    {
      "entropy": 5.272209024429321,
      "epoch": 1.1589817483189242,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000487331495909514,
      "loss": 4.9651,
      "mean_token_accuracy": 0.223458693921566,
      "num_tokens": 27646339.0,
      "step": 12065
    },
    {
      "entropy": 5.150249004364014,
      "epoch": 1.159462055715658,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048732015192156383,
      "loss": 4.7061,
      "mean_token_accuracy": 0.24871332049369813,
      "num_tokens": 27657314.0,
      "step": 12070
    },
    {
      "entropy": 5.2587024688720705,
      "epoch": 1.159942363112392,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048730880300417015,
      "loss": 4.9228,
      "mean_token_accuracy": 0.22458624243736267,
      "num_tokens": 27668680.0,
      "step": 12075
    },
    {
      "entropy": 5.278244972229004,
      "epoch": 1.1604226705091258,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048729744915759657,
      "loss": 4.9006,
      "mean_token_accuracy": 0.2240198701620102,
      "num_tokens": 27679628.0,
      "step": 12080
    },
    {
      "entropy": 5.304752397537231,
      "epoch": 1.1609029779058597,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048728609038210655,
      "loss": 4.9541,
      "mean_token_accuracy": 0.22977682054042817,
      "num_tokens": 27691205.0,
      "step": 12085
    },
    {
      "entropy": 5.291910028457641,
      "epoch": 1.1613832853025936,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048727472667796395,
      "loss": 4.8958,
      "mean_token_accuracy": 0.22529777586460115,
      "num_tokens": 27702903.0,
      "step": 12090
    },
    {
      "entropy": 5.322388744354248,
      "epoch": 1.1618635926993275,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004872633580454325,
      "loss": 5.0194,
      "mean_token_accuracy": 0.2105068549513817,
      "num_tokens": 27716554.0,
      "step": 12095
    },
    {
      "entropy": 5.29278335571289,
      "epoch": 1.1623439000960616,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048725198448477616,
      "loss": 4.861,
      "mean_token_accuracy": 0.23332352489233016,
      "num_tokens": 27727931.0,
      "step": 12100
    },
    {
      "entropy": 5.284858226776123,
      "epoch": 1.1628242074927955,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048724060599625893,
      "loss": 4.9603,
      "mean_token_accuracy": 0.22537014186382293,
      "num_tokens": 27739483.0,
      "step": 12105
    },
    {
      "entropy": 5.317119455337524,
      "epoch": 1.1633045148895294,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048722922258014506,
      "loss": 4.9175,
      "mean_token_accuracy": 0.22129613906145096,
      "num_tokens": 27751157.0,
      "step": 12110
    },
    {
      "entropy": 5.375982475280762,
      "epoch": 1.1637848222862632,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004872178342366989,
      "loss": 5.001,
      "mean_token_accuracy": 0.2176753893494606,
      "num_tokens": 27762190.0,
      "step": 12115
    },
    {
      "entropy": 5.222200632095337,
      "epoch": 1.1642651296829971,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048720644096618475,
      "loss": 4.87,
      "mean_token_accuracy": 0.22698808163404466,
      "num_tokens": 27773684.0,
      "step": 12120
    },
    {
      "entropy": 5.3288298606872555,
      "epoch": 1.164745437079731,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004871950427688672,
      "loss": 4.9353,
      "mean_token_accuracy": 0.21303969472646714,
      "num_tokens": 27785388.0,
      "step": 12125
    },
    {
      "entropy": 5.222191619873047,
      "epoch": 1.1652257444764649,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048718363964501087,
      "loss": 4.8452,
      "mean_token_accuracy": 0.22868625819683075,
      "num_tokens": 27797603.0,
      "step": 12130
    },
    {
      "entropy": 5.338804292678833,
      "epoch": 1.1657060518731988,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004871722315948805,
      "loss": 4.987,
      "mean_token_accuracy": 0.22358833700418473,
      "num_tokens": 27809444.0,
      "step": 12135
    },
    {
      "entropy": 5.38521089553833,
      "epoch": 1.1661863592699326,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004871608186187408,
      "loss": 5.0112,
      "mean_token_accuracy": 0.21391109526157379,
      "num_tokens": 27820213.0,
      "step": 12140
    },
    {
      "entropy": 5.242646312713623,
      "epoch": 1.1666666666666667,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048714940071685703,
      "loss": 4.8644,
      "mean_token_accuracy": 0.22669464200735093,
      "num_tokens": 27831564.0,
      "step": 12145
    },
    {
      "entropy": 5.3327796459198,
      "epoch": 1.1671469740634006,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048713797788949405,
      "loss": 4.9512,
      "mean_token_accuracy": 0.22145850360393524,
      "num_tokens": 27843686.0,
      "step": 12150
    },
    {
      "entropy": 5.364661741256714,
      "epoch": 1.1676272814601345,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048712655013691714,
      "loss": 4.9596,
      "mean_token_accuracy": 0.21912906169891358,
      "num_tokens": 27855428.0,
      "step": 12155
    },
    {
      "entropy": 5.349461221694947,
      "epoch": 1.1681075888568684,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048711511745939165,
      "loss": 5.0769,
      "mean_token_accuracy": 0.21880918145179748,
      "num_tokens": 27866733.0,
      "step": 12160
    },
    {
      "entropy": 5.28728756904602,
      "epoch": 1.1685878962536023,
      "grad_norm": 1.015625,
      "learning_rate": 0.000487103679857183,
      "loss": 4.923,
      "mean_token_accuracy": 0.2184532880783081,
      "num_tokens": 27878639.0,
      "step": 12165
    },
    {
      "entropy": 5.2340404987335205,
      "epoch": 1.1690682036503361,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004870922373305567,
      "loss": 4.934,
      "mean_token_accuracy": 0.22326090782880784,
      "num_tokens": 27892139.0,
      "step": 12170
    },
    {
      "entropy": 5.379704761505127,
      "epoch": 1.1695485110470702,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048708078987977837,
      "loss": 4.9912,
      "mean_token_accuracy": 0.22242120206356047,
      "num_tokens": 27904395.0,
      "step": 12175
    },
    {
      "entropy": 5.3116871356964115,
      "epoch": 1.1700288184438041,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048706933750511394,
      "loss": 4.9459,
      "mean_token_accuracy": 0.221306449174881,
      "num_tokens": 27916444.0,
      "step": 12180
    },
    {
      "entropy": 5.251688623428345,
      "epoch": 1.170509125840538,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004870578802068292,
      "loss": 4.9299,
      "mean_token_accuracy": 0.23007951825857162,
      "num_tokens": 27926914.0,
      "step": 12185
    },
    {
      "entropy": 5.270897245407104,
      "epoch": 1.170989433237272,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048704641798519006,
      "loss": 4.834,
      "mean_token_accuracy": 0.22656827867031099,
      "num_tokens": 27938203.0,
      "step": 12190
    },
    {
      "entropy": 5.323120832443237,
      "epoch": 1.1714697406340058,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048703495084046286,
      "loss": 4.9339,
      "mean_token_accuracy": 0.22079742103815078,
      "num_tokens": 27950463.0,
      "step": 12195
    },
    {
      "entropy": 5.294434690475464,
      "epoch": 1.1719500480307397,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004870234787729137,
      "loss": 4.8497,
      "mean_token_accuracy": 0.22667129188776017,
      "num_tokens": 27962312.0,
      "step": 12200
    },
    {
      "entropy": 5.2646932125091555,
      "epoch": 1.1724303554274735,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004870120017828089,
      "loss": 4.8616,
      "mean_token_accuracy": 0.22962882220745087,
      "num_tokens": 27973723.0,
      "step": 12205
    },
    {
      "entropy": 5.235909080505371,
      "epoch": 1.1729106628242074,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000487000519870415,
      "loss": 4.8327,
      "mean_token_accuracy": 0.23671973645687103,
      "num_tokens": 27983816.0,
      "step": 12210
    },
    {
      "entropy": 5.176678848266602,
      "epoch": 1.1733909702209413,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004869890330359986,
      "loss": 4.8834,
      "mean_token_accuracy": 0.23029006272554398,
      "num_tokens": 27996516.0,
      "step": 12215
    },
    {
      "entropy": 5.2701457977294925,
      "epoch": 1.1738712776176754,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004869775412798262,
      "loss": 4.9132,
      "mean_token_accuracy": 0.22639169096946715,
      "num_tokens": 28008256.0,
      "step": 12220
    },
    {
      "entropy": 5.233039903640747,
      "epoch": 1.1743515850144093,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048696604460216476,
      "loss": 4.7734,
      "mean_token_accuracy": 0.24480203539133072,
      "num_tokens": 28018882.0,
      "step": 12225
    },
    {
      "entropy": 5.248614978790283,
      "epoch": 1.1748318924111432,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048695454300328123,
      "loss": 4.8777,
      "mean_token_accuracy": 0.226214799284935,
      "num_tokens": 28030399.0,
      "step": 12230
    },
    {
      "entropy": 5.265173816680909,
      "epoch": 1.175312199807877,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048694303648344256,
      "loss": 4.9392,
      "mean_token_accuracy": 0.22445117235183715,
      "num_tokens": 28041621.0,
      "step": 12235
    },
    {
      "entropy": 5.267605304718018,
      "epoch": 1.175792507204611,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048693152504291595,
      "loss": 4.8279,
      "mean_token_accuracy": 0.22724812477827072,
      "num_tokens": 28053198.0,
      "step": 12240
    },
    {
      "entropy": 5.262113475799561,
      "epoch": 1.1762728146013448,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004869200086819686,
      "loss": 4.8646,
      "mean_token_accuracy": 0.22707436978816986,
      "num_tokens": 28064095.0,
      "step": 12245
    },
    {
      "entropy": 5.274620914459229,
      "epoch": 1.1767531219980787,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048690848740086796,
      "loss": 4.9356,
      "mean_token_accuracy": 0.21896139681339263,
      "num_tokens": 28075665.0,
      "step": 12250
    },
    {
      "entropy": 5.298528671264648,
      "epoch": 1.1772334293948128,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004868969611998814,
      "loss": 4.9653,
      "mean_token_accuracy": 0.22376179844141006,
      "num_tokens": 28089166.0,
      "step": 12255
    },
    {
      "entropy": 5.393528127670288,
      "epoch": 1.1777137367915467,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004868854300792767,
      "loss": 4.9779,
      "mean_token_accuracy": 0.21829388439655303,
      "num_tokens": 28101379.0,
      "step": 12260
    },
    {
      "entropy": 5.310004377365113,
      "epoch": 1.1781940441882806,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048687389403932144,
      "loss": 4.8987,
      "mean_token_accuracy": 0.2288082256913185,
      "num_tokens": 28112836.0,
      "step": 12265
    },
    {
      "entropy": 5.333872699737549,
      "epoch": 1.1786743515850144,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004868623530802835,
      "loss": 5.004,
      "mean_token_accuracy": 0.2161359339952469,
      "num_tokens": 28125039.0,
      "step": 12270
    },
    {
      "entropy": 5.265948867797851,
      "epoch": 1.1791546589817483,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00048685080720243086,
      "loss": 4.8682,
      "mean_token_accuracy": 0.22900580167770385,
      "num_tokens": 28137504.0,
      "step": 12275
    },
    {
      "entropy": 5.260558795928955,
      "epoch": 1.1796349663784822,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004868392564060315,
      "loss": 4.9071,
      "mean_token_accuracy": 0.23066666573286057,
      "num_tokens": 28149483.0,
      "step": 12280
    },
    {
      "entropy": 5.284013080596924,
      "epoch": 1.180115273775216,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004868277006913537,
      "loss": 4.8884,
      "mean_token_accuracy": 0.22417343109846116,
      "num_tokens": 28161001.0,
      "step": 12285
    },
    {
      "entropy": 5.339535236358643,
      "epoch": 1.18059558117195,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004868161400586656,
      "loss": 4.9445,
      "mean_token_accuracy": 0.22410396188497544,
      "num_tokens": 28171146.0,
      "step": 12290
    },
    {
      "entropy": 5.344981145858765,
      "epoch": 1.1810758885686838,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004868045745082357,
      "loss": 4.9417,
      "mean_token_accuracy": 0.22021251171827316,
      "num_tokens": 28182870.0,
      "step": 12295
    },
    {
      "entropy": 5.234750699996948,
      "epoch": 1.181556195965418,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004867930040403326,
      "loss": 4.9519,
      "mean_token_accuracy": 0.2250476062297821,
      "num_tokens": 28194457.0,
      "step": 12300
    },
    {
      "entropy": 5.328473234176636,
      "epoch": 1.1820365033621518,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048678142865522475,
      "loss": 4.977,
      "mean_token_accuracy": 0.21873759329319,
      "num_tokens": 28205818.0,
      "step": 12305
    },
    {
      "entropy": 5.293333673477173,
      "epoch": 1.1825168107588857,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000486769848353181,
      "loss": 4.8237,
      "mean_token_accuracy": 0.22565012127161027,
      "num_tokens": 28216124.0,
      "step": 12310
    },
    {
      "entropy": 5.2811681747436525,
      "epoch": 1.1829971181556196,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048675826313447027,
      "loss": 4.9082,
      "mean_token_accuracy": 0.22413806319236756,
      "num_tokens": 28227485.0,
      "step": 12315
    },
    {
      "entropy": 5.3200019836425785,
      "epoch": 1.1834774255523535,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048674667299936135,
      "loss": 4.9564,
      "mean_token_accuracy": 0.22329882234334947,
      "num_tokens": 28239007.0,
      "step": 12320
    },
    {
      "entropy": 5.310481452941895,
      "epoch": 1.1839577329490873,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048673507794812356,
      "loss": 4.9191,
      "mean_token_accuracy": 0.22302255928516387,
      "num_tokens": 28250926.0,
      "step": 12325
    },
    {
      "entropy": 5.323655462265014,
      "epoch": 1.1844380403458212,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004867234779810259,
      "loss": 4.9338,
      "mean_token_accuracy": 0.22216372340917587,
      "num_tokens": 28263182.0,
      "step": 12330
    },
    {
      "entropy": 5.218994903564453,
      "epoch": 1.1849183477425553,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004867118730983378,
      "loss": 4.8719,
      "mean_token_accuracy": 0.23049787878990174,
      "num_tokens": 28274666.0,
      "step": 12335
    },
    {
      "entropy": 5.198007202148437,
      "epoch": 1.1853986551392892,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004867002633003286,
      "loss": 4.7769,
      "mean_token_accuracy": 0.23961021155118942,
      "num_tokens": 28286007.0,
      "step": 12340
    },
    {
      "entropy": 5.235986423492432,
      "epoch": 1.185878962536023,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000486688648587268,
      "loss": 4.8725,
      "mean_token_accuracy": 0.23155883103609085,
      "num_tokens": 28298262.0,
      "step": 12345
    },
    {
      "entropy": 5.193673896789551,
      "epoch": 1.186359269932757,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004866770289594256,
      "loss": 4.8301,
      "mean_token_accuracy": 0.23047973960638046,
      "num_tokens": 28309428.0,
      "step": 12350
    },
    {
      "entropy": 5.322192049026489,
      "epoch": 1.1868395773294909,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048666540441707107,
      "loss": 4.9722,
      "mean_token_accuracy": 0.2173434764146805,
      "num_tokens": 28319609.0,
      "step": 12355
    },
    {
      "entropy": 5.312551546096802,
      "epoch": 1.1873198847262247,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004866537749604744,
      "loss": 4.9458,
      "mean_token_accuracy": 0.21562460064888,
      "num_tokens": 28330526.0,
      "step": 12360
    },
    {
      "entropy": 5.2482860565185545,
      "epoch": 1.1878001921229586,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048664214058990546,
      "loss": 4.9187,
      "mean_token_accuracy": 0.22521594613790513,
      "num_tokens": 28343919.0,
      "step": 12365
    },
    {
      "entropy": 5.334722566604614,
      "epoch": 1.1882804995196925,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004866305013056346,
      "loss": 4.9276,
      "mean_token_accuracy": 0.2259458601474762,
      "num_tokens": 28354850.0,
      "step": 12370
    },
    {
      "entropy": 5.215239858627319,
      "epoch": 1.1887608069164266,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004866188571079318,
      "loss": 4.8302,
      "mean_token_accuracy": 0.23198108673095702,
      "num_tokens": 28365699.0,
      "step": 12375
    },
    {
      "entropy": 5.172500991821289,
      "epoch": 1.1892411143131605,
      "grad_norm": 0.875,
      "learning_rate": 0.0004866072079970676,
      "loss": 4.791,
      "mean_token_accuracy": 0.2376916825771332,
      "num_tokens": 28377053.0,
      "step": 12380
    },
    {
      "entropy": 5.3247617244720455,
      "epoch": 1.1897214217098944,
      "grad_norm": 1.0,
      "learning_rate": 0.00048659555397331236,
      "loss": 4.8871,
      "mean_token_accuracy": 0.22118186503648757,
      "num_tokens": 28387505.0,
      "step": 12385
    },
    {
      "entropy": 5.242983675003051,
      "epoch": 1.1902017291066282,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004865838950369366,
      "loss": 4.8875,
      "mean_token_accuracy": 0.2250484988093376,
      "num_tokens": 28399308.0,
      "step": 12390
    },
    {
      "entropy": 5.2290863513946535,
      "epoch": 1.1906820365033621,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048657223118821116,
      "loss": 4.8871,
      "mean_token_accuracy": 0.22630987465381622,
      "num_tokens": 28410860.0,
      "step": 12395
    },
    {
      "entropy": 5.1448948860168455,
      "epoch": 1.191162343900096,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048656056242740665,
      "loss": 4.7348,
      "mean_token_accuracy": 0.23498868495225905,
      "num_tokens": 28421867.0,
      "step": 12400
    },
    {
      "entropy": 5.256177282333374,
      "epoch": 1.19164265129683,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004865488887547942,
      "loss": 4.8764,
      "mean_token_accuracy": 0.2273006707429886,
      "num_tokens": 28433375.0,
      "step": 12405
    },
    {
      "entropy": 5.3157429695129395,
      "epoch": 1.192122958693564,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004865372101706446,
      "loss": 4.9166,
      "mean_token_accuracy": 0.22432421743869782,
      "num_tokens": 28445242.0,
      "step": 12410
    },
    {
      "entropy": 5.205439805984497,
      "epoch": 1.1926032660902979,
      "grad_norm": 1.0,
      "learning_rate": 0.0004865255266752292,
      "loss": 4.8435,
      "mean_token_accuracy": 0.22836997658014296,
      "num_tokens": 28455904.0,
      "step": 12415
    },
    {
      "entropy": 5.295182371139527,
      "epoch": 1.1930835734870318,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004865138382688191,
      "loss": 4.8886,
      "mean_token_accuracy": 0.22266294360160827,
      "num_tokens": 28465768.0,
      "step": 12420
    },
    {
      "entropy": 5.335388422012329,
      "epoch": 1.1935638808837656,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004865021449516859,
      "loss": 4.9762,
      "mean_token_accuracy": 0.2328620046377182,
      "num_tokens": 28476489.0,
      "step": 12425
    },
    {
      "entropy": 5.283498048782349,
      "epoch": 1.1940441882804995,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004864904467241008,
      "loss": 4.841,
      "mean_token_accuracy": 0.23564036339521408,
      "num_tokens": 28487684.0,
      "step": 12430
    },
    {
      "entropy": 5.283005714416504,
      "epoch": 1.1945244956772334,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048647874358633556,
      "loss": 4.9041,
      "mean_token_accuracy": 0.22081979215145112,
      "num_tokens": 28499083.0,
      "step": 12435
    },
    {
      "entropy": 5.229641246795654,
      "epoch": 1.1950048030739673,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048646703553866183,
      "loss": 4.9016,
      "mean_token_accuracy": 0.2275218740105629,
      "num_tokens": 28510057.0,
      "step": 12440
    },
    {
      "entropy": 5.2103743076324465,
      "epoch": 1.1954851104707012,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004864553225813515,
      "loss": 4.8628,
      "mean_token_accuracy": 0.23284558355808258,
      "num_tokens": 28522787.0,
      "step": 12445
    },
    {
      "entropy": 5.359570074081421,
      "epoch": 1.195965417867435,
      "grad_norm": 1.0,
      "learning_rate": 0.0004864436047146765,
      "loss": 4.9164,
      "mean_token_accuracy": 0.22374020963907243,
      "num_tokens": 28534353.0,
      "step": 12450
    },
    {
      "entropy": 5.386217498779297,
      "epoch": 1.1964457252641691,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048643188193890874,
      "loss": 4.9845,
      "mean_token_accuracy": 0.21465651392936708,
      "num_tokens": 28546079.0,
      "step": 12455
    },
    {
      "entropy": 5.322726488113403,
      "epoch": 1.196926032660903,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004864201542543206,
      "loss": 4.9645,
      "mean_token_accuracy": 0.21505631804466246,
      "num_tokens": 28557637.0,
      "step": 12460
    },
    {
      "entropy": 5.262538003921509,
      "epoch": 1.197406340057637,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004864084216611843,
      "loss": 4.8583,
      "mean_token_accuracy": 0.23588258624076844,
      "num_tokens": 28568809.0,
      "step": 12465
    },
    {
      "entropy": 5.289714527130127,
      "epoch": 1.1978866474543708,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048639668415977207,
      "loss": 4.9646,
      "mean_token_accuracy": 0.22182999849319457,
      "num_tokens": 28579766.0,
      "step": 12470
    },
    {
      "entropy": 5.299949502944946,
      "epoch": 1.1983669548511047,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048638494175035665,
      "loss": 4.8533,
      "mean_token_accuracy": 0.22845515757799148,
      "num_tokens": 28590898.0,
      "step": 12475
    },
    {
      "entropy": 5.207098484039307,
      "epoch": 1.1988472622478386,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004863731944332105,
      "loss": 4.8209,
      "mean_token_accuracy": 0.22317590564489365,
      "num_tokens": 28603380.0,
      "step": 12480
    },
    {
      "entropy": 5.298857355117798,
      "epoch": 1.1993275696445724,
      "grad_norm": 1.0,
      "learning_rate": 0.0004863614422086065,
      "loss": 5.0071,
      "mean_token_accuracy": 0.22128204256296158,
      "num_tokens": 28615509.0,
      "step": 12485
    },
    {
      "entropy": 5.305480003356934,
      "epoch": 1.1998078770413065,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004863496850768174,
      "loss": 4.9193,
      "mean_token_accuracy": 0.22499871999025345,
      "num_tokens": 28627353.0,
      "step": 12490
    },
    {
      "entropy": 5.247653818130493,
      "epoch": 1.2002881844380404,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004863379230381162,
      "loss": 4.7719,
      "mean_token_accuracy": 0.2372029036283493,
      "num_tokens": 28638217.0,
      "step": 12495
    },
    {
      "entropy": 5.253597068786621,
      "epoch": 1.2007684918347743,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000486326156092776,
      "loss": 4.8561,
      "mean_token_accuracy": 0.22997648119926453,
      "num_tokens": 28648862.0,
      "step": 12500
    },
    {
      "entropy": 5.2381373882293705,
      "epoch": 1.2012487992315082,
      "grad_norm": 0.828125,
      "learning_rate": 0.00048631438424106985,
      "loss": 4.91,
      "mean_token_accuracy": 0.22861293107271194,
      "num_tokens": 28661561.0,
      "step": 12505
    },
    {
      "entropy": 5.324862098693847,
      "epoch": 1.201729106628242,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048630260748327124,
      "loss": 4.9472,
      "mean_token_accuracy": 0.2238690510392189,
      "num_tokens": 28674851.0,
      "step": 12510
    },
    {
      "entropy": 5.281457185745239,
      "epoch": 1.202209414024976,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048629082581965355,
      "loss": 4.9002,
      "mean_token_accuracy": 0.22551482766866685,
      "num_tokens": 28688444.0,
      "step": 12515
    },
    {
      "entropy": 5.351586723327637,
      "epoch": 1.2026897214217098,
      "grad_norm": 1.0,
      "learning_rate": 0.00048627903925049033,
      "loss": 4.9675,
      "mean_token_accuracy": 0.21779491007328033,
      "num_tokens": 28699374.0,
      "step": 12520
    },
    {
      "entropy": 5.25614366531372,
      "epoch": 1.2031700288184437,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048626724777605507,
      "loss": 4.9366,
      "mean_token_accuracy": 0.22380530685186387,
      "num_tokens": 28711283.0,
      "step": 12525
    },
    {
      "entropy": 5.30745120048523,
      "epoch": 1.2036503362151778,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004862554513966217,
      "loss": 4.9165,
      "mean_token_accuracy": 0.22111036628484726,
      "num_tokens": 28723445.0,
      "step": 12530
    },
    {
      "entropy": 5.2532140731811525,
      "epoch": 1.2041306436119117,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048624365011246405,
      "loss": 4.8844,
      "mean_token_accuracy": 0.23205445110797882,
      "num_tokens": 28734505.0,
      "step": 12535
    },
    {
      "entropy": 5.191025352478027,
      "epoch": 1.2046109510086456,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004862318439238561,
      "loss": 4.8564,
      "mean_token_accuracy": 0.23237161338329315,
      "num_tokens": 28745476.0,
      "step": 12540
    },
    {
      "entropy": 5.261295652389526,
      "epoch": 1.2050912584053795,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000486220032831072,
      "loss": 4.7914,
      "mean_token_accuracy": 0.2300058811903,
      "num_tokens": 28755645.0,
      "step": 12545
    },
    {
      "entropy": 5.2516826629638675,
      "epoch": 1.2055715658021133,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004862082168343859,
      "loss": 4.8993,
      "mean_token_accuracy": 0.22767243087291716,
      "num_tokens": 28767911.0,
      "step": 12550
    },
    {
      "entropy": 5.270639562606812,
      "epoch": 1.2060518731988472,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004861963959340722,
      "loss": 4.8935,
      "mean_token_accuracy": 0.22679862678050994,
      "num_tokens": 28779418.0,
      "step": 12555
    },
    {
      "entropy": 5.308687829971314,
      "epoch": 1.206532180595581,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004861845701304053,
      "loss": 4.9752,
      "mean_token_accuracy": 0.2188785895705223,
      "num_tokens": 28790857.0,
      "step": 12560
    },
    {
      "entropy": 5.351755952835083,
      "epoch": 1.2070124879923152,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048617273942365977,
      "loss": 4.9573,
      "mean_token_accuracy": 0.22226969450712203,
      "num_tokens": 28802975.0,
      "step": 12565
    },
    {
      "entropy": 5.2965466499328615,
      "epoch": 1.207492795389049,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004861609038141103,
      "loss": 4.8835,
      "mean_token_accuracy": 0.22810556441545488,
      "num_tokens": 28813874.0,
      "step": 12570
    },
    {
      "entropy": 5.238026571273804,
      "epoch": 1.207973102785783,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048614906330203165,
      "loss": 4.8559,
      "mean_token_accuracy": 0.22807337194681168,
      "num_tokens": 28826278.0,
      "step": 12575
    },
    {
      "entropy": 5.317562818527222,
      "epoch": 1.2084534101825168,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004861372178876987,
      "loss": 4.9416,
      "mean_token_accuracy": 0.22654914408922194,
      "num_tokens": 28838607.0,
      "step": 12580
    },
    {
      "entropy": 5.249324321746826,
      "epoch": 1.2089337175792507,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048612536757138653,
      "loss": 4.8821,
      "mean_token_accuracy": 0.22991496324539185,
      "num_tokens": 28849801.0,
      "step": 12585
    },
    {
      "entropy": 5.191409254074097,
      "epoch": 1.2094140249759846,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004861135123533702,
      "loss": 4.7788,
      "mean_token_accuracy": 0.2366494819521904,
      "num_tokens": 28861312.0,
      "step": 12590
    },
    {
      "entropy": 5.25629620552063,
      "epoch": 1.2098943323727185,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048610165223392503,
      "loss": 4.9192,
      "mean_token_accuracy": 0.22597109377384186,
      "num_tokens": 28872337.0,
      "step": 12595
    },
    {
      "entropy": 5.271780920028687,
      "epoch": 1.2103746397694524,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004860897872133263,
      "loss": 4.7715,
      "mean_token_accuracy": 0.24014756828546524,
      "num_tokens": 28883263.0,
      "step": 12600
    },
    {
      "entropy": 5.265909481048584,
      "epoch": 1.2108549471661862,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004860779172918496,
      "loss": 4.8705,
      "mean_token_accuracy": 0.22546121925115586,
      "num_tokens": 28896166.0,
      "step": 12605
    },
    {
      "entropy": 5.32851734161377,
      "epoch": 1.2113352545629203,
      "grad_norm": 1.125,
      "learning_rate": 0.0004860660424697704,
      "loss": 4.9394,
      "mean_token_accuracy": 0.22308899462223053,
      "num_tokens": 28906928.0,
      "step": 12610
    },
    {
      "entropy": 5.356096696853638,
      "epoch": 1.2118155619596542,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048605416274736434,
      "loss": 4.9882,
      "mean_token_accuracy": 0.2155408188700676,
      "num_tokens": 28918328.0,
      "step": 12615
    },
    {
      "entropy": 5.272354364395142,
      "epoch": 1.2122958693563881,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048604227812490744,
      "loss": 4.8775,
      "mean_token_accuracy": 0.2274181589484215,
      "num_tokens": 28930327.0,
      "step": 12620
    },
    {
      "entropy": 5.271601581573487,
      "epoch": 1.212776176753122,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048603038860267546,
      "loss": 4.8787,
      "mean_token_accuracy": 0.2276204749941826,
      "num_tokens": 28942166.0,
      "step": 12625
    },
    {
      "entropy": 5.241092872619629,
      "epoch": 1.2132564841498559,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004860184941809445,
      "loss": 4.7776,
      "mean_token_accuracy": 0.23694982826709748,
      "num_tokens": 28953159.0,
      "step": 12630
    },
    {
      "entropy": 5.233134078979492,
      "epoch": 1.2137367915465898,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048600659485999073,
      "loss": 4.8325,
      "mean_token_accuracy": 0.22985096871852875,
      "num_tokens": 28963897.0,
      "step": 12635
    },
    {
      "entropy": 5.205732536315918,
      "epoch": 1.2142170989433236,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048599469064009027,
      "loss": 4.8239,
      "mean_token_accuracy": 0.23175725489854812,
      "num_tokens": 28974452.0,
      "step": 12640
    },
    {
      "entropy": 5.185286140441894,
      "epoch": 1.2146974063400577,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048598278152151974,
      "loss": 4.7972,
      "mean_token_accuracy": 0.22976201325654982,
      "num_tokens": 28984664.0,
      "step": 12645
    },
    {
      "entropy": 5.306240367889404,
      "epoch": 1.2151777137367916,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004859708675045555,
      "loss": 4.8741,
      "mean_token_accuracy": 0.23290131837129593,
      "num_tokens": 28994992.0,
      "step": 12650
    },
    {
      "entropy": 5.352015209197998,
      "epoch": 1.2156580211335255,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004859589485894741,
      "loss": 4.8965,
      "mean_token_accuracy": 0.22114158272743226,
      "num_tokens": 29006190.0,
      "step": 12655
    },
    {
      "entropy": 5.25178713798523,
      "epoch": 1.2161383285302594,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004859470247765524,
      "loss": 4.8635,
      "mean_token_accuracy": 0.2257370322942734,
      "num_tokens": 29017987.0,
      "step": 12660
    },
    {
      "entropy": 5.203616809844971,
      "epoch": 1.2166186359269933,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004859350960660671,
      "loss": 4.8857,
      "mean_token_accuracy": 0.2310155600309372,
      "num_tokens": 29030012.0,
      "step": 12665
    },
    {
      "entropy": 5.340834665298462,
      "epoch": 1.2170989433237271,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004859231624582953,
      "loss": 4.9462,
      "mean_token_accuracy": 0.22174693793058395,
      "num_tokens": 29041684.0,
      "step": 12670
    },
    {
      "entropy": 5.288881587982178,
      "epoch": 1.217579250720461,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048591122395351394,
      "loss": 4.8277,
      "mean_token_accuracy": 0.22464753687381744,
      "num_tokens": 29052525.0,
      "step": 12675
    },
    {
      "entropy": 5.213456296920777,
      "epoch": 1.218059558117195,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004858992805520003,
      "loss": 4.8571,
      "mean_token_accuracy": 0.22601708322763442,
      "num_tokens": 29063471.0,
      "step": 12680
    },
    {
      "entropy": 5.187864685058594,
      "epoch": 1.218539865513929,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048588733225403153,
      "loss": 4.8013,
      "mean_token_accuracy": 0.23218270093202592,
      "num_tokens": 29074616.0,
      "step": 12685
    },
    {
      "entropy": 5.277328538894653,
      "epoch": 1.219020172910663,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004858753790598851,
      "loss": 4.9057,
      "mean_token_accuracy": 0.22922273725271225,
      "num_tokens": 29086603.0,
      "step": 12690
    },
    {
      "entropy": 5.277042865753174,
      "epoch": 1.2195004803073968,
      "grad_norm": 1.0,
      "learning_rate": 0.0004858634209698386,
      "loss": 4.8853,
      "mean_token_accuracy": 0.23373661041259766,
      "num_tokens": 29097232.0,
      "step": 12695
    },
    {
      "entropy": 5.387946224212646,
      "epoch": 1.2199807877041307,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048585145798416956,
      "loss": 4.9745,
      "mean_token_accuracy": 0.21698843836784362,
      "num_tokens": 29107453.0,
      "step": 12700
    },
    {
      "entropy": 5.210908603668213,
      "epoch": 1.2204610951008645,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004858394901031558,
      "loss": 4.8773,
      "mean_token_accuracy": 0.22787757813930512,
      "num_tokens": 29120255.0,
      "step": 12705
    },
    {
      "entropy": 5.293791484832764,
      "epoch": 1.2209414024975984,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004858275173270751,
      "loss": 4.9045,
      "mean_token_accuracy": 0.22155086547136307,
      "num_tokens": 29130892.0,
      "step": 12710
    },
    {
      "entropy": 5.329501819610596,
      "epoch": 1.2214217098943323,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048581553965620553,
      "loss": 4.9256,
      "mean_token_accuracy": 0.2243715286254883,
      "num_tokens": 29142613.0,
      "step": 12715
    },
    {
      "entropy": 5.280271339416504,
      "epoch": 1.2219020172910664,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048580355709082506,
      "loss": 4.8353,
      "mean_token_accuracy": 0.23253411501646043,
      "num_tokens": 29153045.0,
      "step": 12720
    },
    {
      "entropy": 5.1936657428741455,
      "epoch": 1.2223823246878003,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000485791569631212,
      "loss": 4.769,
      "mean_token_accuracy": 0.2366034746170044,
      "num_tokens": 29163767.0,
      "step": 12725
    },
    {
      "entropy": 5.269258594512939,
      "epoch": 1.2228626320845342,
      "grad_norm": 1.0,
      "learning_rate": 0.0004857795772776446,
      "loss": 4.9323,
      "mean_token_accuracy": 0.22075918912887574,
      "num_tokens": 29174277.0,
      "step": 12730
    },
    {
      "entropy": 5.344747400283813,
      "epoch": 1.223342939481268,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048576758003040127,
      "loss": 4.9586,
      "mean_token_accuracy": 0.2221354752779007,
      "num_tokens": 29186078.0,
      "step": 12735
    },
    {
      "entropy": 5.23181562423706,
      "epoch": 1.223823246878002,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048575557788976066,
      "loss": 4.8353,
      "mean_token_accuracy": 0.23407133668661118,
      "num_tokens": 29198056.0,
      "step": 12740
    },
    {
      "entropy": 5.34853892326355,
      "epoch": 1.2243035542747358,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004857435708560013,
      "loss": 5.0375,
      "mean_token_accuracy": 0.2107453837990761,
      "num_tokens": 29209892.0,
      "step": 12745
    },
    {
      "entropy": 5.35208215713501,
      "epoch": 1.2247838616714697,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048573155892940204,
      "loss": 4.9161,
      "mean_token_accuracy": 0.2290266752243042,
      "num_tokens": 29220813.0,
      "step": 12750
    },
    {
      "entropy": 5.310712146759033,
      "epoch": 1.2252641690682036,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048571954211024164,
      "loss": 4.9037,
      "mean_token_accuracy": 0.2247047618031502,
      "num_tokens": 29232807.0,
      "step": 12755
    },
    {
      "entropy": 5.319050645828247,
      "epoch": 1.2257444764649374,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048570752039879924,
      "loss": 4.89,
      "mean_token_accuracy": 0.22319486141204833,
      "num_tokens": 29243721.0,
      "step": 12760
    },
    {
      "entropy": 5.268907308578491,
      "epoch": 1.2262247838616716,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004856954937953539,
      "loss": 4.9138,
      "mean_token_accuracy": 0.22432439774274826,
      "num_tokens": 29255783.0,
      "step": 12765
    },
    {
      "entropy": 5.204079246520996,
      "epoch": 1.2267050912584054,
      "grad_norm": 1.0,
      "learning_rate": 0.0004856834623001848,
      "loss": 4.8171,
      "mean_token_accuracy": 0.23253057897090912,
      "num_tokens": 29266857.0,
      "step": 12770
    },
    {
      "entropy": 5.172171497344971,
      "epoch": 1.2271853986551393,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004856714259135713,
      "loss": 4.7635,
      "mean_token_accuracy": 0.23912411779165268,
      "num_tokens": 29278479.0,
      "step": 12775
    },
    {
      "entropy": 5.327312326431274,
      "epoch": 1.2276657060518732,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004856593846357929,
      "loss": 4.8687,
      "mean_token_accuracy": 0.23330814391374588,
      "num_tokens": 29289369.0,
      "step": 12780
    },
    {
      "entropy": 5.232531213760376,
      "epoch": 1.228146013448607,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004856473384671291,
      "loss": 4.8573,
      "mean_token_accuracy": 0.2212492898106575,
      "num_tokens": 29301015.0,
      "step": 12785
    },
    {
      "entropy": 5.318277215957641,
      "epoch": 1.228626320845341,
      "grad_norm": 1.0,
      "learning_rate": 0.00048563528740785955,
      "loss": 4.9434,
      "mean_token_accuracy": 0.22422471195459365,
      "num_tokens": 29312638.0,
      "step": 12790
    },
    {
      "entropy": 5.306363344192505,
      "epoch": 1.2291066282420748,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048562323145826414,
      "loss": 4.9034,
      "mean_token_accuracy": 0.22704463303089142,
      "num_tokens": 29323540.0,
      "step": 12795
    },
    {
      "entropy": 5.285631608963013,
      "epoch": 1.229586935638809,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004856111706186227,
      "loss": 4.9227,
      "mean_token_accuracy": 0.22471648901700975,
      "num_tokens": 29334287.0,
      "step": 12800
    },
    {
      "entropy": 5.310602521896362,
      "epoch": 1.2300672430355428,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048559910488921534,
      "loss": 4.8656,
      "mean_token_accuracy": 0.2221406877040863,
      "num_tokens": 29345278.0,
      "step": 12805
    },
    {
      "entropy": 5.345884895324707,
      "epoch": 1.2305475504322767,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000485587034270322,
      "loss": 4.9344,
      "mean_token_accuracy": 0.2225649818778038,
      "num_tokens": 29357297.0,
      "step": 12810
    },
    {
      "entropy": 5.269227838516235,
      "epoch": 1.2310278578290106,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000485574958762223,
      "loss": 4.8776,
      "mean_token_accuracy": 0.23230814784765244,
      "num_tokens": 29368513.0,
      "step": 12815
    },
    {
      "entropy": 5.274504566192627,
      "epoch": 1.2315081652257445,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048556287836519886,
      "loss": 4.9084,
      "mean_token_accuracy": 0.22138010263442992,
      "num_tokens": 29379227.0,
      "step": 12820
    },
    {
      "entropy": 5.203992557525635,
      "epoch": 1.2319884726224783,
      "grad_norm": 1.0,
      "learning_rate": 0.0004855507930795299,
      "loss": 4.8879,
      "mean_token_accuracy": 0.22839482575654985,
      "num_tokens": 29390048.0,
      "step": 12825
    },
    {
      "entropy": 5.3048095703125,
      "epoch": 1.2324687800192122,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048553870290549665,
      "loss": 4.9006,
      "mean_token_accuracy": 0.22254465967416764,
      "num_tokens": 29401252.0,
      "step": 12830
    },
    {
      "entropy": 5.276021146774292,
      "epoch": 1.232949087415946,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048552660784338,
      "loss": 4.8457,
      "mean_token_accuracy": 0.2306983768939972,
      "num_tokens": 29413122.0,
      "step": 12835
    },
    {
      "entropy": 5.396855640411377,
      "epoch": 1.23342939481268,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004855145078934606,
      "loss": 4.9688,
      "mean_token_accuracy": 0.22051817327737808,
      "num_tokens": 29423240.0,
      "step": 12840
    },
    {
      "entropy": 5.265974760055542,
      "epoch": 1.233909702209414,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004855024030560195,
      "loss": 4.8195,
      "mean_token_accuracy": 0.23162656426429748,
      "num_tokens": 29433818.0,
      "step": 12845
    },
    {
      "entropy": 5.291182947158814,
      "epoch": 1.234390009606148,
      "grad_norm": 1.0,
      "learning_rate": 0.0004854902933313376,
      "loss": 4.9507,
      "mean_token_accuracy": 0.2251957580447197,
      "num_tokens": 29446927.0,
      "step": 12850
    },
    {
      "entropy": 5.313303852081299,
      "epoch": 1.2348703170028819,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048547817871969607,
      "loss": 4.9157,
      "mean_token_accuracy": 0.22301144003868104,
      "num_tokens": 29458092.0,
      "step": 12855
    },
    {
      "entropy": 5.30733551979065,
      "epoch": 1.2353506243996157,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048546605922137633,
      "loss": 4.9824,
      "mean_token_accuracy": 0.21393632441759108,
      "num_tokens": 29470395.0,
      "step": 12860
    },
    {
      "entropy": 5.326679801940918,
      "epoch": 1.2358309317963496,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004854539348366596,
      "loss": 4.9479,
      "mean_token_accuracy": 0.22279625684022902,
      "num_tokens": 29481177.0,
      "step": 12865
    },
    {
      "entropy": 5.290243434906006,
      "epoch": 1.2363112391930835,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004854418055658274,
      "loss": 4.8398,
      "mean_token_accuracy": 0.22957303673028945,
      "num_tokens": 29492305.0,
      "step": 12870
    },
    {
      "entropy": 5.320012044906616,
      "epoch": 1.2367915465898176,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048542967140916134,
      "loss": 4.9253,
      "mean_token_accuracy": 0.2301064610481262,
      "num_tokens": 29503237.0,
      "step": 12875
    },
    {
      "entropy": 5.273182535171509,
      "epoch": 1.2372718539865515,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004854175323669432,
      "loss": 4.8751,
      "mean_token_accuracy": 0.2261396676301956,
      "num_tokens": 29514195.0,
      "step": 12880
    },
    {
      "entropy": 5.305807733535767,
      "epoch": 1.2377521613832854,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004854053884394547,
      "loss": 5.0007,
      "mean_token_accuracy": 0.21619590669870375,
      "num_tokens": 29526110.0,
      "step": 12885
    },
    {
      "entropy": 5.259426212310791,
      "epoch": 1.2382324687800192,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048539323962697796,
      "loss": 4.8806,
      "mean_token_accuracy": 0.23018712550401688,
      "num_tokens": 29536912.0,
      "step": 12890
    },
    {
      "entropy": 5.330661106109619,
      "epoch": 1.2387127761767531,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004853810859297949,
      "loss": 4.8908,
      "mean_token_accuracy": 0.2255253940820694,
      "num_tokens": 29548070.0,
      "step": 12895
    },
    {
      "entropy": 5.256996202468872,
      "epoch": 1.239193083573487,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048536892734818773,
      "loss": 4.9433,
      "mean_token_accuracy": 0.21995639353990554,
      "num_tokens": 29560713.0,
      "step": 12900
    },
    {
      "entropy": 5.255424547195434,
      "epoch": 1.239673390970221,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004853567638824387,
      "loss": 4.8572,
      "mean_token_accuracy": 0.24065370112657547,
      "num_tokens": 29571269.0,
      "step": 12905
    },
    {
      "entropy": 5.325501203536987,
      "epoch": 1.2401536983669548,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048534459553283026,
      "loss": 4.8894,
      "mean_token_accuracy": 0.23175992369651793,
      "num_tokens": 29584024.0,
      "step": 12910
    },
    {
      "entropy": 5.193266201019287,
      "epoch": 1.2406340057636887,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004853324222996449,
      "loss": 4.7542,
      "mean_token_accuracy": 0.23511122912168503,
      "num_tokens": 29595679.0,
      "step": 12915
    },
    {
      "entropy": 5.27776608467102,
      "epoch": 1.2411143131604228,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048532024418316525,
      "loss": 4.871,
      "mean_token_accuracy": 0.22823090851306915,
      "num_tokens": 29607316.0,
      "step": 12920
    },
    {
      "entropy": 5.264577627182007,
      "epoch": 1.2415946205571566,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004853080611836741,
      "loss": 4.8753,
      "mean_token_accuracy": 0.23004590570926667,
      "num_tokens": 29618929.0,
      "step": 12925
    },
    {
      "entropy": 5.334429836273193,
      "epoch": 1.2420749279538905,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048529587330145427,
      "loss": 4.9533,
      "mean_token_accuracy": 0.22567186653614044,
      "num_tokens": 29630253.0,
      "step": 12930
    },
    {
      "entropy": 5.255400848388672,
      "epoch": 1.2425552353506244,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048528368053678863,
      "loss": 4.8827,
      "mean_token_accuracy": 0.22904572188854216,
      "num_tokens": 29641676.0,
      "step": 12935
    },
    {
      "entropy": 5.22233304977417,
      "epoch": 1.2430355427473583,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004852714828899604,
      "loss": 4.9232,
      "mean_token_accuracy": 0.22467687577009202,
      "num_tokens": 29651942.0,
      "step": 12940
    },
    {
      "entropy": 5.353050804138183,
      "epoch": 1.2435158501440922,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048525928036125264,
      "loss": 4.9633,
      "mean_token_accuracy": 0.22232309430837632,
      "num_tokens": 29663094.0,
      "step": 12945
    },
    {
      "entropy": 5.298305892944336,
      "epoch": 1.243996157540826,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048524707295094884,
      "loss": 4.9134,
      "mean_token_accuracy": 0.22594064921140672,
      "num_tokens": 29675412.0,
      "step": 12950
    },
    {
      "entropy": 5.240254640579224,
      "epoch": 1.2444764649375601,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004852348606593322,
      "loss": 4.8911,
      "mean_token_accuracy": 0.22104781866073608,
      "num_tokens": 29686684.0,
      "step": 12955
    },
    {
      "entropy": 5.237393522262574,
      "epoch": 1.244956772334294,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048522264348668646,
      "loss": 4.8349,
      "mean_token_accuracy": 0.23330250084400178,
      "num_tokens": 29697387.0,
      "step": 12960
    },
    {
      "entropy": 5.293274974822998,
      "epoch": 1.245437079731028,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004852104214332951,
      "loss": 4.7777,
      "mean_token_accuracy": 0.2394947350025177,
      "num_tokens": 29708677.0,
      "step": 12965
    },
    {
      "entropy": 5.287729930877686,
      "epoch": 1.2459173871277618,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048519819449944205,
      "loss": 4.9572,
      "mean_token_accuracy": 0.22048394083976747,
      "num_tokens": 29720430.0,
      "step": 12970
    },
    {
      "entropy": 5.239066648483276,
      "epoch": 1.2463976945244957,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000485185962685411,
      "loss": 4.841,
      "mean_token_accuracy": 0.23739639967679976,
      "num_tokens": 29731370.0,
      "step": 12975
    },
    {
      "entropy": 5.375337076187134,
      "epoch": 1.2468780019212296,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000485173725991486,
      "loss": 4.934,
      "mean_token_accuracy": 0.21622303128242493,
      "num_tokens": 29743399.0,
      "step": 12980
    },
    {
      "entropy": 5.323163270950317,
      "epoch": 1.2473583093179634,
      "grad_norm": 1.0,
      "learning_rate": 0.00048516148441795124,
      "loss": 4.9003,
      "mean_token_accuracy": 0.2182629868388176,
      "num_tokens": 29755693.0,
      "step": 12985
    },
    {
      "entropy": 5.3004638671875,
      "epoch": 1.2478386167146973,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004851492379650908,
      "loss": 4.9359,
      "mean_token_accuracy": 0.2235865354537964,
      "num_tokens": 29766576.0,
      "step": 12990
    },
    {
      "entropy": 5.215474462509155,
      "epoch": 1.2483189241114312,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004851369866331891,
      "loss": 4.8478,
      "mean_token_accuracy": 0.2246627151966095,
      "num_tokens": 29779255.0,
      "step": 12995
    },
    {
      "entropy": 5.236148071289063,
      "epoch": 1.2487992315081653,
      "grad_norm": 1.0,
      "learning_rate": 0.0004851247304225306,
      "loss": 4.843,
      "mean_token_accuracy": 0.23158435821533202,
      "num_tokens": 29790172.0,
      "step": 13000
    },
    {
      "entropy": 5.318034744262695,
      "epoch": 1.2492795389048992,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004851124693333997,
      "loss": 4.9152,
      "mean_token_accuracy": 0.21916219741106033,
      "num_tokens": 29801206.0,
      "step": 13005
    },
    {
      "entropy": 5.381892251968384,
      "epoch": 1.249759846301633,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004851002033660812,
      "loss": 4.9257,
      "mean_token_accuracy": 0.22022076100111007,
      "num_tokens": 29813917.0,
      "step": 13010
    },
    {
      "entropy": 5.207949638366699,
      "epoch": 1.250240153698367,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048508793252085994,
      "loss": 4.7843,
      "mean_token_accuracy": 0.23079826831817626,
      "num_tokens": 29826036.0,
      "step": 13015
    },
    {
      "entropy": 5.176231336593628,
      "epoch": 1.2507204610951008,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004850756567980206,
      "loss": 4.7667,
      "mean_token_accuracy": 0.24254931956529618,
      "num_tokens": 29837901.0,
      "step": 13020
    },
    {
      "entropy": 5.165464544296265,
      "epoch": 1.2512007684918347,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048506337619784836,
      "loss": 4.7892,
      "mean_token_accuracy": 0.2351778656244278,
      "num_tokens": 29850074.0,
      "step": 13025
    },
    {
      "entropy": 5.2281450748443605,
      "epoch": 1.2516810758885688,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004850510907206283,
      "loss": 4.8604,
      "mean_token_accuracy": 0.2298705279827118,
      "num_tokens": 29862447.0,
      "step": 13030
    },
    {
      "entropy": 5.276225614547729,
      "epoch": 1.2521613832853027,
      "grad_norm": 1.0,
      "learning_rate": 0.00048503880036664555,
      "loss": 4.9019,
      "mean_token_accuracy": 0.22614027112722396,
      "num_tokens": 29873483.0,
      "step": 13035
    },
    {
      "entropy": 5.248000240325927,
      "epoch": 1.2526416906820366,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004850265051361857,
      "loss": 4.8249,
      "mean_token_accuracy": 0.2290497824549675,
      "num_tokens": 29884807.0,
      "step": 13040
    },
    {
      "entropy": 5.244644546508789,
      "epoch": 1.2531219980787704,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004850142050295339,
      "loss": 4.8627,
      "mean_token_accuracy": 0.22008608877658845,
      "num_tokens": 29895994.0,
      "step": 13045
    },
    {
      "entropy": 5.273664855957032,
      "epoch": 1.2536023054755043,
      "grad_norm": 1.125,
      "learning_rate": 0.00048500190004697595,
      "loss": 4.9063,
      "mean_token_accuracy": 0.22734262496232988,
      "num_tokens": 29907534.0,
      "step": 13050
    },
    {
      "entropy": 5.278700160980224,
      "epoch": 1.2540826128722382,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004849895901887974,
      "loss": 4.8856,
      "mean_token_accuracy": 0.22726300358772278,
      "num_tokens": 29919747.0,
      "step": 13055
    },
    {
      "entropy": 5.264999485015869,
      "epoch": 1.254562920268972,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004849772754552842,
      "loss": 4.9012,
      "mean_token_accuracy": 0.22810715734958648,
      "num_tokens": 29930254.0,
      "step": 13060
    },
    {
      "entropy": 5.36310305595398,
      "epoch": 1.255043227665706,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048496495584672214,
      "loss": 4.9944,
      "mean_token_accuracy": 0.22049553841352462,
      "num_tokens": 29941232.0,
      "step": 13065
    },
    {
      "entropy": 5.316765213012696,
      "epoch": 1.2555235350624399,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048495263136339725,
      "loss": 4.8402,
      "mean_token_accuracy": 0.22987185567617416,
      "num_tokens": 29952452.0,
      "step": 13070
    },
    {
      "entropy": 5.279210376739502,
      "epoch": 1.2560038424591737,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004849403020055956,
      "loss": 4.8875,
      "mean_token_accuracy": 0.22878491133451462,
      "num_tokens": 29963583.0,
      "step": 13075
    },
    {
      "entropy": 5.242006826400757,
      "epoch": 1.2564841498559078,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048492796777360373,
      "loss": 4.8641,
      "mean_token_accuracy": 0.2311529591679573,
      "num_tokens": 29975186.0,
      "step": 13080
    },
    {
      "entropy": 5.228663110733033,
      "epoch": 1.2569644572526417,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048491562866770767,
      "loss": 4.8401,
      "mean_token_accuracy": 0.23134734779596328,
      "num_tokens": 29987474.0,
      "step": 13085
    },
    {
      "entropy": 5.228747320175171,
      "epoch": 1.2574447646493756,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048490328468819404,
      "loss": 4.844,
      "mean_token_accuracy": 0.2346086844801903,
      "num_tokens": 29999254.0,
      "step": 13090
    },
    {
      "entropy": 5.278165197372436,
      "epoch": 1.2579250720461095,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048489093583534945,
      "loss": 4.9018,
      "mean_token_accuracy": 0.22648218274116516,
      "num_tokens": 30009901.0,
      "step": 13095
    },
    {
      "entropy": 5.3112658023834225,
      "epoch": 1.2584053794428434,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004848785821094606,
      "loss": 4.9789,
      "mean_token_accuracy": 0.2203599512577057,
      "num_tokens": 30022424.0,
      "step": 13100
    },
    {
      "entropy": 5.227879858016967,
      "epoch": 1.2588856868395775,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004848662235108142,
      "loss": 4.9191,
      "mean_token_accuracy": 0.22763861268758773,
      "num_tokens": 30034209.0,
      "step": 13105
    },
    {
      "entropy": 5.256797790527344,
      "epoch": 1.2593659942363113,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004848538600396973,
      "loss": 4.8234,
      "mean_token_accuracy": 0.2310587629675865,
      "num_tokens": 30044885.0,
      "step": 13110
    },
    {
      "entropy": 5.354308271408081,
      "epoch": 1.2598463016330452,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048484149169639694,
      "loss": 4.9017,
      "mean_token_accuracy": 0.2242549777030945,
      "num_tokens": 30056592.0,
      "step": 13115
    },
    {
      "entropy": 5.37705888748169,
      "epoch": 1.260326609029779,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004848291184812003,
      "loss": 5.0261,
      "mean_token_accuracy": 0.21949134171009063,
      "num_tokens": 30068423.0,
      "step": 13120
    },
    {
      "entropy": 5.303070306777954,
      "epoch": 1.260806916426513,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004848167403943945,
      "loss": 4.8947,
      "mean_token_accuracy": 0.22471791058778762,
      "num_tokens": 30079132.0,
      "step": 13125
    },
    {
      "entropy": 5.2165422439575195,
      "epoch": 1.2612872238232469,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00048480435743626703,
      "loss": 4.7979,
      "mean_token_accuracy": 0.23628615587949753,
      "num_tokens": 30090335.0,
      "step": 13130
    },
    {
      "entropy": 5.235983180999756,
      "epoch": 1.2617675312199808,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004847919696071054,
      "loss": 4.8981,
      "mean_token_accuracy": 0.22588902115821838,
      "num_tokens": 30101781.0,
      "step": 13135
    },
    {
      "entropy": 5.246907663345337,
      "epoch": 1.2622478386167146,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048477957690719716,
      "loss": 4.7926,
      "mean_token_accuracy": 0.24108311235904695,
      "num_tokens": 30112486.0,
      "step": 13140
    },
    {
      "entropy": 5.263914680480957,
      "epoch": 1.2627281460134485,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004847671793368301,
      "loss": 4.8928,
      "mean_token_accuracy": 0.2277855709195137,
      "num_tokens": 30123664.0,
      "step": 13145
    },
    {
      "entropy": 5.3062084197998045,
      "epoch": 1.2632084534101824,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000484754776896292,
      "loss": 4.9488,
      "mean_token_accuracy": 0.22482662051916122,
      "num_tokens": 30134139.0,
      "step": 13150
    },
    {
      "entropy": 5.237970209121704,
      "epoch": 1.2636887608069165,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004847423695858708,
      "loss": 4.9023,
      "mean_token_accuracy": 0.22652390599250793,
      "num_tokens": 30146127.0,
      "step": 13155
    },
    {
      "entropy": 5.3226978302001955,
      "epoch": 1.2641690682036504,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048472995740585456,
      "loss": 4.8994,
      "mean_token_accuracy": 0.22508987337350844,
      "num_tokens": 30158177.0,
      "step": 13160
    },
    {
      "entropy": 5.291877603530883,
      "epoch": 1.2646493756003843,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004847175403565316,
      "loss": 4.8899,
      "mean_token_accuracy": 0.22736635357141494,
      "num_tokens": 30168930.0,
      "step": 13165
    },
    {
      "entropy": 5.325614070892334,
      "epoch": 1.2651296829971181,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048470511843818996,
      "loss": 4.9052,
      "mean_token_accuracy": 0.22320545017719268,
      "num_tokens": 30181206.0,
      "step": 13170
    },
    {
      "entropy": 5.296067237854004,
      "epoch": 1.265609990393852,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004846926916511182,
      "loss": 4.929,
      "mean_token_accuracy": 0.23548732697963715,
      "num_tokens": 30193093.0,
      "step": 13175
    },
    {
      "entropy": 5.219576025009156,
      "epoch": 1.266090297790586,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004846802599956048,
      "loss": 4.7864,
      "mean_token_accuracy": 0.23758206963539125,
      "num_tokens": 30204864.0,
      "step": 13180
    },
    {
      "entropy": 5.3023045539855955,
      "epoch": 1.26657060518732,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048466782347193847,
      "loss": 4.8866,
      "mean_token_accuracy": 0.22498230934143065,
      "num_tokens": 30216312.0,
      "step": 13185
    },
    {
      "entropy": 5.330937242507934,
      "epoch": 1.267050912584054,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048465538208040775,
      "loss": 4.9543,
      "mean_token_accuracy": 0.2200424090027809,
      "num_tokens": 30227824.0,
      "step": 13190
    },
    {
      "entropy": 5.309969139099121,
      "epoch": 1.2675312199807878,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048464293582130166,
      "loss": 4.9425,
      "mean_token_accuracy": 0.21994371116161346,
      "num_tokens": 30239065.0,
      "step": 13195
    },
    {
      "entropy": 5.2293332576751705,
      "epoch": 1.2680115273775217,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004846304846949091,
      "loss": 4.7587,
      "mean_token_accuracy": 0.22892523556947708,
      "num_tokens": 30250684.0,
      "step": 13200
    },
    {
      "entropy": 5.258307266235351,
      "epoch": 1.2684918347742555,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048461802870151916,
      "loss": 4.8754,
      "mean_token_accuracy": 0.22773925215005875,
      "num_tokens": 30261376.0,
      "step": 13205
    },
    {
      "entropy": 5.271623325347901,
      "epoch": 1.2689721421709894,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048460556784142106,
      "loss": 4.8638,
      "mean_token_accuracy": 0.23187339752912522,
      "num_tokens": 30272441.0,
      "step": 13210
    },
    {
      "entropy": 5.238671112060547,
      "epoch": 1.2694524495677233,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048459310211490406,
      "loss": 4.869,
      "mean_token_accuracy": 0.2318838655948639,
      "num_tokens": 30284446.0,
      "step": 13215
    },
    {
      "entropy": 5.307750844955445,
      "epoch": 1.2699327569644572,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004845806315222576,
      "loss": 4.969,
      "mean_token_accuracy": 0.22499972581863403,
      "num_tokens": 30295666.0,
      "step": 13220
    },
    {
      "entropy": 5.317988157272339,
      "epoch": 1.270413064361191,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004845681560637711,
      "loss": 4.9167,
      "mean_token_accuracy": 0.2272246763110161,
      "num_tokens": 30307663.0,
      "step": 13225
    },
    {
      "entropy": 5.310830307006836,
      "epoch": 1.270893371757925,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004845556757397344,
      "loss": 4.9121,
      "mean_token_accuracy": 0.2235385522246361,
      "num_tokens": 30318914.0,
      "step": 13230
    },
    {
      "entropy": 5.288627481460571,
      "epoch": 1.271373679154659,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004845431905504372,
      "loss": 4.9715,
      "mean_token_accuracy": 0.2188763439655304,
      "num_tokens": 30331499.0,
      "step": 13235
    },
    {
      "entropy": 5.353644990921021,
      "epoch": 1.271853986551393,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048453070049616926,
      "loss": 4.9059,
      "mean_token_accuracy": 0.22854580879211425,
      "num_tokens": 30342186.0,
      "step": 13240
    },
    {
      "entropy": 5.339401006698608,
      "epoch": 1.2723342939481268,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048451820557722064,
      "loss": 4.9171,
      "mean_token_accuracy": 0.22544696182012558,
      "num_tokens": 30353423.0,
      "step": 13245
    },
    {
      "entropy": 5.234491539001465,
      "epoch": 1.2728146013448607,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004845057057938815,
      "loss": 4.8435,
      "mean_token_accuracy": 0.23332713842391967,
      "num_tokens": 30365008.0,
      "step": 13250
    },
    {
      "entropy": 5.175595140457153,
      "epoch": 1.2732949087415946,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048449320114644185,
      "loss": 4.7786,
      "mean_token_accuracy": 0.23910536915063857,
      "num_tokens": 30376589.0,
      "step": 13255
    },
    {
      "entropy": 5.256163311004639,
      "epoch": 1.2737752161383284,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004844806916351922,
      "loss": 4.8776,
      "mean_token_accuracy": 0.2265567034482956,
      "num_tokens": 30388228.0,
      "step": 13260
    },
    {
      "entropy": 5.242117691040039,
      "epoch": 1.2742555235350626,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004844681772604229,
      "loss": 4.9142,
      "mean_token_accuracy": 0.22831425666809083,
      "num_tokens": 30400225.0,
      "step": 13265
    },
    {
      "entropy": 5.2744872093200685,
      "epoch": 1.2747358309317964,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048445565802242454,
      "loss": 4.8852,
      "mean_token_accuracy": 0.22790091931819917,
      "num_tokens": 30411913.0,
      "step": 13270
    },
    {
      "entropy": 5.258826971054077,
      "epoch": 1.2752161383285303,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004844431339214878,
      "loss": 4.8494,
      "mean_token_accuracy": 0.2292488768696785,
      "num_tokens": 30423891.0,
      "step": 13275
    },
    {
      "entropy": 5.315962553024292,
      "epoch": 1.2756964457252642,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004844306049579034,
      "loss": 4.9033,
      "mean_token_accuracy": 0.2280384734272957,
      "num_tokens": 30434144.0,
      "step": 13280
    },
    {
      "entropy": 5.2179931640625,
      "epoch": 1.276176753121998,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048441807113196216,
      "loss": 4.8592,
      "mean_token_accuracy": 0.22284265756607055,
      "num_tokens": 30446628.0,
      "step": 13285
    },
    {
      "entropy": 5.3117717742919925,
      "epoch": 1.276657060518732,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048440553244395517,
      "loss": 4.855,
      "mean_token_accuracy": 0.2293342962861061,
      "num_tokens": 30456885.0,
      "step": 13290
    },
    {
      "entropy": 5.314032363891601,
      "epoch": 1.2771373679154658,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048439298889417357,
      "loss": 4.9928,
      "mean_token_accuracy": 0.22162097841501235,
      "num_tokens": 30468642.0,
      "step": 13295
    },
    {
      "entropy": 5.296462297439575,
      "epoch": 1.2776176753121997,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048438044048290847,
      "loss": 4.8714,
      "mean_token_accuracy": 0.21795614361763,
      "num_tokens": 30480393.0,
      "step": 13300
    },
    {
      "entropy": 5.24354248046875,
      "epoch": 1.2780979827089336,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048436788721045135,
      "loss": 4.8161,
      "mean_token_accuracy": 0.23068082183599473,
      "num_tokens": 30491918.0,
      "step": 13305
    },
    {
      "entropy": 5.28349027633667,
      "epoch": 1.2785782901056677,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004843553290770935,
      "loss": 4.8142,
      "mean_token_accuracy": 0.2344738945364952,
      "num_tokens": 30502250.0,
      "step": 13310
    },
    {
      "entropy": 5.23266749382019,
      "epoch": 1.2790585975024016,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004843427660831266,
      "loss": 4.8511,
      "mean_token_accuracy": 0.2237774148583412,
      "num_tokens": 30514499.0,
      "step": 13315
    },
    {
      "entropy": 5.22802619934082,
      "epoch": 1.2795389048991355,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048433019822884235,
      "loss": 4.8903,
      "mean_token_accuracy": 0.22280975431203842,
      "num_tokens": 30525133.0,
      "step": 13320
    },
    {
      "entropy": 5.224843597412109,
      "epoch": 1.2800192122958693,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004843176255145325,
      "loss": 4.7935,
      "mean_token_accuracy": 0.23035313040018082,
      "num_tokens": 30535778.0,
      "step": 13325
    },
    {
      "entropy": 5.24747109413147,
      "epoch": 1.2804995196926032,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004843050479404888,
      "loss": 4.7896,
      "mean_token_accuracy": 0.24365344047546386,
      "num_tokens": 30547711.0,
      "step": 13330
    },
    {
      "entropy": 5.174214458465576,
      "epoch": 1.280979827089337,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048429246550700343,
      "loss": 4.7556,
      "mean_token_accuracy": 0.2336757943034172,
      "num_tokens": 30558719.0,
      "step": 13335
    },
    {
      "entropy": 5.196592712402344,
      "epoch": 1.2814601344860712,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004842798782143686,
      "loss": 4.8827,
      "mean_token_accuracy": 0.228485107421875,
      "num_tokens": 30570173.0,
      "step": 13340
    },
    {
      "entropy": 5.320303630828858,
      "epoch": 1.281940441882805,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048426728606287627,
      "loss": 4.88,
      "mean_token_accuracy": 0.23605505079030992,
      "num_tokens": 30580438.0,
      "step": 13345
    },
    {
      "entropy": 5.29133358001709,
      "epoch": 1.282420749279539,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004842546890528191,
      "loss": 4.8564,
      "mean_token_accuracy": 0.2245273694396019,
      "num_tokens": 30592191.0,
      "step": 13350
    },
    {
      "entropy": 5.2436981201171875,
      "epoch": 1.2829010566762729,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004842420871844893,
      "loss": 4.9288,
      "mean_token_accuracy": 0.22842509001493455,
      "num_tokens": 30604041.0,
      "step": 13355
    },
    {
      "entropy": 5.199397087097168,
      "epoch": 1.2833813640730067,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004842294804581796,
      "loss": 4.8302,
      "mean_token_accuracy": 0.24013564586639405,
      "num_tokens": 30616087.0,
      "step": 13360
    },
    {
      "entropy": 5.274821090698242,
      "epoch": 1.2838616714697406,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048421686887418266,
      "loss": 4.8743,
      "mean_token_accuracy": 0.229492712020874,
      "num_tokens": 30627150.0,
      "step": 13365
    },
    {
      "entropy": 5.283530902862549,
      "epoch": 1.2843419788664745,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004842042524327912,
      "loss": 4.8725,
      "mean_token_accuracy": 0.22866750955581666,
      "num_tokens": 30638627.0,
      "step": 13370
    },
    {
      "entropy": 5.2765127658844,
      "epoch": 1.2848222862632084,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004841916311342983,
      "loss": 4.9139,
      "mean_token_accuracy": 0.22479262202978134,
      "num_tokens": 30649257.0,
      "step": 13375
    },
    {
      "entropy": 5.240016031265259,
      "epoch": 1.2853025936599423,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004841790049789969,
      "loss": 4.9057,
      "mean_token_accuracy": 0.23450375497341155,
      "num_tokens": 30659765.0,
      "step": 13380
    },
    {
      "entropy": 5.270106935501099,
      "epoch": 1.2857829010566761,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048416637396718004,
      "loss": 4.8508,
      "mean_token_accuracy": 0.2234204888343811,
      "num_tokens": 30672921.0,
      "step": 13385
    },
    {
      "entropy": 5.299501180648804,
      "epoch": 1.2862632084534102,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004841537380991411,
      "loss": 4.8674,
      "mean_token_accuracy": 0.23164343535900117,
      "num_tokens": 30684994.0,
      "step": 13390
    },
    {
      "entropy": 5.304014730453491,
      "epoch": 1.2867435158501441,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048414109737517346,
      "loss": 4.894,
      "mean_token_accuracy": 0.22779832929372787,
      "num_tokens": 30698094.0,
      "step": 13395
    },
    {
      "entropy": 5.275398635864258,
      "epoch": 1.287223823246878,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004841284517955706,
      "loss": 4.8865,
      "mean_token_accuracy": 0.22490549832582474,
      "num_tokens": 30710716.0,
      "step": 13400
    },
    {
      "entropy": 5.275581693649292,
      "epoch": 1.2877041306436119,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000484115801360626,
      "loss": 4.8957,
      "mean_token_accuracy": 0.22964722514152527,
      "num_tokens": 30721910.0,
      "step": 13405
    },
    {
      "entropy": 5.246979522705078,
      "epoch": 1.2881844380403458,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004841031460706335,
      "loss": 4.8413,
      "mean_token_accuracy": 0.2317227989435196,
      "num_tokens": 30734445.0,
      "step": 13410
    },
    {
      "entropy": 5.296369552612305,
      "epoch": 1.2886647454370797,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048409048592588683,
      "loss": 4.8059,
      "mean_token_accuracy": 0.23733747154474258,
      "num_tokens": 30745131.0,
      "step": 13415
    },
    {
      "entropy": 5.276104593276978,
      "epoch": 1.2891450528338138,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004840778209266799,
      "loss": 4.8585,
      "mean_token_accuracy": 0.22796287685632705,
      "num_tokens": 30757714.0,
      "step": 13420
    },
    {
      "entropy": 5.142038106918335,
      "epoch": 1.2896253602305476,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048406515107330685,
      "loss": 4.8362,
      "mean_token_accuracy": 0.23704024255275727,
      "num_tokens": 30768585.0,
      "step": 13425
    },
    {
      "entropy": 5.282596015930176,
      "epoch": 1.2901056676272815,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048405247636606173,
      "loss": 4.8905,
      "mean_token_accuracy": 0.2345744028687477,
      "num_tokens": 30779866.0,
      "step": 13430
    },
    {
      "entropy": 5.311311483383179,
      "epoch": 1.2905859750240154,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048403979680523894,
      "loss": 4.8527,
      "mean_token_accuracy": 0.22467785328626633,
      "num_tokens": 30791252.0,
      "step": 13435
    },
    {
      "entropy": 5.297217178344726,
      "epoch": 1.2910662824207493,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004840271123911328,
      "loss": 4.8845,
      "mean_token_accuracy": 0.237962906062603,
      "num_tokens": 30802938.0,
      "step": 13440
    },
    {
      "entropy": 5.343379974365234,
      "epoch": 1.2915465898174832,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004840144231240377,
      "loss": 5.0211,
      "mean_token_accuracy": 0.21456135511398317,
      "num_tokens": 30816382.0,
      "step": 13445
    },
    {
      "entropy": 5.291771411895752,
      "epoch": 1.292026897214217,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004840017290042484,
      "loss": 4.8065,
      "mean_token_accuracy": 0.2282480329275131,
      "num_tokens": 30826402.0,
      "step": 13450
    },
    {
      "entropy": 5.254259538650513,
      "epoch": 1.292507204610951,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048398903003205957,
      "loss": 4.8169,
      "mean_token_accuracy": 0.23328054547309876,
      "num_tokens": 30837862.0,
      "step": 13455
    },
    {
      "entropy": 5.229075384140015,
      "epoch": 1.2929875120076848,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048397632620776604,
      "loss": 4.7951,
      "mean_token_accuracy": 0.230962772667408,
      "num_tokens": 30848351.0,
      "step": 13460
    },
    {
      "entropy": 5.206089735031128,
      "epoch": 1.293467819404419,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048396361753166276,
      "loss": 4.8748,
      "mean_token_accuracy": 0.2301948517560959,
      "num_tokens": 30858854.0,
      "step": 13465
    },
    {
      "entropy": 5.22732892036438,
      "epoch": 1.2939481268011528,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048395090400404466,
      "loss": 4.8849,
      "mean_token_accuracy": 0.22781737446784972,
      "num_tokens": 30869977.0,
      "step": 13470
    },
    {
      "entropy": 5.364018201828003,
      "epoch": 1.2944284341978867,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048393818562520715,
      "loss": 4.8931,
      "mean_token_accuracy": 0.23006015568971633,
      "num_tokens": 30881620.0,
      "step": 13475
    },
    {
      "entropy": 5.311263084411621,
      "epoch": 1.2949087415946205,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048392546239544535,
      "loss": 4.9228,
      "mean_token_accuracy": 0.22454346269369124,
      "num_tokens": 30893575.0,
      "step": 13480
    },
    {
      "entropy": 5.233764171600342,
      "epoch": 1.2953890489913544,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004839127343150547,
      "loss": 4.8659,
      "mean_token_accuracy": 0.23050324618816376,
      "num_tokens": 30904489.0,
      "step": 13485
    },
    {
      "entropy": 5.319433784484863,
      "epoch": 1.2958693563880883,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004839000013843307,
      "loss": 4.9404,
      "mean_token_accuracy": 0.2247033193707466,
      "num_tokens": 30915513.0,
      "step": 13490
    },
    {
      "entropy": 5.306324625015259,
      "epoch": 1.2963496637848224,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048388726360356894,
      "loss": 4.8713,
      "mean_token_accuracy": 0.2301812142133713,
      "num_tokens": 30927633.0,
      "step": 13495
    },
    {
      "entropy": 5.381054973602295,
      "epoch": 1.2968299711815563,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004838745209730653,
      "loss": 4.9567,
      "mean_token_accuracy": 0.21706148833036423,
      "num_tokens": 30938937.0,
      "step": 13500
    },
    {
      "entropy": 5.25718240737915,
      "epoch": 1.2973102785782902,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048386177349311535,
      "loss": 4.8625,
      "mean_token_accuracy": 0.23206737488508225,
      "num_tokens": 30950847.0,
      "step": 13505
    },
    {
      "entropy": 5.265108823776245,
      "epoch": 1.297790585975024,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004838490211640153,
      "loss": 4.8138,
      "mean_token_accuracy": 0.23250877857208252,
      "num_tokens": 30962583.0,
      "step": 13510
    },
    {
      "entropy": 5.302864742279053,
      "epoch": 1.298270893371758,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004838362639860611,
      "loss": 4.8988,
      "mean_token_accuracy": 0.22301805019378662,
      "num_tokens": 30973487.0,
      "step": 13515
    },
    {
      "entropy": 5.196312999725341,
      "epoch": 1.2987512007684918,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000483823501959549,
      "loss": 4.7893,
      "mean_token_accuracy": 0.23483429700136185,
      "num_tokens": 30985176.0,
      "step": 13520
    },
    {
      "entropy": 5.282163763046265,
      "epoch": 1.2992315081652257,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048381073508477527,
      "loss": 4.9684,
      "mean_token_accuracy": 0.2225658416748047,
      "num_tokens": 30997239.0,
      "step": 13525
    },
    {
      "entropy": 5.354573202133179,
      "epoch": 1.2997118155619596,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048379796336203625,
      "loss": 4.9174,
      "mean_token_accuracy": 0.22067348659038544,
      "num_tokens": 31009381.0,
      "step": 13530
    },
    {
      "entropy": 5.272178983688354,
      "epoch": 1.3001921229586935,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004837851867916286,
      "loss": 4.8306,
      "mean_token_accuracy": 0.23224907368421555,
      "num_tokens": 31020728.0,
      "step": 13535
    },
    {
      "entropy": 5.291997194290161,
      "epoch": 1.3006724303554273,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004837724053738488,
      "loss": 4.943,
      "mean_token_accuracy": 0.22559564411640168,
      "num_tokens": 31032933.0,
      "step": 13540
    },
    {
      "entropy": 5.284105205535889,
      "epoch": 1.3011527377521614,
      "grad_norm": 1.0,
      "learning_rate": 0.00048375961910899373,
      "loss": 4.9074,
      "mean_token_accuracy": 0.22861146777868271,
      "num_tokens": 31044919.0,
      "step": 13545
    },
    {
      "entropy": 5.303400897979737,
      "epoch": 1.3016330451488953,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004837468279973602,
      "loss": 4.8926,
      "mean_token_accuracy": 0.22767747193574905,
      "num_tokens": 31057140.0,
      "step": 13550
    },
    {
      "entropy": 5.266713953018188,
      "epoch": 1.3021133525456292,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004837340320392451,
      "loss": 4.8152,
      "mean_token_accuracy": 0.22857447415590287,
      "num_tokens": 31068035.0,
      "step": 13555
    },
    {
      "entropy": 5.276372814178467,
      "epoch": 1.302593659942363,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048372123123494563,
      "loss": 4.8978,
      "mean_token_accuracy": 0.22378070801496505,
      "num_tokens": 31079819.0,
      "step": 13560
    },
    {
      "entropy": 5.156053590774536,
      "epoch": 1.303073967339097,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000483708425584759,
      "loss": 4.7929,
      "mean_token_accuracy": 0.23344516307115554,
      "num_tokens": 31091272.0,
      "step": 13565
    },
    {
      "entropy": 5.292856454849243,
      "epoch": 1.3035542747358309,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004836956150889825,
      "loss": 4.9014,
      "mean_token_accuracy": 0.22871409058570863,
      "num_tokens": 31101513.0,
      "step": 13570
    },
    {
      "entropy": 5.331047630310058,
      "epoch": 1.304034582132565,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004836827997479134,
      "loss": 4.9479,
      "mean_token_accuracy": 0.21857866197824477,
      "num_tokens": 31112937.0,
      "step": 13575
    },
    {
      "entropy": 5.311564683914185,
      "epoch": 1.3045148895292988,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004836699795618494,
      "loss": 4.9346,
      "mean_token_accuracy": 0.23036938458681105,
      "num_tokens": 31123445.0,
      "step": 13580
    },
    {
      "entropy": 5.260301208496093,
      "epoch": 1.3049951969260327,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004836571545310881,
      "loss": 4.8772,
      "mean_token_accuracy": 0.22796281427145004,
      "num_tokens": 31135016.0,
      "step": 13585
    },
    {
      "entropy": 5.22722978591919,
      "epoch": 1.3054755043227666,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048364432465592723,
      "loss": 4.846,
      "mean_token_accuracy": 0.22718782275915145,
      "num_tokens": 31146772.0,
      "step": 13590
    },
    {
      "entropy": 5.245557498931885,
      "epoch": 1.3059558117195005,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004836314899366647,
      "loss": 4.7975,
      "mean_token_accuracy": 0.23855511248111724,
      "num_tokens": 31157977.0,
      "step": 13595
    },
    {
      "entropy": 5.215791988372803,
      "epoch": 1.3064361191162344,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048361865037359846,
      "loss": 4.8236,
      "mean_token_accuracy": 0.23507911562919617,
      "num_tokens": 31170693.0,
      "step": 13600
    },
    {
      "entropy": 5.249704504013062,
      "epoch": 1.3069164265129682,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048360580596702664,
      "loss": 4.8516,
      "mean_token_accuracy": 0.23068362176418306,
      "num_tokens": 31181276.0,
      "step": 13605
    },
    {
      "entropy": 5.3251903533935545,
      "epoch": 1.3073967339097021,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048359295671724744,
      "loss": 4.9628,
      "mean_token_accuracy": 0.22135923504829408,
      "num_tokens": 31192740.0,
      "step": 13610
    },
    {
      "entropy": 5.286228656768799,
      "epoch": 1.307877041306436,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004835801026245592,
      "loss": 4.8653,
      "mean_token_accuracy": 0.22010988742113113,
      "num_tokens": 31204075.0,
      "step": 13615
    },
    {
      "entropy": 5.314511251449585,
      "epoch": 1.30835734870317,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048356724368926035,
      "loss": 4.9503,
      "mean_token_accuracy": 0.21232653558254241,
      "num_tokens": 31215590.0,
      "step": 13620
    },
    {
      "entropy": 5.312569189071655,
      "epoch": 1.308837656099904,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048355437991164937,
      "loss": 4.9418,
      "mean_token_accuracy": 0.22125620394945145,
      "num_tokens": 31228427.0,
      "step": 13625
    },
    {
      "entropy": 5.315005826950073,
      "epoch": 1.3093179634966379,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000483541511292025,
      "loss": 4.9527,
      "mean_token_accuracy": 0.2255345582962036,
      "num_tokens": 31240348.0,
      "step": 13630
    },
    {
      "entropy": 5.333975791931152,
      "epoch": 1.3097982708933718,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048352863783068594,
      "loss": 4.9649,
      "mean_token_accuracy": 0.22510398924350739,
      "num_tokens": 31253028.0,
      "step": 13635
    },
    {
      "entropy": 5.358338022232056,
      "epoch": 1.3102785782901056,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048351575952793117,
      "loss": 4.9408,
      "mean_token_accuracy": 0.2209530383348465,
      "num_tokens": 31265242.0,
      "step": 13640
    },
    {
      "entropy": 5.290958881378174,
      "epoch": 1.3107588856868395,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004835028763840595,
      "loss": 4.8778,
      "mean_token_accuracy": 0.22581043988466262,
      "num_tokens": 31277248.0,
      "step": 13645
    },
    {
      "entropy": 5.287257814407349,
      "epoch": 1.3112391930835736,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004834899883993703,
      "loss": 4.9284,
      "mean_token_accuracy": 0.22163309007883072,
      "num_tokens": 31289631.0,
      "step": 13650
    },
    {
      "entropy": 5.185016345977783,
      "epoch": 1.3117195004803075,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048347709557416263,
      "loss": 4.7749,
      "mean_token_accuracy": 0.23895397484302522,
      "num_tokens": 31301006.0,
      "step": 13655
    },
    {
      "entropy": 5.315671396255493,
      "epoch": 1.3121998078770414,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004834641979087359,
      "loss": 4.9538,
      "mean_token_accuracy": 0.22419577091932297,
      "num_tokens": 31312621.0,
      "step": 13660
    },
    {
      "entropy": 5.3225000381469725,
      "epoch": 1.3126801152737753,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004834512954033894,
      "loss": 4.9295,
      "mean_token_accuracy": 0.22379688620567323,
      "num_tokens": 31324173.0,
      "step": 13665
    },
    {
      "entropy": 5.2718583106994625,
      "epoch": 1.3131604226705091,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048343838805842284,
      "loss": 4.8759,
      "mean_token_accuracy": 0.23028386384248734,
      "num_tokens": 31336872.0,
      "step": 13670
    },
    {
      "entropy": 5.337061262130737,
      "epoch": 1.313640730067243,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048342547587413583,
      "loss": 4.9203,
      "mean_token_accuracy": 0.22404322475194932,
      "num_tokens": 31348760.0,
      "step": 13675
    },
    {
      "entropy": 5.296292543411255,
      "epoch": 1.314121037463977,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004834125588508282,
      "loss": 4.8681,
      "mean_token_accuracy": 0.22850588113069534,
      "num_tokens": 31360591.0,
      "step": 13680
    },
    {
      "entropy": 5.368846321105957,
      "epoch": 1.3146013448607108,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004833996369887998,
      "loss": 5.0486,
      "mean_token_accuracy": 0.22744137793779373,
      "num_tokens": 31372273.0,
      "step": 13685
    },
    {
      "entropy": 5.265228796005249,
      "epoch": 1.3150816522574447,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048338671028835063,
      "loss": 4.8712,
      "mean_token_accuracy": 0.23031177520751953,
      "num_tokens": 31382569.0,
      "step": 13690
    },
    {
      "entropy": 5.3737880229949955,
      "epoch": 1.3155619596541785,
      "grad_norm": 1.0,
      "learning_rate": 0.00048337377874978086,
      "loss": 4.9436,
      "mean_token_accuracy": 0.21900282204151153,
      "num_tokens": 31393913.0,
      "step": 13695
    },
    {
      "entropy": 5.221459484100341,
      "epoch": 1.3160422670509127,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048336084237339067,
      "loss": 4.8312,
      "mean_token_accuracy": 0.23084236234426497,
      "num_tokens": 31405645.0,
      "step": 13700
    },
    {
      "entropy": 5.156911373138428,
      "epoch": 1.3165225744476465,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004833479011594804,
      "loss": 4.7949,
      "mean_token_accuracy": 0.23470553606748581,
      "num_tokens": 31418728.0,
      "step": 13705
    },
    {
      "entropy": 5.236312675476074,
      "epoch": 1.3170028818443804,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048333495510835057,
      "loss": 4.8474,
      "mean_token_accuracy": 0.23132781982421874,
      "num_tokens": 31430226.0,
      "step": 13710
    },
    {
      "entropy": 5.278441667556763,
      "epoch": 1.3174831892411143,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048332200422030163,
      "loss": 4.8362,
      "mean_token_accuracy": 0.2332627460360527,
      "num_tokens": 31441751.0,
      "step": 13715
    },
    {
      "entropy": 5.221287059783935,
      "epoch": 1.3179634966378482,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004833090484956345,
      "loss": 4.9034,
      "mean_token_accuracy": 0.22543685138225555,
      "num_tokens": 31455413.0,
      "step": 13720
    },
    {
      "entropy": 5.2819760799407955,
      "epoch": 1.318443804034582,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048329608793464966,
      "loss": 4.7789,
      "mean_token_accuracy": 0.24309951066970825,
      "num_tokens": 31466486.0,
      "step": 13725
    },
    {
      "entropy": 5.2563316822052,
      "epoch": 1.3189241114313162,
      "grad_norm": 1.0,
      "learning_rate": 0.0004832831225376482,
      "loss": 4.8745,
      "mean_token_accuracy": 0.23323155641555787,
      "num_tokens": 31477048.0,
      "step": 13730
    },
    {
      "entropy": 5.2812220573425295,
      "epoch": 1.31940441882805,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004832701523049312,
      "loss": 4.9172,
      "mean_token_accuracy": 0.2238553583621979,
      "num_tokens": 31488015.0,
      "step": 13735
    },
    {
      "entropy": 5.276743030548095,
      "epoch": 1.319884726224784,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004832571772367997,
      "loss": 4.8203,
      "mean_token_accuracy": 0.22813392281532288,
      "num_tokens": 31499735.0,
      "step": 13740
    },
    {
      "entropy": 5.13890700340271,
      "epoch": 1.3203650336215178,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048324419733355485,
      "loss": 4.6679,
      "mean_token_accuracy": 0.23637510389089583,
      "num_tokens": 31511970.0,
      "step": 13745
    },
    {
      "entropy": 5.268892288208008,
      "epoch": 1.3208453410182517,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048323121259549805,
      "loss": 4.9027,
      "mean_token_accuracy": 0.22958308160305024,
      "num_tokens": 31522780.0,
      "step": 13750
    },
    {
      "entropy": 5.309873914718628,
      "epoch": 1.3213256484149856,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048321822302293095,
      "loss": 4.8871,
      "mean_token_accuracy": 0.2281516522169113,
      "num_tokens": 31535578.0,
      "step": 13755
    },
    {
      "entropy": 5.346133041381836,
      "epoch": 1.3218059558117194,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004832052286161549,
      "loss": 4.9676,
      "mean_token_accuracy": 0.21998259127140046,
      "num_tokens": 31546623.0,
      "step": 13760
    },
    {
      "entropy": 5.3297850608825685,
      "epoch": 1.3222862632084533,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048319222937547176,
      "loss": 5.02,
      "mean_token_accuracy": 0.2194166287779808,
      "num_tokens": 31559334.0,
      "step": 13765
    },
    {
      "entropy": 5.296007251739502,
      "epoch": 1.3227665706051872,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048317922530118323,
      "loss": 4.8636,
      "mean_token_accuracy": 0.23496466726064683,
      "num_tokens": 31570431.0,
      "step": 13770
    },
    {
      "entropy": 5.2247161865234375,
      "epoch": 1.3232468780019213,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004831662163935912,
      "loss": 4.7576,
      "mean_token_accuracy": 0.23651786893606186,
      "num_tokens": 31580314.0,
      "step": 13775
    },
    {
      "entropy": 5.232627248764038,
      "epoch": 1.3237271853986552,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048315320265299784,
      "loss": 4.8272,
      "mean_token_accuracy": 0.23777729421854019,
      "num_tokens": 31592071.0,
      "step": 13780
    },
    {
      "entropy": 5.356109189987182,
      "epoch": 1.324207492795389,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048314018407970516,
      "loss": 4.9917,
      "mean_token_accuracy": 0.22196717113256453,
      "num_tokens": 31603143.0,
      "step": 13785
    },
    {
      "entropy": 5.354069709777832,
      "epoch": 1.324687800192123,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048312716067401535,
      "loss": 4.9925,
      "mean_token_accuracy": 0.22331462651491166,
      "num_tokens": 31613979.0,
      "step": 13790
    },
    {
      "entropy": 5.267085123062134,
      "epoch": 1.3251681075888568,
      "grad_norm": 0.921875,
      "learning_rate": 0.000483114132436231,
      "loss": 4.8323,
      "mean_token_accuracy": 0.230244480073452,
      "num_tokens": 31625206.0,
      "step": 13795
    },
    {
      "entropy": 5.346383428573608,
      "epoch": 1.3256484149855907,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004831010993666545,
      "loss": 4.9406,
      "mean_token_accuracy": 0.2333788201212883,
      "num_tokens": 31637709.0,
      "step": 13800
    },
    {
      "entropy": 5.295392847061157,
      "epoch": 1.3261287223823248,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004830880614655884,
      "loss": 4.9125,
      "mean_token_accuracy": 0.22467359900474548,
      "num_tokens": 31648754.0,
      "step": 13805
    },
    {
      "entropy": 5.215533065795898,
      "epoch": 1.3266090297790587,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048307501873333527,
      "loss": 4.7763,
      "mean_token_accuracy": 0.23736656457185745,
      "num_tokens": 31660177.0,
      "step": 13810
    },
    {
      "entropy": 5.2773754596710205,
      "epoch": 1.3270893371757926,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004830619711701982,
      "loss": 4.9246,
      "mean_token_accuracy": 0.22303781062364578,
      "num_tokens": 31671467.0,
      "step": 13815
    },
    {
      "entropy": 5.328343343734741,
      "epoch": 1.3275696445725265,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004830489187764799,
      "loss": 4.9032,
      "mean_token_accuracy": 0.22712527960538864,
      "num_tokens": 31683073.0,
      "step": 13820
    },
    {
      "entropy": 5.215532112121582,
      "epoch": 1.3280499519692603,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004830358615524835,
      "loss": 4.8395,
      "mean_token_accuracy": 0.23019295036792756,
      "num_tokens": 31694661.0,
      "step": 13825
    },
    {
      "entropy": 5.2070960521698,
      "epoch": 1.3285302593659942,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048302279949851215,
      "loss": 4.871,
      "mean_token_accuracy": 0.22348390221595765,
      "num_tokens": 31706904.0,
      "step": 13830
    },
    {
      "entropy": 5.235240030288696,
      "epoch": 1.329010566762728,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048300973261486906,
      "loss": 4.8296,
      "mean_token_accuracy": 0.22953573018312454,
      "num_tokens": 31718087.0,
      "step": 13835
    },
    {
      "entropy": 5.232860660552978,
      "epoch": 1.329490874159462,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004829966609018577,
      "loss": 4.7375,
      "mean_token_accuracy": 0.23485807776451112,
      "num_tokens": 31728960.0,
      "step": 13840
    },
    {
      "entropy": 5.3118226528167725,
      "epoch": 1.3299711815561959,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048298358435978146,
      "loss": 4.9063,
      "mean_token_accuracy": 0.22645290940999985,
      "num_tokens": 31741461.0,
      "step": 13845
    },
    {
      "entropy": 5.2232616424560545,
      "epoch": 1.3304514889529298,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048297050298894394,
      "loss": 4.8371,
      "mean_token_accuracy": 0.22565432339906694,
      "num_tokens": 31753582.0,
      "step": 13850
    },
    {
      "entropy": 5.2139747619628904,
      "epoch": 1.3309317963496639,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004829574167896489,
      "loss": 4.7916,
      "mean_token_accuracy": 0.23458465039730073,
      "num_tokens": 31764354.0,
      "step": 13855
    },
    {
      "entropy": 5.263726043701172,
      "epoch": 1.3314121037463977,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048294432576220027,
      "loss": 4.837,
      "mean_token_accuracy": 0.22190757393836974,
      "num_tokens": 31775860.0,
      "step": 13860
    },
    {
      "entropy": 5.278409194946289,
      "epoch": 1.3318924111431316,
      "grad_norm": 0.875,
      "learning_rate": 0.0004829312299069017,
      "loss": 4.8754,
      "mean_token_accuracy": 0.2319222003221512,
      "num_tokens": 31787606.0,
      "step": 13865
    },
    {
      "entropy": 5.268970775604248,
      "epoch": 1.3323727185398655,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048291812922405755,
      "loss": 4.8531,
      "mean_token_accuracy": 0.23262912333011626,
      "num_tokens": 31798601.0,
      "step": 13870
    },
    {
      "entropy": 5.248348617553711,
      "epoch": 1.3328530259365994,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004829050237139717,
      "loss": 4.8118,
      "mean_token_accuracy": 0.23685635179281234,
      "num_tokens": 31809843.0,
      "step": 13875
    },
    {
      "entropy": 5.303222370147705,
      "epoch": 1.3333333333333333,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004828919133769486,
      "loss": 4.9187,
      "mean_token_accuracy": 0.22461518943309783,
      "num_tokens": 31821309.0,
      "step": 13880
    },
    {
      "entropy": 5.296580839157104,
      "epoch": 1.3338136407300674,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004828787982132926,
      "loss": 4.8433,
      "mean_token_accuracy": 0.23161176592111588,
      "num_tokens": 31832215.0,
      "step": 13885
    },
    {
      "entropy": 5.2321268081665036,
      "epoch": 1.3342939481268012,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048286567822330815,
      "loss": 4.8059,
      "mean_token_accuracy": 0.24145893454551698,
      "num_tokens": 31843254.0,
      "step": 13890
    },
    {
      "entropy": 5.176913833618164,
      "epoch": 1.3347742555235351,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004828525534072999,
      "loss": 4.8254,
      "mean_token_accuracy": 0.2281007319688797,
      "num_tokens": 31855678.0,
      "step": 13895
    },
    {
      "entropy": 5.278821992874145,
      "epoch": 1.335254562920269,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048283942376557254,
      "loss": 4.905,
      "mean_token_accuracy": 0.22992416024208068,
      "num_tokens": 31867264.0,
      "step": 13900
    },
    {
      "entropy": 5.2324436664581295,
      "epoch": 1.3357348703170029,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048282628929843097,
      "loss": 4.7977,
      "mean_token_accuracy": 0.2360186144709587,
      "num_tokens": 31878908.0,
      "step": 13905
    },
    {
      "entropy": 5.267749834060669,
      "epoch": 1.3362151777137368,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048281315000617996,
      "loss": 4.8786,
      "mean_token_accuracy": 0.22786884158849716,
      "num_tokens": 31890299.0,
      "step": 13910
    },
    {
      "entropy": 5.2349811553955075,
      "epoch": 1.3366954851104706,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004828000058891248,
      "loss": 4.8472,
      "mean_token_accuracy": 0.23382498621940612,
      "num_tokens": 31901279.0,
      "step": 13915
    },
    {
      "entropy": 5.294612693786621,
      "epoch": 1.3371757925072045,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004827868569475706,
      "loss": 4.8611,
      "mean_token_accuracy": 0.23315293192863465,
      "num_tokens": 31912851.0,
      "step": 13920
    },
    {
      "entropy": 5.246100425720215,
      "epoch": 1.3376560999039384,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048277370318182243,
      "loss": 4.8296,
      "mean_token_accuracy": 0.23309915661811828,
      "num_tokens": 31923479.0,
      "step": 13925
    },
    {
      "entropy": 5.2802966117858885,
      "epoch": 1.3381364073006723,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048276054459218596,
      "loss": 4.9483,
      "mean_token_accuracy": 0.22937109768390657,
      "num_tokens": 31935035.0,
      "step": 13930
    },
    {
      "entropy": 5.321371030807495,
      "epoch": 1.3386167146974064,
      "grad_norm": 1.0,
      "learning_rate": 0.00048274738117896643,
      "loss": 4.9326,
      "mean_token_accuracy": 0.22418441623449326,
      "num_tokens": 31946140.0,
      "step": 13935
    },
    {
      "entropy": 5.287405681610108,
      "epoch": 1.3390970220941403,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048273421294246966,
      "loss": 4.8418,
      "mean_token_accuracy": 0.2242326706647873,
      "num_tokens": 31957315.0,
      "step": 13940
    },
    {
      "entropy": 5.245630073547363,
      "epoch": 1.3395773294908742,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048272103988300134,
      "loss": 4.8011,
      "mean_token_accuracy": 0.23547552824020385,
      "num_tokens": 31967773.0,
      "step": 13945
    },
    {
      "entropy": 5.323209190368653,
      "epoch": 1.340057636887608,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004827078620008672,
      "loss": 4.9778,
      "mean_token_accuracy": 0.21868592947721482,
      "num_tokens": 31980112.0,
      "step": 13950
    },
    {
      "entropy": 5.2626134872436525,
      "epoch": 1.340537944284342,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048269467929637337,
      "loss": 4.7726,
      "mean_token_accuracy": 0.24294343888759612,
      "num_tokens": 31991655.0,
      "step": 13955
    },
    {
      "entropy": 5.2057271003723145,
      "epoch": 1.341018251681076,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048268149176982576,
      "loss": 4.8131,
      "mean_token_accuracy": 0.23746686726808547,
      "num_tokens": 32003551.0,
      "step": 13960
    },
    {
      "entropy": 5.167764616012573,
      "epoch": 1.34149855907781,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048266829942153055,
      "loss": 4.7511,
      "mean_token_accuracy": 0.2367914229631424,
      "num_tokens": 32015096.0,
      "step": 13965
    },
    {
      "entropy": 5.255722188949585,
      "epoch": 1.3419788664745438,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048265510225179413,
      "loss": 4.8992,
      "mean_token_accuracy": 0.22987070828676223,
      "num_tokens": 32026219.0,
      "step": 13970
    },
    {
      "entropy": 5.287816953659058,
      "epoch": 1.3424591738712777,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004826419002609229,
      "loss": 4.9145,
      "mean_token_accuracy": 0.22713105827569963,
      "num_tokens": 32037301.0,
      "step": 13975
    },
    {
      "entropy": 5.235630893707276,
      "epoch": 1.3429394812680115,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048262869344922326,
      "loss": 4.7896,
      "mean_token_accuracy": 0.23497015237808228,
      "num_tokens": 32048707.0,
      "step": 13980
    },
    {
      "entropy": 5.351365852355957,
      "epoch": 1.3434197886647454,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048261548181700186,
      "loss": 4.9532,
      "mean_token_accuracy": 0.21933018714189528,
      "num_tokens": 32061008.0,
      "step": 13985
    },
    {
      "entropy": 5.212152624130249,
      "epoch": 1.3439000960614793,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004826022653645655,
      "loss": 4.7487,
      "mean_token_accuracy": 0.24278560876846314,
      "num_tokens": 32071360.0,
      "step": 13990
    },
    {
      "entropy": 5.204442548751831,
      "epoch": 1.3443804034582132,
      "grad_norm": 0.9375,
      "learning_rate": 0.000482589044092221,
      "loss": 4.8498,
      "mean_token_accuracy": 0.22874338030815125,
      "num_tokens": 32083441.0,
      "step": 13995
    },
    {
      "entropy": 5.238444375991821,
      "epoch": 1.344860710854947,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048257581800027527,
      "loss": 4.865,
      "mean_token_accuracy": 0.22913867235183716,
      "num_tokens": 32094773.0,
      "step": 14000
    },
    {
      "entropy": 5.233198881149292,
      "epoch": 1.345341018251681,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004825625870890354,
      "loss": 4.8569,
      "mean_token_accuracy": 0.23018885850906373,
      "num_tokens": 32107076.0,
      "step": 14005
    },
    {
      "entropy": 5.410954093933105,
      "epoch": 1.345821325648415,
      "grad_norm": 1.0,
      "learning_rate": 0.0004825493513588086,
      "loss": 5.0057,
      "mean_token_accuracy": 0.21275368481874465,
      "num_tokens": 32118781.0,
      "step": 14010
    },
    {
      "entropy": 5.254666900634765,
      "epoch": 1.346301633045149,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048253611080990226,
      "loss": 4.8052,
      "mean_token_accuracy": 0.23767612278461456,
      "num_tokens": 32130181.0,
      "step": 14015
    },
    {
      "entropy": 5.277146339416504,
      "epoch": 1.3467819404418828,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004825228654426236,
      "loss": 4.8421,
      "mean_token_accuracy": 0.23379273861646652,
      "num_tokens": 32142576.0,
      "step": 14020
    },
    {
      "entropy": 5.269073390960694,
      "epoch": 1.3472622478386167,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004825096152572803,
      "loss": 4.8355,
      "mean_token_accuracy": 0.24063177704811095,
      "num_tokens": 32152413.0,
      "step": 14025
    },
    {
      "entropy": 5.224738454818725,
      "epoch": 1.3477425552353506,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048249636025417974,
      "loss": 4.8619,
      "mean_token_accuracy": 0.2243833974003792,
      "num_tokens": 32163539.0,
      "step": 14030
    },
    {
      "entropy": 5.245755434036255,
      "epoch": 1.3482228626320845,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048248310043362997,
      "loss": 4.8292,
      "mean_token_accuracy": 0.2268417149782181,
      "num_tokens": 32174662.0,
      "step": 14035
    },
    {
      "entropy": 5.208165168762207,
      "epoch": 1.3487031700288186,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004824698357959386,
      "loss": 4.8465,
      "mean_token_accuracy": 0.23347193449735643,
      "num_tokens": 32185950.0,
      "step": 14040
    },
    {
      "entropy": 5.332347440719604,
      "epoch": 1.3491834774255524,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048245656634141385,
      "loss": 4.8979,
      "mean_token_accuracy": 0.22907765656709672,
      "num_tokens": 32198076.0,
      "step": 14045
    },
    {
      "entropy": 5.307613801956177,
      "epoch": 1.3496637848222863,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048244329207036354,
      "loss": 4.8885,
      "mean_token_accuracy": 0.22530339658260345,
      "num_tokens": 32208739.0,
      "step": 14050
    },
    {
      "entropy": 5.3410704135894775,
      "epoch": 1.3501440922190202,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048243001298309604,
      "loss": 5.0471,
      "mean_token_accuracy": 0.21779277175664902,
      "num_tokens": 32221494.0,
      "step": 14055
    },
    {
      "entropy": 5.250935983657837,
      "epoch": 1.350624399615754,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048241672907991954,
      "loss": 4.8309,
      "mean_token_accuracy": 0.2291671022772789,
      "num_tokens": 32232991.0,
      "step": 14060
    },
    {
      "entropy": 5.294940996170044,
      "epoch": 1.351104707012488,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004824034403611424,
      "loss": 4.894,
      "mean_token_accuracy": 0.22370037138462068,
      "num_tokens": 32244022.0,
      "step": 14065
    },
    {
      "entropy": 5.243889427185058,
      "epoch": 1.3515850144092219,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004823901468270733,
      "loss": 4.8549,
      "mean_token_accuracy": 0.22475606501102446,
      "num_tokens": 32255779.0,
      "step": 14070
    },
    {
      "entropy": 5.22629222869873,
      "epoch": 1.3520653218059557,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004823768484780209,
      "loss": 4.7957,
      "mean_token_accuracy": 0.2338210642337799,
      "num_tokens": 32267505.0,
      "step": 14075
    },
    {
      "entropy": 5.250562381744385,
      "epoch": 1.3525456292026896,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048236354531429375,
      "loss": 4.8918,
      "mean_token_accuracy": 0.23539865016937256,
      "num_tokens": 32278675.0,
      "step": 14080
    },
    {
      "entropy": 5.26241102218628,
      "epoch": 1.3530259365994235,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004823502373362009,
      "loss": 4.8498,
      "mean_token_accuracy": 0.2336789295077324,
      "num_tokens": 32289372.0,
      "step": 14085
    },
    {
      "entropy": 5.273185682296753,
      "epoch": 1.3535062439961576,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004823369245440512,
      "loss": 4.8336,
      "mean_token_accuracy": 0.23090527057647706,
      "num_tokens": 32301281.0,
      "step": 14090
    },
    {
      "entropy": 5.30419921875,
      "epoch": 1.3539865513928915,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048232360693815387,
      "loss": 4.8668,
      "mean_token_accuracy": 0.23849944323301314,
      "num_tokens": 32312204.0,
      "step": 14095
    },
    {
      "entropy": 5.255583238601685,
      "epoch": 1.3544668587896254,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048231028451881786,
      "loss": 4.8431,
      "mean_token_accuracy": 0.2269610807299614,
      "num_tokens": 32323607.0,
      "step": 14100
    },
    {
      "entropy": 5.290121984481812,
      "epoch": 1.3549471661863592,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004822969572863527,
      "loss": 4.8794,
      "mean_token_accuracy": 0.23749024718999862,
      "num_tokens": 32334864.0,
      "step": 14105
    },
    {
      "entropy": 5.262269401550293,
      "epoch": 1.3554274735830931,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048228362524106776,
      "loss": 4.9045,
      "mean_token_accuracy": 0.22222439646720887,
      "num_tokens": 32345664.0,
      "step": 14110
    },
    {
      "entropy": 5.268278884887695,
      "epoch": 1.3559077809798272,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048227028838327253,
      "loss": 4.8453,
      "mean_token_accuracy": 0.23738322854042054,
      "num_tokens": 32355727.0,
      "step": 14115
    },
    {
      "entropy": 5.248799514770508,
      "epoch": 1.356388088376561,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048225694671327665,
      "loss": 4.7319,
      "mean_token_accuracy": 0.24155064970254897,
      "num_tokens": 32367132.0,
      "step": 14120
    },
    {
      "entropy": 5.242038106918335,
      "epoch": 1.356868395773295,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004822436002313899,
      "loss": 4.7517,
      "mean_token_accuracy": 0.2403169572353363,
      "num_tokens": 32377599.0,
      "step": 14125
    },
    {
      "entropy": 5.31554946899414,
      "epoch": 1.3573487031700289,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004822302489379222,
      "loss": 4.9315,
      "mean_token_accuracy": 0.21961532533168793,
      "num_tokens": 32389659.0,
      "step": 14130
    },
    {
      "entropy": 5.268999290466309,
      "epoch": 1.3578290105667628,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048221689283318335,
      "loss": 4.8344,
      "mean_token_accuracy": 0.23346930146217346,
      "num_tokens": 32400573.0,
      "step": 14135
    },
    {
      "entropy": 5.301863527297973,
      "epoch": 1.3583093179634966,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004822035319174837,
      "loss": 4.8641,
      "mean_token_accuracy": 0.22924353927373886,
      "num_tokens": 32411009.0,
      "step": 14140
    },
    {
      "entropy": 5.213248825073242,
      "epoch": 1.3587896253602305,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004821901661911332,
      "loss": 4.8617,
      "mean_token_accuracy": 0.23757787942886352,
      "num_tokens": 32422489.0,
      "step": 14145
    },
    {
      "entropy": 5.202754211425781,
      "epoch": 1.3592699327569644,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004821767956544423,
      "loss": 4.7303,
      "mean_token_accuracy": 0.23667351752519608,
      "num_tokens": 32432613.0,
      "step": 14150
    },
    {
      "entropy": 5.218092775344848,
      "epoch": 1.3597502401536983,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004821634203077214,
      "loss": 4.8461,
      "mean_token_accuracy": 0.23408012241125106,
      "num_tokens": 32444556.0,
      "step": 14155
    },
    {
      "entropy": 5.298800468444824,
      "epoch": 1.3602305475504322,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000482150040151281,
      "loss": 4.8173,
      "mean_token_accuracy": 0.23423594683408738,
      "num_tokens": 32455143.0,
      "step": 14160
    },
    {
      "entropy": 5.227648210525513,
      "epoch": 1.3607108549471663,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004821366551854318,
      "loss": 4.8732,
      "mean_token_accuracy": 0.23143816292285918,
      "num_tokens": 32466966.0,
      "step": 14165
    },
    {
      "entropy": 5.177139139175415,
      "epoch": 1.3611911623439001,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004821232654104845,
      "loss": 4.7707,
      "mean_token_accuracy": 0.2374436303973198,
      "num_tokens": 32479435.0,
      "step": 14170
    },
    {
      "entropy": 5.22120885848999,
      "epoch": 1.361671469740634,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048210987082675005,
      "loss": 4.8595,
      "mean_token_accuracy": 0.22958492785692214,
      "num_tokens": 32490923.0,
      "step": 14175
    },
    {
      "entropy": 5.349956083297729,
      "epoch": 1.362151777137368,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048209647143453946,
      "loss": 4.9029,
      "mean_token_accuracy": 0.23199021816253662,
      "num_tokens": 32500939.0,
      "step": 14180
    },
    {
      "entropy": 5.183615589141846,
      "epoch": 1.3626320845341018,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048208306723416356,
      "loss": 4.7773,
      "mean_token_accuracy": 0.23598341047763824,
      "num_tokens": 32513260.0,
      "step": 14185
    },
    {
      "entropy": 5.331999635696411,
      "epoch": 1.3631123919308357,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004820696582259339,
      "loss": 4.9214,
      "mean_token_accuracy": 0.22551199793815613,
      "num_tokens": 32524813.0,
      "step": 14190
    },
    {
      "entropy": 5.298294448852539,
      "epoch": 1.3635926993275698,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004820562444101616,
      "loss": 4.8787,
      "mean_token_accuracy": 0.2262905940413475,
      "num_tokens": 32536904.0,
      "step": 14195
    },
    {
      "entropy": 5.195955610275268,
      "epoch": 1.3640730067243036,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004820428257871581,
      "loss": 4.7836,
      "mean_token_accuracy": 0.2348126217722893,
      "num_tokens": 32547419.0,
      "step": 14200
    },
    {
      "entropy": 5.194369888305664,
      "epoch": 1.3645533141210375,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004820294023572351,
      "loss": 4.7171,
      "mean_token_accuracy": 0.23784016966819763,
      "num_tokens": 32558644.0,
      "step": 14205
    },
    {
      "entropy": 5.210418319702148,
      "epoch": 1.3650336215177714,
      "grad_norm": 0.890625,
      "learning_rate": 0.000482015974120704,
      "loss": 4.798,
      "mean_token_accuracy": 0.233004891872406,
      "num_tokens": 32569629.0,
      "step": 14210
    },
    {
      "entropy": 5.206024503707885,
      "epoch": 1.3655139289145053,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048200254107787677,
      "loss": 4.8799,
      "mean_token_accuracy": 0.2323005348443985,
      "num_tokens": 32581850.0,
      "step": 14215
    },
    {
      "entropy": 5.366559457778931,
      "epoch": 1.3659942363112392,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048198910322906516,
      "loss": 4.9467,
      "mean_token_accuracy": 0.22264957875013353,
      "num_tokens": 32592478.0,
      "step": 14220
    },
    {
      "entropy": 5.393599939346314,
      "epoch": 1.366474543707973,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048197566057458125,
      "loss": 5.0389,
      "mean_token_accuracy": 0.21567141860723496,
      "num_tokens": 32603099.0,
      "step": 14225
    },
    {
      "entropy": 5.33137879371643,
      "epoch": 1.366954851104707,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000481962213114737,
      "loss": 4.9361,
      "mean_token_accuracy": 0.2191356047987938,
      "num_tokens": 32613415.0,
      "step": 14230
    },
    {
      "entropy": 5.267699384689331,
      "epoch": 1.3674351585014408,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004819487608498448,
      "loss": 4.8692,
      "mean_token_accuracy": 0.23249793648719788,
      "num_tokens": 32625348.0,
      "step": 14235
    },
    {
      "entropy": 5.226198482513428,
      "epoch": 1.3679154658981747,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048193530378021687,
      "loss": 4.7967,
      "mean_token_accuracy": 0.23113684207201005,
      "num_tokens": 32636878.0,
      "step": 14240
    },
    {
      "entropy": 5.299793434143067,
      "epoch": 1.3683957732949088,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048192184190616567,
      "loss": 4.9714,
      "mean_token_accuracy": 0.22451183944940567,
      "num_tokens": 32647808.0,
      "step": 14245
    },
    {
      "entropy": 5.47935152053833,
      "epoch": 1.3688760806916427,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004819083752280037,
      "loss": 5.0449,
      "mean_token_accuracy": 0.2145488291978836,
      "num_tokens": 32659425.0,
      "step": 14250
    },
    {
      "entropy": 5.279067707061768,
      "epoch": 1.3693563880883766,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048189490374604373,
      "loss": 4.8786,
      "mean_token_accuracy": 0.22696252018213273,
      "num_tokens": 32671797.0,
      "step": 14255
    },
    {
      "entropy": 5.2361222267150875,
      "epoch": 1.3698366954851104,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004818814274605983,
      "loss": 4.8584,
      "mean_token_accuracy": 0.22831996232271196,
      "num_tokens": 32683506.0,
      "step": 14260
    },
    {
      "entropy": 5.308003759384155,
      "epoch": 1.3703170028818443,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004818679463719805,
      "loss": 4.9597,
      "mean_token_accuracy": 0.21624817997217177,
      "num_tokens": 32695047.0,
      "step": 14265
    },
    {
      "entropy": 5.251731920242309,
      "epoch": 1.3707973102785782,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004818544604805033,
      "loss": 4.8131,
      "mean_token_accuracy": 0.23654665350914,
      "num_tokens": 32705433.0,
      "step": 14270
    },
    {
      "entropy": 5.273219585418701,
      "epoch": 1.3712776176753123,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004818409697864798,
      "loss": 4.9439,
      "mean_token_accuracy": 0.2227967768907547,
      "num_tokens": 32717315.0,
      "step": 14275
    },
    {
      "entropy": 5.248053073883057,
      "epoch": 1.3717579250720462,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048182747429022303,
      "loss": 4.8137,
      "mean_token_accuracy": 0.23951840549707412,
      "num_tokens": 32728817.0,
      "step": 14280
    },
    {
      "entropy": 5.293307781219482,
      "epoch": 1.37223823246878,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004818139739920465,
      "loss": 4.9392,
      "mean_token_accuracy": 0.22159569710493088,
      "num_tokens": 32741030.0,
      "step": 14285
    },
    {
      "entropy": 5.335179233551026,
      "epoch": 1.372718539865514,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004818004688922637,
      "loss": 4.9843,
      "mean_token_accuracy": 0.21638571470975876,
      "num_tokens": 32752212.0,
      "step": 14290
    },
    {
      "entropy": 5.262773227691651,
      "epoch": 1.3731988472622478,
      "grad_norm": 0.984375,
      "learning_rate": 0.000481786958991188,
      "loss": 4.8669,
      "mean_token_accuracy": 0.22861707657575608,
      "num_tokens": 32764581.0,
      "step": 14295
    },
    {
      "entropy": 5.275610685348511,
      "epoch": 1.3736791546589817,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048177344428913316,
      "loss": 4.8801,
      "mean_token_accuracy": 0.22652165740728378,
      "num_tokens": 32775992.0,
      "step": 14300
    },
    {
      "entropy": 5.306463289260864,
      "epoch": 1.3741594620557156,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048175992478641293,
      "loss": 4.9193,
      "mean_token_accuracy": 0.2293963074684143,
      "num_tokens": 32787164.0,
      "step": 14305
    },
    {
      "entropy": 5.209454011917114,
      "epoch": 1.3746397694524495,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004817464004833412,
      "loss": 4.8304,
      "mean_token_accuracy": 0.2383353278040886,
      "num_tokens": 32798190.0,
      "step": 14310
    },
    {
      "entropy": 5.2276331901550295,
      "epoch": 1.3751200768491834,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048173287138023204,
      "loss": 4.7891,
      "mean_token_accuracy": 0.23949553370475768,
      "num_tokens": 32809559.0,
      "step": 14315
    },
    {
      "entropy": 5.303402996063232,
      "epoch": 1.3756003842459175,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004817193374773993,
      "loss": 4.8985,
      "mean_token_accuracy": 0.22408312559127808,
      "num_tokens": 32820799.0,
      "step": 14320
    },
    {
      "entropy": 5.261199617385865,
      "epoch": 1.3760806916426513,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048170579877515753,
      "loss": 4.8569,
      "mean_token_accuracy": 0.22669730931520463,
      "num_tokens": 32833971.0,
      "step": 14325
    },
    {
      "entropy": 5.194620370864868,
      "epoch": 1.3765609990393852,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004816922552738209,
      "loss": 4.7966,
      "mean_token_accuracy": 0.2329302966594696,
      "num_tokens": 32845237.0,
      "step": 14330
    },
    {
      "entropy": 5.306460571289063,
      "epoch": 1.377041306436119,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048167870697370373,
      "loss": 4.9154,
      "mean_token_accuracy": 0.2251480296254158,
      "num_tokens": 32857960.0,
      "step": 14335
    },
    {
      "entropy": 5.222204732894897,
      "epoch": 1.377521613832853,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004816651538751207,
      "loss": 4.7876,
      "mean_token_accuracy": 0.23665741831064224,
      "num_tokens": 32868801.0,
      "step": 14340
    },
    {
      "entropy": 5.311174488067627,
      "epoch": 1.3780019212295869,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048165159597838664,
      "loss": 4.8697,
      "mean_token_accuracy": 0.2292311355471611,
      "num_tokens": 32880544.0,
      "step": 14345
    },
    {
      "entropy": 5.383266448974609,
      "epoch": 1.378482228626321,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000481638033283816,
      "loss": 4.9321,
      "mean_token_accuracy": 0.22215671092271805,
      "num_tokens": 32892179.0,
      "step": 14350
    },
    {
      "entropy": 5.366191148757935,
      "epoch": 1.3789625360230549,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048162446579172387,
      "loss": 4.9058,
      "mean_token_accuracy": 0.23384683728218078,
      "num_tokens": 32903854.0,
      "step": 14355
    },
    {
      "entropy": 5.30956654548645,
      "epoch": 1.3794428434197887,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004816108935024252,
      "loss": 4.8955,
      "mean_token_accuracy": 0.2325608640909195,
      "num_tokens": 32914731.0,
      "step": 14360
    },
    {
      "entropy": 5.201639318466187,
      "epoch": 1.3799231508165226,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048159731641623507,
      "loss": 4.7824,
      "mean_token_accuracy": 0.2309311881661415,
      "num_tokens": 32925894.0,
      "step": 14365
    },
    {
      "entropy": 5.223157072067261,
      "epoch": 1.3804034582132565,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004815837345334687,
      "loss": 4.8235,
      "mean_token_accuracy": 0.2395944729447365,
      "num_tokens": 32937463.0,
      "step": 14370
    },
    {
      "entropy": 5.363404655456543,
      "epoch": 1.3808837656099904,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004815701478544415,
      "loss": 4.9191,
      "mean_token_accuracy": 0.22365123331546782,
      "num_tokens": 32949635.0,
      "step": 14375
    },
    {
      "entropy": 5.2645069599151615,
      "epoch": 1.3813640730067243,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048155655637946876,
      "loss": 4.8869,
      "mean_token_accuracy": 0.23108889013528824,
      "num_tokens": 32961043.0,
      "step": 14380
    },
    {
      "entropy": 5.310822916030884,
      "epoch": 1.3818443804034581,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004815429601088662,
      "loss": 4.9496,
      "mean_token_accuracy": 0.22986554801464082,
      "num_tokens": 32972218.0,
      "step": 14385
    },
    {
      "entropy": 5.249128580093384,
      "epoch": 1.382324687800192,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004815293590429494,
      "loss": 4.7534,
      "mean_token_accuracy": 0.2324935168027878,
      "num_tokens": 32983157.0,
      "step": 14390
    },
    {
      "entropy": 5.305128002166748,
      "epoch": 1.382804995196926,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048151575318203417,
      "loss": 4.906,
      "mean_token_accuracy": 0.2300250068306923,
      "num_tokens": 32994524.0,
      "step": 14395
    },
    {
      "entropy": 5.30927267074585,
      "epoch": 1.38328530259366,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048150214252643637,
      "loss": 4.9093,
      "mean_token_accuracy": 0.22820817083120346,
      "num_tokens": 33006477.0,
      "step": 14400
    },
    {
      "entropy": 5.309292507171631,
      "epoch": 1.3837656099903939,
      "grad_norm": 0.875,
      "learning_rate": 0.000481488527076472,
      "loss": 4.8824,
      "mean_token_accuracy": 0.23085667043924332,
      "num_tokens": 33018067.0,
      "step": 14405
    },
    {
      "entropy": 5.328941726684571,
      "epoch": 1.3842459173871278,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004814749068324571,
      "loss": 4.9195,
      "mean_token_accuracy": 0.2299826294183731,
      "num_tokens": 33031927.0,
      "step": 14410
    },
    {
      "entropy": 5.293738603591919,
      "epoch": 1.3847262247838616,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048146128179470804,
      "loss": 4.9301,
      "mean_token_accuracy": 0.22454020828008653,
      "num_tokens": 33042516.0,
      "step": 14415
    },
    {
      "entropy": 5.271289205551147,
      "epoch": 1.3852065321805955,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004814476519635411,
      "loss": 4.7925,
      "mean_token_accuracy": 0.23073556572198867,
      "num_tokens": 33054931.0,
      "step": 14420
    },
    {
      "entropy": 5.369097375869751,
      "epoch": 1.3856868395773294,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048143401733927274,
      "loss": 4.9589,
      "mean_token_accuracy": 0.21802951097488404,
      "num_tokens": 33067333.0,
      "step": 14425
    },
    {
      "entropy": 5.284532308578491,
      "epoch": 1.3861671469740635,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048142037792221943,
      "loss": 4.8372,
      "mean_token_accuracy": 0.2308535248041153,
      "num_tokens": 33078469.0,
      "step": 14430
    },
    {
      "entropy": 5.311257028579712,
      "epoch": 1.3866474543707974,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004814067337126978,
      "loss": 4.8551,
      "mean_token_accuracy": 0.22540211528539658,
      "num_tokens": 33089993.0,
      "step": 14435
    },
    {
      "entropy": 5.2838396549224855,
      "epoch": 1.3871277617675313,
      "grad_norm": 1.0,
      "learning_rate": 0.0004813930847110248,
      "loss": 4.9039,
      "mean_token_accuracy": 0.2232636496424675,
      "num_tokens": 33101837.0,
      "step": 14440
    },
    {
      "entropy": 5.290772008895874,
      "epoch": 1.3876080691642652,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004813794309175173,
      "loss": 4.8542,
      "mean_token_accuracy": 0.2303632378578186,
      "num_tokens": 33113036.0,
      "step": 14445
    },
    {
      "entropy": 5.243567419052124,
      "epoch": 1.388088376560999,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048136577233249205,
      "loss": 4.8277,
      "mean_token_accuracy": 0.2345754012465477,
      "num_tokens": 33124826.0,
      "step": 14450
    },
    {
      "entropy": 5.186099863052368,
      "epoch": 1.388568683957733,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004813521089562666,
      "loss": 4.6892,
      "mean_token_accuracy": 0.2406955912709236,
      "num_tokens": 33135878.0,
      "step": 14455
    },
    {
      "entropy": 5.235904216766357,
      "epoch": 1.3890489913544668,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004813384407891577,
      "loss": 4.8685,
      "mean_token_accuracy": 0.23606669157743454,
      "num_tokens": 33146938.0,
      "step": 14460
    },
    {
      "entropy": 5.290914106369018,
      "epoch": 1.3895292987512007,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000481324767831483,
      "loss": 4.8433,
      "mean_token_accuracy": 0.22966043800115585,
      "num_tokens": 33159223.0,
      "step": 14465
    },
    {
      "entropy": 5.28332290649414,
      "epoch": 1.3900096061479346,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004813110900835598,
      "loss": 4.9689,
      "mean_token_accuracy": 0.2238354817032814,
      "num_tokens": 33170235.0,
      "step": 14470
    },
    {
      "entropy": 5.238030576705933,
      "epoch": 1.3904899135446687,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004812974075457058,
      "loss": 4.7911,
      "mean_token_accuracy": 0.23845456391572953,
      "num_tokens": 33181193.0,
      "step": 14475
    },
    {
      "entropy": 5.287032747268677,
      "epoch": 1.3909702209414025,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048128372021823845,
      "loss": 4.8398,
      "mean_token_accuracy": 0.22615164518356323,
      "num_tokens": 33192505.0,
      "step": 14480
    },
    {
      "entropy": 5.276429510116577,
      "epoch": 1.3914505283381364,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048127002810147574,
      "loss": 4.8319,
      "mean_token_accuracy": 0.23183957785367965,
      "num_tokens": 33204113.0,
      "step": 14485
    },
    {
      "entropy": 5.4057153224945065,
      "epoch": 1.3919308357348703,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004812563311957355,
      "loss": 5.0438,
      "mean_token_accuracy": 0.21650493890047073,
      "num_tokens": 33216175.0,
      "step": 14490
    },
    {
      "entropy": 5.334194707870483,
      "epoch": 1.3924111431316042,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004812426295013356,
      "loss": 4.895,
      "mean_token_accuracy": 0.22490763962268828,
      "num_tokens": 33226916.0,
      "step": 14495
    },
    {
      "entropy": 5.308570194244385,
      "epoch": 1.392891450528338,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048122892301859433,
      "loss": 4.9282,
      "mean_token_accuracy": 0.2309838518500328,
      "num_tokens": 33238020.0,
      "step": 14500
    },
    {
      "entropy": 5.2876951694488525,
      "epoch": 1.3933717579250722,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048121521174782983,
      "loss": 4.786,
      "mean_token_accuracy": 0.23338018506765365,
      "num_tokens": 33248343.0,
      "step": 14505
    },
    {
      "entropy": 5.358735752105713,
      "epoch": 1.393852065321806,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048120149568936044,
      "loss": 4.9082,
      "mean_token_accuracy": 0.2233003705739975,
      "num_tokens": 33259206.0,
      "step": 14510
    },
    {
      "entropy": 5.248608922958374,
      "epoch": 1.39433237271854,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004811877748435046,
      "loss": 4.9301,
      "mean_token_accuracy": 0.2223157688975334,
      "num_tokens": 33271020.0,
      "step": 14515
    },
    {
      "entropy": 5.250840330123902,
      "epoch": 1.3948126801152738,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004811740492105809,
      "loss": 4.8032,
      "mean_token_accuracy": 0.23761081993579863,
      "num_tokens": 33282603.0,
      "step": 14520
    },
    {
      "entropy": 5.369240045547485,
      "epoch": 1.3952929875120077,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000481160318790908,
      "loss": 4.9145,
      "mean_token_accuracy": 0.2313394919037819,
      "num_tokens": 33292877.0,
      "step": 14525
    },
    {
      "entropy": 5.37590389251709,
      "epoch": 1.3957732949087416,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048114658358480467,
      "loss": 5.1003,
      "mean_token_accuracy": 0.22240544259548187,
      "num_tokens": 33304835.0,
      "step": 14530
    },
    {
      "entropy": 5.351807260513306,
      "epoch": 1.3962536023054755,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048113284359258977,
      "loss": 4.9051,
      "mean_token_accuracy": 0.2280913546681404,
      "num_tokens": 33315028.0,
      "step": 14535
    },
    {
      "entropy": 5.240409183502197,
      "epoch": 1.3967339097022093,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048111909881458234,
      "loss": 4.8345,
      "mean_token_accuracy": 0.22851356118917465,
      "num_tokens": 33325321.0,
      "step": 14540
    },
    {
      "entropy": 5.288677310943603,
      "epoch": 1.3972142170989432,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048110534925110146,
      "loss": 4.9152,
      "mean_token_accuracy": 0.2316841810941696,
      "num_tokens": 33337677.0,
      "step": 14545
    },
    {
      "entropy": 5.2590460777282715,
      "epoch": 1.397694524495677,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004810915949024664,
      "loss": 4.8386,
      "mean_token_accuracy": 0.22854605466127395,
      "num_tokens": 33347613.0,
      "step": 14550
    },
    {
      "entropy": 5.265868139266968,
      "epoch": 1.3981748318924112,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004810778357689965,
      "loss": 4.8372,
      "mean_token_accuracy": 0.22973479926586152,
      "num_tokens": 33358912.0,
      "step": 14555
    },
    {
      "entropy": 5.321146392822266,
      "epoch": 1.398655139289145,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048106407185101116,
      "loss": 4.978,
      "mean_token_accuracy": 0.2258977010846138,
      "num_tokens": 33371635.0,
      "step": 14560
    },
    {
      "entropy": 5.277064323425293,
      "epoch": 1.399135446685879,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048105030314883,
      "loss": 4.7449,
      "mean_token_accuracy": 0.2476913496851921,
      "num_tokens": 33383521.0,
      "step": 14565
    },
    {
      "entropy": 5.24110689163208,
      "epoch": 1.3996157540826129,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004810365296627725,
      "loss": 4.8371,
      "mean_token_accuracy": 0.23030897378921508,
      "num_tokens": 33395297.0,
      "step": 14570
    },
    {
      "entropy": 5.161647653579712,
      "epoch": 1.4000960614793467,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004810227513931587,
      "loss": 4.7304,
      "mean_token_accuracy": 0.24421052634716034,
      "num_tokens": 33407168.0,
      "step": 14575
    },
    {
      "entropy": 5.224277591705322,
      "epoch": 1.4005763688760806,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004810089683403084,
      "loss": 4.7827,
      "mean_token_accuracy": 0.23269509077072142,
      "num_tokens": 33418568.0,
      "step": 14580
    },
    {
      "entropy": 5.246663236618042,
      "epoch": 1.4010566762728147,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004809951805045415,
      "loss": 4.8808,
      "mean_token_accuracy": 0.22451444566249848,
      "num_tokens": 33428822.0,
      "step": 14585
    },
    {
      "entropy": 5.213614797592163,
      "epoch": 1.4015369836695486,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048098138788617815,
      "loss": 4.8107,
      "mean_token_accuracy": 0.2407457172870636,
      "num_tokens": 33441547.0,
      "step": 14590
    },
    {
      "entropy": 5.322092151641845,
      "epoch": 1.4020172910662825,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004809675904855387,
      "loss": 4.8731,
      "mean_token_accuracy": 0.23454676419496537,
      "num_tokens": 33452445.0,
      "step": 14595
    },
    {
      "entropy": 5.218497180938721,
      "epoch": 1.4024975984630164,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048095378830294343,
      "loss": 4.8618,
      "mean_token_accuracy": 0.2360140338540077,
      "num_tokens": 33463845.0,
      "step": 14600
    },
    {
      "entropy": 5.2939229011535645,
      "epoch": 1.4029779058597502,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048093998133871276,
      "loss": 4.8415,
      "mean_token_accuracy": 0.23074757009744645,
      "num_tokens": 33474950.0,
      "step": 14605
    },
    {
      "entropy": 5.2231122016906735,
      "epoch": 1.4034582132564841,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004809261695931671,
      "loss": 4.8021,
      "mean_token_accuracy": 0.229495570063591,
      "num_tokens": 33486225.0,
      "step": 14610
    },
    {
      "entropy": 5.276988935470581,
      "epoch": 1.403938520653218,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004809123530666273,
      "loss": 4.8595,
      "mean_token_accuracy": 0.2307976648211479,
      "num_tokens": 33498125.0,
      "step": 14615
    },
    {
      "entropy": 5.201292991638184,
      "epoch": 1.4044188280499519,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004808985317594142,
      "loss": 4.7481,
      "mean_token_accuracy": 0.23890497237443925,
      "num_tokens": 33508731.0,
      "step": 14620
    },
    {
      "entropy": 5.2356650829315186,
      "epoch": 1.4048991354466858,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048088470567184854,
      "loss": 4.7717,
      "mean_token_accuracy": 0.24011076539754866,
      "num_tokens": 33520356.0,
      "step": 14625
    },
    {
      "entropy": 5.2776771068573,
      "epoch": 1.4053794428434199,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048087087480425133,
      "loss": 4.912,
      "mean_token_accuracy": 0.23246455639600755,
      "num_tokens": 33533179.0,
      "step": 14630
    },
    {
      "entropy": 5.146895551681519,
      "epoch": 1.4058597502401537,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004808570391569437,
      "loss": 4.75,
      "mean_token_accuracy": 0.2353711649775505,
      "num_tokens": 33544700.0,
      "step": 14635
    },
    {
      "entropy": 5.321113157272339,
      "epoch": 1.4063400576368876,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048084319873024694,
      "loss": 4.9817,
      "mean_token_accuracy": 0.22343434244394303,
      "num_tokens": 33558781.0,
      "step": 14640
    },
    {
      "entropy": 5.288047742843628,
      "epoch": 1.4068203650336215,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004808293535244823,
      "loss": 4.8419,
      "mean_token_accuracy": 0.23676956593990325,
      "num_tokens": 33569508.0,
      "step": 14645
    },
    {
      "entropy": 5.308267450332641,
      "epoch": 1.4073006724303554,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004808155035399712,
      "loss": 4.8065,
      "mean_token_accuracy": 0.2326262563467026,
      "num_tokens": 33580977.0,
      "step": 14650
    },
    {
      "entropy": 5.253983354568481,
      "epoch": 1.4077809798270893,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004808016487770354,
      "loss": 4.8901,
      "mean_token_accuracy": 0.23006363213062286,
      "num_tokens": 33593757.0,
      "step": 14655
    },
    {
      "entropy": 5.347203779220581,
      "epoch": 1.4082612872238234,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048078778923599637,
      "loss": 4.9995,
      "mean_token_accuracy": 0.22084451764822005,
      "num_tokens": 33604903.0,
      "step": 14660
    },
    {
      "entropy": 5.269168329238892,
      "epoch": 1.4087415946205573,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048077392491717593,
      "loss": 4.8174,
      "mean_token_accuracy": 0.2396525427699089,
      "num_tokens": 33616976.0,
      "step": 14665
    },
    {
      "entropy": 5.295386886596679,
      "epoch": 1.4092219020172911,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048076005582089597,
      "loss": 4.8481,
      "mean_token_accuracy": 0.23314570784568786,
      "num_tokens": 33627654.0,
      "step": 14670
    },
    {
      "entropy": 5.223801231384277,
      "epoch": 1.409702209414025,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048074618194747845,
      "loss": 4.8027,
      "mean_token_accuracy": 0.23672956377267837,
      "num_tokens": 33638377.0,
      "step": 14675
    },
    {
      "entropy": 5.256158542633057,
      "epoch": 1.410182516810759,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004807323032972456,
      "loss": 4.8273,
      "mean_token_accuracy": 0.22545233219861985,
      "num_tokens": 33650959.0,
      "step": 14680
    },
    {
      "entropy": 5.355327177047729,
      "epoch": 1.4106628242074928,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004807184198705196,
      "loss": 4.9308,
      "mean_token_accuracy": 0.22346847802400588,
      "num_tokens": 33663069.0,
      "step": 14685
    },
    {
      "entropy": 5.253476476669311,
      "epoch": 1.4111431316042267,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004807045316676226,
      "loss": 4.9004,
      "mean_token_accuracy": 0.22809505760669707,
      "num_tokens": 33674868.0,
      "step": 14690
    },
    {
      "entropy": 5.224125576019287,
      "epoch": 1.4116234390009605,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004806906386888773,
      "loss": 4.8405,
      "mean_token_accuracy": 0.23165194392204286,
      "num_tokens": 33685482.0,
      "step": 14695
    },
    {
      "entropy": 5.26125202178955,
      "epoch": 1.4121037463976944,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048067674093460607,
      "loss": 4.823,
      "mean_token_accuracy": 0.23270794451236726,
      "num_tokens": 33697518.0,
      "step": 14700
    },
    {
      "entropy": 5.28419246673584,
      "epoch": 1.4125840537944283,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048066283840513175,
      "loss": 4.8438,
      "mean_token_accuracy": 0.23216162472963334,
      "num_tokens": 33708827.0,
      "step": 14705
    },
    {
      "entropy": 5.253975772857666,
      "epoch": 1.4130643611911624,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004806489311007769,
      "loss": 4.8696,
      "mean_token_accuracy": 0.22728635966777802,
      "num_tokens": 33719452.0,
      "step": 14710
    },
    {
      "entropy": 5.3042981147766115,
      "epoch": 1.4135446685878963,
      "grad_norm": 1.234375,
      "learning_rate": 0.00048063501902186463,
      "loss": 4.9091,
      "mean_token_accuracy": 0.23347591310739518,
      "num_tokens": 33730630.0,
      "step": 14715
    },
    {
      "entropy": 5.233272552490234,
      "epoch": 1.4140249759846302,
      "grad_norm": 1.0,
      "learning_rate": 0.00048062110216871775,
      "loss": 4.8146,
      "mean_token_accuracy": 0.23465735614299774,
      "num_tokens": 33742696.0,
      "step": 14720
    },
    {
      "entropy": 5.2451986312866214,
      "epoch": 1.414505283381364,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048060718054165945,
      "loss": 4.8485,
      "mean_token_accuracy": 0.23154646903276443,
      "num_tokens": 33753818.0,
      "step": 14725
    },
    {
      "entropy": 5.378828048706055,
      "epoch": 1.414985590778098,
      "grad_norm": 0.953125,
      "learning_rate": 0.000480593254141013,
      "loss": 5.0334,
      "mean_token_accuracy": 0.22373906373977662,
      "num_tokens": 33766847.0,
      "step": 14730
    },
    {
      "entropy": 5.29672155380249,
      "epoch": 1.4154658981748318,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048057932296710165,
      "loss": 4.8217,
      "mean_token_accuracy": 0.22844433933496475,
      "num_tokens": 33777070.0,
      "step": 14735
    },
    {
      "entropy": 5.291281747817993,
      "epoch": 1.415946205571566,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004805653870202489,
      "loss": 4.8282,
      "mean_token_accuracy": 0.2350221425294876,
      "num_tokens": 33788826.0,
      "step": 14740
    },
    {
      "entropy": 5.195329236984253,
      "epoch": 1.4164265129682998,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048055144630077825,
      "loss": 4.8009,
      "mean_token_accuracy": 0.2325834274291992,
      "num_tokens": 33801564.0,
      "step": 14745
    },
    {
      "entropy": 5.241344833374024,
      "epoch": 1.4169068203650337,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048053750080901336,
      "loss": 4.8403,
      "mean_token_accuracy": 0.2388267546892166,
      "num_tokens": 33813389.0,
      "step": 14750
    },
    {
      "entropy": 5.265572261810303,
      "epoch": 1.4173871277617676,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048052355054527794,
      "loss": 4.8209,
      "mean_token_accuracy": 0.22890737950801848,
      "num_tokens": 33824585.0,
      "step": 14755
    },
    {
      "entropy": 5.331241083145142,
      "epoch": 1.4178674351585014,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048050959550989606,
      "loss": 4.8985,
      "mean_token_accuracy": 0.2246099054813385,
      "num_tokens": 33837995.0,
      "step": 14760
    },
    {
      "entropy": 5.266156148910523,
      "epoch": 1.4183477425552353,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004804956357031916,
      "loss": 4.8211,
      "mean_token_accuracy": 0.23727236390113832,
      "num_tokens": 33848849.0,
      "step": 14765
    },
    {
      "entropy": 5.232215452194214,
      "epoch": 1.4188280499519692,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048048167112548873,
      "loss": 4.8207,
      "mean_token_accuracy": 0.23983878642320633,
      "num_tokens": 33859711.0,
      "step": 14770
    },
    {
      "entropy": 5.301053524017334,
      "epoch": 1.419308357348703,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048046770177711157,
      "loss": 4.8941,
      "mean_token_accuracy": 0.22355230152606964,
      "num_tokens": 33872613.0,
      "step": 14775
    },
    {
      "entropy": 5.190593576431274,
      "epoch": 1.419788664745437,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004804537276583844,
      "loss": 4.7451,
      "mean_token_accuracy": 0.23809601068496705,
      "num_tokens": 33883759.0,
      "step": 14780
    },
    {
      "entropy": 5.29960150718689,
      "epoch": 1.420268972142171,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004804397487696319,
      "loss": 4.8951,
      "mean_token_accuracy": 0.23068181574344634,
      "num_tokens": 33894448.0,
      "step": 14785
    },
    {
      "entropy": 5.328697156906128,
      "epoch": 1.420749279538905,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004804257651111783,
      "loss": 4.9258,
      "mean_token_accuracy": 0.22691653668880463,
      "num_tokens": 33906976.0,
      "step": 14790
    },
    {
      "entropy": 5.23808012008667,
      "epoch": 1.4212295869356388,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048041177668334853,
      "loss": 4.873,
      "mean_token_accuracy": 0.22597451955080033,
      "num_tokens": 33919384.0,
      "step": 14795
    },
    {
      "entropy": 5.2597955703735355,
      "epoch": 1.4217098943323727,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004803977834864672,
      "loss": 4.7851,
      "mean_token_accuracy": 0.23521350920200348,
      "num_tokens": 33931080.0,
      "step": 14800
    },
    {
      "entropy": 5.258185005187988,
      "epoch": 1.4221902017291066,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048038378552085927,
      "loss": 4.8498,
      "mean_token_accuracy": 0.22891440093517304,
      "num_tokens": 33943021.0,
      "step": 14805
    },
    {
      "entropy": 5.277006959915161,
      "epoch": 1.4226705091258405,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048036978278684974,
      "loss": 4.8629,
      "mean_token_accuracy": 0.22910385131835936,
      "num_tokens": 33953609.0,
      "step": 14810
    },
    {
      "entropy": 5.293630933761596,
      "epoch": 1.4231508165225746,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004803557752847636,
      "loss": 4.9024,
      "mean_token_accuracy": 0.23216885179281235,
      "num_tokens": 33966985.0,
      "step": 14815
    },
    {
      "entropy": 5.291018724441528,
      "epoch": 1.4236311239193085,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048034176301492616,
      "loss": 4.8241,
      "mean_token_accuracy": 0.23419436812400818,
      "num_tokens": 33976536.0,
      "step": 14820
    },
    {
      "entropy": 5.253659629821778,
      "epoch": 1.4241114313160423,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004803277459776628,
      "loss": 4.8335,
      "mean_token_accuracy": 0.23280467838048935,
      "num_tokens": 33988000.0,
      "step": 14825
    },
    {
      "entropy": 5.30514235496521,
      "epoch": 1.4245917387127762,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048031372417329875,
      "loss": 4.9388,
      "mean_token_accuracy": 0.22937967926263808,
      "num_tokens": 33998944.0,
      "step": 14830
    },
    {
      "entropy": 5.331828880310058,
      "epoch": 1.42507204610951,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004802996976021598,
      "loss": 4.9208,
      "mean_token_accuracy": 0.23327389657497405,
      "num_tokens": 34010440.0,
      "step": 14835
    },
    {
      "entropy": 5.3007629871368405,
      "epoch": 1.425552353506244,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048028566626457145,
      "loss": 4.8842,
      "mean_token_accuracy": 0.22803917527198792,
      "num_tokens": 34023222.0,
      "step": 14840
    },
    {
      "entropy": 5.241163635253907,
      "epoch": 1.4260326609029779,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048027163016085947,
      "loss": 4.7625,
      "mean_token_accuracy": 0.2354234278202057,
      "num_tokens": 34035413.0,
      "step": 14845
    },
    {
      "entropy": 5.255193519592285,
      "epoch": 1.4265129682997117,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048025758929134976,
      "loss": 4.8683,
      "mean_token_accuracy": 0.23054903894662857,
      "num_tokens": 34046700.0,
      "step": 14850
    },
    {
      "entropy": 5.235014343261719,
      "epoch": 1.4269932756964456,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004802435436563684,
      "loss": 4.8527,
      "mean_token_accuracy": 0.23582883626222612,
      "num_tokens": 34058039.0,
      "step": 14855
    },
    {
      "entropy": 5.3029368877410885,
      "epoch": 1.4274735830931795,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048022949325624134,
      "loss": 4.8984,
      "mean_token_accuracy": 0.2258405789732933,
      "num_tokens": 34069602.0,
      "step": 14860
    },
    {
      "entropy": 5.337306165695191,
      "epoch": 1.4279538904899136,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048021543809129483,
      "loss": 4.8928,
      "mean_token_accuracy": 0.22030269503593444,
      "num_tokens": 34080514.0,
      "step": 14865
    },
    {
      "entropy": 5.238902282714844,
      "epoch": 1.4284341978866475,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004802013781618552,
      "loss": 4.8227,
      "mean_token_accuracy": 0.23522070944309234,
      "num_tokens": 34091416.0,
      "step": 14870
    },
    {
      "entropy": 5.290758800506592,
      "epoch": 1.4289145052833814,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048018731346824895,
      "loss": 4.8451,
      "mean_token_accuracy": 0.22924076169729232,
      "num_tokens": 34102153.0,
      "step": 14875
    },
    {
      "entropy": 5.338241195678711,
      "epoch": 1.4293948126801153,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004801732440108026,
      "loss": 4.9138,
      "mean_token_accuracy": 0.23116883337497712,
      "num_tokens": 34111917.0,
      "step": 14880
    },
    {
      "entropy": 5.285934114456177,
      "epoch": 1.4298751200768491,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004801591697898427,
      "loss": 4.9015,
      "mean_token_accuracy": 0.22817098051309587,
      "num_tokens": 34124038.0,
      "step": 14885
    },
    {
      "entropy": 5.326124858856201,
      "epoch": 1.430355427473583,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004801450908056961,
      "loss": 4.9476,
      "mean_token_accuracy": 0.22845155149698257,
      "num_tokens": 34134503.0,
      "step": 14890
    },
    {
      "entropy": 5.226578140258789,
      "epoch": 1.4308357348703171,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004801310070586896,
      "loss": 4.764,
      "mean_token_accuracy": 0.23322044163942338,
      "num_tokens": 34146145.0,
      "step": 14895
    },
    {
      "entropy": 5.27427487373352,
      "epoch": 1.431316042267051,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004801169185491503,
      "loss": 4.8323,
      "mean_token_accuracy": 0.22649282664060594,
      "num_tokens": 34156672.0,
      "step": 14900
    },
    {
      "entropy": 5.2918929100036625,
      "epoch": 1.4317963496637849,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048010282527740516,
      "loss": 4.9991,
      "mean_token_accuracy": 0.21704571545124055,
      "num_tokens": 34168817.0,
      "step": 14905
    },
    {
      "entropy": 5.306361341476441,
      "epoch": 1.4322766570605188,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048008872724378146,
      "loss": 4.8316,
      "mean_token_accuracy": 0.23712263107299805,
      "num_tokens": 34179804.0,
      "step": 14910
    },
    {
      "entropy": 5.422490882873535,
      "epoch": 1.4327569644572526,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004800746244486065,
      "loss": 5.0828,
      "mean_token_accuracy": 0.21361998915672303,
      "num_tokens": 34192293.0,
      "step": 14915
    },
    {
      "entropy": 5.3148863315582275,
      "epoch": 1.4332372718539865,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004800605168922077,
      "loss": 4.9252,
      "mean_token_accuracy": 0.22784679234027863,
      "num_tokens": 34203331.0,
      "step": 14920
    },
    {
      "entropy": 5.3051979541778564,
      "epoch": 1.4337175792507204,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048004640457491267,
      "loss": 4.8409,
      "mean_token_accuracy": 0.23164405524730683,
      "num_tokens": 34213554.0,
      "step": 14925
    },
    {
      "entropy": 5.236341285705566,
      "epoch": 1.4341978866474543,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000480032287497049,
      "loss": 4.8002,
      "mean_token_accuracy": 0.2356983333826065,
      "num_tokens": 34225824.0,
      "step": 14930
    },
    {
      "entropy": 5.232591199874878,
      "epoch": 1.4346781940441882,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048001816565894427,
      "loss": 4.8615,
      "mean_token_accuracy": 0.23376472294330597,
      "num_tokens": 34237479.0,
      "step": 14935
    },
    {
      "entropy": 5.315116119384766,
      "epoch": 1.435158501440922,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004800040390609267,
      "loss": 4.927,
      "mean_token_accuracy": 0.2250313028693199,
      "num_tokens": 34248706.0,
      "step": 14940
    },
    {
      "entropy": 5.400471639633179,
      "epoch": 1.4356388088376562,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047998990770332396,
      "loss": 4.9307,
      "mean_token_accuracy": 0.2287221819162369,
      "num_tokens": 34258957.0,
      "step": 14945
    },
    {
      "entropy": 5.268484258651734,
      "epoch": 1.43611911623439,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004799757715864643,
      "loss": 4.89,
      "mean_token_accuracy": 0.22701438516378403,
      "num_tokens": 34270635.0,
      "step": 14950
    },
    {
      "entropy": 5.32386794090271,
      "epoch": 1.436599423631124,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004799616307106759,
      "loss": 4.8357,
      "mean_token_accuracy": 0.23101067692041397,
      "num_tokens": 34282491.0,
      "step": 14955
    },
    {
      "entropy": 5.284918642044067,
      "epoch": 1.4370797310278578,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000479947485076287,
      "loss": 4.7796,
      "mean_token_accuracy": 0.2362533301115036,
      "num_tokens": 34294666.0,
      "step": 14960
    },
    {
      "entropy": 5.258307933807373,
      "epoch": 1.4375600384245917,
      "grad_norm": 1.0,
      "learning_rate": 0.00047993333468362607,
      "loss": 4.7866,
      "mean_token_accuracy": 0.23238448351621627,
      "num_tokens": 34305065.0,
      "step": 14965
    },
    {
      "entropy": 5.2995987892150875,
      "epoch": 1.4380403458213258,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047991917953302173,
      "loss": 4.9231,
      "mean_token_accuracy": 0.22311837375164031,
      "num_tokens": 34315955.0,
      "step": 14970
    },
    {
      "entropy": 5.240520000457764,
      "epoch": 1.4385206532180597,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047990501962480236,
      "loss": 4.7967,
      "mean_token_accuracy": 0.2370201274752617,
      "num_tokens": 34327011.0,
      "step": 14975
    },
    {
      "entropy": 5.226964044570923,
      "epoch": 1.4390009606147935,
      "grad_norm": 0.984375,
      "learning_rate": 0.000479890854959297,
      "loss": 4.8123,
      "mean_token_accuracy": 0.23723654597997665,
      "num_tokens": 34338011.0,
      "step": 14980
    },
    {
      "entropy": 5.242180204391479,
      "epoch": 1.4394812680115274,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004798766855368344,
      "loss": 4.7209,
      "mean_token_accuracy": 0.23904554098844527,
      "num_tokens": 34348957.0,
      "step": 14985
    },
    {
      "entropy": 5.233296871185303,
      "epoch": 1.4399615754082613,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047986251135774343,
      "loss": 4.7946,
      "mean_token_accuracy": 0.2277716413140297,
      "num_tokens": 34359442.0,
      "step": 14990
    },
    {
      "entropy": 5.248393583297729,
      "epoch": 1.4404418828049952,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004798483324223533,
      "loss": 4.8526,
      "mean_token_accuracy": 0.23301923722028733,
      "num_tokens": 34371129.0,
      "step": 14995
    },
    {
      "entropy": 5.2576741695404055,
      "epoch": 1.440922190201729,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004798341487309932,
      "loss": 4.8517,
      "mean_token_accuracy": 0.23518690317869187,
      "num_tokens": 34381565.0,
      "step": 15000
    },
    {
      "epoch": 1.440922190201729,
      "eval_entropy": 5.1062841545599715,
      "eval_loss": 4.916818141937256,
      "eval_mean_token_accuracy": 0.23677928874978404,
      "eval_num_tokens": 34381565.0,
      "eval_runtime": 20.5184,
      "eval_samples_per_second": 1599.3,
      "eval_steps_per_second": 199.919,
      "step": 15000
    },
    {
      "entropy": 5.3089972019195555,
      "epoch": 1.441402497598463,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047981996028399233,
      "loss": 4.7965,
      "mean_token_accuracy": 0.23553545027971268,
      "num_tokens": 34392145.0,
      "step": 15005
    },
    {
      "entropy": 5.269730806350708,
      "epoch": 1.4418828049951968,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004798057670816802,
      "loss": 4.9399,
      "mean_token_accuracy": 0.22369291633367538,
      "num_tokens": 34403193.0,
      "step": 15010
    },
    {
      "entropy": 5.218481969833374,
      "epoch": 1.4423631123919307,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004797915691243863,
      "loss": 4.7951,
      "mean_token_accuracy": 0.22646999955177308,
      "num_tokens": 34414289.0,
      "step": 15015
    },
    {
      "entropy": 5.358546543121338,
      "epoch": 1.4428434197886648,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004797773664124403,
      "loss": 4.926,
      "mean_token_accuracy": 0.22156135141849517,
      "num_tokens": 34424439.0,
      "step": 15020
    },
    {
      "entropy": 5.204573106765747,
      "epoch": 1.4433237271853987,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047976315894617195,
      "loss": 4.7797,
      "mean_token_accuracy": 0.22922777235507966,
      "num_tokens": 34437299.0,
      "step": 15025
    },
    {
      "entropy": 5.241663122177124,
      "epoch": 1.4438040345821326,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000479748946725911,
      "loss": 4.7016,
      "mean_token_accuracy": 0.24454511553049088,
      "num_tokens": 34448397.0,
      "step": 15030
    },
    {
      "entropy": 5.2181273937225345,
      "epoch": 1.4442843419788665,
      "grad_norm": 1.0,
      "learning_rate": 0.0004797347297519875,
      "loss": 4.7759,
      "mean_token_accuracy": 0.23646046072244645,
      "num_tokens": 34459389.0,
      "step": 15035
    },
    {
      "entropy": 5.302580499649048,
      "epoch": 1.4447646493756003,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047972050802473154,
      "loss": 4.8633,
      "mean_token_accuracy": 0.23215881884098052,
      "num_tokens": 34470610.0,
      "step": 15040
    },
    {
      "entropy": 5.205289649963379,
      "epoch": 1.4452449567723342,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004797062815444733,
      "loss": 4.8188,
      "mean_token_accuracy": 0.23764987885951996,
      "num_tokens": 34482100.0,
      "step": 15045
    },
    {
      "entropy": 5.262757921218872,
      "epoch": 1.4457252641690683,
      "grad_norm": 0.953125,
      "learning_rate": 0.000479692050311543,
      "loss": 4.8451,
      "mean_token_accuracy": 0.22255848050117494,
      "num_tokens": 34493086.0,
      "step": 15050
    },
    {
      "entropy": 5.197891283035278,
      "epoch": 1.4462055715658022,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004796778143262711,
      "loss": 4.7682,
      "mean_token_accuracy": 0.24151909053325654,
      "num_tokens": 34505030.0,
      "step": 15055
    },
    {
      "entropy": 5.29539270401001,
      "epoch": 1.446685878962536,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004796635735889882,
      "loss": 4.9369,
      "mean_token_accuracy": 0.22116416692733765,
      "num_tokens": 34516790.0,
      "step": 15060
    },
    {
      "entropy": 5.381803035736084,
      "epoch": 1.44716618635927,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047964932810002476,
      "loss": 4.9345,
      "mean_token_accuracy": 0.22388008534908294,
      "num_tokens": 34530333.0,
      "step": 15065
    },
    {
      "entropy": 5.249238729476929,
      "epoch": 1.4476464937560038,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004796350778597117,
      "loss": 4.8304,
      "mean_token_accuracy": 0.23576582968235016,
      "num_tokens": 34540945.0,
      "step": 15070
    },
    {
      "entropy": 5.340104246139527,
      "epoch": 1.4481268011527377,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004796208228683796,
      "loss": 4.8486,
      "mean_token_accuracy": 0.22635007202625274,
      "num_tokens": 34551449.0,
      "step": 15075
    },
    {
      "entropy": 5.266274261474609,
      "epoch": 1.4486071085494716,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047960656312635977,
      "loss": 4.8827,
      "mean_token_accuracy": 0.2264124259352684,
      "num_tokens": 34562841.0,
      "step": 15080
    },
    {
      "entropy": 5.2730859279632565,
      "epoch": 1.4490874159462055,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004795922986339831,
      "loss": 4.8701,
      "mean_token_accuracy": 0.22762938737869262,
      "num_tokens": 34574408.0,
      "step": 15085
    },
    {
      "entropy": 5.332771062850952,
      "epoch": 1.4495677233429394,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047957802939158057,
      "loss": 4.8828,
      "mean_token_accuracy": 0.23444176018238067,
      "num_tokens": 34586243.0,
      "step": 15090
    },
    {
      "entropy": 5.2862144947052006,
      "epoch": 1.4500480307396733,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004795637553994838,
      "loss": 4.8491,
      "mean_token_accuracy": 0.2338779717683792,
      "num_tokens": 34596631.0,
      "step": 15095
    },
    {
      "entropy": 5.23897271156311,
      "epoch": 1.4505283381364074,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047954947665802404,
      "loss": 4.802,
      "mean_token_accuracy": 0.23816151171922684,
      "num_tokens": 34607952.0,
      "step": 15100
    },
    {
      "entropy": 5.205160522460938,
      "epoch": 1.4510086455331412,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004795351931675329,
      "loss": 4.7479,
      "mean_token_accuracy": 0.2416735291481018,
      "num_tokens": 34618355.0,
      "step": 15105
    },
    {
      "entropy": 5.25368332862854,
      "epoch": 1.4514889529298751,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004795209049283419,
      "loss": 4.8368,
      "mean_token_accuracy": 0.23243466019630432,
      "num_tokens": 34629402.0,
      "step": 15110
    },
    {
      "entropy": 5.350461816787719,
      "epoch": 1.451969260326609,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004795066119407827,
      "loss": 4.9465,
      "mean_token_accuracy": 0.23017913848161697,
      "num_tokens": 34642523.0,
      "step": 15115
    },
    {
      "entropy": 5.195211172103882,
      "epoch": 1.4524495677233429,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004794923142051873,
      "loss": 4.7573,
      "mean_token_accuracy": 0.24226665496826172,
      "num_tokens": 34653714.0,
      "step": 15120
    },
    {
      "entropy": 5.202053022384644,
      "epoch": 1.452929875120077,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047947801172188755,
      "loss": 4.8238,
      "mean_token_accuracy": 0.23381618410348892,
      "num_tokens": 34664899.0,
      "step": 15125
    },
    {
      "entropy": 5.327687311172485,
      "epoch": 1.4534101825168109,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004794637044912155,
      "loss": 4.8705,
      "mean_token_accuracy": 0.22637833207845687,
      "num_tokens": 34674888.0,
      "step": 15130
    },
    {
      "entropy": 5.216417026519776,
      "epoch": 1.4538904899135447,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004794493925135034,
      "loss": 4.7814,
      "mean_token_accuracy": 0.23789752572774886,
      "num_tokens": 34686446.0,
      "step": 15135
    },
    {
      "entropy": 5.333182859420776,
      "epoch": 1.4543707973102786,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047943507578908357,
      "loss": 4.9489,
      "mean_token_accuracy": 0.23625344038009644,
      "num_tokens": 34697592.0,
      "step": 15140
    },
    {
      "entropy": 5.256245803833008,
      "epoch": 1.4548511047070125,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004794207543182883,
      "loss": 4.8423,
      "mean_token_accuracy": 0.23413599133491517,
      "num_tokens": 34709305.0,
      "step": 15145
    },
    {
      "entropy": 5.288630962371826,
      "epoch": 1.4553314121037464,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047940642810145005,
      "loss": 4.9026,
      "mean_token_accuracy": 0.2314820721745491,
      "num_tokens": 34721606.0,
      "step": 15150
    },
    {
      "entropy": 5.27018461227417,
      "epoch": 1.4558117195004803,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047939209713890156,
      "loss": 4.798,
      "mean_token_accuracy": 0.23480991423130035,
      "num_tokens": 34732125.0,
      "step": 15155
    },
    {
      "entropy": 5.23961033821106,
      "epoch": 1.4562920268972142,
      "grad_norm": 1.0,
      "learning_rate": 0.00047937776143097547,
      "loss": 4.8263,
      "mean_token_accuracy": 0.2285424754023552,
      "num_tokens": 34743317.0,
      "step": 15160
    },
    {
      "entropy": 5.212748050689697,
      "epoch": 1.456772334293948,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004793634209780047,
      "loss": 4.8308,
      "mean_token_accuracy": 0.22530567049980163,
      "num_tokens": 34754546.0,
      "step": 15165
    },
    {
      "entropy": 5.245177555084228,
      "epoch": 1.457252641690682,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004793490757803221,
      "loss": 4.7654,
      "mean_token_accuracy": 0.2355531170964241,
      "num_tokens": 34766027.0,
      "step": 15170
    },
    {
      "entropy": 5.297322416305542,
      "epoch": 1.457732949087416,
      "grad_norm": 0.84375,
      "learning_rate": 0.00047933472583826063,
      "loss": 4.832,
      "mean_token_accuracy": 0.22845411747694017,
      "num_tokens": 34777750.0,
      "step": 15175
    },
    {
      "entropy": 5.252485752105713,
      "epoch": 1.45821325648415,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004793203711521537,
      "loss": 4.8847,
      "mean_token_accuracy": 0.2348158538341522,
      "num_tokens": 34788924.0,
      "step": 15180
    },
    {
      "entropy": 5.265337181091309,
      "epoch": 1.4586935638808838,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047930601172233446,
      "loss": 4.8223,
      "mean_token_accuracy": 0.23620762974023818,
      "num_tokens": 34800153.0,
      "step": 15185
    },
    {
      "entropy": 5.248081922531128,
      "epoch": 1.4591738712776177,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047929164754913624,
      "loss": 4.8619,
      "mean_token_accuracy": 0.23097277581691741,
      "num_tokens": 34811597.0,
      "step": 15190
    },
    {
      "entropy": 5.254304695129394,
      "epoch": 1.4596541786743515,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004792772786328926,
      "loss": 4.759,
      "mean_token_accuracy": 0.24488801658153533,
      "num_tokens": 34822661.0,
      "step": 15195
    },
    {
      "entropy": 5.360959005355835,
      "epoch": 1.4601344860710854,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047926290497393714,
      "loss": 4.9239,
      "mean_token_accuracy": 0.22714297324419022,
      "num_tokens": 34833554.0,
      "step": 15200
    },
    {
      "entropy": 5.253881788253784,
      "epoch": 1.4606147934678195,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004792485265726036,
      "loss": 4.875,
      "mean_token_accuracy": 0.23483344465494155,
      "num_tokens": 34846060.0,
      "step": 15205
    },
    {
      "entropy": 5.2821228981018065,
      "epoch": 1.4610951008645534,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004792341434292257,
      "loss": 4.8246,
      "mean_token_accuracy": 0.23266631811857225,
      "num_tokens": 34857324.0,
      "step": 15210
    },
    {
      "entropy": 5.23723840713501,
      "epoch": 1.4615754082612873,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004792197555441374,
      "loss": 4.7766,
      "mean_token_accuracy": 0.24426010400056838,
      "num_tokens": 34869317.0,
      "step": 15215
    },
    {
      "entropy": 5.223677253723144,
      "epoch": 1.4620557156580212,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004792053629176729,
      "loss": 4.8173,
      "mean_token_accuracy": 0.23519230931997298,
      "num_tokens": 34880961.0,
      "step": 15220
    },
    {
      "entropy": 5.26459493637085,
      "epoch": 1.462536023054755,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004791909655501662,
      "loss": 4.8287,
      "mean_token_accuracy": 0.2391995146870613,
      "num_tokens": 34893518.0,
      "step": 15225
    },
    {
      "entropy": 5.316538858413696,
      "epoch": 1.463016330451489,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004791765634419516,
      "loss": 4.8712,
      "mean_token_accuracy": 0.22929371893405914,
      "num_tokens": 34906201.0,
      "step": 15230
    },
    {
      "entropy": 5.286941623687744,
      "epoch": 1.4634966378482228,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047916215659336343,
      "loss": 4.8096,
      "mean_token_accuracy": 0.23142716735601426,
      "num_tokens": 34916841.0,
      "step": 15235
    },
    {
      "entropy": 5.342752265930176,
      "epoch": 1.4639769452449567,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004791477450047363,
      "loss": 4.9266,
      "mean_token_accuracy": 0.22135408073663712,
      "num_tokens": 34928081.0,
      "step": 15240
    },
    {
      "entropy": 5.28982400894165,
      "epoch": 1.4644572526416906,
      "grad_norm": 1.0,
      "learning_rate": 0.00047913332867640464,
      "loss": 4.8269,
      "mean_token_accuracy": 0.23381050378084184,
      "num_tokens": 34939072.0,
      "step": 15245
    },
    {
      "entropy": 5.291661643981934,
      "epoch": 1.4649375600384245,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004791189076087033,
      "loss": 4.8582,
      "mean_token_accuracy": 0.22659847140312195,
      "num_tokens": 34950898.0,
      "step": 15250
    },
    {
      "entropy": 5.360901832580566,
      "epoch": 1.4654178674351586,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047910448180196703,
      "loss": 4.8897,
      "mean_token_accuracy": 0.22808808982372283,
      "num_tokens": 34961560.0,
      "step": 15255
    },
    {
      "entropy": 5.254022455215454,
      "epoch": 1.4658981748318924,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004790900512565307,
      "loss": 4.8707,
      "mean_token_accuracy": 0.24144495278596878,
      "num_tokens": 34973604.0,
      "step": 15260
    },
    {
      "entropy": 5.2547706127166744,
      "epoch": 1.4663784822286263,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004790756159727294,
      "loss": 4.9053,
      "mean_token_accuracy": 0.2245004191994667,
      "num_tokens": 34985369.0,
      "step": 15265
    },
    {
      "entropy": 5.413816356658936,
      "epoch": 1.4668587896253602,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047906117595089835,
      "loss": 4.9465,
      "mean_token_accuracy": 0.2200206473469734,
      "num_tokens": 34996746.0,
      "step": 15270
    },
    {
      "entropy": 5.260503482818604,
      "epoch": 1.467339097022094,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004790467311913727,
      "loss": 4.869,
      "mean_token_accuracy": 0.23899111300706863,
      "num_tokens": 35008032.0,
      "step": 15275
    },
    {
      "entropy": 5.261763286590576,
      "epoch": 1.4678194044188282,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004790322816944879,
      "loss": 4.9207,
      "mean_token_accuracy": 0.23015994429588318,
      "num_tokens": 35019423.0,
      "step": 15280
    },
    {
      "entropy": 5.231019449234009,
      "epoch": 1.468299711815562,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004790178274605793,
      "loss": 4.7726,
      "mean_token_accuracy": 0.24148558527231218,
      "num_tokens": 35029985.0,
      "step": 15285
    },
    {
      "entropy": 5.313793659210205,
      "epoch": 1.468780019212296,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047900336848998254,
      "loss": 4.8768,
      "mean_token_accuracy": 0.233432637155056,
      "num_tokens": 35040580.0,
      "step": 15290
    },
    {
      "entropy": 5.325662851333618,
      "epoch": 1.4692603266090298,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004789889047830334,
      "loss": 4.9091,
      "mean_token_accuracy": 0.23505545407533646,
      "num_tokens": 35051414.0,
      "step": 15295
    },
    {
      "entropy": 5.281435108184814,
      "epoch": 1.4697406340057637,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047897443634006766,
      "loss": 4.8281,
      "mean_token_accuracy": 0.23224953711032867,
      "num_tokens": 35062483.0,
      "step": 15300
    },
    {
      "entropy": 5.273050260543823,
      "epoch": 1.4702209414024976,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004789599631614211,
      "loss": 4.8235,
      "mean_token_accuracy": 0.234444759786129,
      "num_tokens": 35073536.0,
      "step": 15305
    },
    {
      "entropy": 5.320491695404053,
      "epoch": 1.4707012487992315,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004789454852474298,
      "loss": 4.8551,
      "mean_token_accuracy": 0.22591530084609984,
      "num_tokens": 35085636.0,
      "step": 15310
    },
    {
      "entropy": 5.28409104347229,
      "epoch": 1.4711815561959654,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004789310025984299,
      "loss": 4.8182,
      "mean_token_accuracy": 0.22730459570884703,
      "num_tokens": 35096660.0,
      "step": 15315
    },
    {
      "entropy": 5.280424165725708,
      "epoch": 1.4716618635926992,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047891651521475776,
      "loss": 4.9423,
      "mean_token_accuracy": 0.22655123621225357,
      "num_tokens": 35109657.0,
      "step": 15320
    },
    {
      "entropy": 5.268303346633911,
      "epoch": 1.4721421709894331,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047890202309674963,
      "loss": 4.7813,
      "mean_token_accuracy": 0.23058037012815474,
      "num_tokens": 35120399.0,
      "step": 15325
    },
    {
      "entropy": 5.258905649185181,
      "epoch": 1.4726224783861672,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047888752624474195,
      "loss": 4.7492,
      "mean_token_accuracy": 0.23731111288070678,
      "num_tokens": 35130105.0,
      "step": 15330
    },
    {
      "entropy": 5.182507658004761,
      "epoch": 1.473102785782901,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004788730246590714,
      "loss": 4.7981,
      "mean_token_accuracy": 0.23485278189182282,
      "num_tokens": 35142547.0,
      "step": 15335
    },
    {
      "entropy": 5.230224990844727,
      "epoch": 1.473583093179635,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047885851834007456,
      "loss": 4.7395,
      "mean_token_accuracy": 0.237178798019886,
      "num_tokens": 35154603.0,
      "step": 15340
    },
    {
      "entropy": 5.2403429508209225,
      "epoch": 1.4740634005763689,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047884400728808824,
      "loss": 4.8078,
      "mean_token_accuracy": 0.23285267651081085,
      "num_tokens": 35165727.0,
      "step": 15345
    },
    {
      "entropy": 5.227528095245361,
      "epoch": 1.4745437079731027,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004788294915034494,
      "loss": 4.8709,
      "mean_token_accuracy": 0.2308126598596573,
      "num_tokens": 35177642.0,
      "step": 15350
    },
    {
      "entropy": 5.24815263748169,
      "epoch": 1.4750240153698366,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000478814970986495,
      "loss": 4.8476,
      "mean_token_accuracy": 0.2356531247496605,
      "num_tokens": 35189377.0,
      "step": 15355
    },
    {
      "entropy": 5.283090543746948,
      "epoch": 1.4755043227665707,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047880044573756213,
      "loss": 4.9037,
      "mean_token_accuracy": 0.22783648669719697,
      "num_tokens": 35201244.0,
      "step": 15360
    },
    {
      "entropy": 5.2860297679901125,
      "epoch": 1.4759846301633046,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047878591575698816,
      "loss": 4.8488,
      "mean_token_accuracy": 0.23592701852321624,
      "num_tokens": 35212491.0,
      "step": 15365
    },
    {
      "entropy": 5.194163274765015,
      "epoch": 1.4764649375600385,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004787713810451103,
      "loss": 4.8073,
      "mean_token_accuracy": 0.23318872600793839,
      "num_tokens": 35225552.0,
      "step": 15370
    },
    {
      "entropy": 5.225209331512451,
      "epoch": 1.4769452449567724,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047875684160226606,
      "loss": 4.81,
      "mean_token_accuracy": 0.2377237781882286,
      "num_tokens": 35237961.0,
      "step": 15375
    },
    {
      "entropy": 5.259709596633911,
      "epoch": 1.4774255523535063,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000478742297428793,
      "loss": 4.8452,
      "mean_token_accuracy": 0.2303194910287857,
      "num_tokens": 35249502.0,
      "step": 15380
    },
    {
      "entropy": 5.283733177185058,
      "epoch": 1.4779058597502401,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047872774852502877,
      "loss": 4.8924,
      "mean_token_accuracy": 0.22718950659036635,
      "num_tokens": 35262978.0,
      "step": 15385
    },
    {
      "entropy": 5.323650360107422,
      "epoch": 1.478386167146974,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004787131948913112,
      "loss": 4.9389,
      "mean_token_accuracy": 0.2291031762957573,
      "num_tokens": 35273289.0,
      "step": 15390
    },
    {
      "entropy": 5.293175601959229,
      "epoch": 1.478866474543708,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047869863652797806,
      "loss": 4.8447,
      "mean_token_accuracy": 0.22917833179235458,
      "num_tokens": 35283827.0,
      "step": 15395
    },
    {
      "entropy": 5.3472906112670895,
      "epoch": 1.4793467819404418,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004786840734353675,
      "loss": 4.8865,
      "mean_token_accuracy": 0.2306977778673172,
      "num_tokens": 35293264.0,
      "step": 15400
    },
    {
      "entropy": 5.243727970123291,
      "epoch": 1.4798270893371757,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047866950561381756,
      "loss": 4.8335,
      "mean_token_accuracy": 0.23660454601049424,
      "num_tokens": 35304380.0,
      "step": 15405
    },
    {
      "entropy": 5.222856092453003,
      "epoch": 1.4803073967339098,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004786549330636665,
      "loss": 4.7896,
      "mean_token_accuracy": 0.23484990745782852,
      "num_tokens": 35316805.0,
      "step": 15410
    },
    {
      "entropy": 5.245000267028809,
      "epoch": 1.4807877041306436,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047864035578525256,
      "loss": 4.8331,
      "mean_token_accuracy": 0.2321616977453232,
      "num_tokens": 35328645.0,
      "step": 15415
    },
    {
      "entropy": 5.238785648345948,
      "epoch": 1.4812680115273775,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004786257737789143,
      "loss": 4.8318,
      "mean_token_accuracy": 0.22881092578172685,
      "num_tokens": 35339983.0,
      "step": 15420
    },
    {
      "entropy": 5.373903846740722,
      "epoch": 1.4817483189241114,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004786111870449902,
      "loss": 4.9581,
      "mean_token_accuracy": 0.22171321362257004,
      "num_tokens": 35350720.0,
      "step": 15425
    },
    {
      "entropy": 5.299764156341553,
      "epoch": 1.4822286263208453,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047859659558381894,
      "loss": 4.8929,
      "mean_token_accuracy": 0.22602088153362274,
      "num_tokens": 35361435.0,
      "step": 15430
    },
    {
      "entropy": 5.254367113113403,
      "epoch": 1.4827089337175792,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047858199939573935,
      "loss": 4.9468,
      "mean_token_accuracy": 0.22568488270044326,
      "num_tokens": 35372829.0,
      "step": 15435
    },
    {
      "entropy": 5.321689701080322,
      "epoch": 1.4831892411143133,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047856739848109014,
      "loss": 4.8483,
      "mean_token_accuracy": 0.22880145013332367,
      "num_tokens": 35383270.0,
      "step": 15440
    },
    {
      "entropy": 5.365650701522827,
      "epoch": 1.4836695485110472,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047855279284021046,
      "loss": 4.939,
      "mean_token_accuracy": 0.22530481368303298,
      "num_tokens": 35396308.0,
      "step": 15445
    },
    {
      "entropy": 5.265581703186035,
      "epoch": 1.484149855907781,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047853818247343933,
      "loss": 4.8667,
      "mean_token_accuracy": 0.23275803178548812,
      "num_tokens": 35408944.0,
      "step": 15450
    },
    {
      "entropy": 5.3016375541687015,
      "epoch": 1.484630163304515,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047852356738111606,
      "loss": 4.9031,
      "mean_token_accuracy": 0.22381726503372193,
      "num_tokens": 35422227.0,
      "step": 15455
    },
    {
      "entropy": 5.315992832183838,
      "epoch": 1.4851104707012488,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004785089475635799,
      "loss": 4.897,
      "mean_token_accuracy": 0.22315810471773148,
      "num_tokens": 35434110.0,
      "step": 15460
    },
    {
      "entropy": 5.31408200263977,
      "epoch": 1.4855907780979827,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047849432302117024,
      "loss": 4.8617,
      "mean_token_accuracy": 0.2364679142832756,
      "num_tokens": 35445194.0,
      "step": 15465
    },
    {
      "entropy": 5.328440380096436,
      "epoch": 1.4860710854947166,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047847969375422656,
      "loss": 4.8643,
      "mean_token_accuracy": 0.23172784000635147,
      "num_tokens": 35455953.0,
      "step": 15470
    },
    {
      "entropy": 5.25609245300293,
      "epoch": 1.4865513928914504,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004784650597630887,
      "loss": 4.7515,
      "mean_token_accuracy": 0.2390329509973526,
      "num_tokens": 35466932.0,
      "step": 15475
    },
    {
      "entropy": 5.208602046966552,
      "epoch": 1.4870317002881843,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047845042104809635,
      "loss": 4.8244,
      "mean_token_accuracy": 0.23395713567733764,
      "num_tokens": 35477755.0,
      "step": 15480
    },
    {
      "entropy": 5.210040712356568,
      "epoch": 1.4875120076849184,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004784357776095892,
      "loss": 4.7748,
      "mean_token_accuracy": 0.23724135905504226,
      "num_tokens": 35489913.0,
      "step": 15485
    },
    {
      "entropy": 5.26903395652771,
      "epoch": 1.4879923150816523,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004784211294479075,
      "loss": 4.877,
      "mean_token_accuracy": 0.23402811884880065,
      "num_tokens": 35500314.0,
      "step": 15490
    },
    {
      "entropy": 5.34130859375,
      "epoch": 1.4884726224783862,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004784064765633912,
      "loss": 4.9529,
      "mean_token_accuracy": 0.22417826801538468,
      "num_tokens": 35514717.0,
      "step": 15495
    },
    {
      "entropy": 5.259320831298828,
      "epoch": 1.48895292987512,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047839181895638057,
      "loss": 4.8409,
      "mean_token_accuracy": 0.23184080719947814,
      "num_tokens": 35526122.0,
      "step": 15500
    },
    {
      "entropy": 5.2683515548706055,
      "epoch": 1.489433237271854,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047837715662721575,
      "loss": 4.8603,
      "mean_token_accuracy": 0.2255132168531418,
      "num_tokens": 35538270.0,
      "step": 15505
    },
    {
      "entropy": 5.359716939926147,
      "epoch": 1.4899135446685878,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004783624895762372,
      "loss": 4.9276,
      "mean_token_accuracy": 0.22777810245752333,
      "num_tokens": 35549598.0,
      "step": 15510
    },
    {
      "entropy": 5.309580230712891,
      "epoch": 1.490393852065322,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047834781780378563,
      "loss": 4.8487,
      "mean_token_accuracy": 0.23110825270414354,
      "num_tokens": 35559623.0,
      "step": 15515
    },
    {
      "entropy": 5.2623217582702635,
      "epoch": 1.4908741594620558,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004783331413102015,
      "loss": 4.8368,
      "mean_token_accuracy": 0.23138171434402466,
      "num_tokens": 35571578.0,
      "step": 15520
    },
    {
      "entropy": 5.266974973678589,
      "epoch": 1.4913544668587897,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047831846009582557,
      "loss": 4.9035,
      "mean_token_accuracy": 0.22843256294727327,
      "num_tokens": 35583570.0,
      "step": 15525
    },
    {
      "entropy": 5.288261985778808,
      "epoch": 1.4918347742555236,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004783037741609988,
      "loss": 4.894,
      "mean_token_accuracy": 0.2274307683110237,
      "num_tokens": 35595360.0,
      "step": 15530
    },
    {
      "entropy": 5.245399475097656,
      "epoch": 1.4923150816522575,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004782890835060621,
      "loss": 4.8366,
      "mean_token_accuracy": 0.2288857191801071,
      "num_tokens": 35607326.0,
      "step": 15535
    },
    {
      "entropy": 5.361185312271118,
      "epoch": 1.4927953890489913,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004782743881313564,
      "loss": 4.8908,
      "mean_token_accuracy": 0.22909253984689712,
      "num_tokens": 35618504.0,
      "step": 15540
    },
    {
      "entropy": 5.243128299713135,
      "epoch": 1.4932756964457252,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047825968803722315,
      "loss": 4.8055,
      "mean_token_accuracy": 0.23964255452156066,
      "num_tokens": 35631111.0,
      "step": 15545
    },
    {
      "entropy": 5.257733297348023,
      "epoch": 1.493756003842459,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004782449832240035,
      "loss": 4.85,
      "mean_token_accuracy": 0.23779718279838563,
      "num_tokens": 35643434.0,
      "step": 15550
    },
    {
      "entropy": 5.346720886230469,
      "epoch": 1.494236311239193,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004782302736920387,
      "loss": 4.9232,
      "mean_token_accuracy": 0.22857324481010438,
      "num_tokens": 35654531.0,
      "step": 15555
    },
    {
      "entropy": 5.300565528869629,
      "epoch": 1.4947166186359269,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004782155594416705,
      "loss": 4.8491,
      "mean_token_accuracy": 0.22991302311420442,
      "num_tokens": 35665488.0,
      "step": 15560
    },
    {
      "entropy": 5.316967344284057,
      "epoch": 1.495196926032661,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047820084047324045,
      "loss": 4.8979,
      "mean_token_accuracy": 0.22683439701795577,
      "num_tokens": 35677110.0,
      "step": 15565
    },
    {
      "entropy": 5.353493690490723,
      "epoch": 1.4956772334293948,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047818611678709027,
      "loss": 4.9286,
      "mean_token_accuracy": 0.23094941824674606,
      "num_tokens": 35688694.0,
      "step": 15570
    },
    {
      "entropy": 5.247624731063842,
      "epoch": 1.4961575408261287,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004781713883835618,
      "loss": 4.8255,
      "mean_token_accuracy": 0.23079129457473754,
      "num_tokens": 35699497.0,
      "step": 15575
    },
    {
      "entropy": 5.3361663818359375,
      "epoch": 1.4966378482228626,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047815665526299695,
      "loss": 4.883,
      "mean_token_accuracy": 0.22902489304542542,
      "num_tokens": 35711034.0,
      "step": 15580
    },
    {
      "entropy": 5.2670492172241214,
      "epoch": 1.4971181556195965,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004781419174257378,
      "loss": 4.8013,
      "mean_token_accuracy": 0.23513386994600297,
      "num_tokens": 35721804.0,
      "step": 15585
    },
    {
      "entropy": 5.205309295654297,
      "epoch": 1.4975984630163304,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004781271748721266,
      "loss": 4.8097,
      "mean_token_accuracy": 0.23864074051380157,
      "num_tokens": 35732897.0,
      "step": 15590
    },
    {
      "entropy": 5.151329278945923,
      "epoch": 1.4980787704130645,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004781124276025055,
      "loss": 4.7196,
      "mean_token_accuracy": 0.24566929042339325,
      "num_tokens": 35743792.0,
      "step": 15595
    },
    {
      "entropy": 5.303867340087891,
      "epoch": 1.4985590778097984,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000478097675617217,
      "loss": 4.866,
      "mean_token_accuracy": 0.22827015072107315,
      "num_tokens": 35754659.0,
      "step": 15600
    },
    {
      "entropy": 5.334948396682739,
      "epoch": 1.4990393852065322,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047808291891660357,
      "loss": 4.8663,
      "mean_token_accuracy": 0.2350357711315155,
      "num_tokens": 35766269.0,
      "step": 15605
    },
    {
      "entropy": 5.30242691040039,
      "epoch": 1.4995196926032661,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047806815750100774,
      "loss": 4.8869,
      "mean_token_accuracy": 0.23172952681779863,
      "num_tokens": 35777634.0,
      "step": 15610
    },
    {
      "entropy": 5.273705005645752,
      "epoch": 1.5,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004780533913707723,
      "loss": 4.8843,
      "mean_token_accuracy": 0.2284536123275757,
      "num_tokens": 35788689.0,
      "step": 15615
    },
    {
      "entropy": 5.271505737304688,
      "epoch": 1.5004803073967339,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047803862052624006,
      "loss": 4.8099,
      "mean_token_accuracy": 0.23418974578380586,
      "num_tokens": 35799365.0,
      "step": 15620
    },
    {
      "entropy": 5.229896163940429,
      "epoch": 1.5009606147934678,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047802384496775397,
      "loss": 4.7698,
      "mean_token_accuracy": 0.23959048986434936,
      "num_tokens": 35810682.0,
      "step": 15625
    },
    {
      "entropy": 5.271739482879639,
      "epoch": 1.5014409221902016,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004780090646956571,
      "loss": 4.8584,
      "mean_token_accuracy": 0.22535869777202605,
      "num_tokens": 35822193.0,
      "step": 15630
    },
    {
      "entropy": 5.226168251037597,
      "epoch": 1.5019212295869355,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047799427971029245,
      "loss": 4.8645,
      "mean_token_accuracy": 0.22821758091449737,
      "num_tokens": 35835486.0,
      "step": 15635
    },
    {
      "entropy": 5.301498746871948,
      "epoch": 1.5024015369836694,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004779794900120034,
      "loss": 4.8513,
      "mean_token_accuracy": 0.22891197949647904,
      "num_tokens": 35846245.0,
      "step": 15640
    },
    {
      "entropy": 5.344848251342773,
      "epoch": 1.5028818443804035,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004779646956011334,
      "loss": 4.9737,
      "mean_token_accuracy": 0.23060237765312194,
      "num_tokens": 35857914.0,
      "step": 15645
    },
    {
      "entropy": 5.320210886001587,
      "epoch": 1.5033621517771374,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047794989647802574,
      "loss": 4.8672,
      "mean_token_accuracy": 0.22457896918058395,
      "num_tokens": 35870235.0,
      "step": 15650
    },
    {
      "entropy": 5.270398902893066,
      "epoch": 1.5038424591738713,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047793509264302424,
      "loss": 4.9085,
      "mean_token_accuracy": 0.2301445409655571,
      "num_tokens": 35882154.0,
      "step": 15655
    },
    {
      "entropy": 5.268236112594605,
      "epoch": 1.5043227665706052,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047792028409647237,
      "loss": 4.8517,
      "mean_token_accuracy": 0.2426189512014389,
      "num_tokens": 35894192.0,
      "step": 15660
    },
    {
      "entropy": 5.306827783584595,
      "epoch": 1.5048030739673393,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047790547083871414,
      "loss": 4.8468,
      "mean_token_accuracy": 0.22934847474098205,
      "num_tokens": 35906017.0,
      "step": 15665
    },
    {
      "entropy": 5.328298282623291,
      "epoch": 1.5052833813640731,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047789065287009335,
      "loss": 4.9165,
      "mean_token_accuracy": 0.22652230262756348,
      "num_tokens": 35917671.0,
      "step": 15670
    },
    {
      "entropy": 5.169233131408691,
      "epoch": 1.505763688760807,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004778758301909542,
      "loss": 4.7888,
      "mean_token_accuracy": 0.23823485523462296,
      "num_tokens": 35928421.0,
      "step": 15675
    },
    {
      "entropy": 5.324822568893433,
      "epoch": 1.506243996157541,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004778610028016405,
      "loss": 4.9225,
      "mean_token_accuracy": 0.22632264345884323,
      "num_tokens": 35939689.0,
      "step": 15680
    },
    {
      "entropy": 5.332788276672363,
      "epoch": 1.5067243035542748,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004778461707024967,
      "loss": 4.9483,
      "mean_token_accuracy": 0.2270145073533058,
      "num_tokens": 35952342.0,
      "step": 15685
    },
    {
      "entropy": 5.418045568466186,
      "epoch": 1.5072046109510087,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004778313338938672,
      "loss": 4.9852,
      "mean_token_accuracy": 0.21400177776813506,
      "num_tokens": 35964726.0,
      "step": 15690
    },
    {
      "entropy": 5.317668294906616,
      "epoch": 1.5076849183477425,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047781649237609643,
      "loss": 4.8284,
      "mean_token_accuracy": 0.24259402751922607,
      "num_tokens": 35976932.0,
      "step": 15695
    },
    {
      "entropy": 5.2004941463470455,
      "epoch": 1.5081652257444764,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004778016461495289,
      "loss": 4.7695,
      "mean_token_accuracy": 0.2373267874121666,
      "num_tokens": 35988371.0,
      "step": 15700
    },
    {
      "entropy": 5.154518461227417,
      "epoch": 1.5086455331412103,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004777867952145094,
      "loss": 4.7694,
      "mean_token_accuracy": 0.24194466024637223,
      "num_tokens": 36001310.0,
      "step": 15705
    },
    {
      "entropy": 5.379926252365112,
      "epoch": 1.5091258405379442,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004777719395713826,
      "loss": 4.9808,
      "mean_token_accuracy": 0.2202196165919304,
      "num_tokens": 36011975.0,
      "step": 15710
    },
    {
      "entropy": 5.293229722976685,
      "epoch": 1.509606147934678,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047775707922049354,
      "loss": 4.891,
      "mean_token_accuracy": 0.22999646067619323,
      "num_tokens": 36024987.0,
      "step": 15715
    },
    {
      "entropy": 5.270932388305664,
      "epoch": 1.510086455331412,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004777422141621871,
      "loss": 4.8531,
      "mean_token_accuracy": 0.23180584162473677,
      "num_tokens": 36036003.0,
      "step": 15720
    },
    {
      "entropy": 5.308686304092407,
      "epoch": 1.510566762728146,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004777273443968085,
      "loss": 4.8898,
      "mean_token_accuracy": 0.22871946543455124,
      "num_tokens": 36047948.0,
      "step": 15725
    },
    {
      "entropy": 5.254090213775635,
      "epoch": 1.51104707012488,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004777124699247029,
      "loss": 4.8206,
      "mean_token_accuracy": 0.23502880930900574,
      "num_tokens": 36059197.0,
      "step": 15730
    },
    {
      "entropy": 5.34067816734314,
      "epoch": 1.5115273775216138,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004776975907462157,
      "loss": 4.9297,
      "mean_token_accuracy": 0.2249864548444748,
      "num_tokens": 36072167.0,
      "step": 15735
    },
    {
      "entropy": 5.403582191467285,
      "epoch": 1.5120076849183477,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004776827068616924,
      "loss": 4.9768,
      "mean_token_accuracy": 0.22703094482421876,
      "num_tokens": 36083805.0,
      "step": 15740
    },
    {
      "entropy": 5.245254325866699,
      "epoch": 1.5124879923150818,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004776678182714785,
      "loss": 4.8214,
      "mean_token_accuracy": 0.23898584842681886,
      "num_tokens": 36095450.0,
      "step": 15745
    },
    {
      "entropy": 5.2480080127716064,
      "epoch": 1.5129682997118157,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047765292497591955,
      "loss": 4.8286,
      "mean_token_accuracy": 0.24391250163316727,
      "num_tokens": 36106319.0,
      "step": 15750
    },
    {
      "entropy": 5.221454620361328,
      "epoch": 1.5134486071085496,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047763802697536146,
      "loss": 4.6802,
      "mean_token_accuracy": 0.2463770717382431,
      "num_tokens": 36117015.0,
      "step": 15755
    },
    {
      "entropy": 5.250518369674682,
      "epoch": 1.5139289145052834,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047762312427015015,
      "loss": 4.8521,
      "mean_token_accuracy": 0.22387492060661315,
      "num_tokens": 36127308.0,
      "step": 15760
    },
    {
      "entropy": 5.216578149795533,
      "epoch": 1.5144092219020173,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047760821686063153,
      "loss": 4.7729,
      "mean_token_accuracy": 0.2362382397055626,
      "num_tokens": 36139811.0,
      "step": 15765
    },
    {
      "entropy": 5.345195770263672,
      "epoch": 1.5148895292987512,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047759330474715173,
      "loss": 4.8982,
      "mean_token_accuracy": 0.2262015625834465,
      "num_tokens": 36151330.0,
      "step": 15770
    },
    {
      "entropy": 5.267193031311035,
      "epoch": 1.515369836695485,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047757838793005704,
      "loss": 4.8226,
      "mean_token_accuracy": 0.23103735893964766,
      "num_tokens": 36162436.0,
      "step": 15775
    },
    {
      "entropy": 5.371133184432983,
      "epoch": 1.515850144092219,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047756346640969366,
      "loss": 5.0195,
      "mean_token_accuracy": 0.21407540738582612,
      "num_tokens": 36175642.0,
      "step": 15780
    },
    {
      "entropy": 5.27635555267334,
      "epoch": 1.5163304514889528,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047754854018640803,
      "loss": 4.7891,
      "mean_token_accuracy": 0.23149359822273255,
      "num_tokens": 36186925.0,
      "step": 15785
    },
    {
      "entropy": 5.2756870746612545,
      "epoch": 1.5168107588856867,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047753360926054684,
      "loss": 4.9028,
      "mean_token_accuracy": 0.22638735324144363,
      "num_tokens": 36199280.0,
      "step": 15790
    },
    {
      "entropy": 5.230637311935425,
      "epoch": 1.5172910662824206,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047751867363245653,
      "loss": 4.778,
      "mean_token_accuracy": 0.237116976082325,
      "num_tokens": 36210276.0,
      "step": 15795
    },
    {
      "entropy": 5.363597106933594,
      "epoch": 1.5177713736791547,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004775037333024841,
      "loss": 4.8761,
      "mean_token_accuracy": 0.23268525153398514,
      "num_tokens": 36221211.0,
      "step": 15800
    },
    {
      "entropy": 5.209538984298706,
      "epoch": 1.5182516810758886,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004774887882709762,
      "loss": 4.7324,
      "mean_token_accuracy": 0.2428848221898079,
      "num_tokens": 36233465.0,
      "step": 15805
    },
    {
      "entropy": 5.213021230697632,
      "epoch": 1.5187319884726225,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047747383853827995,
      "loss": 4.7852,
      "mean_token_accuracy": 0.2360478848218918,
      "num_tokens": 36245594.0,
      "step": 15810
    },
    {
      "entropy": 5.300924968719483,
      "epoch": 1.5192122958693564,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004774588841047424,
      "loss": 4.9498,
      "mean_token_accuracy": 0.22377603203058244,
      "num_tokens": 36257776.0,
      "step": 15815
    },
    {
      "entropy": 5.304608249664307,
      "epoch": 1.5196926032660905,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004774439249707108,
      "loss": 4.8087,
      "mean_token_accuracy": 0.22620662152767182,
      "num_tokens": 36269056.0,
      "step": 15820
    },
    {
      "entropy": 5.291815614700317,
      "epoch": 1.5201729106628243,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004774289611365323,
      "loss": 4.8182,
      "mean_token_accuracy": 0.22645604014396667,
      "num_tokens": 36280700.0,
      "step": 15825
    },
    {
      "entropy": 5.3109039783477785,
      "epoch": 1.5206532180595582,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047741399260255434,
      "loss": 4.8916,
      "mean_token_accuracy": 0.22616513818502426,
      "num_tokens": 36291708.0,
      "step": 15830
    },
    {
      "entropy": 5.264336729049683,
      "epoch": 1.521133525456292,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047739901936912467,
      "loss": 4.8041,
      "mean_token_accuracy": 0.2315405935049057,
      "num_tokens": 36303830.0,
      "step": 15835
    },
    {
      "entropy": 5.270111227035523,
      "epoch": 1.521613832853026,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004773840414365907,
      "loss": 4.8396,
      "mean_token_accuracy": 0.23100537359714507,
      "num_tokens": 36316514.0,
      "step": 15840
    },
    {
      "entropy": 5.352181053161621,
      "epoch": 1.5220941402497599,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047736905880530026,
      "loss": 4.8867,
      "mean_token_accuracy": 0.2303168222308159,
      "num_tokens": 36328234.0,
      "step": 15845
    },
    {
      "entropy": 5.251244354248047,
      "epoch": 1.5225744476464937,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004773540714756012,
      "loss": 4.8424,
      "mean_token_accuracy": 0.23335151076316835,
      "num_tokens": 36340383.0,
      "step": 15850
    },
    {
      "entropy": 5.2775732517242435,
      "epoch": 1.5230547550432276,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047733907944784144,
      "loss": 4.823,
      "mean_token_accuracy": 0.23859311640262604,
      "num_tokens": 36351970.0,
      "step": 15855
    },
    {
      "entropy": 5.240417909622193,
      "epoch": 1.5235350624399615,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004773240827223691,
      "loss": 4.8191,
      "mean_token_accuracy": 0.2383805826306343,
      "num_tokens": 36363163.0,
      "step": 15860
    },
    {
      "entropy": 5.303683233261109,
      "epoch": 1.5240153698366954,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004773090812995323,
      "loss": 4.8935,
      "mean_token_accuracy": 0.22770759463310242,
      "num_tokens": 36375079.0,
      "step": 15865
    },
    {
      "entropy": 5.3330199241638185,
      "epoch": 1.5244956772334293,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047729407517967945,
      "loss": 4.8075,
      "mean_token_accuracy": 0.23506415486335755,
      "num_tokens": 36386402.0,
      "step": 15870
    },
    {
      "entropy": 5.202378225326538,
      "epoch": 1.5249759846301632,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047727906436315884,
      "loss": 4.76,
      "mean_token_accuracy": 0.24095804393291473,
      "num_tokens": 36397973.0,
      "step": 15875
    },
    {
      "entropy": 5.288751935958862,
      "epoch": 1.5254562920268973,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047726404885031895,
      "loss": 4.935,
      "mean_token_accuracy": 0.22643866240978242,
      "num_tokens": 36409941.0,
      "step": 15880
    },
    {
      "entropy": 5.347469997406006,
      "epoch": 1.5259365994236311,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047724902864150845,
      "loss": 4.9064,
      "mean_token_accuracy": 0.22833145707845687,
      "num_tokens": 36420609.0,
      "step": 15885
    },
    {
      "entropy": 5.283560132980346,
      "epoch": 1.526416906820365,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047723400373707607,
      "loss": 4.8133,
      "mean_token_accuracy": 0.23954736292362214,
      "num_tokens": 36430854.0,
      "step": 15890
    },
    {
      "entropy": 5.191967916488648,
      "epoch": 1.526897214217099,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004772189741373707,
      "loss": 4.7552,
      "mean_token_accuracy": 0.23748446702957154,
      "num_tokens": 36442225.0,
      "step": 15895
    },
    {
      "entropy": 5.408220863342285,
      "epoch": 1.527377521613833,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047720393984274117,
      "loss": 4.9669,
      "mean_token_accuracy": 0.2276992306113243,
      "num_tokens": 36454488.0,
      "step": 15900
    },
    {
      "entropy": 5.320884132385254,
      "epoch": 1.5278578290105669,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047718890085353654,
      "loss": 4.7946,
      "mean_token_accuracy": 0.23915526568889617,
      "num_tokens": 36466197.0,
      "step": 15905
    },
    {
      "entropy": 5.315441370010376,
      "epoch": 1.5283381364073008,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000477173857170106,
      "loss": 4.9037,
      "mean_token_accuracy": 0.2261955514550209,
      "num_tokens": 36478159.0,
      "step": 15910
    },
    {
      "entropy": 5.328295516967773,
      "epoch": 1.5288184438040346,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047715880879279894,
      "loss": 4.9719,
      "mean_token_accuracy": 0.22678953409194946,
      "num_tokens": 36489621.0,
      "step": 15915
    },
    {
      "entropy": 5.341373586654663,
      "epoch": 1.5292987512007685,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004771437557219646,
      "loss": 4.9165,
      "mean_token_accuracy": 0.22716021686792373,
      "num_tokens": 36500629.0,
      "step": 15920
    },
    {
      "entropy": 5.315032958984375,
      "epoch": 1.5297790585975024,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004771286979579524,
      "loss": 4.8782,
      "mean_token_accuracy": 0.2258809596300125,
      "num_tokens": 36512110.0,
      "step": 15925
    },
    {
      "entropy": 5.211618709564209,
      "epoch": 1.5302593659942363,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004771136355011121,
      "loss": 4.7317,
      "mean_token_accuracy": 0.2420870617032051,
      "num_tokens": 36522499.0,
      "step": 15930
    },
    {
      "entropy": 5.283588981628418,
      "epoch": 1.5307396733909702,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047709856835179333,
      "loss": 4.8872,
      "mean_token_accuracy": 0.22645323127508163,
      "num_tokens": 36533465.0,
      "step": 15935
    },
    {
      "entropy": 5.272268867492675,
      "epoch": 1.531219980787704,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047708349651034586,
      "loss": 4.8237,
      "mean_token_accuracy": 0.23559676110744476,
      "num_tokens": 36545213.0,
      "step": 15940
    },
    {
      "entropy": 5.193998622894287,
      "epoch": 1.531700288184438,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047706841997711974,
      "loss": 4.7328,
      "mean_token_accuracy": 0.24212900251150132,
      "num_tokens": 36555656.0,
      "step": 15945
    },
    {
      "entropy": 5.246431446075439,
      "epoch": 1.5321805955811718,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047705333875246495,
      "loss": 4.8448,
      "mean_token_accuracy": 0.2265054389834404,
      "num_tokens": 36568027.0,
      "step": 15950
    },
    {
      "entropy": 5.245044422149658,
      "epoch": 1.532660902977906,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047703825283673153,
      "loss": 4.7815,
      "mean_token_accuracy": 0.23472793996334076,
      "num_tokens": 36579434.0,
      "step": 15955
    },
    {
      "entropy": 5.289803886413575,
      "epoch": 1.5331412103746398,
      "grad_norm": 0.875,
      "learning_rate": 0.0004770231622302699,
      "loss": 4.827,
      "mean_token_accuracy": 0.22790939062833787,
      "num_tokens": 36591068.0,
      "step": 15960
    },
    {
      "entropy": 5.272497701644897,
      "epoch": 1.5336215177713737,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047700806693343016,
      "loss": 4.8079,
      "mean_token_accuracy": 0.22981563955545425,
      "num_tokens": 36603013.0,
      "step": 15965
    },
    {
      "entropy": 5.32612886428833,
      "epoch": 1.5341018251681076,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047699296694656316,
      "loss": 4.9292,
      "mean_token_accuracy": 0.21965130865573884,
      "num_tokens": 36614565.0,
      "step": 15970
    },
    {
      "entropy": 5.243256330490112,
      "epoch": 1.5345821325648417,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004769778622700192,
      "loss": 4.7985,
      "mean_token_accuracy": 0.23676676154136658,
      "num_tokens": 36625685.0,
      "step": 15975
    },
    {
      "entropy": 5.158999395370484,
      "epoch": 1.5350624399615755,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047696275290414885,
      "loss": 4.7655,
      "mean_token_accuracy": 0.23583659827709197,
      "num_tokens": 36637278.0,
      "step": 15980
    },
    {
      "entropy": 5.233465337753296,
      "epoch": 1.5355427473583094,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047694763884930324,
      "loss": 4.8265,
      "mean_token_accuracy": 0.23431949317455292,
      "num_tokens": 36649527.0,
      "step": 15985
    },
    {
      "entropy": 5.295036792755127,
      "epoch": 1.5360230547550433,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047693252010583314,
      "loss": 4.8264,
      "mean_token_accuracy": 0.2301815554499626,
      "num_tokens": 36660653.0,
      "step": 15990
    },
    {
      "entropy": 5.289225339889526,
      "epoch": 1.5365033621517772,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004769173966740895,
      "loss": 4.8186,
      "mean_token_accuracy": 0.23856332153081894,
      "num_tokens": 36672340.0,
      "step": 15995
    },
    {
      "entropy": 5.3356081485748295,
      "epoch": 1.536983669548511,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047690226855442346,
      "loss": 4.9377,
      "mean_token_accuracy": 0.22040853500366211,
      "num_tokens": 36684468.0,
      "step": 16000
    },
    {
      "entropy": 5.26833062171936,
      "epoch": 1.537463976945245,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004768871357471863,
      "loss": 4.8184,
      "mean_token_accuracy": 0.23044191151857377,
      "num_tokens": 36695921.0,
      "step": 16005
    },
    {
      "entropy": 5.209887218475342,
      "epoch": 1.5379442843419788,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047687199825272936,
      "loss": 4.7663,
      "mean_token_accuracy": 0.23483604341745376,
      "num_tokens": 36707029.0,
      "step": 16010
    },
    {
      "entropy": 5.1660240650177,
      "epoch": 1.5384245917387127,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047685685607140403,
      "loss": 4.7455,
      "mean_token_accuracy": 0.24010062962770462,
      "num_tokens": 36717899.0,
      "step": 16015
    },
    {
      "entropy": 5.233793687820435,
      "epoch": 1.5389048991354466,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047684170920356185,
      "loss": 4.7782,
      "mean_token_accuracy": 0.23116395175457,
      "num_tokens": 36728691.0,
      "step": 16020
    },
    {
      "entropy": 5.334080839157105,
      "epoch": 1.5393852065321805,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004768265576495546,
      "loss": 4.9352,
      "mean_token_accuracy": 0.22542988806962966,
      "num_tokens": 36741122.0,
      "step": 16025
    },
    {
      "entropy": 5.334004688262939,
      "epoch": 1.5398655139289144,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047681140140973396,
      "loss": 4.9616,
      "mean_token_accuracy": 0.2304704502224922,
      "num_tokens": 36752841.0,
      "step": 16030
    },
    {
      "entropy": 5.282239246368408,
      "epoch": 1.5403458213256485,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004767962404844517,
      "loss": 4.8454,
      "mean_token_accuracy": 0.23762848377227783,
      "num_tokens": 36763030.0,
      "step": 16035
    },
    {
      "entropy": 5.313391399383545,
      "epoch": 1.5408261287223823,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047678107487406015,
      "loss": 4.9312,
      "mean_token_accuracy": 0.22512030899524688,
      "num_tokens": 36774442.0,
      "step": 16040
    },
    {
      "entropy": 5.294203281402588,
      "epoch": 1.5413064361191162,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047676590457891116,
      "loss": 4.8566,
      "mean_token_accuracy": 0.2289793774485588,
      "num_tokens": 36785650.0,
      "step": 16045
    },
    {
      "entropy": 5.2122739315032955,
      "epoch": 1.54178674351585,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004767507295993569,
      "loss": 4.7813,
      "mean_token_accuracy": 0.23470618426799775,
      "num_tokens": 36797021.0,
      "step": 16050
    },
    {
      "entropy": 5.259646940231323,
      "epoch": 1.5422670509125842,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004767355499357498,
      "loss": 4.8305,
      "mean_token_accuracy": 0.23597093522548676,
      "num_tokens": 36807593.0,
      "step": 16055
    },
    {
      "entropy": 5.2321226596832275,
      "epoch": 1.542747358309318,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004767203655884423,
      "loss": 4.7938,
      "mean_token_accuracy": 0.2345506340265274,
      "num_tokens": 36820185.0,
      "step": 16060
    },
    {
      "entropy": 5.194590330123901,
      "epoch": 1.543227665706052,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004767051765577869,
      "loss": 4.7372,
      "mean_token_accuracy": 0.24312084466218947,
      "num_tokens": 36831350.0,
      "step": 16065
    },
    {
      "entropy": 5.196734476089477,
      "epoch": 1.5437079731027858,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047668998284413624,
      "loss": 4.7219,
      "mean_token_accuracy": 0.23675792664289474,
      "num_tokens": 36842475.0,
      "step": 16070
    },
    {
      "entropy": 5.258511304855347,
      "epoch": 1.5441882804995197,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047667478444784306,
      "loss": 4.8013,
      "mean_token_accuracy": 0.2387152299284935,
      "num_tokens": 36853385.0,
      "step": 16075
    },
    {
      "entropy": 5.212092351913452,
      "epoch": 1.5446685878962536,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004766595813692602,
      "loss": 4.8241,
      "mean_token_accuracy": 0.23064738363027573,
      "num_tokens": 36866339.0,
      "step": 16080
    },
    {
      "entropy": 5.299106884002685,
      "epoch": 1.5451488952929875,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047664437360874076,
      "loss": 4.8102,
      "mean_token_accuracy": 0.23212082535028458,
      "num_tokens": 36877943.0,
      "step": 16085
    },
    {
      "entropy": 5.3122962474822994,
      "epoch": 1.5456292026897214,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047662916116663766,
      "loss": 4.8589,
      "mean_token_accuracy": 0.23288143873214723,
      "num_tokens": 36889343.0,
      "step": 16090
    },
    {
      "entropy": 5.321204042434692,
      "epoch": 1.5461095100864553,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047661394404330417,
      "loss": 4.8529,
      "mean_token_accuracy": 0.23032553046941756,
      "num_tokens": 36900727.0,
      "step": 16095
    },
    {
      "entropy": 5.221599054336548,
      "epoch": 1.5465898174831891,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047659872223909357,
      "loss": 4.8082,
      "mean_token_accuracy": 0.23681075125932693,
      "num_tokens": 36911959.0,
      "step": 16100
    },
    {
      "entropy": 5.233597326278686,
      "epoch": 1.547070124879923,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004765834957543592,
      "loss": 4.7613,
      "mean_token_accuracy": 0.23680628836154938,
      "num_tokens": 36924964.0,
      "step": 16105
    },
    {
      "entropy": 5.308232212066651,
      "epoch": 1.547550432276657,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047656826458945475,
      "loss": 4.9328,
      "mean_token_accuracy": 0.21821384876966476,
      "num_tokens": 36937033.0,
      "step": 16110
    },
    {
      "entropy": 5.269179821014404,
      "epoch": 1.548030739673391,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047655302874473365,
      "loss": 4.7953,
      "mean_token_accuracy": 0.2393047407269478,
      "num_tokens": 36948745.0,
      "step": 16115
    },
    {
      "entropy": 5.244120836257935,
      "epoch": 1.5485110470701249,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004765377882205498,
      "loss": 4.7613,
      "mean_token_accuracy": 0.23593603819608688,
      "num_tokens": 36961210.0,
      "step": 16120
    },
    {
      "entropy": 5.2728547096252445,
      "epoch": 1.5489913544668588,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004765225430172568,
      "loss": 4.8617,
      "mean_token_accuracy": 0.23709393590688704,
      "num_tokens": 36973720.0,
      "step": 16125
    },
    {
      "entropy": 5.331624984741211,
      "epoch": 1.5494716618635929,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004765072931352089,
      "loss": 4.9018,
      "mean_token_accuracy": 0.23136783242225648,
      "num_tokens": 36986899.0,
      "step": 16130
    },
    {
      "entropy": 5.2788454532623295,
      "epoch": 1.5499519692603267,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047649203857476,
      "loss": 4.811,
      "mean_token_accuracy": 0.2330128699541092,
      "num_tokens": 36998324.0,
      "step": 16135
    },
    {
      "entropy": 5.278350067138672,
      "epoch": 1.5504322766570606,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047647677933626423,
      "loss": 4.8844,
      "mean_token_accuracy": 0.234572134912014,
      "num_tokens": 37009779.0,
      "step": 16140
    },
    {
      "entropy": 5.301263332366943,
      "epoch": 1.5509125840537945,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047646151542007583,
      "loss": 4.885,
      "mean_token_accuracy": 0.23913077861070633,
      "num_tokens": 37022473.0,
      "step": 16145
    },
    {
      "entropy": 5.387713289260864,
      "epoch": 1.5513928914505284,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004764462468265494,
      "loss": 4.9667,
      "mean_token_accuracy": 0.22508137673139572,
      "num_tokens": 37033456.0,
      "step": 16150
    },
    {
      "entropy": 5.302484941482544,
      "epoch": 1.5518731988472623,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047643097355603913,
      "loss": 4.8763,
      "mean_token_accuracy": 0.22586306035518647,
      "num_tokens": 37044983.0,
      "step": 16155
    },
    {
      "entropy": 5.263723468780517,
      "epoch": 1.5523535062439962,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004764156956088999,
      "loss": 4.8055,
      "mean_token_accuracy": 0.2337541788816452,
      "num_tokens": 37055992.0,
      "step": 16160
    },
    {
      "entropy": 5.31610426902771,
      "epoch": 1.55283381364073,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004764004129854863,
      "loss": 4.8498,
      "mean_token_accuracy": 0.24172571301460266,
      "num_tokens": 37068227.0,
      "step": 16165
    },
    {
      "entropy": 5.237093448638916,
      "epoch": 1.553314121037464,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047638512568615307,
      "loss": 4.8627,
      "mean_token_accuracy": 0.2273416966199875,
      "num_tokens": 37080221.0,
      "step": 16170
    },
    {
      "entropy": 5.2570640563964846,
      "epoch": 1.5537944284341978,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004763698337112553,
      "loss": 4.8437,
      "mean_token_accuracy": 0.23079460710287095,
      "num_tokens": 37092223.0,
      "step": 16175
    },
    {
      "entropy": 5.332653903961182,
      "epoch": 1.5542747358309317,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004763545370611479,
      "loss": 4.8671,
      "mean_token_accuracy": 0.22601209282875062,
      "num_tokens": 37104169.0,
      "step": 16180
    },
    {
      "entropy": 5.305645513534546,
      "epoch": 1.5547550432276656,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047633923573618605,
      "loss": 4.8961,
      "mean_token_accuracy": 0.23408455401659012,
      "num_tokens": 37115784.0,
      "step": 16185
    },
    {
      "entropy": 5.214728546142578,
      "epoch": 1.5552353506243997,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000476323929736725,
      "loss": 4.7382,
      "mean_token_accuracy": 0.23884457498788833,
      "num_tokens": 37128018.0,
      "step": 16190
    },
    {
      "entropy": 5.273625469207763,
      "epoch": 1.5557156580211335,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047630861906312004,
      "loss": 4.7982,
      "mean_token_accuracy": 0.22727231532335282,
      "num_tokens": 37139330.0,
      "step": 16195
    },
    {
      "entropy": 5.178512334823608,
      "epoch": 1.5561959654178674,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004762933037157268,
      "loss": 4.7061,
      "mean_token_accuracy": 0.24603790193796157,
      "num_tokens": 37151531.0,
      "step": 16200
    },
    {
      "entropy": 5.343041610717774,
      "epoch": 1.5566762728146013,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047627798369490076,
      "loss": 4.8599,
      "mean_token_accuracy": 0.23688916265964508,
      "num_tokens": 37162442.0,
      "step": 16205
    },
    {
      "entropy": 5.27861967086792,
      "epoch": 1.5571565802113354,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047626265900099757,
      "loss": 4.8302,
      "mean_token_accuracy": 0.23903474658727647,
      "num_tokens": 37173737.0,
      "step": 16210
    },
    {
      "entropy": 5.214116096496582,
      "epoch": 1.5576368876080693,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047624732963437314,
      "loss": 4.7937,
      "mean_token_accuracy": 0.23798494637012482,
      "num_tokens": 37185983.0,
      "step": 16215
    },
    {
      "entropy": 5.330527305603027,
      "epoch": 1.5581171950048032,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047623199559538324,
      "loss": 4.9273,
      "mean_token_accuracy": 0.22173206955194474,
      "num_tokens": 37197119.0,
      "step": 16220
    },
    {
      "entropy": 5.406496667861939,
      "epoch": 1.558597502401537,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000476216656884384,
      "loss": 4.9717,
      "mean_token_accuracy": 0.22645564675331115,
      "num_tokens": 37209516.0,
      "step": 16225
    },
    {
      "entropy": 5.204306364059448,
      "epoch": 1.559077809798271,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047620131350173135,
      "loss": 4.7968,
      "mean_token_accuracy": 0.23470272421836852,
      "num_tokens": 37220559.0,
      "step": 16230
    },
    {
      "entropy": 5.335011196136475,
      "epoch": 1.5595581171950048,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004761859654477817,
      "loss": 4.8522,
      "mean_token_accuracy": 0.2340918704867363,
      "num_tokens": 37232232.0,
      "step": 16235
    },
    {
      "entropy": 5.184920692443848,
      "epoch": 1.5600384245917387,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004761706127228914,
      "loss": 4.7499,
      "mean_token_accuracy": 0.23985494971275328,
      "num_tokens": 37242819.0,
      "step": 16240
    },
    {
      "entropy": 5.27667875289917,
      "epoch": 1.5605187319884726,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004761552553274168,
      "loss": 4.8506,
      "mean_token_accuracy": 0.23264015167951585,
      "num_tokens": 37254940.0,
      "step": 16245
    },
    {
      "entropy": 5.355356740951538,
      "epoch": 1.5609990393852065,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004761398932617144,
      "loss": 4.9081,
      "mean_token_accuracy": 0.22893626987934113,
      "num_tokens": 37266220.0,
      "step": 16250
    },
    {
      "entropy": 5.32624192237854,
      "epoch": 1.5614793467819403,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000476124526526141,
      "loss": 4.8866,
      "mean_token_accuracy": 0.2279340296983719,
      "num_tokens": 37278136.0,
      "step": 16255
    },
    {
      "entropy": 5.292405319213867,
      "epoch": 1.5619596541786742,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047610915512105327,
      "loss": 4.8454,
      "mean_token_accuracy": 0.23132514357566833,
      "num_tokens": 37290358.0,
      "step": 16260
    },
    {
      "entropy": 5.371146821975708,
      "epoch": 1.562439961575408,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004760937790468082,
      "loss": 4.9308,
      "mean_token_accuracy": 0.23171036690473557,
      "num_tokens": 37300719.0,
      "step": 16265
    },
    {
      "entropy": 5.262251186370849,
      "epoch": 1.5629202689721422,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004760783983037627,
      "loss": 4.8749,
      "mean_token_accuracy": 0.22787132263183593,
      "num_tokens": 37311917.0,
      "step": 16270
    },
    {
      "entropy": 5.2522783279418945,
      "epoch": 1.563400576368876,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004760630128922738,
      "loss": 4.8309,
      "mean_token_accuracy": 0.23736237734556198,
      "num_tokens": 37325076.0,
      "step": 16275
    },
    {
      "entropy": 5.321431589126587,
      "epoch": 1.56388088376561,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004760476228126989,
      "loss": 4.8547,
      "mean_token_accuracy": 0.22863038182258605,
      "num_tokens": 37336064.0,
      "step": 16280
    },
    {
      "entropy": 5.315920877456665,
      "epoch": 1.5643611911623438,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004760322280653951,
      "loss": 4.8466,
      "mean_token_accuracy": 0.2350413665175438,
      "num_tokens": 37347343.0,
      "step": 16285
    },
    {
      "entropy": 5.274502420425415,
      "epoch": 1.564841498559078,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004760168286507199,
      "loss": 4.798,
      "mean_token_accuracy": 0.24012629240751265,
      "num_tokens": 37359067.0,
      "step": 16290
    },
    {
      "entropy": 5.2849808692932125,
      "epoch": 1.5653218059558118,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047600142456903085,
      "loss": 4.8403,
      "mean_token_accuracy": 0.22946687638759614,
      "num_tokens": 37369811.0,
      "step": 16295
    },
    {
      "entropy": 5.173256254196167,
      "epoch": 1.5658021133525457,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047598601582068555,
      "loss": 4.6998,
      "mean_token_accuracy": 0.246090866625309,
      "num_tokens": 37380944.0,
      "step": 16300
    },
    {
      "entropy": 5.298480463027954,
      "epoch": 1.5662824207492796,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004759706024060418,
      "loss": 4.883,
      "mean_token_accuracy": 0.2183980479836464,
      "num_tokens": 37392099.0,
      "step": 16305
    },
    {
      "entropy": 5.2360748767852785,
      "epoch": 1.5667627281460135,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004759551843254575,
      "loss": 4.8343,
      "mean_token_accuracy": 0.23808059245347976,
      "num_tokens": 37404254.0,
      "step": 16310
    },
    {
      "entropy": 5.24685115814209,
      "epoch": 1.5672430355427474,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047593976157929034,
      "loss": 4.7812,
      "mean_token_accuracy": 0.23662374019622803,
      "num_tokens": 37416112.0,
      "step": 16315
    },
    {
      "entropy": 5.226862668991089,
      "epoch": 1.5677233429394812,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004759243341678987,
      "loss": 4.7915,
      "mean_token_accuracy": 0.2387004852294922,
      "num_tokens": 37428350.0,
      "step": 16320
    },
    {
      "entropy": 5.284838771820068,
      "epoch": 1.5682036503362151,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004759089020916407,
      "loss": 4.8869,
      "mean_token_accuracy": 0.23095106035470964,
      "num_tokens": 37440059.0,
      "step": 16325
    },
    {
      "entropy": 5.357075977325439,
      "epoch": 1.568683957732949,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047589346535087444,
      "loss": 4.8849,
      "mean_token_accuracy": 0.228093121945858,
      "num_tokens": 37450880.0,
      "step": 16330
    },
    {
      "entropy": 5.248473834991455,
      "epoch": 1.5691642651296829,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004758780239459586,
      "loss": 4.8022,
      "mean_token_accuracy": 0.24024815410375594,
      "num_tokens": 37463484.0,
      "step": 16335
    },
    {
      "entropy": 5.2829248905181885,
      "epoch": 1.5696445725264168,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004758625778772514,
      "loss": 4.8283,
      "mean_token_accuracy": 0.23267146199941635,
      "num_tokens": 37474952.0,
      "step": 16340
    },
    {
      "entropy": 5.32055549621582,
      "epoch": 1.5701248799231509,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047584712714511166,
      "loss": 4.9232,
      "mean_token_accuracy": 0.22827216535806655,
      "num_tokens": 37486858.0,
      "step": 16345
    },
    {
      "entropy": 5.2834312915802,
      "epoch": 1.5706051873198847,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047583167174989797,
      "loss": 4.7921,
      "mean_token_accuracy": 0.2348979413509369,
      "num_tokens": 37497852.0,
      "step": 16350
    },
    {
      "entropy": 5.276198530197144,
      "epoch": 1.5710854947166186,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004758162116919692,
      "loss": 4.8995,
      "mean_token_accuracy": 0.2292618602514267,
      "num_tokens": 37509051.0,
      "step": 16355
    },
    {
      "entropy": 5.345890378952026,
      "epoch": 1.5715658021133525,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047580074697168434,
      "loss": 4.9755,
      "mean_token_accuracy": 0.23094065934419633,
      "num_tokens": 37520415.0,
      "step": 16360
    },
    {
      "entropy": 5.297778701782226,
      "epoch": 1.5720461095100866,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047578527758940236,
      "loss": 4.8692,
      "mean_token_accuracy": 0.23180274814367294,
      "num_tokens": 37532087.0,
      "step": 16365
    },
    {
      "entropy": 5.30827054977417,
      "epoch": 1.5725264169068205,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004757698035454825,
      "loss": 4.891,
      "mean_token_accuracy": 0.22990359216928483,
      "num_tokens": 37545219.0,
      "step": 16370
    },
    {
      "entropy": 5.343153572082519,
      "epoch": 1.5730067243035544,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004757543248402839,
      "loss": 4.8209,
      "mean_token_accuracy": 0.23362624794244766,
      "num_tokens": 37558138.0,
      "step": 16375
    },
    {
      "entropy": 5.2158032894134525,
      "epoch": 1.5734870317002883,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047573884147416597,
      "loss": 4.7249,
      "mean_token_accuracy": 0.23674222230911254,
      "num_tokens": 37568975.0,
      "step": 16380
    },
    {
      "entropy": 5.233182048797607,
      "epoch": 1.5739673390970221,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004757233534474883,
      "loss": 4.7579,
      "mean_token_accuracy": 0.239777509868145,
      "num_tokens": 37580774.0,
      "step": 16385
    },
    {
      "entropy": 5.248457670211792,
      "epoch": 1.574447646493756,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004757078607606103,
      "loss": 4.7797,
      "mean_token_accuracy": 0.23446904718875886,
      "num_tokens": 37592858.0,
      "step": 16390
    },
    {
      "entropy": 5.289395713806153,
      "epoch": 1.57492795389049,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004756923634138918,
      "loss": 4.8703,
      "mean_token_accuracy": 0.2285671830177307,
      "num_tokens": 37604005.0,
      "step": 16395
    },
    {
      "entropy": 5.280482625961303,
      "epoch": 1.5754082612872238,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047567686140769264,
      "loss": 4.8125,
      "mean_token_accuracy": 0.23447229862213134,
      "num_tokens": 37614629.0,
      "step": 16400
    },
    {
      "entropy": 5.288556814193726,
      "epoch": 1.5758885686839577,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047566135474237247,
      "loss": 4.8092,
      "mean_token_accuracy": 0.23866589963436127,
      "num_tokens": 37626118.0,
      "step": 16405
    },
    {
      "entropy": 5.352845954895019,
      "epoch": 1.5763688760806915,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047564584341829166,
      "loss": 4.9003,
      "mean_token_accuracy": 0.22150313407182692,
      "num_tokens": 37638514.0,
      "step": 16410
    },
    {
      "entropy": 5.326574325561523,
      "epoch": 1.5768491834774254,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047563032743581,
      "loss": 4.9138,
      "mean_token_accuracy": 0.2237405449151993,
      "num_tokens": 37649602.0,
      "step": 16415
    },
    {
      "entropy": 5.295902252197266,
      "epoch": 1.5773294908741593,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047561480679528804,
      "loss": 4.8638,
      "mean_token_accuracy": 0.23048643618822098,
      "num_tokens": 37661428.0,
      "step": 16420
    },
    {
      "entropy": 5.325285720825195,
      "epoch": 1.5778097982708934,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004755992814970859,
      "loss": 4.8692,
      "mean_token_accuracy": 0.23158176690340043,
      "num_tokens": 37673608.0,
      "step": 16425
    },
    {
      "entropy": 5.361770868301392,
      "epoch": 1.5782901056676273,
      "grad_norm": 1.046875,
      "learning_rate": 0.000475583751541564,
      "loss": 4.9911,
      "mean_token_accuracy": 0.2259400874376297,
      "num_tokens": 37684957.0,
      "step": 16430
    },
    {
      "entropy": 5.243374061584473,
      "epoch": 1.5787704130643612,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047556821692908315,
      "loss": 4.8539,
      "mean_token_accuracy": 0.22926443666219712,
      "num_tokens": 37697773.0,
      "step": 16435
    },
    {
      "entropy": 5.31532850265503,
      "epoch": 1.579250720461095,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004755526776600038,
      "loss": 4.8771,
      "mean_token_accuracy": 0.2342222586274147,
      "num_tokens": 37708315.0,
      "step": 16440
    },
    {
      "entropy": 5.281141042709351,
      "epoch": 1.5797310278578292,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047553713373468684,
      "loss": 4.8466,
      "mean_token_accuracy": 0.23364761471748352,
      "num_tokens": 37720440.0,
      "step": 16445
    },
    {
      "entropy": 5.152264070510864,
      "epoch": 1.580211335254563,
      "grad_norm": 1.125,
      "learning_rate": 0.00047552158515349306,
      "loss": 4.7375,
      "mean_token_accuracy": 0.23636302798986436,
      "num_tokens": 37732332.0,
      "step": 16450
    },
    {
      "entropy": 5.252034139633179,
      "epoch": 1.580691642651297,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047550603191678356,
      "loss": 4.8439,
      "mean_token_accuracy": 0.22869304716587066,
      "num_tokens": 37744770.0,
      "step": 16455
    },
    {
      "entropy": 5.246133136749267,
      "epoch": 1.5811719500480308,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004754904740249194,
      "loss": 4.7717,
      "mean_token_accuracy": 0.23857891857624053,
      "num_tokens": 37755958.0,
      "step": 16460
    },
    {
      "entropy": 5.287131834030151,
      "epoch": 1.5816522574447647,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047547491147826156,
      "loss": 4.8023,
      "mean_token_accuracy": 0.2371246486902237,
      "num_tokens": 37767182.0,
      "step": 16465
    },
    {
      "entropy": 5.3246485710144045,
      "epoch": 1.5821325648414986,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004754593442771718,
      "loss": 4.9071,
      "mean_token_accuracy": 0.2292221188545227,
      "num_tokens": 37778299.0,
      "step": 16470
    },
    {
      "entropy": 5.3743329524993895,
      "epoch": 1.5826128722382324,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047544377242201115,
      "loss": 4.9272,
      "mean_token_accuracy": 0.22904543876647948,
      "num_tokens": 37789879.0,
      "step": 16475
    },
    {
      "entropy": 5.334837293624878,
      "epoch": 1.5830931796349663,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047542819591314136,
      "loss": 4.8718,
      "mean_token_accuracy": 0.22846264094114305,
      "num_tokens": 37800726.0,
      "step": 16480
    },
    {
      "entropy": 5.2914385318756105,
      "epoch": 1.5835734870317002,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004754126147509241,
      "loss": 4.8428,
      "mean_token_accuracy": 0.23620485365390778,
      "num_tokens": 37812298.0,
      "step": 16485
    },
    {
      "entropy": 5.179229784011841,
      "epoch": 1.584053794428434,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047539702893572086,
      "loss": 4.7339,
      "mean_token_accuracy": 0.242834235727787,
      "num_tokens": 37824433.0,
      "step": 16490
    },
    {
      "entropy": 5.278210163116455,
      "epoch": 1.584534101825168,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047538143846789376,
      "loss": 4.8605,
      "mean_token_accuracy": 0.23785120099782944,
      "num_tokens": 37835647.0,
      "step": 16495
    },
    {
      "entropy": 5.329925155639648,
      "epoch": 1.585014409221902,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004753658433478047,
      "loss": 4.9171,
      "mean_token_accuracy": 0.23150623738765716,
      "num_tokens": 37847721.0,
      "step": 16500
    },
    {
      "entropy": 5.2358307361602785,
      "epoch": 1.585494716618636,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047535024357581564,
      "loss": 4.7432,
      "mean_token_accuracy": 0.23617057651281356,
      "num_tokens": 37860111.0,
      "step": 16505
    },
    {
      "entropy": 5.260824060440063,
      "epoch": 1.5859750240153698,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004753346391522889,
      "loss": 4.8966,
      "mean_token_accuracy": 0.2243965223431587,
      "num_tokens": 37872321.0,
      "step": 16510
    },
    {
      "entropy": 5.348930835723877,
      "epoch": 1.5864553314121037,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047531903007758667,
      "loss": 4.9633,
      "mean_token_accuracy": 0.22363830357789993,
      "num_tokens": 37885195.0,
      "step": 16515
    },
    {
      "entropy": 5.3212896347045895,
      "epoch": 1.5869356388088378,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004753034163520714,
      "loss": 4.8088,
      "mean_token_accuracy": 0.23546413481235504,
      "num_tokens": 37895487.0,
      "step": 16520
    },
    {
      "entropy": 5.207644748687744,
      "epoch": 1.5874159462055717,
      "grad_norm": 1.0,
      "learning_rate": 0.00047528779797610557,
      "loss": 4.891,
      "mean_token_accuracy": 0.22984413504600526,
      "num_tokens": 37908344.0,
      "step": 16525
    },
    {
      "entropy": 5.276592683792114,
      "epoch": 1.5878962536023056,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047527217495005184,
      "loss": 4.813,
      "mean_token_accuracy": 0.23313235342502595,
      "num_tokens": 37920220.0,
      "step": 16530
    },
    {
      "entropy": 5.235566902160644,
      "epoch": 1.5883765609990395,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047525654727427285,
      "loss": 4.662,
      "mean_token_accuracy": 0.24705688804388046,
      "num_tokens": 37930874.0,
      "step": 16535
    },
    {
      "entropy": 5.231817626953125,
      "epoch": 1.5888568683957733,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004752409149491315,
      "loss": 4.8113,
      "mean_token_accuracy": 0.2331054389476776,
      "num_tokens": 37942501.0,
      "step": 16540
    },
    {
      "entropy": 5.2487547397613525,
      "epoch": 1.5893371757925072,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047522527797499075,
      "loss": 4.8082,
      "mean_token_accuracy": 0.23301116079092027,
      "num_tokens": 37954408.0,
      "step": 16545
    },
    {
      "entropy": 5.309425354003906,
      "epoch": 1.589817483189241,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004752096363522135,
      "loss": 4.8551,
      "mean_token_accuracy": 0.2288050666451454,
      "num_tokens": 37966088.0,
      "step": 16550
    },
    {
      "entropy": 5.304376459121704,
      "epoch": 1.590297790585975,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047519399008116305,
      "loss": 4.8536,
      "mean_token_accuracy": 0.22188859432935715,
      "num_tokens": 37977726.0,
      "step": 16555
    },
    {
      "entropy": 5.279466962814331,
      "epoch": 1.5907780979827089,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004751783391622026,
      "loss": 4.8766,
      "mean_token_accuracy": 0.22447636276483535,
      "num_tokens": 37989024.0,
      "step": 16560
    },
    {
      "entropy": 5.3251718044281,
      "epoch": 1.5912584053794427,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004751626835956955,
      "loss": 4.9014,
      "mean_token_accuracy": 0.22831487506628037,
      "num_tokens": 37999870.0,
      "step": 16565
    },
    {
      "entropy": 5.340056419372559,
      "epoch": 1.5917387127761766,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004751470233820053,
      "loss": 4.9216,
      "mean_token_accuracy": 0.22760978639125823,
      "num_tokens": 38010945.0,
      "step": 16570
    },
    {
      "entropy": 5.262466287612915,
      "epoch": 1.5922190201729105,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004751313585214955,
      "loss": 4.8138,
      "mean_token_accuracy": 0.235890132188797,
      "num_tokens": 38021584.0,
      "step": 16575
    },
    {
      "entropy": 5.291036748886109,
      "epoch": 1.5926993275696446,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004751156890145298,
      "loss": 4.8942,
      "mean_token_accuracy": 0.22563716471195222,
      "num_tokens": 38034407.0,
      "step": 16580
    },
    {
      "entropy": 5.345171546936035,
      "epoch": 1.5931796349663785,
      "grad_norm": 1.0625,
      "learning_rate": 0.000475100014861472,
      "loss": 4.8976,
      "mean_token_accuracy": 0.2261899843811989,
      "num_tokens": 38047061.0,
      "step": 16585
    },
    {
      "entropy": 5.417765140533447,
      "epoch": 1.5936599423631124,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004750843360626861,
      "loss": 4.9446,
      "mean_token_accuracy": 0.23201363831758498,
      "num_tokens": 38059168.0,
      "step": 16590
    },
    {
      "entropy": 5.3309228897094725,
      "epoch": 1.5941402497598463,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004750686526185359,
      "loss": 4.8192,
      "mean_token_accuracy": 0.23829794377088548,
      "num_tokens": 38070514.0,
      "step": 16595
    },
    {
      "entropy": 5.380277633666992,
      "epoch": 1.5946205571565804,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047505296452938584,
      "loss": 4.9304,
      "mean_token_accuracy": 0.22936539053916932,
      "num_tokens": 38081995.0,
      "step": 16600
    },
    {
      "entropy": 5.25193829536438,
      "epoch": 1.5951008645533142,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047503727179559995,
      "loss": 4.7968,
      "mean_token_accuracy": 0.241678586602211,
      "num_tokens": 38094371.0,
      "step": 16605
    },
    {
      "entropy": 5.209079933166504,
      "epoch": 1.5955811719500481,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047502157441754256,
      "loss": 4.769,
      "mean_token_accuracy": 0.23597528040409088,
      "num_tokens": 38106340.0,
      "step": 16610
    },
    {
      "entropy": 5.297894668579102,
      "epoch": 1.596061479346782,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004750058723955781,
      "loss": 4.8274,
      "mean_token_accuracy": 0.23046587258577347,
      "num_tokens": 38117641.0,
      "step": 16615
    },
    {
      "entropy": 5.342281913757324,
      "epoch": 1.5965417867435159,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004749901657300713,
      "loss": 4.828,
      "mean_token_accuracy": 0.23156410604715347,
      "num_tokens": 38129628.0,
      "step": 16620
    },
    {
      "entropy": 5.2355721950531,
      "epoch": 1.5970220941402498,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047497445442138667,
      "loss": 4.7978,
      "mean_token_accuracy": 0.2323486849665642,
      "num_tokens": 38140843.0,
      "step": 16625
    },
    {
      "entropy": 5.239694690704345,
      "epoch": 1.5975024015369836,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047495873846988896,
      "loss": 4.7695,
      "mean_token_accuracy": 0.24266054928302766,
      "num_tokens": 38152666.0,
      "step": 16630
    },
    {
      "entropy": 5.23565821647644,
      "epoch": 1.5979827089337175,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004749430178759431,
      "loss": 4.7608,
      "mean_token_accuracy": 0.24042134433984758,
      "num_tokens": 38164102.0,
      "step": 16635
    },
    {
      "entropy": 5.252330350875854,
      "epoch": 1.5984630163304514,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047492729263991413,
      "loss": 4.8163,
      "mean_token_accuracy": 0.23836100697517396,
      "num_tokens": 38175211.0,
      "step": 16640
    },
    {
      "entropy": 5.26444787979126,
      "epoch": 1.5989433237271853,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047491156276216695,
      "loss": 4.8039,
      "mean_token_accuracy": 0.24152205884456635,
      "num_tokens": 38186968.0,
      "step": 16645
    },
    {
      "entropy": 5.30109133720398,
      "epoch": 1.5994236311239192,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047489582824306704,
      "loss": 4.8378,
      "mean_token_accuracy": 0.22817151844501496,
      "num_tokens": 38198740.0,
      "step": 16650
    },
    {
      "entropy": 5.2768653392791744,
      "epoch": 1.5999039385206533,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047488008908297955,
      "loss": 4.7642,
      "mean_token_accuracy": 0.2458020642399788,
      "num_tokens": 38209110.0,
      "step": 16655
    },
    {
      "entropy": 5.213055515289307,
      "epoch": 1.6003842459173871,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004748643452822699,
      "loss": 4.7488,
      "mean_token_accuracy": 0.23486157655715942,
      "num_tokens": 38221840.0,
      "step": 16660
    },
    {
      "entropy": 5.335714817047119,
      "epoch": 1.600864553314121,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004748485968413036,
      "loss": 4.8805,
      "mean_token_accuracy": 0.22909073084592818,
      "num_tokens": 38232752.0,
      "step": 16665
    },
    {
      "entropy": 5.259540033340454,
      "epoch": 1.601344860710855,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047483284376044634,
      "loss": 4.8235,
      "mean_token_accuracy": 0.22780465632677077,
      "num_tokens": 38243394.0,
      "step": 16670
    },
    {
      "entropy": 5.3146092891693115,
      "epoch": 1.601825168107589,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004748170860400638,
      "loss": 4.917,
      "mean_token_accuracy": 0.23039546757936477,
      "num_tokens": 38254317.0,
      "step": 16675
    },
    {
      "entropy": 5.2726329326629635,
      "epoch": 1.602305475504323,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047480132368052185,
      "loss": 4.7713,
      "mean_token_accuracy": 0.24397177547216414,
      "num_tokens": 38266886.0,
      "step": 16680
    },
    {
      "entropy": 5.3518760204315186,
      "epoch": 1.6027857829010568,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047478555668218643,
      "loss": 4.9697,
      "mean_token_accuracy": 0.22689229100942612,
      "num_tokens": 38279975.0,
      "step": 16685
    },
    {
      "entropy": 5.335694217681885,
      "epoch": 1.6032660902977907,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004747697850454237,
      "loss": 4.9324,
      "mean_token_accuracy": 0.22228170782327653,
      "num_tokens": 38290237.0,
      "step": 16690
    },
    {
      "entropy": 5.3273704051971436,
      "epoch": 1.6037463976945245,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004747540087705997,
      "loss": 4.7574,
      "mean_token_accuracy": 0.2411135420203209,
      "num_tokens": 38300833.0,
      "step": 16695
    },
    {
      "entropy": 5.233847522735596,
      "epoch": 1.6042267050912584,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004747382278580808,
      "loss": 4.7554,
      "mean_token_accuracy": 0.23897157460451127,
      "num_tokens": 38311926.0,
      "step": 16700
    },
    {
      "entropy": 5.3257678031921385,
      "epoch": 1.6047070124879923,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004747224423082333,
      "loss": 4.9438,
      "mean_token_accuracy": 0.22443341165781022,
      "num_tokens": 38323053.0,
      "step": 16705
    },
    {
      "entropy": 5.332638883590699,
      "epoch": 1.6051873198847262,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047470665212142384,
      "loss": 4.9101,
      "mean_token_accuracy": 0.2259608030319214,
      "num_tokens": 38334959.0,
      "step": 16710
    },
    {
      "entropy": 5.350841808319092,
      "epoch": 1.60566762728146,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047469085729801887,
      "loss": 4.8946,
      "mean_token_accuracy": 0.22436099201440812,
      "num_tokens": 38346402.0,
      "step": 16715
    },
    {
      "entropy": 5.316497707366944,
      "epoch": 1.606147934678194,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047467505783838515,
      "loss": 4.8485,
      "mean_token_accuracy": 0.22884937971830369,
      "num_tokens": 38358549.0,
      "step": 16720
    },
    {
      "entropy": 5.3227774143219,
      "epoch": 1.6066282420749278,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004746592537428895,
      "loss": 4.8491,
      "mean_token_accuracy": 0.23267987221479416,
      "num_tokens": 38370840.0,
      "step": 16725
    },
    {
      "entropy": 5.285579824447632,
      "epoch": 1.6071085494716617,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047464344501189877,
      "loss": 4.8471,
      "mean_token_accuracy": 0.22761898189783097,
      "num_tokens": 38382980.0,
      "step": 16730
    },
    {
      "entropy": 5.308327722549438,
      "epoch": 1.6075888568683958,
      "grad_norm": 1.0,
      "learning_rate": 0.00047462763164578015,
      "loss": 4.7767,
      "mean_token_accuracy": 0.24118322134017944,
      "num_tokens": 38393836.0,
      "step": 16735
    },
    {
      "entropy": 5.202812910079956,
      "epoch": 1.6080691642651297,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004746118136449007,
      "loss": 4.7597,
      "mean_token_accuracy": 0.23757877349853515,
      "num_tokens": 38405770.0,
      "step": 16740
    },
    {
      "entropy": 5.279361295700073,
      "epoch": 1.6085494716618636,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004745959910096276,
      "loss": 4.9252,
      "mean_token_accuracy": 0.22540036141872405,
      "num_tokens": 38418847.0,
      "step": 16745
    },
    {
      "entropy": 5.239596223831176,
      "epoch": 1.6090297790585975,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047458016374032837,
      "loss": 4.7593,
      "mean_token_accuracy": 0.2380074992775917,
      "num_tokens": 38431532.0,
      "step": 16750
    },
    {
      "entropy": 5.3871252059936525,
      "epoch": 1.6095100864553316,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004745643318373703,
      "loss": 4.986,
      "mean_token_accuracy": 0.22898949533700944,
      "num_tokens": 38442071.0,
      "step": 16755
    },
    {
      "entropy": 5.279758405685425,
      "epoch": 1.6099903938520654,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047454849530112106,
      "loss": 4.8919,
      "mean_token_accuracy": 0.22520833760499953,
      "num_tokens": 38454979.0,
      "step": 16760
    },
    {
      "entropy": 5.223357629776001,
      "epoch": 1.6104707012487993,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047453265413194826,
      "loss": 4.7366,
      "mean_token_accuracy": 0.24730508476495744,
      "num_tokens": 38465579.0,
      "step": 16765
    },
    {
      "entropy": 5.222699356079102,
      "epoch": 1.6109510086455332,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047451680833021973,
      "loss": 4.814,
      "mean_token_accuracy": 0.23311687558889388,
      "num_tokens": 38477139.0,
      "step": 16770
    },
    {
      "entropy": 5.265465354919433,
      "epoch": 1.611431316042267,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004745009578963034,
      "loss": 4.7695,
      "mean_token_accuracy": 0.237275393307209,
      "num_tokens": 38487638.0,
      "step": 16775
    },
    {
      "entropy": 5.126733589172363,
      "epoch": 1.611911623439001,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047448510283056716,
      "loss": 4.6448,
      "mean_token_accuracy": 0.25042389780282975,
      "num_tokens": 38498395.0,
      "step": 16780
    },
    {
      "entropy": 5.27101583480835,
      "epoch": 1.6123919308357348,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047446924313337925,
      "loss": 4.7551,
      "mean_token_accuracy": 0.23945056051015853,
      "num_tokens": 38509372.0,
      "step": 16785
    },
    {
      "entropy": 5.281351232528687,
      "epoch": 1.6128722382324687,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047445337880510773,
      "loss": 4.8017,
      "mean_token_accuracy": 0.23201658576726913,
      "num_tokens": 38520651.0,
      "step": 16790
    },
    {
      "entropy": 5.307695579528809,
      "epoch": 1.6133525456292026,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004744375098461211,
      "loss": 4.9131,
      "mean_token_accuracy": 0.2256884902715683,
      "num_tokens": 38531461.0,
      "step": 16795
    },
    {
      "entropy": 5.247994041442871,
      "epoch": 1.6138328530259365,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004744216362567876,
      "loss": 4.7364,
      "mean_token_accuracy": 0.24733877629041673,
      "num_tokens": 38542707.0,
      "step": 16800
    },
    {
      "entropy": 5.270821857452392,
      "epoch": 1.6143131604226704,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047440575803747595,
      "loss": 4.7555,
      "mean_token_accuracy": 0.238628089427948,
      "num_tokens": 38553664.0,
      "step": 16805
    },
    {
      "entropy": 5.298038005828857,
      "epoch": 1.6147934678194045,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047438987518855463,
      "loss": 4.825,
      "mean_token_accuracy": 0.22924889177083968,
      "num_tokens": 38564891.0,
      "step": 16810
    },
    {
      "entropy": 5.25895881652832,
      "epoch": 1.6152737752161384,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004743739877103926,
      "loss": 4.8362,
      "mean_token_accuracy": 0.2294268324971199,
      "num_tokens": 38576535.0,
      "step": 16815
    },
    {
      "entropy": 5.293157386779785,
      "epoch": 1.6157540826128722,
      "grad_norm": 1.0,
      "learning_rate": 0.0004743580956033585,
      "loss": 4.8374,
      "mean_token_accuracy": 0.22955019026994705,
      "num_tokens": 38588160.0,
      "step": 16820
    },
    {
      "entropy": 5.236642646789551,
      "epoch": 1.6162343900096061,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047434219886782135,
      "loss": 4.7539,
      "mean_token_accuracy": 0.2365494728088379,
      "num_tokens": 38599281.0,
      "step": 16825
    },
    {
      "entropy": 5.2818420886993405,
      "epoch": 1.6167146974063402,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004743262975041504,
      "loss": 4.8447,
      "mean_token_accuracy": 0.23492764234542846,
      "num_tokens": 38611823.0,
      "step": 16830
    },
    {
      "entropy": 5.27856936454773,
      "epoch": 1.617195004803074,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004743103915127146,
      "loss": 4.8372,
      "mean_token_accuracy": 0.2312195137143135,
      "num_tokens": 38623181.0,
      "step": 16835
    },
    {
      "entropy": 5.29249529838562,
      "epoch": 1.617675312199808,
      "grad_norm": 1.0,
      "learning_rate": 0.00047429448089388336,
      "loss": 4.8463,
      "mean_token_accuracy": 0.23271494060754777,
      "num_tokens": 38633149.0,
      "step": 16840
    },
    {
      "entropy": 5.280625247955323,
      "epoch": 1.6181556195965419,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047427856564802605,
      "loss": 4.8278,
      "mean_token_accuracy": 0.23377834260463715,
      "num_tokens": 38644204.0,
      "step": 16845
    },
    {
      "entropy": 5.339758253097534,
      "epoch": 1.6186359269932757,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004742626457755122,
      "loss": 4.8087,
      "mean_token_accuracy": 0.23879957050085068,
      "num_tokens": 38655287.0,
      "step": 16850
    },
    {
      "entropy": 5.220594358444214,
      "epoch": 1.6191162343900096,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004742467212767114,
      "loss": 4.7222,
      "mean_token_accuracy": 0.2392561301589012,
      "num_tokens": 38665732.0,
      "step": 16855
    },
    {
      "entropy": 5.305705451965332,
      "epoch": 1.6195965417867435,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004742307921519933,
      "loss": 4.9196,
      "mean_token_accuracy": 0.21834482997655869,
      "num_tokens": 38677289.0,
      "step": 16860
    },
    {
      "entropy": 5.278922843933105,
      "epoch": 1.6200768491834774,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047421485840172794,
      "loss": 4.8041,
      "mean_token_accuracy": 0.2338162049651146,
      "num_tokens": 38689581.0,
      "step": 16865
    },
    {
      "entropy": 5.336023664474487,
      "epoch": 1.6205571565802113,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004741989200262851,
      "loss": 4.8771,
      "mean_token_accuracy": 0.22397912740707399,
      "num_tokens": 38701327.0,
      "step": 16870
    },
    {
      "entropy": 5.293549728393555,
      "epoch": 1.6210374639769451,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004741829770260347,
      "loss": 4.8638,
      "mean_token_accuracy": 0.2303344264626503,
      "num_tokens": 38713431.0,
      "step": 16875
    },
    {
      "entropy": 5.31175422668457,
      "epoch": 1.621517771373679,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047416702940134714,
      "loss": 4.8783,
      "mean_token_accuracy": 0.23053094297647475,
      "num_tokens": 38724091.0,
      "step": 16880
    },
    {
      "entropy": 5.292208051681518,
      "epoch": 1.621998078770413,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047415107715259255,
      "loss": 4.9167,
      "mean_token_accuracy": 0.22979729026556014,
      "num_tokens": 38736219.0,
      "step": 16885
    },
    {
      "entropy": 5.248180532455445,
      "epoch": 1.622478386167147,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047413512028014125,
      "loss": 4.7822,
      "mean_token_accuracy": 0.2350028231739998,
      "num_tokens": 38747512.0,
      "step": 16890
    },
    {
      "entropy": 5.334419298171997,
      "epoch": 1.622958693563881,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004741191587843638,
      "loss": 4.9178,
      "mean_token_accuracy": 0.22611849755048752,
      "num_tokens": 38758777.0,
      "step": 16895
    },
    {
      "entropy": 5.314997005462646,
      "epoch": 1.6234390009606148,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004741031926656308,
      "loss": 4.8671,
      "mean_token_accuracy": 0.23647585958242417,
      "num_tokens": 38770094.0,
      "step": 16900
    },
    {
      "entropy": 5.342555236816406,
      "epoch": 1.6239193083573487,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004740872219243128,
      "loss": 4.8203,
      "mean_token_accuracy": 0.23662331700325012,
      "num_tokens": 38781419.0,
      "step": 16905
    },
    {
      "entropy": 5.262478399276733,
      "epoch": 1.6243996157540828,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004740712465607807,
      "loss": 4.8169,
      "mean_token_accuracy": 0.2306710287928581,
      "num_tokens": 38794303.0,
      "step": 16910
    },
    {
      "entropy": 5.251261758804321,
      "epoch": 1.6248799231508166,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004740552665754054,
      "loss": 4.8665,
      "mean_token_accuracy": 0.23374342769384385,
      "num_tokens": 38805387.0,
      "step": 16915
    },
    {
      "entropy": 5.3208996772766115,
      "epoch": 1.6253602305475505,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047403928196855776,
      "loss": 4.846,
      "mean_token_accuracy": 0.2340749830007553,
      "num_tokens": 38815307.0,
      "step": 16920
    },
    {
      "entropy": 5.270659875869751,
      "epoch": 1.6258405379442844,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047402329274060916,
      "loss": 4.8192,
      "mean_token_accuracy": 0.23876880407333373,
      "num_tokens": 38826585.0,
      "step": 16925
    },
    {
      "entropy": 5.254394817352295,
      "epoch": 1.6263208453410183,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004740072988919306,
      "loss": 4.7654,
      "mean_token_accuracy": 0.24280460625886918,
      "num_tokens": 38838406.0,
      "step": 16930
    },
    {
      "entropy": 5.295915222167968,
      "epoch": 1.6268011527377522,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004739913004228936,
      "loss": 4.7855,
      "mean_token_accuracy": 0.23738071769475938,
      "num_tokens": 38849952.0,
      "step": 16935
    },
    {
      "entropy": 5.300760412216187,
      "epoch": 1.627281460134486,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004739752973338694,
      "loss": 4.883,
      "mean_token_accuracy": 0.22993996292352675,
      "num_tokens": 38862761.0,
      "step": 16940
    },
    {
      "entropy": 5.283096075057983,
      "epoch": 1.62776176753122,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047395928962522965,
      "loss": 4.7976,
      "mean_token_accuracy": 0.23106745183467864,
      "num_tokens": 38874422.0,
      "step": 16945
    },
    {
      "entropy": 5.321087741851807,
      "epoch": 1.6282420749279538,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047394327729734595,
      "loss": 4.8554,
      "mean_token_accuracy": 0.2296795815229416,
      "num_tokens": 38885751.0,
      "step": 16950
    },
    {
      "entropy": 5.277218341827393,
      "epoch": 1.6287223823246877,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004739272603505901,
      "loss": 4.8335,
      "mean_token_accuracy": 0.24297627061605453,
      "num_tokens": 38897539.0,
      "step": 16955
    },
    {
      "entropy": 5.327167606353759,
      "epoch": 1.6292026897214216,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000473911238785334,
      "loss": 4.8285,
      "mean_token_accuracy": 0.2292473569512367,
      "num_tokens": 38908439.0,
      "step": 16960
    },
    {
      "entropy": 5.268059206008911,
      "epoch": 1.6296829971181557,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004738952126019496,
      "loss": 4.8131,
      "mean_token_accuracy": 0.22853629887104035,
      "num_tokens": 38920881.0,
      "step": 16965
    },
    {
      "entropy": 5.314772844314575,
      "epoch": 1.6301633045148896,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004738791818008089,
      "loss": 4.8348,
      "mean_token_accuracy": 0.2392116069793701,
      "num_tokens": 38932097.0,
      "step": 16970
    },
    {
      "entropy": 5.348698377609253,
      "epoch": 1.6306436119116234,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004738631463822841,
      "loss": 4.9983,
      "mean_token_accuracy": 0.22705250084400178,
      "num_tokens": 38945122.0,
      "step": 16975
    },
    {
      "entropy": 5.261803245544433,
      "epoch": 1.6311239193083573,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047384710634674766,
      "loss": 4.8574,
      "mean_token_accuracy": 0.24270246177911758,
      "num_tokens": 38956136.0,
      "step": 16980
    },
    {
      "entropy": 5.352550029754639,
      "epoch": 1.6316042267050914,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047383106169457184,
      "loss": 4.8129,
      "mean_token_accuracy": 0.2338670462369919,
      "num_tokens": 38968650.0,
      "step": 16985
    },
    {
      "entropy": 5.230259037017822,
      "epoch": 1.6320845341018253,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004738150124261292,
      "loss": 4.7626,
      "mean_token_accuracy": 0.24106219559907913,
      "num_tokens": 38979834.0,
      "step": 16990
    },
    {
      "entropy": 5.218370771408081,
      "epoch": 1.6325648414985592,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047379895854179226,
      "loss": 4.7448,
      "mean_token_accuracy": 0.2433784171938896,
      "num_tokens": 38991548.0,
      "step": 16995
    },
    {
      "entropy": 5.28368649482727,
      "epoch": 1.633045148895293,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004737829000419338,
      "loss": 4.8158,
      "mean_token_accuracy": 0.24034103751182556,
      "num_tokens": 39003344.0,
      "step": 17000
    },
    {
      "entropy": 5.280467176437378,
      "epoch": 1.633525456292027,
      "grad_norm": 1.0,
      "learning_rate": 0.00047376683692692666,
      "loss": 4.8545,
      "mean_token_accuracy": 0.23758715391159058,
      "num_tokens": 39015278.0,
      "step": 17005
    },
    {
      "entropy": 5.270189714431763,
      "epoch": 1.6340057636887608,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004737507691971439,
      "loss": 4.8057,
      "mean_token_accuracy": 0.23310541808605195,
      "num_tokens": 39027634.0,
      "step": 17010
    },
    {
      "entropy": 5.337384605407715,
      "epoch": 1.6344860710854947,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047373469685295833,
      "loss": 4.9187,
      "mean_token_accuracy": 0.2329920545220375,
      "num_tokens": 39038226.0,
      "step": 17015
    },
    {
      "entropy": 5.304053735733032,
      "epoch": 1.6349663784822286,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047371861989474326,
      "loss": 4.8786,
      "mean_token_accuracy": 0.2289386510848999,
      "num_tokens": 39050627.0,
      "step": 17020
    },
    {
      "entropy": 5.336245918273926,
      "epoch": 1.6354466858789625,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004737025383228719,
      "loss": 4.8617,
      "mean_token_accuracy": 0.2300426885485649,
      "num_tokens": 39061734.0,
      "step": 17025
    },
    {
      "entropy": 5.274497556686401,
      "epoch": 1.6359269932756964,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047368645213771764,
      "loss": 4.8826,
      "mean_token_accuracy": 0.22388777732849122,
      "num_tokens": 39074288.0,
      "step": 17030
    },
    {
      "entropy": 5.303309202194214,
      "epoch": 1.6364073006724302,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000473670361339654,
      "loss": 4.8932,
      "mean_token_accuracy": 0.2246177986264229,
      "num_tokens": 39084967.0,
      "step": 17035
    },
    {
      "entropy": 5.298746824264526,
      "epoch": 1.6368876080691641,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004736542659290544,
      "loss": 4.7699,
      "mean_token_accuracy": 0.24357028901576996,
      "num_tokens": 39095093.0,
      "step": 17040
    },
    {
      "entropy": 5.336689043045044,
      "epoch": 1.6373679154658982,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004736381659062927,
      "loss": 4.8673,
      "mean_token_accuracy": 0.24015413373708724,
      "num_tokens": 39106952.0,
      "step": 17045
    },
    {
      "entropy": 5.253575468063355,
      "epoch": 1.637848222862632,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047362206127174255,
      "loss": 4.747,
      "mean_token_accuracy": 0.23223066031932832,
      "num_tokens": 39117682.0,
      "step": 17050
    },
    {
      "entropy": 5.328846645355225,
      "epoch": 1.638328530259366,
      "grad_norm": 1.0,
      "learning_rate": 0.00047360595202577786,
      "loss": 4.9111,
      "mean_token_accuracy": 0.22425315529108047,
      "num_tokens": 39129767.0,
      "step": 17055
    },
    {
      "entropy": 5.227824449539185,
      "epoch": 1.6388088376560999,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047358983816877284,
      "loss": 4.825,
      "mean_token_accuracy": 0.23652805685997008,
      "num_tokens": 39140906.0,
      "step": 17060
    },
    {
      "entropy": 5.313734149932861,
      "epoch": 1.639289145052834,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004735737197011015,
      "loss": 4.7661,
      "mean_token_accuracy": 0.2278988242149353,
      "num_tokens": 39151884.0,
      "step": 17065
    },
    {
      "entropy": 5.2261711120605465,
      "epoch": 1.6397694524495678,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047355759662313793,
      "loss": 4.7764,
      "mean_token_accuracy": 0.2382149338722229,
      "num_tokens": 39162986.0,
      "step": 17070
    },
    {
      "entropy": 5.273455619812012,
      "epoch": 1.6402497598463017,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004735414689352566,
      "loss": 4.8637,
      "mean_token_accuracy": 0.23604640513658523,
      "num_tokens": 39174280.0,
      "step": 17075
    },
    {
      "entropy": 5.296095848083496,
      "epoch": 1.6407300672430356,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004735253366378318,
      "loss": 4.8475,
      "mean_token_accuracy": 0.2344707190990448,
      "num_tokens": 39185499.0,
      "step": 17080
    },
    {
      "entropy": 5.231544399261475,
      "epoch": 1.6412103746397695,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004735091997312383,
      "loss": 4.7789,
      "mean_token_accuracy": 0.23479196727275847,
      "num_tokens": 39196322.0,
      "step": 17085
    },
    {
      "entropy": 5.257634162902832,
      "epoch": 1.6416906820365034,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047349305821585067,
      "loss": 4.8352,
      "mean_token_accuracy": 0.2348909005522728,
      "num_tokens": 39207332.0,
      "step": 17090
    },
    {
      "entropy": 5.325033712387085,
      "epoch": 1.6421709894332372,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004734769120920435,
      "loss": 4.8831,
      "mean_token_accuracy": 0.23408539295196534,
      "num_tokens": 39217305.0,
      "step": 17095
    },
    {
      "entropy": 5.280045461654663,
      "epoch": 1.6426512968299711,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004734607613601919,
      "loss": 4.8015,
      "mean_token_accuracy": 0.2351818710565567,
      "num_tokens": 39228895.0,
      "step": 17100
    },
    {
      "entropy": 5.245759201049805,
      "epoch": 1.643131604226705,
      "grad_norm": 1.0,
      "learning_rate": 0.00047344460602067077,
      "loss": 4.8325,
      "mean_token_accuracy": 0.2330389842391014,
      "num_tokens": 39240859.0,
      "step": 17105
    },
    {
      "entropy": 5.226689577102661,
      "epoch": 1.643611911623439,
      "grad_norm": 1.078125,
      "learning_rate": 0.000473428446073855,
      "loss": 4.7372,
      "mean_token_accuracy": 0.24550260305404664,
      "num_tokens": 39251556.0,
      "step": 17110
    },
    {
      "entropy": 5.233819103240966,
      "epoch": 1.6440922190201728,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047341228152012003,
      "loss": 4.6972,
      "mean_token_accuracy": 0.24142104089260102,
      "num_tokens": 39263919.0,
      "step": 17115
    },
    {
      "entropy": 5.303416395187378,
      "epoch": 1.6445725264169067,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000473396112359841,
      "loss": 4.9713,
      "mean_token_accuracy": 0.2290106788277626,
      "num_tokens": 39276873.0,
      "step": 17120
    },
    {
      "entropy": 5.24555459022522,
      "epoch": 1.6450528338136408,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047337993859339334,
      "loss": 4.7539,
      "mean_token_accuracy": 0.24779917001724244,
      "num_tokens": 39288912.0,
      "step": 17125
    },
    {
      "entropy": 5.224681711196899,
      "epoch": 1.6455331412103746,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047336376022115255,
      "loss": 4.7918,
      "mean_token_accuracy": 0.23710740655660628,
      "num_tokens": 39301115.0,
      "step": 17130
    },
    {
      "entropy": 5.289428567886352,
      "epoch": 1.6460134486071085,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047334757724349437,
      "loss": 4.8808,
      "mean_token_accuracy": 0.22677421420812607,
      "num_tokens": 39313535.0,
      "step": 17135
    },
    {
      "entropy": 5.338725805282593,
      "epoch": 1.6464937560038426,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004733313896607943,
      "loss": 4.8284,
      "mean_token_accuracy": 0.23595817536115646,
      "num_tokens": 39325289.0,
      "step": 17140
    },
    {
      "entropy": 5.276720952987671,
      "epoch": 1.6469740634005765,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004733151974734284,
      "loss": 4.8006,
      "mean_token_accuracy": 0.23439791649580002,
      "num_tokens": 39337382.0,
      "step": 17145
    },
    {
      "entropy": 5.25616421699524,
      "epoch": 1.6474543707973104,
      "grad_norm": 1.0,
      "learning_rate": 0.00047329900068177245,
      "loss": 4.7841,
      "mean_token_accuracy": 0.23078353255987166,
      "num_tokens": 39348548.0,
      "step": 17150
    },
    {
      "entropy": 5.2883923053741455,
      "epoch": 1.6479346781940443,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047328279928620244,
      "loss": 4.8092,
      "mean_token_accuracy": 0.23197428435087203,
      "num_tokens": 39359788.0,
      "step": 17155
    },
    {
      "entropy": 5.2945960521697994,
      "epoch": 1.6484149855907781,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004732665932870947,
      "loss": 4.8223,
      "mean_token_accuracy": 0.2353326439857483,
      "num_tokens": 39370533.0,
      "step": 17160
    },
    {
      "entropy": 5.273328828811645,
      "epoch": 1.648895292987512,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047325038268482544,
      "loss": 4.8574,
      "mean_token_accuracy": 0.23233963996171952,
      "num_tokens": 39381157.0,
      "step": 17165
    },
    {
      "entropy": 5.318806314468384,
      "epoch": 1.649375600384246,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004732341674797709,
      "loss": 4.8182,
      "mean_token_accuracy": 0.23263005763292313,
      "num_tokens": 39391280.0,
      "step": 17170
    },
    {
      "entropy": 5.210747909545899,
      "epoch": 1.6498559077809798,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047321794767230766,
      "loss": 4.724,
      "mean_token_accuracy": 0.235298191010952,
      "num_tokens": 39402097.0,
      "step": 17175
    },
    {
      "entropy": 5.29261040687561,
      "epoch": 1.6503362151777137,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047320172326281224,
      "loss": 4.8943,
      "mean_token_accuracy": 0.23430612832307815,
      "num_tokens": 39414590.0,
      "step": 17180
    },
    {
      "entropy": 5.287702417373657,
      "epoch": 1.6508165225744476,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047318549425166134,
      "loss": 4.7922,
      "mean_token_accuracy": 0.23366756439208985,
      "num_tokens": 39425933.0,
      "step": 17185
    },
    {
      "entropy": 5.248785209655762,
      "epoch": 1.6512968299711814,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004731692606392318,
      "loss": 4.8192,
      "mean_token_accuracy": 0.23816737979650499,
      "num_tokens": 39437059.0,
      "step": 17190
    },
    {
      "entropy": 5.285198593139649,
      "epoch": 1.6517771373679153,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004731530224259004,
      "loss": 4.8612,
      "mean_token_accuracy": 0.23232388645410537,
      "num_tokens": 39449582.0,
      "step": 17195
    },
    {
      "entropy": 5.4044722557067875,
      "epoch": 1.6522574447646494,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004731367796120442,
      "loss": 4.869,
      "mean_token_accuracy": 0.23114705532789231,
      "num_tokens": 39461147.0,
      "step": 17200
    },
    {
      "entropy": 5.32912802696228,
      "epoch": 1.6527377521613833,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004731205321980404,
      "loss": 4.8613,
      "mean_token_accuracy": 0.2251904308795929,
      "num_tokens": 39471881.0,
      "step": 17205
    },
    {
      "entropy": 5.224696016311645,
      "epoch": 1.6532180595581172,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047310428018426616,
      "loss": 4.863,
      "mean_token_accuracy": 0.23291049152612686,
      "num_tokens": 39483632.0,
      "step": 17210
    },
    {
      "entropy": 5.290295839309692,
      "epoch": 1.653698366954851,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004730880235710987,
      "loss": 4.8508,
      "mean_token_accuracy": 0.22858270406723022,
      "num_tokens": 39495587.0,
      "step": 17215
    },
    {
      "entropy": 5.297774791717529,
      "epoch": 1.6541786743515852,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004730717623589155,
      "loss": 4.8564,
      "mean_token_accuracy": 0.23003783077001572,
      "num_tokens": 39505503.0,
      "step": 17220
    },
    {
      "entropy": 5.264372158050537,
      "epoch": 1.654658981748319,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004730554965480942,
      "loss": 4.8294,
      "mean_token_accuracy": 0.23507195115089416,
      "num_tokens": 39516681.0,
      "step": 17225
    },
    {
      "entropy": 5.247220277786255,
      "epoch": 1.655139289145053,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004730392261390124,
      "loss": 4.7711,
      "mean_token_accuracy": 0.2359684005379677,
      "num_tokens": 39528451.0,
      "step": 17230
    },
    {
      "entropy": 5.282897138595581,
      "epoch": 1.6556195965417868,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004730229511320478,
      "loss": 4.8791,
      "mean_token_accuracy": 0.2269144907593727,
      "num_tokens": 39540306.0,
      "step": 17235
    },
    {
      "entropy": 5.2785238265991214,
      "epoch": 1.6560999039385207,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047300667152757827,
      "loss": 4.7318,
      "mean_token_accuracy": 0.2378073200583458,
      "num_tokens": 39551260.0,
      "step": 17240
    },
    {
      "entropy": 5.318511152267456,
      "epoch": 1.6565802113352546,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047299038732598184,
      "loss": 4.8758,
      "mean_token_accuracy": 0.2283339112997055,
      "num_tokens": 39562991.0,
      "step": 17245
    },
    {
      "entropy": 5.234248876571655,
      "epoch": 1.6570605187319885,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047297409852763644,
      "loss": 4.7595,
      "mean_token_accuracy": 0.2427535444498062,
      "num_tokens": 39574629.0,
      "step": 17250
    },
    {
      "entropy": 5.289076566696167,
      "epoch": 1.6575408261287223,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004729578051329204,
      "loss": 4.8872,
      "mean_token_accuracy": 0.22994613498449326,
      "num_tokens": 39585599.0,
      "step": 17255
    },
    {
      "entropy": 5.30312876701355,
      "epoch": 1.6580211335254562,
      "grad_norm": 1.0,
      "learning_rate": 0.00047294150714221185,
      "loss": 4.8264,
      "mean_token_accuracy": 0.23337082862854003,
      "num_tokens": 39597225.0,
      "step": 17260
    },
    {
      "entropy": 5.288875675201416,
      "epoch": 1.65850144092219,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004729252045558894,
      "loss": 4.8609,
      "mean_token_accuracy": 0.23354512602090835,
      "num_tokens": 39609296.0,
      "step": 17265
    },
    {
      "entropy": 5.226981496810913,
      "epoch": 1.658981748318924,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047290889737433133,
      "loss": 4.7006,
      "mean_token_accuracy": 0.24376299530267714,
      "num_tokens": 39619369.0,
      "step": 17270
    },
    {
      "entropy": 5.264016485214233,
      "epoch": 1.6594620557156579,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047289258559791633,
      "loss": 4.7704,
      "mean_token_accuracy": 0.23674922585487365,
      "num_tokens": 39630438.0,
      "step": 17275
    },
    {
      "entropy": 5.283062219619751,
      "epoch": 1.659942363112392,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047287626922702317,
      "loss": 4.843,
      "mean_token_accuracy": 0.2338909015059471,
      "num_tokens": 39643339.0,
      "step": 17280
    },
    {
      "entropy": 5.306666040420533,
      "epoch": 1.6604226705091258,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047285994826203054,
      "loss": 4.8114,
      "mean_token_accuracy": 0.23733173608779906,
      "num_tokens": 39654495.0,
      "step": 17285
    },
    {
      "entropy": 5.247898530960083,
      "epoch": 1.6609029779058597,
      "grad_norm": 1.0,
      "learning_rate": 0.0004728436227033175,
      "loss": 4.7919,
      "mean_token_accuracy": 0.23866392374038697,
      "num_tokens": 39665870.0,
      "step": 17290
    },
    {
      "entropy": 5.20890326499939,
      "epoch": 1.6613832853025938,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047282729255126294,
      "loss": 4.7528,
      "mean_token_accuracy": 0.23772606551647185,
      "num_tokens": 39676761.0,
      "step": 17295
    },
    {
      "entropy": 5.299675750732422,
      "epoch": 1.6618635926993277,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004728109578062461,
      "loss": 4.8103,
      "mean_token_accuracy": 0.23806842267513276,
      "num_tokens": 39687515.0,
      "step": 17300
    },
    {
      "entropy": 5.24761471748352,
      "epoch": 1.6623439000960616,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047279461846864626,
      "loss": 4.7881,
      "mean_token_accuracy": 0.23927445113658904,
      "num_tokens": 39699599.0,
      "step": 17305
    },
    {
      "entropy": 5.2638795375823975,
      "epoch": 1.6628242074927955,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047277827453884265,
      "loss": 4.7883,
      "mean_token_accuracy": 0.248645056784153,
      "num_tokens": 39710122.0,
      "step": 17310
    },
    {
      "entropy": 5.323542165756225,
      "epoch": 1.6633045148895294,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047276192601721477,
      "loss": 4.8864,
      "mean_token_accuracy": 0.2293317511677742,
      "num_tokens": 39722884.0,
      "step": 17315
    },
    {
      "entropy": 5.231382846832275,
      "epoch": 1.6637848222862632,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004727455729041422,
      "loss": 4.7929,
      "mean_token_accuracy": 0.23803787678480148,
      "num_tokens": 39734464.0,
      "step": 17320
    },
    {
      "entropy": 5.27784948348999,
      "epoch": 1.6642651296829971,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047272921520000465,
      "loss": 4.8319,
      "mean_token_accuracy": 0.23267358541488647,
      "num_tokens": 39745948.0,
      "step": 17325
    },
    {
      "entropy": 5.29174427986145,
      "epoch": 1.664745437079731,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004727128529051819,
      "loss": 4.813,
      "mean_token_accuracy": 0.23168134689331055,
      "num_tokens": 39757024.0,
      "step": 17330
    },
    {
      "entropy": 5.228216361999512,
      "epoch": 1.6652257444764649,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004726964860200537,
      "loss": 4.7701,
      "mean_token_accuracy": 0.23889886736869811,
      "num_tokens": 39767205.0,
      "step": 17335
    },
    {
      "entropy": 5.259266471862793,
      "epoch": 1.6657060518731988,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004726801145450002,
      "loss": 4.8315,
      "mean_token_accuracy": 0.2339811861515045,
      "num_tokens": 39779543.0,
      "step": 17340
    },
    {
      "entropy": 5.22501425743103,
      "epoch": 1.6661863592699326,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004726637384804014,
      "loss": 4.7258,
      "mean_token_accuracy": 0.24173194617033006,
      "num_tokens": 39790827.0,
      "step": 17345
    },
    {
      "entropy": 5.19028673171997,
      "epoch": 1.6666666666666665,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004726473578266375,
      "loss": 4.6903,
      "mean_token_accuracy": 0.23690952658653258,
      "num_tokens": 39801995.0,
      "step": 17350
    },
    {
      "entropy": 5.252016353607178,
      "epoch": 1.6671469740634006,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047263097258408893,
      "loss": 4.8112,
      "mean_token_accuracy": 0.23439155220985414,
      "num_tokens": 39813319.0,
      "step": 17355
    },
    {
      "entropy": 5.286487293243408,
      "epoch": 1.6676272814601345,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004726145827531359,
      "loss": 4.7715,
      "mean_token_accuracy": 0.24489095211029052,
      "num_tokens": 39824178.0,
      "step": 17360
    },
    {
      "entropy": 5.221025466918945,
      "epoch": 1.6681075888568684,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047259818833415916,
      "loss": 4.7774,
      "mean_token_accuracy": 0.23541236966848372,
      "num_tokens": 39835452.0,
      "step": 17365
    },
    {
      "entropy": 5.189358758926391,
      "epoch": 1.6685878962536023,
      "grad_norm": 1.0,
      "learning_rate": 0.00047258178932753917,
      "loss": 4.7969,
      "mean_token_accuracy": 0.23077984303236007,
      "num_tokens": 39848083.0,
      "step": 17370
    },
    {
      "entropy": 5.299428796768188,
      "epoch": 1.6690682036503364,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047256538573365675,
      "loss": 4.8045,
      "mean_token_accuracy": 0.23056688904762268,
      "num_tokens": 39859132.0,
      "step": 17375
    },
    {
      "entropy": 5.231441831588745,
      "epoch": 1.6695485110470702,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004725489775528928,
      "loss": 4.7979,
      "mean_token_accuracy": 0.2341003328561783,
      "num_tokens": 39870237.0,
      "step": 17380
    },
    {
      "entropy": 5.199922657012939,
      "epoch": 1.6700288184438041,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047253256478562805,
      "loss": 4.7789,
      "mean_token_accuracy": 0.23821650296449662,
      "num_tokens": 39883009.0,
      "step": 17385
    },
    {
      "entropy": 5.253457498550415,
      "epoch": 1.670509125840538,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047251614743224374,
      "loss": 4.8319,
      "mean_token_accuracy": 0.23514513224363326,
      "num_tokens": 39894444.0,
      "step": 17390
    },
    {
      "entropy": 5.31554536819458,
      "epoch": 1.670989433237272,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047249972549312107,
      "loss": 4.8776,
      "mean_token_accuracy": 0.2256608635187149,
      "num_tokens": 39905312.0,
      "step": 17395
    },
    {
      "entropy": 5.31590461730957,
      "epoch": 1.6714697406340058,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004724832989686411,
      "loss": 4.8574,
      "mean_token_accuracy": 0.2366069436073303,
      "num_tokens": 39916724.0,
      "step": 17400
    },
    {
      "entropy": 5.282752799987793,
      "epoch": 1.6719500480307397,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047246686785918545,
      "loss": 4.834,
      "mean_token_accuracy": 0.24160135686397552,
      "num_tokens": 39928919.0,
      "step": 17405
    },
    {
      "entropy": 5.2985063076019285,
      "epoch": 1.6724303554274735,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047245043216513546,
      "loss": 4.9293,
      "mean_token_accuracy": 0.22465210258960724,
      "num_tokens": 39940301.0,
      "step": 17410
    },
    {
      "entropy": 5.259132671356201,
      "epoch": 1.6729106628242074,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004724339918868727,
      "loss": 4.7614,
      "mean_token_accuracy": 0.23712130039930343,
      "num_tokens": 39951269.0,
      "step": 17415
    },
    {
      "entropy": 5.353413486480713,
      "epoch": 1.6733909702209413,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000472417547024779,
      "loss": 4.8638,
      "mean_token_accuracy": 0.23807845562696456,
      "num_tokens": 39963057.0,
      "step": 17420
    },
    {
      "entropy": 5.367051076889038,
      "epoch": 1.6738712776176752,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047240109757923593,
      "loss": 4.9293,
      "mean_token_accuracy": 0.23074073642492293,
      "num_tokens": 39973876.0,
      "step": 17425
    },
    {
      "entropy": 5.3430609703063965,
      "epoch": 1.674351585014409,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004723846435506256,
      "loss": 4.8685,
      "mean_token_accuracy": 0.23180557936429977,
      "num_tokens": 39984477.0,
      "step": 17430
    },
    {
      "entropy": 5.2966526508331295,
      "epoch": 1.6748318924111432,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047236818493932994,
      "loss": 4.8089,
      "mean_token_accuracy": 0.2382492184638977,
      "num_tokens": 39996266.0,
      "step": 17435
    },
    {
      "entropy": 5.2843156337738035,
      "epoch": 1.675312199807877,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004723517217457311,
      "loss": 4.8835,
      "mean_token_accuracy": 0.2324286624789238,
      "num_tokens": 40008618.0,
      "step": 17440
    },
    {
      "entropy": 5.28943362236023,
      "epoch": 1.675792507204611,
      "grad_norm": 1.0,
      "learning_rate": 0.0004723352539702113,
      "loss": 4.7711,
      "mean_token_accuracy": 0.23619728833436965,
      "num_tokens": 40020290.0,
      "step": 17445
    },
    {
      "entropy": 5.256688451766967,
      "epoch": 1.6762728146013448,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004723187816131529,
      "loss": 4.7219,
      "mean_token_accuracy": 0.23738554269075393,
      "num_tokens": 40031167.0,
      "step": 17450
    },
    {
      "entropy": 5.211318874359131,
      "epoch": 1.676753121998079,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004723023046749383,
      "loss": 4.8016,
      "mean_token_accuracy": 0.24337275624275206,
      "num_tokens": 40043278.0,
      "step": 17455
    },
    {
      "entropy": 5.261746120452881,
      "epoch": 1.6772334293948128,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047228582315595,
      "loss": 4.778,
      "mean_token_accuracy": 0.23787106424570084,
      "num_tokens": 40055022.0,
      "step": 17460
    },
    {
      "entropy": 5.269498634338379,
      "epoch": 1.6777137367915467,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004722693370565708,
      "loss": 4.873,
      "mean_token_accuracy": 0.23438012599945068,
      "num_tokens": 40066556.0,
      "step": 17465
    },
    {
      "entropy": 5.401852130889893,
      "epoch": 1.6781940441882806,
      "grad_norm": 1.0,
      "learning_rate": 0.00047225284637718323,
      "loss": 4.9049,
      "mean_token_accuracy": 0.22581236064434052,
      "num_tokens": 40078015.0,
      "step": 17470
    },
    {
      "entropy": 5.245467329025269,
      "epoch": 1.6786743515850144,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004722363511181703,
      "loss": 4.8059,
      "mean_token_accuracy": 0.2353228971362114,
      "num_tokens": 40090061.0,
      "step": 17475
    },
    {
      "entropy": 5.307917356491089,
      "epoch": 1.6791546589817483,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000472219851279915,
      "loss": 4.8737,
      "mean_token_accuracy": 0.2314113348722458,
      "num_tokens": 40102569.0,
      "step": 17480
    },
    {
      "entropy": 5.296753787994385,
      "epoch": 1.6796349663784822,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004722033468628004,
      "loss": 4.8149,
      "mean_token_accuracy": 0.23382521271705628,
      "num_tokens": 40114883.0,
      "step": 17485
    },
    {
      "entropy": 5.2546069622039795,
      "epoch": 1.680115273775216,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004721868378672098,
      "loss": 4.7415,
      "mean_token_accuracy": 0.24794315844774245,
      "num_tokens": 40126337.0,
      "step": 17490
    },
    {
      "entropy": 5.218559074401855,
      "epoch": 1.68059558117195,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004721703242935261,
      "loss": 4.788,
      "mean_token_accuracy": 0.2437625989317894,
      "num_tokens": 40137438.0,
      "step": 17495
    },
    {
      "entropy": 5.270514440536499,
      "epoch": 1.6810758885686838,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004721538061421331,
      "loss": 4.8041,
      "mean_token_accuracy": 0.23207644671201705,
      "num_tokens": 40148061.0,
      "step": 17500
    },
    {
      "entropy": 5.339199447631836,
      "epoch": 1.6815561959654177,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047213728341341407,
      "loss": 4.845,
      "mean_token_accuracy": 0.23061527460813522,
      "num_tokens": 40159118.0,
      "step": 17505
    },
    {
      "entropy": 5.251945543289184,
      "epoch": 1.6820365033621518,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004721207561077527,
      "loss": 4.7716,
      "mean_token_accuracy": 0.23612982481718064,
      "num_tokens": 40170027.0,
      "step": 17510
    },
    {
      "entropy": 5.2080975532531735,
      "epoch": 1.6825168107588857,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004721042242255327,
      "loss": 4.8021,
      "mean_token_accuracy": 0.23276917189359664,
      "num_tokens": 40181819.0,
      "step": 17515
    },
    {
      "entropy": 5.312236547470093,
      "epoch": 1.6829971181556196,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047208768776713805,
      "loss": 4.7289,
      "mean_token_accuracy": 0.23787782788276673,
      "num_tokens": 40191870.0,
      "step": 17520
    },
    {
      "entropy": 5.300977611541748,
      "epoch": 1.6834774255523535,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004720711467329523,
      "loss": 4.8192,
      "mean_token_accuracy": 0.23521960377693177,
      "num_tokens": 40203191.0,
      "step": 17525
    },
    {
      "entropy": 5.37005205154419,
      "epoch": 1.6839577329490876,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004720546011233599,
      "loss": 4.9356,
      "mean_token_accuracy": 0.2223026216030121,
      "num_tokens": 40214217.0,
      "step": 17530
    },
    {
      "entropy": 5.277394866943359,
      "epoch": 1.6844380403458215,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004720380509387446,
      "loss": 4.8419,
      "mean_token_accuracy": 0.23431964069604874,
      "num_tokens": 40225711.0,
      "step": 17535
    },
    {
      "entropy": 5.383616590499878,
      "epoch": 1.6849183477425553,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000472021496179491,
      "loss": 4.9326,
      "mean_token_accuracy": 0.2249226003885269,
      "num_tokens": 40235748.0,
      "step": 17540
    },
    {
      "entropy": 5.3722234725952145,
      "epoch": 1.6853986551392892,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047200493684598316,
      "loss": 4.8984,
      "mean_token_accuracy": 0.22594718635082245,
      "num_tokens": 40246090.0,
      "step": 17545
    },
    {
      "entropy": 5.293075323104858,
      "epoch": 1.685878962536023,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047198837293860573,
      "loss": 4.8936,
      "mean_token_accuracy": 0.23168880641460418,
      "num_tokens": 40257709.0,
      "step": 17550
    },
    {
      "entropy": 5.281067752838135,
      "epoch": 1.686359269932757,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004719718044577432,
      "loss": 4.8277,
      "mean_token_accuracy": 0.23278246074914932,
      "num_tokens": 40268670.0,
      "step": 17555
    },
    {
      "entropy": 5.2510580062866214,
      "epoch": 1.6868395773294909,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047195523140378034,
      "loss": 4.7546,
      "mean_token_accuracy": 0.24131402522325515,
      "num_tokens": 40279289.0,
      "step": 17560
    },
    {
      "entropy": 5.343555450439453,
      "epoch": 1.6873198847262247,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047193865377710177,
      "loss": 4.9473,
      "mean_token_accuracy": 0.22606462836265565,
      "num_tokens": 40289953.0,
      "step": 17565
    },
    {
      "entropy": 5.340047883987427,
      "epoch": 1.6878001921229586,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047192207157809246,
      "loss": 4.892,
      "mean_token_accuracy": 0.22798294723033904,
      "num_tokens": 40302180.0,
      "step": 17570
    },
    {
      "entropy": 5.2548707008361815,
      "epoch": 1.6882804995196925,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047190548480713736,
      "loss": 4.758,
      "mean_token_accuracy": 0.23494012653827667,
      "num_tokens": 40312556.0,
      "step": 17575
    },
    {
      "entropy": 5.286889123916626,
      "epoch": 1.6887608069164264,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047188889346462163,
      "loss": 4.8537,
      "mean_token_accuracy": 0.23204318135976792,
      "num_tokens": 40323213.0,
      "step": 17580
    },
    {
      "entropy": 5.29881181716919,
      "epoch": 1.6892411143131603,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047187229755093037,
      "loss": 4.9005,
      "mean_token_accuracy": 0.22363416999578475,
      "num_tokens": 40334567.0,
      "step": 17585
    },
    {
      "entropy": 5.333898735046387,
      "epoch": 1.6897214217098944,
      "grad_norm": 1.046875,
      "learning_rate": 0.000471855697066449,
      "loss": 4.8611,
      "mean_token_accuracy": 0.22984439879655838,
      "num_tokens": 40345606.0,
      "step": 17590
    },
    {
      "entropy": 5.316702365875244,
      "epoch": 1.6902017291066282,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047183909201156297,
      "loss": 4.8424,
      "mean_token_accuracy": 0.23505552262067794,
      "num_tokens": 40356366.0,
      "step": 17595
    },
    {
      "entropy": 5.21933274269104,
      "epoch": 1.6906820365033621,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004718224823866576,
      "loss": 4.7626,
      "mean_token_accuracy": 0.24183897525072098,
      "num_tokens": 40367451.0,
      "step": 17600
    },
    {
      "entropy": 5.172420406341553,
      "epoch": 1.691162343900096,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004718058681921186,
      "loss": 4.7039,
      "mean_token_accuracy": 0.24331605285406113,
      "num_tokens": 40379377.0,
      "step": 17605
    },
    {
      "entropy": 5.268290042877197,
      "epoch": 1.6916426512968301,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047178924942833185,
      "loss": 4.8368,
      "mean_token_accuracy": 0.2342387169599533,
      "num_tokens": 40390615.0,
      "step": 17610
    },
    {
      "entropy": 5.338238620758057,
      "epoch": 1.692122958693564,
      "grad_norm": 1.0,
      "learning_rate": 0.0004717726260956831,
      "loss": 4.8794,
      "mean_token_accuracy": 0.22813993394374849,
      "num_tokens": 40402475.0,
      "step": 17615
    },
    {
      "entropy": 5.290013027191162,
      "epoch": 1.6926032660902979,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004717559981945581,
      "loss": 4.8602,
      "mean_token_accuracy": 0.2240104779601097,
      "num_tokens": 40414311.0,
      "step": 17620
    },
    {
      "entropy": 5.2338401794433596,
      "epoch": 1.6930835734870318,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004717393657253432,
      "loss": 4.7559,
      "mean_token_accuracy": 0.2371399775147438,
      "num_tokens": 40425859.0,
      "step": 17625
    },
    {
      "entropy": 5.370133543014527,
      "epoch": 1.6935638808837656,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004717227286884243,
      "loss": 4.8724,
      "mean_token_accuracy": 0.2263252019882202,
      "num_tokens": 40436679.0,
      "step": 17630
    },
    {
      "entropy": 5.357480716705322,
      "epoch": 1.6940441882804995,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004717060870841879,
      "loss": 4.9168,
      "mean_token_accuracy": 0.22431331127882004,
      "num_tokens": 40446985.0,
      "step": 17635
    },
    {
      "entropy": 5.2554051876068115,
      "epoch": 1.6945244956772334,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004716894409130202,
      "loss": 4.7912,
      "mean_token_accuracy": 0.2356422573328018,
      "num_tokens": 40458777.0,
      "step": 17640
    },
    {
      "entropy": 5.332077360153198,
      "epoch": 1.6950048030739673,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004716727901753078,
      "loss": 4.9105,
      "mean_token_accuracy": 0.22311581671237946,
      "num_tokens": 40469493.0,
      "step": 17645
    },
    {
      "entropy": 5.281788349151611,
      "epoch": 1.6954851104707012,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004716561348714371,
      "loss": 4.9048,
      "mean_token_accuracy": 0.23791031390428544,
      "num_tokens": 40480889.0,
      "step": 17650
    },
    {
      "entropy": 5.288267755508423,
      "epoch": 1.695965417867435,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047163947500179494,
      "loss": 4.7331,
      "mean_token_accuracy": 0.2359095260500908,
      "num_tokens": 40491630.0,
      "step": 17655
    },
    {
      "entropy": 5.418720245361328,
      "epoch": 1.696445725264169,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004716228105667681,
      "loss": 4.9548,
      "mean_token_accuracy": 0.21702479422092438,
      "num_tokens": 40503204.0,
      "step": 17660
    },
    {
      "entropy": 5.266919040679932,
      "epoch": 1.696926032660903,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004716061415667435,
      "loss": 4.8325,
      "mean_token_accuracy": 0.23248408287763594,
      "num_tokens": 40513917.0,
      "step": 17665
    },
    {
      "entropy": 5.331394958496094,
      "epoch": 1.697406340057637,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000471589468002108,
      "loss": 4.879,
      "mean_token_accuracy": 0.231434266269207,
      "num_tokens": 40526501.0,
      "step": 17670
    },
    {
      "entropy": 5.248792123794556,
      "epoch": 1.6978866474543708,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004715727898732488,
      "loss": 4.7561,
      "mean_token_accuracy": 0.23905529975891113,
      "num_tokens": 40536641.0,
      "step": 17675
    },
    {
      "entropy": 5.280260133743286,
      "epoch": 1.6983669548511047,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047155610718055315,
      "loss": 4.8333,
      "mean_token_accuracy": 0.23579123616218567,
      "num_tokens": 40547497.0,
      "step": 17680
    },
    {
      "entropy": 5.290646076202393,
      "epoch": 1.6988472622478388,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047153941992440833,
      "loss": 4.7803,
      "mean_token_accuracy": 0.23134051114320756,
      "num_tokens": 40559212.0,
      "step": 17685
    },
    {
      "entropy": 5.217078351974488,
      "epoch": 1.6993275696445727,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004715227281052018,
      "loss": 4.7454,
      "mean_token_accuracy": 0.2350364089012146,
      "num_tokens": 40571692.0,
      "step": 17690
    },
    {
      "entropy": 5.2451056957244875,
      "epoch": 1.6998078770413065,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004715060317233211,
      "loss": 4.8155,
      "mean_token_accuracy": 0.23100574910640717,
      "num_tokens": 40583751.0,
      "step": 17695
    },
    {
      "entropy": 5.351688003540039,
      "epoch": 1.7002881844380404,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004714893307791538,
      "loss": 4.8372,
      "mean_token_accuracy": 0.2337740659713745,
      "num_tokens": 40594782.0,
      "step": 17700
    },
    {
      "entropy": 5.307193040847778,
      "epoch": 1.7007684918347743,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047147262527308766,
      "loss": 4.8087,
      "mean_token_accuracy": 0.2371048167347908,
      "num_tokens": 40604583.0,
      "step": 17705
    },
    {
      "entropy": 5.268728017807007,
      "epoch": 1.7012487992315082,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004714559152055106,
      "loss": 4.8432,
      "mean_token_accuracy": 0.23025344759225846,
      "num_tokens": 40617957.0,
      "step": 17710
    },
    {
      "entropy": 5.297297239303589,
      "epoch": 1.701729106628242,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004714392005768106,
      "loss": 4.7929,
      "mean_token_accuracy": 0.23549143970012665,
      "num_tokens": 40628016.0,
      "step": 17715
    },
    {
      "entropy": 5.362607431411743,
      "epoch": 1.702209414024976,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004714224813873756,
      "loss": 4.9539,
      "mean_token_accuracy": 0.22495478689670562,
      "num_tokens": 40639339.0,
      "step": 17720
    },
    {
      "entropy": 5.339702558517456,
      "epoch": 1.7026897214217098,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047140575763759393,
      "loss": 4.9093,
      "mean_token_accuracy": 0.232338947057724,
      "num_tokens": 40651634.0,
      "step": 17725
    },
    {
      "entropy": 5.341545963287354,
      "epoch": 1.7031700288184437,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047138902932785363,
      "loss": 4.8797,
      "mean_token_accuracy": 0.22912612557411194,
      "num_tokens": 40662742.0,
      "step": 17730
    },
    {
      "entropy": 5.340724802017212,
      "epoch": 1.7036503362151776,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047137229645854333,
      "loss": 4.8815,
      "mean_token_accuracy": 0.23204363286495208,
      "num_tokens": 40673586.0,
      "step": 17735
    },
    {
      "entropy": 5.239948034286499,
      "epoch": 1.7041306436119115,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004713555590300513,
      "loss": 4.7269,
      "mean_token_accuracy": 0.24017900377511978,
      "num_tokens": 40684238.0,
      "step": 17740
    },
    {
      "entropy": 5.322170257568359,
      "epoch": 1.7046109510086456,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004713388170427664,
      "loss": 4.9427,
      "mean_token_accuracy": 0.22665225714445114,
      "num_tokens": 40696154.0,
      "step": 17745
    },
    {
      "entropy": 5.2989830493927,
      "epoch": 1.7050912584053795,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004713220704970771,
      "loss": 4.844,
      "mean_token_accuracy": 0.23638058304786683,
      "num_tokens": 40708328.0,
      "step": 17750
    },
    {
      "entropy": 5.355513191223144,
      "epoch": 1.7055715658021133,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047130531939337236,
      "loss": 4.9504,
      "mean_token_accuracy": 0.2250259816646576,
      "num_tokens": 40719067.0,
      "step": 17755
    },
    {
      "entropy": 5.170226097106934,
      "epoch": 1.7060518731988472,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047128856373204086,
      "loss": 4.6114,
      "mean_token_accuracy": 0.2503987610340118,
      "num_tokens": 40730638.0,
      "step": 17760
    },
    {
      "entropy": 5.297075653076172,
      "epoch": 1.7065321805955813,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047127180351347184,
      "loss": 4.8254,
      "mean_token_accuracy": 0.23135687857866288,
      "num_tokens": 40743245.0,
      "step": 17765
    },
    {
      "entropy": 5.348367357254029,
      "epoch": 1.7070124879923152,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004712550387380544,
      "loss": 4.826,
      "mean_token_accuracy": 0.234270641207695,
      "num_tokens": 40753769.0,
      "step": 17770
    },
    {
      "entropy": 5.243228483200073,
      "epoch": 1.707492795389049,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004712382694061776,
      "loss": 4.8528,
      "mean_token_accuracy": 0.23835416585206987,
      "num_tokens": 40765470.0,
      "step": 17775
    },
    {
      "entropy": 5.308897018432617,
      "epoch": 1.707973102785783,
      "grad_norm": 1.0,
      "learning_rate": 0.00047122149551823096,
      "loss": 4.8388,
      "mean_token_accuracy": 0.23048177510499954,
      "num_tokens": 40776427.0,
      "step": 17780
    },
    {
      "entropy": 5.273647212982178,
      "epoch": 1.7084534101825168,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004712047170746039,
      "loss": 4.7942,
      "mean_token_accuracy": 0.23018195778131484,
      "num_tokens": 40788707.0,
      "step": 17785
    },
    {
      "entropy": 5.395482921600342,
      "epoch": 1.7089337175792507,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047118793407568586,
      "loss": 4.9497,
      "mean_token_accuracy": 0.22381114363670349,
      "num_tokens": 40800092.0,
      "step": 17790
    },
    {
      "entropy": 5.239621496200561,
      "epoch": 1.7094140249759846,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047117114652186657,
      "loss": 4.7596,
      "mean_token_accuracy": 0.23928821682929993,
      "num_tokens": 40811389.0,
      "step": 17795
    },
    {
      "entropy": 5.282537603378296,
      "epoch": 1.7098943323727185,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047115435441353573,
      "loss": 4.8319,
      "mean_token_accuracy": 0.2312431052327156,
      "num_tokens": 40821438.0,
      "step": 17800
    },
    {
      "entropy": 5.232645273208618,
      "epoch": 1.7103746397694524,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047113755775108333,
      "loss": 4.7921,
      "mean_token_accuracy": 0.233128622174263,
      "num_tokens": 40833335.0,
      "step": 17805
    },
    {
      "entropy": 5.422710990905761,
      "epoch": 1.7108549471661862,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047112075653489913,
      "loss": 4.9396,
      "mean_token_accuracy": 0.22812143117189407,
      "num_tokens": 40844090.0,
      "step": 17810
    },
    {
      "entropy": 5.31012864112854,
      "epoch": 1.7113352545629201,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004711039507653734,
      "loss": 4.8195,
      "mean_token_accuracy": 0.23034635484218596,
      "num_tokens": 40855176.0,
      "step": 17815
    },
    {
      "entropy": 5.253414869308472,
      "epoch": 1.7118155619596542,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004710871404428961,
      "loss": 4.9051,
      "mean_token_accuracy": 0.22907905131578446,
      "num_tokens": 40866190.0,
      "step": 17820
    },
    {
      "entropy": 5.2994619846344,
      "epoch": 1.7122958693563881,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047107032556785786,
      "loss": 4.7988,
      "mean_token_accuracy": 0.22820111066102983,
      "num_tokens": 40876273.0,
      "step": 17825
    },
    {
      "entropy": 5.242632293701172,
      "epoch": 1.712776176753122,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047105350614064874,
      "loss": 4.7044,
      "mean_token_accuracy": 0.241782546043396,
      "num_tokens": 40887843.0,
      "step": 17830
    },
    {
      "entropy": 5.290749549865723,
      "epoch": 1.7132564841498559,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047103668216165944,
      "loss": 4.8198,
      "mean_token_accuracy": 0.2335355117917061,
      "num_tokens": 40898648.0,
      "step": 17835
    },
    {
      "entropy": 5.242843532562256,
      "epoch": 1.71373679154659,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047101985363128045,
      "loss": 4.8092,
      "mean_token_accuracy": 0.23674969226121903,
      "num_tokens": 40910817.0,
      "step": 17840
    },
    {
      "entropy": 5.308571243286133,
      "epoch": 1.7142170989433239,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047100302054990255,
      "loss": 4.87,
      "mean_token_accuracy": 0.23674565106630324,
      "num_tokens": 40922520.0,
      "step": 17845
    },
    {
      "entropy": 5.301729965209961,
      "epoch": 1.7146974063400577,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004709861829179165,
      "loss": 4.8525,
      "mean_token_accuracy": 0.23260308057069778,
      "num_tokens": 40934188.0,
      "step": 17850
    },
    {
      "entropy": 5.317915725708008,
      "epoch": 1.7151777137367916,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047096934073571325,
      "loss": 4.8677,
      "mean_token_accuracy": 0.23010713011026382,
      "num_tokens": 40944520.0,
      "step": 17855
    },
    {
      "entropy": 5.328940439224243,
      "epoch": 1.7156580211335255,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047095249400368384,
      "loss": 4.8523,
      "mean_token_accuracy": 0.22281205803155898,
      "num_tokens": 40955874.0,
      "step": 17860
    },
    {
      "entropy": 5.291602373123169,
      "epoch": 1.7161383285302594,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047093564272221927,
      "loss": 4.8397,
      "mean_token_accuracy": 0.23875494301319122,
      "num_tokens": 40966697.0,
      "step": 17865
    },
    {
      "entropy": 5.238680076599121,
      "epoch": 1.7166186359269933,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047091878689171105,
      "loss": 4.8305,
      "mean_token_accuracy": 0.23257206678390502,
      "num_tokens": 40978107.0,
      "step": 17870
    },
    {
      "entropy": 5.297430753707886,
      "epoch": 1.7170989433237271,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004709019265125502,
      "loss": 4.7995,
      "mean_token_accuracy": 0.2322564795613289,
      "num_tokens": 40991535.0,
      "step": 17875
    },
    {
      "entropy": 5.3958351612091064,
      "epoch": 1.717579250720461,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047088506158512837,
      "loss": 4.8409,
      "mean_token_accuracy": 0.23020493984222412,
      "num_tokens": 41001333.0,
      "step": 17880
    },
    {
      "entropy": 5.198616933822632,
      "epoch": 1.718059558117195,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047086819210983714,
      "loss": 4.7391,
      "mean_token_accuracy": 0.23855255395174027,
      "num_tokens": 41012854.0,
      "step": 17885
    },
    {
      "entropy": 5.221627521514892,
      "epoch": 1.7185398655139288,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047085131808706813,
      "loss": 4.7779,
      "mean_token_accuracy": 0.2391347125172615,
      "num_tokens": 41024041.0,
      "step": 17890
    },
    {
      "entropy": 5.313968801498413,
      "epoch": 1.7190201729106627,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004708344395172129,
      "loss": 4.8255,
      "mean_token_accuracy": 0.2361692875623703,
      "num_tokens": 41034788.0,
      "step": 17895
    },
    {
      "entropy": 5.342522192001343,
      "epoch": 1.7195004803073968,
      "grad_norm": 1.0,
      "learning_rate": 0.0004708175564006636,
      "loss": 4.8211,
      "mean_token_accuracy": 0.2417980432510376,
      "num_tokens": 41046513.0,
      "step": 17900
    },
    {
      "entropy": 5.366668033599853,
      "epoch": 1.7199807877041307,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004708006687378121,
      "loss": 4.9448,
      "mean_token_accuracy": 0.22209766507148743,
      "num_tokens": 41059054.0,
      "step": 17905
    },
    {
      "entropy": 5.294875288009644,
      "epoch": 1.7204610951008645,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004707837765290505,
      "loss": 4.8081,
      "mean_token_accuracy": 0.2352616786956787,
      "num_tokens": 41070171.0,
      "step": 17910
    },
    {
      "entropy": 5.261094093322754,
      "epoch": 1.7209414024975984,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004707668797747709,
      "loss": 4.7849,
      "mean_token_accuracy": 0.2335084304213524,
      "num_tokens": 41080963.0,
      "step": 17915
    },
    {
      "entropy": 5.277729082107544,
      "epoch": 1.7214217098943325,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004707499784753657,
      "loss": 4.8467,
      "mean_token_accuracy": 0.2320716604590416,
      "num_tokens": 41092658.0,
      "step": 17920
    },
    {
      "entropy": 5.315527677536011,
      "epoch": 1.7219020172910664,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004707330726312273,
      "loss": 4.8695,
      "mean_token_accuracy": 0.2275842934846878,
      "num_tokens": 41104521.0,
      "step": 17925
    },
    {
      "entropy": 5.299657154083252,
      "epoch": 1.7223823246878003,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047071616224274803,
      "loss": 4.8138,
      "mean_token_accuracy": 0.23888926804065705,
      "num_tokens": 41116420.0,
      "step": 17930
    },
    {
      "entropy": 5.27760910987854,
      "epoch": 1.7228626320845342,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004706992473103207,
      "loss": 4.8244,
      "mean_token_accuracy": 0.23134843856096268,
      "num_tokens": 41127747.0,
      "step": 17935
    },
    {
      "entropy": 5.29104790687561,
      "epoch": 1.723342939481268,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047068232783433806,
      "loss": 4.7918,
      "mean_token_accuracy": 0.2397273525595665,
      "num_tokens": 41138498.0,
      "step": 17940
    },
    {
      "entropy": 5.343942546844483,
      "epoch": 1.723823246878002,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004706654038151927,
      "loss": 4.8933,
      "mean_token_accuracy": 0.23278069794178008,
      "num_tokens": 41149781.0,
      "step": 17945
    },
    {
      "entropy": 5.309473943710327,
      "epoch": 1.7243035542747358,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004706484752532777,
      "loss": 4.8571,
      "mean_token_accuracy": 0.23345241844654083,
      "num_tokens": 41160421.0,
      "step": 17950
    },
    {
      "entropy": 5.246830606460572,
      "epoch": 1.7247838616714697,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004706315421489861,
      "loss": 4.795,
      "mean_token_accuracy": 0.2468930274248123,
      "num_tokens": 41171192.0,
      "step": 17955
    },
    {
      "entropy": 5.234622669219971,
      "epoch": 1.7252641690682036,
      "grad_norm": 1.0,
      "learning_rate": 0.0004706146045027109,
      "loss": 4.8026,
      "mean_token_accuracy": 0.24054929465055466,
      "num_tokens": 41182396.0,
      "step": 17960
    },
    {
      "entropy": 5.3794722080230715,
      "epoch": 1.7257444764649374,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004705976623148455,
      "loss": 4.9256,
      "mean_token_accuracy": 0.2197549059987068,
      "num_tokens": 41194939.0,
      "step": 17965
    },
    {
      "entropy": 5.281379842758179,
      "epoch": 1.7262247838616713,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047058071558578324,
      "loss": 4.8319,
      "mean_token_accuracy": 0.22983901649713517,
      "num_tokens": 41206670.0,
      "step": 17970
    },
    {
      "entropy": 5.348994970321655,
      "epoch": 1.7267050912584054,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004705637643159175,
      "loss": 4.9878,
      "mean_token_accuracy": 0.22610980421304702,
      "num_tokens": 41220200.0,
      "step": 17975
    },
    {
      "entropy": 5.321943235397339,
      "epoch": 1.7271853986551393,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047054680850564185,
      "loss": 4.785,
      "mean_token_accuracy": 0.238235904276371,
      "num_tokens": 41231697.0,
      "step": 17980
    },
    {
      "entropy": 5.270865154266358,
      "epoch": 1.7276657060518732,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004705298481553499,
      "loss": 4.8353,
      "mean_token_accuracy": 0.23461008965969085,
      "num_tokens": 41243835.0,
      "step": 17985
    },
    {
      "entropy": 5.36062068939209,
      "epoch": 1.728146013448607,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047051288326543553,
      "loss": 4.8468,
      "mean_token_accuracy": 0.23668460100889205,
      "num_tokens": 41255147.0,
      "step": 17990
    },
    {
      "entropy": 5.412624311447144,
      "epoch": 1.7286263208453412,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047049591383629247,
      "loss": 4.9027,
      "mean_token_accuracy": 0.2219374731183052,
      "num_tokens": 41267863.0,
      "step": 17995
    },
    {
      "entropy": 5.3648681640625,
      "epoch": 1.729106628242075,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047047893986831493,
      "loss": 4.8566,
      "mean_token_accuracy": 0.23286622911691665,
      "num_tokens": 41278506.0,
      "step": 18000
    },
    {
      "epoch": 1.729106628242075,
      "eval_entropy": 5.088965035949551,
      "eval_loss": 4.871064186096191,
      "eval_mean_token_accuracy": 0.24155001379940014,
      "eval_num_tokens": 41278506.0,
      "eval_runtime": 20.304,
      "eval_samples_per_second": 1616.183,
      "eval_steps_per_second": 202.029,
      "step": 18000
    },
    {
      "entropy": 5.315106630325317,
      "epoch": 1.729586935638809,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047046196136189686,
      "loss": 4.8502,
      "mean_token_accuracy": 0.23052018880844116,
      "num_tokens": 41289825.0,
      "step": 18005
    },
    {
      "entropy": 5.2751627445220945,
      "epoch": 1.7300672430355428,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004704449783174323,
      "loss": 4.8126,
      "mean_token_accuracy": 0.2360662952065468,
      "num_tokens": 41300451.0,
      "step": 18010
    },
    {
      "entropy": 5.265012598037719,
      "epoch": 1.7305475504322767,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004704279907353158,
      "loss": 4.765,
      "mean_token_accuracy": 0.24258455783128738,
      "num_tokens": 41312762.0,
      "step": 18015
    },
    {
      "entropy": 5.2626911163330075,
      "epoch": 1.7310278578290106,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047041099861594167,
      "loss": 4.8337,
      "mean_token_accuracy": 0.2369905889034271,
      "num_tokens": 41323464.0,
      "step": 18020
    },
    {
      "entropy": 5.300305986404419,
      "epoch": 1.7315081652257445,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004703940019597044,
      "loss": 4.8532,
      "mean_token_accuracy": 0.23861521780490874,
      "num_tokens": 41335361.0,
      "step": 18025
    },
    {
      "entropy": 5.325695276260376,
      "epoch": 1.7319884726224783,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047037700076699857,
      "loss": 4.8072,
      "mean_token_accuracy": 0.23519962579011916,
      "num_tokens": 41345353.0,
      "step": 18030
    },
    {
      "entropy": 5.284633207321167,
      "epoch": 1.7324687800192122,
      "grad_norm": 1.09375,
      "learning_rate": 0.000470359995038219,
      "loss": 4.8902,
      "mean_token_accuracy": 0.23670542985200882,
      "num_tokens": 41356578.0,
      "step": 18035
    },
    {
      "entropy": 5.324893999099731,
      "epoch": 1.732949087415946,
      "grad_norm": 1.0,
      "learning_rate": 0.0004703429847737604,
      "loss": 4.8494,
      "mean_token_accuracy": 0.23545071184635163,
      "num_tokens": 41367668.0,
      "step": 18040
    },
    {
      "entropy": 5.294162893295288,
      "epoch": 1.73342939481268,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004703259699740177,
      "loss": 4.7903,
      "mean_token_accuracy": 0.2369881883263588,
      "num_tokens": 41379746.0,
      "step": 18045
    },
    {
      "entropy": 5.354740333557129,
      "epoch": 1.7339097022094139,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047030895063938607,
      "loss": 4.904,
      "mean_token_accuracy": 0.2298465386033058,
      "num_tokens": 41390251.0,
      "step": 18050
    },
    {
      "entropy": 5.336814832687378,
      "epoch": 1.734390009606148,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047029192677026043,
      "loss": 4.8091,
      "mean_token_accuracy": 0.23657422065734862,
      "num_tokens": 41401689.0,
      "step": 18055
    },
    {
      "entropy": 5.302208280563354,
      "epoch": 1.7348703170028819,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004702748983670363,
      "loss": 4.8199,
      "mean_token_accuracy": 0.23283260613679885,
      "num_tokens": 41414001.0,
      "step": 18060
    },
    {
      "entropy": 5.394041061401367,
      "epoch": 1.7353506243996157,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004702578654301088,
      "loss": 4.929,
      "mean_token_accuracy": 0.23248368948698045,
      "num_tokens": 41426012.0,
      "step": 18065
    },
    {
      "entropy": 5.346261644363404,
      "epoch": 1.7358309317963496,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004702408279598734,
      "loss": 4.9066,
      "mean_token_accuracy": 0.22950783520936965,
      "num_tokens": 41436799.0,
      "step": 18070
    },
    {
      "entropy": 5.272062206268311,
      "epoch": 1.7363112391930837,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004702237859567258,
      "loss": 4.801,
      "mean_token_accuracy": 0.24468848556280137,
      "num_tokens": 41448802.0,
      "step": 18075
    },
    {
      "entropy": 5.218347454071045,
      "epoch": 1.7367915465898176,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004702067394210616,
      "loss": 4.731,
      "mean_token_accuracy": 0.24317118376493455,
      "num_tokens": 41459576.0,
      "step": 18080
    },
    {
      "entropy": 5.297867250442505,
      "epoch": 1.7372718539865515,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047018968835327643,
      "loss": 4.8269,
      "mean_token_accuracy": 0.2299545332789421,
      "num_tokens": 41471105.0,
      "step": 18085
    },
    {
      "entropy": 5.319504928588867,
      "epoch": 1.7377521613832854,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004701726327537664,
      "loss": 4.8107,
      "mean_token_accuracy": 0.2380877286195755,
      "num_tokens": 41482747.0,
      "step": 18090
    },
    {
      "entropy": 5.309560966491699,
      "epoch": 1.7382324687800192,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004701555726229274,
      "loss": 4.8492,
      "mean_token_accuracy": 0.22702578604221343,
      "num_tokens": 41494268.0,
      "step": 18095
    },
    {
      "entropy": 5.287116479873657,
      "epoch": 1.7387127761767531,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004701385079611555,
      "loss": 4.8196,
      "mean_token_accuracy": 0.23849975019693376,
      "num_tokens": 41506144.0,
      "step": 18100
    },
    {
      "entropy": 5.248522758483887,
      "epoch": 1.739193083573487,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047012143876884677,
      "loss": 4.7119,
      "mean_token_accuracy": 0.2532498463988304,
      "num_tokens": 41518165.0,
      "step": 18105
    },
    {
      "entropy": 5.277455615997314,
      "epoch": 1.739673390970221,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004701043650463977,
      "loss": 4.7846,
      "mean_token_accuracy": 0.24438977539539336,
      "num_tokens": 41529644.0,
      "step": 18110
    },
    {
      "entropy": 5.271407651901245,
      "epoch": 1.7401536983669548,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004700872867942046,
      "loss": 4.794,
      "mean_token_accuracy": 0.237464801967144,
      "num_tokens": 41540540.0,
      "step": 18115
    },
    {
      "entropy": 5.331207036972046,
      "epoch": 1.7406340057636887,
      "grad_norm": 1.09375,
      "learning_rate": 0.000470070204012664,
      "loss": 4.8004,
      "mean_token_accuracy": 0.2298593506217003,
      "num_tokens": 41551937.0,
      "step": 18120
    },
    {
      "entropy": 5.273026800155639,
      "epoch": 1.7411143131604225,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047005311670217256,
      "loss": 4.7627,
      "mean_token_accuracy": 0.24444033950567245,
      "num_tokens": 41561630.0,
      "step": 18125
    },
    {
      "entropy": 5.230842590332031,
      "epoch": 1.7415946205571564,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047003602486312687,
      "loss": 4.7445,
      "mean_token_accuracy": 0.24038286358118058,
      "num_tokens": 41573152.0,
      "step": 18130
    },
    {
      "entropy": 5.3150684356689455,
      "epoch": 1.7420749279538905,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004700189284959238,
      "loss": 4.8479,
      "mean_token_accuracy": 0.23051564246416092,
      "num_tokens": 41585280.0,
      "step": 18135
    },
    {
      "entropy": 5.292421150207519,
      "epoch": 1.7425552353506244,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047000182760096037,
      "loss": 4.7665,
      "mean_token_accuracy": 0.23996683955192566,
      "num_tokens": 41596308.0,
      "step": 18140
    },
    {
      "entropy": 5.232572793960571,
      "epoch": 1.7430355427473583,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004699847221786335,
      "loss": 4.7902,
      "mean_token_accuracy": 0.23406947404146194,
      "num_tokens": 41607479.0,
      "step": 18145
    },
    {
      "entropy": 5.204199171066284,
      "epoch": 1.7435158501440924,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004699676122293403,
      "loss": 4.7725,
      "mean_token_accuracy": 0.2320700541138649,
      "num_tokens": 41619832.0,
      "step": 18150
    },
    {
      "entropy": 5.347791242599487,
      "epoch": 1.7439961575408263,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004699504977534782,
      "loss": 4.9007,
      "mean_token_accuracy": 0.22537073940038682,
      "num_tokens": 41631715.0,
      "step": 18155
    },
    {
      "entropy": 5.371211051940918,
      "epoch": 1.7444764649375601,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004699333787514444,
      "loss": 4.8875,
      "mean_token_accuracy": 0.23146500438451767,
      "num_tokens": 41643454.0,
      "step": 18160
    },
    {
      "entropy": 5.286026334762573,
      "epoch": 1.744956772334294,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004699162552236363,
      "loss": 4.8205,
      "mean_token_accuracy": 0.23681031167507172,
      "num_tokens": 41655319.0,
      "step": 18165
    },
    {
      "entropy": 5.26386866569519,
      "epoch": 1.745437079731028,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046989912717045165,
      "loss": 4.8321,
      "mean_token_accuracy": 0.23480662405490876,
      "num_tokens": 41667082.0,
      "step": 18170
    },
    {
      "entropy": 5.33060393333435,
      "epoch": 1.7459173871277618,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046988199459228793,
      "loss": 4.7772,
      "mean_token_accuracy": 0.23703452199697495,
      "num_tokens": 41678281.0,
      "step": 18175
    },
    {
      "entropy": 5.26322751045227,
      "epoch": 1.7463976945244957,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004698648574895429,
      "loss": 4.7796,
      "mean_token_accuracy": 0.24271055907011033,
      "num_tokens": 41690596.0,
      "step": 18180
    },
    {
      "entropy": 5.325496673583984,
      "epoch": 1.7468780019212296,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046984771586261465,
      "loss": 4.8267,
      "mean_token_accuracy": 0.23277984857559203,
      "num_tokens": 41702441.0,
      "step": 18185
    },
    {
      "entropy": 5.331269979476929,
      "epoch": 1.7473583093179634,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000469830569711901,
      "loss": 4.8417,
      "mean_token_accuracy": 0.23301418274641036,
      "num_tokens": 41714235.0,
      "step": 18190
    },
    {
      "entropy": 5.291389608383179,
      "epoch": 1.7478386167146973,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004698134190377999,
      "loss": 4.8179,
      "mean_token_accuracy": 0.23679810762405396,
      "num_tokens": 41725870.0,
      "step": 18195
    },
    {
      "entropy": 5.268979024887085,
      "epoch": 1.7483189241114312,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046979626384070983,
      "loss": 4.7973,
      "mean_token_accuracy": 0.24038981646299362,
      "num_tokens": 41736500.0,
      "step": 18200
    },
    {
      "entropy": 5.249798154830932,
      "epoch": 1.748799231508165,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000469779104121029,
      "loss": 4.7721,
      "mean_token_accuracy": 0.2382185935974121,
      "num_tokens": 41748466.0,
      "step": 18205
    },
    {
      "entropy": 5.337527751922607,
      "epoch": 1.7492795389048992,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046976193987915553,
      "loss": 4.9325,
      "mean_token_accuracy": 0.23020538240671157,
      "num_tokens": 41759728.0,
      "step": 18210
    },
    {
      "entropy": 5.290850067138672,
      "epoch": 1.749759846301633,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004697447711154883,
      "loss": 4.8254,
      "mean_token_accuracy": 0.2352230280637741,
      "num_tokens": 41771628.0,
      "step": 18215
    },
    {
      "entropy": 5.345867156982422,
      "epoch": 1.750240153698367,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046972759783042576,
      "loss": 4.8069,
      "mean_token_accuracy": 0.23992937803268433,
      "num_tokens": 41783224.0,
      "step": 18220
    },
    {
      "entropy": 5.212275266647339,
      "epoch": 1.7507204610951008,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004697104200243666,
      "loss": 4.7696,
      "mean_token_accuracy": 0.24328659176826478,
      "num_tokens": 41794899.0,
      "step": 18225
    },
    {
      "entropy": 5.2412882328033445,
      "epoch": 1.751200768491835,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004696932376977096,
      "loss": 4.7566,
      "mean_token_accuracy": 0.24016034752130508,
      "num_tokens": 41806336.0,
      "step": 18230
    },
    {
      "entropy": 5.327985334396362,
      "epoch": 1.7516810758885688,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004696760508508538,
      "loss": 4.8226,
      "mean_token_accuracy": 0.24013479948043823,
      "num_tokens": 41817443.0,
      "step": 18235
    },
    {
      "entropy": 5.289219045639038,
      "epoch": 1.7521613832853027,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046965885948419814,
      "loss": 4.8768,
      "mean_token_accuracy": 0.23399290293455124,
      "num_tokens": 41829740.0,
      "step": 18240
    },
    {
      "entropy": 5.254031753540039,
      "epoch": 1.7526416906820366,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004696416635981418,
      "loss": 4.7851,
      "mean_token_accuracy": 0.24435230642557143,
      "num_tokens": 41840519.0,
      "step": 18245
    },
    {
      "entropy": 5.213440847396851,
      "epoch": 1.7531219980787704,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000469624463193084,
      "loss": 4.7375,
      "mean_token_accuracy": 0.24318318516016008,
      "num_tokens": 41850975.0,
      "step": 18250
    },
    {
      "entropy": 5.255119466781617,
      "epoch": 1.7536023054755043,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000469607258269424,
      "loss": 4.7874,
      "mean_token_accuracy": 0.2373722344636917,
      "num_tokens": 41861854.0,
      "step": 18255
    },
    {
      "entropy": 5.337873125076294,
      "epoch": 1.7540826128722382,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004695900488275614,
      "loss": 4.8991,
      "mean_token_accuracy": 0.23282883167266846,
      "num_tokens": 41873279.0,
      "step": 18260
    },
    {
      "entropy": 5.329365205764771,
      "epoch": 1.754562920268972,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004695728348678957,
      "loss": 4.818,
      "mean_token_accuracy": 0.2313791498541832,
      "num_tokens": 41885545.0,
      "step": 18265
    },
    {
      "entropy": 5.2372079372406,
      "epoch": 1.755043227665706,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004695556163908265,
      "loss": 4.7747,
      "mean_token_accuracy": 0.24084445983171462,
      "num_tokens": 41896168.0,
      "step": 18270
    },
    {
      "entropy": 5.240135765075683,
      "epoch": 1.7555235350624399,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004695383933967536,
      "loss": 4.766,
      "mean_token_accuracy": 0.24127921909093858,
      "num_tokens": 41907423.0,
      "step": 18275
    },
    {
      "entropy": 5.237374210357666,
      "epoch": 1.7560038424591737,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046952116588607694,
      "loss": 4.7543,
      "mean_token_accuracy": 0.24173429906368255,
      "num_tokens": 41918617.0,
      "step": 18280
    },
    {
      "entropy": 5.2626618385314945,
      "epoch": 1.7564841498559076,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004695039338591963,
      "loss": 4.7475,
      "mean_token_accuracy": 0.23882779031991958,
      "num_tokens": 41930235.0,
      "step": 18285
    },
    {
      "entropy": 5.261363983154297,
      "epoch": 1.7569644572526417,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000469486697316512,
      "loss": 4.8586,
      "mean_token_accuracy": 0.23583266884088516,
      "num_tokens": 41941308.0,
      "step": 18290
    },
    {
      "entropy": 5.370620203018189,
      "epoch": 1.7574447646493756,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000469469456258424,
      "loss": 4.9509,
      "mean_token_accuracy": 0.2299128517508507,
      "num_tokens": 41952406.0,
      "step": 18295
    },
    {
      "entropy": 5.349005699157715,
      "epoch": 1.7579250720461095,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004694522106853327,
      "loss": 4.8477,
      "mean_token_accuracy": 0.22449478954076768,
      "num_tokens": 41964642.0,
      "step": 18300
    },
    {
      "entropy": 5.340524673461914,
      "epoch": 1.7584053794428436,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046943496059763845,
      "loss": 4.8507,
      "mean_token_accuracy": 0.23351938724517823,
      "num_tokens": 41975761.0,
      "step": 18305
    },
    {
      "entropy": 5.252025461196899,
      "epoch": 1.7588856868395775,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046941770599574176,
      "loss": 4.7822,
      "mean_token_accuracy": 0.23447389155626297,
      "num_tokens": 41987596.0,
      "step": 18310
    },
    {
      "entropy": 5.26167368888855,
      "epoch": 1.7593659942363113,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004694004468800433,
      "loss": 4.7728,
      "mean_token_accuracy": 0.23380352258682252,
      "num_tokens": 41999148.0,
      "step": 18315
    },
    {
      "entropy": 5.26645975112915,
      "epoch": 1.7598463016330452,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004693831832509437,
      "loss": 4.7691,
      "mean_token_accuracy": 0.2422672525048256,
      "num_tokens": 42010030.0,
      "step": 18320
    },
    {
      "entropy": 5.30007586479187,
      "epoch": 1.760326609029779,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046936591510884375,
      "loss": 4.8881,
      "mean_token_accuracy": 0.23055338561534883,
      "num_tokens": 42022009.0,
      "step": 18325
    },
    {
      "entropy": 5.31683759689331,
      "epoch": 1.760806916426513,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046934864245414443,
      "loss": 4.8477,
      "mean_token_accuracy": 0.23198751956224442,
      "num_tokens": 42033654.0,
      "step": 18330
    },
    {
      "entropy": 5.24234676361084,
      "epoch": 1.7612872238232469,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00046933136528724676,
      "loss": 4.7771,
      "mean_token_accuracy": 0.2364094763994217,
      "num_tokens": 42045521.0,
      "step": 18335
    },
    {
      "entropy": 5.210381889343262,
      "epoch": 1.7617675312199808,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004693140836085518,
      "loss": 4.7932,
      "mean_token_accuracy": 0.23608895838260652,
      "num_tokens": 42057751.0,
      "step": 18340
    },
    {
      "entropy": 5.293923425674438,
      "epoch": 1.7622478386167146,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046929679741846076,
      "loss": 4.8669,
      "mean_token_accuracy": 0.23913259506225587,
      "num_tokens": 42068823.0,
      "step": 18345
    },
    {
      "entropy": 5.262626600265503,
      "epoch": 1.7627281460134485,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046927950671737505,
      "loss": 4.7592,
      "mean_token_accuracy": 0.24107017666101455,
      "num_tokens": 42079856.0,
      "step": 18350
    },
    {
      "entropy": 5.37905797958374,
      "epoch": 1.7632084534101824,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046926221150569617,
      "loss": 4.807,
      "mean_token_accuracy": 0.23149582445621492,
      "num_tokens": 42090560.0,
      "step": 18355
    },
    {
      "entropy": 5.322408723831177,
      "epoch": 1.7636887608069163,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004692449117838255,
      "loss": 4.8873,
      "mean_token_accuracy": 0.2286471724510193,
      "num_tokens": 42102252.0,
      "step": 18360
    },
    {
      "entropy": 5.323352813720703,
      "epoch": 1.7641690682036504,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004692276075521648,
      "loss": 4.9334,
      "mean_token_accuracy": 0.22538111358880997,
      "num_tokens": 42113679.0,
      "step": 18365
    },
    {
      "entropy": 5.282343006134033,
      "epoch": 1.7646493756003843,
      "grad_norm": 1.0,
      "learning_rate": 0.0004692102988111158,
      "loss": 4.8451,
      "mean_token_accuracy": 0.23533279001712798,
      "num_tokens": 42125919.0,
      "step": 18370
    },
    {
      "entropy": 5.431267404556275,
      "epoch": 1.7651296829971181,
      "grad_norm": 1.0,
      "learning_rate": 0.00046919298556108023,
      "loss": 4.9633,
      "mean_token_accuracy": 0.22476842254400253,
      "num_tokens": 42138189.0,
      "step": 18375
    },
    {
      "entropy": 5.371929359436035,
      "epoch": 1.765609990393852,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046917566780246036,
      "loss": 4.872,
      "mean_token_accuracy": 0.236562879383564,
      "num_tokens": 42150232.0,
      "step": 18380
    },
    {
      "entropy": 5.306816577911377,
      "epoch": 1.7660902977905861,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046915834553565793,
      "loss": 4.8287,
      "mean_token_accuracy": 0.23263147175312043,
      "num_tokens": 42160740.0,
      "step": 18385
    },
    {
      "entropy": 5.327568435668946,
      "epoch": 1.76657060518732,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004691410187610753,
      "loss": 4.8268,
      "mean_token_accuracy": 0.24698745012283324,
      "num_tokens": 42171761.0,
      "step": 18390
    },
    {
      "entropy": 5.256658697128296,
      "epoch": 1.767050912584054,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046912368747911465,
      "loss": 4.7815,
      "mean_token_accuracy": 0.23880992233753204,
      "num_tokens": 42183362.0,
      "step": 18395
    },
    {
      "entropy": 5.3579058170318605,
      "epoch": 1.7675312199807878,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046910635169017845,
      "loss": 4.9329,
      "mean_token_accuracy": 0.22689748406410218,
      "num_tokens": 42195610.0,
      "step": 18400
    },
    {
      "entropy": 5.313549518585205,
      "epoch": 1.7680115273775217,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004690890113946691,
      "loss": 4.8132,
      "mean_token_accuracy": 0.23060692548751832,
      "num_tokens": 42207489.0,
      "step": 18405
    },
    {
      "entropy": 5.340543127059936,
      "epoch": 1.7684918347742555,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004690716665929893,
      "loss": 4.8559,
      "mean_token_accuracy": 0.23181567937135697,
      "num_tokens": 42217300.0,
      "step": 18410
    },
    {
      "entropy": 5.328533601760864,
      "epoch": 1.7689721421709894,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046905431728554164,
      "loss": 4.8432,
      "mean_token_accuracy": 0.2365052044391632,
      "num_tokens": 42229031.0,
      "step": 18415
    },
    {
      "entropy": 5.130498361587525,
      "epoch": 1.7694524495677233,
      "grad_norm": 1.0,
      "learning_rate": 0.00046903696347272894,
      "loss": 4.6027,
      "mean_token_accuracy": 0.25382308214902877,
      "num_tokens": 42239102.0,
      "step": 18420
    },
    {
      "entropy": 5.324645233154297,
      "epoch": 1.7699327569644572,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046901960515495413,
      "loss": 4.8063,
      "mean_token_accuracy": 0.23776207864284515,
      "num_tokens": 42250436.0,
      "step": 18425
    },
    {
      "entropy": 5.303100442886352,
      "epoch": 1.770413064361191,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004690022423326202,
      "loss": 4.7607,
      "mean_token_accuracy": 0.24483394473791123,
      "num_tokens": 42260544.0,
      "step": 18430
    },
    {
      "entropy": 5.233433103561401,
      "epoch": 1.770893371757925,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004689848750061303,
      "loss": 4.803,
      "mean_token_accuracy": 0.23276572674512863,
      "num_tokens": 42272519.0,
      "step": 18435
    },
    {
      "entropy": 5.278465318679809,
      "epoch": 1.7713736791546588,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004689675031758876,
      "loss": 4.7713,
      "mean_token_accuracy": 0.24080827385187148,
      "num_tokens": 42281722.0,
      "step": 18440
    },
    {
      "entropy": 5.284533166885376,
      "epoch": 1.771853986551393,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004689501268422954,
      "loss": 4.8687,
      "mean_token_accuracy": 0.24007919877767564,
      "num_tokens": 42294067.0,
      "step": 18445
    },
    {
      "entropy": 5.272848081588745,
      "epoch": 1.7723342939481268,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046893274600575725,
      "loss": 4.794,
      "mean_token_accuracy": 0.2346661478281021,
      "num_tokens": 42305547.0,
      "step": 18450
    },
    {
      "entropy": 5.331240606307984,
      "epoch": 1.7728146013448607,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004689153606666765,
      "loss": 4.8454,
      "mean_token_accuracy": 0.22267400324344636,
      "num_tokens": 42315557.0,
      "step": 18455
    },
    {
      "entropy": 5.336933279037476,
      "epoch": 1.7732949087415946,
      "grad_norm": 0.9375,
      "learning_rate": 0.000468897970825457,
      "loss": 4.9182,
      "mean_token_accuracy": 0.22592617869377135,
      "num_tokens": 42327374.0,
      "step": 18460
    },
    {
      "entropy": 5.253955268859864,
      "epoch": 1.7737752161383287,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046888057648250233,
      "loss": 4.7839,
      "mean_token_accuracy": 0.23818233907222747,
      "num_tokens": 42338889.0,
      "step": 18465
    },
    {
      "entropy": 5.295960712432861,
      "epoch": 1.7742555235350626,
      "grad_norm": 1.125,
      "learning_rate": 0.0004688631776382164,
      "loss": 4.7975,
      "mean_token_accuracy": 0.24139987528324128,
      "num_tokens": 42348678.0,
      "step": 18470
    },
    {
      "entropy": 5.256591558456421,
      "epoch": 1.7747358309317964,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046884577429300305,
      "loss": 4.8556,
      "mean_token_accuracy": 0.2365972802042961,
      "num_tokens": 42359090.0,
      "step": 18475
    },
    {
      "entropy": 5.301103830337524,
      "epoch": 1.7752161383285303,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004688283664472665,
      "loss": 4.8024,
      "mean_token_accuracy": 0.22968147844076156,
      "num_tokens": 42370947.0,
      "step": 18480
    },
    {
      "entropy": 5.28152904510498,
      "epoch": 1.7756964457252642,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046881095410141084,
      "loss": 4.8265,
      "mean_token_accuracy": 0.23526912480592727,
      "num_tokens": 42381597.0,
      "step": 18485
    },
    {
      "entropy": 5.310621976852417,
      "epoch": 1.776176753121998,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046879353725584036,
      "loss": 4.7855,
      "mean_token_accuracy": 0.23630943298339843,
      "num_tokens": 42391877.0,
      "step": 18490
    },
    {
      "entropy": 5.268912363052368,
      "epoch": 1.776657060518732,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046877611591095923,
      "loss": 4.7752,
      "mean_token_accuracy": 0.23460371196269988,
      "num_tokens": 42403916.0,
      "step": 18495
    },
    {
      "entropy": 5.313749933242798,
      "epoch": 1.7771373679154658,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046875869006717224,
      "loss": 4.8437,
      "mean_token_accuracy": 0.23076467216014862,
      "num_tokens": 42415624.0,
      "step": 18500
    },
    {
      "entropy": 5.304334115982056,
      "epoch": 1.7776176753121997,
      "grad_norm": 1.0,
      "learning_rate": 0.00046874125972488375,
      "loss": 4.8113,
      "mean_token_accuracy": 0.23889951556921005,
      "num_tokens": 42427228.0,
      "step": 18505
    },
    {
      "entropy": 5.283728694915771,
      "epoch": 1.7780979827089336,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046872382488449853,
      "loss": 4.8517,
      "mean_token_accuracy": 0.23035478442907334,
      "num_tokens": 42440023.0,
      "step": 18510
    },
    {
      "entropy": 5.28707447052002,
      "epoch": 1.7785782901056675,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046870638554642133,
      "loss": 4.7628,
      "mean_token_accuracy": 0.2403624564409256,
      "num_tokens": 42451126.0,
      "step": 18515
    },
    {
      "entropy": 5.249801397323608,
      "epoch": 1.7790585975024016,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046868894171105704,
      "loss": 4.8511,
      "mean_token_accuracy": 0.22726238667964935,
      "num_tokens": 42462975.0,
      "step": 18520
    },
    {
      "entropy": 5.237307929992676,
      "epoch": 1.7795389048991355,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004686714933788107,
      "loss": 4.7623,
      "mean_token_accuracy": 0.24028223007917404,
      "num_tokens": 42474716.0,
      "step": 18525
    },
    {
      "entropy": 5.3228363513946535,
      "epoch": 1.7800192122958693,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004686540405500873,
      "loss": 4.8225,
      "mean_token_accuracy": 0.24037611931562425,
      "num_tokens": 42484616.0,
      "step": 18530
    },
    {
      "entropy": 5.292559957504272,
      "epoch": 1.7804995196926032,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004686365832252922,
      "loss": 4.7714,
      "mean_token_accuracy": 0.23498354107141495,
      "num_tokens": 42494840.0,
      "step": 18535
    },
    {
      "entropy": 5.301965761184692,
      "epoch": 1.7809798270893373,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046861912140483056,
      "loss": 4.8259,
      "mean_token_accuracy": 0.23265385627746582,
      "num_tokens": 42505957.0,
      "step": 18540
    },
    {
      "entropy": 5.355477333068848,
      "epoch": 1.7814601344860712,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046860165508910787,
      "loss": 4.9811,
      "mean_token_accuracy": 0.22426350563764572,
      "num_tokens": 42517861.0,
      "step": 18545
    },
    {
      "entropy": 5.239961910247803,
      "epoch": 1.781940441882805,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004685841842785296,
      "loss": 4.7521,
      "mean_token_accuracy": 0.23338561952114106,
      "num_tokens": 42529986.0,
      "step": 18550
    },
    {
      "entropy": 5.251360082626343,
      "epoch": 1.782420749279539,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004685667089735014,
      "loss": 4.7606,
      "mean_token_accuracy": 0.2402243435382843,
      "num_tokens": 42541778.0,
      "step": 18555
    },
    {
      "entropy": 5.314540576934815,
      "epoch": 1.7829010566762729,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046854922917442907,
      "loss": 4.8304,
      "mean_token_accuracy": 0.232177209854126,
      "num_tokens": 42553776.0,
      "step": 18560
    },
    {
      "entropy": 5.306165885925293,
      "epoch": 1.7833813640730067,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004685317448817182,
      "loss": 4.8469,
      "mean_token_accuracy": 0.2351295158267021,
      "num_tokens": 42566277.0,
      "step": 18565
    },
    {
      "entropy": 5.169776248931885,
      "epoch": 1.7838616714697406,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004685142560957751,
      "loss": 4.6872,
      "mean_token_accuracy": 0.24271001070737838,
      "num_tokens": 42576598.0,
      "step": 18570
    },
    {
      "entropy": 5.240270185470581,
      "epoch": 1.7843419788664745,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004684967628170054,
      "loss": 4.7677,
      "mean_token_accuracy": 0.23577314168214797,
      "num_tokens": 42587850.0,
      "step": 18575
    },
    {
      "entropy": 5.25745644569397,
      "epoch": 1.7848222862632084,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046847926504581553,
      "loss": 4.7944,
      "mean_token_accuracy": 0.23540612012147905,
      "num_tokens": 42598754.0,
      "step": 18580
    },
    {
      "entropy": 5.329287719726563,
      "epoch": 1.7853025936599423,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004684617627826116,
      "loss": 4.8251,
      "mean_token_accuracy": 0.2394539847970009,
      "num_tokens": 42609842.0,
      "step": 18585
    },
    {
      "entropy": 5.302327299118042,
      "epoch": 1.7857829010566761,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004684442560278001,
      "loss": 4.7382,
      "mean_token_accuracy": 0.23938652127981186,
      "num_tokens": 42621641.0,
      "step": 18590
    },
    {
      "entropy": 5.261514902114868,
      "epoch": 1.78626320845341,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046842674478178727,
      "loss": 4.76,
      "mean_token_accuracy": 0.23567525297403336,
      "num_tokens": 42632523.0,
      "step": 18595
    },
    {
      "entropy": 5.339730215072632,
      "epoch": 1.7867435158501441,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004684092290449798,
      "loss": 4.8568,
      "mean_token_accuracy": 0.23160876333713531,
      "num_tokens": 42643511.0,
      "step": 18600
    },
    {
      "entropy": 5.284294557571411,
      "epoch": 1.787223823246878,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004683917088177844,
      "loss": 4.7673,
      "mean_token_accuracy": 0.2420673757791519,
      "num_tokens": 42655077.0,
      "step": 18605
    },
    {
      "entropy": 5.350880718231201,
      "epoch": 1.7877041306436119,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004683741841006077,
      "loss": 4.9066,
      "mean_token_accuracy": 0.23345645070075988,
      "num_tokens": 42667454.0,
      "step": 18610
    },
    {
      "entropy": 5.3063802242279055,
      "epoch": 1.7881844380403458,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004683566548938567,
      "loss": 4.7838,
      "mean_token_accuracy": 0.23543807566165925,
      "num_tokens": 42678214.0,
      "step": 18615
    },
    {
      "entropy": 5.233328819274902,
      "epoch": 1.7886647454370799,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004683391211979383,
      "loss": 4.7916,
      "mean_token_accuracy": 0.23738272190093995,
      "num_tokens": 42689424.0,
      "step": 18620
    },
    {
      "entropy": 5.406211042404175,
      "epoch": 1.7891450528338138,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004683215830132597,
      "loss": 4.9222,
      "mean_token_accuracy": 0.22522022128105162,
      "num_tokens": 42701679.0,
      "step": 18625
    },
    {
      "entropy": 5.306813478469849,
      "epoch": 1.7896253602305476,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046830404034022786,
      "loss": 4.797,
      "mean_token_accuracy": 0.23889809250831603,
      "num_tokens": 42711931.0,
      "step": 18630
    },
    {
      "entropy": 5.330230236053467,
      "epoch": 1.7901056676272815,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004682864931792502,
      "loss": 4.8833,
      "mean_token_accuracy": 0.23262309581041335,
      "num_tokens": 42723460.0,
      "step": 18635
    },
    {
      "entropy": 5.299175882339478,
      "epoch": 1.7905859750240154,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004682689415307342,
      "loss": 4.8046,
      "mean_token_accuracy": 0.23687044829130172,
      "num_tokens": 42735389.0,
      "step": 18640
    },
    {
      "entropy": 5.259147882461548,
      "epoch": 1.7910662824207493,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004682513853950872,
      "loss": 4.7857,
      "mean_token_accuracy": 0.2377074956893921,
      "num_tokens": 42747779.0,
      "step": 18645
    },
    {
      "entropy": 5.2360069274902346,
      "epoch": 1.7915465898174832,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000468233824772717,
      "loss": 4.7969,
      "mean_token_accuracy": 0.23320552408695222,
      "num_tokens": 42758776.0,
      "step": 18650
    },
    {
      "entropy": 5.434389257431031,
      "epoch": 1.792026897214217,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000468216259664031,
      "loss": 4.9199,
      "mean_token_accuracy": 0.22862340658903121,
      "num_tokens": 42770234.0,
      "step": 18655
    },
    {
      "entropy": 5.316122198104859,
      "epoch": 1.792507204610951,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046819869006943727,
      "loss": 4.8127,
      "mean_token_accuracy": 0.23929618597030639,
      "num_tokens": 42782076.0,
      "step": 18660
    },
    {
      "entropy": 5.249883937835693,
      "epoch": 1.7929875120076848,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004681811159893436,
      "loss": 4.7815,
      "mean_token_accuracy": 0.24016362726688384,
      "num_tokens": 42793620.0,
      "step": 18665
    },
    {
      "entropy": 5.311093139648437,
      "epoch": 1.7934678194044187,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046816353742415814,
      "loss": 4.8574,
      "mean_token_accuracy": 0.23997083455324172,
      "num_tokens": 42805125.0,
      "step": 18670
    },
    {
      "entropy": 5.297583293914795,
      "epoch": 1.7939481268011528,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046814595437428885,
      "loss": 4.7514,
      "mean_token_accuracy": 0.23955140113830567,
      "num_tokens": 42815472.0,
      "step": 18675
    },
    {
      "entropy": 5.260460090637207,
      "epoch": 1.7944284341978867,
      "grad_norm": 1.03125,
      "learning_rate": 0.000468128366840144,
      "loss": 4.8231,
      "mean_token_accuracy": 0.2375444918870926,
      "num_tokens": 42826822.0,
      "step": 18680
    },
    {
      "entropy": 5.292391967773438,
      "epoch": 1.7949087415946205,
      "grad_norm": 1.03125,
      "learning_rate": 0.000468110774822132,
      "loss": 4.8232,
      "mean_token_accuracy": 0.23744536340236663,
      "num_tokens": 42839024.0,
      "step": 18685
    },
    {
      "entropy": 5.284437131881714,
      "epoch": 1.7953890489913544,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004680931783206612,
      "loss": 4.7665,
      "mean_token_accuracy": 0.2365342915058136,
      "num_tokens": 42849782.0,
      "step": 18690
    },
    {
      "entropy": 5.298311042785644,
      "epoch": 1.7958693563880885,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046807557733614014,
      "loss": 4.8062,
      "mean_token_accuracy": 0.23243419528007508,
      "num_tokens": 42859792.0,
      "step": 18695
    },
    {
      "entropy": 5.260026979446411,
      "epoch": 1.7963496637848224,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046805797186897757,
      "loss": 4.88,
      "mean_token_accuracy": 0.23045288473367692,
      "num_tokens": 42872031.0,
      "step": 18700
    },
    {
      "entropy": 5.279284191131592,
      "epoch": 1.7968299711815563,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046804036191958206,
      "loss": 4.8343,
      "mean_token_accuracy": 0.23896762430667878,
      "num_tokens": 42882830.0,
      "step": 18705
    },
    {
      "entropy": 5.372193336486816,
      "epoch": 1.7973102785782902,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046802274748836267,
      "loss": 4.8421,
      "mean_token_accuracy": 0.23221004754304886,
      "num_tokens": 42893576.0,
      "step": 18710
    },
    {
      "entropy": 5.286179494857788,
      "epoch": 1.797790585975024,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004680051285757281,
      "loss": 4.7889,
      "mean_token_accuracy": 0.2400351196527481,
      "num_tokens": 42904405.0,
      "step": 18715
    },
    {
      "entropy": 5.272048711776733,
      "epoch": 1.798270893371758,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004679875051820877,
      "loss": 4.8197,
      "mean_token_accuracy": 0.23256647884845733,
      "num_tokens": 42916601.0,
      "step": 18720
    },
    {
      "entropy": 5.273844814300537,
      "epoch": 1.7987512007684918,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004679698773078503,
      "loss": 4.8448,
      "mean_token_accuracy": 0.2285892441868782,
      "num_tokens": 42928977.0,
      "step": 18725
    },
    {
      "entropy": 5.314793872833252,
      "epoch": 1.7992315081652257,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046795224495342554,
      "loss": 4.7507,
      "mean_token_accuracy": 0.23452287912368774,
      "num_tokens": 42940149.0,
      "step": 18730
    },
    {
      "entropy": 5.34348497390747,
      "epoch": 1.7997118155619596,
      "grad_norm": 1.0,
      "learning_rate": 0.00046793460811922255,
      "loss": 4.8794,
      "mean_token_accuracy": 0.22910443991422652,
      "num_tokens": 42952122.0,
      "step": 18735
    },
    {
      "entropy": 5.264488601684571,
      "epoch": 1.8001921229586935,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046791696680565075,
      "loss": 4.7765,
      "mean_token_accuracy": 0.23779892027378083,
      "num_tokens": 42962984.0,
      "step": 18740
    },
    {
      "entropy": 5.366231536865234,
      "epoch": 1.8006724303554273,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046789932101312003,
      "loss": 4.8539,
      "mean_token_accuracy": 0.22653574943542482,
      "num_tokens": 42974409.0,
      "step": 18745
    },
    {
      "entropy": 5.367153692245483,
      "epoch": 1.8011527377521612,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004678816707420397,
      "loss": 4.8827,
      "mean_token_accuracy": 0.22927067577838897,
      "num_tokens": 42986600.0,
      "step": 18750
    },
    {
      "entropy": 5.24708571434021,
      "epoch": 1.8016330451488953,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004678640159928198,
      "loss": 4.7747,
      "mean_token_accuracy": 0.23978012204170226,
      "num_tokens": 42998192.0,
      "step": 18755
    },
    {
      "entropy": 5.285899019241333,
      "epoch": 1.8021133525456292,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004678463567658701,
      "loss": 4.7477,
      "mean_token_accuracy": 0.2445960134267807,
      "num_tokens": 43009680.0,
      "step": 18760
    },
    {
      "entropy": 5.2341468334198,
      "epoch": 1.802593659942363,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004678286930616006,
      "loss": 4.7988,
      "mean_token_accuracy": 0.2358892247080803,
      "num_tokens": 43021031.0,
      "step": 18765
    },
    {
      "entropy": 5.287977457046509,
      "epoch": 1.803073967339097,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004678110248804215,
      "loss": 4.8204,
      "mean_token_accuracy": 0.243240787088871,
      "num_tokens": 43031423.0,
      "step": 18770
    },
    {
      "entropy": 5.302126455307007,
      "epoch": 1.803554274735831,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046779335222274293,
      "loss": 4.8446,
      "mean_token_accuracy": 0.23734294921159743,
      "num_tokens": 43042895.0,
      "step": 18775
    },
    {
      "entropy": 5.317495679855346,
      "epoch": 1.804034582132565,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046777567508897515,
      "loss": 4.9301,
      "mean_token_accuracy": 0.22341910749673843,
      "num_tokens": 43053052.0,
      "step": 18780
    },
    {
      "entropy": 5.3661096572875975,
      "epoch": 1.8045148895292988,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046775799347952864,
      "loss": 4.8847,
      "mean_token_accuracy": 0.22652423828840257,
      "num_tokens": 43066908.0,
      "step": 18785
    },
    {
      "entropy": 5.325060224533081,
      "epoch": 1.8049951969260327,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004677403073948139,
      "loss": 4.8725,
      "mean_token_accuracy": 0.23547277003526687,
      "num_tokens": 43079736.0,
      "step": 18790
    },
    {
      "entropy": 5.308591651916504,
      "epoch": 1.8054755043227666,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004677226168352416,
      "loss": 4.8955,
      "mean_token_accuracy": 0.2287858694791794,
      "num_tokens": 43091600.0,
      "step": 18795
    },
    {
      "entropy": 5.306531381607056,
      "epoch": 1.8059558117195005,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004677049218012223,
      "loss": 4.817,
      "mean_token_accuracy": 0.24084895700216294,
      "num_tokens": 43103544.0,
      "step": 18800
    },
    {
      "entropy": 5.3227454662323,
      "epoch": 1.8064361191162344,
      "grad_norm": 1.0625,
      "learning_rate": 0.000467687222293167,
      "loss": 4.8062,
      "mean_token_accuracy": 0.23683811277151107,
      "num_tokens": 43113533.0,
      "step": 18805
    },
    {
      "entropy": 5.222534561157227,
      "epoch": 1.8069164265129682,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004676695183114866,
      "loss": 4.7797,
      "mean_token_accuracy": 0.2374231368303299,
      "num_tokens": 43124917.0,
      "step": 18810
    },
    {
      "entropy": 5.262121868133545,
      "epoch": 1.8073967339097021,
      "grad_norm": 1.03125,
      "learning_rate": 0.000467651809856592,
      "loss": 4.8037,
      "mean_token_accuracy": 0.23436479419469833,
      "num_tokens": 43136808.0,
      "step": 18815
    },
    {
      "entropy": 5.359499168395996,
      "epoch": 1.807877041306436,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046763409692889446,
      "loss": 4.8759,
      "mean_token_accuracy": 0.22702230364084244,
      "num_tokens": 43148567.0,
      "step": 18820
    },
    {
      "entropy": 5.227659702301025,
      "epoch": 1.8083573487031699,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046761637952880516,
      "loss": 4.7071,
      "mean_token_accuracy": 0.24262206703424455,
      "num_tokens": 43160343.0,
      "step": 18825
    },
    {
      "entropy": 5.32520637512207,
      "epoch": 1.808837656099904,
      "grad_norm": 1.0,
      "learning_rate": 0.00046759865765673555,
      "loss": 4.9728,
      "mean_token_accuracy": 0.22117119431495666,
      "num_tokens": 43173546.0,
      "step": 18830
    },
    {
      "entropy": 5.357563161849976,
      "epoch": 1.8093179634966379,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000467580931313097,
      "loss": 4.8685,
      "mean_token_accuracy": 0.237652887403965,
      "num_tokens": 43183846.0,
      "step": 18835
    },
    {
      "entropy": 5.428438758850097,
      "epoch": 1.8097982708933718,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046756320049830106,
      "loss": 4.8773,
      "mean_token_accuracy": 0.23059234023094177,
      "num_tokens": 43195612.0,
      "step": 18840
    },
    {
      "entropy": 5.229885149002075,
      "epoch": 1.8102785782901056,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004675454652127594,
      "loss": 4.8358,
      "mean_token_accuracy": 0.23173980861902238,
      "num_tokens": 43207439.0,
      "step": 18845
    },
    {
      "entropy": 5.322157096862793,
      "epoch": 1.8107588856868397,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046752772545688377,
      "loss": 4.8285,
      "mean_token_accuracy": 0.22599481791257858,
      "num_tokens": 43218066.0,
      "step": 18850
    },
    {
      "entropy": 5.328280591964722,
      "epoch": 1.8112391930835736,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004675099812310861,
      "loss": 4.8373,
      "mean_token_accuracy": 0.23097915351390838,
      "num_tokens": 43230679.0,
      "step": 18855
    },
    {
      "entropy": 5.299061012268067,
      "epoch": 1.8117195004803075,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004674922325357782,
      "loss": 4.8652,
      "mean_token_accuracy": 0.23573773801326753,
      "num_tokens": 43241566.0,
      "step": 18860
    },
    {
      "entropy": 5.401655960083008,
      "epoch": 1.8121998078770414,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046747447937137235,
      "loss": 4.908,
      "mean_token_accuracy": 0.22701998949050903,
      "num_tokens": 43252630.0,
      "step": 18865
    },
    {
      "entropy": 5.415476322174072,
      "epoch": 1.8126801152737753,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046745672173828057,
      "loss": 5.0022,
      "mean_token_accuracy": 0.22112301886081695,
      "num_tokens": 43264879.0,
      "step": 18870
    },
    {
      "entropy": 5.252551937103272,
      "epoch": 1.8131604226705091,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004674389596369151,
      "loss": 4.8188,
      "mean_token_accuracy": 0.2332765504717827,
      "num_tokens": 43277321.0,
      "step": 18875
    },
    {
      "entropy": 5.330219316482544,
      "epoch": 1.813640730067243,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046742119306768855,
      "loss": 4.8734,
      "mean_token_accuracy": 0.23502857089042664,
      "num_tokens": 43289711.0,
      "step": 18880
    },
    {
      "entropy": 5.332897663116455,
      "epoch": 1.814121037463977,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004674034220310132,
      "loss": 4.8711,
      "mean_token_accuracy": 0.23039526641368865,
      "num_tokens": 43301372.0,
      "step": 18885
    },
    {
      "entropy": 5.286749982833863,
      "epoch": 1.8146013448607108,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046738564652730176,
      "loss": 4.7257,
      "mean_token_accuracy": 0.24138181358575822,
      "num_tokens": 43312960.0,
      "step": 18890
    },
    {
      "entropy": 5.283509731292725,
      "epoch": 1.8150816522574447,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004673678665569669,
      "loss": 4.799,
      "mean_token_accuracy": 0.23158974647521974,
      "num_tokens": 43325039.0,
      "step": 18895
    },
    {
      "entropy": 5.341380405426025,
      "epoch": 1.8155619596541785,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004673500821204213,
      "loss": 4.9071,
      "mean_token_accuracy": 0.23115096390247344,
      "num_tokens": 43337064.0,
      "step": 18900
    },
    {
      "entropy": 5.362489318847656,
      "epoch": 1.8160422670509124,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000467332293218078,
      "loss": 4.8094,
      "mean_token_accuracy": 0.23090253472328187,
      "num_tokens": 43349307.0,
      "step": 18905
    },
    {
      "entropy": 5.257839965820312,
      "epoch": 1.8165225744476465,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046731449985035,
      "loss": 4.7979,
      "mean_token_accuracy": 0.24726089090108871,
      "num_tokens": 43362566.0,
      "step": 18910
    },
    {
      "entropy": 5.295166730880737,
      "epoch": 1.8170028818443804,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046729670201765036,
      "loss": 4.9074,
      "mean_token_accuracy": 0.2280411273241043,
      "num_tokens": 43374516.0,
      "step": 18915
    },
    {
      "entropy": 5.294977378845215,
      "epoch": 1.8174831892411143,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046727889972039227,
      "loss": 4.7086,
      "mean_token_accuracy": 0.24329839050769805,
      "num_tokens": 43386615.0,
      "step": 18920
    },
    {
      "entropy": 5.344839334487915,
      "epoch": 1.8179634966378482,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046726109295898904,
      "loss": 4.8478,
      "mean_token_accuracy": 0.23164843916893005,
      "num_tokens": 43396582.0,
      "step": 18925
    },
    {
      "entropy": 5.245763301849365,
      "epoch": 1.8184438040345823,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004672432817338542,
      "loss": 4.7576,
      "mean_token_accuracy": 0.2402382805943489,
      "num_tokens": 43407617.0,
      "step": 18930
    },
    {
      "entropy": 5.3233483791351315,
      "epoch": 1.8189241114313162,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046722546604540115,
      "loss": 4.8923,
      "mean_token_accuracy": 0.22882914394140244,
      "num_tokens": 43421048.0,
      "step": 18935
    },
    {
      "entropy": 5.38718957901001,
      "epoch": 1.81940441882805,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004672076458940436,
      "loss": 4.8757,
      "mean_token_accuracy": 0.23395396023988724,
      "num_tokens": 43433325.0,
      "step": 18940
    },
    {
      "entropy": 5.249499273300171,
      "epoch": 1.819884726224784,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046718982128019534,
      "loss": 4.7411,
      "mean_token_accuracy": 0.23308519423007965,
      "num_tokens": 43444817.0,
      "step": 18945
    },
    {
      "entropy": 5.333471202850342,
      "epoch": 1.8203650336215178,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046717199220427003,
      "loss": 4.8765,
      "mean_token_accuracy": 0.23212577253580094,
      "num_tokens": 43456295.0,
      "step": 18950
    },
    {
      "entropy": 5.321954298019409,
      "epoch": 1.8208453410182517,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046715415866668163,
      "loss": 4.8074,
      "mean_token_accuracy": 0.2379835531115532,
      "num_tokens": 43468218.0,
      "step": 18955
    },
    {
      "entropy": 5.296234130859375,
      "epoch": 1.8213256484149856,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004671363206678443,
      "loss": 4.8492,
      "mean_token_accuracy": 0.23470766097307205,
      "num_tokens": 43480182.0,
      "step": 18960
    },
    {
      "entropy": 5.339520788192749,
      "epoch": 1.8218059558117194,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046711847820817215,
      "loss": 4.8589,
      "mean_token_accuracy": 0.2305966079235077,
      "num_tokens": 43491108.0,
      "step": 18965
    },
    {
      "entropy": 5.29542646408081,
      "epoch": 1.8222862632084533,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004671006312880794,
      "loss": 4.7747,
      "mean_token_accuracy": 0.24642649590969085,
      "num_tokens": 43503530.0,
      "step": 18970
    },
    {
      "entropy": 5.334644365310669,
      "epoch": 1.8227665706051872,
      "grad_norm": 1.0,
      "learning_rate": 0.0004670827799079805,
      "loss": 4.8835,
      "mean_token_accuracy": 0.23908702433109283,
      "num_tokens": 43514805.0,
      "step": 18975
    },
    {
      "entropy": 5.3363378047943115,
      "epoch": 1.823246878001921,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046706492406828966,
      "loss": 4.9246,
      "mean_token_accuracy": 0.23286825865507127,
      "num_tokens": 43526027.0,
      "step": 18980
    },
    {
      "entropy": 5.3238475799560545,
      "epoch": 1.8237271853986552,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004670470637694217,
      "loss": 4.8904,
      "mean_token_accuracy": 0.23265611827373506,
      "num_tokens": 43537191.0,
      "step": 18985
    },
    {
      "entropy": 5.384231281280518,
      "epoch": 1.824207492795389,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004670291990117912,
      "loss": 4.8383,
      "mean_token_accuracy": 0.2362362787127495,
      "num_tokens": 43547494.0,
      "step": 18990
    },
    {
      "entropy": 5.296970176696777,
      "epoch": 1.824687800192123,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004670113297958128,
      "loss": 4.7774,
      "mean_token_accuracy": 0.24309783428907394,
      "num_tokens": 43559048.0,
      "step": 18995
    },
    {
      "entropy": 5.366056299209594,
      "epoch": 1.8251681075888568,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046699345612190155,
      "loss": 4.8547,
      "mean_token_accuracy": 0.22888772338628768,
      "num_tokens": 43569571.0,
      "step": 19000
    },
    {
      "entropy": 5.305092811584473,
      "epoch": 1.825648414985591,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046697557799047233,
      "loss": 4.7986,
      "mean_token_accuracy": 0.23604153394699096,
      "num_tokens": 43580092.0,
      "step": 19005
    },
    {
      "entropy": 5.2296116828918455,
      "epoch": 1.8261287223823248,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004669576954019403,
      "loss": 4.7545,
      "mean_token_accuracy": 0.2362995058298111,
      "num_tokens": 43592867.0,
      "step": 19010
    },
    {
      "entropy": 5.308857154846192,
      "epoch": 1.8266090297790587,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004669398083567205,
      "loss": 4.8278,
      "mean_token_accuracy": 0.2354874923825264,
      "num_tokens": 43602618.0,
      "step": 19015
    },
    {
      "entropy": 5.353134202957153,
      "epoch": 1.8270893371757926,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004669219168552284,
      "loss": 4.8387,
      "mean_token_accuracy": 0.23541273772716523,
      "num_tokens": 43614425.0,
      "step": 19020
    },
    {
      "entropy": 5.210196685791016,
      "epoch": 1.8275696445725265,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00046690402089787916,
      "loss": 4.7054,
      "mean_token_accuracy": 0.24028224349021912,
      "num_tokens": 43627334.0,
      "step": 19025
    },
    {
      "entropy": 5.2840166091918945,
      "epoch": 1.8280499519692603,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004668861204850884,
      "loss": 4.8075,
      "mean_token_accuracy": 0.23622317612171173,
      "num_tokens": 43638509.0,
      "step": 19030
    },
    {
      "entropy": 5.277458524703979,
      "epoch": 1.8285302593659942,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046686821561727176,
      "loss": 4.7669,
      "mean_token_accuracy": 0.2498920053243637,
      "num_tokens": 43649633.0,
      "step": 19035
    },
    {
      "entropy": 5.3498152732849125,
      "epoch": 1.829010566762728,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004668503062948449,
      "loss": 4.8455,
      "mean_token_accuracy": 0.23232366144657135,
      "num_tokens": 43661175.0,
      "step": 19040
    },
    {
      "entropy": 5.326530027389526,
      "epoch": 1.829490874159462,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004668323925182236,
      "loss": 4.8439,
      "mean_token_accuracy": 0.23198689073324202,
      "num_tokens": 43672000.0,
      "step": 19045
    },
    {
      "entropy": 5.292389392852783,
      "epoch": 1.8299711815561959,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046681447428782377,
      "loss": 4.7801,
      "mean_token_accuracy": 0.24051085561513902,
      "num_tokens": 43682574.0,
      "step": 19050
    },
    {
      "entropy": 5.303547763824463,
      "epoch": 1.8304514889529298,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004667965516040613,
      "loss": 4.8871,
      "mean_token_accuracy": 0.2300344005227089,
      "num_tokens": 43693051.0,
      "step": 19055
    },
    {
      "entropy": 5.386662721633911,
      "epoch": 1.8309317963496636,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004667786244673526,
      "loss": 4.813,
      "mean_token_accuracy": 0.23347243517637253,
      "num_tokens": 43705471.0,
      "step": 19060
    },
    {
      "entropy": 5.265243148803711,
      "epoch": 1.8314121037463977,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046676069287811365,
      "loss": 4.78,
      "mean_token_accuracy": 0.2425997093319893,
      "num_tokens": 43716676.0,
      "step": 19065
    },
    {
      "entropy": 5.374091243743896,
      "epoch": 1.8318924111431316,
      "grad_norm": 1.125,
      "learning_rate": 0.0004667427568367607,
      "loss": 4.8803,
      "mean_token_accuracy": 0.23042716532945634,
      "num_tokens": 43727997.0,
      "step": 19070
    },
    {
      "entropy": 5.301415729522705,
      "epoch": 1.8323727185398655,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046672481634371047,
      "loss": 4.8355,
      "mean_token_accuracy": 0.23420198857784272,
      "num_tokens": 43739163.0,
      "step": 19075
    },
    {
      "entropy": 5.3486591339111325,
      "epoch": 1.8328530259365994,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046670687139937925,
      "loss": 4.8227,
      "mean_token_accuracy": 0.23251170217990874,
      "num_tokens": 43751442.0,
      "step": 19080
    },
    {
      "entropy": 5.290380811691284,
      "epoch": 1.8333333333333335,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004666889220041837,
      "loss": 4.878,
      "mean_token_accuracy": 0.2279914379119873,
      "num_tokens": 43763430.0,
      "step": 19085
    },
    {
      "entropy": 5.288370990753174,
      "epoch": 1.8338136407300674,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046667096815854056,
      "loss": 4.7493,
      "mean_token_accuracy": 0.24771594107151032,
      "num_tokens": 43774660.0,
      "step": 19090
    },
    {
      "entropy": 5.248483085632325,
      "epoch": 1.8342939481268012,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004666530098628667,
      "loss": 4.7351,
      "mean_token_accuracy": 0.24249210506677626,
      "num_tokens": 43785997.0,
      "step": 19095
    },
    {
      "entropy": 5.292431497573853,
      "epoch": 1.8347742555235351,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004666350471175791,
      "loss": 4.8161,
      "mean_token_accuracy": 0.23094668835401536,
      "num_tokens": 43797744.0,
      "step": 19100
    },
    {
      "entropy": 5.325319004058838,
      "epoch": 1.835254562920269,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004666170799230947,
      "loss": 4.8829,
      "mean_token_accuracy": 0.22703109830617904,
      "num_tokens": 43810415.0,
      "step": 19105
    },
    {
      "entropy": 5.21659836769104,
      "epoch": 1.8357348703170029,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004665991082798307,
      "loss": 4.7701,
      "mean_token_accuracy": 0.241365709900856,
      "num_tokens": 43822652.0,
      "step": 19110
    },
    {
      "entropy": 5.3830193996429445,
      "epoch": 1.8362151777137368,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004665811321882043,
      "loss": 4.8619,
      "mean_token_accuracy": 0.23089806139469146,
      "num_tokens": 43834607.0,
      "step": 19115
    },
    {
      "entropy": 5.302348709106445,
      "epoch": 1.8366954851104706,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046656315164863297,
      "loss": 4.8348,
      "mean_token_accuracy": 0.23506847620010377,
      "num_tokens": 43845124.0,
      "step": 19120
    },
    {
      "entropy": 5.3760007381439205,
      "epoch": 1.8371757925072045,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046654516666153403,
      "loss": 4.8945,
      "mean_token_accuracy": 0.23012445718050004,
      "num_tokens": 43855964.0,
      "step": 19125
    },
    {
      "entropy": 5.320136308670044,
      "epoch": 1.8376560999039384,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004665271772273251,
      "loss": 4.7626,
      "mean_token_accuracy": 0.24191076606512069,
      "num_tokens": 43866659.0,
      "step": 19130
    },
    {
      "entropy": 5.334185361862183,
      "epoch": 1.8381364073006723,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004665091833464239,
      "loss": 4.8743,
      "mean_token_accuracy": 0.23396852910518645,
      "num_tokens": 43877410.0,
      "step": 19135
    },
    {
      "entropy": 5.311033916473389,
      "epoch": 1.8386167146974062,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046649118501924805,
      "loss": 4.9047,
      "mean_token_accuracy": 0.2267853155732155,
      "num_tokens": 43889672.0,
      "step": 19140
    },
    {
      "entropy": 5.338738346099854,
      "epoch": 1.8390970220941403,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004664731822462154,
      "loss": 4.8636,
      "mean_token_accuracy": 0.23965682089328766,
      "num_tokens": 43900876.0,
      "step": 19145
    },
    {
      "entropy": 5.31527156829834,
      "epoch": 1.8395773294908742,
      "grad_norm": 1.0,
      "learning_rate": 0.00046645517502774415,
      "loss": 4.7306,
      "mean_token_accuracy": 0.243075168132782,
      "num_tokens": 43911460.0,
      "step": 19150
    },
    {
      "entropy": 5.288602352142334,
      "epoch": 1.840057636887608,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046643716336425224,
      "loss": 4.8176,
      "mean_token_accuracy": 0.2370465248823166,
      "num_tokens": 43922344.0,
      "step": 19155
    },
    {
      "entropy": 5.27722578048706,
      "epoch": 1.8405379442843421,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004664191472561578,
      "loss": 4.7519,
      "mean_token_accuracy": 0.24020490050315857,
      "num_tokens": 43933960.0,
      "step": 19160
    },
    {
      "entropy": 5.330636501312256,
      "epoch": 1.841018251681076,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004664011267038792,
      "loss": 4.9566,
      "mean_token_accuracy": 0.22862802147865297,
      "num_tokens": 43945805.0,
      "step": 19165
    },
    {
      "entropy": 5.334782218933105,
      "epoch": 1.84149855907781,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046638310170783476,
      "loss": 4.8676,
      "mean_token_accuracy": 0.23737662732601167,
      "num_tokens": 43958318.0,
      "step": 19170
    },
    {
      "entropy": 5.282655572891235,
      "epoch": 1.8419788664745438,
      "grad_norm": 1.03125,
      "learning_rate": 0.000466365072268443,
      "loss": 4.7492,
      "mean_token_accuracy": 0.24397037029266358,
      "num_tokens": 43969385.0,
      "step": 19175
    },
    {
      "entropy": 5.276175117492675,
      "epoch": 1.8424591738712777,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004663470383861225,
      "loss": 4.8143,
      "mean_token_accuracy": 0.23683173656463624,
      "num_tokens": 43980383.0,
      "step": 19180
    },
    {
      "entropy": 5.297229146957397,
      "epoch": 1.8429394812680115,
      "grad_norm": 1.0,
      "learning_rate": 0.000466329000061292,
      "loss": 4.8532,
      "mean_token_accuracy": 0.2335292175412178,
      "num_tokens": 43992900.0,
      "step": 19185
    },
    {
      "entropy": 5.392330074310303,
      "epoch": 1.8434197886647454,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004663109572943702,
      "loss": 4.7655,
      "mean_token_accuracy": 0.24124280214309693,
      "num_tokens": 44001823.0,
      "step": 19190
    },
    {
      "entropy": 5.348013877868652,
      "epoch": 1.8439000960614793,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000466292910085776,
      "loss": 4.8933,
      "mean_token_accuracy": 0.22937179505825042,
      "num_tokens": 44014851.0,
      "step": 19195
    },
    {
      "entropy": 5.29071307182312,
      "epoch": 1.8443804034582132,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046627485843592854,
      "loss": 4.8913,
      "mean_token_accuracy": 0.22826795876026154,
      "num_tokens": 44025191.0,
      "step": 19200
    },
    {
      "entropy": 5.3518256664276125,
      "epoch": 1.844860710854947,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046625680234524674,
      "loss": 4.8841,
      "mean_token_accuracy": 0.23155723363161088,
      "num_tokens": 44036319.0,
      "step": 19205
    },
    {
      "entropy": 5.401404571533203,
      "epoch": 1.845341018251681,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046623874181414993,
      "loss": 4.8853,
      "mean_token_accuracy": 0.23193420171737672,
      "num_tokens": 44048618.0,
      "step": 19210
    },
    {
      "entropy": 5.338677787780762,
      "epoch": 1.8458213256484148,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004662206768430574,
      "loss": 4.8371,
      "mean_token_accuracy": 0.2326425850391388,
      "num_tokens": 44060754.0,
      "step": 19215
    },
    {
      "entropy": 5.274700403213501,
      "epoch": 1.846301633045149,
      "grad_norm": 1.125,
      "learning_rate": 0.0004662026074323886,
      "loss": 4.7786,
      "mean_token_accuracy": 0.23874835371971131,
      "num_tokens": 44071239.0,
      "step": 19220
    },
    {
      "entropy": 5.375203561782837,
      "epoch": 1.8467819404418828,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046618453358256303,
      "loss": 4.8695,
      "mean_token_accuracy": 0.23006610125303267,
      "num_tokens": 44082635.0,
      "step": 19225
    },
    {
      "entropy": 5.362653684616089,
      "epoch": 1.8472622478386167,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046616645529400026,
      "loss": 4.7941,
      "mean_token_accuracy": 0.2348039299249649,
      "num_tokens": 44093680.0,
      "step": 19230
    },
    {
      "entropy": 5.30058331489563,
      "epoch": 1.8477425552353506,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004661483725671201,
      "loss": 4.8124,
      "mean_token_accuracy": 0.2354785159230232,
      "num_tokens": 44104490.0,
      "step": 19235
    },
    {
      "entropy": 5.210348272323609,
      "epoch": 1.8482228626320847,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046613028540234226,
      "loss": 4.752,
      "mean_token_accuracy": 0.23877576142549514,
      "num_tokens": 44116961.0,
      "step": 19240
    },
    {
      "entropy": 5.291398668289185,
      "epoch": 1.8487031700288186,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004661121938000867,
      "loss": 4.7807,
      "mean_token_accuracy": 0.23881955444812775,
      "num_tokens": 44127264.0,
      "step": 19245
    },
    {
      "entropy": 5.301933288574219,
      "epoch": 1.8491834774255524,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004660940977607736,
      "loss": 4.7618,
      "mean_token_accuracy": 0.2439037889242172,
      "num_tokens": 44139374.0,
      "step": 19250
    },
    {
      "entropy": 5.212233161926269,
      "epoch": 1.8496637848222863,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046607599728482285,
      "loss": 4.7185,
      "mean_token_accuracy": 0.23728734403848648,
      "num_tokens": 44150416.0,
      "step": 19255
    },
    {
      "entropy": 5.34790072441101,
      "epoch": 1.8501440922190202,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046605789237265496,
      "loss": 4.8514,
      "mean_token_accuracy": 0.2409219115972519,
      "num_tokens": 44162094.0,
      "step": 19260
    },
    {
      "entropy": 5.344606733322143,
      "epoch": 1.850624399615754,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046603978302469,
      "loss": 4.8635,
      "mean_token_accuracy": 0.23740449249744416,
      "num_tokens": 44174526.0,
      "step": 19265
    },
    {
      "entropy": 5.303201627731323,
      "epoch": 1.851104707012488,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004660216692413486,
      "loss": 4.7956,
      "mean_token_accuracy": 0.23825736343860626,
      "num_tokens": 44185961.0,
      "step": 19270
    },
    {
      "entropy": 5.255365514755249,
      "epoch": 1.8515850144092219,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004660035510230513,
      "loss": 4.7918,
      "mean_token_accuracy": 0.23503799736499786,
      "num_tokens": 44198431.0,
      "step": 19275
    },
    {
      "entropy": 5.313368940353394,
      "epoch": 1.8520653218059557,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004659854283702186,
      "loss": 4.84,
      "mean_token_accuracy": 0.24198198467493057,
      "num_tokens": 44210518.0,
      "step": 19280
    },
    {
      "entropy": 5.204764318466187,
      "epoch": 1.8525456292026896,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004659673012832715,
      "loss": 4.7052,
      "mean_token_accuracy": 0.24190931767225266,
      "num_tokens": 44222300.0,
      "step": 19285
    },
    {
      "entropy": 5.341884279251099,
      "epoch": 1.8530259365994235,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004659491697626306,
      "loss": 4.8514,
      "mean_token_accuracy": 0.23240304589271546,
      "num_tokens": 44233142.0,
      "step": 19290
    },
    {
      "entropy": 5.29928412437439,
      "epoch": 1.8535062439961574,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046593103380871705,
      "loss": 4.8635,
      "mean_token_accuracy": 0.23200464248657227,
      "num_tokens": 44246507.0,
      "step": 19295
    },
    {
      "entropy": 5.363926553726197,
      "epoch": 1.8539865513928915,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046591289342195184,
      "loss": 4.8496,
      "mean_token_accuracy": 0.22694653421640396,
      "num_tokens": 44257451.0,
      "step": 19300
    },
    {
      "entropy": 5.2711082935333256,
      "epoch": 1.8544668587896254,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004658947486027562,
      "loss": 4.7557,
      "mean_token_accuracy": 0.24465905725955964,
      "num_tokens": 44269157.0,
      "step": 19305
    },
    {
      "entropy": 5.267344999313354,
      "epoch": 1.8549471661863592,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046587659935155124,
      "loss": 4.7664,
      "mean_token_accuracy": 0.23648781478404998,
      "num_tokens": 44279890.0,
      "step": 19310
    },
    {
      "entropy": 5.240487337112427,
      "epoch": 1.8554274735830933,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046585844566875845,
      "loss": 4.7097,
      "mean_token_accuracy": 0.24358926564455033,
      "num_tokens": 44291457.0,
      "step": 19315
    },
    {
      "entropy": 5.295566368103027,
      "epoch": 1.8559077809798272,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004658402875547993,
      "loss": 4.8344,
      "mean_token_accuracy": 0.23542150408029555,
      "num_tokens": 44301692.0,
      "step": 19320
    },
    {
      "entropy": 5.32517032623291,
      "epoch": 1.856388088376561,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046582212501009533,
      "loss": 4.8424,
      "mean_token_accuracy": 0.23316144496202468,
      "num_tokens": 44313540.0,
      "step": 19325
    },
    {
      "entropy": 5.319202280044555,
      "epoch": 1.856868395773295,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046580395803506825,
      "loss": 4.921,
      "mean_token_accuracy": 0.23138953596353531,
      "num_tokens": 44324970.0,
      "step": 19330
    },
    {
      "entropy": 5.367196083068848,
      "epoch": 1.8573487031700289,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004657857866301399,
      "loss": 4.8321,
      "mean_token_accuracy": 0.23707957267761232,
      "num_tokens": 44336521.0,
      "step": 19335
    },
    {
      "entropy": 5.234256744384766,
      "epoch": 1.8578290105667628,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046576761079573204,
      "loss": 4.7226,
      "mean_token_accuracy": 0.24561231583356857,
      "num_tokens": 44347128.0,
      "step": 19340
    },
    {
      "entropy": 5.33325686454773,
      "epoch": 1.8583093179634966,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004657494305322667,
      "loss": 4.8933,
      "mean_token_accuracy": 0.2317257344722748,
      "num_tokens": 44358503.0,
      "step": 19345
    },
    {
      "entropy": 5.3272277355194095,
      "epoch": 1.8587896253602305,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004657312458401661,
      "loss": 4.8595,
      "mean_token_accuracy": 0.23870461881160737,
      "num_tokens": 44370849.0,
      "step": 19350
    },
    {
      "entropy": 5.3319018363952635,
      "epoch": 1.8592699327569644,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004657130567198522,
      "loss": 4.7581,
      "mean_token_accuracy": 0.23712517619132994,
      "num_tokens": 44383806.0,
      "step": 19355
    },
    {
      "entropy": 5.1844127655029295,
      "epoch": 1.8597502401536983,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046569486317174746,
      "loss": 4.6157,
      "mean_token_accuracy": 0.24969440549612046,
      "num_tokens": 44393937.0,
      "step": 19360
    },
    {
      "entropy": 5.262852001190185,
      "epoch": 1.8602305475504322,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004656766651962742,
      "loss": 4.8501,
      "mean_token_accuracy": 0.23789609968662262,
      "num_tokens": 44404967.0,
      "step": 19365
    },
    {
      "entropy": 5.285940027236938,
      "epoch": 1.860710854947166,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004656584627938551,
      "loss": 4.8409,
      "mean_token_accuracy": 0.23592022210359573,
      "num_tokens": 44416421.0,
      "step": 19370
    },
    {
      "entropy": 5.272666645050049,
      "epoch": 1.8611911623439001,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046564025596491254,
      "loss": 4.8131,
      "mean_token_accuracy": 0.2381610706448555,
      "num_tokens": 44427145.0,
      "step": 19375
    },
    {
      "entropy": 5.305710124969482,
      "epoch": 1.861671469740634,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004656220447098693,
      "loss": 4.7704,
      "mean_token_accuracy": 0.2416410118341446,
      "num_tokens": 44438456.0,
      "step": 19380
    },
    {
      "entropy": 5.272903680801392,
      "epoch": 1.862151777137368,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004656038290291483,
      "loss": 4.7316,
      "mean_token_accuracy": 0.23634986132383345,
      "num_tokens": 44450366.0,
      "step": 19385
    },
    {
      "entropy": 5.250116300582886,
      "epoch": 1.8626320845341018,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004655856089231723,
      "loss": 4.6999,
      "mean_token_accuracy": 0.24142504632472991,
      "num_tokens": 44461246.0,
      "step": 19390
    },
    {
      "entropy": 5.3132905006408695,
      "epoch": 1.8631123919308359,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004655673843923644,
      "loss": 4.8147,
      "mean_token_accuracy": 0.23699672967195512,
      "num_tokens": 44473099.0,
      "step": 19395
    },
    {
      "entropy": 5.322422361373901,
      "epoch": 1.8635926993275698,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004655491554371477,
      "loss": 4.8268,
      "mean_token_accuracy": 0.2349246025085449,
      "num_tokens": 44484516.0,
      "step": 19400
    },
    {
      "entropy": 5.21822304725647,
      "epoch": 1.8640730067243036,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046553092205794543,
      "loss": 4.7126,
      "mean_token_accuracy": 0.24844637364149094,
      "num_tokens": 44495500.0,
      "step": 19405
    },
    {
      "entropy": 5.22309832572937,
      "epoch": 1.8645533141210375,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046551268425518096,
      "loss": 4.7674,
      "mean_token_accuracy": 0.2338672995567322,
      "num_tokens": 44506346.0,
      "step": 19410
    },
    {
      "entropy": 5.357246208190918,
      "epoch": 1.8650336215177714,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004654944420292776,
      "loss": 4.8866,
      "mean_token_accuracy": 0.23477094918489455,
      "num_tokens": 44517234.0,
      "step": 19415
    },
    {
      "entropy": 5.283240556716919,
      "epoch": 1.8655139289145053,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000465476195380659,
      "loss": 4.7122,
      "mean_token_accuracy": 0.24927014261484146,
      "num_tokens": 44528543.0,
      "step": 19420
    },
    {
      "entropy": 5.337205171585083,
      "epoch": 1.8659942363112392,
      "grad_norm": 1.0,
      "learning_rate": 0.0004654579443097487,
      "loss": 4.8788,
      "mean_token_accuracy": 0.23046518862247467,
      "num_tokens": 44539385.0,
      "step": 19425
    },
    {
      "entropy": 5.206415367126465,
      "epoch": 1.866474543707973,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004654396888169705,
      "loss": 4.7097,
      "mean_token_accuracy": 0.23986048698425294,
      "num_tokens": 44550013.0,
      "step": 19430
    },
    {
      "entropy": 5.314290857315063,
      "epoch": 1.866954851104707,
      "grad_norm": 1.0,
      "learning_rate": 0.00046542142890274816,
      "loss": 4.8316,
      "mean_token_accuracy": 0.23400411307811736,
      "num_tokens": 44561926.0,
      "step": 19435
    },
    {
      "entropy": 5.22550220489502,
      "epoch": 1.8674351585014408,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004654031645675057,
      "loss": 4.7792,
      "mean_token_accuracy": 0.23835160583257675,
      "num_tokens": 44574589.0,
      "step": 19440
    },
    {
      "entropy": 5.330621576309204,
      "epoch": 1.8679154658981747,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004653848958116672,
      "loss": 4.8285,
      "mean_token_accuracy": 0.238548943400383,
      "num_tokens": 44586531.0,
      "step": 19445
    },
    {
      "entropy": 5.349969053268433,
      "epoch": 1.8683957732949086,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046536662263565667,
      "loss": 4.7965,
      "mean_token_accuracy": 0.24175089597702026,
      "num_tokens": 44597432.0,
      "step": 19450
    },
    {
      "entropy": 5.342046403884888,
      "epoch": 1.8688760806916427,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004653483450398985,
      "loss": 4.8661,
      "mean_token_accuracy": 0.22933080047369003,
      "num_tokens": 44609264.0,
      "step": 19455
    },
    {
      "entropy": 5.211973810195923,
      "epoch": 1.8693563880883766,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046533006302481694,
      "loss": 4.716,
      "mean_token_accuracy": 0.2466918021440506,
      "num_tokens": 44620410.0,
      "step": 19460
    },
    {
      "entropy": 5.2886152267456055,
      "epoch": 1.8698366954851104,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004653117765908365,
      "loss": 4.8549,
      "mean_token_accuracy": 0.23450296223163605,
      "num_tokens": 44633123.0,
      "step": 19465
    },
    {
      "entropy": 5.3473046779632565,
      "epoch": 1.8703170028818443,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004652934857383816,
      "loss": 4.8484,
      "mean_token_accuracy": 0.23110186457633972,
      "num_tokens": 44644693.0,
      "step": 19470
    },
    {
      "entropy": 5.286482620239258,
      "epoch": 1.8707973102785784,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004652751904678772,
      "loss": 4.7669,
      "mean_token_accuracy": 0.24703191071748734,
      "num_tokens": 44656669.0,
      "step": 19475
    },
    {
      "entropy": 5.277759027481079,
      "epoch": 1.8712776176753123,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046525689077974775,
      "loss": 4.8043,
      "mean_token_accuracy": 0.24176431447267532,
      "num_tokens": 44667938.0,
      "step": 19480
    },
    {
      "entropy": 5.319389200210571,
      "epoch": 1.8717579250720462,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046523858667441834,
      "loss": 4.8461,
      "mean_token_accuracy": 0.23079919070005417,
      "num_tokens": 44678967.0,
      "step": 19485
    },
    {
      "entropy": 5.280212259292602,
      "epoch": 1.87223823246878,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004652202781523138,
      "loss": 4.7044,
      "mean_token_accuracy": 0.24750685542821885,
      "num_tokens": 44689583.0,
      "step": 19490
    },
    {
      "entropy": 5.246271657943725,
      "epoch": 1.872718539865514,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004652019652138592,
      "loss": 4.7421,
      "mean_token_accuracy": 0.24525039345026017,
      "num_tokens": 44700942.0,
      "step": 19495
    },
    {
      "entropy": 5.266356563568115,
      "epoch": 1.8731988472622478,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004651836478594798,
      "loss": 4.7594,
      "mean_token_accuracy": 0.23748984187841415,
      "num_tokens": 44712323.0,
      "step": 19500
    },
    {
      "entropy": 5.350511932373047,
      "epoch": 1.8736791546589817,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004651653260896008,
      "loss": 4.9384,
      "mean_token_accuracy": 0.22690550833940507,
      "num_tokens": 44724033.0,
      "step": 19505
    },
    {
      "entropy": 5.339764642715454,
      "epoch": 1.8741594620557156,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046514699990464763,
      "loss": 4.8767,
      "mean_token_accuracy": 0.23320403844118118,
      "num_tokens": 44735388.0,
      "step": 19510
    },
    {
      "entropy": 5.27905592918396,
      "epoch": 1.8746397694524495,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004651286693050458,
      "loss": 4.7829,
      "mean_token_accuracy": 0.24034113585948944,
      "num_tokens": 44748165.0,
      "step": 19515
    },
    {
      "entropy": 5.313099384307861,
      "epoch": 1.8751200768491834,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004651103342912207,
      "loss": 4.806,
      "mean_token_accuracy": 0.24052114188671112,
      "num_tokens": 44758500.0,
      "step": 19520
    },
    {
      "entropy": 5.374856472015381,
      "epoch": 1.8756003842459172,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046509199486359824,
      "loss": 4.9247,
      "mean_token_accuracy": 0.22504837214946746,
      "num_tokens": 44770347.0,
      "step": 19525
    },
    {
      "entropy": 5.3095354557037355,
      "epoch": 1.8760806916426513,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046507365102260403,
      "loss": 4.7919,
      "mean_token_accuracy": 0.23547566831111907,
      "num_tokens": 44782365.0,
      "step": 19530
    },
    {
      "entropy": 5.273911809921264,
      "epoch": 1.8765609990393852,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046505530276866417,
      "loss": 4.7363,
      "mean_token_accuracy": 0.24241956025362016,
      "num_tokens": 44793128.0,
      "step": 19535
    },
    {
      "entropy": 5.24897608757019,
      "epoch": 1.877041306436119,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046503695010220443,
      "loss": 4.7844,
      "mean_token_accuracy": 0.24308834969997406,
      "num_tokens": 44804236.0,
      "step": 19540
    },
    {
      "entropy": 5.237995624542236,
      "epoch": 1.877521613832853,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000465018593023651,
      "loss": 4.7498,
      "mean_token_accuracy": 0.23783439993858338,
      "num_tokens": 44815983.0,
      "step": 19545
    },
    {
      "entropy": 5.330631685256958,
      "epoch": 1.878001921229587,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004650002315334302,
      "loss": 4.9242,
      "mean_token_accuracy": 0.22773220837116243,
      "num_tokens": 44828267.0,
      "step": 19550
    },
    {
      "entropy": 5.278448534011841,
      "epoch": 1.878482228626321,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000464981865631968,
      "loss": 4.6956,
      "mean_token_accuracy": 0.23960509598255159,
      "num_tokens": 44838668.0,
      "step": 19555
    },
    {
      "entropy": 5.269239234924316,
      "epoch": 1.8789625360230549,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004649634953196912,
      "loss": 4.8346,
      "mean_token_accuracy": 0.23696491718292237,
      "num_tokens": 44849492.0,
      "step": 19560
    },
    {
      "entropy": 5.393634366989136,
      "epoch": 1.8794428434197887,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046494512059702605,
      "loss": 4.9589,
      "mean_token_accuracy": 0.2248767286539078,
      "num_tokens": 44861388.0,
      "step": 19565
    },
    {
      "entropy": 5.285600328445435,
      "epoch": 1.8799231508165226,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004649267414643992,
      "loss": 4.8153,
      "mean_token_accuracy": 0.23808353394269943,
      "num_tokens": 44873353.0,
      "step": 19570
    },
    {
      "entropy": 5.306372451782226,
      "epoch": 1.8804034582132565,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004649083579222374,
      "loss": 4.8501,
      "mean_token_accuracy": 0.23440296500921248,
      "num_tokens": 44885020.0,
      "step": 19575
    },
    {
      "entropy": 5.395475053787232,
      "epoch": 1.8808837656099904,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046488996997096744,
      "loss": 4.9044,
      "mean_token_accuracy": 0.23269910961389542,
      "num_tokens": 44896743.0,
      "step": 19580
    },
    {
      "entropy": 5.282521438598633,
      "epoch": 1.8813640730067243,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004648715776110162,
      "loss": 4.7846,
      "mean_token_accuracy": 0.2366724580526352,
      "num_tokens": 44909138.0,
      "step": 19585
    },
    {
      "entropy": 5.254808473587036,
      "epoch": 1.8818443804034581,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004648531808428108,
      "loss": 4.7956,
      "mean_token_accuracy": 0.2376634731888771,
      "num_tokens": 44921139.0,
      "step": 19590
    },
    {
      "entropy": 5.230926465988159,
      "epoch": 1.882324687800192,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004648347796667782,
      "loss": 4.682,
      "mean_token_accuracy": 0.24234849363565444,
      "num_tokens": 44933282.0,
      "step": 19595
    },
    {
      "entropy": 5.317355298995972,
      "epoch": 1.882804995196926,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004648163740833458,
      "loss": 4.8396,
      "mean_token_accuracy": 0.22794306874275208,
      "num_tokens": 44946350.0,
      "step": 19600
    },
    {
      "entropy": 5.280272626876831,
      "epoch": 1.8832853025936598,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046479796409294076,
      "loss": 4.8171,
      "mean_token_accuracy": 0.23317396640777588,
      "num_tokens": 44958348.0,
      "step": 19605
    },
    {
      "entropy": 5.254821109771728,
      "epoch": 1.8837656099903939,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004647795496959907,
      "loss": 4.7379,
      "mean_token_accuracy": 0.24816108793020247,
      "num_tokens": 44970206.0,
      "step": 19610
    },
    {
      "entropy": 5.355894088745117,
      "epoch": 1.8842459173871278,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046476113089292286,
      "loss": 4.8101,
      "mean_token_accuracy": 0.2425827369093895,
      "num_tokens": 44980814.0,
      "step": 19615
    },
    {
      "entropy": 5.274718523025513,
      "epoch": 1.8847262247838616,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004647427076841651,
      "loss": 4.7724,
      "mean_token_accuracy": 0.23829679787158967,
      "num_tokens": 44991148.0,
      "step": 19620
    },
    {
      "entropy": 5.217650413513184,
      "epoch": 1.8852065321805955,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046472428007014515,
      "loss": 4.7298,
      "mean_token_accuracy": 0.25018097311258314,
      "num_tokens": 45003070.0,
      "step": 19625
    },
    {
      "entropy": 5.257108545303344,
      "epoch": 1.8856868395773296,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004647058480512907,
      "loss": 4.7737,
      "mean_token_accuracy": 0.23411166965961455,
      "num_tokens": 45015760.0,
      "step": 19630
    },
    {
      "entropy": 5.335402393341065,
      "epoch": 1.8861671469740635,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046468741162802987,
      "loss": 4.8257,
      "mean_token_accuracy": 0.23533037751913072,
      "num_tokens": 45026297.0,
      "step": 19635
    },
    {
      "entropy": 5.250051116943359,
      "epoch": 1.8866474543707974,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004646689708007905,
      "loss": 4.7242,
      "mean_token_accuracy": 0.24188960790634156,
      "num_tokens": 45037814.0,
      "step": 19640
    },
    {
      "entropy": 5.21107759475708,
      "epoch": 1.8871277617675313,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046465052557000087,
      "loss": 4.7657,
      "mean_token_accuracy": 0.24286952316761018,
      "num_tokens": 45050632.0,
      "step": 19645
    },
    {
      "entropy": 5.319071578979492,
      "epoch": 1.8876080691642652,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046463207593608916,
      "loss": 4.8097,
      "mean_token_accuracy": 0.23561593145132065,
      "num_tokens": 45061866.0,
      "step": 19650
    },
    {
      "entropy": 5.323861074447632,
      "epoch": 1.888088376560999,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004646136218994837,
      "loss": 4.8747,
      "mean_token_accuracy": 0.22859181463718414,
      "num_tokens": 45074550.0,
      "step": 19655
    },
    {
      "entropy": 5.369389438629151,
      "epoch": 1.888568683957733,
      "grad_norm": 1.125,
      "learning_rate": 0.00046459516346061304,
      "loss": 4.8656,
      "mean_token_accuracy": 0.2316967099905014,
      "num_tokens": 45085710.0,
      "step": 19660
    },
    {
      "entropy": 5.2916826725006105,
      "epoch": 1.8890489913544668,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046457670061990564,
      "loss": 4.8168,
      "mean_token_accuracy": 0.23739214539527892,
      "num_tokens": 45097917.0,
      "step": 19665
    },
    {
      "entropy": 5.242239761352539,
      "epoch": 1.8895292987512007,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046455823337779024,
      "loss": 4.7587,
      "mean_token_accuracy": 0.24410243779420854,
      "num_tokens": 45108455.0,
      "step": 19670
    },
    {
      "entropy": 5.374004697799682,
      "epoch": 1.8900096061479346,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004645397617346954,
      "loss": 4.9605,
      "mean_token_accuracy": 0.2300271898508072,
      "num_tokens": 45120313.0,
      "step": 19675
    },
    {
      "entropy": 5.30882158279419,
      "epoch": 1.8904899135446684,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004645212856910502,
      "loss": 4.7604,
      "mean_token_accuracy": 0.24245844781398773,
      "num_tokens": 45131673.0,
      "step": 19680
    },
    {
      "entropy": 5.259243583679199,
      "epoch": 1.8909702209414025,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004645028052472835,
      "loss": 4.7894,
      "mean_token_accuracy": 0.23718342036008835,
      "num_tokens": 45143589.0,
      "step": 19685
    },
    {
      "entropy": 5.246568489074707,
      "epoch": 1.8914505283381364,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046448432040382444,
      "loss": 4.7606,
      "mean_token_accuracy": 0.23663748800754547,
      "num_tokens": 45155904.0,
      "step": 19690
    },
    {
      "entropy": 5.294768285751343,
      "epoch": 1.8919308357348703,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000464465831161102,
      "loss": 4.7689,
      "mean_token_accuracy": 0.24372636675834655,
      "num_tokens": 45166802.0,
      "step": 19695
    },
    {
      "entropy": 5.378216934204102,
      "epoch": 1.8924111431316042,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004644473375195456,
      "loss": 4.857,
      "mean_token_accuracy": 0.23145218342542648,
      "num_tokens": 45178986.0,
      "step": 19700
    },
    {
      "entropy": 5.290633010864258,
      "epoch": 1.8928914505283383,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046442883947958466,
      "loss": 4.7964,
      "mean_token_accuracy": 0.24802263230085372,
      "num_tokens": 45190814.0,
      "step": 19705
    },
    {
      "entropy": 5.21997857093811,
      "epoch": 1.8933717579250722,
      "grad_norm": 1.0,
      "learning_rate": 0.00046441033704164845,
      "loss": 4.7407,
      "mean_token_accuracy": 0.24865712523460387,
      "num_tokens": 45202553.0,
      "step": 19710
    },
    {
      "entropy": 5.326942396163941,
      "epoch": 1.893852065321806,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004643918302061666,
      "loss": 4.8876,
      "mean_token_accuracy": 0.2304786905646324,
      "num_tokens": 45214482.0,
      "step": 19715
    },
    {
      "entropy": 5.293378734588623,
      "epoch": 1.89433237271854,
      "grad_norm": 1.0,
      "learning_rate": 0.0004643733189735689,
      "loss": 4.7906,
      "mean_token_accuracy": 0.24117643535137176,
      "num_tokens": 45226675.0,
      "step": 19720
    },
    {
      "entropy": 5.283992910385132,
      "epoch": 1.8948126801152738,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000464354803344285,
      "loss": 4.8165,
      "mean_token_accuracy": 0.2365666076540947,
      "num_tokens": 45238995.0,
      "step": 19725
    },
    {
      "entropy": 5.272534513473511,
      "epoch": 1.8952929875120077,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046433628331874496,
      "loss": 4.7794,
      "mean_token_accuracy": 0.23809199333190917,
      "num_tokens": 45250696.0,
      "step": 19730
    },
    {
      "entropy": 5.292238426208496,
      "epoch": 1.8957732949087416,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004643177588973785,
      "loss": 4.7062,
      "mean_token_accuracy": 0.2421130582690239,
      "num_tokens": 45262434.0,
      "step": 19735
    },
    {
      "entropy": 5.300869941711426,
      "epoch": 1.8962536023054755,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004642992300806159,
      "loss": 4.8763,
      "mean_token_accuracy": 0.233504419028759,
      "num_tokens": 45273267.0,
      "step": 19740
    },
    {
      "entropy": 5.272695112228393,
      "epoch": 1.8967339097022093,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004642806968688873,
      "loss": 4.7895,
      "mean_token_accuracy": 0.2388200730085373,
      "num_tokens": 45284601.0,
      "step": 19745
    },
    {
      "entropy": 5.231877851486206,
      "epoch": 1.8972142170989432,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046426215926262295,
      "loss": 4.7851,
      "mean_token_accuracy": 0.24400290846824646,
      "num_tokens": 45296504.0,
      "step": 19750
    },
    {
      "entropy": 5.253055810928345,
      "epoch": 1.897694524495677,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004642436172622532,
      "loss": 4.7305,
      "mean_token_accuracy": 0.24054066985845565,
      "num_tokens": 45307232.0,
      "step": 19755
    },
    {
      "entropy": 5.2410712242126465,
      "epoch": 1.898174831892411,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004642250708682086,
      "loss": 4.6903,
      "mean_token_accuracy": 0.24905335009098054,
      "num_tokens": 45317768.0,
      "step": 19760
    },
    {
      "entropy": 5.3370246410369875,
      "epoch": 1.898655139289145,
      "grad_norm": 3.734375,
      "learning_rate": 0.00046420652008091984,
      "loss": 4.8322,
      "mean_token_accuracy": 0.2377060979604721,
      "num_tokens": 45329172.0,
      "step": 19765
    },
    {
      "entropy": 5.3679407119750975,
      "epoch": 1.899135446685879,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004641879649008174,
      "loss": 4.9105,
      "mean_token_accuracy": 0.2285564973950386,
      "num_tokens": 45339840.0,
      "step": 19770
    },
    {
      "entropy": 5.313313293457031,
      "epoch": 1.8996157540826129,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004641694053283323,
      "loss": 4.8687,
      "mean_token_accuracy": 0.23572106808423995,
      "num_tokens": 45351532.0,
      "step": 19775
    },
    {
      "entropy": 5.296630859375,
      "epoch": 1.9000960614793467,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046415084136389525,
      "loss": 4.901,
      "mean_token_accuracy": 0.22739094495773315,
      "num_tokens": 45362872.0,
      "step": 19780
    },
    {
      "entropy": 5.49705138206482,
      "epoch": 1.9005763688760808,
      "grad_norm": 1.125,
      "learning_rate": 0.0004641322730079374,
      "loss": 4.9408,
      "mean_token_accuracy": 0.22601778507232667,
      "num_tokens": 45374877.0,
      "step": 19785
    },
    {
      "entropy": 5.34177827835083,
      "epoch": 1.9010566762728147,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004641137002608897,
      "loss": 4.8463,
      "mean_token_accuracy": 0.23458772897720337,
      "num_tokens": 45387136.0,
      "step": 19790
    },
    {
      "entropy": 5.222240781784057,
      "epoch": 1.9015369836695486,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046409512312318345,
      "loss": 4.7777,
      "mean_token_accuracy": 0.24144722819328307,
      "num_tokens": 45398310.0,
      "step": 19795
    },
    {
      "entropy": 5.261081409454346,
      "epoch": 1.9020172910662825,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046407654159524994,
      "loss": 4.7961,
      "mean_token_accuracy": 0.22935667037963867,
      "num_tokens": 45410456.0,
      "step": 19800
    },
    {
      "entropy": 5.301387596130371,
      "epoch": 1.9024975984630164,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046405795567752055,
      "loss": 4.7214,
      "mean_token_accuracy": 0.23707799315452577,
      "num_tokens": 45423292.0,
      "step": 19805
    },
    {
      "entropy": 5.247507381439209,
      "epoch": 1.9029779058597502,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046403936537042686,
      "loss": 4.794,
      "mean_token_accuracy": 0.2480373278260231,
      "num_tokens": 45434699.0,
      "step": 19810
    },
    {
      "entropy": 5.23076057434082,
      "epoch": 1.9034582132564841,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046402077067440043,
      "loss": 4.7821,
      "mean_token_accuracy": 0.23256126642227173,
      "num_tokens": 45446849.0,
      "step": 19815
    },
    {
      "entropy": 5.3062803745269775,
      "epoch": 1.903938520653218,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046400217158987293,
      "loss": 4.8402,
      "mean_token_accuracy": 0.23410187810659408,
      "num_tokens": 45458070.0,
      "step": 19820
    },
    {
      "entropy": 5.337236928939819,
      "epoch": 1.9044188280499519,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046398356811727626,
      "loss": 4.8445,
      "mean_token_accuracy": 0.23652203381061554,
      "num_tokens": 45470829.0,
      "step": 19825
    },
    {
      "entropy": 5.25196852684021,
      "epoch": 1.9048991354466858,
      "grad_norm": 1.0,
      "learning_rate": 0.0004639649602570423,
      "loss": 4.7365,
      "mean_token_accuracy": 0.24370529055595397,
      "num_tokens": 45482354.0,
      "step": 19830
    },
    {
      "entropy": 5.25517840385437,
      "epoch": 1.9053794428434196,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046394634800960314,
      "loss": 4.7515,
      "mean_token_accuracy": 0.23573573380708696,
      "num_tokens": 45493819.0,
      "step": 19835
    },
    {
      "entropy": 5.3322751998901365,
      "epoch": 1.9058597502401537,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046392773137539074,
      "loss": 4.8605,
      "mean_token_accuracy": 0.23356813490390776,
      "num_tokens": 45504877.0,
      "step": 19840
    },
    {
      "entropy": 5.280382919311523,
      "epoch": 1.9063400576368876,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046390911035483744,
      "loss": 4.8035,
      "mean_token_accuracy": 0.23491956740617753,
      "num_tokens": 45516160.0,
      "step": 19845
    },
    {
      "entropy": 5.257534837722778,
      "epoch": 1.9068203650336215,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004638904849483756,
      "loss": 4.7387,
      "mean_token_accuracy": 0.24389853030443193,
      "num_tokens": 45529028.0,
      "step": 19850
    },
    {
      "entropy": 5.268261671066284,
      "epoch": 1.9073006724303554,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046387185515643756,
      "loss": 4.7879,
      "mean_token_accuracy": 0.2397521808743477,
      "num_tokens": 45541507.0,
      "step": 19855
    },
    {
      "entropy": 5.371911287307739,
      "epoch": 1.9077809798270895,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004638532209794559,
      "loss": 4.9108,
      "mean_token_accuracy": 0.22509036660194398,
      "num_tokens": 45553379.0,
      "step": 19860
    },
    {
      "entropy": 5.264101266860962,
      "epoch": 1.9082612872238234,
      "grad_norm": 1.0,
      "learning_rate": 0.0004638345824178631,
      "loss": 4.7316,
      "mean_token_accuracy": 0.24028994292020797,
      "num_tokens": 45564962.0,
      "step": 19865
    },
    {
      "entropy": 5.345943880081177,
      "epoch": 1.9087415946205573,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046381593947209215,
      "loss": 4.8916,
      "mean_token_accuracy": 0.2305974096059799,
      "num_tokens": 45575270.0,
      "step": 19870
    },
    {
      "entropy": 5.2859108448028564,
      "epoch": 1.9092219020172911,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004637972921425757,
      "loss": 4.7475,
      "mean_token_accuracy": 0.24355046600103378,
      "num_tokens": 45587446.0,
      "step": 19875
    },
    {
      "entropy": 5.380211400985718,
      "epoch": 1.909702209414025,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046377864042974675,
      "loss": 4.9243,
      "mean_token_accuracy": 0.2380807340145111,
      "num_tokens": 45599598.0,
      "step": 19880
    },
    {
      "entropy": 5.298010206222534,
      "epoch": 1.910182516810759,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004637599843340384,
      "loss": 4.7246,
      "mean_token_accuracy": 0.23891299813985825,
      "num_tokens": 45610862.0,
      "step": 19885
    },
    {
      "entropy": 5.2459314346313475,
      "epoch": 1.9106628242074928,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046374132385588356,
      "loss": 4.7384,
      "mean_token_accuracy": 0.25386798232793806,
      "num_tokens": 45621398.0,
      "step": 19890
    },
    {
      "entropy": 5.26541690826416,
      "epoch": 1.9111431316042267,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046372265899571576,
      "loss": 4.7891,
      "mean_token_accuracy": 0.2350035399198532,
      "num_tokens": 45631139.0,
      "step": 19895
    },
    {
      "entropy": 5.293388223648071,
      "epoch": 1.9116234390009605,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046370398975396817,
      "loss": 4.8596,
      "mean_token_accuracy": 0.23679569214582444,
      "num_tokens": 45642439.0,
      "step": 19900
    },
    {
      "entropy": 5.315039777755738,
      "epoch": 1.9121037463976944,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004636853161310743,
      "loss": 4.7196,
      "mean_token_accuracy": 0.24566220343112946,
      "num_tokens": 45653701.0,
      "step": 19905
    },
    {
      "entropy": 5.261599826812744,
      "epoch": 1.9125840537944283,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046366663812746764,
      "loss": 4.773,
      "mean_token_accuracy": 0.2441158652305603,
      "num_tokens": 45663870.0,
      "step": 19910
    },
    {
      "entropy": 5.2807472229003904,
      "epoch": 1.9130643611911622,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004636479557435818,
      "loss": 4.7927,
      "mean_token_accuracy": 0.2333766281604767,
      "num_tokens": 45674597.0,
      "step": 19915
    },
    {
      "entropy": 5.336585283279419,
      "epoch": 1.9135446685878963,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046362926897985067,
      "loss": 4.852,
      "mean_token_accuracy": 0.23466378450393677,
      "num_tokens": 45685871.0,
      "step": 19920
    },
    {
      "entropy": 5.262584209442139,
      "epoch": 1.9140249759846302,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000463610577836708,
      "loss": 4.8556,
      "mean_token_accuracy": 0.23962184339761733,
      "num_tokens": 45696742.0,
      "step": 19925
    },
    {
      "entropy": 5.3265135288238525,
      "epoch": 1.914505283381364,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046359188231458783,
      "loss": 4.8362,
      "mean_token_accuracy": 0.24112892299890518,
      "num_tokens": 45708398.0,
      "step": 19930
    },
    {
      "entropy": 5.315515804290771,
      "epoch": 1.914985590778098,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046357318241392414,
      "loss": 4.8621,
      "mean_token_accuracy": 0.23687999695539474,
      "num_tokens": 45721525.0,
      "step": 19935
    },
    {
      "entropy": 5.2079651832580565,
      "epoch": 1.915465898174832,
      "grad_norm": 1.078125,
      "learning_rate": 0.000463554478135151,
      "loss": 4.7355,
      "mean_token_accuracy": 0.24251578450202943,
      "num_tokens": 45732089.0,
      "step": 19940
    },
    {
      "entropy": 5.295583009719849,
      "epoch": 1.915946205571566,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004635357694787029,
      "loss": 4.7326,
      "mean_token_accuracy": 0.24486218243837357,
      "num_tokens": 45743243.0,
      "step": 19945
    },
    {
      "entropy": 5.286836910247803,
      "epoch": 1.9164265129682998,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000463517056445014,
      "loss": 4.7879,
      "mean_token_accuracy": 0.2344546765089035,
      "num_tokens": 45754113.0,
      "step": 19950
    },
    {
      "entropy": 5.204514455795288,
      "epoch": 1.9169068203650337,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046349833903451884,
      "loss": 4.7397,
      "mean_token_accuracy": 0.24145962297916412,
      "num_tokens": 45766190.0,
      "step": 19955
    },
    {
      "entropy": 5.37755856513977,
      "epoch": 1.9173871277617676,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046347961724765196,
      "loss": 4.8843,
      "mean_token_accuracy": 0.23093047440052034,
      "num_tokens": 45777736.0,
      "step": 19960
    },
    {
      "entropy": 5.370610237121582,
      "epoch": 1.9178674351585014,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046346089108484806,
      "loss": 4.8386,
      "mean_token_accuracy": 0.23455018699169158,
      "num_tokens": 45788813.0,
      "step": 19965
    },
    {
      "entropy": 5.310533809661865,
      "epoch": 1.9183477425552353,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046344216054654193,
      "loss": 4.8589,
      "mean_token_accuracy": 0.22861503511667253,
      "num_tokens": 45800671.0,
      "step": 19970
    },
    {
      "entropy": 5.28101167678833,
      "epoch": 1.9188280499519692,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046342342563316833,
      "loss": 4.8023,
      "mean_token_accuracy": 0.23479206413030623,
      "num_tokens": 45811587.0,
      "step": 19975
    },
    {
      "entropy": 5.248469686508178,
      "epoch": 1.919308357348703,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00046340468634516223,
      "loss": 4.7044,
      "mean_token_accuracy": 0.24759768694639206,
      "num_tokens": 45823381.0,
      "step": 19980
    },
    {
      "entropy": 5.369346952438354,
      "epoch": 1.919788664745437,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046338594268295884,
      "loss": 4.8204,
      "mean_token_accuracy": 0.23964275866746904,
      "num_tokens": 45833560.0,
      "step": 19985
    },
    {
      "entropy": 5.300217199325561,
      "epoch": 1.9202689721421708,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004633671946469933,
      "loss": 4.7877,
      "mean_token_accuracy": 0.24546139240264891,
      "num_tokens": 45843823.0,
      "step": 19990
    },
    {
      "entropy": 5.266721105575561,
      "epoch": 1.920749279538905,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046334844223770076,
      "loss": 4.8382,
      "mean_token_accuracy": 0.23099299818277358,
      "num_tokens": 45855122.0,
      "step": 19995
    },
    {
      "entropy": 5.280814790725708,
      "epoch": 1.9212295869356388,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046332968545551674,
      "loss": 4.7988,
      "mean_token_accuracy": 0.24577650129795076,
      "num_tokens": 45866638.0,
      "step": 20000
    },
    {
      "entropy": 5.391184091567993,
      "epoch": 1.9217098943323727,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004633109243008765,
      "loss": 4.8864,
      "mean_token_accuracy": 0.22343889623880386,
      "num_tokens": 45878654.0,
      "step": 20005
    },
    {
      "entropy": 5.33652892112732,
      "epoch": 1.9221902017291066,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004632921587742159,
      "loss": 4.7968,
      "mean_token_accuracy": 0.23492608219385147,
      "num_tokens": 45889466.0,
      "step": 20010
    },
    {
      "entropy": 5.298433685302735,
      "epoch": 1.9226705091258407,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046327338887597043,
      "loss": 4.7954,
      "mean_token_accuracy": 0.24324515610933303,
      "num_tokens": 45900750.0,
      "step": 20015
    },
    {
      "entropy": 5.348216247558594,
      "epoch": 1.9231508165225746,
      "grad_norm": 1.0,
      "learning_rate": 0.000463254614606576,
      "loss": 4.8439,
      "mean_token_accuracy": 0.2304997131228447,
      "num_tokens": 45912351.0,
      "step": 20020
    },
    {
      "entropy": 5.25238823890686,
      "epoch": 1.9236311239193085,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004632358359664683,
      "loss": 4.74,
      "mean_token_accuracy": 0.244276662170887,
      "num_tokens": 45923484.0,
      "step": 20025
    },
    {
      "entropy": 5.255502414703369,
      "epoch": 1.9241114313160423,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046321705295608356,
      "loss": 4.792,
      "mean_token_accuracy": 0.24216009229421614,
      "num_tokens": 45935512.0,
      "step": 20030
    },
    {
      "entropy": 5.27532114982605,
      "epoch": 1.9245917387127762,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046319826557585764,
      "loss": 4.7462,
      "mean_token_accuracy": 0.24406860023736954,
      "num_tokens": 45946353.0,
      "step": 20035
    },
    {
      "entropy": 5.30128026008606,
      "epoch": 1.92507204610951,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004631794738262269,
      "loss": 4.7549,
      "mean_token_accuracy": 0.23518649339675904,
      "num_tokens": 45958281.0,
      "step": 20040
    },
    {
      "entropy": 5.259121751785278,
      "epoch": 1.925552353506244,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004631606777076275,
      "loss": 4.7607,
      "mean_token_accuracy": 0.23999468684196473,
      "num_tokens": 45969548.0,
      "step": 20045
    },
    {
      "entropy": 5.287804651260376,
      "epoch": 1.9260326609029779,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046314187722049587,
      "loss": 4.7863,
      "mean_token_accuracy": 0.23125479966402054,
      "num_tokens": 45980300.0,
      "step": 20050
    },
    {
      "entropy": 5.2970648288726805,
      "epoch": 1.9265129682997117,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046312307236526863,
      "loss": 4.7533,
      "mean_token_accuracy": 0.23575175553560257,
      "num_tokens": 45991280.0,
      "step": 20055
    },
    {
      "entropy": 5.352742099761963,
      "epoch": 1.9269932756964456,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046310426314238217,
      "loss": 4.8629,
      "mean_token_accuracy": 0.2357402503490448,
      "num_tokens": 46003074.0,
      "step": 20060
    },
    {
      "entropy": 5.185699415206909,
      "epoch": 1.9274735830931795,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004630854495522733,
      "loss": 4.7361,
      "mean_token_accuracy": 0.24750707745552064,
      "num_tokens": 46014784.0,
      "step": 20065
    },
    {
      "entropy": 5.23555588722229,
      "epoch": 1.9279538904899134,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046306663159537874,
      "loss": 4.7156,
      "mean_token_accuracy": 0.250432924926281,
      "num_tokens": 46025403.0,
      "step": 20070
    },
    {
      "entropy": 5.312037944793701,
      "epoch": 1.9284341978866475,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046304780927213554,
      "loss": 4.8176,
      "mean_token_accuracy": 0.2369946926832199,
      "num_tokens": 46037171.0,
      "step": 20075
    },
    {
      "entropy": 5.365404891967773,
      "epoch": 1.9289145052833814,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046302898258298046,
      "loss": 4.8562,
      "mean_token_accuracy": 0.22835959494113922,
      "num_tokens": 46049219.0,
      "step": 20080
    },
    {
      "entropy": 5.22703104019165,
      "epoch": 1.9293948126801153,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004630101515283509,
      "loss": 4.7432,
      "mean_token_accuracy": 0.2455800086259842,
      "num_tokens": 46061068.0,
      "step": 20085
    },
    {
      "entropy": 5.295748710632324,
      "epoch": 1.9298751200768491,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046299131610868377,
      "loss": 4.7185,
      "mean_token_accuracy": 0.24337782263755797,
      "num_tokens": 46073752.0,
      "step": 20090
    },
    {
      "entropy": 5.332561111450195,
      "epoch": 1.9303554274735832,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004629724763244165,
      "loss": 4.8387,
      "mean_token_accuracy": 0.23857258409261703,
      "num_tokens": 46085908.0,
      "step": 20095
    },
    {
      "entropy": 5.249281024932861,
      "epoch": 1.9308357348703171,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004629536321759866,
      "loss": 4.7562,
      "mean_token_accuracy": 0.24551275074481965,
      "num_tokens": 46097467.0,
      "step": 20100
    },
    {
      "entropy": 5.2716902732849125,
      "epoch": 1.931316042267051,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046293478366383133,
      "loss": 4.8023,
      "mean_token_accuracy": 0.24109130948781968,
      "num_tokens": 46107951.0,
      "step": 20105
    },
    {
      "entropy": 5.331090688705444,
      "epoch": 1.9317963496637849,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004629159307883885,
      "loss": 4.8348,
      "mean_token_accuracy": 0.23130493760108947,
      "num_tokens": 46119862.0,
      "step": 20110
    },
    {
      "entropy": 5.265496397018433,
      "epoch": 1.9322766570605188,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004628970735500958,
      "loss": 4.7543,
      "mean_token_accuracy": 0.23699362128973006,
      "num_tokens": 46130625.0,
      "step": 20115
    },
    {
      "entropy": 5.313340377807617,
      "epoch": 1.9327569644572526,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046287821194939094,
      "loss": 4.7997,
      "mean_token_accuracy": 0.2391948476433754,
      "num_tokens": 46141819.0,
      "step": 20120
    },
    {
      "entropy": 5.366791820526123,
      "epoch": 1.9332372718539865,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046285934598671186,
      "loss": 4.8986,
      "mean_token_accuracy": 0.22765258699655533,
      "num_tokens": 46154474.0,
      "step": 20125
    },
    {
      "entropy": 5.3378785133361815,
      "epoch": 1.9337175792507204,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046284047566249665,
      "loss": 4.8128,
      "mean_token_accuracy": 0.2289728969335556,
      "num_tokens": 46165397.0,
      "step": 20130
    },
    {
      "entropy": 5.2365484714508055,
      "epoch": 1.9341978866474543,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046282160097718336,
      "loss": 4.7581,
      "mean_token_accuracy": 0.24380571842193605,
      "num_tokens": 46177202.0,
      "step": 20135
    },
    {
      "entropy": 5.237990856170654,
      "epoch": 1.9346781940441882,
      "grad_norm": 1.0,
      "learning_rate": 0.0004628027219312102,
      "loss": 4.799,
      "mean_token_accuracy": 0.23872011452913283,
      "num_tokens": 46189879.0,
      "step": 20140
    },
    {
      "entropy": 5.4038971900939945,
      "epoch": 1.935158501440922,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004627838385250155,
      "loss": 4.9015,
      "mean_token_accuracy": 0.2283681035041809,
      "num_tokens": 46201788.0,
      "step": 20145
    },
    {
      "entropy": 5.368332958221435,
      "epoch": 1.9356388088376562,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046276495075903764,
      "loss": 4.8053,
      "mean_token_accuracy": 0.2394896686077118,
      "num_tokens": 46212316.0,
      "step": 20150
    },
    {
      "entropy": 5.257050800323486,
      "epoch": 1.93611911623439,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046274605863371517,
      "loss": 4.7562,
      "mean_token_accuracy": 0.24538294821977616,
      "num_tokens": 46223192.0,
      "step": 20155
    },
    {
      "entropy": 5.3432317733764645,
      "epoch": 1.936599423631124,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004627271621494868,
      "loss": 4.8293,
      "mean_token_accuracy": 0.2308499351143837,
      "num_tokens": 46234404.0,
      "step": 20160
    },
    {
      "entropy": 5.3414098739624025,
      "epoch": 1.9370797310278578,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000462708261306791,
      "loss": 4.7852,
      "mean_token_accuracy": 0.23400003463029861,
      "num_tokens": 46246653.0,
      "step": 20165
    },
    {
      "entropy": 5.3490723133087155,
      "epoch": 1.937560038424592,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004626893561060669,
      "loss": 4.8688,
      "mean_token_accuracy": 0.23196227848529816,
      "num_tokens": 46258145.0,
      "step": 20170
    },
    {
      "entropy": 5.295912790298462,
      "epoch": 1.9380403458213258,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046267044654775324,
      "loss": 4.7604,
      "mean_token_accuracy": 0.2444765478372574,
      "num_tokens": 46269021.0,
      "step": 20175
    },
    {
      "entropy": 5.244674825668335,
      "epoch": 1.9385206532180597,
      "grad_norm": 1.0,
      "learning_rate": 0.000462651532632289,
      "loss": 4.7674,
      "mean_token_accuracy": 0.23751564472913742,
      "num_tokens": 46280729.0,
      "step": 20180
    },
    {
      "entropy": 5.2774210453033445,
      "epoch": 1.9390009606147935,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046263261436011344,
      "loss": 4.832,
      "mean_token_accuracy": 0.2324442148208618,
      "num_tokens": 46291999.0,
      "step": 20185
    },
    {
      "entropy": 5.347610282897949,
      "epoch": 1.9394812680115274,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004626136917316657,
      "loss": 4.8559,
      "mean_token_accuracy": 0.23299823105335235,
      "num_tokens": 46302740.0,
      "step": 20190
    },
    {
      "entropy": 5.293584632873535,
      "epoch": 1.9399615754082613,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046259476474738514,
      "loss": 4.7475,
      "mean_token_accuracy": 0.24390690475702287,
      "num_tokens": 46313700.0,
      "step": 20195
    },
    {
      "entropy": 5.310221576690674,
      "epoch": 1.9404418828049952,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046257583340771123,
      "loss": 4.8312,
      "mean_token_accuracy": 0.23549176156520843,
      "num_tokens": 46324445.0,
      "step": 20200
    },
    {
      "entropy": 5.218066453933716,
      "epoch": 1.940922190201729,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004625568977130835,
      "loss": 4.7725,
      "mean_token_accuracy": 0.23843089789152144,
      "num_tokens": 46335192.0,
      "step": 20205
    },
    {
      "entropy": 5.358770799636841,
      "epoch": 1.941402497598463,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004625379576639414,
      "loss": 4.8444,
      "mean_token_accuracy": 0.2357109397649765,
      "num_tokens": 46346067.0,
      "step": 20210
    },
    {
      "entropy": 5.313088130950928,
      "epoch": 1.9418828049951968,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046251901326072487,
      "loss": 4.7392,
      "mean_token_accuracy": 0.24330868870019912,
      "num_tokens": 46356736.0,
      "step": 20215
    },
    {
      "entropy": 5.285254240036011,
      "epoch": 1.9423631123919307,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046250006450387367,
      "loss": 4.8031,
      "mean_token_accuracy": 0.2409905582666397,
      "num_tokens": 46368250.0,
      "step": 20220
    },
    {
      "entropy": 5.350107192993164,
      "epoch": 1.9428434197886646,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004624811113938277,
      "loss": 4.8975,
      "mean_token_accuracy": 0.22637644708156585,
      "num_tokens": 46379662.0,
      "step": 20225
    },
    {
      "entropy": 5.256504249572754,
      "epoch": 1.9433237271853987,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004624621539310271,
      "loss": 4.7435,
      "mean_token_accuracy": 0.24251626133918763,
      "num_tokens": 46391485.0,
      "step": 20230
    },
    {
      "entropy": 5.327595472335815,
      "epoch": 1.9438040345821326,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046244319211591193,
      "loss": 4.9534,
      "mean_token_accuracy": 0.2282435715198517,
      "num_tokens": 46403608.0,
      "step": 20235
    },
    {
      "entropy": 5.3063617706298825,
      "epoch": 1.9442843419788665,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004624242259489223,
      "loss": 4.7479,
      "mean_token_accuracy": 0.24111166298389436,
      "num_tokens": 46416154.0,
      "step": 20240
    },
    {
      "entropy": 5.326380014419556,
      "epoch": 1.9447646493756003,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046240525543049884,
      "loss": 4.7851,
      "mean_token_accuracy": 0.2428573340177536,
      "num_tokens": 46427654.0,
      "step": 20245
    },
    {
      "entropy": 5.281146383285522,
      "epoch": 1.9452449567723344,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046238628056108176,
      "loss": 4.7944,
      "mean_token_accuracy": 0.2364579290151596,
      "num_tokens": 46438074.0,
      "step": 20250
    },
    {
      "entropy": 5.223665952682495,
      "epoch": 1.9457252641690683,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046236730134111166,
      "loss": 4.7312,
      "mean_token_accuracy": 0.24495338052511215,
      "num_tokens": 46449002.0,
      "step": 20255
    },
    {
      "entropy": 5.30032639503479,
      "epoch": 1.9462055715658022,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004623483177710291,
      "loss": 4.7974,
      "mean_token_accuracy": 0.238482666015625,
      "num_tokens": 46460576.0,
      "step": 20260
    },
    {
      "entropy": 5.391250038146973,
      "epoch": 1.946685878962536,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004623293298512751,
      "loss": 4.8379,
      "mean_token_accuracy": 0.23680564314126967,
      "num_tokens": 46471636.0,
      "step": 20265
    },
    {
      "entropy": 5.22972846031189,
      "epoch": 1.94716618635927,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046231033758229026,
      "loss": 4.7343,
      "mean_token_accuracy": 0.2380402699112892,
      "num_tokens": 46482364.0,
      "step": 20270
    },
    {
      "entropy": 5.31588191986084,
      "epoch": 1.9476464937560038,
      "grad_norm": 1.0,
      "learning_rate": 0.0004622913409645154,
      "loss": 4.853,
      "mean_token_accuracy": 0.23046952486038208,
      "num_tokens": 46493612.0,
      "step": 20275
    },
    {
      "entropy": 5.270754003524781,
      "epoch": 1.9481268011527377,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004622723399983919,
      "loss": 4.7823,
      "mean_token_accuracy": 0.23998814225196838,
      "num_tokens": 46504825.0,
      "step": 20280
    },
    {
      "entropy": 5.261887502670288,
      "epoch": 1.9486071085494716,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046225333468436077,
      "loss": 4.7359,
      "mean_token_accuracy": 0.24565144032239913,
      "num_tokens": 46515094.0,
      "step": 20285
    },
    {
      "entropy": 5.321150302886963,
      "epoch": 1.9490874159462055,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046223432502286323,
      "loss": 4.8051,
      "mean_token_accuracy": 0.23454590141773224,
      "num_tokens": 46527956.0,
      "step": 20290
    },
    {
      "entropy": 5.239185333251953,
      "epoch": 1.9495677233429394,
      "grad_norm": 1.0,
      "learning_rate": 0.00046221531101434056,
      "loss": 4.7464,
      "mean_token_accuracy": 0.24362321943044662,
      "num_tokens": 46539446.0,
      "step": 20295
    },
    {
      "entropy": 5.373226976394653,
      "epoch": 1.9500480307396733,
      "grad_norm": 1.0,
      "learning_rate": 0.0004621962926592343,
      "loss": 4.8438,
      "mean_token_accuracy": 0.23991678059101104,
      "num_tokens": 46551008.0,
      "step": 20300
    },
    {
      "entropy": 5.184166288375854,
      "epoch": 1.9505283381364071,
      "grad_norm": 1.0625,
      "learning_rate": 0.000462177269957986,
      "loss": 4.6623,
      "mean_token_accuracy": 0.24832851588726043,
      "num_tokens": 46563904.0,
      "step": 20305
    },
    {
      "entropy": 5.31282606124878,
      "epoch": 1.9510086455331412,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004621582429110373,
      "loss": 4.8227,
      "mean_token_accuracy": 0.2324594795703888,
      "num_tokens": 46575065.0,
      "step": 20310
    },
    {
      "entropy": 5.263319969177246,
      "epoch": 1.9514889529298751,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046213921151883,
      "loss": 4.7457,
      "mean_token_accuracy": 0.242560575902462,
      "num_tokens": 46587683.0,
      "step": 20315
    },
    {
      "entropy": 5.263176536560058,
      "epoch": 1.951969260326609,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004621201757818059,
      "loss": 4.7414,
      "mean_token_accuracy": 0.2498372808098793,
      "num_tokens": 46600060.0,
      "step": 20320
    },
    {
      "entropy": 5.313458442687988,
      "epoch": 1.952449567723343,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046210113570040683,
      "loss": 4.8435,
      "mean_token_accuracy": 0.23395503014326097,
      "num_tokens": 46611619.0,
      "step": 20325
    },
    {
      "entropy": 5.251535367965698,
      "epoch": 1.952929875120077,
      "grad_norm": 1.078125,
      "learning_rate": 0.000462082091275075,
      "loss": 4.7094,
      "mean_token_accuracy": 0.24208119213581086,
      "num_tokens": 46622286.0,
      "step": 20330
    },
    {
      "entropy": 5.349640321731568,
      "epoch": 1.9534101825168109,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004620630425062526,
      "loss": 4.8573,
      "mean_token_accuracy": 0.23010816276073456,
      "num_tokens": 46633402.0,
      "step": 20335
    },
    {
      "entropy": 5.18636908531189,
      "epoch": 1.9538904899135447,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004620439893943817,
      "loss": 4.7371,
      "mean_token_accuracy": 0.2438398703932762,
      "num_tokens": 46644889.0,
      "step": 20340
    },
    {
      "entropy": 5.261611461639404,
      "epoch": 1.9543707973102786,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004620249319399049,
      "loss": 4.7856,
      "mean_token_accuracy": 0.24520948380231858,
      "num_tokens": 46656355.0,
      "step": 20345
    },
    {
      "entropy": 5.2842724323272705,
      "epoch": 1.9548511047070125,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046200587014326455,
      "loss": 4.8179,
      "mean_token_accuracy": 0.23087991625070572,
      "num_tokens": 46668570.0,
      "step": 20350
    },
    {
      "entropy": 5.300688457489014,
      "epoch": 1.9553314121037464,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004619868040049031,
      "loss": 4.7758,
      "mean_token_accuracy": 0.23492701947689057,
      "num_tokens": 46680015.0,
      "step": 20355
    },
    {
      "entropy": 5.29052700996399,
      "epoch": 1.9558117195004803,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004619677335252633,
      "loss": 4.8115,
      "mean_token_accuracy": 0.23414433002471924,
      "num_tokens": 46690872.0,
      "step": 20360
    },
    {
      "entropy": 5.33471417427063,
      "epoch": 1.9562920268972142,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046194865870478793,
      "loss": 4.7922,
      "mean_token_accuracy": 0.2351495534181595,
      "num_tokens": 46702760.0,
      "step": 20365
    },
    {
      "entropy": 5.299415063858032,
      "epoch": 1.956772334293948,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046192957954391983,
      "loss": 4.8468,
      "mean_token_accuracy": 0.2306791588664055,
      "num_tokens": 46715038.0,
      "step": 20370
    },
    {
      "entropy": 5.329598379135132,
      "epoch": 1.957252641690682,
      "grad_norm": 1.046875,
      "learning_rate": 0.000461910496043102,
      "loss": 4.8101,
      "mean_token_accuracy": 0.23794560730457306,
      "num_tokens": 46725851.0,
      "step": 20375
    },
    {
      "entropy": 5.281465339660644,
      "epoch": 1.9577329490874158,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004618914082027773,
      "loss": 4.7317,
      "mean_token_accuracy": 0.2420224130153656,
      "num_tokens": 46736380.0,
      "step": 20380
    },
    {
      "entropy": 5.199045991897583,
      "epoch": 1.95821325648415,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046187231602338926,
      "loss": 4.7071,
      "mean_token_accuracy": 0.24608220010995865,
      "num_tokens": 46747872.0,
      "step": 20385
    },
    {
      "entropy": 5.253033399581909,
      "epoch": 1.9586935638808838,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046185321950538086,
      "loss": 4.6968,
      "mean_token_accuracy": 0.23988113552331924,
      "num_tokens": 46759704.0,
      "step": 20390
    },
    {
      "entropy": 5.306225919723511,
      "epoch": 1.9591738712776177,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004618341186491955,
      "loss": 4.8009,
      "mean_token_accuracy": 0.24106279015541077,
      "num_tokens": 46770895.0,
      "step": 20395
    },
    {
      "entropy": 5.324343872070313,
      "epoch": 1.9596541786743515,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004618150134552768,
      "loss": 4.7945,
      "mean_token_accuracy": 0.23780770152807235,
      "num_tokens": 46781190.0,
      "step": 20400
    },
    {
      "entropy": 5.259131717681885,
      "epoch": 1.9601344860710856,
      "grad_norm": 1.125,
      "learning_rate": 0.000461795903924068,
      "loss": 4.7837,
      "mean_token_accuracy": 0.24124163538217544,
      "num_tokens": 46792172.0,
      "step": 20405
    },
    {
      "entropy": 5.2903639793396,
      "epoch": 1.9606147934678195,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046177679005601313,
      "loss": 4.875,
      "mean_token_accuracy": 0.2260477915406227,
      "num_tokens": 46804189.0,
      "step": 20410
    },
    {
      "entropy": 5.339194965362549,
      "epoch": 1.9610951008645534,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004617576718515558,
      "loss": 4.8309,
      "mean_token_accuracy": 0.2312603324651718,
      "num_tokens": 46815127.0,
      "step": 20415
    },
    {
      "entropy": 5.29616641998291,
      "epoch": 1.9615754082612873,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004617385493111399,
      "loss": 4.829,
      "mean_token_accuracy": 0.23669522255659103,
      "num_tokens": 46825927.0,
      "step": 20420
    },
    {
      "entropy": 5.35881781578064,
      "epoch": 1.9620557156580212,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004617194224352093,
      "loss": 4.8672,
      "mean_token_accuracy": 0.23118499666452408,
      "num_tokens": 46836689.0,
      "step": 20425
    },
    {
      "entropy": 5.272569990158081,
      "epoch": 1.962536023054755,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004617002912242083,
      "loss": 4.729,
      "mean_token_accuracy": 0.2475740447640419,
      "num_tokens": 46847650.0,
      "step": 20430
    },
    {
      "entropy": 5.206133079528809,
      "epoch": 1.963016330451489,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046168115567858084,
      "loss": 4.7696,
      "mean_token_accuracy": 0.24303744584321976,
      "num_tokens": 46859469.0,
      "step": 20435
    },
    {
      "entropy": 5.265421056747437,
      "epoch": 1.9634966378482228,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046166201579877125,
      "loss": 4.7409,
      "mean_token_accuracy": 0.24277161657810212,
      "num_tokens": 46869583.0,
      "step": 20440
    },
    {
      "entropy": 5.332756233215332,
      "epoch": 1.9639769452449567,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004616428715852239,
      "loss": 4.7491,
      "mean_token_accuracy": 0.24648621082305908,
      "num_tokens": 46881357.0,
      "step": 20445
    },
    {
      "entropy": 5.2328558444976805,
      "epoch": 1.9644572526416906,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004616237230383833,
      "loss": 4.7972,
      "mean_token_accuracy": 0.23858246207237244,
      "num_tokens": 46893627.0,
      "step": 20450
    },
    {
      "entropy": 5.299405574798584,
      "epoch": 1.9649375600384245,
      "grad_norm": 1.125,
      "learning_rate": 0.00046160457015869414,
      "loss": 4.8287,
      "mean_token_accuracy": 0.23319570422172547,
      "num_tokens": 46904621.0,
      "step": 20455
    },
    {
      "entropy": 5.366876268386841,
      "epoch": 1.9654178674351583,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046158541294660083,
      "loss": 4.8203,
      "mean_token_accuracy": 0.23175910264253616,
      "num_tokens": 46914381.0,
      "step": 20460
    },
    {
      "entropy": 5.363978433609009,
      "epoch": 1.9658981748318924,
      "grad_norm": 1.0,
      "learning_rate": 0.0004615662514025483,
      "loss": 4.7627,
      "mean_token_accuracy": 0.2364461228251457,
      "num_tokens": 46925139.0,
      "step": 20465
    },
    {
      "entropy": 5.326973104476929,
      "epoch": 1.9663784822286263,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046154708552698147,
      "loss": 4.7977,
      "mean_token_accuracy": 0.23422815054655075,
      "num_tokens": 46937638.0,
      "step": 20470
    },
    {
      "entropy": 5.209340763092041,
      "epoch": 1.9668587896253602,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046152791532034517,
      "loss": 4.8183,
      "mean_token_accuracy": 0.23915256708860397,
      "num_tokens": 46949460.0,
      "step": 20475
    },
    {
      "entropy": 5.330128574371338,
      "epoch": 1.967339097022094,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046150874078308463,
      "loss": 4.8265,
      "mean_token_accuracy": 0.22940124720335006,
      "num_tokens": 46961047.0,
      "step": 20480
    },
    {
      "entropy": 5.310081958770752,
      "epoch": 1.9678194044188282,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004614895619156449,
      "loss": 4.7756,
      "mean_token_accuracy": 0.24143534898757935,
      "num_tokens": 46972009.0,
      "step": 20485
    },
    {
      "entropy": 5.210238027572632,
      "epoch": 1.968299711815562,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046147037871847125,
      "loss": 4.7214,
      "mean_token_accuracy": 0.2419111967086792,
      "num_tokens": 46982911.0,
      "step": 20490
    },
    {
      "entropy": 5.294147062301636,
      "epoch": 1.968780019212296,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004614511911920092,
      "loss": 4.7907,
      "mean_token_accuracy": 0.23172235637903213,
      "num_tokens": 46995096.0,
      "step": 20495
    },
    {
      "entropy": 5.370435476303101,
      "epoch": 1.9692603266090298,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046143199933670404,
      "loss": 4.8725,
      "mean_token_accuracy": 0.23388825207948685,
      "num_tokens": 47006814.0,
      "step": 20500
    },
    {
      "entropy": 5.249778985977173,
      "epoch": 1.9697406340057637,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004614128031530015,
      "loss": 4.7013,
      "mean_token_accuracy": 0.24599230140447617,
      "num_tokens": 47018204.0,
      "step": 20505
    },
    {
      "entropy": 5.243410873413086,
      "epoch": 1.9702209414024976,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046139360264134724,
      "loss": 4.8011,
      "mean_token_accuracy": 0.24027661234140396,
      "num_tokens": 47029655.0,
      "step": 20510
    },
    {
      "entropy": 5.341069173812866,
      "epoch": 1.9707012487992315,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004613743978021869,
      "loss": 4.8907,
      "mean_token_accuracy": 0.22968657910823823,
      "num_tokens": 47041830.0,
      "step": 20515
    },
    {
      "entropy": 5.328192949295044,
      "epoch": 1.9711815561959654,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046135518863596654,
      "loss": 4.8048,
      "mean_token_accuracy": 0.24456742107868196,
      "num_tokens": 47052817.0,
      "step": 20520
    },
    {
      "entropy": 5.287215328216552,
      "epoch": 1.9716618635926992,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046133597514313204,
      "loss": 4.8124,
      "mean_token_accuracy": 0.23518126010894774,
      "num_tokens": 47064452.0,
      "step": 20525
    },
    {
      "entropy": 5.366016817092896,
      "epoch": 1.9721421709894331,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004613167573241295,
      "loss": 4.8669,
      "mean_token_accuracy": 0.22863273322582245,
      "num_tokens": 47074952.0,
      "step": 20530
    },
    {
      "entropy": 5.35166368484497,
      "epoch": 1.972622478386167,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004612975351794051,
      "loss": 4.8129,
      "mean_token_accuracy": 0.2361438527703285,
      "num_tokens": 47085496.0,
      "step": 20535
    },
    {
      "entropy": 5.273460912704468,
      "epoch": 1.973102785782901,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004612783087094051,
      "loss": 4.8256,
      "mean_token_accuracy": 0.23278567045927048,
      "num_tokens": 47097798.0,
      "step": 20540
    },
    {
      "entropy": 5.367456769943237,
      "epoch": 1.973583093179635,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046125907791457594,
      "loss": 4.8867,
      "mean_token_accuracy": 0.23561851978302,
      "num_tokens": 47109798.0,
      "step": 20545
    },
    {
      "entropy": 5.35615496635437,
      "epoch": 1.9740634005763689,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046123984279536405,
      "loss": 4.8191,
      "mean_token_accuracy": 0.23806620240211487,
      "num_tokens": 47120073.0,
      "step": 20550
    },
    {
      "entropy": 5.3219382762908936,
      "epoch": 1.9745437079731027,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046122060335221604,
      "loss": 4.801,
      "mean_token_accuracy": 0.2353764593601227,
      "num_tokens": 47130596.0,
      "step": 20555
    },
    {
      "entropy": 5.215368938446045,
      "epoch": 1.9750240153698368,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046120135958557855,
      "loss": 4.7181,
      "mean_token_accuracy": 0.25189929753541945,
      "num_tokens": 47141660.0,
      "step": 20560
    },
    {
      "entropy": 5.45718092918396,
      "epoch": 1.9755043227665707,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046118211149589843,
      "loss": 4.961,
      "mean_token_accuracy": 0.2179691895842552,
      "num_tokens": 47153046.0,
      "step": 20565
    },
    {
      "entropy": 5.406516885757446,
      "epoch": 1.9759846301633046,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004611628590836225,
      "loss": 4.913,
      "mean_token_accuracy": 0.2295879602432251,
      "num_tokens": 47162706.0,
      "step": 20570
    },
    {
      "entropy": 5.278098297119141,
      "epoch": 1.9764649375600385,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004611436023491979,
      "loss": 4.8039,
      "mean_token_accuracy": 0.23424024283885955,
      "num_tokens": 47174508.0,
      "step": 20575
    },
    {
      "entropy": 5.290502691268921,
      "epoch": 1.9769452449567724,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004611243412930714,
      "loss": 4.7265,
      "mean_token_accuracy": 0.24112486988306045,
      "num_tokens": 47185019.0,
      "step": 20580
    },
    {
      "entropy": 5.261726522445679,
      "epoch": 1.9774255523535063,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046110507591569047,
      "loss": 4.7568,
      "mean_token_accuracy": 0.24300662577152252,
      "num_tokens": 47195538.0,
      "step": 20585
    },
    {
      "entropy": 5.326708078384399,
      "epoch": 1.9779058597502401,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004610858062175023,
      "loss": 4.7985,
      "mean_token_accuracy": 0.23911076486110688,
      "num_tokens": 47206398.0,
      "step": 20590
    },
    {
      "entropy": 5.3092142105102536,
      "epoch": 1.978386167146974,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046106653219895417,
      "loss": 4.8528,
      "mean_token_accuracy": 0.23061247617006303,
      "num_tokens": 47217775.0,
      "step": 20595
    },
    {
      "entropy": 5.314103651046753,
      "epoch": 1.978866474543708,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004610472538604938,
      "loss": 4.8279,
      "mean_token_accuracy": 0.2323402836918831,
      "num_tokens": 47230179.0,
      "step": 20600
    },
    {
      "entropy": 5.270280456542968,
      "epoch": 1.9793467819404418,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046102797120256854,
      "loss": 4.6989,
      "mean_token_accuracy": 0.24970816522836686,
      "num_tokens": 47241416.0,
      "step": 20605
    },
    {
      "entropy": 5.240595388412475,
      "epoch": 1.9798270893371757,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004610086842256262,
      "loss": 4.764,
      "mean_token_accuracy": 0.23522841185331345,
      "num_tokens": 47253916.0,
      "step": 20610
    },
    {
      "entropy": 5.362278604507447,
      "epoch": 1.9803073967339095,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004609893929301146,
      "loss": 4.8662,
      "mean_token_accuracy": 0.2305807501077652,
      "num_tokens": 47265176.0,
      "step": 20615
    },
    {
      "entropy": 5.318528747558593,
      "epoch": 1.9807877041306436,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004609700973164816,
      "loss": 4.7574,
      "mean_token_accuracy": 0.2331048220396042,
      "num_tokens": 47276178.0,
      "step": 20620
    },
    {
      "entropy": 5.2082854270935055,
      "epoch": 1.9812680115273775,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004609507973851751,
      "loss": 4.7692,
      "mean_token_accuracy": 0.23514583259820937,
      "num_tokens": 47287607.0,
      "step": 20625
    },
    {
      "entropy": 5.337826251983643,
      "epoch": 1.9817483189241114,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046093149313664316,
      "loss": 4.9091,
      "mean_token_accuracy": 0.22546651661396028,
      "num_tokens": 47300069.0,
      "step": 20630
    },
    {
      "entropy": 5.2967884063720705,
      "epoch": 1.9822286263208453,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004609121845713342,
      "loss": 4.7482,
      "mean_token_accuracy": 0.24628988802433013,
      "num_tokens": 47312349.0,
      "step": 20635
    },
    {
      "entropy": 5.242399215698242,
      "epoch": 1.9827089337175794,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004608928716896963,
      "loss": 4.7824,
      "mean_token_accuracy": 0.24382131546735764,
      "num_tokens": 47323171.0,
      "step": 20640
    },
    {
      "entropy": 5.326991891860962,
      "epoch": 1.9831892411143133,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004608735544921778,
      "loss": 4.8502,
      "mean_token_accuracy": 0.2334582582116127,
      "num_tokens": 47333990.0,
      "step": 20645
    },
    {
      "entropy": 5.250008964538575,
      "epoch": 1.9836695485110472,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046085423297922745,
      "loss": 4.7646,
      "mean_token_accuracy": 0.2363168329000473,
      "num_tokens": 47346081.0,
      "step": 20650
    },
    {
      "entropy": 5.424335861206055,
      "epoch": 1.984149855907781,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046083490715129367,
      "loss": 4.9624,
      "mean_token_accuracy": 0.22970577478408813,
      "num_tokens": 47357979.0,
      "step": 20655
    },
    {
      "entropy": 5.454973554611206,
      "epoch": 1.984630163304515,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004608155770088251,
      "loss": 4.8765,
      "mean_token_accuracy": 0.22337692528963088,
      "num_tokens": 47369066.0,
      "step": 20660
    },
    {
      "entropy": 5.325919342041016,
      "epoch": 1.9851104707012488,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046079624255227066,
      "loss": 4.8389,
      "mean_token_accuracy": 0.2337113305926323,
      "num_tokens": 47380316.0,
      "step": 20665
    },
    {
      "entropy": 5.273400974273682,
      "epoch": 1.9855907780979827,
      "grad_norm": 1.0,
      "learning_rate": 0.0004607769037820791,
      "loss": 4.8181,
      "mean_token_accuracy": 0.23587167114019394,
      "num_tokens": 47392040.0,
      "step": 20670
    },
    {
      "entropy": 5.2806511402130125,
      "epoch": 1.9860710854947166,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004607575606986995,
      "loss": 4.8117,
      "mean_token_accuracy": 0.23829571455717086,
      "num_tokens": 47404114.0,
      "step": 20675
    },
    {
      "entropy": 5.2495733261108395,
      "epoch": 1.9865513928914504,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004607382133025809,
      "loss": 4.7612,
      "mean_token_accuracy": 0.24319465607404708,
      "num_tokens": 47416581.0,
      "step": 20680
    },
    {
      "entropy": 5.261407566070557,
      "epoch": 1.9870317002881843,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046071886159417257,
      "loss": 4.6824,
      "mean_token_accuracy": 0.2505770429968834,
      "num_tokens": 47427314.0,
      "step": 20685
    },
    {
      "entropy": 5.366486215591431,
      "epoch": 1.9875120076849182,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004606995055739238,
      "loss": 4.8772,
      "mean_token_accuracy": 0.22765283286571503,
      "num_tokens": 47438672.0,
      "step": 20690
    },
    {
      "entropy": 5.362815523147583,
      "epoch": 1.9879923150816523,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046068014524228374,
      "loss": 4.8754,
      "mean_token_accuracy": 0.23716043829917907,
      "num_tokens": 47449901.0,
      "step": 20695
    },
    {
      "entropy": 5.293164253234863,
      "epoch": 1.9884726224783862,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046066078059970217,
      "loss": 4.8609,
      "mean_token_accuracy": 0.23298921585083007,
      "num_tokens": 47462831.0,
      "step": 20700
    },
    {
      "entropy": 5.326147222518921,
      "epoch": 1.98895292987512,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004606414116466286,
      "loss": 4.8129,
      "mean_token_accuracy": 0.2397354692220688,
      "num_tokens": 47474075.0,
      "step": 20705
    },
    {
      "entropy": 5.348658657073974,
      "epoch": 1.989433237271854,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046062203838351267,
      "loss": 4.9219,
      "mean_token_accuracy": 0.2303904563188553,
      "num_tokens": 47486893.0,
      "step": 20710
    },
    {
      "entropy": 5.345154476165772,
      "epoch": 1.989913544668588,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046060266081080414,
      "loss": 4.8318,
      "mean_token_accuracy": 0.23605775684118271,
      "num_tokens": 47497969.0,
      "step": 20715
    },
    {
      "entropy": 5.418742370605469,
      "epoch": 1.990393852065322,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004605832789289531,
      "loss": 4.8849,
      "mean_token_accuracy": 0.23864722549915313,
      "num_tokens": 47509476.0,
      "step": 20720
    },
    {
      "entropy": 5.283522367477417,
      "epoch": 1.9908741594620558,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004605638927384093,
      "loss": 4.7438,
      "mean_token_accuracy": 0.24272841066122056,
      "num_tokens": 47520124.0,
      "step": 20725
    },
    {
      "entropy": 5.2711906909942625,
      "epoch": 1.9913544668587897,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046054450223962284,
      "loss": 4.7471,
      "mean_token_accuracy": 0.24267758131027223,
      "num_tokens": 47532079.0,
      "step": 20730
    },
    {
      "entropy": 5.310142374038696,
      "epoch": 1.9918347742555236,
      "grad_norm": 1.0,
      "learning_rate": 0.00046052510743304405,
      "loss": 4.7798,
      "mean_token_accuracy": 0.24473518133163452,
      "num_tokens": 47544142.0,
      "step": 20735
    },
    {
      "entropy": 5.291308212280273,
      "epoch": 1.9923150816522575,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004605057083191232,
      "loss": 4.7966,
      "mean_token_accuracy": 0.23718674778938292,
      "num_tokens": 47555688.0,
      "step": 20740
    },
    {
      "entropy": 5.28134708404541,
      "epoch": 1.9927953890489913,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004604863048983106,
      "loss": 4.7609,
      "mean_token_accuracy": 0.24227488189935684,
      "num_tokens": 47566590.0,
      "step": 20745
    },
    {
      "entropy": 5.328578281402588,
      "epoch": 1.9932756964457252,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004604668971710568,
      "loss": 4.826,
      "mean_token_accuracy": 0.246590256690979,
      "num_tokens": 47577781.0,
      "step": 20750
    },
    {
      "entropy": 5.308039903640747,
      "epoch": 1.993756003842459,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004604474851378124,
      "loss": 4.742,
      "mean_token_accuracy": 0.25048361122608187,
      "num_tokens": 47589074.0,
      "step": 20755
    },
    {
      "entropy": 5.282664823532104,
      "epoch": 1.994236311239193,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046042806879902803,
      "loss": 4.809,
      "mean_token_accuracy": 0.24113133251667024,
      "num_tokens": 47600841.0,
      "step": 20760
    },
    {
      "entropy": 5.344459676742554,
      "epoch": 1.9947166186359269,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004604086481551546,
      "loss": 4.8102,
      "mean_token_accuracy": 0.23721210807561874,
      "num_tokens": 47612332.0,
      "step": 20765
    },
    {
      "entropy": 5.301406955718994,
      "epoch": 1.9951969260326607,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004603892232066428,
      "loss": 4.7844,
      "mean_token_accuracy": 0.23577394485473632,
      "num_tokens": 47623773.0,
      "step": 20770
    },
    {
      "entropy": 5.297901916503906,
      "epoch": 1.9956772334293948,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046036979395394374,
      "loss": 4.7663,
      "mean_token_accuracy": 0.23800451606512069,
      "num_tokens": 47634802.0,
      "step": 20775
    },
    {
      "entropy": 5.362395715713501,
      "epoch": 1.9961575408261287,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004603503603975085,
      "loss": 4.8924,
      "mean_token_accuracy": 0.23082840889692308,
      "num_tokens": 47646772.0,
      "step": 20780
    },
    {
      "entropy": 5.363853979110718,
      "epoch": 1.9966378482228626,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004603309225377883,
      "loss": 4.8238,
      "mean_token_accuracy": 0.23597587794065475,
      "num_tokens": 47658321.0,
      "step": 20785
    },
    {
      "entropy": 5.239433431625367,
      "epoch": 1.9971181556195965,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004603114803752344,
      "loss": 4.7351,
      "mean_token_accuracy": 0.247407765686512,
      "num_tokens": 47668996.0,
      "step": 20790
    },
    {
      "entropy": 5.268589639663697,
      "epoch": 1.9975984630163306,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046029203391029813,
      "loss": 4.7913,
      "mean_token_accuracy": 0.24017405211925508,
      "num_tokens": 47679091.0,
      "step": 20795
    },
    {
      "entropy": 5.319162845611572,
      "epoch": 1.9980787704130645,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046027258314343107,
      "loss": 4.7927,
      "mean_token_accuracy": 0.24045733064413072,
      "num_tokens": 47690740.0,
      "step": 20800
    },
    {
      "entropy": 5.279374599456787,
      "epoch": 1.9985590778097984,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046025312807508487,
      "loss": 4.7599,
      "mean_token_accuracy": 0.23997929245233535,
      "num_tokens": 47703514.0,
      "step": 20805
    },
    {
      "entropy": 5.235484504699707,
      "epoch": 1.9990393852065322,
      "grad_norm": 1.0,
      "learning_rate": 0.00046023366870571097,
      "loss": 4.7812,
      "mean_token_accuracy": 0.23539731949567794,
      "num_tokens": 47715220.0,
      "step": 20810
    },
    {
      "entropy": 5.329309558868408,
      "epoch": 1.9995196926032661,
      "grad_norm": 1.0,
      "learning_rate": 0.00046021420503576145,
      "loss": 4.8093,
      "mean_token_accuracy": 0.24090158492326735,
      "num_tokens": 47727387.0,
      "step": 20815
    },
    {
      "entropy": 5.398097324371338,
      "epoch": 2.0,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004601947370656879,
      "loss": 4.8802,
      "mean_token_accuracy": 0.2300075799226761,
      "num_tokens": 47737072.0,
      "step": 20820
    },
    {
      "entropy": 5.2393756866455075,
      "epoch": 2.000480307396734,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004601752647959426,
      "loss": 4.6699,
      "mean_token_accuracy": 0.25638543218374255,
      "num_tokens": 47748516.0,
      "step": 20825
    },
    {
      "entropy": 5.217671966552734,
      "epoch": 2.0009606147934678,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004601557882269775,
      "loss": 4.6903,
      "mean_token_accuracy": 0.2439010888338089,
      "num_tokens": 47760482.0,
      "step": 20830
    },
    {
      "entropy": 5.335058069229126,
      "epoch": 2.0014409221902016,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004601363073592447,
      "loss": 4.7219,
      "mean_token_accuracy": 0.24766440093517303,
      "num_tokens": 47771204.0,
      "step": 20835
    },
    {
      "entropy": 5.267731142044068,
      "epoch": 2.0019212295869355,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004601168221931967,
      "loss": 4.7156,
      "mean_token_accuracy": 0.2452890008687973,
      "num_tokens": 47782404.0,
      "step": 20840
    },
    {
      "entropy": 5.313049936294556,
      "epoch": 2.0024015369836694,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004600973327292857,
      "loss": 4.766,
      "mean_token_accuracy": 0.2383619248867035,
      "num_tokens": 47793612.0,
      "step": 20845
    },
    {
      "entropy": 5.2973119735717775,
      "epoch": 2.0028818443804033,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046007783896796436,
      "loss": 4.6805,
      "mean_token_accuracy": 0.24738150238990783,
      "num_tokens": 47804467.0,
      "step": 20850
    },
    {
      "entropy": 5.394306230545044,
      "epoch": 2.003362151777137,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004600583409096851,
      "loss": 4.8114,
      "mean_token_accuracy": 0.2319738432765007,
      "num_tokens": 47815654.0,
      "step": 20855
    },
    {
      "entropy": 5.342439413070679,
      "epoch": 2.0038424591738715,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046003883855490066,
      "loss": 4.7928,
      "mean_token_accuracy": 0.23847392350435256,
      "num_tokens": 47827630.0,
      "step": 20860
    },
    {
      "entropy": 5.2972267150878904,
      "epoch": 2.0043227665706054,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004600193319040638,
      "loss": 4.7102,
      "mean_token_accuracy": 0.24335910379886627,
      "num_tokens": 47838046.0,
      "step": 20865
    },
    {
      "entropy": 5.317156791687012,
      "epoch": 2.0048030739673393,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045999982095762756,
      "loss": 4.7679,
      "mean_token_accuracy": 0.2419344052672386,
      "num_tokens": 47849506.0,
      "step": 20870
    },
    {
      "entropy": 5.229816865921021,
      "epoch": 2.005283381364073,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045998030571604473,
      "loss": 4.6493,
      "mean_token_accuracy": 0.2509633332490921,
      "num_tokens": 47860689.0,
      "step": 20875
    },
    {
      "entropy": 5.2823402881622314,
      "epoch": 2.005763688760807,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004599607861797685,
      "loss": 4.6776,
      "mean_token_accuracy": 0.2523679196834564,
      "num_tokens": 47872215.0,
      "step": 20880
    },
    {
      "entropy": 5.298958587646484,
      "epoch": 2.006243996157541,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045994126234925203,
      "loss": 4.8355,
      "mean_token_accuracy": 0.2349218264222145,
      "num_tokens": 47883726.0,
      "step": 20885
    },
    {
      "entropy": 5.355041265487671,
      "epoch": 2.006724303554275,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045992173422494865,
      "loss": 4.7293,
      "mean_token_accuracy": 0.23454415649175644,
      "num_tokens": 47893915.0,
      "step": 20890
    },
    {
      "entropy": 5.346019887924195,
      "epoch": 2.0072046109510087,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004599022018073117,
      "loss": 4.8173,
      "mean_token_accuracy": 0.24080432802438737,
      "num_tokens": 47905026.0,
      "step": 20895
    },
    {
      "entropy": 5.306896162033081,
      "epoch": 2.0076849183477425,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004598826650967946,
      "loss": 4.7938,
      "mean_token_accuracy": 0.24149762988090515,
      "num_tokens": 47918410.0,
      "step": 20900
    },
    {
      "entropy": 5.352611351013183,
      "epoch": 2.0081652257444764,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045986312409385105,
      "loss": 4.7958,
      "mean_token_accuracy": 0.23382787257432938,
      "num_tokens": 47929584.0,
      "step": 20905
    },
    {
      "entropy": 5.2985076904296875,
      "epoch": 2.0086455331412103,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004598435787989347,
      "loss": 4.7829,
      "mean_token_accuracy": 0.23342761546373367,
      "num_tokens": 47940585.0,
      "step": 20910
    },
    {
      "entropy": 5.283606147766113,
      "epoch": 2.009125840537944,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045982402921249934,
      "loss": 4.7545,
      "mean_token_accuracy": 0.23447577506303788,
      "num_tokens": 47953951.0,
      "step": 20915
    },
    {
      "entropy": 5.220671510696411,
      "epoch": 2.009606147934678,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004598044753349988,
      "loss": 4.6676,
      "mean_token_accuracy": 0.2408484250307083,
      "num_tokens": 47965783.0,
      "step": 20920
    },
    {
      "entropy": 5.262368774414062,
      "epoch": 2.010086455331412,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00045978491716688706,
      "loss": 4.7107,
      "mean_token_accuracy": 0.23940306305885314,
      "num_tokens": 47978509.0,
      "step": 20925
    },
    {
      "entropy": 5.361297750473023,
      "epoch": 2.010566762728146,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004597653547086184,
      "loss": 4.7978,
      "mean_token_accuracy": 0.24140390902757644,
      "num_tokens": 47989828.0,
      "step": 20930
    },
    {
      "entropy": 5.313862085342407,
      "epoch": 2.0110470701248797,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004597457879606467,
      "loss": 4.6588,
      "mean_token_accuracy": 0.24536834806203842,
      "num_tokens": 47999683.0,
      "step": 20935
    },
    {
      "entropy": 5.225770902633667,
      "epoch": 2.011527377521614,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045972621692342636,
      "loss": 4.5956,
      "mean_token_accuracy": 0.25099281072616575,
      "num_tokens": 48010526.0,
      "step": 20940
    },
    {
      "entropy": 5.176472234725952,
      "epoch": 2.012007684918348,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045970664159741186,
      "loss": 4.6314,
      "mean_token_accuracy": 0.25751496851444244,
      "num_tokens": 48021383.0,
      "step": 20945
    },
    {
      "entropy": 5.147475337982177,
      "epoch": 2.012487992315082,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045968706198305765,
      "loss": 4.5698,
      "mean_token_accuracy": 0.2541489452123642,
      "num_tokens": 48031459.0,
      "step": 20950
    },
    {
      "entropy": 5.353528118133545,
      "epoch": 2.0129682997118157,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045966747808081824,
      "loss": 4.8296,
      "mean_token_accuracy": 0.23401206135749816,
      "num_tokens": 48043590.0,
      "step": 20955
    },
    {
      "entropy": 5.230110502243042,
      "epoch": 2.0134486071085496,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004596478898911483,
      "loss": 4.6318,
      "mean_token_accuracy": 0.248113414645195,
      "num_tokens": 48054715.0,
      "step": 20960
    },
    {
      "entropy": 5.294203090667724,
      "epoch": 2.0139289145052834,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045962829741450265,
      "loss": 4.7405,
      "mean_token_accuracy": 0.23833052963018417,
      "num_tokens": 48065606.0,
      "step": 20965
    },
    {
      "entropy": 5.361145782470703,
      "epoch": 2.0144092219020173,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004596087006513361,
      "loss": 4.8134,
      "mean_token_accuracy": 0.23604495376348494,
      "num_tokens": 48076097.0,
      "step": 20970
    },
    {
      "entropy": 5.393516159057617,
      "epoch": 2.014889529298751,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045958909960210385,
      "loss": 4.8173,
      "mean_token_accuracy": 0.23486543595790862,
      "num_tokens": 48087429.0,
      "step": 20975
    },
    {
      "entropy": 5.173222208023072,
      "epoch": 2.015369836695485,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045956949426726075,
      "loss": 4.6711,
      "mean_token_accuracy": 0.24674167931079866,
      "num_tokens": 48098347.0,
      "step": 20980
    },
    {
      "entropy": 5.175881004333496,
      "epoch": 2.015850144092219,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045954988464726203,
      "loss": 4.5804,
      "mean_token_accuracy": 0.25335948914289474,
      "num_tokens": 48109355.0,
      "step": 20985
    },
    {
      "entropy": 5.304176998138428,
      "epoch": 2.016330451488953,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000459530270742563,
      "loss": 4.7802,
      "mean_token_accuracy": 0.2360669270157814,
      "num_tokens": 48121875.0,
      "step": 20990
    },
    {
      "entropy": 5.3589019775390625,
      "epoch": 2.0168107588856867,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045951065255361905,
      "loss": 4.8127,
      "mean_token_accuracy": 0.23470491468906401,
      "num_tokens": 48133192.0,
      "step": 20995
    },
    {
      "entropy": 5.293544054031372,
      "epoch": 2.0172910662824206,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004594910300808856,
      "loss": 4.6861,
      "mean_token_accuracy": 0.24817626178264618,
      "num_tokens": 48143621.0,
      "step": 21000
    },
    {
      "epoch": 2.0172910662824206,
      "eval_entropy": 5.050515402241023,
      "eval_loss": 4.840777397155762,
      "eval_mean_token_accuracy": 0.24551627228625922,
      "eval_num_tokens": 48143621.0,
      "eval_runtime": 20.4746,
      "eval_samples_per_second": 1602.715,
      "eval_steps_per_second": 200.345,
      "step": 21000
    },
    {
      "entropy": 5.2131256580352785,
      "epoch": 2.0177713736791545,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004594714033248183,
      "loss": 4.6281,
      "mean_token_accuracy": 0.24833381921052933,
      "num_tokens": 48153803.0,
      "step": 21005
    },
    {
      "entropy": 5.247600126266479,
      "epoch": 2.0182516810758884,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004594517722858728,
      "loss": 4.6993,
      "mean_token_accuracy": 0.24191970974206925,
      "num_tokens": 48165742.0,
      "step": 21010
    },
    {
      "entropy": 5.245131731033325,
      "epoch": 2.0187319884726227,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045943213696450475,
      "loss": 4.6817,
      "mean_token_accuracy": 0.24770814329385757,
      "num_tokens": 48177894.0,
      "step": 21015
    },
    {
      "entropy": 5.339456033706665,
      "epoch": 2.0192122958693566,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045941249736117023,
      "loss": 4.7707,
      "mean_token_accuracy": 0.2386813059449196,
      "num_tokens": 48188536.0,
      "step": 21020
    },
    {
      "entropy": 5.304665279388428,
      "epoch": 2.0196926032660905,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004593928534763251,
      "loss": 4.685,
      "mean_token_accuracy": 0.247812482714653,
      "num_tokens": 48199276.0,
      "step": 21025
    },
    {
      "entropy": 5.372749757766724,
      "epoch": 2.0201729106628243,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004593732053104254,
      "loss": 4.8106,
      "mean_token_accuracy": 0.24008578956127166,
      "num_tokens": 48212479.0,
      "step": 21030
    },
    {
      "entropy": 5.272351312637329,
      "epoch": 2.020653218059558,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045935355286392735,
      "loss": 4.6255,
      "mean_token_accuracy": 0.2511127695441246,
      "num_tokens": 48224279.0,
      "step": 21035
    },
    {
      "entropy": 5.287754154205322,
      "epoch": 2.021133525456292,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004593338961372873,
      "loss": 4.8012,
      "mean_token_accuracy": 0.2331051081418991,
      "num_tokens": 48235943.0,
      "step": 21040
    },
    {
      "entropy": 5.207994413375855,
      "epoch": 2.021613832853026,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004593142351309614,
      "loss": 4.596,
      "mean_token_accuracy": 0.25519100576639175,
      "num_tokens": 48246407.0,
      "step": 21045
    },
    {
      "entropy": 5.354035329818726,
      "epoch": 2.02209414024976,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004592945698454064,
      "loss": 4.7875,
      "mean_token_accuracy": 0.2387877494096756,
      "num_tokens": 48256957.0,
      "step": 21050
    },
    {
      "entropy": 5.195591783523559,
      "epoch": 2.0225744476464937,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045927490028107866,
      "loss": 4.6389,
      "mean_token_accuracy": 0.2535992920398712,
      "num_tokens": 48268253.0,
      "step": 21055
    },
    {
      "entropy": 5.332375478744507,
      "epoch": 2.0230547550432276,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000459255226438435,
      "loss": 4.8422,
      "mean_token_accuracy": 0.24251485019922256,
      "num_tokens": 48279980.0,
      "step": 21060
    },
    {
      "entropy": 5.380788850784302,
      "epoch": 2.0235350624399615,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000459235548317932,
      "loss": 4.7991,
      "mean_token_accuracy": 0.23677869588136674,
      "num_tokens": 48291199.0,
      "step": 21065
    },
    {
      "entropy": 5.38641209602356,
      "epoch": 2.0240153698366954,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045921586592002667,
      "loss": 4.7801,
      "mean_token_accuracy": 0.2365005925297737,
      "num_tokens": 48301610.0,
      "step": 21070
    },
    {
      "entropy": 5.281110239028931,
      "epoch": 2.0244956772334293,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000459196179245176,
      "loss": 4.7559,
      "mean_token_accuracy": 0.24173133373260497,
      "num_tokens": 48313655.0,
      "step": 21075
    },
    {
      "entropy": 5.2674657821655275,
      "epoch": 2.024975984630163,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004591764882938369,
      "loss": 4.6756,
      "mean_token_accuracy": 0.24614075273275376,
      "num_tokens": 48325385.0,
      "step": 21080
    },
    {
      "entropy": 5.2751976490020756,
      "epoch": 2.025456292026897,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004591567930664667,
      "loss": 4.6434,
      "mean_token_accuracy": 0.24680763781070708,
      "num_tokens": 48335986.0,
      "step": 21085
    },
    {
      "entropy": 5.301749229431152,
      "epoch": 2.025936599423631,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004591370935635226,
      "loss": 4.7928,
      "mean_token_accuracy": 0.24147737473249437,
      "num_tokens": 48347196.0,
      "step": 21090
    },
    {
      "entropy": 5.3768494606018065,
      "epoch": 2.0264169068203652,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004591173897854619,
      "loss": 4.8158,
      "mean_token_accuracy": 0.23178545981645585,
      "num_tokens": 48357944.0,
      "step": 21095
    },
    {
      "entropy": 5.3935768604278564,
      "epoch": 2.026897214217099,
      "grad_norm": 1.0,
      "learning_rate": 0.0004590976817327422,
      "loss": 4.8897,
      "mean_token_accuracy": 0.23097368329763412,
      "num_tokens": 48369838.0,
      "step": 21100
    },
    {
      "entropy": 5.300244998931885,
      "epoch": 2.027377521613833,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004590779694058209,
      "loss": 4.7085,
      "mean_token_accuracy": 0.24285829216241836,
      "num_tokens": 48381107.0,
      "step": 21105
    },
    {
      "entropy": 5.249950933456421,
      "epoch": 2.027857829010567,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045905825280515586,
      "loss": 4.7499,
      "mean_token_accuracy": 0.24618389308452607,
      "num_tokens": 48392699.0,
      "step": 21110
    },
    {
      "entropy": 5.349729776382446,
      "epoch": 2.0283381364073008,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045903853193120464,
      "loss": 4.8267,
      "mean_token_accuracy": 0.2364930361509323,
      "num_tokens": 48403836.0,
      "step": 21115
    },
    {
      "entropy": 5.236342716217041,
      "epoch": 2.0288184438040346,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045901880678442524,
      "loss": 4.6745,
      "mean_token_accuracy": 0.2515877917408943,
      "num_tokens": 48417373.0,
      "step": 21120
    },
    {
      "entropy": 5.317943572998047,
      "epoch": 2.0292987512007685,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045899907736527556,
      "loss": 4.7413,
      "mean_token_accuracy": 0.24447320103645326,
      "num_tokens": 48428033.0,
      "step": 21125
    },
    {
      "entropy": 5.281838083267212,
      "epoch": 2.0297790585975024,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045897934367421364,
      "loss": 4.6607,
      "mean_token_accuracy": 0.2499845340847969,
      "num_tokens": 48440147.0,
      "step": 21130
    },
    {
      "entropy": 5.197733783721924,
      "epoch": 2.0302593659942363,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004589596057116977,
      "loss": 4.7034,
      "mean_token_accuracy": 0.2443346381187439,
      "num_tokens": 48451406.0,
      "step": 21135
    },
    {
      "entropy": 5.29351315498352,
      "epoch": 2.03073967339097,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00045893986347818593,
      "loss": 4.6968,
      "mean_token_accuracy": 0.24932251423597335,
      "num_tokens": 48464393.0,
      "step": 21140
    },
    {
      "entropy": 5.343116521835327,
      "epoch": 2.031219980787704,
      "grad_norm": 1.0,
      "learning_rate": 0.0004589201169741368,
      "loss": 4.794,
      "mean_token_accuracy": 0.23875813633203508,
      "num_tokens": 48476433.0,
      "step": 21145
    },
    {
      "entropy": 5.303962850570679,
      "epoch": 2.031700288184438,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045890036620000856,
      "loss": 4.7276,
      "mean_token_accuracy": 0.23886261582374574,
      "num_tokens": 48488372.0,
      "step": 21150
    },
    {
      "entropy": 5.247672462463379,
      "epoch": 2.032180595581172,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045888061115626,
      "loss": 4.6806,
      "mean_token_accuracy": 0.24850929379463196,
      "num_tokens": 48498834.0,
      "step": 21155
    },
    {
      "entropy": 5.36145052909851,
      "epoch": 2.0326609029779057,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004588608518433496,
      "loss": 4.7021,
      "mean_token_accuracy": 0.2456900417804718,
      "num_tokens": 48509160.0,
      "step": 21160
    },
    {
      "entropy": 5.317860460281372,
      "epoch": 2.0331412103746396,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004588410882617362,
      "loss": 4.746,
      "mean_token_accuracy": 0.23960973769426347,
      "num_tokens": 48522005.0,
      "step": 21165
    },
    {
      "entropy": 5.246332120895386,
      "epoch": 2.0336215177713735,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004588213204118786,
      "loss": 4.7066,
      "mean_token_accuracy": 0.24482183605432511,
      "num_tokens": 48534238.0,
      "step": 21170
    },
    {
      "entropy": 5.286562061309814,
      "epoch": 2.034101825168108,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045880154829423586,
      "loss": 4.7244,
      "mean_token_accuracy": 0.24422499686479568,
      "num_tokens": 48546440.0,
      "step": 21175
    },
    {
      "entropy": 5.35047721862793,
      "epoch": 2.0345821325648417,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004587817719092668,
      "loss": 4.7217,
      "mean_token_accuracy": 0.2526050701737404,
      "num_tokens": 48556893.0,
      "step": 21180
    },
    {
      "entropy": 5.276201868057251,
      "epoch": 2.0350624399615755,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045876199125743087,
      "loss": 4.722,
      "mean_token_accuracy": 0.2422921136021614,
      "num_tokens": 48568483.0,
      "step": 21185
    },
    {
      "entropy": 5.355493545532227,
      "epoch": 2.0355427473583094,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004587422063391871,
      "loss": 4.7607,
      "mean_token_accuracy": 0.23293207287788392,
      "num_tokens": 48579352.0,
      "step": 21190
    },
    {
      "entropy": 5.357388639450074,
      "epoch": 2.0360230547550433,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004587224171549949,
      "loss": 4.7693,
      "mean_token_accuracy": 0.23299713730812072,
      "num_tokens": 48591447.0,
      "step": 21195
    },
    {
      "entropy": 5.309848880767822,
      "epoch": 2.036503362151777,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045870262370531376,
      "loss": 4.7551,
      "mean_token_accuracy": 0.23572982400655745,
      "num_tokens": 48603724.0,
      "step": 21200
    },
    {
      "entropy": 5.380260896682739,
      "epoch": 2.036983669548511,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045868282599060314,
      "loss": 4.778,
      "mean_token_accuracy": 0.2327711433172226,
      "num_tokens": 48616319.0,
      "step": 21205
    },
    {
      "entropy": 5.28028302192688,
      "epoch": 2.037463976945245,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004586630240113227,
      "loss": 4.7451,
      "mean_token_accuracy": 0.24800250977277755,
      "num_tokens": 48627121.0,
      "step": 21210
    },
    {
      "entropy": 5.249201345443725,
      "epoch": 2.037944284341979,
      "grad_norm": 1.0,
      "learning_rate": 0.0004586432177679322,
      "loss": 4.6532,
      "mean_token_accuracy": 0.25377386510372163,
      "num_tokens": 48638687.0,
      "step": 21215
    },
    {
      "entropy": 5.312063837051392,
      "epoch": 2.0384245917387127,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045862340726089153,
      "loss": 4.7951,
      "mean_token_accuracy": 0.23486456423997878,
      "num_tokens": 48649480.0,
      "step": 21220
    },
    {
      "entropy": 5.397331380844117,
      "epoch": 2.0389048991354466,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004586035924906606,
      "loss": 4.8817,
      "mean_token_accuracy": 0.22936951518058776,
      "num_tokens": 48660197.0,
      "step": 21225
    },
    {
      "entropy": 5.2396492004394535,
      "epoch": 2.0393852065321805,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045858377345769946,
      "loss": 4.6339,
      "mean_token_accuracy": 0.24354548305273055,
      "num_tokens": 48671711.0,
      "step": 21230
    },
    {
      "entropy": 5.238779592514038,
      "epoch": 2.0398655139289144,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004585639501624682,
      "loss": 4.6586,
      "mean_token_accuracy": 0.24729667454957963,
      "num_tokens": 48683285.0,
      "step": 21235
    },
    {
      "entropy": 5.328602600097656,
      "epoch": 2.0403458213256482,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000458544122605427,
      "loss": 4.7252,
      "mean_token_accuracy": 0.24833420515060425,
      "num_tokens": 48696308.0,
      "step": 21240
    },
    {
      "entropy": 5.2382895946502686,
      "epoch": 2.040826128722382,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045852429078703646,
      "loss": 4.581,
      "mean_token_accuracy": 0.25525871813297274,
      "num_tokens": 48706999.0,
      "step": 21245
    },
    {
      "entropy": 5.2776055335998535,
      "epoch": 2.0413064361191164,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045850445470775673,
      "loss": 4.7285,
      "mean_token_accuracy": 0.247298364341259,
      "num_tokens": 48717248.0,
      "step": 21250
    },
    {
      "entropy": 5.272793436050415,
      "epoch": 2.0417867435158503,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004584846143680485,
      "loss": 4.8259,
      "mean_token_accuracy": 0.24017725735902787,
      "num_tokens": 48727948.0,
      "step": 21255
    },
    {
      "entropy": 5.283115196228027,
      "epoch": 2.042267050912584,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004584647697683723,
      "loss": 4.7261,
      "mean_token_accuracy": 0.24804541170597078,
      "num_tokens": 48740474.0,
      "step": 21260
    },
    {
      "entropy": 5.345682096481323,
      "epoch": 2.042747358309318,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045844492090918904,
      "loss": 4.7774,
      "mean_token_accuracy": 0.23484760522842407,
      "num_tokens": 48753415.0,
      "step": 21265
    },
    {
      "entropy": 5.281098413467407,
      "epoch": 2.043227665706052,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045842506779095936,
      "loss": 4.6809,
      "mean_token_accuracy": 0.2510722652077675,
      "num_tokens": 48764466.0,
      "step": 21270
    },
    {
      "entropy": 5.223915863037109,
      "epoch": 2.043707973102786,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004584052104141443,
      "loss": 4.6892,
      "mean_token_accuracy": 0.24604349285364152,
      "num_tokens": 48777513.0,
      "step": 21275
    },
    {
      "entropy": 5.3682575702667235,
      "epoch": 2.0441882804995197,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004583853487792048,
      "loss": 4.788,
      "mean_token_accuracy": 0.24231015145778656,
      "num_tokens": 48788732.0,
      "step": 21280
    },
    {
      "entropy": 5.27666163444519,
      "epoch": 2.0446685878962536,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004583654828866021,
      "loss": 4.6805,
      "mean_token_accuracy": 0.2417672336101532,
      "num_tokens": 48799620.0,
      "step": 21285
    },
    {
      "entropy": 5.217077589035034,
      "epoch": 2.0451488952929875,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004583456127367973,
      "loss": 4.7034,
      "mean_token_accuracy": 0.2435172453522682,
      "num_tokens": 48811914.0,
      "step": 21290
    },
    {
      "entropy": 5.2331499576568605,
      "epoch": 2.0456292026897214,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004583257383302519,
      "loss": 4.7083,
      "mean_token_accuracy": 0.24649409651756288,
      "num_tokens": 48822610.0,
      "step": 21295
    },
    {
      "entropy": 5.329961919784546,
      "epoch": 2.0461095100864553,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004583058596674271,
      "loss": 4.8139,
      "mean_token_accuracy": 0.23668624311685563,
      "num_tokens": 48834296.0,
      "step": 21300
    },
    {
      "entropy": 5.256770753860474,
      "epoch": 2.046589817483189,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004582859767487846,
      "loss": 4.6279,
      "mean_token_accuracy": 0.25544952899217604,
      "num_tokens": 48846455.0,
      "step": 21305
    },
    {
      "entropy": 5.34183988571167,
      "epoch": 2.047070124879923,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045826608957478604,
      "loss": 4.8316,
      "mean_token_accuracy": 0.23577709794044494,
      "num_tokens": 48858364.0,
      "step": 21310
    },
    {
      "entropy": 5.287413549423218,
      "epoch": 2.047550432276657,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045824619814589297,
      "loss": 4.7485,
      "mean_token_accuracy": 0.2447037324309349,
      "num_tokens": 48869829.0,
      "step": 21315
    },
    {
      "entropy": 5.30897798538208,
      "epoch": 2.0480307396733908,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004582263024625674,
      "loss": 4.7278,
      "mean_token_accuracy": 0.244903327524662,
      "num_tokens": 48880951.0,
      "step": 21320
    },
    {
      "entropy": 5.297849702835083,
      "epoch": 2.048511047070125,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004582064025252711,
      "loss": 4.7247,
      "mean_token_accuracy": 0.2432342067360878,
      "num_tokens": 48893367.0,
      "step": 21325
    },
    {
      "entropy": 5.324419546127319,
      "epoch": 2.048991354466859,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004581864983344661,
      "loss": 4.7556,
      "mean_token_accuracy": 0.2403846025466919,
      "num_tokens": 48904400.0,
      "step": 21330
    },
    {
      "entropy": 5.296992015838623,
      "epoch": 2.049471661863593,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004581665898906147,
      "loss": 4.7642,
      "mean_token_accuracy": 0.24044447839260102,
      "num_tokens": 48916305.0,
      "step": 21335
    },
    {
      "entropy": 5.284990310668945,
      "epoch": 2.0499519692603267,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045814667719417887,
      "loss": 4.7419,
      "mean_token_accuracy": 0.23329475820064544,
      "num_tokens": 48927160.0,
      "step": 21340
    },
    {
      "entropy": 5.3198963642120365,
      "epoch": 2.0504322766570606,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000458126760245621,
      "loss": 4.7589,
      "mean_token_accuracy": 0.23398589938879014,
      "num_tokens": 48940776.0,
      "step": 21345
    },
    {
      "entropy": 5.357657814025879,
      "epoch": 2.0509125840537945,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004581068390454036,
      "loss": 4.7697,
      "mean_token_accuracy": 0.2371277242898941,
      "num_tokens": 48952375.0,
      "step": 21350
    },
    {
      "entropy": 5.311276912689209,
      "epoch": 2.0513928914505284,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045808691359398905,
      "loss": 4.731,
      "mean_token_accuracy": 0.24392813593149185,
      "num_tokens": 48964684.0,
      "step": 21355
    },
    {
      "entropy": 5.344949579238891,
      "epoch": 2.0518731988472623,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004580669838918401,
      "loss": 4.7442,
      "mean_token_accuracy": 0.2476413667201996,
      "num_tokens": 48975654.0,
      "step": 21360
    },
    {
      "entropy": 5.291677665710449,
      "epoch": 2.052353506243996,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045804704993941935,
      "loss": 4.7378,
      "mean_token_accuracy": 0.24045889526605607,
      "num_tokens": 48987983.0,
      "step": 21365
    },
    {
      "entropy": 5.210590505599976,
      "epoch": 2.05283381364073,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045802711173718966,
      "loss": 4.6415,
      "mean_token_accuracy": 0.2549999475479126,
      "num_tokens": 48998728.0,
      "step": 21370
    },
    {
      "entropy": 5.306811285018921,
      "epoch": 2.053314121037464,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004580071692856138,
      "loss": 4.754,
      "mean_token_accuracy": 0.2423437252640724,
      "num_tokens": 49009177.0,
      "step": 21375
    },
    {
      "entropy": 5.23294095993042,
      "epoch": 2.053794428434198,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045798722258515504,
      "loss": 4.6872,
      "mean_token_accuracy": 0.24699071645736695,
      "num_tokens": 49021550.0,
      "step": 21380
    },
    {
      "entropy": 5.3400777816772464,
      "epoch": 2.0542747358309317,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045796727163627623,
      "loss": 4.7236,
      "mean_token_accuracy": 0.24428340792655945,
      "num_tokens": 49032349.0,
      "step": 21385
    },
    {
      "entropy": 5.281238651275634,
      "epoch": 2.0547550432276656,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004579473164394408,
      "loss": 4.6766,
      "mean_token_accuracy": 0.2523767903447151,
      "num_tokens": 49044829.0,
      "step": 21390
    },
    {
      "entropy": 5.294853210449219,
      "epoch": 2.0552353506243994,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045792735699511176,
      "loss": 4.7925,
      "mean_token_accuracy": 0.2368115335702896,
      "num_tokens": 49057319.0,
      "step": 21395
    },
    {
      "entropy": 5.302769660949707,
      "epoch": 2.0557156580211333,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045790739330375276,
      "loss": 4.7131,
      "mean_token_accuracy": 0.244924858212471,
      "num_tokens": 49068872.0,
      "step": 21400
    },
    {
      "entropy": 5.244687080383301,
      "epoch": 2.0561959654178676,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045788742536582717,
      "loss": 4.6604,
      "mean_token_accuracy": 0.2407780408859253,
      "num_tokens": 49081080.0,
      "step": 21405
    },
    {
      "entropy": 5.222436714172363,
      "epoch": 2.0566762728146015,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045786745318179866,
      "loss": 4.6236,
      "mean_token_accuracy": 0.2509635269641876,
      "num_tokens": 49092535.0,
      "step": 21410
    },
    {
      "entropy": 5.326153087615967,
      "epoch": 2.0571565802113354,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004578474767521309,
      "loss": 4.8105,
      "mean_token_accuracy": 0.23665284216403962,
      "num_tokens": 49103743.0,
      "step": 21415
    },
    {
      "entropy": 5.325118160247802,
      "epoch": 2.0576368876080693,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045782749607728765,
      "loss": 4.785,
      "mean_token_accuracy": 0.2444726273417473,
      "num_tokens": 49115346.0,
      "step": 21420
    },
    {
      "entropy": 5.2921552658081055,
      "epoch": 2.058117195004803,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045780751115773286,
      "loss": 4.7138,
      "mean_token_accuracy": 0.24619616866111754,
      "num_tokens": 49126000.0,
      "step": 21425
    },
    {
      "entropy": 5.259038734436035,
      "epoch": 2.058597502401537,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004577875219939304,
      "loss": 4.6501,
      "mean_token_accuracy": 0.2525392070412636,
      "num_tokens": 49136617.0,
      "step": 21430
    },
    {
      "entropy": 5.306465291976929,
      "epoch": 2.059077809798271,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004577675285863446,
      "loss": 4.7605,
      "mean_token_accuracy": 0.24181148707866668,
      "num_tokens": 49147804.0,
      "step": 21435
    },
    {
      "entropy": 5.3054028987884525,
      "epoch": 2.059558117195005,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045774753093543943,
      "loss": 4.6525,
      "mean_token_accuracy": 0.2566931486129761,
      "num_tokens": 49158367.0,
      "step": 21440
    },
    {
      "entropy": 5.3125275611877445,
      "epoch": 2.0600384245917387,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004577275290416791,
      "loss": 4.7709,
      "mean_token_accuracy": 0.24303866922855377,
      "num_tokens": 49170236.0,
      "step": 21445
    },
    {
      "entropy": 5.23494930267334,
      "epoch": 2.0605187319884726,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004577075229055283,
      "loss": 4.7226,
      "mean_token_accuracy": 0.2455672025680542,
      "num_tokens": 49183379.0,
      "step": 21450
    },
    {
      "entropy": 5.281590747833252,
      "epoch": 2.0609990393852065,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045768751252745133,
      "loss": 4.7597,
      "mean_token_accuracy": 0.23808769136667252,
      "num_tokens": 49194649.0,
      "step": 21455
    },
    {
      "entropy": 5.321300649642945,
      "epoch": 2.0614793467819403,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045766749790791274,
      "loss": 4.752,
      "mean_token_accuracy": 0.24253710508346557,
      "num_tokens": 49206523.0,
      "step": 21460
    },
    {
      "entropy": 5.3192283630371096,
      "epoch": 2.061959654178674,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004576474790473773,
      "loss": 4.7644,
      "mean_token_accuracy": 0.23562355935573578,
      "num_tokens": 49219020.0,
      "step": 21465
    },
    {
      "entropy": 5.285251808166504,
      "epoch": 2.062439961575408,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045762745594630973,
      "loss": 4.7376,
      "mean_token_accuracy": 0.24406626522541047,
      "num_tokens": 49230751.0,
      "step": 21470
    },
    {
      "entropy": 5.35581374168396,
      "epoch": 2.062920268972142,
      "grad_norm": 1.125,
      "learning_rate": 0.0004576074286051749,
      "loss": 4.7786,
      "mean_token_accuracy": 0.2404780864715576,
      "num_tokens": 49241758.0,
      "step": 21475
    },
    {
      "entropy": 5.264520311355591,
      "epoch": 2.063400576368876,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045758739702443787,
      "loss": 4.7229,
      "mean_token_accuracy": 0.24478927105665207,
      "num_tokens": 49253404.0,
      "step": 21480
    },
    {
      "entropy": 5.333926057815551,
      "epoch": 2.06388088376561,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004575673612045636,
      "loss": 4.8333,
      "mean_token_accuracy": 0.2340583920478821,
      "num_tokens": 49264000.0,
      "step": 21485
    },
    {
      "entropy": 5.3267737865448,
      "epoch": 2.064361191162344,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004575473211460173,
      "loss": 4.7806,
      "mean_token_accuracy": 0.24240600317716599,
      "num_tokens": 49275657.0,
      "step": 21490
    },
    {
      "entropy": 5.342300367355347,
      "epoch": 2.064841498559078,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004575272768492644,
      "loss": 4.8617,
      "mean_token_accuracy": 0.2333188310265541,
      "num_tokens": 49288570.0,
      "step": 21495
    },
    {
      "entropy": 5.326517915725708,
      "epoch": 2.065321805955812,
      "grad_norm": 1.0,
      "learning_rate": 0.00045750722831476993,
      "loss": 4.7631,
      "mean_token_accuracy": 0.23554647266864776,
      "num_tokens": 49300982.0,
      "step": 21500
    },
    {
      "entropy": 5.308995151519776,
      "epoch": 2.0658021133525457,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045748717554299964,
      "loss": 4.7854,
      "mean_token_accuracy": 0.24026532471179962,
      "num_tokens": 49313141.0,
      "step": 21505
    },
    {
      "entropy": 5.221657133102417,
      "epoch": 2.0662824207492796,
      "grad_norm": 1.015625,
      "learning_rate": 0.000457467118534419,
      "loss": 4.727,
      "mean_token_accuracy": 0.24476176798343657,
      "num_tokens": 49324726.0,
      "step": 21510
    },
    {
      "entropy": 5.343070030212402,
      "epoch": 2.0667627281460135,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004574470572894938,
      "loss": 4.8069,
      "mean_token_accuracy": 0.23484398871660234,
      "num_tokens": 49336187.0,
      "step": 21515
    },
    {
      "entropy": 5.380567598342895,
      "epoch": 2.0672430355427474,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004574269918086895,
      "loss": 4.755,
      "mean_token_accuracy": 0.24480170756578445,
      "num_tokens": 49348075.0,
      "step": 21520
    },
    {
      "entropy": 5.300129175186157,
      "epoch": 2.0677233429394812,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004574069220924722,
      "loss": 4.7203,
      "mean_token_accuracy": 0.24345349222421647,
      "num_tokens": 49358339.0,
      "step": 21525
    },
    {
      "entropy": 5.338204479217529,
      "epoch": 2.068203650336215,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004573868481413079,
      "loss": 4.7645,
      "mean_token_accuracy": 0.24014970064163207,
      "num_tokens": 49369650.0,
      "step": 21530
    },
    {
      "entropy": 5.309175682067871,
      "epoch": 2.068683957732949,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045736676995566244,
      "loss": 4.7817,
      "mean_token_accuracy": 0.24259265661239623,
      "num_tokens": 49381982.0,
      "step": 21535
    },
    {
      "entropy": 5.30123496055603,
      "epoch": 2.069164265129683,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045734668753600217,
      "loss": 4.7583,
      "mean_token_accuracy": 0.2383550986647606,
      "num_tokens": 49393442.0,
      "step": 21540
    },
    {
      "entropy": 5.316585493087769,
      "epoch": 2.0696445725264168,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045732660088279326,
      "loss": 4.7351,
      "mean_token_accuracy": 0.24808476418256759,
      "num_tokens": 49404311.0,
      "step": 21545
    },
    {
      "entropy": 5.382693910598755,
      "epoch": 2.0701248799231506,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045730650999650216,
      "loss": 4.8114,
      "mean_token_accuracy": 0.2402074694633484,
      "num_tokens": 49417587.0,
      "step": 21550
    },
    {
      "entropy": 5.3292749404907225,
      "epoch": 2.0706051873198845,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045728641487759506,
      "loss": 4.7856,
      "mean_token_accuracy": 0.23542287647724153,
      "num_tokens": 49429119.0,
      "step": 21555
    },
    {
      "entropy": 5.295337104797364,
      "epoch": 2.071085494716619,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004572663155265388,
      "loss": 4.7269,
      "mean_token_accuracy": 0.23720415085554122,
      "num_tokens": 49440734.0,
      "step": 21560
    },
    {
      "entropy": 5.263858079910278,
      "epoch": 2.0715658021133527,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004572462119437999,
      "loss": 4.6743,
      "mean_token_accuracy": 0.24271910786628723,
      "num_tokens": 49452830.0,
      "step": 21565
    },
    {
      "entropy": 5.304459857940674,
      "epoch": 2.0720461095100866,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045722610412984513,
      "loss": 4.7352,
      "mean_token_accuracy": 0.2327562838792801,
      "num_tokens": 49464380.0,
      "step": 21570
    },
    {
      "entropy": 5.25035605430603,
      "epoch": 2.0725264169068205,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004572059920851412,
      "loss": 4.6939,
      "mean_token_accuracy": 0.24941890686750412,
      "num_tokens": 49475436.0,
      "step": 21575
    },
    {
      "entropy": 5.2804283618927,
      "epoch": 2.0730067243035544,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045718587581015534,
      "loss": 4.749,
      "mean_token_accuracy": 0.24071856439113617,
      "num_tokens": 49487015.0,
      "step": 21580
    },
    {
      "entropy": 5.286914253234864,
      "epoch": 2.0734870317002883,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004571657553053544,
      "loss": 4.6588,
      "mean_token_accuracy": 0.24484362751245498,
      "num_tokens": 49498485.0,
      "step": 21585
    },
    {
      "entropy": 5.3448570728302,
      "epoch": 2.073967339097022,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004571456305712055,
      "loss": 4.7729,
      "mean_token_accuracy": 0.2410779967904091,
      "num_tokens": 49508899.0,
      "step": 21590
    },
    {
      "entropy": 5.233684396743774,
      "epoch": 2.074447646493756,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004571255016081759,
      "loss": 4.675,
      "mean_token_accuracy": 0.24303788989782332,
      "num_tokens": 49520956.0,
      "step": 21595
    },
    {
      "entropy": 5.3841126441955565,
      "epoch": 2.07492795389049,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000457105368416733,
      "loss": 4.8138,
      "mean_token_accuracy": 0.23349129408597946,
      "num_tokens": 49532726.0,
      "step": 21600
    },
    {
      "entropy": 5.3276495933532715,
      "epoch": 2.0754082612872238,
      "grad_norm": 1.0,
      "learning_rate": 0.00045708523099734417,
      "loss": 4.7976,
      "mean_token_accuracy": 0.2356851264834404,
      "num_tokens": 49545156.0,
      "step": 21605
    },
    {
      "entropy": 5.26768388748169,
      "epoch": 2.0758885686839577,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045706508935047693,
      "loss": 4.751,
      "mean_token_accuracy": 0.24465580731630326,
      "num_tokens": 49557552.0,
      "step": 21610
    },
    {
      "entropy": 5.254490184783935,
      "epoch": 2.0763688760806915,
      "grad_norm": 1.03125,
      "learning_rate": 0.000457044943476599,
      "loss": 4.7502,
      "mean_token_accuracy": 0.23972570151090622,
      "num_tokens": 49567906.0,
      "step": 21615
    },
    {
      "entropy": 5.287671089172363,
      "epoch": 2.0768491834774254,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045702479337617795,
      "loss": 4.6701,
      "mean_token_accuracy": 0.24972226470708847,
      "num_tokens": 49578641.0,
      "step": 21620
    },
    {
      "entropy": 5.315793657302857,
      "epoch": 2.0773294908741593,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004570046390496818,
      "loss": 4.7143,
      "mean_token_accuracy": 0.24299610108137132,
      "num_tokens": 49590449.0,
      "step": 21625
    },
    {
      "entropy": 5.311933517456055,
      "epoch": 2.077809798270893,
      "grad_norm": 1.0,
      "learning_rate": 0.0004569844804975783,
      "loss": 4.767,
      "mean_token_accuracy": 0.23858040124177932,
      "num_tokens": 49600877.0,
      "step": 21630
    },
    {
      "entropy": 5.222575378417969,
      "epoch": 2.0782901056676275,
      "grad_norm": 1.125,
      "learning_rate": 0.0004569643177203356,
      "loss": 4.7656,
      "mean_token_accuracy": 0.24054000824689864,
      "num_tokens": 49612310.0,
      "step": 21635
    },
    {
      "entropy": 5.302777671813965,
      "epoch": 2.0787704130643614,
      "grad_norm": 1.125,
      "learning_rate": 0.0004569441507184218,
      "loss": 4.7516,
      "mean_token_accuracy": 0.23073697239160537,
      "num_tokens": 49625347.0,
      "step": 21640
    },
    {
      "entropy": 5.362421894073487,
      "epoch": 2.0792507204610953,
      "grad_norm": 1.0,
      "learning_rate": 0.00045692397949230495,
      "loss": 4.8389,
      "mean_token_accuracy": 0.234532867372036,
      "num_tokens": 49637943.0,
      "step": 21645
    },
    {
      "entropy": 5.2874127388000485,
      "epoch": 2.079731027857829,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045690380404245364,
      "loss": 4.66,
      "mean_token_accuracy": 0.2552524894475937,
      "num_tokens": 49649318.0,
      "step": 21650
    },
    {
      "entropy": 5.238890027999878,
      "epoch": 2.080211335254563,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045688362436933607,
      "loss": 4.7219,
      "mean_token_accuracy": 0.24781393259763718,
      "num_tokens": 49661368.0,
      "step": 21655
    },
    {
      "entropy": 5.296549558639526,
      "epoch": 2.080691642651297,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004568634404734208,
      "loss": 4.6585,
      "mean_token_accuracy": 0.25265467166900635,
      "num_tokens": 49673566.0,
      "step": 21660
    },
    {
      "entropy": 5.248905515670776,
      "epoch": 2.081171950048031,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004568432523551765,
      "loss": 4.6973,
      "mean_token_accuracy": 0.24557511508464813,
      "num_tokens": 49685151.0,
      "step": 21665
    },
    {
      "entropy": 5.317660903930664,
      "epoch": 2.0816522574447647,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004568230600150718,
      "loss": 4.7138,
      "mean_token_accuracy": 0.2454763725399971,
      "num_tokens": 49696409.0,
      "step": 21670
    },
    {
      "entropy": 5.254036474227905,
      "epoch": 2.0821325648414986,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004568028634535757,
      "loss": 4.6908,
      "mean_token_accuracy": 0.2472606435418129,
      "num_tokens": 49709031.0,
      "step": 21675
    },
    {
      "entropy": 5.2052867889404295,
      "epoch": 2.0826128722382324,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004567826626711568,
      "loss": 4.6418,
      "mean_token_accuracy": 0.24797616004943848,
      "num_tokens": 49719646.0,
      "step": 21680
    },
    {
      "entropy": 5.249859237670899,
      "epoch": 2.0830931796349663,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004567624576682843,
      "loss": 4.6837,
      "mean_token_accuracy": 0.2504208728671074,
      "num_tokens": 49730384.0,
      "step": 21685
    },
    {
      "entropy": 5.290834426879883,
      "epoch": 2.0835734870317,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004567422484454273,
      "loss": 4.8083,
      "mean_token_accuracy": 0.2391546666622162,
      "num_tokens": 49742256.0,
      "step": 21690
    },
    {
      "entropy": 5.284143304824829,
      "epoch": 2.084053794428434,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045672203500305493,
      "loss": 4.6793,
      "mean_token_accuracy": 0.24890456795692445,
      "num_tokens": 49753547.0,
      "step": 21695
    },
    {
      "entropy": 5.329063272476196,
      "epoch": 2.084534101825168,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045670181734163654,
      "loss": 4.824,
      "mean_token_accuracy": 0.2361891970038414,
      "num_tokens": 49765472.0,
      "step": 21700
    },
    {
      "entropy": 5.23595175743103,
      "epoch": 2.085014409221902,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004566815954616414,
      "loss": 4.62,
      "mean_token_accuracy": 0.24499037861824036,
      "num_tokens": 49777031.0,
      "step": 21705
    },
    {
      "entropy": 5.299614429473877,
      "epoch": 2.0854947166186357,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045666136936353913,
      "loss": 4.7295,
      "mean_token_accuracy": 0.24597701281309128,
      "num_tokens": 49787923.0,
      "step": 21710
    },
    {
      "entropy": 5.27898383140564,
      "epoch": 2.08597502401537,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004566411390477993,
      "loss": 4.7096,
      "mean_token_accuracy": 0.24009291529655458,
      "num_tokens": 49799652.0,
      "step": 21715
    },
    {
      "entropy": 5.238503360748291,
      "epoch": 2.086455331412104,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045662090451489156,
      "loss": 4.7181,
      "mean_token_accuracy": 0.24554968476295472,
      "num_tokens": 49811278.0,
      "step": 21720
    },
    {
      "entropy": 5.3533109664917,
      "epoch": 2.086935638808838,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045660066576528577,
      "loss": 4.7439,
      "mean_token_accuracy": 0.2423337385058403,
      "num_tokens": 49823560.0,
      "step": 21725
    },
    {
      "entropy": 5.275273704528809,
      "epoch": 2.0874159462055717,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004565804227994518,
      "loss": 4.6254,
      "mean_token_accuracy": 0.2527826353907585,
      "num_tokens": 49834662.0,
      "step": 21730
    },
    {
      "entropy": 5.248792886734009,
      "epoch": 2.0878962536023056,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004565601756178595,
      "loss": 4.7103,
      "mean_token_accuracy": 0.24925799816846847,
      "num_tokens": 49845832.0,
      "step": 21735
    },
    {
      "entropy": 5.301216888427734,
      "epoch": 2.0883765609990395,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004565399242209791,
      "loss": 4.7473,
      "mean_token_accuracy": 0.24094882160425185,
      "num_tokens": 49856420.0,
      "step": 21740
    },
    {
      "entropy": 5.247446012496948,
      "epoch": 2.0888568683957733,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004565196686092807,
      "loss": 4.6222,
      "mean_token_accuracy": 0.2541125684976578,
      "num_tokens": 49867680.0,
      "step": 21745
    },
    {
      "entropy": 5.256249713897705,
      "epoch": 2.089337175792507,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004564994087832346,
      "loss": 4.689,
      "mean_token_accuracy": 0.23957178443670274,
      "num_tokens": 49879246.0,
      "step": 21750
    },
    {
      "entropy": 5.261856365203857,
      "epoch": 2.089817483189241,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045647914474331123,
      "loss": 4.6817,
      "mean_token_accuracy": 0.2503194987773895,
      "num_tokens": 49890536.0,
      "step": 21755
    },
    {
      "entropy": 5.36480746269226,
      "epoch": 2.090297790585975,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045645887648998094,
      "loss": 4.8091,
      "mean_token_accuracy": 0.2397552326321602,
      "num_tokens": 49902331.0,
      "step": 21760
    },
    {
      "entropy": 5.265902900695801,
      "epoch": 2.090778097982709,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045643860402371433,
      "loss": 4.6582,
      "mean_token_accuracy": 0.2551946505904198,
      "num_tokens": 49913814.0,
      "step": 21765
    },
    {
      "entropy": 5.262144660949707,
      "epoch": 2.0912584053794427,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004564183273449821,
      "loss": 4.6197,
      "mean_token_accuracy": 0.24861056804656984,
      "num_tokens": 49924592.0,
      "step": 21770
    },
    {
      "entropy": 5.291995096206665,
      "epoch": 2.0917387127761766,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004563980464542551,
      "loss": 4.7671,
      "mean_token_accuracy": 0.23459287583827973,
      "num_tokens": 49936526.0,
      "step": 21775
    },
    {
      "entropy": 5.31781039237976,
      "epoch": 2.0922190201729105,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045637776135200406,
      "loss": 4.7468,
      "mean_token_accuracy": 0.24748677313327788,
      "num_tokens": 49947655.0,
      "step": 21780
    },
    {
      "entropy": 5.2494996070861815,
      "epoch": 2.0926993275696444,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004563574720386999,
      "loss": 4.7428,
      "mean_token_accuracy": 0.2412853866815567,
      "num_tokens": 49959251.0,
      "step": 21785
    },
    {
      "entropy": 5.291286754608154,
      "epoch": 2.0931796349663783,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004563371785148139,
      "loss": 4.7879,
      "mean_token_accuracy": 0.23628169149160386,
      "num_tokens": 49970411.0,
      "step": 21790
    },
    {
      "entropy": 5.3288219451904295,
      "epoch": 2.0936599423631126,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045631688078081695,
      "loss": 4.8533,
      "mean_token_accuracy": 0.2359705314040184,
      "num_tokens": 49983935.0,
      "step": 21795
    },
    {
      "entropy": 5.3239562034606935,
      "epoch": 2.0941402497598465,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004562965788371805,
      "loss": 4.6938,
      "mean_token_accuracy": 0.24751280546188353,
      "num_tokens": 49995278.0,
      "step": 21800
    },
    {
      "entropy": 5.250741004943848,
      "epoch": 2.0946205571565804,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004562762726843758,
      "loss": 4.7163,
      "mean_token_accuracy": 0.24660526365041732,
      "num_tokens": 50005534.0,
      "step": 21805
    },
    {
      "entropy": 5.28363733291626,
      "epoch": 2.0951008645533142,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045625596232287436,
      "loss": 4.7209,
      "mean_token_accuracy": 0.24316229075193405,
      "num_tokens": 50016583.0,
      "step": 21810
    },
    {
      "entropy": 5.249169826507568,
      "epoch": 2.095581171950048,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004562356477531477,
      "loss": 4.6976,
      "mean_token_accuracy": 0.2338692769408226,
      "num_tokens": 50027290.0,
      "step": 21815
    },
    {
      "entropy": 5.328296661376953,
      "epoch": 2.096061479346782,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004562153289756674,
      "loss": 4.7687,
      "mean_token_accuracy": 0.24779041856527328,
      "num_tokens": 50039671.0,
      "step": 21820
    },
    {
      "entropy": 5.308818054199219,
      "epoch": 2.096541786743516,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004561950059909053,
      "loss": 4.702,
      "mean_token_accuracy": 0.24875099956989288,
      "num_tokens": 50051639.0,
      "step": 21825
    },
    {
      "entropy": 5.274341058731079,
      "epoch": 2.0970220941402498,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004561746787993332,
      "loss": 4.7815,
      "mean_token_accuracy": 0.23470915257930755,
      "num_tokens": 50064475.0,
      "step": 21830
    },
    {
      "entropy": 5.31854944229126,
      "epoch": 2.0975024015369836,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045615434740142307,
      "loss": 4.7267,
      "mean_token_accuracy": 0.24249733984470367,
      "num_tokens": 50076038.0,
      "step": 21835
    },
    {
      "entropy": 5.339339685440064,
      "epoch": 2.0979827089337175,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045613401179764686,
      "loss": 4.7196,
      "mean_token_accuracy": 0.24927317947149277,
      "num_tokens": 50086425.0,
      "step": 21840
    },
    {
      "entropy": 5.3063392162323,
      "epoch": 2.0984630163304514,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045611367198847676,
      "loss": 4.7983,
      "mean_token_accuracy": 0.24054252058267594,
      "num_tokens": 50098320.0,
      "step": 21845
    },
    {
      "entropy": 5.246003818511963,
      "epoch": 2.0989433237271853,
      "grad_norm": 1.046875,
      "learning_rate": 0.000456093327974385,
      "loss": 4.7036,
      "mean_token_accuracy": 0.2438790336251259,
      "num_tokens": 50109619.0,
      "step": 21850
    },
    {
      "entropy": 5.261871480941773,
      "epoch": 2.099423631123919,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004560729797558438,
      "loss": 4.6899,
      "mean_token_accuracy": 0.25245976746082305,
      "num_tokens": 50121043.0,
      "step": 21855
    },
    {
      "entropy": 5.3075018405914305,
      "epoch": 2.099903938520653,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004560526273333259,
      "loss": 4.7901,
      "mean_token_accuracy": 0.24278393834829332,
      "num_tokens": 50132673.0,
      "step": 21860
    },
    {
      "entropy": 5.247772836685181,
      "epoch": 2.100384245917387,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045603227070730346,
      "loss": 4.6246,
      "mean_token_accuracy": 0.24685593396425248,
      "num_tokens": 50144393.0,
      "step": 21865
    },
    {
      "entropy": 5.286378240585327,
      "epoch": 2.1008645533141213,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045601190987824933,
      "loss": 4.7081,
      "mean_token_accuracy": 0.23798996657133104,
      "num_tokens": 50156278.0,
      "step": 21870
    },
    {
      "entropy": 5.265440940856934,
      "epoch": 2.101344860710855,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045599154484663606,
      "loss": 4.7413,
      "mean_token_accuracy": 0.24623714685440062,
      "num_tokens": 50167652.0,
      "step": 21875
    },
    {
      "entropy": 5.273383712768554,
      "epoch": 2.101825168107589,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045597117561293663,
      "loss": 4.785,
      "mean_token_accuracy": 0.2476348787546158,
      "num_tokens": 50179482.0,
      "step": 21880
    },
    {
      "entropy": 5.3043347835540775,
      "epoch": 2.102305475504323,
      "grad_norm": 1.125,
      "learning_rate": 0.0004559508021776238,
      "loss": 4.7407,
      "mean_token_accuracy": 0.2462424084544182,
      "num_tokens": 50190137.0,
      "step": 21885
    },
    {
      "entropy": 5.511318016052246,
      "epoch": 2.1027857829010568,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004559304245411707,
      "loss": 4.8931,
      "mean_token_accuracy": 0.23408910036087036,
      "num_tokens": 50201853.0,
      "step": 21890
    },
    {
      "entropy": 5.302522134780884,
      "epoch": 2.1032660902977907,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045591004270405044,
      "loss": 4.7151,
      "mean_token_accuracy": 0.24507634341716766,
      "num_tokens": 50212635.0,
      "step": 21895
    },
    {
      "entropy": 5.227581596374511,
      "epoch": 2.1037463976945245,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004558896566667361,
      "loss": 4.6444,
      "mean_token_accuracy": 0.24966693222522734,
      "num_tokens": 50224842.0,
      "step": 21900
    },
    {
      "entropy": 5.331597948074341,
      "epoch": 2.1042267050912584,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045586926642970113,
      "loss": 4.784,
      "mean_token_accuracy": 0.2435631349682808,
      "num_tokens": 50236741.0,
      "step": 21905
    },
    {
      "entropy": 5.352510023117065,
      "epoch": 2.1047070124879923,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004558488719934188,
      "loss": 4.7091,
      "mean_token_accuracy": 0.2371079817414284,
      "num_tokens": 50247572.0,
      "step": 21910
    },
    {
      "entropy": 5.387078523635864,
      "epoch": 2.105187319884726,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004558284733583627,
      "loss": 4.8135,
      "mean_token_accuracy": 0.24023428857326506,
      "num_tokens": 50258566.0,
      "step": 21915
    },
    {
      "entropy": 5.307300424575805,
      "epoch": 2.10566762728146,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045580807052500645,
      "loss": 4.7293,
      "mean_token_accuracy": 0.2454454243183136,
      "num_tokens": 50270577.0,
      "step": 21920
    },
    {
      "entropy": 5.288446283340454,
      "epoch": 2.106147934678194,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004557876634938236,
      "loss": 4.7484,
      "mean_token_accuracy": 0.2415603369474411,
      "num_tokens": 50281124.0,
      "step": 21925
    },
    {
      "entropy": 5.306158399581909,
      "epoch": 2.106628242074928,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004557672522652881,
      "loss": 4.668,
      "mean_token_accuracy": 0.24641745686531066,
      "num_tokens": 50291945.0,
      "step": 21930
    },
    {
      "entropy": 5.2931924819946286,
      "epoch": 2.1071085494716617,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004557468368398738,
      "loss": 4.6887,
      "mean_token_accuracy": 0.2519678309559822,
      "num_tokens": 50303312.0,
      "step": 21935
    },
    {
      "entropy": 5.2960034847259525,
      "epoch": 2.1075888568683956,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004557264172180546,
      "loss": 4.7545,
      "mean_token_accuracy": 0.23698639124631882,
      "num_tokens": 50314791.0,
      "step": 21940
    },
    {
      "entropy": 5.355816459655761,
      "epoch": 2.1080691642651295,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004557059934003046,
      "loss": 4.7918,
      "mean_token_accuracy": 0.2393517851829529,
      "num_tokens": 50326438.0,
      "step": 21945
    },
    {
      "entropy": 5.17974271774292,
      "epoch": 2.108549471661864,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004556855653870981,
      "loss": 4.5655,
      "mean_token_accuracy": 0.2597981050610542,
      "num_tokens": 50338091.0,
      "step": 21950
    },
    {
      "entropy": 5.231098651885986,
      "epoch": 2.1090297790585977,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004556651331789092,
      "loss": 4.6592,
      "mean_token_accuracy": 0.25470990538597105,
      "num_tokens": 50350387.0,
      "step": 21955
    },
    {
      "entropy": 5.228749370574951,
      "epoch": 2.1095100864553316,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004556446967762125,
      "loss": 4.7123,
      "mean_token_accuracy": 0.246083964407444,
      "num_tokens": 50362689.0,
      "step": 21960
    },
    {
      "entropy": 5.274654531478882,
      "epoch": 2.1099903938520654,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045562425617948226,
      "loss": 4.6906,
      "mean_token_accuracy": 0.24249472320079804,
      "num_tokens": 50373290.0,
      "step": 21965
    },
    {
      "entropy": 5.302180194854737,
      "epoch": 2.1104707012487993,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045560381138919315,
      "loss": 4.7383,
      "mean_token_accuracy": 0.240946027636528,
      "num_tokens": 50385267.0,
      "step": 21970
    },
    {
      "entropy": 5.30516881942749,
      "epoch": 2.110951008645533,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045558336240581984,
      "loss": 4.7296,
      "mean_token_accuracy": 0.24979275465011597,
      "num_tokens": 50398014.0,
      "step": 21975
    },
    {
      "entropy": 5.3501019954681395,
      "epoch": 2.111431316042267,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045556290922983705,
      "loss": 4.7537,
      "mean_token_accuracy": 0.25113250613212584,
      "num_tokens": 50408661.0,
      "step": 21980
    },
    {
      "entropy": 5.3159873485565186,
      "epoch": 2.111911623439001,
      "grad_norm": 1.125,
      "learning_rate": 0.0004555424518617197,
      "loss": 4.7326,
      "mean_token_accuracy": 0.24494149386882783,
      "num_tokens": 50419231.0,
      "step": 21985
    },
    {
      "entropy": 5.28578143119812,
      "epoch": 2.112391930835735,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045552199030194274,
      "loss": 4.7273,
      "mean_token_accuracy": 0.24587590098381043,
      "num_tokens": 50430412.0,
      "step": 21990
    },
    {
      "entropy": 5.421510171890259,
      "epoch": 2.1128722382324687,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045550152455098113,
      "loss": 4.8421,
      "mean_token_accuracy": 0.23911796510219574,
      "num_tokens": 50442431.0,
      "step": 21995
    },
    {
      "entropy": 5.2996422290802006,
      "epoch": 2.1133525456292026,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004554810546093102,
      "loss": 4.746,
      "mean_token_accuracy": 0.2447154000401497,
      "num_tokens": 50454417.0,
      "step": 22000
    },
    {
      "entropy": 5.262757253646851,
      "epoch": 2.1138328530259365,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000455460580477405,
      "loss": 4.73,
      "mean_token_accuracy": 0.24632966965436937,
      "num_tokens": 50466447.0,
      "step": 22005
    },
    {
      "entropy": 5.2943267822265625,
      "epoch": 2.1143131604226704,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000455440102155741,
      "loss": 4.7165,
      "mean_token_accuracy": 0.238439904153347,
      "num_tokens": 50477843.0,
      "step": 22010
    },
    {
      "entropy": 5.3230852603912355,
      "epoch": 2.1147934678194042,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004554196196447937,
      "loss": 4.7543,
      "mean_token_accuracy": 0.24241141825914383,
      "num_tokens": 50489407.0,
      "step": 22015
    },
    {
      "entropy": 5.278720474243164,
      "epoch": 2.115273775216138,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004553991329450385,
      "loss": 4.7146,
      "mean_token_accuracy": 0.24391889125108718,
      "num_tokens": 50499492.0,
      "step": 22020
    },
    {
      "entropy": 5.238530445098877,
      "epoch": 2.115754082612872,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045537864205695116,
      "loss": 4.7644,
      "mean_token_accuracy": 0.23770290613174438,
      "num_tokens": 50511741.0,
      "step": 22025
    },
    {
      "entropy": 5.216525506973267,
      "epoch": 2.1162343900096063,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004553581469810073,
      "loss": 4.667,
      "mean_token_accuracy": 0.2444983094930649,
      "num_tokens": 50523238.0,
      "step": 22030
    },
    {
      "entropy": 5.242811965942383,
      "epoch": 2.11671469740634,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00045533764771768287,
      "loss": 4.692,
      "mean_token_accuracy": 0.24896390587091446,
      "num_tokens": 50534420.0,
      "step": 22035
    },
    {
      "entropy": 5.219944477081299,
      "epoch": 2.117195004803074,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045531714426745373,
      "loss": 4.6644,
      "mean_token_accuracy": 0.2508095756173134,
      "num_tokens": 50545785.0,
      "step": 22040
    },
    {
      "entropy": 5.341212224960327,
      "epoch": 2.117675312199808,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004552966366307959,
      "loss": 4.7439,
      "mean_token_accuracy": 0.2405800297856331,
      "num_tokens": 50556441.0,
      "step": 22045
    },
    {
      "entropy": 5.220398235321045,
      "epoch": 2.118155619596542,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004552761248081856,
      "loss": 4.6762,
      "mean_token_accuracy": 0.24853820204734803,
      "num_tokens": 50568519.0,
      "step": 22050
    },
    {
      "entropy": 5.308399963378906,
      "epoch": 2.1186359269932757,
      "grad_norm": 1.046875,
      "learning_rate": 0.000455255608800099,
      "loss": 4.7519,
      "mean_token_accuracy": 0.2459840103983879,
      "num_tokens": 50579469.0,
      "step": 22055
    },
    {
      "entropy": 5.140761137008667,
      "epoch": 2.1191162343900096,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045523508860701237,
      "loss": 4.5629,
      "mean_token_accuracy": 0.2556856721639633,
      "num_tokens": 50591405.0,
      "step": 22060
    },
    {
      "entropy": 5.280521821975708,
      "epoch": 2.1195965417867435,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004552145642294021,
      "loss": 4.7524,
      "mean_token_accuracy": 0.25324597060680387,
      "num_tokens": 50603742.0,
      "step": 22065
    },
    {
      "entropy": 5.301707124710083,
      "epoch": 2.1200768491834774,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045519403566774493,
      "loss": 4.7102,
      "mean_token_accuracy": 0.24695349335670472,
      "num_tokens": 50615753.0,
      "step": 22070
    },
    {
      "entropy": 5.275036430358886,
      "epoch": 2.1205571565802113,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004551735029225172,
      "loss": 4.641,
      "mean_token_accuracy": 0.2474494457244873,
      "num_tokens": 50626715.0,
      "step": 22075
    },
    {
      "entropy": 5.240623855590821,
      "epoch": 2.121037463976945,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045515296599419583,
      "loss": 4.6691,
      "mean_token_accuracy": 0.2471174567937851,
      "num_tokens": 50637552.0,
      "step": 22080
    },
    {
      "entropy": 5.325067329406738,
      "epoch": 2.121517771373679,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004551324248832574,
      "loss": 4.7978,
      "mean_token_accuracy": 0.23948855102062225,
      "num_tokens": 50648143.0,
      "step": 22085
    },
    {
      "entropy": 5.27941403388977,
      "epoch": 2.121998078770413,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004551118795901791,
      "loss": 4.7957,
      "mean_token_accuracy": 0.23123064935207366,
      "num_tokens": 50658878.0,
      "step": 22090
    },
    {
      "entropy": 5.266353178024292,
      "epoch": 2.122478386167147,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004550913301154376,
      "loss": 4.7248,
      "mean_token_accuracy": 0.2490468814969063,
      "num_tokens": 50670912.0,
      "step": 22095
    },
    {
      "entropy": 5.3011726379394535,
      "epoch": 2.1229586935638807,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004550707764595103,
      "loss": 4.7649,
      "mean_token_accuracy": 0.2393169194459915,
      "num_tokens": 50683030.0,
      "step": 22100
    },
    {
      "entropy": 5.395643711090088,
      "epoch": 2.123439000960615,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045505021862287434,
      "loss": 4.7934,
      "mean_token_accuracy": 0.23369942605495453,
      "num_tokens": 50694654.0,
      "step": 22105
    },
    {
      "entropy": 5.328769540786743,
      "epoch": 2.123919308357349,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045502965660600684,
      "loss": 4.8124,
      "mean_token_accuracy": 0.23730082660913468,
      "num_tokens": 50706277.0,
      "step": 22110
    },
    {
      "entropy": 5.283092737197876,
      "epoch": 2.1243996157540828,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004550090904093853,
      "loss": 4.7169,
      "mean_token_accuracy": 0.2492297649383545,
      "num_tokens": 50717610.0,
      "step": 22115
    },
    {
      "entropy": 5.365875625610352,
      "epoch": 2.1248799231508166,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004549885200334872,
      "loss": 4.8873,
      "mean_token_accuracy": 0.23427749723196029,
      "num_tokens": 50728827.0,
      "step": 22120
    },
    {
      "entropy": 5.27054271697998,
      "epoch": 2.1253602305475505,
      "grad_norm": 1.125,
      "learning_rate": 0.0004549679454787901,
      "loss": 4.693,
      "mean_token_accuracy": 0.24254559874534606,
      "num_tokens": 50740098.0,
      "step": 22125
    },
    {
      "entropy": 5.3034202575683596,
      "epoch": 2.1258405379442844,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045494736674577175,
      "loss": 4.7376,
      "mean_token_accuracy": 0.23904522955417634,
      "num_tokens": 50752166.0,
      "step": 22130
    },
    {
      "entropy": 5.309321546554566,
      "epoch": 2.1263208453410183,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004549267838349099,
      "loss": 4.7596,
      "mean_token_accuracy": 0.24484605193138123,
      "num_tokens": 50763561.0,
      "step": 22135
    },
    {
      "entropy": 5.34655442237854,
      "epoch": 2.126801152737752,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004549061967466823,
      "loss": 4.7907,
      "mean_token_accuracy": 0.24697199910879136,
      "num_tokens": 50774194.0,
      "step": 22140
    },
    {
      "entropy": 5.259202003479004,
      "epoch": 2.127281460134486,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004548856054815671,
      "loss": 4.7517,
      "mean_token_accuracy": 0.24183080196380616,
      "num_tokens": 50784873.0,
      "step": 22145
    },
    {
      "entropy": 5.347904443740845,
      "epoch": 2.12776176753122,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045486501004004225,
      "loss": 4.8218,
      "mean_token_accuracy": 0.23874318301677705,
      "num_tokens": 50795687.0,
      "step": 22150
    },
    {
      "entropy": 5.352289772033691,
      "epoch": 2.128242074927954,
      "grad_norm": 1.0,
      "learning_rate": 0.000454844410422586,
      "loss": 4.7637,
      "mean_token_accuracy": 0.24124373346567154,
      "num_tokens": 50806304.0,
      "step": 22155
    },
    {
      "entropy": 5.194091367721557,
      "epoch": 2.1287223823246877,
      "grad_norm": 1.0,
      "learning_rate": 0.00045482380662967655,
      "loss": 4.6418,
      "mean_token_accuracy": 0.2450547531247139,
      "num_tokens": 50816775.0,
      "step": 22160
    },
    {
      "entropy": 5.318414688110352,
      "epoch": 2.1292026897214216,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004548031986617923,
      "loss": 4.7867,
      "mean_token_accuracy": 0.23867136985063553,
      "num_tokens": 50828266.0,
      "step": 22165
    },
    {
      "entropy": 5.301375913619995,
      "epoch": 2.1296829971181555,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004547825865194117,
      "loss": 4.7417,
      "mean_token_accuracy": 0.24399740099906922,
      "num_tokens": 50840140.0,
      "step": 22170
    },
    {
      "entropy": 5.35536823272705,
      "epoch": 2.1301633045148893,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045476197020301323,
      "loss": 4.7612,
      "mean_token_accuracy": 0.24210839867591857,
      "num_tokens": 50850415.0,
      "step": 22175
    },
    {
      "entropy": 5.295572090148926,
      "epoch": 2.1306436119116237,
      "grad_norm": 1.125,
      "learning_rate": 0.00045474134971307554,
      "loss": 4.7196,
      "mean_token_accuracy": 0.23733197301626205,
      "num_tokens": 50861396.0,
      "step": 22180
    },
    {
      "entropy": 5.28825306892395,
      "epoch": 2.1311239193083575,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004547207250500775,
      "loss": 4.7229,
      "mean_token_accuracy": 0.24447123110294341,
      "num_tokens": 50873931.0,
      "step": 22185
    },
    {
      "entropy": 5.284252023696899,
      "epoch": 2.1316042267050914,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000454700096214498,
      "loss": 4.6982,
      "mean_token_accuracy": 0.23861466199159623,
      "num_tokens": 50884557.0,
      "step": 22190
    },
    {
      "entropy": 5.264255619049072,
      "epoch": 2.1320845341018253,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045467946320681567,
      "loss": 4.6502,
      "mean_token_accuracy": 0.2472236454486847,
      "num_tokens": 50895757.0,
      "step": 22195
    },
    {
      "entropy": 5.307019996643066,
      "epoch": 2.132564841498559,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004546588260275098,
      "loss": 4.7419,
      "mean_token_accuracy": 0.24466263204813005,
      "num_tokens": 50906437.0,
      "step": 22200
    },
    {
      "entropy": 5.242577075958252,
      "epoch": 2.133045148895293,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045463818467705955,
      "loss": 4.7216,
      "mean_token_accuracy": 0.24631764888763427,
      "num_tokens": 50916689.0,
      "step": 22205
    },
    {
      "entropy": 5.382634687423706,
      "epoch": 2.133525456292027,
      "grad_norm": 1.0625,
      "learning_rate": 0.000454617539155944,
      "loss": 4.8037,
      "mean_token_accuracy": 0.23116094172000884,
      "num_tokens": 50928266.0,
      "step": 22210
    },
    {
      "entropy": 5.379155969619751,
      "epoch": 2.134005763688761,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045459688946464255,
      "loss": 4.7849,
      "mean_token_accuracy": 0.2375516876578331,
      "num_tokens": 50940220.0,
      "step": 22215
    },
    {
      "entropy": 5.239936351776123,
      "epoch": 2.1344860710854947,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004545762356036346,
      "loss": 4.6939,
      "mean_token_accuracy": 0.25325721502304077,
      "num_tokens": 50952668.0,
      "step": 22220
    },
    {
      "entropy": 5.2632852554321286,
      "epoch": 2.1349663784822286,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004545555775733998,
      "loss": 4.7292,
      "mean_token_accuracy": 0.2458663761615753,
      "num_tokens": 50964624.0,
      "step": 22225
    },
    {
      "entropy": 5.351213502883911,
      "epoch": 2.1354466858789625,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00045453491537441747,
      "loss": 4.7839,
      "mean_token_accuracy": 0.24230860471725463,
      "num_tokens": 50976704.0,
      "step": 22230
    },
    {
      "entropy": 5.234540605545044,
      "epoch": 2.1359269932756964,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045451424900716763,
      "loss": 4.6506,
      "mean_token_accuracy": 0.24718479067087173,
      "num_tokens": 50987986.0,
      "step": 22235
    },
    {
      "entropy": 5.246134424209595,
      "epoch": 2.1364073006724302,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045449357847212994,
      "loss": 4.6972,
      "mean_token_accuracy": 0.2421519324183464,
      "num_tokens": 50999624.0,
      "step": 22240
    },
    {
      "entropy": 5.302620077133179,
      "epoch": 2.136887608069164,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004544729037697844,
      "loss": 4.7047,
      "mean_token_accuracy": 0.24612161368131638,
      "num_tokens": 51011800.0,
      "step": 22245
    },
    {
      "entropy": 5.260834980010986,
      "epoch": 2.137367915465898,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045445222490061093,
      "loss": 4.6419,
      "mean_token_accuracy": 0.2541304975748062,
      "num_tokens": 51023004.0,
      "step": 22250
    },
    {
      "entropy": 5.185668325424194,
      "epoch": 2.1378482228626323,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004544315418650897,
      "loss": 4.6054,
      "mean_token_accuracy": 0.2484297439455986,
      "num_tokens": 51033993.0,
      "step": 22255
    },
    {
      "entropy": 5.214791917800904,
      "epoch": 2.138328530259366,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004544108546637008,
      "loss": 4.6987,
      "mean_token_accuracy": 0.24965699166059493,
      "num_tokens": 51045269.0,
      "step": 22260
    },
    {
      "entropy": 5.2600038051605225,
      "epoch": 2.1388088376561,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004543901632969247,
      "loss": 4.6402,
      "mean_token_accuracy": 0.2449547529220581,
      "num_tokens": 51056044.0,
      "step": 22265
    },
    {
      "entropy": 5.287258672714233,
      "epoch": 2.139289145052834,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045436946776524157,
      "loss": 4.7022,
      "mean_token_accuracy": 0.25076422542333604,
      "num_tokens": 51067096.0,
      "step": 22270
    },
    {
      "entropy": 5.240266990661621,
      "epoch": 2.139769452449568,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045434876806913204,
      "loss": 4.7052,
      "mean_token_accuracy": 0.24672264009714126,
      "num_tokens": 51080050.0,
      "step": 22275
    },
    {
      "entropy": 5.319568300247193,
      "epoch": 2.1402497598463017,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004543280642090767,
      "loss": 4.7603,
      "mean_token_accuracy": 0.24375136345624923,
      "num_tokens": 51091333.0,
      "step": 22280
    },
    {
      "entropy": 5.31566219329834,
      "epoch": 2.1407300672430356,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004543073561855562,
      "loss": 4.7639,
      "mean_token_accuracy": 0.23998615145683289,
      "num_tokens": 51102508.0,
      "step": 22285
    },
    {
      "entropy": 5.289892530441284,
      "epoch": 2.1412103746397695,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004542866439990513,
      "loss": 4.72,
      "mean_token_accuracy": 0.24208055436611176,
      "num_tokens": 51114427.0,
      "step": 22290
    },
    {
      "entropy": 5.314411973953247,
      "epoch": 2.1416906820365034,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004542659276500429,
      "loss": 4.7435,
      "mean_token_accuracy": 0.2437494158744812,
      "num_tokens": 51126486.0,
      "step": 22295
    },
    {
      "entropy": 5.291698312759399,
      "epoch": 2.1421709894332372,
      "grad_norm": 1.0,
      "learning_rate": 0.00045424520713901204,
      "loss": 4.6763,
      "mean_token_accuracy": 0.2505164578557014,
      "num_tokens": 51136813.0,
      "step": 22300
    },
    {
      "entropy": 5.172976684570313,
      "epoch": 2.142651296829971,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004542244824664396,
      "loss": 4.5927,
      "mean_token_accuracy": 0.25356641709804534,
      "num_tokens": 51148257.0,
      "step": 22305
    },
    {
      "entropy": 5.314541053771973,
      "epoch": 2.143131604226705,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045420375363280696,
      "loss": 4.8103,
      "mean_token_accuracy": 0.23341166824102402,
      "num_tokens": 51159159.0,
      "step": 22310
    },
    {
      "entropy": 5.250764942169189,
      "epoch": 2.143611911623439,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045418302063859526,
      "loss": 4.6774,
      "mean_token_accuracy": 0.2469483107328415,
      "num_tokens": 51170900.0,
      "step": 22315
    },
    {
      "entropy": 5.364605283737182,
      "epoch": 2.1440922190201728,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045416228348428583,
      "loss": 4.8375,
      "mean_token_accuracy": 0.22842242419719697,
      "num_tokens": 51183707.0,
      "step": 22320
    },
    {
      "entropy": 5.253160619735718,
      "epoch": 2.1445725264169067,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045414154217036023,
      "loss": 4.6634,
      "mean_token_accuracy": 0.24858063459396362,
      "num_tokens": 51195114.0,
      "step": 22325
    },
    {
      "entropy": 5.245231866836548,
      "epoch": 2.1450528338136405,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045412079669730006,
      "loss": 4.6204,
      "mean_token_accuracy": 0.25539488047361375,
      "num_tokens": 51206210.0,
      "step": 22330
    },
    {
      "entropy": 5.264296388626098,
      "epoch": 2.1455331412103744,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004541000470655867,
      "loss": 4.731,
      "mean_token_accuracy": 0.24742351770401,
      "num_tokens": 51217657.0,
      "step": 22335
    },
    {
      "entropy": 5.350746154785156,
      "epoch": 2.1460134486071087,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045407929327570215,
      "loss": 4.7919,
      "mean_token_accuracy": 0.23658225536346436,
      "num_tokens": 51227836.0,
      "step": 22340
    },
    {
      "entropy": 5.296799039840698,
      "epoch": 2.1464937560038426,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004540585353281282,
      "loss": 4.7112,
      "mean_token_accuracy": 0.24919286966323853,
      "num_tokens": 51238936.0,
      "step": 22345
    },
    {
      "entropy": 5.375830745697021,
      "epoch": 2.1469740634005765,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004540377732233467,
      "loss": 4.8432,
      "mean_token_accuracy": 0.23809102475643157,
      "num_tokens": 51250696.0,
      "step": 22350
    },
    {
      "entropy": 5.325840473175049,
      "epoch": 2.1474543707973104,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004540170069618397,
      "loss": 4.7402,
      "mean_token_accuracy": 0.24359819144010544,
      "num_tokens": 51262414.0,
      "step": 22355
    },
    {
      "entropy": 5.223852491378784,
      "epoch": 2.1479346781940443,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045399623654408946,
      "loss": 4.667,
      "mean_token_accuracy": 0.24587644189596175,
      "num_tokens": 51273267.0,
      "step": 22360
    },
    {
      "entropy": 5.283436346054077,
      "epoch": 2.148414985590778,
      "grad_norm": 1.015625,
      "learning_rate": 0.000453975461970578,
      "loss": 4.7519,
      "mean_token_accuracy": 0.24452621042728423,
      "num_tokens": 51284783.0,
      "step": 22365
    },
    {
      "entropy": 5.2798819065094,
      "epoch": 2.148895292987512,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004539546832417879,
      "loss": 4.6915,
      "mean_token_accuracy": 0.24683719724416733,
      "num_tokens": 51296048.0,
      "step": 22370
    },
    {
      "entropy": 5.263624668121338,
      "epoch": 2.149375600384246,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045393390035820136,
      "loss": 4.7185,
      "mean_token_accuracy": 0.24233693778514862,
      "num_tokens": 51306731.0,
      "step": 22375
    },
    {
      "entropy": 5.320494079589844,
      "epoch": 2.14985590778098,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000453913113320301,
      "loss": 4.818,
      "mean_token_accuracy": 0.233406962454319,
      "num_tokens": 51318431.0,
      "step": 22380
    },
    {
      "entropy": 5.319278955459595,
      "epoch": 2.1503362151777137,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004538923221285694,
      "loss": 4.7882,
      "mean_token_accuracy": 0.24219376742839813,
      "num_tokens": 51330585.0,
      "step": 22385
    },
    {
      "entropy": 5.2726569175720215,
      "epoch": 2.1508165225744476,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004538715267834893,
      "loss": 4.7002,
      "mean_token_accuracy": 0.2454417899250984,
      "num_tokens": 51341936.0,
      "step": 22390
    },
    {
      "entropy": 5.325822448730468,
      "epoch": 2.1512968299711814,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004538507272855434,
      "loss": 4.7709,
      "mean_token_accuracy": 0.2447100028395653,
      "num_tokens": 51353270.0,
      "step": 22395
    },
    {
      "entropy": 5.411716651916504,
      "epoch": 2.1517771373679153,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045382992363521486,
      "loss": 4.7856,
      "mean_token_accuracy": 0.23666758537292482,
      "num_tokens": 51365470.0,
      "step": 22400
    },
    {
      "entropy": 5.2899489402771,
      "epoch": 2.152257444764649,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045380911583298633,
      "loss": 4.7498,
      "mean_token_accuracy": 0.2401596948504448,
      "num_tokens": 51375911.0,
      "step": 22405
    },
    {
      "entropy": 5.277199935913086,
      "epoch": 2.152737752161383,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045378830387934123,
      "loss": 4.8256,
      "mean_token_accuracy": 0.23822401762008666,
      "num_tokens": 51387082.0,
      "step": 22410
    },
    {
      "entropy": 5.303730821609497,
      "epoch": 2.1532180595581174,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004537674877747626,
      "loss": 4.7117,
      "mean_token_accuracy": 0.2496006891131401,
      "num_tokens": 51398826.0,
      "step": 22415
    },
    {
      "entropy": 5.304609632492065,
      "epoch": 2.1536983669548513,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045374666751973365,
      "loss": 4.7127,
      "mean_token_accuracy": 0.24158405214548112,
      "num_tokens": 51410848.0,
      "step": 22420
    },
    {
      "entropy": 5.265511512756348,
      "epoch": 2.154178674351585,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045372584311473784,
      "loss": 4.6525,
      "mean_token_accuracy": 0.24595007300376892,
      "num_tokens": 51422272.0,
      "step": 22425
    },
    {
      "entropy": 5.212894439697266,
      "epoch": 2.154658981748319,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004537050145602587,
      "loss": 4.6714,
      "mean_token_accuracy": 0.2457035645842552,
      "num_tokens": 51433139.0,
      "step": 22430
    },
    {
      "entropy": 5.262662696838379,
      "epoch": 2.155139289145053,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004536841818567798,
      "loss": 4.665,
      "mean_token_accuracy": 0.2448244422674179,
      "num_tokens": 51445582.0,
      "step": 22435
    },
    {
      "entropy": 5.283078670501709,
      "epoch": 2.155619596541787,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004536633450047847,
      "loss": 4.6309,
      "mean_token_accuracy": 0.25277168601751326,
      "num_tokens": 51456859.0,
      "step": 22440
    },
    {
      "entropy": 5.3205150127410885,
      "epoch": 2.1560999039385207,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045364250400475734,
      "loss": 4.7846,
      "mean_token_accuracy": 0.24019590616226197,
      "num_tokens": 51468939.0,
      "step": 22445
    },
    {
      "entropy": 5.263503551483154,
      "epoch": 2.1565802113352546,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004536216588571814,
      "loss": 4.7213,
      "mean_token_accuracy": 0.23908526599407195,
      "num_tokens": 51480390.0,
      "step": 22450
    },
    {
      "entropy": 5.249623012542725,
      "epoch": 2.1570605187319885,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000453600809562541,
      "loss": 4.6927,
      "mean_token_accuracy": 0.25050421953201296,
      "num_tokens": 51493194.0,
      "step": 22455
    },
    {
      "entropy": 5.261006784439087,
      "epoch": 2.1575408261287223,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004535799561213202,
      "loss": 4.7074,
      "mean_token_accuracy": 0.24548809975385666,
      "num_tokens": 51505084.0,
      "step": 22460
    },
    {
      "entropy": 5.353165960311889,
      "epoch": 2.158021133525456,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000453559098534003,
      "loss": 4.7352,
      "mean_token_accuracy": 0.24411174952983855,
      "num_tokens": 51516474.0,
      "step": 22465
    },
    {
      "entropy": 5.312988710403443,
      "epoch": 2.15850144092219,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004535382368010738,
      "loss": 4.7576,
      "mean_token_accuracy": 0.2444503515958786,
      "num_tokens": 51526734.0,
      "step": 22470
    },
    {
      "entropy": 5.262475824356079,
      "epoch": 2.158981748318924,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045351737092301676,
      "loss": 4.698,
      "mean_token_accuracy": 0.24411442130804062,
      "num_tokens": 51538230.0,
      "step": 22475
    },
    {
      "entropy": 5.303941965103149,
      "epoch": 2.159462055715658,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004534965009003165,
      "loss": 4.7047,
      "mean_token_accuracy": 0.2459572061896324,
      "num_tokens": 51551671.0,
      "step": 22480
    },
    {
      "entropy": 5.21869158744812,
      "epoch": 2.1599423631123917,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004534756267334576,
      "loss": 4.6915,
      "mean_token_accuracy": 0.24792403280735015,
      "num_tokens": 51563034.0,
      "step": 22485
    },
    {
      "entropy": 5.209864616394043,
      "epoch": 2.160422670509126,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045345474842292455,
      "loss": 4.6198,
      "mean_token_accuracy": 0.24936287999153137,
      "num_tokens": 51575210.0,
      "step": 22490
    },
    {
      "entropy": 5.3281248092651365,
      "epoch": 2.16090297790586,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004534338659692022,
      "loss": 4.7904,
      "mean_token_accuracy": 0.2376413807272911,
      "num_tokens": 51585586.0,
      "step": 22495
    },
    {
      "entropy": 5.325833320617676,
      "epoch": 2.161383285302594,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004534129793727753,
      "loss": 4.7698,
      "mean_token_accuracy": 0.24225755482912065,
      "num_tokens": 51596519.0,
      "step": 22500
    },
    {
      "entropy": 5.260570049285889,
      "epoch": 2.1618635926993277,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004533920886341288,
      "loss": 4.716,
      "mean_token_accuracy": 0.24513882845640184,
      "num_tokens": 51608421.0,
      "step": 22505
    },
    {
      "entropy": 5.277367877960205,
      "epoch": 2.1623439000960616,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004533711937537477,
      "loss": 4.7682,
      "mean_token_accuracy": 0.24081301242113112,
      "num_tokens": 51620615.0,
      "step": 22510
    },
    {
      "entropy": 5.331304121017456,
      "epoch": 2.1628242074927955,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004533502947321171,
      "loss": 4.7498,
      "mean_token_accuracy": 0.24359963834285736,
      "num_tokens": 51632797.0,
      "step": 22515
    },
    {
      "entropy": 5.297337818145752,
      "epoch": 2.1633045148895294,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004533293915697223,
      "loss": 4.7736,
      "mean_token_accuracy": 0.23633878231048583,
      "num_tokens": 51644735.0,
      "step": 22520
    },
    {
      "entropy": 5.253624963760376,
      "epoch": 2.1637848222862632,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045330848426704853,
      "loss": 4.7584,
      "mean_token_accuracy": 0.2341844245791435,
      "num_tokens": 51656270.0,
      "step": 22525
    },
    {
      "entropy": 5.252125024795532,
      "epoch": 2.164265129682997,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004532875728245813,
      "loss": 4.6776,
      "mean_token_accuracy": 0.24979602098464965,
      "num_tokens": 51668223.0,
      "step": 22530
    },
    {
      "entropy": 5.2936766147613525,
      "epoch": 2.164745437079731,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045326665724280594,
      "loss": 4.6557,
      "mean_token_accuracy": 0.24832387417554855,
      "num_tokens": 51680298.0,
      "step": 22535
    },
    {
      "entropy": 5.314579963684082,
      "epoch": 2.165225744476465,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045324573752220814,
      "loss": 4.8198,
      "mean_token_accuracy": 0.2314014807343483,
      "num_tokens": 51692803.0,
      "step": 22540
    },
    {
      "entropy": 5.28027606010437,
      "epoch": 2.1657060518731988,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045322481366327365,
      "loss": 4.7137,
      "mean_token_accuracy": 0.2399127259850502,
      "num_tokens": 51704478.0,
      "step": 22545
    },
    {
      "entropy": 5.323043727874756,
      "epoch": 2.1661863592699326,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004532038856664882,
      "loss": 4.706,
      "mean_token_accuracy": 0.25129790753126147,
      "num_tokens": 51714627.0,
      "step": 22550
    },
    {
      "entropy": 5.257886171340942,
      "epoch": 2.1666666666666665,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004531829535323376,
      "loss": 4.6851,
      "mean_token_accuracy": 0.25284804701805114,
      "num_tokens": 51726297.0,
      "step": 22555
    },
    {
      "entropy": 5.219498157501221,
      "epoch": 2.1671469740634004,
      "grad_norm": 1.046875,
      "learning_rate": 0.000453162017261308,
      "loss": 4.6925,
      "mean_token_accuracy": 0.2539189800620079,
      "num_tokens": 51737676.0,
      "step": 22560
    },
    {
      "entropy": 5.265134572982788,
      "epoch": 2.1676272814601343,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004531410768538854,
      "loss": 4.7198,
      "mean_token_accuracy": 0.24545225799083709,
      "num_tokens": 51749532.0,
      "step": 22565
    },
    {
      "entropy": 5.2764945983886715,
      "epoch": 2.168107588856868,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045312013231055596,
      "loss": 4.7256,
      "mean_token_accuracy": 0.24694864749908446,
      "num_tokens": 51761047.0,
      "step": 22570
    },
    {
      "entropy": 5.353229618072509,
      "epoch": 2.1685878962536025,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045309918363180593,
      "loss": 4.738,
      "mean_token_accuracy": 0.24682846665382385,
      "num_tokens": 51772134.0,
      "step": 22575
    },
    {
      "entropy": 5.417838144302368,
      "epoch": 2.1690682036503364,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045307823081812166,
      "loss": 4.8027,
      "mean_token_accuracy": 0.23643009960651398,
      "num_tokens": 51785775.0,
      "step": 22580
    },
    {
      "entropy": 5.247251510620117,
      "epoch": 2.1695485110470702,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045305727386998977,
      "loss": 4.6921,
      "mean_token_accuracy": 0.24238763600587845,
      "num_tokens": 51797306.0,
      "step": 22585
    },
    {
      "entropy": 5.3044782161712645,
      "epoch": 2.170028818443804,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004530363127878966,
      "loss": 4.7885,
      "mean_token_accuracy": 0.23526365906000138,
      "num_tokens": 51809312.0,
      "step": 22590
    },
    {
      "entropy": 5.314849233627319,
      "epoch": 2.170509125840538,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045301534757232885,
      "loss": 4.72,
      "mean_token_accuracy": 0.2482499361038208,
      "num_tokens": 51821646.0,
      "step": 22595
    },
    {
      "entropy": 5.283812856674194,
      "epoch": 2.170989433237272,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004529943782237735,
      "loss": 4.6402,
      "mean_token_accuracy": 0.24811933785676957,
      "num_tokens": 51831546.0,
      "step": 22600
    },
    {
      "entropy": 5.2925879001617435,
      "epoch": 2.1714697406340058,
      "grad_norm": 1.0,
      "learning_rate": 0.00045297340474271717,
      "loss": 4.7676,
      "mean_token_accuracy": 0.24219279438257219,
      "num_tokens": 51842273.0,
      "step": 22605
    },
    {
      "entropy": 5.310692691802979,
      "epoch": 2.1719500480307397,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004529524271296468,
      "loss": 4.7635,
      "mean_token_accuracy": 0.24280039817094803,
      "num_tokens": 51854715.0,
      "step": 22610
    },
    {
      "entropy": 5.293506097793579,
      "epoch": 2.1724303554274735,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045293144538504943,
      "loss": 4.7053,
      "mean_token_accuracy": 0.24559927582740784,
      "num_tokens": 51866819.0,
      "step": 22615
    },
    {
      "entropy": 5.24495496749878,
      "epoch": 2.1729106628242074,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004529104595094124,
      "loss": 4.6943,
      "mean_token_accuracy": 0.25041990727186203,
      "num_tokens": 51878651.0,
      "step": 22620
    },
    {
      "entropy": 5.318528032302856,
      "epoch": 2.1733909702209413,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045288946950322264,
      "loss": 4.7479,
      "mean_token_accuracy": 0.24603442400693892,
      "num_tokens": 51888640.0,
      "step": 22625
    },
    {
      "entropy": 5.282753992080688,
      "epoch": 2.173871277617675,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004528684753669677,
      "loss": 4.6877,
      "mean_token_accuracy": 0.25211218297481536,
      "num_tokens": 51899810.0,
      "step": 22630
    },
    {
      "entropy": 5.3040083885192875,
      "epoch": 2.174351585014409,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004528474771011349,
      "loss": 4.7646,
      "mean_token_accuracy": 0.23556546717882157,
      "num_tokens": 51910985.0,
      "step": 22635
    },
    {
      "entropy": 5.242361354827881,
      "epoch": 2.174831892411143,
      "grad_norm": 1.0,
      "learning_rate": 0.00045282647470621176,
      "loss": 4.6892,
      "mean_token_accuracy": 0.25300425440073016,
      "num_tokens": 51922550.0,
      "step": 22640
    },
    {
      "entropy": 5.261018943786621,
      "epoch": 2.175312199807877,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045280546818268595,
      "loss": 4.5936,
      "mean_token_accuracy": 0.2491880789399147,
      "num_tokens": 51934319.0,
      "step": 22645
    },
    {
      "entropy": 5.350818395614624,
      "epoch": 2.175792507204611,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004527844575310452,
      "loss": 4.8362,
      "mean_token_accuracy": 0.23323967158794404,
      "num_tokens": 51945516.0,
      "step": 22650
    },
    {
      "entropy": 5.347962141036987,
      "epoch": 2.176272814601345,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045276344275177715,
      "loss": 4.7645,
      "mean_token_accuracy": 0.24183813035488128,
      "num_tokens": 51957033.0,
      "step": 22655
    },
    {
      "entropy": 5.273760032653809,
      "epoch": 2.176753121998079,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045274242384536984,
      "loss": 4.7141,
      "mean_token_accuracy": 0.2441483825445175,
      "num_tokens": 51968211.0,
      "step": 22660
    },
    {
      "entropy": 5.212149620056152,
      "epoch": 2.177233429394813,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004527214008123113,
      "loss": 4.6566,
      "mean_token_accuracy": 0.24768031388521194,
      "num_tokens": 51978709.0,
      "step": 22665
    },
    {
      "entropy": 5.268980598449707,
      "epoch": 2.1777137367915467,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004527003736530895,
      "loss": 4.7842,
      "mean_token_accuracy": 0.23967037796974183,
      "num_tokens": 51989997.0,
      "step": 22670
    },
    {
      "entropy": 5.237946605682373,
      "epoch": 2.1781940441882806,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045267934236819265,
      "loss": 4.7068,
      "mean_token_accuracy": 0.2455658346414566,
      "num_tokens": 52001730.0,
      "step": 22675
    },
    {
      "entropy": 5.382737350463867,
      "epoch": 2.1786743515850144,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004526583069581091,
      "loss": 4.7794,
      "mean_token_accuracy": 0.23468698263168336,
      "num_tokens": 52012267.0,
      "step": 22680
    },
    {
      "entropy": 5.368169832229614,
      "epoch": 2.1791546589817483,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004526372674233272,
      "loss": 4.7869,
      "mean_token_accuracy": 0.2380916506052017,
      "num_tokens": 52024700.0,
      "step": 22685
    },
    {
      "entropy": 5.318635272979736,
      "epoch": 2.179634966378482,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045261622376433543,
      "loss": 4.72,
      "mean_token_accuracy": 0.24205984622240068,
      "num_tokens": 52036685.0,
      "step": 22690
    },
    {
      "entropy": 5.286784029006958,
      "epoch": 2.180115273775216,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045259517598162237,
      "loss": 4.7077,
      "mean_token_accuracy": 0.24811702370643615,
      "num_tokens": 52048911.0,
      "step": 22695
    },
    {
      "entropy": 5.283618402481079,
      "epoch": 2.18059558117195,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004525741240756766,
      "loss": 4.7226,
      "mean_token_accuracy": 0.24152926206588746,
      "num_tokens": 52060039.0,
      "step": 22700
    },
    {
      "entropy": 5.2762510776519775,
      "epoch": 2.181075888568684,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004525530680469871,
      "loss": 4.7091,
      "mean_token_accuracy": 0.24889936447143554,
      "num_tokens": 52070549.0,
      "step": 22705
    },
    {
      "entropy": 5.352696704864502,
      "epoch": 2.1815561959654177,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045253200789604245,
      "loss": 4.787,
      "mean_token_accuracy": 0.2337861329317093,
      "num_tokens": 52081573.0,
      "step": 22710
    },
    {
      "entropy": 5.27009596824646,
      "epoch": 2.1820365033621516,
      "grad_norm": 0.90625,
      "learning_rate": 0.00045251094362333186,
      "loss": 4.7193,
      "mean_token_accuracy": 0.2475986585021019,
      "num_tokens": 52093694.0,
      "step": 22715
    },
    {
      "entropy": 5.309299087524414,
      "epoch": 2.1825168107588855,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004524898752293441,
      "loss": 4.8216,
      "mean_token_accuracy": 0.23852483928203583,
      "num_tokens": 52104784.0,
      "step": 22720
    },
    {
      "entropy": 5.3337281227111815,
      "epoch": 2.18299711815562,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045246880271456857,
      "loss": 4.7816,
      "mean_token_accuracy": 0.23403811007738112,
      "num_tokens": 52116444.0,
      "step": 22725
    },
    {
      "entropy": 5.196913719177246,
      "epoch": 2.1834774255523537,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004524477260794944,
      "loss": 4.6757,
      "mean_token_accuracy": 0.24555414766073227,
      "num_tokens": 52130264.0,
      "step": 22730
    },
    {
      "entropy": 5.263083410263062,
      "epoch": 2.1839577329490876,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045242664532461094,
      "loss": 4.7123,
      "mean_token_accuracy": 0.23682308197021484,
      "num_tokens": 52142688.0,
      "step": 22735
    },
    {
      "entropy": 5.397024345397949,
      "epoch": 2.1844380403458215,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045240556045040767,
      "loss": 4.7941,
      "mean_token_accuracy": 0.23754342943429946,
      "num_tokens": 52154823.0,
      "step": 22740
    },
    {
      "entropy": 5.31235523223877,
      "epoch": 2.1849183477425553,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045238447145737397,
      "loss": 4.7718,
      "mean_token_accuracy": 0.24693239480257034,
      "num_tokens": 52165542.0,
      "step": 22745
    },
    {
      "entropy": 5.3412994861602785,
      "epoch": 2.185398655139289,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045236337834599966,
      "loss": 4.7512,
      "mean_token_accuracy": 0.24792954176664353,
      "num_tokens": 52176475.0,
      "step": 22750
    },
    {
      "entropy": 5.278438520431519,
      "epoch": 2.185878962536023,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045234228111677434,
      "loss": 4.6822,
      "mean_token_accuracy": 0.24250692427158355,
      "num_tokens": 52188467.0,
      "step": 22755
    },
    {
      "entropy": 5.287885761260986,
      "epoch": 2.186359269932757,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004523211797701878,
      "loss": 4.6826,
      "mean_token_accuracy": 0.24304610788822173,
      "num_tokens": 52199518.0,
      "step": 22760
    },
    {
      "entropy": 5.217509794235229,
      "epoch": 2.186839577329491,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045230007430673014,
      "loss": 4.7154,
      "mean_token_accuracy": 0.2492004245519638,
      "num_tokens": 52211429.0,
      "step": 22765
    },
    {
      "entropy": 5.387588930130005,
      "epoch": 2.1873198847262247,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004522789647268911,
      "loss": 4.7679,
      "mean_token_accuracy": 0.24505715519189836,
      "num_tokens": 52222258.0,
      "step": 22770
    },
    {
      "entropy": 5.274696111679077,
      "epoch": 2.1878001921229586,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004522578510311609,
      "loss": 4.6754,
      "mean_token_accuracy": 0.24394378960132598,
      "num_tokens": 52233660.0,
      "step": 22775
    },
    {
      "entropy": 5.267175674438477,
      "epoch": 2.1882804995196925,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045223673322002984,
      "loss": 4.6995,
      "mean_token_accuracy": 0.24548177570104598,
      "num_tokens": 52244721.0,
      "step": 22780
    },
    {
      "entropy": 5.338377523422241,
      "epoch": 2.1887608069164264,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045221561129398804,
      "loss": 4.8011,
      "mean_token_accuracy": 0.23844362199306487,
      "num_tokens": 52256099.0,
      "step": 22785
    },
    {
      "entropy": 5.310098600387573,
      "epoch": 2.1892411143131603,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000452194485253526,
      "loss": 4.7237,
      "mean_token_accuracy": 0.24879448115825653,
      "num_tokens": 52266943.0,
      "step": 22790
    },
    {
      "entropy": 5.250206232070923,
      "epoch": 2.189721421709894,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004521733550991342,
      "loss": 4.7073,
      "mean_token_accuracy": 0.2506921783089638,
      "num_tokens": 52279074.0,
      "step": 22795
    },
    {
      "entropy": 5.31021671295166,
      "epoch": 2.1902017291066285,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045215222083130316,
      "loss": 4.7443,
      "mean_token_accuracy": 0.24306874126195907,
      "num_tokens": 52290613.0,
      "step": 22800
    },
    {
      "entropy": 5.330202865600586,
      "epoch": 2.1906820365033624,
      "grad_norm": 1.125,
      "learning_rate": 0.0004521310824505236,
      "loss": 4.7517,
      "mean_token_accuracy": 0.2438317731022835,
      "num_tokens": 52301903.0,
      "step": 22805
    },
    {
      "entropy": 5.270598649978638,
      "epoch": 2.1911623439000962,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045210993995728623,
      "loss": 4.7241,
      "mean_token_accuracy": 0.24446299970149993,
      "num_tokens": 52312738.0,
      "step": 22810
    },
    {
      "entropy": 5.235638523101807,
      "epoch": 2.19164265129683,
      "grad_norm": 0.96875,
      "learning_rate": 0.000452088793352082,
      "loss": 4.6774,
      "mean_token_accuracy": 0.24593191891908645,
      "num_tokens": 52325315.0,
      "step": 22815
    },
    {
      "entropy": 5.364625120162964,
      "epoch": 2.192122958693564,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004520676426354018,
      "loss": 4.818,
      "mean_token_accuracy": 0.23842950612306596,
      "num_tokens": 52338330.0,
      "step": 22820
    },
    {
      "entropy": 5.402494716644287,
      "epoch": 2.192603266090298,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004520464878077368,
      "loss": 4.8601,
      "mean_token_accuracy": 0.22438787221908568,
      "num_tokens": 52349758.0,
      "step": 22825
    },
    {
      "entropy": 5.272943925857544,
      "epoch": 2.1930835734870318,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045202532886957805,
      "loss": 4.6844,
      "mean_token_accuracy": 0.24596937745809555,
      "num_tokens": 52361901.0,
      "step": 22830
    },
    {
      "entropy": 5.3093846321105955,
      "epoch": 2.1935638808837656,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045200416582141676,
      "loss": 4.7185,
      "mean_token_accuracy": 0.24810848087072374,
      "num_tokens": 52372912.0,
      "step": 22835
    },
    {
      "entropy": 5.240019273757935,
      "epoch": 2.1940441882804995,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004519829986637444,
      "loss": 4.7206,
      "mean_token_accuracy": 0.24513311237096785,
      "num_tokens": 52384270.0,
      "step": 22840
    },
    {
      "entropy": 5.311841011047363,
      "epoch": 2.1945244956772334,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004519618273970523,
      "loss": 4.7927,
      "mean_token_accuracy": 0.2374573603272438,
      "num_tokens": 52396555.0,
      "step": 22845
    },
    {
      "entropy": 5.293543910980224,
      "epoch": 2.1950048030739673,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045194065202183205,
      "loss": 4.6975,
      "mean_token_accuracy": 0.24753886461257935,
      "num_tokens": 52408183.0,
      "step": 22850
    },
    {
      "entropy": 5.323305559158325,
      "epoch": 2.195485110470701,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004519194725385752,
      "loss": 4.8453,
      "mean_token_accuracy": 0.24168906658887862,
      "num_tokens": 52420285.0,
      "step": 22855
    },
    {
      "entropy": 5.3261596202850345,
      "epoch": 2.195965417867435,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045189828894777364,
      "loss": 4.7402,
      "mean_token_accuracy": 0.23802688121795654,
      "num_tokens": 52431558.0,
      "step": 22860
    },
    {
      "entropy": 5.372430324554443,
      "epoch": 2.196445725264169,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045187710124991904,
      "loss": 4.8199,
      "mean_token_accuracy": 0.23555926680564881,
      "num_tokens": 52443189.0,
      "step": 22865
    },
    {
      "entropy": 5.283569955825806,
      "epoch": 2.196926032660903,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004518559094455034,
      "loss": 4.7361,
      "mean_token_accuracy": 0.2476087912917137,
      "num_tokens": 52454781.0,
      "step": 22870
    },
    {
      "entropy": 5.277330493927002,
      "epoch": 2.1974063400576367,
      "grad_norm": 1.25,
      "learning_rate": 0.0004518347135350187,
      "loss": 4.666,
      "mean_token_accuracy": 0.2527589350938797,
      "num_tokens": 52466424.0,
      "step": 22875
    },
    {
      "entropy": 5.294313716888428,
      "epoch": 2.1978866474543706,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045181351351895703,
      "loss": 4.6728,
      "mean_token_accuracy": 0.24330192506313325,
      "num_tokens": 52477043.0,
      "step": 22880
    },
    {
      "entropy": 5.199041843414307,
      "epoch": 2.198366954851105,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004517923093978106,
      "loss": 4.635,
      "mean_token_accuracy": 0.24983334690332412,
      "num_tokens": 52488321.0,
      "step": 22885
    },
    {
      "entropy": 5.299348211288452,
      "epoch": 2.1988472622478388,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004517711011720717,
      "loss": 4.7293,
      "mean_token_accuracy": 0.24487604349851608,
      "num_tokens": 52499071.0,
      "step": 22890
    },
    {
      "entropy": 5.277556705474853,
      "epoch": 2.1993275696445727,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004517498888422327,
      "loss": 4.6673,
      "mean_token_accuracy": 0.25531216710805893,
      "num_tokens": 52509846.0,
      "step": 22895
    },
    {
      "entropy": 5.3854772567749025,
      "epoch": 2.1998078770413065,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004517286724087862,
      "loss": 4.8526,
      "mean_token_accuracy": 0.22907119244337082,
      "num_tokens": 52521450.0,
      "step": 22900
    },
    {
      "entropy": 5.291443634033203,
      "epoch": 2.2002881844380404,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004517074518722247,
      "loss": 4.7236,
      "mean_token_accuracy": 0.24836100041866302,
      "num_tokens": 52533115.0,
      "step": 22905
    },
    {
      "entropy": 5.234097957611084,
      "epoch": 2.2007684918347743,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045168622723304084,
      "loss": 4.6572,
      "mean_token_accuracy": 0.2560629412531853,
      "num_tokens": 52543921.0,
      "step": 22910
    },
    {
      "entropy": 5.313203763961792,
      "epoch": 2.201248799231508,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004516649984917274,
      "loss": 4.7022,
      "mean_token_accuracy": 0.24845128804445266,
      "num_tokens": 52554956.0,
      "step": 22915
    },
    {
      "entropy": 5.239618349075317,
      "epoch": 2.201729106628242,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045164376564877734,
      "loss": 4.7012,
      "mean_token_accuracy": 0.2437540203332901,
      "num_tokens": 52567386.0,
      "step": 22920
    },
    {
      "entropy": 5.311718702316284,
      "epoch": 2.202209414024976,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045162252870468354,
      "loss": 4.7194,
      "mean_token_accuracy": 0.25032168328762056,
      "num_tokens": 52578533.0,
      "step": 22925
    },
    {
      "entropy": 5.408133792877197,
      "epoch": 2.20268972142171,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004516012876599391,
      "loss": 4.8725,
      "mean_token_accuracy": 0.23082606196403505,
      "num_tokens": 52590283.0,
      "step": 22930
    },
    {
      "entropy": 5.2207457542419435,
      "epoch": 2.2031700288184437,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045158004251503715,
      "loss": 4.6742,
      "mean_token_accuracy": 0.2527786925435066,
      "num_tokens": 52602362.0,
      "step": 22935
    },
    {
      "entropy": 5.206586790084839,
      "epoch": 2.2036503362151776,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045155879327047087,
      "loss": 4.6423,
      "mean_token_accuracy": 0.25452311187982557,
      "num_tokens": 52614057.0,
      "step": 22940
    },
    {
      "entropy": 5.279662418365478,
      "epoch": 2.2041306436119115,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004515375399267338,
      "loss": 4.7309,
      "mean_token_accuracy": 0.23893888145685196,
      "num_tokens": 52624298.0,
      "step": 22945
    },
    {
      "entropy": 5.297483491897583,
      "epoch": 2.2046109510086453,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045151628248431925,
      "loss": 4.7719,
      "mean_token_accuracy": 0.2376757025718689,
      "num_tokens": 52637043.0,
      "step": 22950
    },
    {
      "entropy": 5.376499366760254,
      "epoch": 2.2050912584053792,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045149502094372077,
      "loss": 4.816,
      "mean_token_accuracy": 0.23812714517116546,
      "num_tokens": 52649326.0,
      "step": 22955
    },
    {
      "entropy": 5.326310396194458,
      "epoch": 2.2055715658021136,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045147375530543195,
      "loss": 4.7146,
      "mean_token_accuracy": 0.24488336145877837,
      "num_tokens": 52660973.0,
      "step": 22960
    },
    {
      "entropy": 5.247778654098511,
      "epoch": 2.2060518731988474,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045145248556994653,
      "loss": 4.6725,
      "mean_token_accuracy": 0.25117991864681244,
      "num_tokens": 52671943.0,
      "step": 22965
    },
    {
      "entropy": 5.301106929779053,
      "epoch": 2.2065321805955813,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004514312117377584,
      "loss": 4.7575,
      "mean_token_accuracy": 0.24096084237098694,
      "num_tokens": 52684634.0,
      "step": 22970
    },
    {
      "entropy": 5.271383380889892,
      "epoch": 2.207012487992315,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004514099338093613,
      "loss": 4.7273,
      "mean_token_accuracy": 0.24697251468896866,
      "num_tokens": 52695681.0,
      "step": 22975
    },
    {
      "entropy": 5.242936277389527,
      "epoch": 2.207492795389049,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004513886517852496,
      "loss": 4.6633,
      "mean_token_accuracy": 0.25209134370088576,
      "num_tokens": 52706967.0,
      "step": 22980
    },
    {
      "entropy": 5.3066308975219725,
      "epoch": 2.207973102785783,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000451367365665917,
      "loss": 4.7232,
      "mean_token_accuracy": 0.2420036256313324,
      "num_tokens": 52719226.0,
      "step": 22985
    },
    {
      "entropy": 5.155986356735229,
      "epoch": 2.208453410182517,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045134607545185785,
      "loss": 4.595,
      "mean_token_accuracy": 0.2528442725539207,
      "num_tokens": 52730141.0,
      "step": 22990
    },
    {
      "entropy": 5.314439916610718,
      "epoch": 2.2089337175792507,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004513247811435666,
      "loss": 4.7936,
      "mean_token_accuracy": 0.2370090290904045,
      "num_tokens": 52741059.0,
      "step": 22995
    },
    {
      "entropy": 5.340946340560913,
      "epoch": 2.2094140249759846,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045130348274153735,
      "loss": 4.7657,
      "mean_token_accuracy": 0.24648419469594957,
      "num_tokens": 52752598.0,
      "step": 23000
    },
    {
      "entropy": 5.221481609344482,
      "epoch": 2.2098943323727185,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045128218024626486,
      "loss": 4.6691,
      "mean_token_accuracy": 0.25029301196336745,
      "num_tokens": 52764620.0,
      "step": 23005
    },
    {
      "entropy": 5.2549004554748535,
      "epoch": 2.2103746397694524,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004512608736582436,
      "loss": 4.6707,
      "mean_token_accuracy": 0.2489864006638527,
      "num_tokens": 52774756.0,
      "step": 23010
    },
    {
      "entropy": 5.349588012695312,
      "epoch": 2.2108549471661862,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004512395629779682,
      "loss": 4.8021,
      "mean_token_accuracy": 0.23652371764183044,
      "num_tokens": 52786742.0,
      "step": 23015
    },
    {
      "entropy": 5.352253675460815,
      "epoch": 2.21133525456292,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004512182482059335,
      "loss": 4.7637,
      "mean_token_accuracy": 0.23621227443218232,
      "num_tokens": 52798303.0,
      "step": 23020
    },
    {
      "entropy": 5.2934692859649655,
      "epoch": 2.211815561959654,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004511969293426343,
      "loss": 4.7212,
      "mean_token_accuracy": 0.24270084053277968,
      "num_tokens": 52808930.0,
      "step": 23025
    },
    {
      "entropy": 5.246779298782348,
      "epoch": 2.212295869356388,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045117560638856567,
      "loss": 4.6982,
      "mean_token_accuracy": 0.2428629770874977,
      "num_tokens": 52820066.0,
      "step": 23030
    },
    {
      "entropy": 5.30494441986084,
      "epoch": 2.212776176753122,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004511542793442225,
      "loss": 4.7251,
      "mean_token_accuracy": 0.24272956401109697,
      "num_tokens": 52830839.0,
      "step": 23035
    },
    {
      "entropy": 5.337722206115723,
      "epoch": 2.213256484149856,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004511329482101001,
      "loss": 4.7267,
      "mean_token_accuracy": 0.23897884786128998,
      "num_tokens": 52843112.0,
      "step": 23040
    },
    {
      "entropy": 5.315442943572998,
      "epoch": 2.21373679154659,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004511116129866936,
      "loss": 4.8085,
      "mean_token_accuracy": 0.23998062908649445,
      "num_tokens": 52854737.0,
      "step": 23045
    },
    {
      "entropy": 5.387732744216919,
      "epoch": 2.214217098943324,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045109027367449845,
      "loss": 4.7972,
      "mean_token_accuracy": 0.24105751216411592,
      "num_tokens": 52866876.0,
      "step": 23050
    },
    {
      "entropy": 5.379046630859375,
      "epoch": 2.2146974063400577,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045106893027400995,
      "loss": 4.7688,
      "mean_token_accuracy": 0.2354642257094383,
      "num_tokens": 52878731.0,
      "step": 23055
    },
    {
      "entropy": 5.329440641403198,
      "epoch": 2.2151777137367916,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045104758278572375,
      "loss": 4.7573,
      "mean_token_accuracy": 0.23865867406129837,
      "num_tokens": 52891011.0,
      "step": 23060
    },
    {
      "entropy": 5.232191801071167,
      "epoch": 2.2156580211335255,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004510262312101355,
      "loss": 4.7145,
      "mean_token_accuracy": 0.24383563697338104,
      "num_tokens": 52902334.0,
      "step": 23065
    },
    {
      "entropy": 5.411802625656128,
      "epoch": 2.2161383285302594,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004510048755477407,
      "loss": 4.8284,
      "mean_token_accuracy": 0.23642483800649644,
      "num_tokens": 52913839.0,
      "step": 23070
    },
    {
      "entropy": 5.368006086349487,
      "epoch": 2.2166186359269933,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004509835157990354,
      "loss": 4.7516,
      "mean_token_accuracy": 0.23811997026205062,
      "num_tokens": 52925588.0,
      "step": 23075
    },
    {
      "entropy": 5.4197088241577145,
      "epoch": 2.217098943323727,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00045096215196451547,
      "loss": 4.8384,
      "mean_token_accuracy": 0.23934966027736665,
      "num_tokens": 52938474.0,
      "step": 23080
    },
    {
      "entropy": 5.217487239837647,
      "epoch": 2.217579250720461,
      "grad_norm": 1.0,
      "learning_rate": 0.00045094078404467683,
      "loss": 4.658,
      "mean_token_accuracy": 0.25314382910728456,
      "num_tokens": 52950157.0,
      "step": 23085
    },
    {
      "entropy": 5.348073244094849,
      "epoch": 2.218059558117195,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045091941204001564,
      "loss": 4.8121,
      "mean_token_accuracy": 0.23515015393495559,
      "num_tokens": 52960334.0,
      "step": 23090
    },
    {
      "entropy": 5.317583703994751,
      "epoch": 2.218539865513929,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000450898035951028,
      "loss": 4.8015,
      "mean_token_accuracy": 0.24372091442346572,
      "num_tokens": 52972251.0,
      "step": 23095
    },
    {
      "entropy": 5.326356506347656,
      "epoch": 2.2190201729106627,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045087665577821034,
      "loss": 4.6993,
      "mean_token_accuracy": 0.24515654742717743,
      "num_tokens": 52983897.0,
      "step": 23100
    },
    {
      "entropy": 5.366090774536133,
      "epoch": 2.2195004803073966,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000450855271522059,
      "loss": 4.7847,
      "mean_token_accuracy": 0.23582209199666976,
      "num_tokens": 52995821.0,
      "step": 23105
    },
    {
      "entropy": 5.258014917373657,
      "epoch": 2.219980787704131,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045083388318307044,
      "loss": 4.7308,
      "mean_token_accuracy": 0.24092326909303666,
      "num_tokens": 53007890.0,
      "step": 23110
    },
    {
      "entropy": 5.302795362472534,
      "epoch": 2.2204610951008648,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004508124907617411,
      "loss": 4.7318,
      "mean_token_accuracy": 0.24515934139490128,
      "num_tokens": 53019390.0,
      "step": 23115
    },
    {
      "entropy": 5.309201908111572,
      "epoch": 2.2209414024975986,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004507910942585679,
      "loss": 4.7224,
      "mean_token_accuracy": 0.2436984211206436,
      "num_tokens": 53031318.0,
      "step": 23120
    },
    {
      "entropy": 5.239513492584228,
      "epoch": 2.2214217098943325,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004507696936740475,
      "loss": 4.6204,
      "mean_token_accuracy": 0.25454881638288496,
      "num_tokens": 53043121.0,
      "step": 23125
    },
    {
      "entropy": 5.282328748703003,
      "epoch": 2.2219020172910664,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004507482890086767,
      "loss": 4.7324,
      "mean_token_accuracy": 0.2465039998292923,
      "num_tokens": 53055600.0,
      "step": 23130
    },
    {
      "entropy": 5.2171876430511475,
      "epoch": 2.2223823246878003,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004507268802629525,
      "loss": 4.6451,
      "mean_token_accuracy": 0.25158182829618453,
      "num_tokens": 53067075.0,
      "step": 23135
    },
    {
      "entropy": 5.202173709869385,
      "epoch": 2.222862632084534,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004507054674373719,
      "loss": 4.7075,
      "mean_token_accuracy": 0.24697208404541016,
      "num_tokens": 53078374.0,
      "step": 23140
    },
    {
      "entropy": 5.287337398529052,
      "epoch": 2.223342939481268,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045068405053243216,
      "loss": 4.7149,
      "mean_token_accuracy": 0.24565015286207198,
      "num_tokens": 53089177.0,
      "step": 23145
    },
    {
      "entropy": 5.214093255996704,
      "epoch": 2.223823246878002,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004506626295486304,
      "loss": 4.5789,
      "mean_token_accuracy": 0.2531770348548889,
      "num_tokens": 53099957.0,
      "step": 23150
    },
    {
      "entropy": 5.360232162475586,
      "epoch": 2.224303554274736,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045064120448646405,
      "loss": 4.7471,
      "mean_token_accuracy": 0.24387151449918748,
      "num_tokens": 53113369.0,
      "step": 23155
    },
    {
      "entropy": 5.309368038177491,
      "epoch": 2.2247838616714697,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004506197753464304,
      "loss": 4.6984,
      "mean_token_accuracy": 0.24477151036262512,
      "num_tokens": 53124647.0,
      "step": 23160
    },
    {
      "entropy": 5.353060054779053,
      "epoch": 2.2252641690682036,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045059834212902707,
      "loss": 4.7862,
      "mean_token_accuracy": 0.23905571550130844,
      "num_tokens": 53135218.0,
      "step": 23165
    },
    {
      "entropy": 5.259757614135742,
      "epoch": 2.2257444764649374,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045057690483475167,
      "loss": 4.7203,
      "mean_token_accuracy": 0.2410741075873375,
      "num_tokens": 53148077.0,
      "step": 23170
    },
    {
      "entropy": 5.336408901214599,
      "epoch": 2.2262247838616713,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004505554634641019,
      "loss": 4.7557,
      "mean_token_accuracy": 0.2440165549516678,
      "num_tokens": 53159866.0,
      "step": 23175
    },
    {
      "entropy": 5.278645753860474,
      "epoch": 2.226705091258405,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045053401801757554,
      "loss": 4.7059,
      "mean_token_accuracy": 0.24735113233327866,
      "num_tokens": 53171151.0,
      "step": 23180
    },
    {
      "entropy": 5.301444625854492,
      "epoch": 2.227185398655139,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045051256849567054,
      "loss": 4.7442,
      "mean_token_accuracy": 0.24075358510017394,
      "num_tokens": 53182400.0,
      "step": 23185
    },
    {
      "entropy": 5.34704327583313,
      "epoch": 2.227665706051873,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045049111489888486,
      "loss": 4.8391,
      "mean_token_accuracy": 0.23457176238298416,
      "num_tokens": 53194248.0,
      "step": 23190
    },
    {
      "entropy": 5.306047677993774,
      "epoch": 2.2281460134486073,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004504696572277165,
      "loss": 4.7419,
      "mean_token_accuracy": 0.24555314481258392,
      "num_tokens": 53205484.0,
      "step": 23195
    },
    {
      "entropy": 5.256046390533447,
      "epoch": 2.228626320845341,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045044819548266385,
      "loss": 4.7099,
      "mean_token_accuracy": 0.24691611528396606,
      "num_tokens": 53217070.0,
      "step": 23200
    },
    {
      "entropy": 5.2438372611999515,
      "epoch": 2.229106628242075,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045042672966422506,
      "loss": 4.7036,
      "mean_token_accuracy": 0.24926642030477525,
      "num_tokens": 53228431.0,
      "step": 23205
    },
    {
      "entropy": 5.3747255325317385,
      "epoch": 2.229586935638809,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045040525977289847,
      "loss": 4.8342,
      "mean_token_accuracy": 0.24039165377616883,
      "num_tokens": 53240718.0,
      "step": 23210
    },
    {
      "entropy": 5.408140087127686,
      "epoch": 2.230067243035543,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004503837858091826,
      "loss": 4.7669,
      "mean_token_accuracy": 0.2458103120326996,
      "num_tokens": 53251542.0,
      "step": 23215
    },
    {
      "entropy": 5.275371789932251,
      "epoch": 2.2305475504322767,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045036230777357604,
      "loss": 4.6986,
      "mean_token_accuracy": 0.25014342814683915,
      "num_tokens": 53261915.0,
      "step": 23220
    },
    {
      "entropy": 5.315931797027588,
      "epoch": 2.2310278578290106,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004503408256665774,
      "loss": 4.7782,
      "mean_token_accuracy": 0.23987528383731843,
      "num_tokens": 53273287.0,
      "step": 23225
    },
    {
      "entropy": 5.317750740051269,
      "epoch": 2.2315081652257445,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045031933948868545,
      "loss": 4.797,
      "mean_token_accuracy": 0.2345612183213234,
      "num_tokens": 53284482.0,
      "step": 23230
    },
    {
      "entropy": 5.2577800273895265,
      "epoch": 2.2319884726224783,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00045029784924039903,
      "loss": 4.6782,
      "mean_token_accuracy": 0.2455463856458664,
      "num_tokens": 53296235.0,
      "step": 23235
    },
    {
      "entropy": 5.447472000122071,
      "epoch": 2.2324687800192122,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045027635492221716,
      "loss": 4.8346,
      "mean_token_accuracy": 0.2321691706776619,
      "num_tokens": 53306784.0,
      "step": 23240
    },
    {
      "entropy": 5.3280942916870115,
      "epoch": 2.232949087415946,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045025485653463866,
      "loss": 4.7145,
      "mean_token_accuracy": 0.24553632438182832,
      "num_tokens": 53318303.0,
      "step": 23245
    },
    {
      "entropy": 5.275581121444702,
      "epoch": 2.23342939481268,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004502333540781628,
      "loss": 4.7043,
      "mean_token_accuracy": 0.2508359238505363,
      "num_tokens": 53330196.0,
      "step": 23250
    },
    {
      "entropy": 5.291199779510498,
      "epoch": 2.233909702209414,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004502118475532888,
      "loss": 4.7153,
      "mean_token_accuracy": 0.24340131580829621,
      "num_tokens": 53342250.0,
      "step": 23255
    },
    {
      "entropy": 5.2425004005432125,
      "epoch": 2.2343900096061478,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000450190336960516,
      "loss": 4.6859,
      "mean_token_accuracy": 0.23767983466386794,
      "num_tokens": 53353050.0,
      "step": 23260
    },
    {
      "entropy": 5.289824342727661,
      "epoch": 2.2348703170028816,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004501688223003438,
      "loss": 4.6786,
      "mean_token_accuracy": 0.2536478891968727,
      "num_tokens": 53364702.0,
      "step": 23265
    },
    {
      "entropy": 5.294997835159302,
      "epoch": 2.235350624399616,
      "grad_norm": 1.125,
      "learning_rate": 0.0004501473035732717,
      "loss": 4.7139,
      "mean_token_accuracy": 0.23794875890016556,
      "num_tokens": 53375790.0,
      "step": 23270
    },
    {
      "entropy": 5.33405556678772,
      "epoch": 2.23583093179635,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004501257807797993,
      "loss": 4.7324,
      "mean_token_accuracy": 0.2435537561774254,
      "num_tokens": 53387206.0,
      "step": 23275
    },
    {
      "entropy": 5.276723432540893,
      "epoch": 2.2363112391930837,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045010425392042624,
      "loss": 4.7482,
      "mean_token_accuracy": 0.24335831105709077,
      "num_tokens": 53399519.0,
      "step": 23280
    },
    {
      "entropy": 5.279264640808106,
      "epoch": 2.2367915465898176,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004500827229956524,
      "loss": 4.6976,
      "mean_token_accuracy": 0.24617983251810074,
      "num_tokens": 53411187.0,
      "step": 23285
    },
    {
      "entropy": 5.377202653884888,
      "epoch": 2.2372718539865515,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045006118800597757,
      "loss": 4.7848,
      "mean_token_accuracy": 0.2386761799454689,
      "num_tokens": 53421328.0,
      "step": 23290
    },
    {
      "entropy": 5.263536214828491,
      "epoch": 2.2377521613832854,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045003964895190177,
      "loss": 4.7745,
      "mean_token_accuracy": 0.23908315896987914,
      "num_tokens": 53433155.0,
      "step": 23295
    },
    {
      "entropy": 5.331100988388061,
      "epoch": 2.2382324687800192,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004500181058339251,
      "loss": 4.7253,
      "mean_token_accuracy": 0.24186987727880477,
      "num_tokens": 53445498.0,
      "step": 23300
    },
    {
      "entropy": 5.311475229263306,
      "epoch": 2.238712776176753,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004499965586525478,
      "loss": 4.7127,
      "mean_token_accuracy": 0.24506797045469284,
      "num_tokens": 53458005.0,
      "step": 23305
    },
    {
      "entropy": 5.317197847366333,
      "epoch": 2.239193083573487,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044997500740826993,
      "loss": 4.784,
      "mean_token_accuracy": 0.2438283920288086,
      "num_tokens": 53470052.0,
      "step": 23310
    },
    {
      "entropy": 5.276613426208496,
      "epoch": 2.239673390970221,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000449953452101592,
      "loss": 4.7412,
      "mean_token_accuracy": 0.2500900238752365,
      "num_tokens": 53480665.0,
      "step": 23315
    },
    {
      "entropy": 5.316168737411499,
      "epoch": 2.2401536983669548,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044993189273301445,
      "loss": 4.7118,
      "mean_token_accuracy": 0.2441806450486183,
      "num_tokens": 53491903.0,
      "step": 23320
    },
    {
      "entropy": 5.271732950210572,
      "epoch": 2.2406340057636887,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004499103293030377,
      "loss": 4.7259,
      "mean_token_accuracy": 0.24758655875921248,
      "num_tokens": 53502806.0,
      "step": 23325
    },
    {
      "entropy": 5.285796737670898,
      "epoch": 2.2411143131604225,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004498887618121625,
      "loss": 4.7263,
      "mean_token_accuracy": 0.23986750841140747,
      "num_tokens": 53513738.0,
      "step": 23330
    },
    {
      "entropy": 5.255324077606201,
      "epoch": 2.2415946205571564,
      "grad_norm": 1.125,
      "learning_rate": 0.0004498671902608897,
      "loss": 4.7059,
      "mean_token_accuracy": 0.24605665653944014,
      "num_tokens": 53525546.0,
      "step": 23335
    },
    {
      "entropy": 5.2924541473388675,
      "epoch": 2.2420749279538903,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004498456146497198,
      "loss": 4.708,
      "mean_token_accuracy": 0.24686607718467712,
      "num_tokens": 53537162.0,
      "step": 23340
    },
    {
      "entropy": 5.301639461517334,
      "epoch": 2.2425552353506246,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044982403497915405,
      "loss": 4.747,
      "mean_token_accuracy": 0.2503662511706352,
      "num_tokens": 53549165.0,
      "step": 23345
    },
    {
      "entropy": 5.297987651824951,
      "epoch": 2.2430355427473585,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044980245124969333,
      "loss": 4.7161,
      "mean_token_accuracy": 0.24856470227241517,
      "num_tokens": 53561641.0,
      "step": 23350
    },
    {
      "entropy": 5.287984085083008,
      "epoch": 2.2435158501440924,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004497808634618386,
      "loss": 4.7055,
      "mean_token_accuracy": 0.24981431066989898,
      "num_tokens": 53572350.0,
      "step": 23355
    },
    {
      "entropy": 5.306265592575073,
      "epoch": 2.2439961575408263,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004497592716160913,
      "loss": 4.812,
      "mean_token_accuracy": 0.23519349247217178,
      "num_tokens": 53584191.0,
      "step": 23360
    },
    {
      "entropy": 5.364340448379517,
      "epoch": 2.24447646493756,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00044973767571295273,
      "loss": 4.7892,
      "mean_token_accuracy": 0.2361203745007515,
      "num_tokens": 53595021.0,
      "step": 23365
    },
    {
      "entropy": 5.397205018997193,
      "epoch": 2.244956772334294,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004497160757529241,
      "loss": 4.7359,
      "mean_token_accuracy": 0.2551219239830971,
      "num_tokens": 53606992.0,
      "step": 23370
    },
    {
      "entropy": 5.299884605407715,
      "epoch": 2.245437079731028,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044969447173650695,
      "loss": 4.6827,
      "mean_token_accuracy": 0.2527642548084259,
      "num_tokens": 53616548.0,
      "step": 23375
    },
    {
      "entropy": 5.264618825912476,
      "epoch": 2.245917387127762,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000449672863664203,
      "loss": 4.6912,
      "mean_token_accuracy": 0.25293937474489214,
      "num_tokens": 53629078.0,
      "step": 23380
    },
    {
      "entropy": 5.292096090316773,
      "epoch": 2.2463976945244957,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044965125153651375,
      "loss": 4.6382,
      "mean_token_accuracy": 0.26025075763463973,
      "num_tokens": 53640059.0,
      "step": 23385
    },
    {
      "entropy": 5.320336771011353,
      "epoch": 2.2468780019212296,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004496296353539411,
      "loss": 4.8426,
      "mean_token_accuracy": 0.2329700455069542,
      "num_tokens": 53651806.0,
      "step": 23390
    },
    {
      "entropy": 5.339899015426636,
      "epoch": 2.2473583093179634,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004496080151169869,
      "loss": 4.8059,
      "mean_token_accuracy": 0.2346579387784004,
      "num_tokens": 53662180.0,
      "step": 23395
    },
    {
      "entropy": 5.304065179824829,
      "epoch": 2.2478386167146973,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044958639082615294,
      "loss": 4.8194,
      "mean_token_accuracy": 0.23745980262756347,
      "num_tokens": 53674520.0,
      "step": 23400
    },
    {
      "entropy": 5.427558183670044,
      "epoch": 2.248318924111431,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004495647624819415,
      "loss": 4.8622,
      "mean_token_accuracy": 0.23387911319732665,
      "num_tokens": 53686051.0,
      "step": 23405
    },
    {
      "entropy": 5.318250370025635,
      "epoch": 2.248799231508165,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044954313008485457,
      "loss": 4.6049,
      "mean_token_accuracy": 0.2579593613743782,
      "num_tokens": 53697171.0,
      "step": 23410
    },
    {
      "entropy": 5.312553644180298,
      "epoch": 2.249279538904899,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044952149363539453,
      "loss": 4.7415,
      "mean_token_accuracy": 0.24664939492940902,
      "num_tokens": 53707776.0,
      "step": 23415
    },
    {
      "entropy": 5.332644510269165,
      "epoch": 2.249759846301633,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004494998531340635,
      "loss": 4.7873,
      "mean_token_accuracy": 0.23493693470954896,
      "num_tokens": 53719430.0,
      "step": 23420
    },
    {
      "entropy": 5.310572147369385,
      "epoch": 2.2502401536983667,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004494782085813641,
      "loss": 4.6623,
      "mean_token_accuracy": 0.25016679912805556,
      "num_tokens": 53730708.0,
      "step": 23425
    },
    {
      "entropy": 5.288370561599732,
      "epoch": 2.250720461095101,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004494565599777988,
      "loss": 4.7349,
      "mean_token_accuracy": 0.2415642261505127,
      "num_tokens": 53741623.0,
      "step": 23430
    },
    {
      "entropy": 5.25747013092041,
      "epoch": 2.251200768491835,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044943490732387025,
      "loss": 4.6603,
      "mean_token_accuracy": 0.2523209661245346,
      "num_tokens": 53752348.0,
      "step": 23435
    },
    {
      "entropy": 5.2717619895935055,
      "epoch": 2.251681075888569,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004494132506200811,
      "loss": 4.7007,
      "mean_token_accuracy": 0.24620282053947448,
      "num_tokens": 53763973.0,
      "step": 23440
    },
    {
      "entropy": 5.218607330322266,
      "epoch": 2.2521613832853027,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004493915898669341,
      "loss": 4.5862,
      "mean_token_accuracy": 0.2531050980091095,
      "num_tokens": 53775139.0,
      "step": 23445
    },
    {
      "entropy": 5.3183520317077635,
      "epoch": 2.2526416906820366,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004493699250649323,
      "loss": 4.7713,
      "mean_token_accuracy": 0.24434444308280945,
      "num_tokens": 53788324.0,
      "step": 23450
    },
    {
      "entropy": 5.271003580093383,
      "epoch": 2.2531219980787704,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004493482562145786,
      "loss": 4.6643,
      "mean_token_accuracy": 0.24646642804145813,
      "num_tokens": 53800019.0,
      "step": 23455
    },
    {
      "entropy": 5.319000864028931,
      "epoch": 2.2536023054755043,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044932658331637605,
      "loss": 4.77,
      "mean_token_accuracy": 0.23903196156024933,
      "num_tokens": 53812155.0,
      "step": 23460
    },
    {
      "entropy": 5.269713497161865,
      "epoch": 2.254082612872238,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004493049063708279,
      "loss": 4.7007,
      "mean_token_accuracy": 0.2512423679232597,
      "num_tokens": 53823770.0,
      "step": 23465
    },
    {
      "entropy": 5.375194644927978,
      "epoch": 2.254562920268972,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044928322537843746,
      "loss": 4.8934,
      "mean_token_accuracy": 0.23239874839782715,
      "num_tokens": 53836037.0,
      "step": 23470
    },
    {
      "entropy": 5.26633038520813,
      "epoch": 2.255043227665706,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044926154033970793,
      "loss": 4.6193,
      "mean_token_accuracy": 0.2457193523645401,
      "num_tokens": 53848517.0,
      "step": 23475
    },
    {
      "entropy": 5.249071836471558,
      "epoch": 2.25552353506244,
      "grad_norm": 1.078125,
      "learning_rate": 0.000449239851255143,
      "loss": 4.6834,
      "mean_token_accuracy": 0.2534642741084099,
      "num_tokens": 53860143.0,
      "step": 23480
    },
    {
      "entropy": 5.355127286911011,
      "epoch": 2.2560038424591737,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044921815812524606,
      "loss": 4.7918,
      "mean_token_accuracy": 0.2440176248550415,
      "num_tokens": 53871173.0,
      "step": 23485
    },
    {
      "entropy": 5.302151727676391,
      "epoch": 2.2564841498559076,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044919646095052077,
      "loss": 4.7194,
      "mean_token_accuracy": 0.24665238708257675,
      "num_tokens": 53882791.0,
      "step": 23490
    },
    {
      "entropy": 5.298785305023193,
      "epoch": 2.2569644572526415,
      "grad_norm": 1.03125,
      "learning_rate": 0.000449174759731471,
      "loss": 4.7403,
      "mean_token_accuracy": 0.2477159470319748,
      "num_tokens": 53894785.0,
      "step": 23495
    },
    {
      "entropy": 5.33782262802124,
      "epoch": 2.2574447646493754,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044915305446860046,
      "loss": 4.7579,
      "mean_token_accuracy": 0.24116982370615006,
      "num_tokens": 53905989.0,
      "step": 23500
    },
    {
      "entropy": 5.302945566177368,
      "epoch": 2.2579250720461097,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044913134516241305,
      "loss": 4.6605,
      "mean_token_accuracy": 0.24165676534175873,
      "num_tokens": 53918233.0,
      "step": 23505
    },
    {
      "entropy": 5.269454622268677,
      "epoch": 2.2584053794428436,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004491096318134129,
      "loss": 4.7404,
      "mean_token_accuracy": 0.2393736183643341,
      "num_tokens": 53929886.0,
      "step": 23510
    },
    {
      "entropy": 5.387959289550781,
      "epoch": 2.2588856868395775,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004490879144221042,
      "loss": 4.9049,
      "mean_token_accuracy": 0.22529622167348862,
      "num_tokens": 53941788.0,
      "step": 23515
    },
    {
      "entropy": 5.344743251800537,
      "epoch": 2.2593659942363113,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044906619298899097,
      "loss": 4.7007,
      "mean_token_accuracy": 0.24699675887823105,
      "num_tokens": 53952753.0,
      "step": 23520
    },
    {
      "entropy": 5.257972478866577,
      "epoch": 2.2598463016330452,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004490444675145777,
      "loss": 4.7243,
      "mean_token_accuracy": 0.2433695450425148,
      "num_tokens": 53964435.0,
      "step": 23525
    },
    {
      "entropy": 5.296572208404541,
      "epoch": 2.260326609029779,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004490227379993686,
      "loss": 4.7683,
      "mean_token_accuracy": 0.24431150406599045,
      "num_tokens": 53976906.0,
      "step": 23530
    },
    {
      "entropy": 5.347026824951172,
      "epoch": 2.260806916426513,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004490010044438683,
      "loss": 4.7205,
      "mean_token_accuracy": 0.2482174053788185,
      "num_tokens": 53988076.0,
      "step": 23535
    },
    {
      "entropy": 5.196344709396362,
      "epoch": 2.261287223823247,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044897926684858133,
      "loss": 4.6232,
      "mean_token_accuracy": 0.25222315788269045,
      "num_tokens": 53998896.0,
      "step": 23540
    },
    {
      "entropy": 5.186874961853027,
      "epoch": 2.2617675312199808,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044895752521401246,
      "loss": 4.6977,
      "mean_token_accuracy": 0.2477954313158989,
      "num_tokens": 54010855.0,
      "step": 23545
    },
    {
      "entropy": 5.328880786895752,
      "epoch": 2.2622478386167146,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004489357795406663,
      "loss": 4.7352,
      "mean_token_accuracy": 0.25101886242628096,
      "num_tokens": 54021630.0,
      "step": 23550
    },
    {
      "entropy": 5.232352161407471,
      "epoch": 2.2627281460134485,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004489140298290479,
      "loss": 4.659,
      "mean_token_accuracy": 0.24886443167924882,
      "num_tokens": 54033715.0,
      "step": 23555
    },
    {
      "entropy": 5.22718620300293,
      "epoch": 2.2632084534101824,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044889227607966217,
      "loss": 4.7677,
      "mean_token_accuracy": 0.23796481639146805,
      "num_tokens": 54046633.0,
      "step": 23560
    },
    {
      "entropy": 5.307505702972412,
      "epoch": 2.2636887608069163,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044887051829301406,
      "loss": 4.7207,
      "mean_token_accuracy": 0.243051441013813,
      "num_tokens": 54058389.0,
      "step": 23565
    },
    {
      "entropy": 5.3533624649047855,
      "epoch": 2.26416906820365,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044884875646960886,
      "loss": 4.8039,
      "mean_token_accuracy": 0.24171649813652038,
      "num_tokens": 54070338.0,
      "step": 23570
    },
    {
      "entropy": 5.2887989521026615,
      "epoch": 2.264649375600384,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044882699060995175,
      "loss": 4.6848,
      "mean_token_accuracy": 0.24668814688920976,
      "num_tokens": 54082511.0,
      "step": 23575
    },
    {
      "entropy": 5.258115339279175,
      "epoch": 2.2651296829971184,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004488052207145481,
      "loss": 4.6822,
      "mean_token_accuracy": 0.24922045618295668,
      "num_tokens": 54094141.0,
      "step": 23580
    },
    {
      "entropy": 5.304886102676392,
      "epoch": 2.2656099903938522,
      "grad_norm": 1.46875,
      "learning_rate": 0.00044878344678390324,
      "loss": 4.7874,
      "mean_token_accuracy": 0.24073251187801362,
      "num_tokens": 54104837.0,
      "step": 23585
    },
    {
      "entropy": 5.408927345275879,
      "epoch": 2.266090297790586,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044876166881852286,
      "loss": 4.7853,
      "mean_token_accuracy": 0.23102782368659974,
      "num_tokens": 54116924.0,
      "step": 23590
    },
    {
      "entropy": 5.292192602157593,
      "epoch": 2.26657060518732,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004487398868189125,
      "loss": 4.7255,
      "mean_token_accuracy": 0.24867661595344542,
      "num_tokens": 54127187.0,
      "step": 23595
    },
    {
      "entropy": 5.279629421234131,
      "epoch": 2.267050912584054,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044871810078557777,
      "loss": 4.723,
      "mean_token_accuracy": 0.24722511172294617,
      "num_tokens": 54139024.0,
      "step": 23600
    },
    {
      "entropy": 5.2939409732818605,
      "epoch": 2.2675312199807878,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004486963107190247,
      "loss": 4.801,
      "mean_token_accuracy": 0.23807460069656372,
      "num_tokens": 54150643.0,
      "step": 23605
    },
    {
      "entropy": 5.31705002784729,
      "epoch": 2.2680115273775217,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044867451661975894,
      "loss": 4.7212,
      "mean_token_accuracy": 0.2463079571723938,
      "num_tokens": 54162171.0,
      "step": 23610
    },
    {
      "entropy": 5.322376251220703,
      "epoch": 2.2684918347742555,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044865271848828673,
      "loss": 4.7354,
      "mean_token_accuracy": 0.24130878299474717,
      "num_tokens": 54173104.0,
      "step": 23615
    },
    {
      "entropy": 5.277387046813965,
      "epoch": 2.2689721421709894,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000448630916325114,
      "loss": 4.7235,
      "mean_token_accuracy": 0.24356458932161332,
      "num_tokens": 54183956.0,
      "step": 23620
    },
    {
      "entropy": 5.3291465759277346,
      "epoch": 2.2694524495677233,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000448609110130747,
      "loss": 4.7658,
      "mean_token_accuracy": 0.24457595497369766,
      "num_tokens": 54195087.0,
      "step": 23625
    },
    {
      "entropy": 5.250939130783081,
      "epoch": 2.269932756964457,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044858729990569193,
      "loss": 4.6826,
      "mean_token_accuracy": 0.2500224307179451,
      "num_tokens": 54206684.0,
      "step": 23630
    },
    {
      "entropy": 5.242085027694702,
      "epoch": 2.270413064361191,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044856548565045523,
      "loss": 4.7118,
      "mean_token_accuracy": 0.23984963446855545,
      "num_tokens": 54218155.0,
      "step": 23635
    },
    {
      "entropy": 5.354400587081909,
      "epoch": 2.270893371757925,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044854366736554323,
      "loss": 4.7722,
      "mean_token_accuracy": 0.2354020357131958,
      "num_tokens": 54229124.0,
      "step": 23640
    },
    {
      "entropy": 5.289257907867432,
      "epoch": 2.271373679154659,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044852184505146274,
      "loss": 4.7139,
      "mean_token_accuracy": 0.24935157150030135,
      "num_tokens": 54240889.0,
      "step": 23645
    },
    {
      "entropy": 5.2968395233154295,
      "epoch": 2.2718539865513927,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004485000187087202,
      "loss": 4.6758,
      "mean_token_accuracy": 0.2501367077231407,
      "num_tokens": 54252388.0,
      "step": 23650
    },
    {
      "entropy": 5.289164352416992,
      "epoch": 2.272334293948127,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004484781883378224,
      "loss": 4.7119,
      "mean_token_accuracy": 0.2454767107963562,
      "num_tokens": 54264702.0,
      "step": 23655
    },
    {
      "entropy": 5.30821795463562,
      "epoch": 2.2728146013448605,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044845635393927623,
      "loss": 4.7782,
      "mean_token_accuracy": 0.24056392759084702,
      "num_tokens": 54277064.0,
      "step": 23660
    },
    {
      "entropy": 5.249357843399048,
      "epoch": 2.273294908741595,
      "grad_norm": 1.0,
      "learning_rate": 0.0004484345155135886,
      "loss": 4.6966,
      "mean_token_accuracy": 0.24782235473394393,
      "num_tokens": 54287958.0,
      "step": 23665
    },
    {
      "entropy": 5.252449417114258,
      "epoch": 2.2737752161383287,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004484126730612664,
      "loss": 4.6949,
      "mean_token_accuracy": 0.2475246384739876,
      "num_tokens": 54299428.0,
      "step": 23670
    },
    {
      "entropy": 5.335153818130493,
      "epoch": 2.2742555235350626,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000448390826582817,
      "loss": 4.7726,
      "mean_token_accuracy": 0.2398653119802475,
      "num_tokens": 54312249.0,
      "step": 23675
    },
    {
      "entropy": 5.418244075775147,
      "epoch": 2.2747358309317964,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044836897607874744,
      "loss": 4.8354,
      "mean_token_accuracy": 0.23584618866443635,
      "num_tokens": 54324408.0,
      "step": 23680
    },
    {
      "entropy": 5.3501660346984865,
      "epoch": 2.2752161383285303,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000448347121549565,
      "loss": 4.7876,
      "mean_token_accuracy": 0.23711136728525162,
      "num_tokens": 54336092.0,
      "step": 23685
    },
    {
      "entropy": 5.246084499359131,
      "epoch": 2.275696445725264,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004483252629957771,
      "loss": 4.6544,
      "mean_token_accuracy": 0.25517176687717436,
      "num_tokens": 54347231.0,
      "step": 23690
    },
    {
      "entropy": 5.289004993438721,
      "epoch": 2.276176753121998,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044830340041789133,
      "loss": 4.7096,
      "mean_token_accuracy": 0.2473297029733658,
      "num_tokens": 54359096.0,
      "step": 23695
    },
    {
      "entropy": 5.272753095626831,
      "epoch": 2.276657060518732,
      "grad_norm": 1.25,
      "learning_rate": 0.0004482815338164152,
      "loss": 4.6184,
      "mean_token_accuracy": 0.26073159873485563,
      "num_tokens": 54370433.0,
      "step": 23700
    },
    {
      "entropy": 5.345137119293213,
      "epoch": 2.277137367915466,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004482596631918564,
      "loss": 4.7625,
      "mean_token_accuracy": 0.23737241476774215,
      "num_tokens": 54381430.0,
      "step": 23705
    },
    {
      "entropy": 5.283606386184692,
      "epoch": 2.2776176753121997,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044823778854472267,
      "loss": 4.7327,
      "mean_token_accuracy": 0.24008925557136535,
      "num_tokens": 54392840.0,
      "step": 23710
    },
    {
      "entropy": 5.387982606887817,
      "epoch": 2.2780979827089336,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004482159098755219,
      "loss": 4.841,
      "mean_token_accuracy": 0.23426577001810073,
      "num_tokens": 54405776.0,
      "step": 23715
    },
    {
      "entropy": 5.2987137794494625,
      "epoch": 2.2785782901056675,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004481940271847621,
      "loss": 4.676,
      "mean_token_accuracy": 0.24977568089962005,
      "num_tokens": 54417392.0,
      "step": 23720
    },
    {
      "entropy": 5.341944646835327,
      "epoch": 2.2790585975024014,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004481721404729513,
      "loss": 4.827,
      "mean_token_accuracy": 0.2330741196870804,
      "num_tokens": 54428407.0,
      "step": 23725
    },
    {
      "entropy": 5.343556690216064,
      "epoch": 2.2795389048991357,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004481502497405975,
      "loss": 4.7428,
      "mean_token_accuracy": 0.24325593709945678,
      "num_tokens": 54439667.0,
      "step": 23730
    },
    {
      "entropy": 5.27226300239563,
      "epoch": 2.280019212295869,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004481283549882091,
      "loss": 4.6814,
      "mean_token_accuracy": 0.24621856212615967,
      "num_tokens": 54451966.0,
      "step": 23735
    },
    {
      "entropy": 5.372122955322266,
      "epoch": 2.2804995196926034,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044810645621629443,
      "loss": 4.8553,
      "mean_token_accuracy": 0.22755274623632432,
      "num_tokens": 54462896.0,
      "step": 23740
    },
    {
      "entropy": 5.346497917175293,
      "epoch": 2.2809798270893373,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044808455342536176,
      "loss": 4.7745,
      "mean_token_accuracy": 0.24732649475336074,
      "num_tokens": 54474056.0,
      "step": 23745
    },
    {
      "entropy": 5.36555552482605,
      "epoch": 2.281460134486071,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044806264661591976,
      "loss": 4.7344,
      "mean_token_accuracy": 0.24422607719898223,
      "num_tokens": 54485846.0,
      "step": 23750
    },
    {
      "entropy": 5.294140863418579,
      "epoch": 2.281940441882805,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004480407357884771,
      "loss": 4.6758,
      "mean_token_accuracy": 0.25133625119924546,
      "num_tokens": 54496020.0,
      "step": 23755
    },
    {
      "entropy": 5.377021646499633,
      "epoch": 2.282420749279539,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044801882094354226,
      "loss": 4.802,
      "mean_token_accuracy": 0.23548173308372497,
      "num_tokens": 54507388.0,
      "step": 23760
    },
    {
      "entropy": 5.325430822372437,
      "epoch": 2.282901056676273,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004479969020816242,
      "loss": 4.7266,
      "mean_token_accuracy": 0.2511107861995697,
      "num_tokens": 54519159.0,
      "step": 23765
    },
    {
      "entropy": 5.345074462890625,
      "epoch": 2.2833813640730067,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044797497920323175,
      "loss": 4.8825,
      "mean_token_accuracy": 0.23569864332675933,
      "num_tokens": 54530258.0,
      "step": 23770
    },
    {
      "entropy": 5.357535934448242,
      "epoch": 2.2838616714697406,
      "grad_norm": 1.125,
      "learning_rate": 0.0004479530523088739,
      "loss": 4.8399,
      "mean_token_accuracy": 0.2322676107287407,
      "num_tokens": 54542626.0,
      "step": 23775
    },
    {
      "entropy": 5.361540222167969,
      "epoch": 2.2843419788664745,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004479311213990598,
      "loss": 4.7418,
      "mean_token_accuracy": 0.24246080368757247,
      "num_tokens": 54553762.0,
      "step": 23780
    },
    {
      "entropy": 5.312752676010132,
      "epoch": 2.2848222862632084,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044790918647429854,
      "loss": 4.6722,
      "mean_token_accuracy": 0.24680483043193818,
      "num_tokens": 54565484.0,
      "step": 23785
    },
    {
      "entropy": 5.2591307163238525,
      "epoch": 2.2853025936599423,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044788724753509935,
      "loss": 4.691,
      "mean_token_accuracy": 0.25331258624792097,
      "num_tokens": 54577321.0,
      "step": 23790
    },
    {
      "entropy": 5.226526737213135,
      "epoch": 2.285782901056676,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004478653045819717,
      "loss": 4.6643,
      "mean_token_accuracy": 0.2575864434242249,
      "num_tokens": 54587796.0,
      "step": 23795
    },
    {
      "entropy": 5.354264879226685,
      "epoch": 2.28626320845341,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004478433576154249,
      "loss": 4.8032,
      "mean_token_accuracy": 0.2339572489261627,
      "num_tokens": 54600637.0,
      "step": 23800
    },
    {
      "entropy": 5.393707990646362,
      "epoch": 2.286743515850144,
      "grad_norm": 1.125,
      "learning_rate": 0.0004478214066359687,
      "loss": 4.762,
      "mean_token_accuracy": 0.2479168325662613,
      "num_tokens": 54610690.0,
      "step": 23805
    },
    {
      "entropy": 5.336318397521973,
      "epoch": 2.287223823246878,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044779945164411254,
      "loss": 4.7835,
      "mean_token_accuracy": 0.24065859466791154,
      "num_tokens": 54622417.0,
      "step": 23810
    },
    {
      "entropy": 5.258180952072143,
      "epoch": 2.287704130643612,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004477774926403662,
      "loss": 4.7036,
      "mean_token_accuracy": 0.24893718659877778,
      "num_tokens": 54633190.0,
      "step": 23815
    },
    {
      "entropy": 5.377865266799927,
      "epoch": 2.288184438040346,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004477555296252396,
      "loss": 4.728,
      "mean_token_accuracy": 0.24289763420820237,
      "num_tokens": 54643507.0,
      "step": 23820
    },
    {
      "entropy": 5.224619388580322,
      "epoch": 2.28866474543708,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044773356259924255,
      "loss": 4.6535,
      "mean_token_accuracy": 0.2510466814041138,
      "num_tokens": 54653835.0,
      "step": 23825
    },
    {
      "entropy": 5.3213219165802,
      "epoch": 2.2891450528338138,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044771159156288505,
      "loss": 4.7535,
      "mean_token_accuracy": 0.24260659366846085,
      "num_tokens": 54664908.0,
      "step": 23830
    },
    {
      "entropy": 5.259343957901001,
      "epoch": 2.2896253602305476,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004476896165166772,
      "loss": 4.7263,
      "mean_token_accuracy": 0.24055345803499223,
      "num_tokens": 54677714.0,
      "step": 23835
    },
    {
      "entropy": 5.312827920913696,
      "epoch": 2.2901056676272815,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044766763746112936,
      "loss": 4.756,
      "mean_token_accuracy": 0.2364394634962082,
      "num_tokens": 54688919.0,
      "step": 23840
    },
    {
      "entropy": 5.318038845062256,
      "epoch": 2.2905859750240154,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004476456543967517,
      "loss": 4.7455,
      "mean_token_accuracy": 0.2443494141101837,
      "num_tokens": 54700089.0,
      "step": 23845
    },
    {
      "entropy": 5.3033520698547365,
      "epoch": 2.2910662824207493,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044762366732405454,
      "loss": 4.6952,
      "mean_token_accuracy": 0.2537821754813194,
      "num_tokens": 54711233.0,
      "step": 23850
    },
    {
      "entropy": 5.356963634490967,
      "epoch": 2.291546589817483,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004476016762435484,
      "loss": 4.8111,
      "mean_token_accuracy": 0.23749470114707946,
      "num_tokens": 54723140.0,
      "step": 23855
    },
    {
      "entropy": 5.307627773284912,
      "epoch": 2.292026897214217,
      "grad_norm": 1.0,
      "learning_rate": 0.0004475796811557439,
      "loss": 4.7437,
      "mean_token_accuracy": 0.24778575599193572,
      "num_tokens": 54734196.0,
      "step": 23860
    },
    {
      "entropy": 5.367501592636108,
      "epoch": 2.292507204610951,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044755768206115155,
      "loss": 4.8025,
      "mean_token_accuracy": 0.2300773248076439,
      "num_tokens": 54747395.0,
      "step": 23865
    },
    {
      "entropy": 5.321346569061279,
      "epoch": 2.292987512007685,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004475356789602823,
      "loss": 4.7936,
      "mean_token_accuracy": 0.23810062259435655,
      "num_tokens": 54761137.0,
      "step": 23870
    },
    {
      "entropy": 5.3148151397705075,
      "epoch": 2.2934678194044187,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044751367185364696,
      "loss": 4.677,
      "mean_token_accuracy": 0.2478257820010185,
      "num_tokens": 54772181.0,
      "step": 23875
    },
    {
      "entropy": 5.255773305892944,
      "epoch": 2.2939481268011526,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044749166074175634,
      "loss": 4.7169,
      "mean_token_accuracy": 0.25041214376688004,
      "num_tokens": 54783759.0,
      "step": 23880
    },
    {
      "entropy": 5.22601170539856,
      "epoch": 2.2944284341978864,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044746964562512154,
      "loss": 4.6336,
      "mean_token_accuracy": 0.24969541579484938,
      "num_tokens": 54796265.0,
      "step": 23885
    },
    {
      "entropy": 5.262799549102783,
      "epoch": 2.2949087415946208,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044744762650425376,
      "loss": 4.7331,
      "mean_token_accuracy": 0.24573244005441666,
      "num_tokens": 54808159.0,
      "step": 23890
    },
    {
      "entropy": 5.370202684402466,
      "epoch": 2.2953890489913547,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044742560337966415,
      "loss": 4.8281,
      "mean_token_accuracy": 0.24137523472309114,
      "num_tokens": 54819070.0,
      "step": 23895
    },
    {
      "entropy": 5.259682750701904,
      "epoch": 2.2958693563880885,
      "grad_norm": 1.234375,
      "learning_rate": 0.000447403576251864,
      "loss": 4.7124,
      "mean_token_accuracy": 0.2363220915198326,
      "num_tokens": 54829210.0,
      "step": 23900
    },
    {
      "entropy": 5.269902992248535,
      "epoch": 2.2963496637848224,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004473815451213648,
      "loss": 4.7057,
      "mean_token_accuracy": 0.24514997899532318,
      "num_tokens": 54841238.0,
      "step": 23905
    },
    {
      "entropy": 5.2738148212432865,
      "epoch": 2.2968299711815563,
      "grad_norm": 1.125,
      "learning_rate": 0.0004473595099886779,
      "loss": 4.7572,
      "mean_token_accuracy": 0.246664659678936,
      "num_tokens": 54852531.0,
      "step": 23910
    },
    {
      "entropy": 5.349978637695313,
      "epoch": 2.29731027857829,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004473374708543151,
      "loss": 4.7192,
      "mean_token_accuracy": 0.24332005828619002,
      "num_tokens": 54864089.0,
      "step": 23915
    },
    {
      "entropy": 5.306702375411987,
      "epoch": 2.297790585975024,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004473154277187878,
      "loss": 4.7047,
      "mean_token_accuracy": 0.24160427004098892,
      "num_tokens": 54874956.0,
      "step": 23920
    },
    {
      "entropy": 5.281047344207764,
      "epoch": 2.298270893371758,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044729338058260805,
      "loss": 4.7322,
      "mean_token_accuracy": 0.2469619482755661,
      "num_tokens": 54885608.0,
      "step": 23925
    },
    {
      "entropy": 5.2714485168457035,
      "epoch": 2.298751200768492,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004472713294462876,
      "loss": 4.6846,
      "mean_token_accuracy": 0.24792144894599916,
      "num_tokens": 54896782.0,
      "step": 23930
    },
    {
      "entropy": 5.298613023757935,
      "epoch": 2.2992315081652257,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044724927431033843,
      "loss": 4.7535,
      "mean_token_accuracy": 0.243763467669487,
      "num_tokens": 54908179.0,
      "step": 23935
    },
    {
      "entropy": 5.295199871063232,
      "epoch": 2.2997118155619596,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004472272151752725,
      "loss": 4.7097,
      "mean_token_accuracy": 0.2461494192481041,
      "num_tokens": 54921114.0,
      "step": 23940
    },
    {
      "entropy": 5.322293329238891,
      "epoch": 2.3001921229586935,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004472051520416022,
      "loss": 4.7336,
      "mean_token_accuracy": 0.2401993155479431,
      "num_tokens": 54931398.0,
      "step": 23945
    },
    {
      "entropy": 5.375329971313477,
      "epoch": 2.3006724303554273,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004471830849098395,
      "loss": 4.7819,
      "mean_token_accuracy": 0.2399084150791168,
      "num_tokens": 54942354.0,
      "step": 23950
    },
    {
      "entropy": 5.293951177597046,
      "epoch": 2.3011527377521612,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044716101378049683,
      "loss": 4.7145,
      "mean_token_accuracy": 0.2516032546758652,
      "num_tokens": 54953700.0,
      "step": 23955
    },
    {
      "entropy": 5.16375150680542,
      "epoch": 2.301633045148895,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044713893865408667,
      "loss": 4.6395,
      "mean_token_accuracy": 0.251749324798584,
      "num_tokens": 54965544.0,
      "step": 23960
    },
    {
      "entropy": 5.279995059967041,
      "epoch": 2.3021133525456294,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004471168595311215,
      "loss": 4.6828,
      "mean_token_accuracy": 0.25274141430854796,
      "num_tokens": 54976811.0,
      "step": 23965
    },
    {
      "entropy": 5.305463647842407,
      "epoch": 2.302593659942363,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044709477641211395,
      "loss": 4.6907,
      "mean_token_accuracy": 0.25068259835243223,
      "num_tokens": 54987508.0,
      "step": 23970
    },
    {
      "entropy": 5.401940202713012,
      "epoch": 2.303073967339097,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004470726892975767,
      "loss": 4.88,
      "mean_token_accuracy": 0.2293749675154686,
      "num_tokens": 55000276.0,
      "step": 23975
    },
    {
      "entropy": 5.306238842010498,
      "epoch": 2.303554274735831,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044705059818802255,
      "loss": 4.7042,
      "mean_token_accuracy": 0.24988709837198259,
      "num_tokens": 55012397.0,
      "step": 23980
    },
    {
      "entropy": 5.265456104278565,
      "epoch": 2.304034582132565,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004470285030839644,
      "loss": 4.6633,
      "mean_token_accuracy": 0.25028522610664367,
      "num_tokens": 55024358.0,
      "step": 23985
    },
    {
      "entropy": 5.2347277164459225,
      "epoch": 2.304514889529299,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044700640398591526,
      "loss": 4.726,
      "mean_token_accuracy": 0.2473867878317833,
      "num_tokens": 55035887.0,
      "step": 23990
    },
    {
      "entropy": 5.349122476577759,
      "epoch": 2.3049951969260327,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004469843008943881,
      "loss": 4.7984,
      "mean_token_accuracy": 0.235358726978302,
      "num_tokens": 55048775.0,
      "step": 23995
    },
    {
      "entropy": 5.337689590454102,
      "epoch": 2.3054755043227666,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004469621938098962,
      "loss": 4.7633,
      "mean_token_accuracy": 0.24535427689552308,
      "num_tokens": 55059994.0,
      "step": 24000
    },
    {
      "epoch": 2.3054755043227666,
      "eval_entropy": 5.09432132728852,
      "eval_loss": 4.819790840148926,
      "eval_mean_token_accuracy": 0.24871662934776168,
      "eval_num_tokens": 55059994.0,
      "eval_runtime": 20.4886,
      "eval_samples_per_second": 1601.626,
      "eval_steps_per_second": 200.209,
      "step": 24000
    },
    {
      "entropy": 5.319386959075928,
      "epoch": 2.3059558117195005,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004469400827329528,
      "loss": 4.7698,
      "mean_token_accuracy": 0.2408551439642906,
      "num_tokens": 55071950.0,
      "step": 24005
    },
    {
      "entropy": 5.200397157669068,
      "epoch": 2.3064361191162344,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004469179676640712,
      "loss": 4.6275,
      "mean_token_accuracy": 0.2567056745290756,
      "num_tokens": 55085260.0,
      "step": 24010
    },
    {
      "entropy": 5.363962650299072,
      "epoch": 2.3069164265129682,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004468958486037649,
      "loss": 4.6557,
      "mean_token_accuracy": 0.24390911757946016,
      "num_tokens": 55095794.0,
      "step": 24015
    },
    {
      "entropy": 5.271401309967041,
      "epoch": 2.307396733909702,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004468737255525474,
      "loss": 4.7394,
      "mean_token_accuracy": 0.24086216241121292,
      "num_tokens": 55106476.0,
      "step": 24020
    },
    {
      "entropy": 5.248284482955933,
      "epoch": 2.307877041306436,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004468515985109324,
      "loss": 4.7485,
      "mean_token_accuracy": 0.24627834409475327,
      "num_tokens": 55118277.0,
      "step": 24025
    },
    {
      "entropy": 5.323884773254394,
      "epoch": 2.30835734870317,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004468294674794335,
      "loss": 4.7426,
      "mean_token_accuracy": 0.24354281574487685,
      "num_tokens": 55129902.0,
      "step": 24030
    },
    {
      "entropy": 5.267329263687134,
      "epoch": 2.3088376560999038,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004468073324585647,
      "loss": 4.6907,
      "mean_token_accuracy": 0.2386388063430786,
      "num_tokens": 55143342.0,
      "step": 24035
    },
    {
      "entropy": 5.361611032485962,
      "epoch": 2.309317963496638,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004467851934488398,
      "loss": 4.7515,
      "mean_token_accuracy": 0.24653020352125168,
      "num_tokens": 55154116.0,
      "step": 24040
    },
    {
      "entropy": 5.320466232299805,
      "epoch": 2.3097982708933715,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004467630504507727,
      "loss": 4.7267,
      "mean_token_accuracy": 0.24481357932090758,
      "num_tokens": 55165548.0,
      "step": 24045
    },
    {
      "entropy": 5.3319411277771,
      "epoch": 2.310278578290106,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004467409034648776,
      "loss": 4.7988,
      "mean_token_accuracy": 0.2380385383963585,
      "num_tokens": 55176047.0,
      "step": 24050
    },
    {
      "entropy": 5.271692419052124,
      "epoch": 2.3107588856868397,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004467187524916688,
      "loss": 4.7515,
      "mean_token_accuracy": 0.24562847912311553,
      "num_tokens": 55188715.0,
      "step": 24055
    },
    {
      "entropy": 5.205599594116211,
      "epoch": 2.3112391930835736,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004466965975316604,
      "loss": 4.6155,
      "mean_token_accuracy": 0.2557682454586029,
      "num_tokens": 55198988.0,
      "step": 24060
    },
    {
      "entropy": 5.188669013977051,
      "epoch": 2.3117195004803075,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044667443858536685,
      "loss": 4.6024,
      "mean_token_accuracy": 0.2570546671748161,
      "num_tokens": 55210000.0,
      "step": 24065
    },
    {
      "entropy": 5.312769031524658,
      "epoch": 2.3121998078770414,
      "grad_norm": 1.125,
      "learning_rate": 0.0004466522756533026,
      "loss": 4.7358,
      "mean_token_accuracy": 0.23673695027828218,
      "num_tokens": 55221794.0,
      "step": 24070
    },
    {
      "entropy": 5.35266695022583,
      "epoch": 2.3126801152737753,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004466301087359822,
      "loss": 4.8041,
      "mean_token_accuracy": 0.2374489739537239,
      "num_tokens": 55233166.0,
      "step": 24075
    },
    {
      "entropy": 5.275799417495728,
      "epoch": 2.313160422670509,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044660793783392035,
      "loss": 4.685,
      "mean_token_accuracy": 0.24702323824167252,
      "num_tokens": 55244179.0,
      "step": 24080
    },
    {
      "entropy": 5.249599552154541,
      "epoch": 2.313640730067243,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004465857629476317,
      "loss": 4.6993,
      "mean_token_accuracy": 0.24720760881900788,
      "num_tokens": 55255556.0,
      "step": 24085
    },
    {
      "entropy": 5.407215261459351,
      "epoch": 2.314121037463977,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004465635840776312,
      "loss": 4.8433,
      "mean_token_accuracy": 0.23900545090436937,
      "num_tokens": 55267001.0,
      "step": 24090
    },
    {
      "entropy": 5.397832250595092,
      "epoch": 2.314601344860711,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044654140122443373,
      "loss": 4.7476,
      "mean_token_accuracy": 0.24598277360200882,
      "num_tokens": 55278253.0,
      "step": 24095
    },
    {
      "entropy": 5.319160652160645,
      "epoch": 2.3150816522574447,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004465192143885542,
      "loss": 4.7203,
      "mean_token_accuracy": 0.24726214706897737,
      "num_tokens": 55289601.0,
      "step": 24100
    },
    {
      "entropy": 5.2887006282806395,
      "epoch": 2.3155619596541785,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044649702357050787,
      "loss": 4.7494,
      "mean_token_accuracy": 0.238137823343277,
      "num_tokens": 55300661.0,
      "step": 24105
    },
    {
      "entropy": 5.28918309211731,
      "epoch": 2.3160422670509124,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004464748287708099,
      "loss": 4.7058,
      "mean_token_accuracy": 0.24940220266580582,
      "num_tokens": 55311890.0,
      "step": 24110
    },
    {
      "entropy": 5.236471319198609,
      "epoch": 2.3165225744476463,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044645262998997557,
      "loss": 4.6607,
      "mean_token_accuracy": 0.25559800714254377,
      "num_tokens": 55322241.0,
      "step": 24115
    },
    {
      "entropy": 5.278766775131226,
      "epoch": 2.31700288184438,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044643042722852024,
      "loss": 4.6377,
      "mean_token_accuracy": 0.25591217130422594,
      "num_tokens": 55334064.0,
      "step": 24120
    },
    {
      "entropy": 5.303686857223511,
      "epoch": 2.3174831892411145,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004464082204869595,
      "loss": 4.7536,
      "mean_token_accuracy": 0.23903511911630632,
      "num_tokens": 55346266.0,
      "step": 24125
    },
    {
      "entropy": 5.2643648147583,
      "epoch": 2.3179634966378484,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004463860097658088,
      "loss": 4.6916,
      "mean_token_accuracy": 0.24787720441818237,
      "num_tokens": 55357395.0,
      "step": 24130
    },
    {
      "entropy": 5.297639036178589,
      "epoch": 2.3184438040345823,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004463637950655839,
      "loss": 4.6936,
      "mean_token_accuracy": 0.24257504642009736,
      "num_tokens": 55367982.0,
      "step": 24135
    },
    {
      "entropy": 5.303102111816406,
      "epoch": 2.318924111431316,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044634157638680054,
      "loss": 4.7736,
      "mean_token_accuracy": 0.24268539100885392,
      "num_tokens": 55379822.0,
      "step": 24140
    },
    {
      "entropy": 5.322305488586426,
      "epoch": 2.31940441882805,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044631935372997455,
      "loss": 4.7169,
      "mean_token_accuracy": 0.23912623524665833,
      "num_tokens": 55391271.0,
      "step": 24145
    },
    {
      "entropy": 5.310552406311035,
      "epoch": 2.319884726224784,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004462971270956219,
      "loss": 4.6636,
      "mean_token_accuracy": 0.24339904636144638,
      "num_tokens": 55401514.0,
      "step": 24150
    },
    {
      "entropy": 5.36877212524414,
      "epoch": 2.320365033621518,
      "grad_norm": 1.125,
      "learning_rate": 0.0004462748964842586,
      "loss": 4.7745,
      "mean_token_accuracy": 0.2372725188732147,
      "num_tokens": 55412979.0,
      "step": 24155
    },
    {
      "entropy": 5.276529932022095,
      "epoch": 2.3208453410182517,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004462526618964008,
      "loss": 4.7027,
      "mean_token_accuracy": 0.24246210008859634,
      "num_tokens": 55423985.0,
      "step": 24160
    },
    {
      "entropy": 5.346647500991821,
      "epoch": 2.3213256484149856,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004462304233325647,
      "loss": 4.7236,
      "mean_token_accuracy": 0.24347291886806488,
      "num_tokens": 55434638.0,
      "step": 24165
    },
    {
      "entropy": 5.321153259277343,
      "epoch": 2.3218059558117194,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004462081807932666,
      "loss": 4.7858,
      "mean_token_accuracy": 0.2367343083024025,
      "num_tokens": 55445402.0,
      "step": 24170
    },
    {
      "entropy": 5.322035312652588,
      "epoch": 2.3222862632084533,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004461859342790229,
      "loss": 4.7063,
      "mean_token_accuracy": 0.2439023122191429,
      "num_tokens": 55456707.0,
      "step": 24175
    },
    {
      "entropy": 5.269166374206543,
      "epoch": 2.322766570605187,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004461636837903501,
      "loss": 4.6942,
      "mean_token_accuracy": 0.2436886191368103,
      "num_tokens": 55468473.0,
      "step": 24180
    },
    {
      "entropy": 5.188522577285767,
      "epoch": 2.323246878001921,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004461414293277649,
      "loss": 4.5976,
      "mean_token_accuracy": 0.25284337252378464,
      "num_tokens": 55479623.0,
      "step": 24185
    },
    {
      "entropy": 5.324103927612304,
      "epoch": 2.323727185398655,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004461191708917839,
      "loss": 4.7955,
      "mean_token_accuracy": 0.23878859281539916,
      "num_tokens": 55490036.0,
      "step": 24190
    },
    {
      "entropy": 5.267402315139771,
      "epoch": 2.324207492795389,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044609690848292376,
      "loss": 4.7388,
      "mean_token_accuracy": 0.24750769585371019,
      "num_tokens": 55501605.0,
      "step": 24195
    },
    {
      "entropy": 5.297731876373291,
      "epoch": 2.324687800192123,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004460746421017016,
      "loss": 4.6546,
      "mean_token_accuracy": 0.2548551172018051,
      "num_tokens": 55512029.0,
      "step": 24200
    },
    {
      "entropy": 5.323645114898682,
      "epoch": 2.325168107588857,
      "grad_norm": 1.0,
      "learning_rate": 0.00044605237174863405,
      "loss": 4.7465,
      "mean_token_accuracy": 0.24834782183170317,
      "num_tokens": 55524260.0,
      "step": 24205
    },
    {
      "entropy": 5.208985280990601,
      "epoch": 2.325648414985591,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004460300974242386,
      "loss": 4.7471,
      "mean_token_accuracy": 0.24232691377401352,
      "num_tokens": 55536520.0,
      "step": 24210
    },
    {
      "entropy": 5.34546012878418,
      "epoch": 2.326128722382325,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004460078191290319,
      "loss": 4.7946,
      "mean_token_accuracy": 0.23477506190538405,
      "num_tokens": 55547896.0,
      "step": 24215
    },
    {
      "entropy": 5.367196130752563,
      "epoch": 2.3266090297790587,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044598553686353153,
      "loss": 4.7244,
      "mean_token_accuracy": 0.23851590603590012,
      "num_tokens": 55559229.0,
      "step": 24220
    },
    {
      "entropy": 5.3526793956756595,
      "epoch": 2.3270893371757926,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044596325062825476,
      "loss": 4.7459,
      "mean_token_accuracy": 0.24217483550310134,
      "num_tokens": 55569531.0,
      "step": 24225
    },
    {
      "entropy": 5.304793214797973,
      "epoch": 2.3275696445725265,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004459409604237189,
      "loss": 4.7009,
      "mean_token_accuracy": 0.2449586421251297,
      "num_tokens": 55580454.0,
      "step": 24230
    },
    {
      "entropy": 5.275087928771972,
      "epoch": 2.3280499519692603,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044591866625044154,
      "loss": 4.7016,
      "mean_token_accuracy": 0.2568980470299721,
      "num_tokens": 55591336.0,
      "step": 24235
    },
    {
      "entropy": 5.282370519638062,
      "epoch": 2.3285302593659942,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004458963681089403,
      "loss": 4.7558,
      "mean_token_accuracy": 0.2458904579281807,
      "num_tokens": 55601955.0,
      "step": 24240
    },
    {
      "entropy": 5.345089864730835,
      "epoch": 2.329010566762728,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004458740659997328,
      "loss": 4.8252,
      "mean_token_accuracy": 0.2332942619919777,
      "num_tokens": 55614130.0,
      "step": 24245
    },
    {
      "entropy": 5.355337619781494,
      "epoch": 2.329490874159462,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004458517599233369,
      "loss": 4.7346,
      "mean_token_accuracy": 0.24347249269485474,
      "num_tokens": 55625447.0,
      "step": 24250
    },
    {
      "entropy": 5.297988843917847,
      "epoch": 2.329971181556196,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004458294498802706,
      "loss": 4.706,
      "mean_token_accuracy": 0.24658190608024597,
      "num_tokens": 55636426.0,
      "step": 24255
    },
    {
      "entropy": 5.237401628494263,
      "epoch": 2.3304514889529298,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004458071358710516,
      "loss": 4.6846,
      "mean_token_accuracy": 0.23808805048465728,
      "num_tokens": 55647095.0,
      "step": 24260
    },
    {
      "entropy": 5.241651344299316,
      "epoch": 2.3309317963496636,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004457848178961981,
      "loss": 4.6331,
      "mean_token_accuracy": 0.24853117913007736,
      "num_tokens": 55660362.0,
      "step": 24265
    },
    {
      "entropy": 5.302249622344971,
      "epoch": 2.3314121037463975,
      "grad_norm": 1.0,
      "learning_rate": 0.00044576249595622833,
      "loss": 4.6714,
      "mean_token_accuracy": 0.2489761620759964,
      "num_tokens": 55671760.0,
      "step": 24270
    },
    {
      "entropy": 5.2812254428863525,
      "epoch": 2.331892411143132,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004457401700516603,
      "loss": 4.7316,
      "mean_token_accuracy": 0.24539994597434997,
      "num_tokens": 55683826.0,
      "step": 24275
    },
    {
      "entropy": 5.332390022277832,
      "epoch": 2.3323727185398653,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044571784018301267,
      "loss": 4.7233,
      "mean_token_accuracy": 0.24767063558101654,
      "num_tokens": 55694939.0,
      "step": 24280
    },
    {
      "entropy": 5.296097993850708,
      "epoch": 2.3328530259365996,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00044569550635080365,
      "loss": 4.752,
      "mean_token_accuracy": 0.23870059847831726,
      "num_tokens": 55707101.0,
      "step": 24285
    },
    {
      "entropy": 5.247024583816528,
      "epoch": 2.3333333333333335,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044567316855555184,
      "loss": 4.6538,
      "mean_token_accuracy": 0.2454217940568924,
      "num_tokens": 55718418.0,
      "step": 24290
    },
    {
      "entropy": 5.352167320251465,
      "epoch": 2.3338136407300674,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004456508267977759,
      "loss": 4.8748,
      "mean_token_accuracy": 0.22981469035148622,
      "num_tokens": 55729229.0,
      "step": 24295
    },
    {
      "entropy": 5.341595792770386,
      "epoch": 2.3342939481268012,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044562848107799444,
      "loss": 4.7528,
      "mean_token_accuracy": 0.2419998377561569,
      "num_tokens": 55740719.0,
      "step": 24300
    },
    {
      "entropy": 5.308290910720825,
      "epoch": 2.334774255523535,
      "grad_norm": 1.78125,
      "learning_rate": 0.00044560613139672627,
      "loss": 4.6369,
      "mean_token_accuracy": 0.25181944668293,
      "num_tokens": 55752363.0,
      "step": 24305
    },
    {
      "entropy": 5.226222038269043,
      "epoch": 2.335254562920269,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044558377775449036,
      "loss": 4.6716,
      "mean_token_accuracy": 0.24895276725292206,
      "num_tokens": 55763851.0,
      "step": 24310
    },
    {
      "entropy": 5.290854549407959,
      "epoch": 2.335734870317003,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044556142015180573,
      "loss": 4.6985,
      "mean_token_accuracy": 0.2560324013233185,
      "num_tokens": 55776118.0,
      "step": 24315
    },
    {
      "entropy": 5.220709323883057,
      "epoch": 2.3362151777137368,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044553905858919134,
      "loss": 4.6312,
      "mean_token_accuracy": 0.2525891125202179,
      "num_tokens": 55788106.0,
      "step": 24320
    },
    {
      "entropy": 5.283627653121949,
      "epoch": 2.3366954851104706,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004455166930671664,
      "loss": 4.7493,
      "mean_token_accuracy": 0.243920961022377,
      "num_tokens": 55798608.0,
      "step": 24325
    },
    {
      "entropy": 5.369166707992553,
      "epoch": 2.3371757925072045,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044549432358625014,
      "loss": 4.7579,
      "mean_token_accuracy": 0.24213359802961348,
      "num_tokens": 55810963.0,
      "step": 24330
    },
    {
      "entropy": 5.275917291641235,
      "epoch": 2.3376560999039384,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000445471950146962,
      "loss": 4.6255,
      "mean_token_accuracy": 0.24952723681926728,
      "num_tokens": 55821593.0,
      "step": 24335
    },
    {
      "entropy": 5.3198895931243895,
      "epoch": 2.3381364073006723,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004454495727498214,
      "loss": 4.806,
      "mean_token_accuracy": 0.23801144063472748,
      "num_tokens": 55833602.0,
      "step": 24340
    },
    {
      "entropy": 5.434912919998169,
      "epoch": 2.338616714697406,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004454271913953478,
      "loss": 4.8458,
      "mean_token_accuracy": 0.23343184888362883,
      "num_tokens": 55844745.0,
      "step": 24345
    },
    {
      "entropy": 5.412781715393066,
      "epoch": 2.3390970220941405,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044540480608406093,
      "loss": 4.8202,
      "mean_token_accuracy": 0.23942200243473052,
      "num_tokens": 55857193.0,
      "step": 24350
    },
    {
      "entropy": 5.3081365585327145,
      "epoch": 2.339577329490874,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004453824168164804,
      "loss": 4.7195,
      "mean_token_accuracy": 0.24173368215560914,
      "num_tokens": 55868954.0,
      "step": 24355
    },
    {
      "entropy": 5.24794750213623,
      "epoch": 2.3400576368876083,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004453600235931261,
      "loss": 4.7159,
      "mean_token_accuracy": 0.24526629447937012,
      "num_tokens": 55880485.0,
      "step": 24360
    },
    {
      "entropy": 5.3143946647644045,
      "epoch": 2.340537944284342,
      "grad_norm": 1.03125,
      "learning_rate": 0.000445337626414518,
      "loss": 4.7583,
      "mean_token_accuracy": 0.2390078440308571,
      "num_tokens": 55893043.0,
      "step": 24365
    },
    {
      "entropy": 5.2398120880126955,
      "epoch": 2.341018251681076,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044531522528117593,
      "loss": 4.6456,
      "mean_token_accuracy": 0.2491727888584137,
      "num_tokens": 55905776.0,
      "step": 24370
    },
    {
      "entropy": 5.304390811920166,
      "epoch": 2.34149855907781,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044529282019362007,
      "loss": 4.7184,
      "mean_token_accuracy": 0.2516643449664116,
      "num_tokens": 55916424.0,
      "step": 24375
    },
    {
      "entropy": 5.384766530990601,
      "epoch": 2.341978866474544,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044527041115237056,
      "loss": 4.8549,
      "mean_token_accuracy": 0.22950814068317413,
      "num_tokens": 55928510.0,
      "step": 24380
    },
    {
      "entropy": 5.385864925384522,
      "epoch": 2.3424591738712777,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004452479981579477,
      "loss": 4.8178,
      "mean_token_accuracy": 0.2360099047422409,
      "num_tokens": 55940536.0,
      "step": 24385
    },
    {
      "entropy": 5.35517635345459,
      "epoch": 2.3429394812680115,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004452255812108719,
      "loss": 4.7595,
      "mean_token_accuracy": 0.24470617920160292,
      "num_tokens": 55951837.0,
      "step": 24390
    },
    {
      "entropy": 5.184165096282959,
      "epoch": 2.3434197886647454,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004452031603116636,
      "loss": 4.6113,
      "mean_token_accuracy": 0.25965113937854767,
      "num_tokens": 55963641.0,
      "step": 24395
    },
    {
      "entropy": 5.304730224609375,
      "epoch": 2.3439000960614793,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044518073546084326,
      "loss": 4.7791,
      "mean_token_accuracy": 0.23964662253856658,
      "num_tokens": 55975205.0,
      "step": 24400
    },
    {
      "entropy": 5.379022216796875,
      "epoch": 2.344380403458213,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004451583066589316,
      "loss": 4.7966,
      "mean_token_accuracy": 0.23726546615362168,
      "num_tokens": 55985456.0,
      "step": 24405
    },
    {
      "entropy": 5.355392551422119,
      "epoch": 2.344860710854947,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044513587390644925,
      "loss": 4.7682,
      "mean_token_accuracy": 0.24538686275482177,
      "num_tokens": 55997946.0,
      "step": 24410
    },
    {
      "entropy": 5.357404899597168,
      "epoch": 2.345341018251681,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044511343720391724,
      "loss": 4.764,
      "mean_token_accuracy": 0.23850201964378356,
      "num_tokens": 56008907.0,
      "step": 24415
    },
    {
      "entropy": 5.298089408874512,
      "epoch": 2.345821325648415,
      "grad_norm": 1.125,
      "learning_rate": 0.0004450909965518563,
      "loss": 4.7084,
      "mean_token_accuracy": 0.2375366434454918,
      "num_tokens": 56019153.0,
      "step": 24420
    },
    {
      "entropy": 5.206905126571655,
      "epoch": 2.3463016330451487,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044506855195078755,
      "loss": 4.6121,
      "mean_token_accuracy": 0.24850079119205476,
      "num_tokens": 56029944.0,
      "step": 24425
    },
    {
      "entropy": 5.33498592376709,
      "epoch": 2.3467819404418826,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044504610340123185,
      "loss": 4.6842,
      "mean_token_accuracy": 0.24486488550901414,
      "num_tokens": 56041782.0,
      "step": 24430
    },
    {
      "entropy": 5.3754478931427006,
      "epoch": 2.347262247838617,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044502365090371066,
      "loss": 4.8131,
      "mean_token_accuracy": 0.2384071931242943,
      "num_tokens": 56052803.0,
      "step": 24435
    },
    {
      "entropy": 5.356517648696899,
      "epoch": 2.347742555235351,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004450011944587452,
      "loss": 4.7525,
      "mean_token_accuracy": 0.2419370397925377,
      "num_tokens": 56062503.0,
      "step": 24440
    },
    {
      "entropy": 5.279032945632935,
      "epoch": 2.3482228626320847,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044497873406685673,
      "loss": 4.7156,
      "mean_token_accuracy": 0.24802958220243454,
      "num_tokens": 56073147.0,
      "step": 24445
    },
    {
      "entropy": 5.3301468849182125,
      "epoch": 2.3487031700288186,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004449562697285669,
      "loss": 4.7719,
      "mean_token_accuracy": 0.24263350665569305,
      "num_tokens": 56084450.0,
      "step": 24450
    },
    {
      "entropy": 5.251854944229126,
      "epoch": 2.3491834774255524,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044493380144439707,
      "loss": 4.6452,
      "mean_token_accuracy": 0.2492106944322586,
      "num_tokens": 56094869.0,
      "step": 24455
    },
    {
      "entropy": 5.27655029296875,
      "epoch": 2.3496637848222863,
      "grad_norm": 1.03125,
      "learning_rate": 0.000444911329214869,
      "loss": 4.6865,
      "mean_token_accuracy": 0.25324206352233886,
      "num_tokens": 56105640.0,
      "step": 24460
    },
    {
      "entropy": 5.353990840911865,
      "epoch": 2.35014409221902,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044488885304050434,
      "loss": 4.8122,
      "mean_token_accuracy": 0.23658599704504013,
      "num_tokens": 56116756.0,
      "step": 24465
    },
    {
      "entropy": 5.357531261444092,
      "epoch": 2.350624399615754,
      "grad_norm": 0.921875,
      "learning_rate": 0.000444866372921825,
      "loss": 4.7385,
      "mean_token_accuracy": 0.24589592069387436,
      "num_tokens": 56129674.0,
      "step": 24470
    },
    {
      "entropy": 5.3400355815887455,
      "epoch": 2.351104707012488,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044484388885935287,
      "loss": 4.8028,
      "mean_token_accuracy": 0.2389134630560875,
      "num_tokens": 56141360.0,
      "step": 24475
    },
    {
      "entropy": 5.271697854995727,
      "epoch": 2.351585014409222,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044482140085361005,
      "loss": 4.6319,
      "mean_token_accuracy": 0.2594841971993446,
      "num_tokens": 56152995.0,
      "step": 24480
    },
    {
      "entropy": 5.360607862472534,
      "epoch": 2.3520653218059557,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044479890890511853,
      "loss": 4.8093,
      "mean_token_accuracy": 0.24257292449474335,
      "num_tokens": 56164515.0,
      "step": 24485
    },
    {
      "entropy": 5.299129390716553,
      "epoch": 2.3525456292026896,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044477641301440054,
      "loss": 4.7526,
      "mean_token_accuracy": 0.24799326211214065,
      "num_tokens": 56175983.0,
      "step": 24490
    },
    {
      "entropy": 5.21643099784851,
      "epoch": 2.3530259365994235,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004447539131819784,
      "loss": 4.6554,
      "mean_token_accuracy": 0.24597966372966767,
      "num_tokens": 56187176.0,
      "step": 24495
    },
    {
      "entropy": 5.302491283416748,
      "epoch": 2.3535062439961574,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044473140940837436,
      "loss": 4.7144,
      "mean_token_accuracy": 0.2426817774772644,
      "num_tokens": 56198713.0,
      "step": 24500
    },
    {
      "entropy": 5.294138336181641,
      "epoch": 2.3539865513928913,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044470890169411107,
      "loss": 4.7011,
      "mean_token_accuracy": 0.2445197209715843,
      "num_tokens": 56210473.0,
      "step": 24505
    },
    {
      "entropy": 5.312314176559449,
      "epoch": 2.3544668587896256,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000444686390039711,
      "loss": 4.8149,
      "mean_token_accuracy": 0.23844933956861497,
      "num_tokens": 56221850.0,
      "step": 24510
    },
    {
      "entropy": 5.291109943389893,
      "epoch": 2.3549471661863595,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004446638744456968,
      "loss": 4.7049,
      "mean_token_accuracy": 0.24648405015468597,
      "num_tokens": 56232359.0,
      "step": 24515
    },
    {
      "entropy": 5.254538536071777,
      "epoch": 2.3554274735830933,
      "grad_norm": 1.125,
      "learning_rate": 0.00044464135491259135,
      "loss": 4.6883,
      "mean_token_accuracy": 0.2443451017141342,
      "num_tokens": 56243385.0,
      "step": 24520
    },
    {
      "entropy": 5.281840181350708,
      "epoch": 2.3559077809798272,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004446188314409172,
      "loss": 4.6781,
      "mean_token_accuracy": 0.257035468518734,
      "num_tokens": 56254713.0,
      "step": 24525
    },
    {
      "entropy": 5.278661394119263,
      "epoch": 2.356388088376561,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004445963040311975,
      "loss": 4.6579,
      "mean_token_accuracy": 0.2557175800204277,
      "num_tokens": 56266227.0,
      "step": 24530
    },
    {
      "entropy": 5.384924459457397,
      "epoch": 2.356868395773295,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044457377268395526,
      "loss": 4.8991,
      "mean_token_accuracy": 0.234905669093132,
      "num_tokens": 56276258.0,
      "step": 24535
    },
    {
      "entropy": 5.400746917724609,
      "epoch": 2.357348703170029,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044455123739971355,
      "loss": 4.7915,
      "mean_token_accuracy": 0.2405368134379387,
      "num_tokens": 56287909.0,
      "step": 24540
    },
    {
      "entropy": 5.3661597728729244,
      "epoch": 2.3578290105667628,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044452869817899554,
      "loss": 4.776,
      "mean_token_accuracy": 0.24490271061658858,
      "num_tokens": 56299161.0,
      "step": 24545
    },
    {
      "entropy": 5.325969266891479,
      "epoch": 2.3583093179634966,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004445061550223246,
      "loss": 4.7496,
      "mean_token_accuracy": 0.24578150361776352,
      "num_tokens": 56311333.0,
      "step": 24550
    },
    {
      "entropy": 5.296340560913086,
      "epoch": 2.3587896253602305,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044448360793022403,
      "loss": 4.6377,
      "mean_token_accuracy": 0.2574515253305435,
      "num_tokens": 56322349.0,
      "step": 24555
    },
    {
      "entropy": 5.309181976318359,
      "epoch": 2.3592699327569644,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004444610569032174,
      "loss": 4.7703,
      "mean_token_accuracy": 0.24408458471298217,
      "num_tokens": 56333169.0,
      "step": 24560
    },
    {
      "entropy": 5.196787929534912,
      "epoch": 2.3597502401536983,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004444385019418281,
      "loss": 4.635,
      "mean_token_accuracy": 0.2514422595500946,
      "num_tokens": 56345419.0,
      "step": 24565
    },
    {
      "entropy": 5.281092834472656,
      "epoch": 2.360230547550432,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044441594304658004,
      "loss": 4.6777,
      "mean_token_accuracy": 0.25005222409963607,
      "num_tokens": 56355848.0,
      "step": 24570
    },
    {
      "entropy": 5.370937728881836,
      "epoch": 2.360710854947166,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004443933802179968,
      "loss": 4.8702,
      "mean_token_accuracy": 0.23157764226198196,
      "num_tokens": 56368369.0,
      "step": 24575
    },
    {
      "entropy": 5.272527694702148,
      "epoch": 2.3611911623439,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044437081345660224,
      "loss": 4.6865,
      "mean_token_accuracy": 0.24206824153661727,
      "num_tokens": 56379069.0,
      "step": 24580
    },
    {
      "entropy": 5.287461996078491,
      "epoch": 2.3616714697406342,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004443482427629204,
      "loss": 4.7212,
      "mean_token_accuracy": 0.24425991475582123,
      "num_tokens": 56390325.0,
      "step": 24585
    },
    {
      "entropy": 5.355063533782959,
      "epoch": 2.3621517771373677,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004443256681374751,
      "loss": 4.7846,
      "mean_token_accuracy": 0.24420422017574311,
      "num_tokens": 56401336.0,
      "step": 24590
    },
    {
      "entropy": 5.34642972946167,
      "epoch": 2.362632084534102,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004443030895807907,
      "loss": 4.8028,
      "mean_token_accuracy": 0.24331697076559067,
      "num_tokens": 56412547.0,
      "step": 24595
    },
    {
      "entropy": 5.398087215423584,
      "epoch": 2.363112391930836,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044428050709339117,
      "loss": 4.8082,
      "mean_token_accuracy": 0.24045567810535431,
      "num_tokens": 56423757.0,
      "step": 24600
    },
    {
      "entropy": 5.264205455780029,
      "epoch": 2.3635926993275698,
      "grad_norm": 1.109375,
      "learning_rate": 0.000444257920675801,
      "loss": 4.7457,
      "mean_token_accuracy": 0.24282570481300353,
      "num_tokens": 56433936.0,
      "step": 24605
    },
    {
      "entropy": 5.228886413574219,
      "epoch": 2.3640730067243036,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044423533032854454,
      "loss": 4.6759,
      "mean_token_accuracy": 0.24755157828330993,
      "num_tokens": 56446504.0,
      "step": 24610
    },
    {
      "entropy": 5.402956151962281,
      "epoch": 2.3645533141210375,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004442127360521462,
      "loss": 4.8094,
      "mean_token_accuracy": 0.23901001513004302,
      "num_tokens": 56459019.0,
      "step": 24615
    },
    {
      "entropy": 5.324404764175415,
      "epoch": 2.3650336215177714,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004441901378471306,
      "loss": 4.7366,
      "mean_token_accuracy": 0.23820848912000656,
      "num_tokens": 56470208.0,
      "step": 24620
    },
    {
      "entropy": 5.290033769607544,
      "epoch": 2.3655139289145053,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044416753571402233,
      "loss": 4.7433,
      "mean_token_accuracy": 0.23930860161781312,
      "num_tokens": 56482188.0,
      "step": 24625
    },
    {
      "entropy": 5.253919219970703,
      "epoch": 2.365994236311239,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004441449296533462,
      "loss": 4.6902,
      "mean_token_accuracy": 0.2419720023870468,
      "num_tokens": 56493599.0,
      "step": 24630
    },
    {
      "entropy": 5.350423574447632,
      "epoch": 2.366474543707973,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044412231966562717,
      "loss": 4.8314,
      "mean_token_accuracy": 0.23774529993534088,
      "num_tokens": 56504448.0,
      "step": 24635
    },
    {
      "entropy": 5.322364473342896,
      "epoch": 2.366954851104707,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044409970575139,
      "loss": 4.7613,
      "mean_token_accuracy": 0.2395089790225029,
      "num_tokens": 56517361.0,
      "step": 24640
    },
    {
      "entropy": 5.2697406768798825,
      "epoch": 2.367435158501441,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004440770879111598,
      "loss": 4.6074,
      "mean_token_accuracy": 0.2534013271331787,
      "num_tokens": 56530152.0,
      "step": 24645
    },
    {
      "entropy": 5.399232339859009,
      "epoch": 2.3679154658981747,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044405446614546163,
      "loss": 4.8213,
      "mean_token_accuracy": 0.2416141226887703,
      "num_tokens": 56541121.0,
      "step": 24650
    },
    {
      "entropy": 5.26408634185791,
      "epoch": 2.3683957732949086,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004440318404548208,
      "loss": 4.6382,
      "mean_token_accuracy": 0.2528198048472404,
      "num_tokens": 56551697.0,
      "step": 24655
    },
    {
      "entropy": 5.434298849105835,
      "epoch": 2.3688760806916425,
      "grad_norm": 1.125,
      "learning_rate": 0.00044400921083976246,
      "loss": 4.9193,
      "mean_token_accuracy": 0.2276859387755394,
      "num_tokens": 56564250.0,
      "step": 24660
    },
    {
      "entropy": 5.372978734970093,
      "epoch": 2.3693563880883763,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004439865773008122,
      "loss": 4.7531,
      "mean_token_accuracy": 0.24134962558746337,
      "num_tokens": 56574963.0,
      "step": 24665
    },
    {
      "entropy": 5.282676792144775,
      "epoch": 2.3698366954851107,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004439639398384953,
      "loss": 4.711,
      "mean_token_accuracy": 0.25025250017642975,
      "num_tokens": 56585516.0,
      "step": 24670
    },
    {
      "entropy": 5.243211555480957,
      "epoch": 2.3703170028818445,
      "grad_norm": 1.125,
      "learning_rate": 0.00044394129845333756,
      "loss": 4.6431,
      "mean_token_accuracy": 0.25044596791267393,
      "num_tokens": 56597459.0,
      "step": 24675
    },
    {
      "entropy": 5.366539764404297,
      "epoch": 2.3707973102785784,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004439186531458645,
      "loss": 4.7387,
      "mean_token_accuracy": 0.2466416910290718,
      "num_tokens": 56608942.0,
      "step": 24680
    },
    {
      "entropy": 5.346301317214966,
      "epoch": 2.3712776176753123,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044389600391660185,
      "loss": 4.8009,
      "mean_token_accuracy": 0.23713631331920623,
      "num_tokens": 56621444.0,
      "step": 24685
    },
    {
      "entropy": 5.305496120452881,
      "epoch": 2.371757925072046,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044387335076607554,
      "loss": 4.6636,
      "mean_token_accuracy": 0.24667196422815324,
      "num_tokens": 56631734.0,
      "step": 24690
    },
    {
      "entropy": 5.261005496978759,
      "epoch": 2.37223823246878,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004438506936948115,
      "loss": 4.7051,
      "mean_token_accuracy": 0.24396854937076567,
      "num_tokens": 56644140.0,
      "step": 24695
    },
    {
      "entropy": 5.2474618434906,
      "epoch": 2.372718539865514,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044382803270333565,
      "loss": 4.6645,
      "mean_token_accuracy": 0.2521482765674591,
      "num_tokens": 56655071.0,
      "step": 24700
    },
    {
      "entropy": 5.312582063674927,
      "epoch": 2.373198847262248,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004438053677921743,
      "loss": 4.7037,
      "mean_token_accuracy": 0.248298479616642,
      "num_tokens": 56667434.0,
      "step": 24705
    },
    {
      "entropy": 5.367788696289063,
      "epoch": 2.3736791546589817,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044378269896185344,
      "loss": 4.7815,
      "mean_token_accuracy": 0.23650476038455964,
      "num_tokens": 56677675.0,
      "step": 24710
    },
    {
      "entropy": 5.2717485427856445,
      "epoch": 2.3741594620557156,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004437600262128996,
      "loss": 4.7183,
      "mean_token_accuracy": 0.24639626145362853,
      "num_tokens": 56687849.0,
      "step": 24715
    },
    {
      "entropy": 5.309392023086548,
      "epoch": 2.3746397694524495,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000443737349545839,
      "loss": 4.7197,
      "mean_token_accuracy": 0.2439948484301567,
      "num_tokens": 56699637.0,
      "step": 24720
    },
    {
      "entropy": 5.399775743484497,
      "epoch": 2.3751200768491834,
      "grad_norm": 1.25,
      "learning_rate": 0.00044371466896119823,
      "loss": 4.7892,
      "mean_token_accuracy": 0.24400611966848373,
      "num_tokens": 56711412.0,
      "step": 24725
    },
    {
      "entropy": 5.40480375289917,
      "epoch": 2.3756003842459172,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044369198445950384,
      "loss": 4.8086,
      "mean_token_accuracy": 0.24251364171504974,
      "num_tokens": 56723379.0,
      "step": 24730
    },
    {
      "entropy": 5.365699720382691,
      "epoch": 2.376080691642651,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004436692960412824,
      "loss": 4.8316,
      "mean_token_accuracy": 0.24031717479228973,
      "num_tokens": 56734936.0,
      "step": 24735
    },
    {
      "entropy": 5.2785743236541744,
      "epoch": 2.376560999039385,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004436466037070608,
      "loss": 4.6678,
      "mean_token_accuracy": 0.256518816947937,
      "num_tokens": 56746925.0,
      "step": 24740
    },
    {
      "entropy": 5.351843595504761,
      "epoch": 2.3770413064361193,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044362390745736585,
      "loss": 4.8479,
      "mean_token_accuracy": 0.2421711951494217,
      "num_tokens": 56759034.0,
      "step": 24745
    },
    {
      "entropy": 5.356195068359375,
      "epoch": 2.377521613832853,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004436012072927245,
      "loss": 4.747,
      "mean_token_accuracy": 0.24249434918165208,
      "num_tokens": 56770882.0,
      "step": 24750
    },
    {
      "entropy": 5.288758754730225,
      "epoch": 2.378001921229587,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044357850321366375,
      "loss": 4.681,
      "mean_token_accuracy": 0.25317375361919403,
      "num_tokens": 56782110.0,
      "step": 24755
    },
    {
      "entropy": 5.301182794570923,
      "epoch": 2.378482228626321,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004435557952207107,
      "loss": 4.7091,
      "mean_token_accuracy": 0.2440403550863266,
      "num_tokens": 56793689.0,
      "step": 24760
    },
    {
      "entropy": 5.253468179702759,
      "epoch": 2.378962536023055,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044353308331439257,
      "loss": 4.7189,
      "mean_token_accuracy": 0.24761935472488403,
      "num_tokens": 56805315.0,
      "step": 24765
    },
    {
      "entropy": 5.3520286083221436,
      "epoch": 2.3794428434197887,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004435103674952367,
      "loss": 4.8043,
      "mean_token_accuracy": 0.23752714842557907,
      "num_tokens": 56816738.0,
      "step": 24770
    },
    {
      "entropy": 5.424173355102539,
      "epoch": 2.3799231508165226,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044348764776377047,
      "loss": 4.9294,
      "mean_token_accuracy": 0.23300515562295915,
      "num_tokens": 56828613.0,
      "step": 24775
    },
    {
      "entropy": 5.400389671325684,
      "epoch": 2.3804034582132565,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004434649241205214,
      "loss": 4.7619,
      "mean_token_accuracy": 0.2414322316646576,
      "num_tokens": 56840373.0,
      "step": 24780
    },
    {
      "entropy": 5.377579975128174,
      "epoch": 2.3808837656099904,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044344219656601704,
      "loss": 4.812,
      "mean_token_accuracy": 0.23200325816869735,
      "num_tokens": 56850581.0,
      "step": 24785
    },
    {
      "entropy": 5.3050120830535885,
      "epoch": 2.3813640730067243,
      "grad_norm": 1.078125,
      "learning_rate": 0.000443419465100785,
      "loss": 4.7171,
      "mean_token_accuracy": 0.25232468396425245,
      "num_tokens": 56861576.0,
      "step": 24790
    },
    {
      "entropy": 5.362205600738525,
      "epoch": 2.381844380403458,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004433967297253531,
      "loss": 4.829,
      "mean_token_accuracy": 0.2457115799188614,
      "num_tokens": 56872833.0,
      "step": 24795
    },
    {
      "entropy": 5.336214399337768,
      "epoch": 2.382324687800192,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044337399044024924,
      "loss": 4.7769,
      "mean_token_accuracy": 0.2474267989397049,
      "num_tokens": 56884367.0,
      "step": 24800
    },
    {
      "entropy": 5.236234712600708,
      "epoch": 2.382804995196926,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004433512472460012,
      "loss": 4.633,
      "mean_token_accuracy": 0.2524049639701843,
      "num_tokens": 56895233.0,
      "step": 24805
    },
    {
      "entropy": 5.241886568069458,
      "epoch": 2.38328530259366,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044332850014313713,
      "loss": 4.7542,
      "mean_token_accuracy": 0.2420166626572609,
      "num_tokens": 56907133.0,
      "step": 24810
    },
    {
      "entropy": 5.394517469406128,
      "epoch": 2.3837656099903937,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004433057491321851,
      "loss": 4.7876,
      "mean_token_accuracy": 0.246514031291008,
      "num_tokens": 56919372.0,
      "step": 24815
    },
    {
      "entropy": 5.363408994674683,
      "epoch": 2.384245917387128,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044328299421367333,
      "loss": 4.7961,
      "mean_token_accuracy": 0.23655386120080948,
      "num_tokens": 56930523.0,
      "step": 24820
    },
    {
      "entropy": 5.2548243522644045,
      "epoch": 2.3847262247838614,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004432602353881302,
      "loss": 4.6467,
      "mean_token_accuracy": 0.2473378747701645,
      "num_tokens": 56941324.0,
      "step": 24825
    },
    {
      "entropy": 5.18313512802124,
      "epoch": 2.3852065321805958,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044323747265608395,
      "loss": 4.6343,
      "mean_token_accuracy": 0.2520028933882713,
      "num_tokens": 56952359.0,
      "step": 24830
    },
    {
      "entropy": 5.300753021240235,
      "epoch": 2.3856868395773296,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004432147060180632,
      "loss": 4.7039,
      "mean_token_accuracy": 0.24968957155942917,
      "num_tokens": 56963456.0,
      "step": 24835
    },
    {
      "entropy": 5.23161735534668,
      "epoch": 2.3861671469740635,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044319193547459645,
      "loss": 4.6299,
      "mean_token_accuracy": 0.24805161505937576,
      "num_tokens": 56975029.0,
      "step": 24840
    },
    {
      "entropy": 5.272592163085937,
      "epoch": 2.3866474543707974,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004431691610262124,
      "loss": 4.7303,
      "mean_token_accuracy": 0.2413897469639778,
      "num_tokens": 56987159.0,
      "step": 24845
    },
    {
      "entropy": 5.340608263015747,
      "epoch": 2.3871277617675313,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044314638267343976,
      "loss": 4.8026,
      "mean_token_accuracy": 0.23855309784412385,
      "num_tokens": 56998359.0,
      "step": 24850
    },
    {
      "entropy": 5.316206264495849,
      "epoch": 2.387608069164265,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004431236004168075,
      "loss": 4.7562,
      "mean_token_accuracy": 0.23839999884366989,
      "num_tokens": 57009980.0,
      "step": 24855
    },
    {
      "entropy": 5.294656133651733,
      "epoch": 2.388088376560999,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004431008142568444,
      "loss": 4.7229,
      "mean_token_accuracy": 0.24308698922395705,
      "num_tokens": 57022531.0,
      "step": 24860
    },
    {
      "entropy": 5.357480382919311,
      "epoch": 2.388568683957733,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044307802419407954,
      "loss": 4.8265,
      "mean_token_accuracy": 0.23649962544441222,
      "num_tokens": 57034746.0,
      "step": 24865
    },
    {
      "entropy": 5.286729526519776,
      "epoch": 2.389048991354467,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004430552302290421,
      "loss": 4.7819,
      "mean_token_accuracy": 0.23839926421642305,
      "num_tokens": 57046496.0,
      "step": 24870
    },
    {
      "entropy": 5.241340208053589,
      "epoch": 2.3895292987512007,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004430324323622611,
      "loss": 4.6977,
      "mean_token_accuracy": 0.23738086968660355,
      "num_tokens": 57059083.0,
      "step": 24875
    },
    {
      "entropy": 5.37035870552063,
      "epoch": 2.3900096061479346,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044300963059426605,
      "loss": 4.7276,
      "mean_token_accuracy": 0.2481667920947075,
      "num_tokens": 57070077.0,
      "step": 24880
    },
    {
      "entropy": 5.34260630607605,
      "epoch": 2.3904899135446684,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044298682492558637,
      "loss": 4.6833,
      "mean_token_accuracy": 0.24787929952144622,
      "num_tokens": 57081397.0,
      "step": 24885
    },
    {
      "entropy": 5.315635919570923,
      "epoch": 2.3909702209414023,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044296401535675136,
      "loss": 4.7649,
      "mean_token_accuracy": 0.24136567264795303,
      "num_tokens": 57093882.0,
      "step": 24890
    },
    {
      "entropy": 5.387095928192139,
      "epoch": 2.3914505283381366,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044294120188829056,
      "loss": 4.8356,
      "mean_token_accuracy": 0.23451861292123793,
      "num_tokens": 57104603.0,
      "step": 24895
    },
    {
      "entropy": 5.394546413421631,
      "epoch": 2.39193083573487,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004429183845207339,
      "loss": 4.792,
      "mean_token_accuracy": 0.24283212572336196,
      "num_tokens": 57115512.0,
      "step": 24900
    },
    {
      "entropy": 5.336180830001831,
      "epoch": 2.3924111431316044,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004428955632546108,
      "loss": 4.7348,
      "mean_token_accuracy": 0.24574765264987947,
      "num_tokens": 57125934.0,
      "step": 24905
    },
    {
      "entropy": 5.285605573654175,
      "epoch": 2.3928914505283383,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004428727380904514,
      "loss": 4.6501,
      "mean_token_accuracy": 0.25493136793375015,
      "num_tokens": 57136437.0,
      "step": 24910
    },
    {
      "entropy": 5.306923294067383,
      "epoch": 2.393371757925072,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044284990902878545,
      "loss": 4.7403,
      "mean_token_accuracy": 0.23555824905633926,
      "num_tokens": 57147542.0,
      "step": 24915
    },
    {
      "entropy": 5.271802473068237,
      "epoch": 2.393852065321806,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044282707607014304,
      "loss": 4.7197,
      "mean_token_accuracy": 0.24439281672239305,
      "num_tokens": 57159248.0,
      "step": 24920
    },
    {
      "entropy": 5.312189817428589,
      "epoch": 2.39433237271854,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044280423921505427,
      "loss": 4.7241,
      "mean_token_accuracy": 0.2419796958565712,
      "num_tokens": 57171627.0,
      "step": 24925
    },
    {
      "entropy": 5.369476747512818,
      "epoch": 2.394812680115274,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004427813984640493,
      "loss": 4.7406,
      "mean_token_accuracy": 0.2477199450135231,
      "num_tokens": 57181604.0,
      "step": 24930
    },
    {
      "entropy": 5.249745607376099,
      "epoch": 2.3952929875120077,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004427585538176585,
      "loss": 4.662,
      "mean_token_accuracy": 0.24878149926662446,
      "num_tokens": 57191409.0,
      "step": 24935
    },
    {
      "entropy": 5.222042512893677,
      "epoch": 2.3957732949087416,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00044273570527641223,
      "loss": 4.6774,
      "mean_token_accuracy": 0.2496369257569313,
      "num_tokens": 57204606.0,
      "step": 24940
    },
    {
      "entropy": 5.382649707794189,
      "epoch": 2.3962536023054755,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044271285284084097,
      "loss": 4.7974,
      "mean_token_accuracy": 0.24180198907852174,
      "num_tokens": 57215641.0,
      "step": 24945
    },
    {
      "entropy": 5.347118377685547,
      "epoch": 2.3967339097022093,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004426899965114752,
      "loss": 4.7769,
      "mean_token_accuracy": 0.23654577434062957,
      "num_tokens": 57226520.0,
      "step": 24950
    },
    {
      "entropy": 5.3507496356964115,
      "epoch": 2.3972142170989432,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044266713628884566,
      "loss": 4.8078,
      "mean_token_accuracy": 0.23640327900648117,
      "num_tokens": 57237859.0,
      "step": 24955
    },
    {
      "entropy": 5.327178287506103,
      "epoch": 2.397694524495677,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044264427217348315,
      "loss": 4.6662,
      "mean_token_accuracy": 0.25302400141954423,
      "num_tokens": 57248727.0,
      "step": 24960
    },
    {
      "entropy": 5.283007287979126,
      "epoch": 2.398174831892411,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004426214041659184,
      "loss": 4.7128,
      "mean_token_accuracy": 0.2495455786585808,
      "num_tokens": 57259644.0,
      "step": 24965
    },
    {
      "entropy": 5.358208703994751,
      "epoch": 2.398655139289145,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004425985322666824,
      "loss": 4.7721,
      "mean_token_accuracy": 0.23833487629890443,
      "num_tokens": 57271140.0,
      "step": 24970
    },
    {
      "entropy": 5.370613861083984,
      "epoch": 2.3991354466858787,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004425756564763061,
      "loss": 4.731,
      "mean_token_accuracy": 0.2471403658390045,
      "num_tokens": 57281205.0,
      "step": 24975
    },
    {
      "entropy": 5.303550386428833,
      "epoch": 2.399615754082613,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044255277679532075,
      "loss": 4.7183,
      "mean_token_accuracy": 0.24629422426223754,
      "num_tokens": 57292056.0,
      "step": 24980
    },
    {
      "entropy": 5.311012172698975,
      "epoch": 2.400096061479347,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044252989322425735,
      "loss": 4.7037,
      "mean_token_accuracy": 0.24770190417766572,
      "num_tokens": 57303444.0,
      "step": 24985
    },
    {
      "entropy": 5.3094964027404785,
      "epoch": 2.400576368876081,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044250700576364734,
      "loss": 4.723,
      "mean_token_accuracy": 0.241867034137249,
      "num_tokens": 57314784.0,
      "step": 24990
    },
    {
      "entropy": 5.259481477737427,
      "epoch": 2.4010566762728147,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000442484114414022,
      "loss": 4.6594,
      "mean_token_accuracy": 0.2501977175474167,
      "num_tokens": 57325896.0,
      "step": 24995
    },
    {
      "entropy": 5.272733116149903,
      "epoch": 2.4015369836695486,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004424612191759129,
      "loss": 4.755,
      "mean_token_accuracy": 0.24074311852455138,
      "num_tokens": 57338338.0,
      "step": 25000
    },
    {
      "entropy": 5.343039464950562,
      "epoch": 2.4020172910662825,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004424383200498515,
      "loss": 4.7578,
      "mean_token_accuracy": 0.24488132148981095,
      "num_tokens": 57350038.0,
      "step": 25005
    },
    {
      "entropy": 5.355789613723755,
      "epoch": 2.4024975984630164,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004424154170363696,
      "loss": 4.805,
      "mean_token_accuracy": 0.23854247629642486,
      "num_tokens": 57362371.0,
      "step": 25010
    },
    {
      "entropy": 5.361563873291016,
      "epoch": 2.4029779058597502,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004423925101359987,
      "loss": 4.7714,
      "mean_token_accuracy": 0.24081692099571228,
      "num_tokens": 57374032.0,
      "step": 25015
    },
    {
      "entropy": 5.288333129882813,
      "epoch": 2.403458213256484,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004423695993492709,
      "loss": 4.6897,
      "mean_token_accuracy": 0.24492595046758653,
      "num_tokens": 57385705.0,
      "step": 25020
    },
    {
      "entropy": 5.258714008331299,
      "epoch": 2.403938520653218,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004423466846767179,
      "loss": 4.6703,
      "mean_token_accuracy": 0.2541018784046173,
      "num_tokens": 57396863.0,
      "step": 25025
    },
    {
      "entropy": 5.450971412658691,
      "epoch": 2.404418828049952,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044232376611887185,
      "loss": 4.9439,
      "mean_token_accuracy": 0.228681680560112,
      "num_tokens": 57408486.0,
      "step": 25030
    },
    {
      "entropy": 5.388098478317261,
      "epoch": 2.4048991354466858,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044230084367626477,
      "loss": 4.749,
      "mean_token_accuracy": 0.2412605866789818,
      "num_tokens": 57419899.0,
      "step": 25035
    },
    {
      "entropy": 5.377523517608642,
      "epoch": 2.4053794428434196,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004422779173494288,
      "loss": 4.7275,
      "mean_token_accuracy": 0.24325777888298034,
      "num_tokens": 57431620.0,
      "step": 25040
    },
    {
      "entropy": 5.269879674911499,
      "epoch": 2.4058597502401535,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004422549871388965,
      "loss": 4.7426,
      "mean_token_accuracy": 0.23894322216510772,
      "num_tokens": 57443985.0,
      "step": 25045
    },
    {
      "entropy": 5.314147424697876,
      "epoch": 2.4063400576368874,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044223205304519994,
      "loss": 4.7343,
      "mean_token_accuracy": 0.2431662157177925,
      "num_tokens": 57455160.0,
      "step": 25050
    },
    {
      "entropy": 5.32756552696228,
      "epoch": 2.4068203650336217,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004422091150688717,
      "loss": 4.7562,
      "mean_token_accuracy": 0.24914292842149735,
      "num_tokens": 57467557.0,
      "step": 25055
    },
    {
      "entropy": 5.382518482208252,
      "epoch": 2.4073006724303556,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004421861732104443,
      "loss": 4.7837,
      "mean_token_accuracy": 0.24013908058404923,
      "num_tokens": 57478749.0,
      "step": 25060
    },
    {
      "entropy": 5.289538097381592,
      "epoch": 2.4077809798270895,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004421632274704504,
      "loss": 4.705,
      "mean_token_accuracy": 0.24280474036931993,
      "num_tokens": 57489899.0,
      "step": 25065
    },
    {
      "entropy": 5.378178691864013,
      "epoch": 2.4082612872238234,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004421402778494227,
      "loss": 4.7629,
      "mean_token_accuracy": 0.24052033871412276,
      "num_tokens": 57502086.0,
      "step": 25070
    },
    {
      "entropy": 5.304590463638306,
      "epoch": 2.4087415946205573,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004421173243478941,
      "loss": 4.6757,
      "mean_token_accuracy": 0.24463892877101898,
      "num_tokens": 57513555.0,
      "step": 25075
    },
    {
      "entropy": 5.369786167144776,
      "epoch": 2.409221902017291,
      "grad_norm": 1.125,
      "learning_rate": 0.00044209436696639745,
      "loss": 4.7528,
      "mean_token_accuracy": 0.24104636311531066,
      "num_tokens": 57525100.0,
      "step": 25080
    },
    {
      "entropy": 5.308739709854126,
      "epoch": 2.409702209414025,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044207140570546574,
      "loss": 4.7479,
      "mean_token_accuracy": 0.24346123337745668,
      "num_tokens": 57536355.0,
      "step": 25085
    },
    {
      "entropy": 5.287871074676514,
      "epoch": 2.410182516810759,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044204844056563216,
      "loss": 4.693,
      "mean_token_accuracy": 0.24295339733362198,
      "num_tokens": 57547260.0,
      "step": 25090
    },
    {
      "entropy": 5.316242790222168,
      "epoch": 2.410662824207493,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004420254715474297,
      "loss": 4.7248,
      "mean_token_accuracy": 0.24350830018520356,
      "num_tokens": 57558299.0,
      "step": 25095
    },
    {
      "entropy": 5.254494476318359,
      "epoch": 2.4111431316042267,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044200249865139187,
      "loss": 4.7344,
      "mean_token_accuracy": 0.24233859926462173,
      "num_tokens": 57569393.0,
      "step": 25100
    },
    {
      "entropy": 5.335854005813599,
      "epoch": 2.4116234390009605,
      "grad_norm": 1.125,
      "learning_rate": 0.00044197952187805185,
      "loss": 4.7767,
      "mean_token_accuracy": 0.23999830782413484,
      "num_tokens": 57580585.0,
      "step": 25105
    },
    {
      "entropy": 5.394869565963745,
      "epoch": 2.4121037463976944,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044195654122794324,
      "loss": 4.8086,
      "mean_token_accuracy": 0.24255528748035432,
      "num_tokens": 57592887.0,
      "step": 25110
    },
    {
      "entropy": 5.400080585479737,
      "epoch": 2.4125840537944283,
      "grad_norm": 1.125,
      "learning_rate": 0.0004419335567015994,
      "loss": 4.8016,
      "mean_token_accuracy": 0.2378072440624237,
      "num_tokens": 57605129.0,
      "step": 25115
    },
    {
      "entropy": 5.195873546600342,
      "epoch": 2.413064361191162,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004419105682995542,
      "loss": 4.6376,
      "mean_token_accuracy": 0.2507411792874336,
      "num_tokens": 57616852.0,
      "step": 25120
    },
    {
      "entropy": 5.290833806991577,
      "epoch": 2.413544668587896,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004418875760223411,
      "loss": 4.7609,
      "mean_token_accuracy": 0.23958763480186462,
      "num_tokens": 57629808.0,
      "step": 25125
    },
    {
      "entropy": 5.403853559494019,
      "epoch": 2.4140249759846304,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044186457987049405,
      "loss": 4.8055,
      "mean_token_accuracy": 0.24072101712226868,
      "num_tokens": 57641025.0,
      "step": 25130
    },
    {
      "entropy": 5.316455936431884,
      "epoch": 2.414505283381364,
      "grad_norm": 1.109375,
      "learning_rate": 0.000441841579844547,
      "loss": 4.6974,
      "mean_token_accuracy": 0.24361866414546968,
      "num_tokens": 57652240.0,
      "step": 25135
    },
    {
      "entropy": 5.199885034561158,
      "epoch": 2.414985590778098,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004418185759450338,
      "loss": 4.658,
      "mean_token_accuracy": 0.2586354464292526,
      "num_tokens": 57663842.0,
      "step": 25140
    },
    {
      "entropy": 5.2571769714355465,
      "epoch": 2.415465898174832,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004417955681724887,
      "loss": 4.6621,
      "mean_token_accuracy": 0.2524684131145477,
      "num_tokens": 57674299.0,
      "step": 25145
    },
    {
      "entropy": 5.247325754165649,
      "epoch": 2.415946205571566,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044177255652744576,
      "loss": 4.6947,
      "mean_token_accuracy": 0.24355719834566117,
      "num_tokens": 57684858.0,
      "step": 25150
    },
    {
      "entropy": 5.375805330276489,
      "epoch": 2.4164265129683,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044174954101043926,
      "loss": 4.7823,
      "mean_token_accuracy": 0.23420143574476243,
      "num_tokens": 57696542.0,
      "step": 25155
    },
    {
      "entropy": 5.331563901901245,
      "epoch": 2.4169068203650337,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044172652162200354,
      "loss": 4.6554,
      "mean_token_accuracy": 0.25573578029870986,
      "num_tokens": 57708603.0,
      "step": 25160
    },
    {
      "entropy": 5.288554239273071,
      "epoch": 2.4173871277617676,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004417034983626731,
      "loss": 4.7488,
      "mean_token_accuracy": 0.24313832223415374,
      "num_tokens": 57720273.0,
      "step": 25165
    },
    {
      "entropy": 5.327586936950683,
      "epoch": 2.4178674351585014,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004416804712329825,
      "loss": 4.6825,
      "mean_token_accuracy": 0.24420003443956376,
      "num_tokens": 57730915.0,
      "step": 25170
    },
    {
      "entropy": 5.267145347595215,
      "epoch": 2.4183477425552353,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044165744023346614,
      "loss": 4.6842,
      "mean_token_accuracy": 0.25624350607395174,
      "num_tokens": 57743379.0,
      "step": 25175
    },
    {
      "entropy": 5.238912200927734,
      "epoch": 2.418828049951969,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044163440536465904,
      "loss": 4.7435,
      "mean_token_accuracy": 0.24218726456165313,
      "num_tokens": 57754441.0,
      "step": 25180
    },
    {
      "entropy": 5.366569995880127,
      "epoch": 2.419308357348703,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044161136662709577,
      "loss": 4.7594,
      "mean_token_accuracy": 0.23845661282539368,
      "num_tokens": 57765701.0,
      "step": 25185
    },
    {
      "entropy": 5.373220920562744,
      "epoch": 2.419788664745437,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044158832402131133,
      "loss": 4.8292,
      "mean_token_accuracy": 0.24170973002910615,
      "num_tokens": 57776774.0,
      "step": 25190
    },
    {
      "entropy": 5.349977350234985,
      "epoch": 2.420268972142171,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044156527754784066,
      "loss": 4.7867,
      "mean_token_accuracy": 0.2462119311094284,
      "num_tokens": 57787618.0,
      "step": 25195
    },
    {
      "entropy": 5.411642551422119,
      "epoch": 2.4207492795389047,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044154222720721887,
      "loss": 4.8089,
      "mean_token_accuracy": 0.24109580665826796,
      "num_tokens": 57800133.0,
      "step": 25200
    },
    {
      "entropy": 5.324774122238159,
      "epoch": 2.421229586935639,
      "grad_norm": 1.046875,
      "learning_rate": 0.000441519172999981,
      "loss": 4.6886,
      "mean_token_accuracy": 0.24810314923524857,
      "num_tokens": 57812736.0,
      "step": 25205
    },
    {
      "entropy": 5.296200466156006,
      "epoch": 2.4217098943323725,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004414961149266625,
      "loss": 4.7108,
      "mean_token_accuracy": 0.2469944640994072,
      "num_tokens": 57824459.0,
      "step": 25210
    },
    {
      "entropy": 5.376498508453369,
      "epoch": 2.422190201729107,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044147305298779856,
      "loss": 4.7787,
      "mean_token_accuracy": 0.2427724227309227,
      "num_tokens": 57836292.0,
      "step": 25215
    },
    {
      "entropy": 5.299994945526123,
      "epoch": 2.4226705091258407,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004414499871839247,
      "loss": 4.7427,
      "mean_token_accuracy": 0.24081757366657258,
      "num_tokens": 57848197.0,
      "step": 25220
    },
    {
      "entropy": 5.30347580909729,
      "epoch": 2.4231508165225746,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004414269175155763,
      "loss": 4.7302,
      "mean_token_accuracy": 0.23920983374118804,
      "num_tokens": 57859491.0,
      "step": 25225
    },
    {
      "entropy": 5.274920129776001,
      "epoch": 2.4236311239193085,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004414038439832891,
      "loss": 4.6496,
      "mean_token_accuracy": 0.25318737924098966,
      "num_tokens": 57870199.0,
      "step": 25230
    },
    {
      "entropy": 5.320224380493164,
      "epoch": 2.4241114313160423,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004413807665875988,
      "loss": 4.8617,
      "mean_token_accuracy": 0.24047745764255524,
      "num_tokens": 57882059.0,
      "step": 25235
    },
    {
      "entropy": 5.319404935836792,
      "epoch": 2.4245917387127762,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044135768532904104,
      "loss": 4.744,
      "mean_token_accuracy": 0.2456517070531845,
      "num_tokens": 57892316.0,
      "step": 25240
    },
    {
      "entropy": 5.287698602676391,
      "epoch": 2.42507204610951,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000441334600208152,
      "loss": 4.6888,
      "mean_token_accuracy": 0.25140420347452164,
      "num_tokens": 57903273.0,
      "step": 25245
    },
    {
      "entropy": 5.341436100006104,
      "epoch": 2.425552353506244,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044131151122546724,
      "loss": 4.8173,
      "mean_token_accuracy": 0.23762896358966829,
      "num_tokens": 57914484.0,
      "step": 25250
    },
    {
      "entropy": 5.269693040847779,
      "epoch": 2.426032660902978,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044128841838152313,
      "loss": 4.646,
      "mean_token_accuracy": 0.24709265530109406,
      "num_tokens": 57925213.0,
      "step": 25255
    },
    {
      "entropy": 5.301406574249268,
      "epoch": 2.4265129682997117,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004412653216768558,
      "loss": 4.719,
      "mean_token_accuracy": 0.23942028731107712,
      "num_tokens": 57937422.0,
      "step": 25260
    },
    {
      "entropy": 5.31001844406128,
      "epoch": 2.4269932756964456,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004412422211120013,
      "loss": 4.7344,
      "mean_token_accuracy": 0.24815135896205903,
      "num_tokens": 57947576.0,
      "step": 25265
    },
    {
      "entropy": 5.292912912368775,
      "epoch": 2.4274735830931795,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004412191166874961,
      "loss": 4.7081,
      "mean_token_accuracy": 0.24855644702911378,
      "num_tokens": 57958998.0,
      "step": 25270
    },
    {
      "entropy": 5.3419169902801515,
      "epoch": 2.4279538904899134,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004411960084038766,
      "loss": 4.7923,
      "mean_token_accuracy": 0.24063989967107774,
      "num_tokens": 57970126.0,
      "step": 25275
    },
    {
      "entropy": 5.306371259689331,
      "epoch": 2.4284341978866473,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044117289626167917,
      "loss": 4.732,
      "mean_token_accuracy": 0.24650708138942717,
      "num_tokens": 57981523.0,
      "step": 25280
    },
    {
      "entropy": 5.300368213653565,
      "epoch": 2.428914505283381,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004411497802614406,
      "loss": 4.691,
      "mean_token_accuracy": 0.2474402442574501,
      "num_tokens": 57992113.0,
      "step": 25285
    },
    {
      "entropy": 5.357346296310425,
      "epoch": 2.4293948126801155,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004411266604036975,
      "loss": 4.7568,
      "mean_token_accuracy": 0.2368482917547226,
      "num_tokens": 58005163.0,
      "step": 25290
    },
    {
      "entropy": 5.370837926864624,
      "epoch": 2.4298751200768494,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044110353668898674,
      "loss": 4.7623,
      "mean_token_accuracy": 0.236603944003582,
      "num_tokens": 58015095.0,
      "step": 25295
    },
    {
      "entropy": 5.3170051097869875,
      "epoch": 2.4303554274735832,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004410804091178449,
      "loss": 4.6825,
      "mean_token_accuracy": 0.24206698983907698,
      "num_tokens": 58025522.0,
      "step": 25300
    },
    {
      "entropy": 5.298336362838745,
      "epoch": 2.430835734870317,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004410572776908092,
      "loss": 4.75,
      "mean_token_accuracy": 0.23824716061353685,
      "num_tokens": 58036567.0,
      "step": 25305
    },
    {
      "entropy": 5.303626918792725,
      "epoch": 2.431316042267051,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044103414240841664,
      "loss": 4.6927,
      "mean_token_accuracy": 0.24832114577293396,
      "num_tokens": 58048321.0,
      "step": 25310
    },
    {
      "entropy": 5.3116217136383055,
      "epoch": 2.431796349663785,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004410110032712043,
      "loss": 4.7485,
      "mean_token_accuracy": 0.24119576960802078,
      "num_tokens": 58060505.0,
      "step": 25315
    },
    {
      "entropy": 5.250061893463135,
      "epoch": 2.4322766570605188,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004409878602797094,
      "loss": 4.6949,
      "mean_token_accuracy": 0.2538239449262619,
      "num_tokens": 58071926.0,
      "step": 25320
    },
    {
      "entropy": 5.315074634552002,
      "epoch": 2.4327569644572526,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044096471343446923,
      "loss": 4.7687,
      "mean_token_accuracy": 0.2478758916258812,
      "num_tokens": 58082963.0,
      "step": 25325
    },
    {
      "entropy": 5.334404945373535,
      "epoch": 2.4332372718539865,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004409415627360213,
      "loss": 4.7302,
      "mean_token_accuracy": 0.24438008964061736,
      "num_tokens": 58095343.0,
      "step": 25330
    },
    {
      "entropy": 5.261237382888794,
      "epoch": 2.4337175792507204,
      "grad_norm": 1.125,
      "learning_rate": 0.00044091840818490303,
      "loss": 4.6834,
      "mean_token_accuracy": 0.25101089775562285,
      "num_tokens": 58106763.0,
      "step": 25335
    },
    {
      "entropy": 5.264698219299317,
      "epoch": 2.4341978866474543,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044089524978165197,
      "loss": 4.6749,
      "mean_token_accuracy": 0.24682711362838744,
      "num_tokens": 58116942.0,
      "step": 25340
    },
    {
      "entropy": 5.224446153640747,
      "epoch": 2.434678194044188,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00044087208752680577,
      "loss": 4.6493,
      "mean_token_accuracy": 0.2566885486245155,
      "num_tokens": 58127365.0,
      "step": 25345
    },
    {
      "entropy": 5.363105726242066,
      "epoch": 2.435158501440922,
      "grad_norm": 1.125,
      "learning_rate": 0.0004408489214209023,
      "loss": 4.7853,
      "mean_token_accuracy": 0.23956508189439774,
      "num_tokens": 58137572.0,
      "step": 25350
    },
    {
      "entropy": 5.3921154022216795,
      "epoch": 2.435638808837656,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004408257514644793,
      "loss": 4.8453,
      "mean_token_accuracy": 0.24004860371351242,
      "num_tokens": 58150012.0,
      "step": 25355
    },
    {
      "entropy": 5.248890018463134,
      "epoch": 2.43611911623439,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044080257765807476,
      "loss": 4.5785,
      "mean_token_accuracy": 0.25723277777433395,
      "num_tokens": 58160611.0,
      "step": 25360
    },
    {
      "entropy": 5.360936021804809,
      "epoch": 2.436599423631124,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004407794000022267,
      "loss": 4.8488,
      "mean_token_accuracy": 0.22983704507350922,
      "num_tokens": 58172113.0,
      "step": 25365
    },
    {
      "entropy": 5.300959777832031,
      "epoch": 2.437079731027858,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004407562184974732,
      "loss": 4.7559,
      "mean_token_accuracy": 0.24200656563043593,
      "num_tokens": 58183118.0,
      "step": 25370
    },
    {
      "entropy": 5.274296236038208,
      "epoch": 2.437560038424592,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004407330331443526,
      "loss": 4.7221,
      "mean_token_accuracy": 0.24329158365726472,
      "num_tokens": 58194534.0,
      "step": 25375
    },
    {
      "entropy": 5.364896583557129,
      "epoch": 2.438040345821326,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000440709843943403,
      "loss": 4.79,
      "mean_token_accuracy": 0.2384905681014061,
      "num_tokens": 58205952.0,
      "step": 25380
    },
    {
      "entropy": 5.32056040763855,
      "epoch": 2.4385206532180597,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000440686650895163,
      "loss": 4.704,
      "mean_token_accuracy": 0.24888446778059006,
      "num_tokens": 58217811.0,
      "step": 25385
    },
    {
      "entropy": 5.398961305618286,
      "epoch": 2.4390009606147935,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044066345400017084,
      "loss": 4.8286,
      "mean_token_accuracy": 0.24138305932283402,
      "num_tokens": 58228658.0,
      "step": 25390
    },
    {
      "entropy": 5.295126152038574,
      "epoch": 2.4394812680115274,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044064025325896524,
      "loss": 4.7611,
      "mean_token_accuracy": 0.24430646747350693,
      "num_tokens": 58242130.0,
      "step": 25395
    },
    {
      "entropy": 5.334532308578491,
      "epoch": 2.4399615754082613,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044061704867208484,
      "loss": 4.7174,
      "mean_token_accuracy": 0.24795260280370712,
      "num_tokens": 58253533.0,
      "step": 25400
    },
    {
      "entropy": 5.327444839477539,
      "epoch": 2.440441882804995,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044059384024006825,
      "loss": 4.7454,
      "mean_token_accuracy": 0.241423162817955,
      "num_tokens": 58264381.0,
      "step": 25405
    },
    {
      "entropy": 5.320031213760376,
      "epoch": 2.440922190201729,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004405706279634545,
      "loss": 4.7921,
      "mean_token_accuracy": 0.23945262283086777,
      "num_tokens": 58275114.0,
      "step": 25410
    },
    {
      "entropy": 5.268980169296265,
      "epoch": 2.441402497598463,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00044054741184278243,
      "loss": 4.6846,
      "mean_token_accuracy": 0.2464488387107849,
      "num_tokens": 58287349.0,
      "step": 25415
    },
    {
      "entropy": 5.281464672088623,
      "epoch": 2.441882804995197,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044052419187859095,
      "loss": 4.6811,
      "mean_token_accuracy": 0.2559520170092583,
      "num_tokens": 58298787.0,
      "step": 25420
    },
    {
      "entropy": 5.315772247314453,
      "epoch": 2.4423631123919307,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004405009680714193,
      "loss": 4.79,
      "mean_token_accuracy": 0.23647845834493636,
      "num_tokens": 58311019.0,
      "step": 25425
    },
    {
      "entropy": 5.279531478881836,
      "epoch": 2.4428434197886646,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004404777404218065,
      "loss": 4.7119,
      "mean_token_accuracy": 0.24744521230459213,
      "num_tokens": 58322692.0,
      "step": 25430
    },
    {
      "entropy": 5.350232172012329,
      "epoch": 2.4433237271853985,
      "grad_norm": 1.125,
      "learning_rate": 0.000440454508930292,
      "loss": 4.769,
      "mean_token_accuracy": 0.24129513949155806,
      "num_tokens": 58333638.0,
      "step": 25435
    },
    {
      "entropy": 5.276310777664184,
      "epoch": 2.443804034582133,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004404312735974152,
      "loss": 4.7401,
      "mean_token_accuracy": 0.2404444247484207,
      "num_tokens": 58345291.0,
      "step": 25440
    },
    {
      "entropy": 5.356992101669311,
      "epoch": 2.4442843419788662,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044040803442371533,
      "loss": 4.794,
      "mean_token_accuracy": 0.24358667582273483,
      "num_tokens": 58357303.0,
      "step": 25445
    },
    {
      "entropy": 5.2982625484466555,
      "epoch": 2.4447646493756006,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004403847914097321,
      "loss": 4.7195,
      "mean_token_accuracy": 0.2450895056128502,
      "num_tokens": 58369804.0,
      "step": 25450
    },
    {
      "entropy": 5.311979627609253,
      "epoch": 2.4452449567723344,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044036154455600517,
      "loss": 4.7509,
      "mean_token_accuracy": 0.24426581114530563,
      "num_tokens": 58381027.0,
      "step": 25455
    },
    {
      "entropy": 5.32955756187439,
      "epoch": 2.4457252641690683,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004403382938630741,
      "loss": 4.7149,
      "mean_token_accuracy": 0.23961149901151657,
      "num_tokens": 58392165.0,
      "step": 25460
    },
    {
      "entropy": 5.351481866836548,
      "epoch": 2.446205571565802,
      "grad_norm": 1.125,
      "learning_rate": 0.00044031503933147887,
      "loss": 4.7359,
      "mean_token_accuracy": 0.24242564737796785,
      "num_tokens": 58402494.0,
      "step": 25465
    },
    {
      "entropy": 5.252609348297119,
      "epoch": 2.446685878962536,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044029178096175934,
      "loss": 4.6829,
      "mean_token_accuracy": 0.2483374923467636,
      "num_tokens": 58413735.0,
      "step": 25470
    },
    {
      "entropy": 5.266089820861817,
      "epoch": 2.44716618635927,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004402685187544554,
      "loss": 4.7265,
      "mean_token_accuracy": 0.2409334123134613,
      "num_tokens": 58425977.0,
      "step": 25475
    },
    {
      "entropy": 5.308540534973145,
      "epoch": 2.447646493756004,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004402452527101072,
      "loss": 4.7252,
      "mean_token_accuracy": 0.24392176568508148,
      "num_tokens": 58436845.0,
      "step": 25480
    },
    {
      "entropy": 5.284183931350708,
      "epoch": 2.4481268011527377,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004402219828292549,
      "loss": 4.6753,
      "mean_token_accuracy": 0.2521177351474762,
      "num_tokens": 58448113.0,
      "step": 25485
    },
    {
      "entropy": 5.313370656967163,
      "epoch": 2.4486071085494716,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004401987091124388,
      "loss": 4.7504,
      "mean_token_accuracy": 0.23798300474882125,
      "num_tokens": 58459226.0,
      "step": 25490
    },
    {
      "entropy": 5.359014225006104,
      "epoch": 2.4490874159462055,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004401754315601992,
      "loss": 4.7588,
      "mean_token_accuracy": 0.24397291690111161,
      "num_tokens": 58471354.0,
      "step": 25495
    },
    {
      "entropy": 5.340893316268921,
      "epoch": 2.4495677233429394,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004401521501730765,
      "loss": 4.7495,
      "mean_token_accuracy": 0.24825492650270461,
      "num_tokens": 58482859.0,
      "step": 25500
    },
    {
      "entropy": 5.364076042175293,
      "epoch": 2.4500480307396733,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044012886495161144,
      "loss": 4.7636,
      "mean_token_accuracy": 0.2383489117026329,
      "num_tokens": 58496351.0,
      "step": 25505
    },
    {
      "entropy": 5.320231771469116,
      "epoch": 2.450528338136407,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004401055758963443,
      "loss": 4.776,
      "mean_token_accuracy": 0.24675821363925934,
      "num_tokens": 58508055.0,
      "step": 25510
    },
    {
      "entropy": 5.316258144378662,
      "epoch": 2.451008645533141,
      "grad_norm": 1.0625,
      "learning_rate": 0.000440082283007816,
      "loss": 4.7098,
      "mean_token_accuracy": 0.2427606463432312,
      "num_tokens": 58519344.0,
      "step": 25515
    },
    {
      "entropy": 5.417582845687866,
      "epoch": 2.451488952929875,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044005898628656734,
      "loss": 4.8353,
      "mean_token_accuracy": 0.2340197056531906,
      "num_tokens": 58530141.0,
      "step": 25520
    },
    {
      "entropy": 5.345852518081665,
      "epoch": 2.4519692603266092,
      "grad_norm": 1.125,
      "learning_rate": 0.000440035685733139,
      "loss": 4.7117,
      "mean_token_accuracy": 0.2468792900443077,
      "num_tokens": 58541527.0,
      "step": 25525
    },
    {
      "entropy": 5.38276309967041,
      "epoch": 2.452449567723343,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004400123813480722,
      "loss": 4.8278,
      "mean_token_accuracy": 0.24342640042304992,
      "num_tokens": 58552205.0,
      "step": 25530
    },
    {
      "entropy": 5.295437860488891,
      "epoch": 2.452929875120077,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043998907313190787,
      "loss": 4.6965,
      "mean_token_accuracy": 0.24514851719141006,
      "num_tokens": 58563563.0,
      "step": 25535
    },
    {
      "entropy": 5.315817499160767,
      "epoch": 2.453410182516811,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004399657610851873,
      "loss": 4.7304,
      "mean_token_accuracy": 0.24284296482801437,
      "num_tokens": 58575605.0,
      "step": 25540
    },
    {
      "entropy": 5.385099220275879,
      "epoch": 2.4538904899135447,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043994244520845146,
      "loss": 4.854,
      "mean_token_accuracy": 0.2324341505765915,
      "num_tokens": 58587761.0,
      "step": 25545
    },
    {
      "entropy": 5.425414514541626,
      "epoch": 2.4543707973102786,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004399191255022418,
      "loss": 4.7097,
      "mean_token_accuracy": 0.24549288898706437,
      "num_tokens": 58598541.0,
      "step": 25550
    },
    {
      "entropy": 5.293618440628052,
      "epoch": 2.4548511047070125,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004398958019670998,
      "loss": 4.7304,
      "mean_token_accuracy": 0.24603156596422196,
      "num_tokens": 58610833.0,
      "step": 25555
    },
    {
      "entropy": 5.392272520065307,
      "epoch": 2.4553314121037464,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043987247460356696,
      "loss": 4.7843,
      "mean_token_accuracy": 0.2381092056632042,
      "num_tokens": 58620857.0,
      "step": 25560
    },
    {
      "entropy": 5.335360670089722,
      "epoch": 2.4558117195004803,
      "grad_norm": 1.125,
      "learning_rate": 0.0004398491434121847,
      "loss": 4.7547,
      "mean_token_accuracy": 0.24810679703950883,
      "num_tokens": 58631245.0,
      "step": 25565
    },
    {
      "entropy": 5.310660123825073,
      "epoch": 2.456292026897214,
      "grad_norm": 1.140625,
      "learning_rate": 0.000439825808393495,
      "loss": 4.7369,
      "mean_token_accuracy": 0.2515259176492691,
      "num_tokens": 58642995.0,
      "step": 25570
    },
    {
      "entropy": 5.255844402313232,
      "epoch": 2.456772334293948,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004398024695480394,
      "loss": 4.6818,
      "mean_token_accuracy": 0.2528055995702744,
      "num_tokens": 58654817.0,
      "step": 25575
    },
    {
      "entropy": 5.319554233551026,
      "epoch": 2.457252641690682,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004397791268763598,
      "loss": 4.7343,
      "mean_token_accuracy": 0.24372208267450332,
      "num_tokens": 58665912.0,
      "step": 25580
    },
    {
      "entropy": 5.315913343429566,
      "epoch": 2.457732949087416,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043975578037899814,
      "loss": 4.723,
      "mean_token_accuracy": 0.24786821901798248,
      "num_tokens": 58678080.0,
      "step": 25585
    },
    {
      "entropy": 5.324902582168579,
      "epoch": 2.4582132564841497,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004397324300564966,
      "loss": 4.6695,
      "mean_token_accuracy": 0.24725355505943297,
      "num_tokens": 58689902.0,
      "step": 25590
    },
    {
      "entropy": 5.3046801567077635,
      "epoch": 2.4586935638808836,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004397090759093971,
      "loss": 4.7016,
      "mean_token_accuracy": 0.24582220166921614,
      "num_tokens": 58701857.0,
      "step": 25595
    },
    {
      "entropy": 5.321281385421753,
      "epoch": 2.459173871277618,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043968571793824194,
      "loss": 4.7418,
      "mean_token_accuracy": 0.24548054933547975,
      "num_tokens": 58712775.0,
      "step": 25600
    },
    {
      "entropy": 5.309778928756714,
      "epoch": 2.4596541786743518,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004396623561435734,
      "loss": 4.6709,
      "mean_token_accuracy": 0.24500538408756256,
      "num_tokens": 58723583.0,
      "step": 25605
    },
    {
      "entropy": 5.376312351226806,
      "epoch": 2.4601344860710856,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004396389905259339,
      "loss": 4.8029,
      "mean_token_accuracy": 0.23986284881830217,
      "num_tokens": 58734187.0,
      "step": 25610
    },
    {
      "entropy": 5.303905916213989,
      "epoch": 2.4606147934678195,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043961562108586603,
      "loss": 4.7461,
      "mean_token_accuracy": 0.2365572616457939,
      "num_tokens": 58746138.0,
      "step": 25615
    },
    {
      "entropy": 5.381664514541626,
      "epoch": 2.4610951008645534,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043959224782391215,
      "loss": 4.8581,
      "mean_token_accuracy": 0.2387454867362976,
      "num_tokens": 58757765.0,
      "step": 25620
    },
    {
      "entropy": 5.495850944519043,
      "epoch": 2.4615754082612873,
      "grad_norm": 1.015625,
      "learning_rate": 0.000439568870740615,
      "loss": 4.9289,
      "mean_token_accuracy": 0.2210085093975067,
      "num_tokens": 58769957.0,
      "step": 25625
    },
    {
      "entropy": 5.435769510269165,
      "epoch": 2.462055715658021,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004395454898365174,
      "loss": 4.8398,
      "mean_token_accuracy": 0.23920833319425583,
      "num_tokens": 58781019.0,
      "step": 25630
    },
    {
      "entropy": 5.333022594451904,
      "epoch": 2.462536023054755,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043952210511216205,
      "loss": 4.7627,
      "mean_token_accuracy": 0.2468622624874115,
      "num_tokens": 58793427.0,
      "step": 25635
    },
    {
      "entropy": 5.37119688987732,
      "epoch": 2.463016330451489,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043949871656809205,
      "loss": 4.8098,
      "mean_token_accuracy": 0.23768462389707565,
      "num_tokens": 58804406.0,
      "step": 25640
    },
    {
      "entropy": 5.340428447723388,
      "epoch": 2.463496637848223,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043947532420485024,
      "loss": 4.7374,
      "mean_token_accuracy": 0.24678486287593843,
      "num_tokens": 58815783.0,
      "step": 25645
    },
    {
      "entropy": 5.364830207824707,
      "epoch": 2.4639769452449567,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004394519280229798,
      "loss": 4.8423,
      "mean_token_accuracy": 0.2400517702102661,
      "num_tokens": 58826968.0,
      "step": 25650
    },
    {
      "entropy": 5.331212759017944,
      "epoch": 2.4644572526416906,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043942852802302397,
      "loss": 4.7067,
      "mean_token_accuracy": 0.2456907108426094,
      "num_tokens": 58839090.0,
      "step": 25655
    },
    {
      "entropy": 5.313432455062866,
      "epoch": 2.4649375600384245,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004394051242055259,
      "loss": 4.6874,
      "mean_token_accuracy": 0.24315191954374313,
      "num_tokens": 58851636.0,
      "step": 25660
    },
    {
      "entropy": 5.35497350692749,
      "epoch": 2.4654178674351583,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004393817165710291,
      "loss": 4.7717,
      "mean_token_accuracy": 0.2458059534430504,
      "num_tokens": 58863777.0,
      "step": 25665
    },
    {
      "entropy": 5.322840309143066,
      "epoch": 2.465898174831892,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043935830512007687,
      "loss": 4.7541,
      "mean_token_accuracy": 0.23888702243566512,
      "num_tokens": 58875454.0,
      "step": 25670
    },
    {
      "entropy": 5.31825361251831,
      "epoch": 2.4663784822286265,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043933488985321286,
      "loss": 4.7045,
      "mean_token_accuracy": 0.2516183242201805,
      "num_tokens": 58887932.0,
      "step": 25675
    },
    {
      "entropy": 5.345677328109741,
      "epoch": 2.46685878962536,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004393114707709807,
      "loss": 4.7582,
      "mean_token_accuracy": 0.2339325577020645,
      "num_tokens": 58900167.0,
      "step": 25680
    },
    {
      "entropy": 5.2724871158599855,
      "epoch": 2.4673390970220943,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004392880478739241,
      "loss": 4.6772,
      "mean_token_accuracy": 0.2501827567815781,
      "num_tokens": 58911769.0,
      "step": 25685
    },
    {
      "entropy": 5.304034996032715,
      "epoch": 2.467819404418828,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004392646211625869,
      "loss": 4.7545,
      "mean_token_accuracy": 0.24374611526727677,
      "num_tokens": 58923747.0,
      "step": 25690
    },
    {
      "entropy": 5.373422145843506,
      "epoch": 2.468299711815562,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004392411906375129,
      "loss": 4.7671,
      "mean_token_accuracy": 0.2433740183711052,
      "num_tokens": 58934801.0,
      "step": 25695
    },
    {
      "entropy": 5.3240196228027346,
      "epoch": 2.468780019212296,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043921775629924615,
      "loss": 4.7392,
      "mean_token_accuracy": 0.24527580440044403,
      "num_tokens": 58946566.0,
      "step": 25700
    },
    {
      "entropy": 5.301040124893189,
      "epoch": 2.46926032660903,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043919431814833077,
      "loss": 4.7106,
      "mean_token_accuracy": 0.2434443339705467,
      "num_tokens": 58957066.0,
      "step": 25705
    },
    {
      "entropy": 5.278025913238525,
      "epoch": 2.4697406340057637,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043917087618531084,
      "loss": 4.6543,
      "mean_token_accuracy": 0.2529663428664207,
      "num_tokens": 58967553.0,
      "step": 25710
    },
    {
      "entropy": 5.342777347564697,
      "epoch": 2.4702209414024976,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004391474304107307,
      "loss": 4.7613,
      "mean_token_accuracy": 0.2459619954228401,
      "num_tokens": 58979038.0,
      "step": 25715
    },
    {
      "entropy": 5.333079099655151,
      "epoch": 2.4707012487992315,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043912398082513463,
      "loss": 4.714,
      "mean_token_accuracy": 0.24155760258436204,
      "num_tokens": 58989564.0,
      "step": 25720
    },
    {
      "entropy": 5.28240818977356,
      "epoch": 2.4711815561959654,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000439100527429067,
      "loss": 4.7454,
      "mean_token_accuracy": 0.24918168932199478,
      "num_tokens": 59001789.0,
      "step": 25725
    },
    {
      "entropy": 5.322026014328003,
      "epoch": 2.4716618635926992,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043907707022307243,
      "loss": 4.7586,
      "mean_token_accuracy": 0.24148885160684586,
      "num_tokens": 59013518.0,
      "step": 25730
    },
    {
      "entropy": 5.30854549407959,
      "epoch": 2.472142170989433,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043905360920769553,
      "loss": 4.6743,
      "mean_token_accuracy": 0.24739443510770798,
      "num_tokens": 59024739.0,
      "step": 25735
    },
    {
      "entropy": 5.339999341964722,
      "epoch": 2.472622478386167,
      "grad_norm": 1.0625,
      "learning_rate": 0.000439030144383481,
      "loss": 4.7125,
      "mean_token_accuracy": 0.24554741978645325,
      "num_tokens": 59036142.0,
      "step": 25740
    },
    {
      "entropy": 5.379006052017212,
      "epoch": 2.473102785782901,
      "grad_norm": 1.125,
      "learning_rate": 0.00043900667575097355,
      "loss": 4.7763,
      "mean_token_accuracy": 0.24523380398750305,
      "num_tokens": 59046675.0,
      "step": 25745
    },
    {
      "entropy": 5.345741510391235,
      "epoch": 2.473583093179635,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004389832033107181,
      "loss": 4.7102,
      "mean_token_accuracy": 0.24085948765277862,
      "num_tokens": 59057193.0,
      "step": 25750
    },
    {
      "entropy": 5.2872076511383055,
      "epoch": 2.4740634005763686,
      "grad_norm": 1.125,
      "learning_rate": 0.00043895972706325953,
      "loss": 4.7178,
      "mean_token_accuracy": 0.24243101626634597,
      "num_tokens": 59069501.0,
      "step": 25755
    },
    {
      "entropy": 5.356794357299805,
      "epoch": 2.474543707973103,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000438936247009143,
      "loss": 4.8538,
      "mean_token_accuracy": 0.23741938620805741,
      "num_tokens": 59080973.0,
      "step": 25760
    },
    {
      "entropy": 5.324906349182129,
      "epoch": 2.475024015369837,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043891276314891365,
      "loss": 4.7081,
      "mean_token_accuracy": 0.24455763399600983,
      "num_tokens": 59092541.0,
      "step": 25765
    },
    {
      "entropy": 5.2702559471130375,
      "epoch": 2.4755043227665707,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004388892754831166,
      "loss": 4.7093,
      "mean_token_accuracy": 0.2462815523147583,
      "num_tokens": 59103936.0,
      "step": 25770
    },
    {
      "entropy": 5.293495798110962,
      "epoch": 2.4759846301633046,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004388657840122973,
      "loss": 4.7327,
      "mean_token_accuracy": 0.24519333243370056,
      "num_tokens": 59114714.0,
      "step": 25775
    },
    {
      "entropy": 5.39832763671875,
      "epoch": 2.4764649375600385,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000438842288737001,
      "loss": 4.7529,
      "mean_token_accuracy": 0.24127866476774215,
      "num_tokens": 59126678.0,
      "step": 25780
    },
    {
      "entropy": 5.3149408340454105,
      "epoch": 2.4769452449567724,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043881878965777325,
      "loss": 4.7426,
      "mean_token_accuracy": 0.24553004503250123,
      "num_tokens": 59137732.0,
      "step": 25785
    },
    {
      "entropy": 5.281337833404541,
      "epoch": 2.4774255523535063,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043879528677515973,
      "loss": 4.7071,
      "mean_token_accuracy": 0.24374909102916717,
      "num_tokens": 59149231.0,
      "step": 25790
    },
    {
      "entropy": 5.406073999404907,
      "epoch": 2.47790585975024,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043877178008970596,
      "loss": 4.8428,
      "mean_token_accuracy": 0.2221219226717949,
      "num_tokens": 59161033.0,
      "step": 25795
    },
    {
      "entropy": 5.295688438415527,
      "epoch": 2.478386167146974,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004387482696019578,
      "loss": 4.6707,
      "mean_token_accuracy": 0.25502333492040635,
      "num_tokens": 59171906.0,
      "step": 25800
    },
    {
      "entropy": 5.340134334564209,
      "epoch": 2.478866474543708,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043872475531246105,
      "loss": 4.8231,
      "mean_token_accuracy": 0.2402649700641632,
      "num_tokens": 59185036.0,
      "step": 25805
    },
    {
      "entropy": 5.335235452651977,
      "epoch": 2.479346781940442,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043870123722176166,
      "loss": 4.7982,
      "mean_token_accuracy": 0.2416560485959053,
      "num_tokens": 59196397.0,
      "step": 25810
    },
    {
      "entropy": 5.381109952926636,
      "epoch": 2.4798270893371757,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004386777153304056,
      "loss": 4.8129,
      "mean_token_accuracy": 0.2365442097187042,
      "num_tokens": 59208180.0,
      "step": 25815
    },
    {
      "entropy": 5.467008113861084,
      "epoch": 2.4803073967339095,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043865418963893896,
      "loss": 4.8947,
      "mean_token_accuracy": 0.23552970141172408,
      "num_tokens": 59219651.0,
      "step": 25820
    },
    {
      "entropy": 5.234236097335815,
      "epoch": 2.4807877041306434,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004386306601479081,
      "loss": 4.6383,
      "mean_token_accuracy": 0.25324489623308183,
      "num_tokens": 59231327.0,
      "step": 25825
    },
    {
      "entropy": 5.296954107284546,
      "epoch": 2.4812680115273773,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004386071268578591,
      "loss": 4.7501,
      "mean_token_accuracy": 0.2398039534687996,
      "num_tokens": 59243606.0,
      "step": 25830
    },
    {
      "entropy": 5.361504554748535,
      "epoch": 2.4817483189241116,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043858358976933844,
      "loss": 4.7776,
      "mean_token_accuracy": 0.2350335732102394,
      "num_tokens": 59255824.0,
      "step": 25835
    },
    {
      "entropy": 5.322793817520141,
      "epoch": 2.4822286263208455,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043856004888289264,
      "loss": 4.6872,
      "mean_token_accuracy": 0.24600597470998764,
      "num_tokens": 59267405.0,
      "step": 25840
    },
    {
      "entropy": 5.3333124160766605,
      "epoch": 2.4827089337175794,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004385365041990681,
      "loss": 4.7715,
      "mean_token_accuracy": 0.23978419303894044,
      "num_tokens": 59279063.0,
      "step": 25845
    },
    {
      "entropy": 5.164304304122925,
      "epoch": 2.4831892411143133,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004385129557184116,
      "loss": 4.5412,
      "mean_token_accuracy": 0.25962081700563433,
      "num_tokens": 59289838.0,
      "step": 25850
    },
    {
      "entropy": 5.253693294525147,
      "epoch": 2.483669548511047,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043848940344146976,
      "loss": 4.6671,
      "mean_token_accuracy": 0.24894091486930847,
      "num_tokens": 59301505.0,
      "step": 25855
    },
    {
      "entropy": 5.2601540088653564,
      "epoch": 2.484149855907781,
      "grad_norm": 1.0,
      "learning_rate": 0.0004384658473687894,
      "loss": 4.6428,
      "mean_token_accuracy": 0.25309713929891586,
      "num_tokens": 59312934.0,
      "step": 25860
    },
    {
      "entropy": 5.342837238311768,
      "epoch": 2.484630163304515,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004384422875009176,
      "loss": 4.729,
      "mean_token_accuracy": 0.2470986559987068,
      "num_tokens": 59324104.0,
      "step": 25865
    },
    {
      "entropy": 5.3536763191223145,
      "epoch": 2.485110470701249,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004384187238384011,
      "loss": 4.7662,
      "mean_token_accuracy": 0.2439442753791809,
      "num_tokens": 59334978.0,
      "step": 25870
    },
    {
      "entropy": 5.289494943618775,
      "epoch": 2.4855907780979827,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004383951563817871,
      "loss": 4.6985,
      "mean_token_accuracy": 0.2487751916050911,
      "num_tokens": 59345779.0,
      "step": 25875
    },
    {
      "entropy": 5.422065687179566,
      "epoch": 2.4860710854947166,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004383715851316227,
      "loss": 4.8176,
      "mean_token_accuracy": 0.24043326824903488,
      "num_tokens": 59358870.0,
      "step": 25880
    },
    {
      "entropy": 5.425562524795533,
      "epoch": 2.4865513928914504,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043834801008845527,
      "loss": 4.8041,
      "mean_token_accuracy": 0.2361508995294571,
      "num_tokens": 59369831.0,
      "step": 25885
    },
    {
      "entropy": 5.2538093566894535,
      "epoch": 2.4870317002881843,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004383244312528321,
      "loss": 4.6503,
      "mean_token_accuracy": 0.25211168378591536,
      "num_tokens": 59381105.0,
      "step": 25890
    },
    {
      "entropy": 5.333580160140992,
      "epoch": 2.487512007684918,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004383008486253006,
      "loss": 4.7484,
      "mean_token_accuracy": 0.2430980995297432,
      "num_tokens": 59391923.0,
      "step": 25895
    },
    {
      "entropy": 5.372943019866943,
      "epoch": 2.487992315081652,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043827726220640827,
      "loss": 4.7208,
      "mean_token_accuracy": 0.2524872496724129,
      "num_tokens": 59402986.0,
      "step": 25900
    },
    {
      "entropy": 5.311718368530274,
      "epoch": 2.488472622478386,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043825367199670274,
      "loss": 4.7371,
      "mean_token_accuracy": 0.24806678891181946,
      "num_tokens": 59413656.0,
      "step": 25905
    },
    {
      "entropy": 5.238162088394165,
      "epoch": 2.4889529298751203,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004382300779967318,
      "loss": 4.6693,
      "mean_token_accuracy": 0.2481812983751297,
      "num_tokens": 59425396.0,
      "step": 25910
    },
    {
      "entropy": 5.371428155899048,
      "epoch": 2.489433237271854,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043820648020704303,
      "loss": 4.7868,
      "mean_token_accuracy": 0.23978421539068223,
      "num_tokens": 59436859.0,
      "step": 25915
    },
    {
      "entropy": 5.370865726470948,
      "epoch": 2.489913544668588,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043818287862818444,
      "loss": 4.8172,
      "mean_token_accuracy": 0.24050249308347701,
      "num_tokens": 59448415.0,
      "step": 25920
    },
    {
      "entropy": 5.259892702102661,
      "epoch": 2.490393852065322,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000438159273260704,
      "loss": 4.6445,
      "mean_token_accuracy": 0.25458006411790846,
      "num_tokens": 59459477.0,
      "step": 25925
    },
    {
      "entropy": 5.211852264404297,
      "epoch": 2.490874159462056,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004381356641051497,
      "loss": 4.6703,
      "mean_token_accuracy": 0.25583205819129945,
      "num_tokens": 59471184.0,
      "step": 25930
    },
    {
      "entropy": 5.276208162307739,
      "epoch": 2.4913544668587897,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004381120511620697,
      "loss": 4.6926,
      "mean_token_accuracy": 0.2502207100391388,
      "num_tokens": 59482706.0,
      "step": 25935
    },
    {
      "entropy": 5.331433439254761,
      "epoch": 2.4918347742555236,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043808843443201217,
      "loss": 4.7638,
      "mean_token_accuracy": 0.23743786215782164,
      "num_tokens": 59493822.0,
      "step": 25940
    },
    {
      "entropy": 5.319265604019165,
      "epoch": 2.4923150816522575,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004380648139155255,
      "loss": 4.6814,
      "mean_token_accuracy": 0.24544309973716735,
      "num_tokens": 59504577.0,
      "step": 25945
    },
    {
      "entropy": 5.283628654479981,
      "epoch": 2.4927953890489913,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004380411896131581,
      "loss": 4.7467,
      "mean_token_accuracy": 0.2357115179300308,
      "num_tokens": 59517395.0,
      "step": 25950
    },
    {
      "entropy": 5.338270759582519,
      "epoch": 2.493275696445725,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043801756152545836,
      "loss": 4.7564,
      "mean_token_accuracy": 0.2431965410709381,
      "num_tokens": 59529270.0,
      "step": 25955
    },
    {
      "entropy": 5.394161081314087,
      "epoch": 2.493756003842459,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043799392965297496,
      "loss": 4.7499,
      "mean_token_accuracy": 0.2392943322658539,
      "num_tokens": 59540866.0,
      "step": 25960
    },
    {
      "entropy": 5.3835591793060305,
      "epoch": 2.494236311239193,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004379702939962564,
      "loss": 4.8307,
      "mean_token_accuracy": 0.23529308587312697,
      "num_tokens": 59553201.0,
      "step": 25965
    },
    {
      "entropy": 5.286488819122314,
      "epoch": 2.494716618635927,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004379466545558516,
      "loss": 4.7254,
      "mean_token_accuracy": 0.24861264675855638,
      "num_tokens": 59563054.0,
      "step": 25970
    },
    {
      "entropy": 5.237975978851319,
      "epoch": 2.4951969260326607,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043792301133230933,
      "loss": 4.6823,
      "mean_token_accuracy": 0.2467556595802307,
      "num_tokens": 59574849.0,
      "step": 25975
    },
    {
      "entropy": 5.222704267501831,
      "epoch": 2.4956772334293946,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004378993643261785,
      "loss": 4.6286,
      "mean_token_accuracy": 0.253773906826973,
      "num_tokens": 59585567.0,
      "step": 25980
    },
    {
      "entropy": 5.312535285949707,
      "epoch": 2.496157540826129,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043787571353800814,
      "loss": 4.7231,
      "mean_token_accuracy": 0.25244301855564116,
      "num_tokens": 59596810.0,
      "step": 25985
    },
    {
      "entropy": 5.298408699035645,
      "epoch": 2.4966378482228624,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004378520589683475,
      "loss": 4.7157,
      "mean_token_accuracy": 0.24529744535684586,
      "num_tokens": 59608704.0,
      "step": 25990
    },
    {
      "entropy": 5.296883964538575,
      "epoch": 2.4971181556195967,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043782840061774544,
      "loss": 4.663,
      "mean_token_accuracy": 0.25600039660930635,
      "num_tokens": 59619325.0,
      "step": 25995
    },
    {
      "entropy": 5.37685112953186,
      "epoch": 2.4975984630163306,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043780473848675143,
      "loss": 4.8201,
      "mean_token_accuracy": 0.24163515418767928,
      "num_tokens": 59630181.0,
      "step": 26000
    },
    {
      "entropy": 5.2821837902069095,
      "epoch": 2.4980787704130645,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004377810725759149,
      "loss": 4.748,
      "mean_token_accuracy": 0.24482716619968414,
      "num_tokens": 59642722.0,
      "step": 26005
    },
    {
      "entropy": 5.369703149795532,
      "epoch": 2.4985590778097984,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043775740288578516,
      "loss": 4.7629,
      "mean_token_accuracy": 0.24332213401794434,
      "num_tokens": 59653675.0,
      "step": 26010
    },
    {
      "entropy": 5.41936731338501,
      "epoch": 2.4990393852065322,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004377337294169118,
      "loss": 4.8464,
      "mean_token_accuracy": 0.235074420273304,
      "num_tokens": 59664897.0,
      "step": 26015
    },
    {
      "entropy": 5.315204000473022,
      "epoch": 2.499519692603266,
      "grad_norm": 1.125,
      "learning_rate": 0.00043771005216984457,
      "loss": 4.6899,
      "mean_token_accuracy": 0.25260454714298247,
      "num_tokens": 59676554.0,
      "step": 26020
    },
    {
      "entropy": 5.366802930831909,
      "epoch": 2.5,
      "grad_norm": 1.09375,
      "learning_rate": 0.000437686371145133,
      "loss": 4.8123,
      "mean_token_accuracy": 0.23660745471715927,
      "num_tokens": 59688030.0,
      "step": 26025
    },
    {
      "entropy": 5.269101476669311,
      "epoch": 2.500480307396734,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000437662686343327,
      "loss": 4.7096,
      "mean_token_accuracy": 0.24654360860586166,
      "num_tokens": 59699721.0,
      "step": 26030
    },
    {
      "entropy": 5.276818323135376,
      "epoch": 2.5009606147934678,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004376389977649764,
      "loss": 4.6991,
      "mean_token_accuracy": 0.241880464553833,
      "num_tokens": 59710426.0,
      "step": 26035
    },
    {
      "entropy": 5.276800632476807,
      "epoch": 2.5014409221902016,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004376153054106313,
      "loss": 4.6986,
      "mean_token_accuracy": 0.245375956594944,
      "num_tokens": 59720837.0,
      "step": 26040
    },
    {
      "entropy": 5.3313572883605955,
      "epoch": 2.5019212295869355,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004375916092808416,
      "loss": 4.746,
      "mean_token_accuracy": 0.24724450558423997,
      "num_tokens": 59732045.0,
      "step": 26045
    },
    {
      "entropy": 5.341475963592529,
      "epoch": 2.5024015369836694,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004375679093761575,
      "loss": 4.7336,
      "mean_token_accuracy": 0.2376833364367485,
      "num_tokens": 59744735.0,
      "step": 26050
    },
    {
      "entropy": 5.326226806640625,
      "epoch": 2.5028818443804033,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043754420569712925,
      "loss": 4.702,
      "mean_token_accuracy": 0.24566367715597154,
      "num_tokens": 59755604.0,
      "step": 26055
    },
    {
      "entropy": 5.283274793624878,
      "epoch": 2.5033621517771376,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043752049824430736,
      "loss": 4.7167,
      "mean_token_accuracy": 0.24623547494411469,
      "num_tokens": 59765892.0,
      "step": 26060
    },
    {
      "entropy": 5.280462694168091,
      "epoch": 2.503842459173871,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043749678701824197,
      "loss": 4.7374,
      "mean_token_accuracy": 0.24311803579330443,
      "num_tokens": 59779068.0,
      "step": 26065
    },
    {
      "entropy": 5.365776443481446,
      "epoch": 2.5043227665706054,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004374730720194837,
      "loss": 4.7882,
      "mean_token_accuracy": 0.2459059700369835,
      "num_tokens": 59789894.0,
      "step": 26070
    },
    {
      "entropy": 5.257040643692017,
      "epoch": 2.5048030739673393,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004374493532485832,
      "loss": 4.6898,
      "mean_token_accuracy": 0.24925874918699265,
      "num_tokens": 59801384.0,
      "step": 26075
    },
    {
      "entropy": 5.339774179458618,
      "epoch": 2.505283381364073,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000437425630706091,
      "loss": 4.7406,
      "mean_token_accuracy": 0.23940975219011307,
      "num_tokens": 59813471.0,
      "step": 26080
    },
    {
      "entropy": 5.379665422439575,
      "epoch": 2.505763688760807,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000437401904392558,
      "loss": 4.7993,
      "mean_token_accuracy": 0.2416891410946846,
      "num_tokens": 59824495.0,
      "step": 26085
    },
    {
      "entropy": 5.427485084533691,
      "epoch": 2.506243996157541,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043737817430853504,
      "loss": 4.8121,
      "mean_token_accuracy": 0.23500901311635972,
      "num_tokens": 59835588.0,
      "step": 26090
    },
    {
      "entropy": 5.339417171478272,
      "epoch": 2.506724303554275,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043735444045457303,
      "loss": 4.7531,
      "mean_token_accuracy": 0.23791597485542298,
      "num_tokens": 59847093.0,
      "step": 26095
    },
    {
      "entropy": 5.283724737167359,
      "epoch": 2.5072046109510087,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043733070283122306,
      "loss": 4.6367,
      "mean_token_accuracy": 0.25868443101644517,
      "num_tokens": 59858109.0,
      "step": 26100
    },
    {
      "entropy": 5.265320491790772,
      "epoch": 2.5076849183477425,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043730696143903607,
      "loss": 4.6788,
      "mean_token_accuracy": 0.24752366691827773,
      "num_tokens": 59870161.0,
      "step": 26105
    },
    {
      "entropy": 5.330035543441772,
      "epoch": 2.5081652257444764,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004372832162785635,
      "loss": 4.7365,
      "mean_token_accuracy": 0.24311739206314087,
      "num_tokens": 59882068.0,
      "step": 26110
    },
    {
      "entropy": 5.349405241012573,
      "epoch": 2.5086455331412103,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004372594673503565,
      "loss": 4.7084,
      "mean_token_accuracy": 0.2499666541814804,
      "num_tokens": 59892418.0,
      "step": 26115
    },
    {
      "entropy": 5.370205307006836,
      "epoch": 2.509125840537944,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004372357146549664,
      "loss": 4.7485,
      "mean_token_accuracy": 0.24805428683757783,
      "num_tokens": 59902790.0,
      "step": 26120
    },
    {
      "entropy": 5.2655243396759035,
      "epoch": 2.509606147934678,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043721195819294487,
      "loss": 4.6969,
      "mean_token_accuracy": 0.24771212339401244,
      "num_tokens": 59914982.0,
      "step": 26125
    },
    {
      "entropy": 5.323065280914307,
      "epoch": 2.510086455331412,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004371881979648433,
      "loss": 4.848,
      "mean_token_accuracy": 0.23386870920658112,
      "num_tokens": 59925961.0,
      "step": 26130
    },
    {
      "entropy": 5.341041612625122,
      "epoch": 2.5105667627281463,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004371644339712133,
      "loss": 4.7853,
      "mean_token_accuracy": 0.2394346371293068,
      "num_tokens": 59938709.0,
      "step": 26135
    },
    {
      "entropy": 5.355858898162841,
      "epoch": 2.5110470701248797,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004371406662126067,
      "loss": 4.7162,
      "mean_token_accuracy": 0.2448517307639122,
      "num_tokens": 59949981.0,
      "step": 26140
    },
    {
      "entropy": 5.264013051986694,
      "epoch": 2.511527377521614,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043711689468957534,
      "loss": 4.6664,
      "mean_token_accuracy": 0.2536080330610275,
      "num_tokens": 59961741.0,
      "step": 26145
    },
    {
      "entropy": 5.251559686660767,
      "epoch": 2.5120076849183475,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043709311940267107,
      "loss": 4.7004,
      "mean_token_accuracy": 0.24508706629276275,
      "num_tokens": 59973399.0,
      "step": 26150
    },
    {
      "entropy": 5.293568849563599,
      "epoch": 2.512487992315082,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004370693403524458,
      "loss": 4.6955,
      "mean_token_accuracy": 0.2521614834666252,
      "num_tokens": 59985180.0,
      "step": 26155
    },
    {
      "entropy": 5.348215389251709,
      "epoch": 2.5129682997118157,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004370455575394518,
      "loss": 4.7254,
      "mean_token_accuracy": 0.24375040084123611,
      "num_tokens": 59995391.0,
      "step": 26160
    },
    {
      "entropy": 5.307676696777344,
      "epoch": 2.5134486071085496,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004370217709642411,
      "loss": 4.7502,
      "mean_token_accuracy": 0.24538978487253188,
      "num_tokens": 60006092.0,
      "step": 26165
    },
    {
      "entropy": 5.358996963500976,
      "epoch": 2.5139289145052834,
      "grad_norm": 1.015625,
      "learning_rate": 0.000436997980627366,
      "loss": 4.7855,
      "mean_token_accuracy": 0.24371442347764968,
      "num_tokens": 60017615.0,
      "step": 26170
    },
    {
      "entropy": 5.347012186050415,
      "epoch": 2.5144092219020173,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00043697418652937877,
      "loss": 4.7432,
      "mean_token_accuracy": 0.23693998157978058,
      "num_tokens": 60030062.0,
      "step": 26175
    },
    {
      "entropy": 5.311870336532593,
      "epoch": 2.514889529298751,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004369503886708319,
      "loss": 4.6947,
      "mean_token_accuracy": 0.24763541519641877,
      "num_tokens": 60042188.0,
      "step": 26180
    },
    {
      "entropy": 5.332211875915528,
      "epoch": 2.515369836695485,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043692658705227796,
      "loss": 4.7614,
      "mean_token_accuracy": 0.25057604908943176,
      "num_tokens": 60054532.0,
      "step": 26185
    },
    {
      "entropy": 5.4009024143219,
      "epoch": 2.515850144092219,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043690278167426945,
      "loss": 4.8727,
      "mean_token_accuracy": 0.24186428189277648,
      "num_tokens": 60066026.0,
      "step": 26190
    },
    {
      "entropy": 5.298065042495727,
      "epoch": 2.516330451488953,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004368789725373591,
      "loss": 4.6682,
      "mean_token_accuracy": 0.2508356124162674,
      "num_tokens": 60079679.0,
      "step": 26195
    },
    {
      "entropy": 5.347916841506958,
      "epoch": 2.5168107588856867,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043685515964209977,
      "loss": 4.7797,
      "mean_token_accuracy": 0.24222542047500611,
      "num_tokens": 60089971.0,
      "step": 26200
    },
    {
      "entropy": 5.287687683105469,
      "epoch": 2.5172910662824206,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004368313429890441,
      "loss": 4.6714,
      "mean_token_accuracy": 0.2538845047354698,
      "num_tokens": 60100200.0,
      "step": 26205
    },
    {
      "entropy": 5.306947183609009,
      "epoch": 2.517771373679155,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004368075225787453,
      "loss": 4.7297,
      "mean_token_accuracy": 0.24636056125164033,
      "num_tokens": 60112031.0,
      "step": 26210
    },
    {
      "entropy": 5.333997678756714,
      "epoch": 2.5182516810758884,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004367836984117562,
      "loss": 4.728,
      "mean_token_accuracy": 0.24995851516723633,
      "num_tokens": 60122233.0,
      "step": 26215
    },
    {
      "entropy": 5.318245220184326,
      "epoch": 2.5187319884726227,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043675987048863,
      "loss": 4.6786,
      "mean_token_accuracy": 0.24989335089921952,
      "num_tokens": 60133144.0,
      "step": 26220
    },
    {
      "entropy": 5.309215068817139,
      "epoch": 2.519212295869356,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004367360388099201,
      "loss": 4.773,
      "mean_token_accuracy": 0.24428625106811525,
      "num_tokens": 60144052.0,
      "step": 26225
    },
    {
      "entropy": 5.3362658500671385,
      "epoch": 2.5196926032660905,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004367122033761796,
      "loss": 4.7359,
      "mean_token_accuracy": 0.24590540379285813,
      "num_tokens": 60155333.0,
      "step": 26230
    },
    {
      "entropy": 5.328320455551148,
      "epoch": 2.5201729106628243,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004366883641879618,
      "loss": 4.7136,
      "mean_token_accuracy": 0.24899066835641862,
      "num_tokens": 60166473.0,
      "step": 26235
    },
    {
      "entropy": 5.142019271850586,
      "epoch": 2.520653218059558,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043666452124582034,
      "loss": 4.5794,
      "mean_token_accuracy": 0.25726027190685274,
      "num_tokens": 60177914.0,
      "step": 26240
    },
    {
      "entropy": 5.305232667922974,
      "epoch": 2.521133525456292,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004366406745503088,
      "loss": 4.772,
      "mean_token_accuracy": 0.24095385372638703,
      "num_tokens": 60190414.0,
      "step": 26245
    },
    {
      "entropy": 5.429467821121216,
      "epoch": 2.521613832853026,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004366168241019807,
      "loss": 4.8854,
      "mean_token_accuracy": 0.2347080186009407,
      "num_tokens": 60202686.0,
      "step": 26250
    },
    {
      "entropy": 5.379337644577026,
      "epoch": 2.52209414024976,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004365929699013899,
      "loss": 4.7194,
      "mean_token_accuracy": 0.24657713174819945,
      "num_tokens": 60214244.0,
      "step": 26255
    },
    {
      "entropy": 5.325851249694824,
      "epoch": 2.5225744476464937,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004365691119490902,
      "loss": 4.7224,
      "mean_token_accuracy": 0.24218034148216247,
      "num_tokens": 60225057.0,
      "step": 26260
    },
    {
      "entropy": 5.351280069351196,
      "epoch": 2.5230547550432276,
      "grad_norm": 1.125,
      "learning_rate": 0.0004365452502456354,
      "loss": 4.8253,
      "mean_token_accuracy": 0.24040338397026062,
      "num_tokens": 60236060.0,
      "step": 26265
    },
    {
      "entropy": 5.310219430923462,
      "epoch": 2.5235350624399615,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004365213847915796,
      "loss": 4.7354,
      "mean_token_accuracy": 0.24489282667636872,
      "num_tokens": 60247493.0,
      "step": 26270
    },
    {
      "entropy": 5.235272216796875,
      "epoch": 2.5240153698366954,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043649751558747695,
      "loss": 4.5959,
      "mean_token_accuracy": 0.25461949557065966,
      "num_tokens": 60258525.0,
      "step": 26275
    },
    {
      "entropy": 5.343403768539429,
      "epoch": 2.5244956772334293,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043647364263388143,
      "loss": 4.7043,
      "mean_token_accuracy": 0.24826311320066452,
      "num_tokens": 60270466.0,
      "step": 26280
    },
    {
      "entropy": 5.312183094024658,
      "epoch": 2.524975984630163,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004364497659313475,
      "loss": 4.7147,
      "mean_token_accuracy": 0.24077924191951752,
      "num_tokens": 60282078.0,
      "step": 26285
    },
    {
      "entropy": 5.328350687026978,
      "epoch": 2.525456292026897,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004364258854804294,
      "loss": 4.7097,
      "mean_token_accuracy": 0.23813129663467408,
      "num_tokens": 60293569.0,
      "step": 26290
    },
    {
      "entropy": 5.244582653045654,
      "epoch": 2.5259365994236314,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004364020012816815,
      "loss": 4.6711,
      "mean_token_accuracy": 0.2468466579914093,
      "num_tokens": 60305840.0,
      "step": 26295
    },
    {
      "entropy": 5.317078161239624,
      "epoch": 2.526416906820365,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004363781133356584,
      "loss": 4.7329,
      "mean_token_accuracy": 0.24782008677721024,
      "num_tokens": 60318519.0,
      "step": 26300
    },
    {
      "entropy": 5.1975172996521,
      "epoch": 2.526897214217099,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004363542216429147,
      "loss": 4.6231,
      "mean_token_accuracy": 0.25271673798561095,
      "num_tokens": 60328984.0,
      "step": 26305
    },
    {
      "entropy": 5.398141670227051,
      "epoch": 2.527377521613833,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004363303262040051,
      "loss": 4.9054,
      "mean_token_accuracy": 0.2334810122847557,
      "num_tokens": 60339846.0,
      "step": 26310
    },
    {
      "entropy": 5.333591747283935,
      "epoch": 2.527857829010567,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043630642701948446,
      "loss": 4.7549,
      "mean_token_accuracy": 0.2460736393928528,
      "num_tokens": 60351336.0,
      "step": 26315
    },
    {
      "entropy": 5.248623371124268,
      "epoch": 2.5283381364073008,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043628252408990756,
      "loss": 4.6583,
      "mean_token_accuracy": 0.2517656072974205,
      "num_tokens": 60362794.0,
      "step": 26320
    },
    {
      "entropy": 5.283070802688599,
      "epoch": 2.5288184438040346,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043625861741582926,
      "loss": 4.7269,
      "mean_token_accuracy": 0.24751584380865096,
      "num_tokens": 60374176.0,
      "step": 26325
    },
    {
      "entropy": 5.271465921401978,
      "epoch": 2.5292987512007685,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043623470699780483,
      "loss": 4.6684,
      "mean_token_accuracy": 0.25395097136497496,
      "num_tokens": 60385414.0,
      "step": 26330
    },
    {
      "entropy": 5.303861665725708,
      "epoch": 2.5297790585975024,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004362107928363892,
      "loss": 4.7061,
      "mean_token_accuracy": 0.2458325520157814,
      "num_tokens": 60397738.0,
      "step": 26335
    },
    {
      "entropy": 5.21811318397522,
      "epoch": 2.5302593659942363,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004361868749321377,
      "loss": 4.6556,
      "mean_token_accuracy": 0.25265554487705233,
      "num_tokens": 60408554.0,
      "step": 26340
    },
    {
      "entropy": 5.275764799118042,
      "epoch": 2.53073967339097,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004361629532856055,
      "loss": 4.6618,
      "mean_token_accuracy": 0.2566846892237663,
      "num_tokens": 60420000.0,
      "step": 26345
    },
    {
      "entropy": 5.292410802841187,
      "epoch": 2.531219980787704,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00043613902789734816,
      "loss": 4.606,
      "mean_token_accuracy": 0.25728404372930525,
      "num_tokens": 60431639.0,
      "step": 26350
    },
    {
      "entropy": 5.319972324371338,
      "epoch": 2.531700288184438,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000436115098767921,
      "loss": 4.7781,
      "mean_token_accuracy": 0.24153947234153747,
      "num_tokens": 60442401.0,
      "step": 26355
    },
    {
      "entropy": 5.375637674331665,
      "epoch": 2.532180595581172,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043609116589787974,
      "loss": 4.8146,
      "mean_token_accuracy": 0.23950733095407487,
      "num_tokens": 60455530.0,
      "step": 26360
    },
    {
      "entropy": 5.426512956619263,
      "epoch": 2.5326609029779057,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004360672292877799,
      "loss": 4.8262,
      "mean_token_accuracy": 0.2363116830587387,
      "num_tokens": 60466380.0,
      "step": 26365
    },
    {
      "entropy": 5.3498883724212645,
      "epoch": 2.53314121037464,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043604328893817726,
      "loss": 4.7447,
      "mean_token_accuracy": 0.2491679683327675,
      "num_tokens": 60477309.0,
      "step": 26370
    },
    {
      "entropy": 5.303422641754151,
      "epoch": 2.5336215177713735,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043601934484962775,
      "loss": 4.7651,
      "mean_token_accuracy": 0.23997810035943984,
      "num_tokens": 60489181.0,
      "step": 26375
    },
    {
      "entropy": 5.315183877944946,
      "epoch": 2.534101825168108,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004359953970226871,
      "loss": 4.7274,
      "mean_token_accuracy": 0.24943280071020127,
      "num_tokens": 60501994.0,
      "step": 26380
    },
    {
      "entropy": 5.299920845031738,
      "epoch": 2.5345821325648417,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043597144545791134,
      "loss": 4.6497,
      "mean_token_accuracy": 0.24681236147880553,
      "num_tokens": 60512769.0,
      "step": 26385
    },
    {
      "entropy": 5.378996133804321,
      "epoch": 2.5350624399615755,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004359474901558567,
      "loss": 4.7458,
      "mean_token_accuracy": 0.2408579707145691,
      "num_tokens": 60523965.0,
      "step": 26390
    },
    {
      "entropy": 5.3160406112670895,
      "epoch": 2.5355427473583094,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004359235311170792,
      "loss": 4.714,
      "mean_token_accuracy": 0.24881679713726043,
      "num_tokens": 60534892.0,
      "step": 26395
    },
    {
      "entropy": 5.39424090385437,
      "epoch": 2.5360230547550433,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004358995683421352,
      "loss": 4.7585,
      "mean_token_accuracy": 0.247086800634861,
      "num_tokens": 60545744.0,
      "step": 26400
    },
    {
      "entropy": 5.179546642303467,
      "epoch": 2.536503362151777,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043587560183158095,
      "loss": 4.5888,
      "mean_token_accuracy": 0.2515127420425415,
      "num_tokens": 60556830.0,
      "step": 26405
    },
    {
      "entropy": 5.285826349258423,
      "epoch": 2.536983669548511,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000435851631585973,
      "loss": 4.7531,
      "mean_token_accuracy": 0.2382993757724762,
      "num_tokens": 60567360.0,
      "step": 26410
    },
    {
      "entropy": 5.35584774017334,
      "epoch": 2.537463976945245,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004358276576058677,
      "loss": 4.7694,
      "mean_token_accuracy": 0.2382144421339035,
      "num_tokens": 60577915.0,
      "step": 26415
    },
    {
      "entropy": 5.331000947952271,
      "epoch": 2.537944284341979,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004358036798918218,
      "loss": 4.7396,
      "mean_token_accuracy": 0.23793186098337174,
      "num_tokens": 60589160.0,
      "step": 26420
    },
    {
      "entropy": 5.304939794540405,
      "epoch": 2.5384245917387127,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004357796984443919,
      "loss": 4.744,
      "mean_token_accuracy": 0.24244072288274765,
      "num_tokens": 60599573.0,
      "step": 26425
    },
    {
      "entropy": 5.319076013565064,
      "epoch": 2.5389048991354466,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043575571326413484,
      "loss": 4.7484,
      "mean_token_accuracy": 0.2466205582022667,
      "num_tokens": 60610886.0,
      "step": 26430
    },
    {
      "entropy": 5.3571983814239506,
      "epoch": 2.5393852065321805,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004357317243516075,
      "loss": 4.7556,
      "mean_token_accuracy": 0.24913984090089797,
      "num_tokens": 60623056.0,
      "step": 26435
    },
    {
      "entropy": 5.333567428588867,
      "epoch": 2.5398655139289144,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043570773170736676,
      "loss": 4.7338,
      "mean_token_accuracy": 0.24526423960924149,
      "num_tokens": 60634358.0,
      "step": 26440
    },
    {
      "entropy": 5.2998466968536375,
      "epoch": 2.5403458213256487,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043568373533196976,
      "loss": 4.7577,
      "mean_token_accuracy": 0.2429261565208435,
      "num_tokens": 60644611.0,
      "step": 26445
    },
    {
      "entropy": 5.317352199554444,
      "epoch": 2.540826128722382,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043565973522597344,
      "loss": 4.7895,
      "mean_token_accuracy": 0.23669586926698685,
      "num_tokens": 60655873.0,
      "step": 26450
    },
    {
      "entropy": 5.342749881744385,
      "epoch": 2.5413064361191164,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043563573138993524,
      "loss": 4.8005,
      "mean_token_accuracy": 0.24434011727571486,
      "num_tokens": 60668634.0,
      "step": 26455
    },
    {
      "entropy": 5.47437629699707,
      "epoch": 2.54178674351585,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004356117238244123,
      "loss": 4.825,
      "mean_token_accuracy": 0.24129209965467452,
      "num_tokens": 60679460.0,
      "step": 26460
    },
    {
      "entropy": 5.362420988082886,
      "epoch": 2.542267050912584,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043558771252996204,
      "loss": 4.7417,
      "mean_token_accuracy": 0.24551832675933838,
      "num_tokens": 60689704.0,
      "step": 26465
    },
    {
      "entropy": 5.360082197189331,
      "epoch": 2.542747358309318,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000435563697507142,
      "loss": 4.7421,
      "mean_token_accuracy": 0.2414999261498451,
      "num_tokens": 60700838.0,
      "step": 26470
    },
    {
      "entropy": 5.321664714813233,
      "epoch": 2.543227665706052,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004355396787565096,
      "loss": 4.7307,
      "mean_token_accuracy": 0.24391621798276902,
      "num_tokens": 60712048.0,
      "step": 26475
    },
    {
      "entropy": 5.267249631881714,
      "epoch": 2.543707973102786,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043551565627862257,
      "loss": 4.6676,
      "mean_token_accuracy": 0.25565323829650877,
      "num_tokens": 60723451.0,
      "step": 26480
    },
    {
      "entropy": 5.257278347015381,
      "epoch": 2.5441882804995197,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004354916300740387,
      "loss": 4.697,
      "mean_token_accuracy": 0.2566931679844856,
      "num_tokens": 60735992.0,
      "step": 26485
    },
    {
      "entropy": 5.339653587341308,
      "epoch": 2.5446685878962536,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004354676001433157,
      "loss": 4.733,
      "mean_token_accuracy": 0.24183777272701262,
      "num_tokens": 60746653.0,
      "step": 26490
    },
    {
      "entropy": 5.3188355445861815,
      "epoch": 2.5451488952929875,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004354435664870116,
      "loss": 4.7301,
      "mean_token_accuracy": 0.24242707788944245,
      "num_tokens": 60757884.0,
      "step": 26495
    },
    {
      "entropy": 5.300941085815429,
      "epoch": 2.5456292026897214,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043541952910568417,
      "loss": 4.6811,
      "mean_token_accuracy": 0.2488512247800827,
      "num_tokens": 60769243.0,
      "step": 26500
    },
    {
      "entropy": 5.356678581237793,
      "epoch": 2.5461095100864553,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004353954879998916,
      "loss": 4.8564,
      "mean_token_accuracy": 0.23578229397535325,
      "num_tokens": 60781168.0,
      "step": 26505
    },
    {
      "entropy": 5.335352849960327,
      "epoch": 2.546589817483189,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004353714431701922,
      "loss": 4.7006,
      "mean_token_accuracy": 0.25407738238573074,
      "num_tokens": 60792743.0,
      "step": 26510
    },
    {
      "entropy": 5.315325784683227,
      "epoch": 2.547070124879923,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004353473946171441,
      "loss": 4.6658,
      "mean_token_accuracy": 0.25166711807250974,
      "num_tokens": 60804251.0,
      "step": 26515
    },
    {
      "entropy": 5.342681646347046,
      "epoch": 2.547550432276657,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043532334234130547,
      "loss": 4.7436,
      "mean_token_accuracy": 0.24583680778741837,
      "num_tokens": 60815567.0,
      "step": 26520
    },
    {
      "entropy": 5.2973236560821535,
      "epoch": 2.5480307396733908,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043529928634323503,
      "loss": 4.7738,
      "mean_token_accuracy": 0.23924424648284912,
      "num_tokens": 60827028.0,
      "step": 26525
    },
    {
      "entropy": 5.21240086555481,
      "epoch": 2.548511047070125,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043527522662349113,
      "loss": 4.6648,
      "mean_token_accuracy": 0.2517237260937691,
      "num_tokens": 60839158.0,
      "step": 26530
    },
    {
      "entropy": 5.343279790878296,
      "epoch": 2.5489913544668585,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004352511631826324,
      "loss": 4.7894,
      "mean_token_accuracy": 0.23766143023967742,
      "num_tokens": 60850022.0,
      "step": 26535
    },
    {
      "entropy": 5.255306577682495,
      "epoch": 2.549471661863593,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004352270960212175,
      "loss": 4.6437,
      "mean_token_accuracy": 0.25184751152992246,
      "num_tokens": 60861259.0,
      "step": 26540
    },
    {
      "entropy": 5.381740474700928,
      "epoch": 2.5499519692603267,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004352030251398052,
      "loss": 4.7266,
      "mean_token_accuracy": 0.24138081967830657,
      "num_tokens": 60871838.0,
      "step": 26545
    },
    {
      "entropy": 5.263728237152099,
      "epoch": 2.5504322766570606,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043517895053895434,
      "loss": 4.6715,
      "mean_token_accuracy": 0.25306740403175354,
      "num_tokens": 60883939.0,
      "step": 26550
    },
    {
      "entropy": 5.350432348251343,
      "epoch": 2.5509125840537945,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004351548722192239,
      "loss": 4.7756,
      "mean_token_accuracy": 0.24716270416975022,
      "num_tokens": 60895799.0,
      "step": 26555
    },
    {
      "entropy": 5.296579122543335,
      "epoch": 2.5513928914505284,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004351307901811729,
      "loss": 4.6676,
      "mean_token_accuracy": 0.24723878651857376,
      "num_tokens": 60906330.0,
      "step": 26560
    },
    {
      "entropy": 5.277750635147095,
      "epoch": 2.5518731988472623,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004351067044253605,
      "loss": 4.6746,
      "mean_token_accuracy": 0.24803802669048308,
      "num_tokens": 60917613.0,
      "step": 26565
    },
    {
      "entropy": 5.363565921783447,
      "epoch": 2.552353506243996,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043508261495234577,
      "loss": 4.8636,
      "mean_token_accuracy": 0.23242043405771257,
      "num_tokens": 60929116.0,
      "step": 26570
    },
    {
      "entropy": 5.244589757919312,
      "epoch": 2.55283381364073,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000435058521762688,
      "loss": 4.6056,
      "mean_token_accuracy": 0.24782298356294633,
      "num_tokens": 60940455.0,
      "step": 26575
    },
    {
      "entropy": 5.320885419845581,
      "epoch": 2.553314121037464,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004350344248569467,
      "loss": 4.8161,
      "mean_token_accuracy": 0.23570578545331955,
      "num_tokens": 60953606.0,
      "step": 26580
    },
    {
      "entropy": 5.335484457015991,
      "epoch": 2.553794428434198,
      "grad_norm": 1.125,
      "learning_rate": 0.0004350103242356813,
      "loss": 4.7207,
      "mean_token_accuracy": 0.2413208544254303,
      "num_tokens": 60964487.0,
      "step": 26585
    },
    {
      "entropy": 5.425752878189087,
      "epoch": 2.5542747358309317,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004349862198994513,
      "loss": 4.8533,
      "mean_token_accuracy": 0.23646885454654692,
      "num_tokens": 60976249.0,
      "step": 26590
    },
    {
      "entropy": 5.405466604232788,
      "epoch": 2.5547550432276656,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004349621118488163,
      "loss": 4.7153,
      "mean_token_accuracy": 0.2511762842535973,
      "num_tokens": 60987581.0,
      "step": 26595
    },
    {
      "entropy": 5.28603196144104,
      "epoch": 2.5552353506243994,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004349380000843361,
      "loss": 4.7207,
      "mean_token_accuracy": 0.24437691420316696,
      "num_tokens": 60999286.0,
      "step": 26600
    },
    {
      "entropy": 5.224120664596557,
      "epoch": 2.5557156580211338,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004349138846065704,
      "loss": 4.6603,
      "mean_token_accuracy": 0.25302050858736036,
      "num_tokens": 61010207.0,
      "step": 26605
    },
    {
      "entropy": 5.278884029388427,
      "epoch": 2.556195965417867,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004348897654160791,
      "loss": 4.726,
      "mean_token_accuracy": 0.2518224433064461,
      "num_tokens": 61021842.0,
      "step": 26610
    },
    {
      "entropy": 5.363008165359497,
      "epoch": 2.5566762728146015,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004348656425134223,
      "loss": 4.7793,
      "mean_token_accuracy": 0.2354576200246811,
      "num_tokens": 61033344.0,
      "step": 26615
    },
    {
      "entropy": 5.256709718704224,
      "epoch": 2.5571565802113354,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043484151589916,
      "loss": 4.695,
      "mean_token_accuracy": 0.24737578332424165,
      "num_tokens": 61044920.0,
      "step": 26620
    },
    {
      "entropy": 5.309695816040039,
      "epoch": 2.5576368876080693,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004348173855738523,
      "loss": 4.6622,
      "mean_token_accuracy": 0.2545671582221985,
      "num_tokens": 61056132.0,
      "step": 26625
    },
    {
      "entropy": 5.268593263626099,
      "epoch": 2.558117195004803,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004347932515380595,
      "loss": 4.7075,
      "mean_token_accuracy": 0.24917189180850982,
      "num_tokens": 61068590.0,
      "step": 26630
    },
    {
      "entropy": 5.319815921783447,
      "epoch": 2.558597502401537,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004347691137923418,
      "loss": 4.7662,
      "mean_token_accuracy": 0.24972684532403946,
      "num_tokens": 61080011.0,
      "step": 26635
    },
    {
      "entropy": 5.3146037578582765,
      "epoch": 2.559077809798271,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004347449723372598,
      "loss": 4.6418,
      "mean_token_accuracy": 0.25201809108257295,
      "num_tokens": 61091448.0,
      "step": 26640
    },
    {
      "entropy": 5.25793867111206,
      "epoch": 2.559558117195005,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004347208271733738,
      "loss": 4.6191,
      "mean_token_accuracy": 0.2583448812365532,
      "num_tokens": 61102507.0,
      "step": 26645
    },
    {
      "entropy": 5.27109546661377,
      "epoch": 2.5600384245917387,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004346966783012445,
      "loss": 4.7077,
      "mean_token_accuracy": 0.24291618913412094,
      "num_tokens": 61114577.0,
      "step": 26650
    },
    {
      "entropy": 5.262277603149414,
      "epoch": 2.5605187319884726,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043467252572143247,
      "loss": 4.5968,
      "mean_token_accuracy": 0.25360594391822816,
      "num_tokens": 61126044.0,
      "step": 26655
    },
    {
      "entropy": 5.33987455368042,
      "epoch": 2.5609990393852065,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043464836943449866,
      "loss": 4.8276,
      "mean_token_accuracy": 0.2327176347374916,
      "num_tokens": 61137880.0,
      "step": 26660
    },
    {
      "entropy": 5.322337961196899,
      "epoch": 2.5614793467819403,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004346242094410036,
      "loss": 4.7697,
      "mean_token_accuracy": 0.24040174037218093,
      "num_tokens": 61149103.0,
      "step": 26665
    },
    {
      "entropy": 5.272454118728637,
      "epoch": 2.561959654178674,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004346000457415085,
      "loss": 4.681,
      "mean_token_accuracy": 0.2455011487007141,
      "num_tokens": 61161235.0,
      "step": 26670
    },
    {
      "entropy": 5.335565567016602,
      "epoch": 2.562439961575408,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043457587833657424,
      "loss": 4.7369,
      "mean_token_accuracy": 0.24695781767368316,
      "num_tokens": 61172497.0,
      "step": 26675
    },
    {
      "entropy": 5.437325096130371,
      "epoch": 2.5629202689721424,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043455170722676194,
      "loss": 4.9046,
      "mean_token_accuracy": 0.22731609791517257,
      "num_tokens": 61183795.0,
      "step": 26680
    },
    {
      "entropy": 5.353398370742798,
      "epoch": 2.563400576368876,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004345275324126328,
      "loss": 4.689,
      "mean_token_accuracy": 0.2519833788275719,
      "num_tokens": 61195909.0,
      "step": 26685
    },
    {
      "entropy": 5.3333290100097654,
      "epoch": 2.56388088376561,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043450335389474796,
      "loss": 4.705,
      "mean_token_accuracy": 0.24490793347358703,
      "num_tokens": 61207017.0,
      "step": 26690
    },
    {
      "entropy": 5.315013790130616,
      "epoch": 2.5643611911623436,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004344791716736689,
      "loss": 4.7815,
      "mean_token_accuracy": 0.23791081160306932,
      "num_tokens": 61219172.0,
      "step": 26695
    },
    {
      "entropy": 5.337757253646851,
      "epoch": 2.564841498559078,
      "grad_norm": 1.0,
      "learning_rate": 0.00043445498574995705,
      "loss": 4.7632,
      "mean_token_accuracy": 0.248645555973053,
      "num_tokens": 61232612.0,
      "step": 26700
    },
    {
      "entropy": 5.3333416938781735,
      "epoch": 2.565321805955812,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043443079612417394,
      "loss": 4.785,
      "mean_token_accuracy": 0.24183891266584395,
      "num_tokens": 61243627.0,
      "step": 26705
    },
    {
      "entropy": 5.3240118503570555,
      "epoch": 2.5658021133525457,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004344066027968812,
      "loss": 4.789,
      "mean_token_accuracy": 0.2367504358291626,
      "num_tokens": 61254516.0,
      "step": 26710
    },
    {
      "entropy": 5.389010810852051,
      "epoch": 2.5662824207492796,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043438240576864034,
      "loss": 4.7341,
      "mean_token_accuracy": 0.24081854075193404,
      "num_tokens": 61266764.0,
      "step": 26715
    },
    {
      "entropy": 5.272210216522216,
      "epoch": 2.5667627281460135,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004343582050400134,
      "loss": 4.6459,
      "mean_token_accuracy": 0.2552323192358017,
      "num_tokens": 61278088.0,
      "step": 26720
    },
    {
      "entropy": 5.263206195831299,
      "epoch": 2.5672430355427474,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004343340006115621,
      "loss": 4.7193,
      "mean_token_accuracy": 0.24589210003614426,
      "num_tokens": 61289558.0,
      "step": 26725
    },
    {
      "entropy": 5.224081373214721,
      "epoch": 2.5677233429394812,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004343097924838483,
      "loss": 4.658,
      "mean_token_accuracy": 0.2505580008029938,
      "num_tokens": 61301815.0,
      "step": 26730
    },
    {
      "entropy": 5.437985944747925,
      "epoch": 2.568203650336215,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004342855806574344,
      "loss": 4.8207,
      "mean_token_accuracy": 0.23896640539169312,
      "num_tokens": 61312216.0,
      "step": 26735
    },
    {
      "entropy": 5.338290214538574,
      "epoch": 2.568683957732949,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004342613651328822,
      "loss": 4.729,
      "mean_token_accuracy": 0.24321127384901048,
      "num_tokens": 61322485.0,
      "step": 26740
    },
    {
      "entropy": 5.3547046184539795,
      "epoch": 2.569164265129683,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000434237145910754,
      "loss": 4.7679,
      "mean_token_accuracy": 0.2359336331486702,
      "num_tokens": 61334654.0,
      "step": 26745
    },
    {
      "entropy": 5.304648733139038,
      "epoch": 2.5696445725264168,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043421292299161213,
      "loss": 4.6712,
      "mean_token_accuracy": 0.25481452643871305,
      "num_tokens": 61346563.0,
      "step": 26750
    },
    {
      "entropy": 5.2839864730834964,
      "epoch": 2.570124879923151,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043418869637601887,
      "loss": 4.7215,
      "mean_token_accuracy": 0.24390317648649215,
      "num_tokens": 61360121.0,
      "step": 26755
    },
    {
      "entropy": 5.343758487701416,
      "epoch": 2.5706051873198845,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043416446606453686,
      "loss": 4.7048,
      "mean_token_accuracy": 0.2489152356982231,
      "num_tokens": 61371340.0,
      "step": 26760
    },
    {
      "entropy": 5.294671201705933,
      "epoch": 2.571085494716619,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004341402320577285,
      "loss": 4.6943,
      "mean_token_accuracy": 0.24788956493139266,
      "num_tokens": 61382571.0,
      "step": 26765
    },
    {
      "entropy": 5.232373142242432,
      "epoch": 2.5715658021133523,
      "grad_norm": 1.0,
      "learning_rate": 0.0004341159943561566,
      "loss": 4.6049,
      "mean_token_accuracy": 0.25491091310977937,
      "num_tokens": 61394548.0,
      "step": 26770
    },
    {
      "entropy": 5.281132173538208,
      "epoch": 2.5720461095100866,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004340917529603837,
      "loss": 4.6745,
      "mean_token_accuracy": 0.25045210868120193,
      "num_tokens": 61405668.0,
      "step": 26775
    },
    {
      "entropy": 5.3077775001525875,
      "epoch": 2.5725264169068205,
      "grad_norm": 1.125,
      "learning_rate": 0.0004340675078709728,
      "loss": 4.7454,
      "mean_token_accuracy": 0.24092965871095656,
      "num_tokens": 61417396.0,
      "step": 26780
    },
    {
      "entropy": 5.278641700744629,
      "epoch": 2.5730067243035544,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004340432590884866,
      "loss": 4.6486,
      "mean_token_accuracy": 0.2543732956051826,
      "num_tokens": 61427981.0,
      "step": 26785
    },
    {
      "entropy": 5.310888957977295,
      "epoch": 2.5734870317002883,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043401900661348825,
      "loss": 4.7175,
      "mean_token_accuracy": 0.24301644265651703,
      "num_tokens": 61439011.0,
      "step": 26790
    },
    {
      "entropy": 5.3507462501525875,
      "epoch": 2.573967339097022,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043399475044654073,
      "loss": 4.7912,
      "mean_token_accuracy": 0.24851143658161162,
      "num_tokens": 61450567.0,
      "step": 26795
    },
    {
      "entropy": 5.314025068283081,
      "epoch": 2.574447646493756,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043397049058820726,
      "loss": 4.7201,
      "mean_token_accuracy": 0.24077454656362535,
      "num_tokens": 61461563.0,
      "step": 26800
    },
    {
      "entropy": 5.315420532226563,
      "epoch": 2.57492795389049,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000433946227039051,
      "loss": 4.7285,
      "mean_token_accuracy": 0.2428484544157982,
      "num_tokens": 61472697.0,
      "step": 26805
    },
    {
      "entropy": 5.320188856124878,
      "epoch": 2.5754082612872238,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004339219597996353,
      "loss": 4.7176,
      "mean_token_accuracy": 0.24076631367206575,
      "num_tokens": 61485877.0,
      "step": 26810
    },
    {
      "entropy": 5.274716091156006,
      "epoch": 2.5758885686839577,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004338976888705236,
      "loss": 4.7047,
      "mean_token_accuracy": 0.23944891095161439,
      "num_tokens": 61499007.0,
      "step": 26815
    },
    {
      "entropy": 5.332191514968872,
      "epoch": 2.5763688760806915,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043387341425227944,
      "loss": 4.7532,
      "mean_token_accuracy": 0.2413945361971855,
      "num_tokens": 61510901.0,
      "step": 26820
    },
    {
      "entropy": 5.300230216979981,
      "epoch": 2.5768491834774254,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004338491359454664,
      "loss": 4.6689,
      "mean_token_accuracy": 0.25020082890987394,
      "num_tokens": 61523628.0,
      "step": 26825
    },
    {
      "entropy": 5.329717302322388,
      "epoch": 2.5773294908741593,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043382485395064796,
      "loss": 4.7539,
      "mean_token_accuracy": 0.24995806217193603,
      "num_tokens": 61534765.0,
      "step": 26830
    },
    {
      "entropy": 5.290347862243652,
      "epoch": 2.577809798270893,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004338005682683881,
      "loss": 4.7308,
      "mean_token_accuracy": 0.24559413641691208,
      "num_tokens": 61547009.0,
      "step": 26835
    },
    {
      "entropy": 5.408317899703979,
      "epoch": 2.5782901056676275,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043377627889925057,
      "loss": 4.8307,
      "mean_token_accuracy": 0.23214859068393706,
      "num_tokens": 61558713.0,
      "step": 26840
    },
    {
      "entropy": 5.371112442016601,
      "epoch": 2.578770413064361,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004337519858437994,
      "loss": 4.7791,
      "mean_token_accuracy": 0.24165882766246796,
      "num_tokens": 61569841.0,
      "step": 26845
    },
    {
      "entropy": 5.313019895553589,
      "epoch": 2.5792507204610953,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004337276891025984,
      "loss": 4.6614,
      "mean_token_accuracy": 0.2507180854678154,
      "num_tokens": 61580878.0,
      "step": 26850
    },
    {
      "entropy": 5.3685400009155275,
      "epoch": 2.579731027857829,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043370338867621184,
      "loss": 4.8196,
      "mean_token_accuracy": 0.23403155654668809,
      "num_tokens": 61593328.0,
      "step": 26855
    },
    {
      "entropy": 5.35723237991333,
      "epoch": 2.580211335254563,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043367908456520387,
      "loss": 4.7599,
      "mean_token_accuracy": 0.23586350977420806,
      "num_tokens": 61605085.0,
      "step": 26860
    },
    {
      "entropy": 5.353047513961792,
      "epoch": 2.580691642651297,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004336547767701387,
      "loss": 4.7705,
      "mean_token_accuracy": 0.23462273478507994,
      "num_tokens": 61618410.0,
      "step": 26865
    },
    {
      "entropy": 5.37677526473999,
      "epoch": 2.581171950048031,
      "grad_norm": 1.125,
      "learning_rate": 0.00043363046529158077,
      "loss": 4.7246,
      "mean_token_accuracy": 0.2446468859910965,
      "num_tokens": 61628795.0,
      "step": 26870
    },
    {
      "entropy": 5.336270332336426,
      "epoch": 2.5816522574447647,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004336061501300944,
      "loss": 4.7331,
      "mean_token_accuracy": 0.24821080416440963,
      "num_tokens": 61638769.0,
      "step": 26875
    },
    {
      "entropy": 5.28133487701416,
      "epoch": 2.5821325648414986,
      "grad_norm": 1.125,
      "learning_rate": 0.0004335818312862442,
      "loss": 4.6991,
      "mean_token_accuracy": 0.2541353017091751,
      "num_tokens": 61648748.0,
      "step": 26880
    },
    {
      "entropy": 5.270302820205688,
      "epoch": 2.5826128722382324,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043355750876059485,
      "loss": 4.6337,
      "mean_token_accuracy": 0.24787077456712722,
      "num_tokens": 61658733.0,
      "step": 26885
    },
    {
      "entropy": 5.309867858886719,
      "epoch": 2.5830931796349663,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043353318255371086,
      "loss": 4.7392,
      "mean_token_accuracy": 0.2423054590821266,
      "num_tokens": 61669192.0,
      "step": 26890
    },
    {
      "entropy": 5.405073595046997,
      "epoch": 2.5835734870317,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004335088526661571,
      "loss": 4.8051,
      "mean_token_accuracy": 0.2449680671095848,
      "num_tokens": 61679091.0,
      "step": 26895
    },
    {
      "entropy": 5.348251581192017,
      "epoch": 2.584053794428434,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043348451909849855,
      "loss": 4.7495,
      "mean_token_accuracy": 0.24833954423666,
      "num_tokens": 61691143.0,
      "step": 26900
    },
    {
      "entropy": 5.40163950920105,
      "epoch": 2.584534101825168,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043346018185130006,
      "loss": 4.8157,
      "mean_token_accuracy": 0.2482965648174286,
      "num_tokens": 61703549.0,
      "step": 26905
    },
    {
      "entropy": 5.294964599609375,
      "epoch": 2.585014409221902,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043343584092512665,
      "loss": 4.679,
      "mean_token_accuracy": 0.2493676707148552,
      "num_tokens": 61713447.0,
      "step": 26910
    },
    {
      "entropy": 5.352622318267822,
      "epoch": 2.585494716618636,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004334114963205435,
      "loss": 4.7529,
      "mean_token_accuracy": 0.23940616995096206,
      "num_tokens": 61724848.0,
      "step": 26915
    },
    {
      "entropy": 5.309061527252197,
      "epoch": 2.5859750240153696,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004333871480381158,
      "loss": 4.7309,
      "mean_token_accuracy": 0.2491192638874054,
      "num_tokens": 61734624.0,
      "step": 26920
    },
    {
      "entropy": 5.3341552257537845,
      "epoch": 2.586455331412104,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004333627960784088,
      "loss": 4.7306,
      "mean_token_accuracy": 0.2485247179865837,
      "num_tokens": 61745627.0,
      "step": 26925
    },
    {
      "entropy": 5.2857825756073,
      "epoch": 2.586935638808838,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004333384404419879,
      "loss": 4.7419,
      "mean_token_accuracy": 0.24050147086381912,
      "num_tokens": 61757050.0,
      "step": 26930
    },
    {
      "entropy": 5.401601028442383,
      "epoch": 2.5874159462055717,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004333140811294186,
      "loss": 4.845,
      "mean_token_accuracy": 0.2332404837012291,
      "num_tokens": 61768896.0,
      "step": 26935
    },
    {
      "entropy": 5.440938425064087,
      "epoch": 2.5878962536023056,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004332897181412664,
      "loss": 4.7871,
      "mean_token_accuracy": 0.23814336955547333,
      "num_tokens": 61779809.0,
      "step": 26940
    },
    {
      "entropy": 5.276342439651489,
      "epoch": 2.5883765609990395,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043326535147809696,
      "loss": 4.6388,
      "mean_token_accuracy": 0.25412336736917496,
      "num_tokens": 61791737.0,
      "step": 26945
    },
    {
      "entropy": 5.257721471786499,
      "epoch": 2.5888568683957733,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043324098114047604,
      "loss": 4.6472,
      "mean_token_accuracy": 0.25283363461494446,
      "num_tokens": 61804193.0,
      "step": 26950
    },
    {
      "entropy": 5.261097002029419,
      "epoch": 2.589337175792507,
      "grad_norm": 1.125,
      "learning_rate": 0.0004332166071289695,
      "loss": 4.6454,
      "mean_token_accuracy": 0.251523420214653,
      "num_tokens": 61815735.0,
      "step": 26955
    },
    {
      "entropy": 5.316955327987671,
      "epoch": 2.589817483189241,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043319222944414304,
      "loss": 4.7077,
      "mean_token_accuracy": 0.25157896280288694,
      "num_tokens": 61827263.0,
      "step": 26960
    },
    {
      "entropy": 5.3042572975158695,
      "epoch": 2.590297790585975,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043316784808656276,
      "loss": 4.6194,
      "mean_token_accuracy": 0.2500371724367142,
      "num_tokens": 61838734.0,
      "step": 26965
    },
    {
      "entropy": 5.360987758636474,
      "epoch": 2.590778097982709,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043314346305679477,
      "loss": 4.7406,
      "mean_token_accuracy": 0.24216468632221222,
      "num_tokens": 61849935.0,
      "step": 26970
    },
    {
      "entropy": 5.277378702163697,
      "epoch": 2.5912584053794427,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043311907435540517,
      "loss": 4.626,
      "mean_token_accuracy": 0.2600464478135109,
      "num_tokens": 61861849.0,
      "step": 26975
    },
    {
      "entropy": 5.3549243927001955,
      "epoch": 2.5917387127761766,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004330946819829601,
      "loss": 4.705,
      "mean_token_accuracy": 0.248875030875206,
      "num_tokens": 61873017.0,
      "step": 26980
    },
    {
      "entropy": 5.294958162307739,
      "epoch": 2.5922190201729105,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043307028594002597,
      "loss": 4.6925,
      "mean_token_accuracy": 0.24722861498594284,
      "num_tokens": 61884730.0,
      "step": 26985
    },
    {
      "entropy": 5.318822717666626,
      "epoch": 2.592699327569645,
      "grad_norm": 1.0,
      "learning_rate": 0.00043304588622716924,
      "loss": 4.7273,
      "mean_token_accuracy": 0.2513909786939621,
      "num_tokens": 61896726.0,
      "step": 26990
    },
    {
      "entropy": 5.259279823303222,
      "epoch": 2.5931796349663783,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004330214828449563,
      "loss": 4.6855,
      "mean_token_accuracy": 0.2434140995144844,
      "num_tokens": 61907931.0,
      "step": 26995
    },
    {
      "entropy": 5.309835720062256,
      "epoch": 2.5936599423631126,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043299707579395365,
      "loss": 4.6623,
      "mean_token_accuracy": 0.25867488235235214,
      "num_tokens": 61919001.0,
      "step": 27000
    },
    {
      "epoch": 2.5936599423631126,
      "eval_entropy": 5.092046837176653,
      "eval_loss": 4.800987720489502,
      "eval_mean_token_accuracy": 0.2512702957496795,
      "eval_num_tokens": 61919001.0,
      "eval_runtime": 20.4938,
      "eval_samples_per_second": 1601.216,
      "eval_steps_per_second": 200.158,
      "step": 27000
    },
    {
      "entropy": 5.405819272994995,
      "epoch": 2.594140249759846,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004329726650747282,
      "loss": 4.8532,
      "mean_token_accuracy": 0.235319384932518,
      "num_tokens": 61931948.0,
      "step": 27005
    },
    {
      "entropy": 5.412089681625366,
      "epoch": 2.5946205571565804,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004329482506878465,
      "loss": 4.8343,
      "mean_token_accuracy": 0.2318657711148262,
      "num_tokens": 61942781.0,
      "step": 27010
    },
    {
      "entropy": 5.319111585617065,
      "epoch": 2.5951008645533142,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043292383263387536,
      "loss": 4.7172,
      "mean_token_accuracy": 0.24613075107336044,
      "num_tokens": 61953981.0,
      "step": 27015
    },
    {
      "entropy": 5.2932737350463865,
      "epoch": 2.595581171950048,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043289941091338187,
      "loss": 4.6682,
      "mean_token_accuracy": 0.25383370369672775,
      "num_tokens": 61965328.0,
      "step": 27020
    },
    {
      "entropy": 5.3354085922241214,
      "epoch": 2.596061479346782,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004328749855269329,
      "loss": 4.6997,
      "mean_token_accuracy": 0.24310676753520966,
      "num_tokens": 61975427.0,
      "step": 27025
    },
    {
      "entropy": 5.314357948303223,
      "epoch": 2.596541786743516,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004328505564750955,
      "loss": 4.7451,
      "mean_token_accuracy": 0.2426811307668686,
      "num_tokens": 61985724.0,
      "step": 27030
    },
    {
      "entropy": 5.443517065048217,
      "epoch": 2.5970220941402498,
      "grad_norm": 1.078125,
      "learning_rate": 0.000432826123758437,
      "loss": 4.9009,
      "mean_token_accuracy": 0.2325392410159111,
      "num_tokens": 61996382.0,
      "step": 27035
    },
    {
      "entropy": 5.394157552719117,
      "epoch": 2.5975024015369836,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004328016873775244,
      "loss": 4.7958,
      "mean_token_accuracy": 0.24559269100427628,
      "num_tokens": 62008859.0,
      "step": 27040
    },
    {
      "entropy": 5.262844896316528,
      "epoch": 2.5979827089337175,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043277724733292527,
      "loss": 4.6287,
      "mean_token_accuracy": 0.25553482472896577,
      "num_tokens": 62019096.0,
      "step": 27045
    },
    {
      "entropy": 5.315943765640259,
      "epoch": 2.5984630163304514,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004327528036252071,
      "loss": 4.7331,
      "mean_token_accuracy": 0.24756340831518173,
      "num_tokens": 62030874.0,
      "step": 27050
    },
    {
      "entropy": 5.271388196945191,
      "epoch": 2.5989433237271853,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004327283562549371,
      "loss": 4.6671,
      "mean_token_accuracy": 0.24920263290405273,
      "num_tokens": 62042620.0,
      "step": 27055
    },
    {
      "entropy": 5.3485089302062985,
      "epoch": 2.599423631123919,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000432703905222683,
      "loss": 4.7482,
      "mean_token_accuracy": 0.25002919882535934,
      "num_tokens": 62054124.0,
      "step": 27060
    },
    {
      "entropy": 5.294420146942139,
      "epoch": 2.5999039385206535,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043267945052901264,
      "loss": 4.6987,
      "mean_token_accuracy": 0.245651775598526,
      "num_tokens": 62064647.0,
      "step": 27065
    },
    {
      "entropy": 5.321081018447876,
      "epoch": 2.600384245917387,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004326549921744935,
      "loss": 4.7067,
      "mean_token_accuracy": 0.24958719909191132,
      "num_tokens": 62075298.0,
      "step": 27070
    },
    {
      "entropy": 5.373770046234131,
      "epoch": 2.6008645533141213,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004326305301596937,
      "loss": 4.8276,
      "mean_token_accuracy": 0.2350325345993042,
      "num_tokens": 62087875.0,
      "step": 27075
    },
    {
      "entropy": 5.2890888214111325,
      "epoch": 2.6013448607108547,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043260606448518096,
      "loss": 4.6115,
      "mean_token_accuracy": 0.2577334985136986,
      "num_tokens": 62098538.0,
      "step": 27080
    },
    {
      "entropy": 5.267714309692383,
      "epoch": 2.601825168107589,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043258159515152347,
      "loss": 4.6742,
      "mean_token_accuracy": 0.24345507472753525,
      "num_tokens": 62109637.0,
      "step": 27085
    },
    {
      "entropy": 5.276412534713745,
      "epoch": 2.602305475504323,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004325571221592892,
      "loss": 4.6581,
      "mean_token_accuracy": 0.24673401862382888,
      "num_tokens": 62120471.0,
      "step": 27090
    },
    {
      "entropy": 5.342976999282837,
      "epoch": 2.6027857829010568,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043253264550904646,
      "loss": 4.7508,
      "mean_token_accuracy": 0.249012753367424,
      "num_tokens": 62130150.0,
      "step": 27095
    },
    {
      "entropy": 5.346546220779419,
      "epoch": 2.6032660902977907,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004325081652013635,
      "loss": 4.7708,
      "mean_token_accuracy": 0.24535564482212066,
      "num_tokens": 62142382.0,
      "step": 27100
    },
    {
      "entropy": 5.248174715042114,
      "epoch": 2.6037463976945245,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043248368123680855,
      "loss": 4.7043,
      "mean_token_accuracy": 0.248147676885128,
      "num_tokens": 62155398.0,
      "step": 27105
    },
    {
      "entropy": 5.279680013656616,
      "epoch": 2.6042267050912584,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043245919361595026,
      "loss": 4.6622,
      "mean_token_accuracy": 0.2481452479958534,
      "num_tokens": 62167694.0,
      "step": 27110
    },
    {
      "entropy": 5.277118873596192,
      "epoch": 2.6047070124879923,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043243470233935696,
      "loss": 4.6692,
      "mean_token_accuracy": 0.2508812353014946,
      "num_tokens": 62179420.0,
      "step": 27115
    },
    {
      "entropy": 5.311356496810913,
      "epoch": 2.605187319884726,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004324102074075973,
      "loss": 4.7028,
      "mean_token_accuracy": 0.24801961630582808,
      "num_tokens": 62190881.0,
      "step": 27120
    },
    {
      "entropy": 5.268491125106811,
      "epoch": 2.60566762728146,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004323857088212402,
      "loss": 4.6341,
      "mean_token_accuracy": 0.24985069632530213,
      "num_tokens": 62201513.0,
      "step": 27125
    },
    {
      "entropy": 5.264788961410522,
      "epoch": 2.606147934678194,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004323612065808541,
      "loss": 4.6717,
      "mean_token_accuracy": 0.24896518439054488,
      "num_tokens": 62212921.0,
      "step": 27130
    },
    {
      "entropy": 5.246712684631348,
      "epoch": 2.606628242074928,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043233670068700827,
      "loss": 4.7095,
      "mean_token_accuracy": 0.2464730769395828,
      "num_tokens": 62224579.0,
      "step": 27135
    },
    {
      "entropy": 5.247177076339722,
      "epoch": 2.6071085494716617,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004323121911402713,
      "loss": 4.7631,
      "mean_token_accuracy": 0.24534956216812134,
      "num_tokens": 62236181.0,
      "step": 27140
    },
    {
      "entropy": 5.349857902526855,
      "epoch": 2.6075888568683956,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043228767794121245,
      "loss": 4.8188,
      "mean_token_accuracy": 0.2457162007689476,
      "num_tokens": 62247607.0,
      "step": 27145
    },
    {
      "entropy": 5.400425958633423,
      "epoch": 2.60806916426513,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004322631610904006,
      "loss": 4.7308,
      "mean_token_accuracy": 0.24699419736862183,
      "num_tokens": 62259171.0,
      "step": 27150
    },
    {
      "entropy": 5.356380605697632,
      "epoch": 2.6085494716618634,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043223864058840534,
      "loss": 4.7979,
      "mean_token_accuracy": 0.24365384578704835,
      "num_tokens": 62271105.0,
      "step": 27155
    },
    {
      "entropy": 5.2882660865783695,
      "epoch": 2.6090297790585977,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043221411643579557,
      "loss": 4.7243,
      "mean_token_accuracy": 0.24707685858011247,
      "num_tokens": 62282818.0,
      "step": 27160
    },
    {
      "entropy": 5.299183797836304,
      "epoch": 2.6095100864553316,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043218958863314096,
      "loss": 4.7614,
      "mean_token_accuracy": 0.24561866819858552,
      "num_tokens": 62294604.0,
      "step": 27165
    },
    {
      "entropy": 5.351138353347778,
      "epoch": 2.6099903938520654,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004321650571810109,
      "loss": 4.7778,
      "mean_token_accuracy": 0.24886855781078338,
      "num_tokens": 62305871.0,
      "step": 27170
    },
    {
      "entropy": 5.29907374382019,
      "epoch": 2.6104707012487993,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004321405220799747,
      "loss": 4.6905,
      "mean_token_accuracy": 0.24614580124616622,
      "num_tokens": 62317100.0,
      "step": 27175
    },
    {
      "entropy": 5.297683143615723,
      "epoch": 2.610951008645533,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004321159833306024,
      "loss": 4.7035,
      "mean_token_accuracy": 0.2437910795211792,
      "num_tokens": 62328934.0,
      "step": 27180
    },
    {
      "entropy": 5.379034996032715,
      "epoch": 2.611431316042267,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004320914409334634,
      "loss": 4.7802,
      "mean_token_accuracy": 0.24026234745979308,
      "num_tokens": 62340812.0,
      "step": 27185
    },
    {
      "entropy": 5.315815401077271,
      "epoch": 2.611911623439001,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004320668948891276,
      "loss": 4.7241,
      "mean_token_accuracy": 0.24990969002246857,
      "num_tokens": 62352753.0,
      "step": 27190
    },
    {
      "entropy": 5.331965065002441,
      "epoch": 2.612391930835735,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043204234519816486,
      "loss": 4.7497,
      "mean_token_accuracy": 0.2453960135579109,
      "num_tokens": 62362731.0,
      "step": 27195
    },
    {
      "entropy": 5.36860990524292,
      "epoch": 2.6128722382324687,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004320177918611453,
      "loss": 4.7759,
      "mean_token_accuracy": 0.23925052732229232,
      "num_tokens": 62374299.0,
      "step": 27200
    },
    {
      "entropy": 5.340035390853882,
      "epoch": 2.6133525456292026,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043199323487863876,
      "loss": 4.7564,
      "mean_token_accuracy": 0.23846922367811202,
      "num_tokens": 62387708.0,
      "step": 27205
    },
    {
      "entropy": 5.321159982681275,
      "epoch": 2.6138328530259365,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043196867425121554,
      "loss": 4.7093,
      "mean_token_accuracy": 0.24955795854330062,
      "num_tokens": 62399139.0,
      "step": 27210
    },
    {
      "entropy": 5.336646175384521,
      "epoch": 2.6143131604226704,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043194410997944577,
      "loss": 4.7119,
      "mean_token_accuracy": 0.24228151589632035,
      "num_tokens": 62411581.0,
      "step": 27215
    },
    {
      "entropy": 5.3457112312316895,
      "epoch": 2.6147934678194042,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043191954206389985,
      "loss": 4.8261,
      "mean_token_accuracy": 0.23342013061046601,
      "num_tokens": 62423768.0,
      "step": 27220
    },
    {
      "entropy": 5.327397680282592,
      "epoch": 2.6152737752161386,
      "grad_norm": 1.0,
      "learning_rate": 0.0004318949705051481,
      "loss": 4.7267,
      "mean_token_accuracy": 0.25085991322994233,
      "num_tokens": 62435716.0,
      "step": 27225
    },
    {
      "entropy": 5.292001962661743,
      "epoch": 2.615754082612872,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004318703953037609,
      "loss": 4.6622,
      "mean_token_accuracy": 0.24740487337112427,
      "num_tokens": 62447409.0,
      "step": 27230
    },
    {
      "entropy": 5.285010147094726,
      "epoch": 2.6162343900096063,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004318458164603091,
      "loss": 4.6805,
      "mean_token_accuracy": 0.2541782796382904,
      "num_tokens": 62458900.0,
      "step": 27235
    },
    {
      "entropy": 5.319493675231934,
      "epoch": 2.61671469740634,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004318212339753631,
      "loss": 4.7566,
      "mean_token_accuracy": 0.24725858569145204,
      "num_tokens": 62471433.0,
      "step": 27240
    },
    {
      "entropy": 5.33993182182312,
      "epoch": 2.617195004803074,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043179664784949375,
      "loss": 4.779,
      "mean_token_accuracy": 0.2406846195459366,
      "num_tokens": 62482633.0,
      "step": 27245
    },
    {
      "entropy": 5.298340463638306,
      "epoch": 2.617675312199808,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004317720580832717,
      "loss": 4.5995,
      "mean_token_accuracy": 0.25751742720603943,
      "num_tokens": 62492661.0,
      "step": 27250
    },
    {
      "entropy": 5.3028360366821286,
      "epoch": 2.618155619596542,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004317474646772681,
      "loss": 4.6983,
      "mean_token_accuracy": 0.24915977716445922,
      "num_tokens": 62504470.0,
      "step": 27255
    },
    {
      "entropy": 5.461922025680542,
      "epoch": 2.6186359269932757,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004317228676320539,
      "loss": 4.8164,
      "mean_token_accuracy": 0.23563704192638396,
      "num_tokens": 62514711.0,
      "step": 27260
    },
    {
      "entropy": 5.385121154785156,
      "epoch": 2.6191162343900096,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043169826694819987,
      "loss": 4.7305,
      "mean_token_accuracy": 0.2525908827781677,
      "num_tokens": 62525939.0,
      "step": 27265
    },
    {
      "entropy": 5.278012657165528,
      "epoch": 2.6195965417867435,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004316736626262775,
      "loss": 4.7045,
      "mean_token_accuracy": 0.2528697997331619,
      "num_tokens": 62536643.0,
      "step": 27270
    },
    {
      "entropy": 5.312819480895996,
      "epoch": 2.6200768491834774,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004316490546668579,
      "loss": 4.7276,
      "mean_token_accuracy": 0.25017926245927813,
      "num_tokens": 62547699.0,
      "step": 27275
    },
    {
      "entropy": 5.356249475479126,
      "epoch": 2.6205571565802113,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004316244430705124,
      "loss": 4.7396,
      "mean_token_accuracy": 0.24386198371648787,
      "num_tokens": 62557756.0,
      "step": 27280
    },
    {
      "entropy": 5.296036243438721,
      "epoch": 2.621037463976945,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004315998278378123,
      "loss": 4.7325,
      "mean_token_accuracy": 0.24683024734258652,
      "num_tokens": 62570046.0,
      "step": 27285
    },
    {
      "entropy": 5.371774101257325,
      "epoch": 2.621517771373679,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043157520896932943,
      "loss": 4.7758,
      "mean_token_accuracy": 0.23990180641412734,
      "num_tokens": 62580037.0,
      "step": 27290
    },
    {
      "entropy": 5.328418493270874,
      "epoch": 2.621998078770413,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000431550586465635,
      "loss": 4.7362,
      "mean_token_accuracy": 0.24587779343128205,
      "num_tokens": 62591442.0,
      "step": 27295
    },
    {
      "entropy": 5.392735624313355,
      "epoch": 2.6224783861671472,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043152596032730085,
      "loss": 4.7737,
      "mean_token_accuracy": 0.2396452710032463,
      "num_tokens": 62603367.0,
      "step": 27300
    },
    {
      "entropy": 5.318356370925903,
      "epoch": 2.6229586935638807,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043150133055489865,
      "loss": 4.657,
      "mean_token_accuracy": 0.25303096920251844,
      "num_tokens": 62613506.0,
      "step": 27305
    },
    {
      "entropy": 5.265680980682373,
      "epoch": 2.623439000960615,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004314766971490003,
      "loss": 4.6719,
      "mean_token_accuracy": 0.24103525578975676,
      "num_tokens": 62625389.0,
      "step": 27310
    },
    {
      "entropy": 5.3032573699951175,
      "epoch": 2.6239193083573484,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004314520601101776,
      "loss": 4.7251,
      "mean_token_accuracy": 0.2484479919075966,
      "num_tokens": 62637949.0,
      "step": 27315
    },
    {
      "entropy": 5.282613945007324,
      "epoch": 2.6243996157540828,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043142741943900275,
      "loss": 4.6787,
      "mean_token_accuracy": 0.2411600425839424,
      "num_tokens": 62649952.0,
      "step": 27320
    },
    {
      "entropy": 5.382475662231445,
      "epoch": 2.6248799231508166,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043140277513604763,
      "loss": 4.8198,
      "mean_token_accuracy": 0.23738697916269302,
      "num_tokens": 62661152.0,
      "step": 27325
    },
    {
      "entropy": 5.356381893157959,
      "epoch": 2.6253602305475505,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004313781272018845,
      "loss": 4.7833,
      "mean_token_accuracy": 0.2349278524518013,
      "num_tokens": 62671648.0,
      "step": 27330
    },
    {
      "entropy": 5.339990711212158,
      "epoch": 2.6258405379442844,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004313534756370857,
      "loss": 4.7113,
      "mean_token_accuracy": 0.24168251305818558,
      "num_tokens": 62683114.0,
      "step": 27335
    },
    {
      "entropy": 5.316632413864136,
      "epoch": 2.6263208453410183,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043132882044222336,
      "loss": 4.7121,
      "mean_token_accuracy": 0.24867279082536697,
      "num_tokens": 62693921.0,
      "step": 27340
    },
    {
      "entropy": 5.324108839035034,
      "epoch": 2.626801152737752,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043130416161787005,
      "loss": 4.823,
      "mean_token_accuracy": 0.23504832684993743,
      "num_tokens": 62705342.0,
      "step": 27345
    },
    {
      "entropy": 5.385346126556397,
      "epoch": 2.627281460134486,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043127949916459823,
      "loss": 4.8115,
      "mean_token_accuracy": 0.24152714014053345,
      "num_tokens": 62717359.0,
      "step": 27350
    },
    {
      "entropy": 5.378161430358887,
      "epoch": 2.62776176753122,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043125483308298053,
      "loss": 4.7862,
      "mean_token_accuracy": 0.2407100185751915,
      "num_tokens": 62728824.0,
      "step": 27355
    },
    {
      "entropy": 5.2463664531707765,
      "epoch": 2.628242074927954,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004312301633735895,
      "loss": 4.6217,
      "mean_token_accuracy": 0.25343096554279326,
      "num_tokens": 62740909.0,
      "step": 27360
    },
    {
      "entropy": 5.290977478027344,
      "epoch": 2.6287223823246877,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004312054900369981,
      "loss": 4.6676,
      "mean_token_accuracy": 0.24804340153932572,
      "num_tokens": 62751934.0,
      "step": 27365
    },
    {
      "entropy": 5.330070495605469,
      "epoch": 2.6292026897214216,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000431180813073779,
      "loss": 4.7367,
      "mean_token_accuracy": 0.24374857395887375,
      "num_tokens": 62762975.0,
      "step": 27370
    },
    {
      "entropy": 5.377506828308105,
      "epoch": 2.629682997118156,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004311561324845051,
      "loss": 4.7416,
      "mean_token_accuracy": 0.24175915867090225,
      "num_tokens": 62774880.0,
      "step": 27375
    },
    {
      "entropy": 5.345115709304809,
      "epoch": 2.6301633045148893,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004311314482697496,
      "loss": 4.6682,
      "mean_token_accuracy": 0.25776350051164626,
      "num_tokens": 62787372.0,
      "step": 27380
    },
    {
      "entropy": 5.292890214920044,
      "epoch": 2.6306436119116237,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004311067604300855,
      "loss": 4.7181,
      "mean_token_accuracy": 0.2546992376446724,
      "num_tokens": 62798437.0,
      "step": 27385
    },
    {
      "entropy": 5.301122188568115,
      "epoch": 2.631123919308357,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004310820689660859,
      "loss": 4.6741,
      "mean_token_accuracy": 0.2473079428076744,
      "num_tokens": 62810406.0,
      "step": 27390
    },
    {
      "entropy": 5.327690505981446,
      "epoch": 2.6316042267050914,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004310573738783242,
      "loss": 4.7555,
      "mean_token_accuracy": 0.24184834212064743,
      "num_tokens": 62822334.0,
      "step": 27395
    },
    {
      "entropy": 5.339380884170533,
      "epoch": 2.6320845341018253,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004310326751673736,
      "loss": 4.7475,
      "mean_token_accuracy": 0.24217303097248077,
      "num_tokens": 62834601.0,
      "step": 27400
    },
    {
      "entropy": 5.380185079574585,
      "epoch": 2.632564841498559,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043100797283380756,
      "loss": 4.7536,
      "mean_token_accuracy": 0.24449584633111954,
      "num_tokens": 62847086.0,
      "step": 27405
    },
    {
      "entropy": 5.32173900604248,
      "epoch": 2.633045148895293,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043098326687819973,
      "loss": 4.6766,
      "mean_token_accuracy": 0.24590715765953064,
      "num_tokens": 62858658.0,
      "step": 27410
    },
    {
      "entropy": 5.303416538238525,
      "epoch": 2.633525456292027,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004309585573011236,
      "loss": 4.7139,
      "mean_token_accuracy": 0.246023029088974,
      "num_tokens": 62870731.0,
      "step": 27415
    },
    {
      "entropy": 5.3566680431365965,
      "epoch": 2.634005763688761,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004309338441031528,
      "loss": 4.8109,
      "mean_token_accuracy": 0.2412973552942276,
      "num_tokens": 62880848.0,
      "step": 27420
    },
    {
      "entropy": 5.321469926834107,
      "epoch": 2.6344860710854947,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043090912728486135,
      "loss": 4.7359,
      "mean_token_accuracy": 0.2493104189634323,
      "num_tokens": 62892367.0,
      "step": 27425
    },
    {
      "entropy": 5.344194316864014,
      "epoch": 2.6349663784822286,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004308844068468228,
      "loss": 4.7522,
      "mean_token_accuracy": 0.24330608248710633,
      "num_tokens": 62903764.0,
      "step": 27430
    },
    {
      "entropy": 5.386289596557617,
      "epoch": 2.6354466858789625,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043085968278961116,
      "loss": 4.8328,
      "mean_token_accuracy": 0.2321823075413704,
      "num_tokens": 62914502.0,
      "step": 27435
    },
    {
      "entropy": 5.388804626464844,
      "epoch": 2.6359269932756964,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043083495511380055,
      "loss": 4.8925,
      "mean_token_accuracy": 0.2375840038061142,
      "num_tokens": 62924793.0,
      "step": 27440
    },
    {
      "entropy": 5.267092657089234,
      "epoch": 2.6364073006724302,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043081022381996506,
      "loss": 4.7083,
      "mean_token_accuracy": 0.24333903044462205,
      "num_tokens": 62936806.0,
      "step": 27445
    },
    {
      "entropy": 5.325898504257202,
      "epoch": 2.636887608069164,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004307854889086787,
      "loss": 4.7093,
      "mean_token_accuracy": 0.24733633548021317,
      "num_tokens": 62948957.0,
      "step": 27450
    },
    {
      "entropy": 5.332693767547608,
      "epoch": 2.637367915465898,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00043076075038051605,
      "loss": 4.7292,
      "mean_token_accuracy": 0.2434602200984955,
      "num_tokens": 62960983.0,
      "step": 27455
    },
    {
      "entropy": 5.255579948425293,
      "epoch": 2.6378482228626323,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004307360082360512,
      "loss": 4.6621,
      "mean_token_accuracy": 0.255060151219368,
      "num_tokens": 62972155.0,
      "step": 27460
    },
    {
      "entropy": 5.325872087478638,
      "epoch": 2.6383285302593658,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043071126247585866,
      "loss": 4.7715,
      "mean_token_accuracy": 0.24423653185367583,
      "num_tokens": 62982928.0,
      "step": 27465
    },
    {
      "entropy": 5.282705593109131,
      "epoch": 2.6388088376561,
      "grad_norm": 1.109375,
      "learning_rate": 0.000430686513100513,
      "loss": 4.6801,
      "mean_token_accuracy": 0.2431701436638832,
      "num_tokens": 62993326.0,
      "step": 27470
    },
    {
      "entropy": 5.4342974662780765,
      "epoch": 2.639289145052834,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00043066176011058877,
      "loss": 4.8641,
      "mean_token_accuracy": 0.23476433604955674,
      "num_tokens": 63006512.0,
      "step": 27475
    },
    {
      "entropy": 5.401448726654053,
      "epoch": 2.639769452449568,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043063700350666066,
      "loss": 4.7849,
      "mean_token_accuracy": 0.24685115367174149,
      "num_tokens": 63017439.0,
      "step": 27480
    },
    {
      "entropy": 5.350289440155029,
      "epoch": 2.6402497598463017,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004306122432893036,
      "loss": 4.6775,
      "mean_token_accuracy": 0.25639048516750335,
      "num_tokens": 63028186.0,
      "step": 27485
    },
    {
      "entropy": 5.25800518989563,
      "epoch": 2.6407300672430356,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043058747945909224,
      "loss": 4.7277,
      "mean_token_accuracy": 0.2512982502579689,
      "num_tokens": 63039864.0,
      "step": 27490
    },
    {
      "entropy": 5.294002389907837,
      "epoch": 2.6412103746397695,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043056271201660166,
      "loss": 4.7234,
      "mean_token_accuracy": 0.24226932972669601,
      "num_tokens": 63051965.0,
      "step": 27495
    },
    {
      "entropy": 5.314355564117432,
      "epoch": 2.6416906820365034,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004305379409624068,
      "loss": 4.7671,
      "mean_token_accuracy": 0.24250079989433287,
      "num_tokens": 63063116.0,
      "step": 27500
    },
    {
      "entropy": 5.355915260314942,
      "epoch": 2.6421709894332372,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004305131662970827,
      "loss": 4.7729,
      "mean_token_accuracy": 0.23939382880926133,
      "num_tokens": 63074680.0,
      "step": 27505
    },
    {
      "entropy": 5.3225596904754635,
      "epoch": 2.642651296829971,
      "grad_norm": 1.0,
      "learning_rate": 0.0004304883880212048,
      "loss": 4.7012,
      "mean_token_accuracy": 0.25074642598629,
      "num_tokens": 63086470.0,
      "step": 27510
    },
    {
      "entropy": 5.396704387664795,
      "epoch": 2.643131604226705,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004304636061353482,
      "loss": 4.7821,
      "mean_token_accuracy": 0.24321856200695038,
      "num_tokens": 63097616.0,
      "step": 27515
    },
    {
      "entropy": 5.329330635070801,
      "epoch": 2.643611911623439,
      "grad_norm": 1.0,
      "learning_rate": 0.0004304388206400883,
      "loss": 4.7516,
      "mean_token_accuracy": 0.24743762314319612,
      "num_tokens": 63109792.0,
      "step": 27520
    },
    {
      "entropy": 5.381741762161255,
      "epoch": 2.6440922190201728,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004304140315360005,
      "loss": 4.826,
      "mean_token_accuracy": 0.23724850118160248,
      "num_tokens": 63122491.0,
      "step": 27525
    },
    {
      "entropy": 5.394035291671753,
      "epoch": 2.6445725264169067,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004303892388236604,
      "loss": 4.7672,
      "mean_token_accuracy": 0.23946222364902497,
      "num_tokens": 63134042.0,
      "step": 27530
    },
    {
      "entropy": 5.264126968383789,
      "epoch": 2.645052833813641,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004303644425036436,
      "loss": 4.5885,
      "mean_token_accuracy": 0.257209911942482,
      "num_tokens": 63145885.0,
      "step": 27535
    },
    {
      "entropy": 5.301181983947754,
      "epoch": 2.6455331412103744,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043033964257652575,
      "loss": 4.6978,
      "mean_token_accuracy": 0.24689621329307557,
      "num_tokens": 63156995.0,
      "step": 27540
    },
    {
      "entropy": 5.325877618789673,
      "epoch": 2.6460134486071087,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004303148390428827,
      "loss": 4.8013,
      "mean_token_accuracy": 0.23708871454000474,
      "num_tokens": 63168604.0,
      "step": 27545
    },
    {
      "entropy": 5.376401281356811,
      "epoch": 2.6464937560038426,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043029003190329023,
      "loss": 4.6918,
      "mean_token_accuracy": 0.2395616203546524,
      "num_tokens": 63181284.0,
      "step": 27550
    },
    {
      "entropy": 5.2590978145599365,
      "epoch": 2.6469740634005765,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004302652211583244,
      "loss": 4.6553,
      "mean_token_accuracy": 0.24534384161233902,
      "num_tokens": 63193760.0,
      "step": 27555
    },
    {
      "entropy": 5.310172080993652,
      "epoch": 2.6474543707973104,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004302404068085612,
      "loss": 4.7026,
      "mean_token_accuracy": 0.25587007254362104,
      "num_tokens": 63205280.0,
      "step": 27560
    },
    {
      "entropy": 5.243616724014283,
      "epoch": 2.6479346781940443,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004302155888545766,
      "loss": 4.6135,
      "mean_token_accuracy": 0.25216409713029864,
      "num_tokens": 63216253.0,
      "step": 27565
    },
    {
      "entropy": 5.303314971923828,
      "epoch": 2.648414985590778,
      "grad_norm": 1.09375,
      "learning_rate": 0.000430190767296947,
      "loss": 4.7674,
      "mean_token_accuracy": 0.24114762991666794,
      "num_tokens": 63227341.0,
      "step": 27570
    },
    {
      "entropy": 5.299710655212403,
      "epoch": 2.648895292987512,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004301659421362486,
      "loss": 4.7204,
      "mean_token_accuracy": 0.24548285603523254,
      "num_tokens": 63239090.0,
      "step": 27575
    },
    {
      "entropy": 5.280298328399658,
      "epoch": 2.649375600384246,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004301411133730578,
      "loss": 4.7147,
      "mean_token_accuracy": 0.2550721734762192,
      "num_tokens": 63250077.0,
      "step": 27580
    },
    {
      "entropy": 5.2947015285491945,
      "epoch": 2.64985590778098,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043011628100795093,
      "loss": 4.6785,
      "mean_token_accuracy": 0.24215085357427596,
      "num_tokens": 63261177.0,
      "step": 27585
    },
    {
      "entropy": 5.285066175460815,
      "epoch": 2.6503362151777137,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004300914450415047,
      "loss": 4.69,
      "mean_token_accuracy": 0.25387996435165405,
      "num_tokens": 63273176.0,
      "step": 27590
    },
    {
      "entropy": 5.317886829376221,
      "epoch": 2.6508165225744476,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043006660547429565,
      "loss": 4.759,
      "mean_token_accuracy": 0.2457375168800354,
      "num_tokens": 63284529.0,
      "step": 27595
    },
    {
      "entropy": 5.239973354339599,
      "epoch": 2.6512968299711814,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004300417623069005,
      "loss": 4.5741,
      "mean_token_accuracy": 0.25895974934101107,
      "num_tokens": 63295824.0,
      "step": 27600
    },
    {
      "entropy": 5.319367361068726,
      "epoch": 2.6517771373679153,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004300169155398959,
      "loss": 4.7137,
      "mean_token_accuracy": 0.2471233293414116,
      "num_tokens": 63307267.0,
      "step": 27605
    },
    {
      "entropy": 5.397471761703491,
      "epoch": 2.6522574447646496,
      "grad_norm": 1.25,
      "learning_rate": 0.00042999206517385885,
      "loss": 4.7747,
      "mean_token_accuracy": 0.24310313016176224,
      "num_tokens": 63317703.0,
      "step": 27610
    },
    {
      "entropy": 5.32558307647705,
      "epoch": 2.652737752161383,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004299672112093663,
      "loss": 4.7837,
      "mean_token_accuracy": 0.23985424488782883,
      "num_tokens": 63328677.0,
      "step": 27615
    },
    {
      "entropy": 5.44613823890686,
      "epoch": 2.6532180595581174,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042994235364699526,
      "loss": 4.8594,
      "mean_token_accuracy": 0.23630535304546357,
      "num_tokens": 63339604.0,
      "step": 27620
    },
    {
      "entropy": 5.3837018489837645,
      "epoch": 2.653698366954851,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004299174924873229,
      "loss": 4.7881,
      "mean_token_accuracy": 0.2388268604874611,
      "num_tokens": 63351428.0,
      "step": 27625
    },
    {
      "entropy": 5.35583233833313,
      "epoch": 2.654178674351585,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004298926277309263,
      "loss": 4.8109,
      "mean_token_accuracy": 0.24101043939590455,
      "num_tokens": 63364545.0,
      "step": 27630
    },
    {
      "entropy": 5.344295024871826,
      "epoch": 2.654658981748319,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042986775937838283,
      "loss": 4.6977,
      "mean_token_accuracy": 0.24774497300386428,
      "num_tokens": 63376063.0,
      "step": 27635
    },
    {
      "entropy": 5.316325426101685,
      "epoch": 2.655139289145053,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004298428874302699,
      "loss": 4.6689,
      "mean_token_accuracy": 0.2568302467465401,
      "num_tokens": 63387448.0,
      "step": 27640
    },
    {
      "entropy": 5.31105260848999,
      "epoch": 2.655619596541787,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004298180118871649,
      "loss": 4.7696,
      "mean_token_accuracy": 0.24229122549295426,
      "num_tokens": 63399674.0,
      "step": 27645
    },
    {
      "entropy": 5.359382057189942,
      "epoch": 2.6560999039385207,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042979313274964535,
      "loss": 4.7782,
      "mean_token_accuracy": 0.2433603212237358,
      "num_tokens": 63410622.0,
      "step": 27650
    },
    {
      "entropy": 5.368937158584595,
      "epoch": 2.6565802113352546,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042976825001828897,
      "loss": 4.7501,
      "mean_token_accuracy": 0.23858112543821336,
      "num_tokens": 63422354.0,
      "step": 27655
    },
    {
      "entropy": 5.3596720695495605,
      "epoch": 2.6570605187319885,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042974336369367333,
      "loss": 4.737,
      "mean_token_accuracy": 0.24401975125074388,
      "num_tokens": 63433232.0,
      "step": 27660
    },
    {
      "entropy": 5.2315943241119385,
      "epoch": 2.6575408261287223,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004297184737763763,
      "loss": 4.635,
      "mean_token_accuracy": 0.25421824157238004,
      "num_tokens": 63444998.0,
      "step": 27665
    },
    {
      "entropy": 5.315202808380127,
      "epoch": 2.658021133525456,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042969358026697567,
      "loss": 4.7865,
      "mean_token_accuracy": 0.23748118281364441,
      "num_tokens": 63455846.0,
      "step": 27670
    },
    {
      "entropy": 5.362964677810669,
      "epoch": 2.65850144092219,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004296686831660495,
      "loss": 4.7252,
      "mean_token_accuracy": 0.24376165270805358,
      "num_tokens": 63468353.0,
      "step": 27675
    },
    {
      "entropy": 5.354706907272339,
      "epoch": 2.658981748318924,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004296437824741758,
      "loss": 4.7073,
      "mean_token_accuracy": 0.24482696503400803,
      "num_tokens": 63479802.0,
      "step": 27680
    },
    {
      "entropy": 5.373414325714111,
      "epoch": 2.659462055715658,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042961887819193263,
      "loss": 4.7901,
      "mean_token_accuracy": 0.2357308968901634,
      "num_tokens": 63490856.0,
      "step": 27685
    },
    {
      "entropy": 5.409032201766967,
      "epoch": 2.6599423631123917,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004295939703198983,
      "loss": 4.8028,
      "mean_token_accuracy": 0.23570724874734877,
      "num_tokens": 63502294.0,
      "step": 27690
    },
    {
      "entropy": 5.4491020202636715,
      "epoch": 2.660422670509126,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000429569058858651,
      "loss": 4.7935,
      "mean_token_accuracy": 0.24104488044977188,
      "num_tokens": 63513466.0,
      "step": 27695
    },
    {
      "entropy": 5.317182493209839,
      "epoch": 2.6609029779058595,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042954414380876906,
      "loss": 4.6556,
      "mean_token_accuracy": 0.2541769564151764,
      "num_tokens": 63523596.0,
      "step": 27700
    },
    {
      "entropy": 5.2710127353668215,
      "epoch": 2.661383285302594,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042951922517083104,
      "loss": 4.7098,
      "mean_token_accuracy": 0.24407961666584016,
      "num_tokens": 63534468.0,
      "step": 27705
    },
    {
      "entropy": 5.2944169521331785,
      "epoch": 2.6618635926993277,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004294943029454155,
      "loss": 4.7095,
      "mean_token_accuracy": 0.24979802519083022,
      "num_tokens": 63546273.0,
      "step": 27710
    },
    {
      "entropy": 5.26757550239563,
      "epoch": 2.6623439000960616,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042946937713310093,
      "loss": 4.734,
      "mean_token_accuracy": 0.24041973203420638,
      "num_tokens": 63557032.0,
      "step": 27715
    },
    {
      "entropy": 5.363001441955566,
      "epoch": 2.6628242074927955,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004294444477344661,
      "loss": 4.7372,
      "mean_token_accuracy": 0.2537866771221161,
      "num_tokens": 63568358.0,
      "step": 27720
    },
    {
      "entropy": 5.423635196685791,
      "epoch": 2.6633045148895294,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004294195147500898,
      "loss": 4.838,
      "mean_token_accuracy": 0.2417287290096283,
      "num_tokens": 63579716.0,
      "step": 27725
    },
    {
      "entropy": 5.269621515274048,
      "epoch": 2.6637848222862632,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042939457818055095,
      "loss": 4.6992,
      "mean_token_accuracy": 0.2542797863483429,
      "num_tokens": 63590871.0,
      "step": 27730
    },
    {
      "entropy": 5.390863132476807,
      "epoch": 2.664265129682997,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042936963802642843,
      "loss": 4.7908,
      "mean_token_accuracy": 0.24199527949094773,
      "num_tokens": 63600813.0,
      "step": 27735
    },
    {
      "entropy": 5.25482177734375,
      "epoch": 2.664745437079731,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004293446942883013,
      "loss": 4.6211,
      "mean_token_accuracy": 0.25525553673505785,
      "num_tokens": 63611141.0,
      "step": 27740
    },
    {
      "entropy": 5.253128290176392,
      "epoch": 2.665225744476465,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042931974696674866,
      "loss": 4.6482,
      "mean_token_accuracy": 0.2523071378469467,
      "num_tokens": 63622172.0,
      "step": 27745
    },
    {
      "entropy": 5.27330641746521,
      "epoch": 2.6657060518731988,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004292947960623497,
      "loss": 4.7574,
      "mean_token_accuracy": 0.2393648087978363,
      "num_tokens": 63634369.0,
      "step": 27750
    },
    {
      "entropy": 5.300602436065674,
      "epoch": 2.6661863592699326,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042926984157568384,
      "loss": 4.6868,
      "mean_token_accuracy": 0.2528583362698555,
      "num_tokens": 63645142.0,
      "step": 27755
    },
    {
      "entropy": 5.296164274215698,
      "epoch": 2.6666666666666665,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042924488350733024,
      "loss": 4.6985,
      "mean_token_accuracy": 0.2498031049966812,
      "num_tokens": 63656497.0,
      "step": 27760
    },
    {
      "entropy": 5.331110000610352,
      "epoch": 2.6671469740634004,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042921992185786847,
      "loss": 4.7993,
      "mean_token_accuracy": 0.24539343565702437,
      "num_tokens": 63667795.0,
      "step": 27765
    },
    {
      "entropy": 5.274717569351196,
      "epoch": 2.6676272814601347,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042919495662787813,
      "loss": 4.7757,
      "mean_token_accuracy": 0.24714350402355195,
      "num_tokens": 63679676.0,
      "step": 27770
    },
    {
      "entropy": 5.377812194824219,
      "epoch": 2.668107588856868,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004291699878179387,
      "loss": 4.7265,
      "mean_token_accuracy": 0.23858391046524047,
      "num_tokens": 63690749.0,
      "step": 27775
    },
    {
      "entropy": 5.291700220108032,
      "epoch": 2.6685878962536025,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004291450154286299,
      "loss": 4.6466,
      "mean_token_accuracy": 0.2454966515302658,
      "num_tokens": 63702583.0,
      "step": 27780
    },
    {
      "entropy": 5.318862771987915,
      "epoch": 2.6690682036503364,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004291200394605317,
      "loss": 4.7084,
      "mean_token_accuracy": 0.24714806377887727,
      "num_tokens": 63715296.0,
      "step": 27785
    },
    {
      "entropy": 5.331928062438965,
      "epoch": 2.6695485110470702,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004290950599142237,
      "loss": 4.6608,
      "mean_token_accuracy": 0.2539996415376663,
      "num_tokens": 63726261.0,
      "step": 27790
    },
    {
      "entropy": 5.398883962631226,
      "epoch": 2.670028818443804,
      "grad_norm": 1.078125,
      "learning_rate": 0.000429070076790286,
      "loss": 4.8537,
      "mean_token_accuracy": 0.237601175904274,
      "num_tokens": 63738482.0,
      "step": 27795
    },
    {
      "entropy": 5.35823917388916,
      "epoch": 2.670509125840538,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042904509008929873,
      "loss": 4.8108,
      "mean_token_accuracy": 0.2386433020234108,
      "num_tokens": 63748899.0,
      "step": 27800
    },
    {
      "entropy": 5.411176633834839,
      "epoch": 2.670989433237272,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004290200998118417,
      "loss": 4.7457,
      "mean_token_accuracy": 0.24234521985054017,
      "num_tokens": 63760427.0,
      "step": 27805
    },
    {
      "entropy": 5.328832483291626,
      "epoch": 2.6714697406340058,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042899510595849544,
      "loss": 4.7307,
      "mean_token_accuracy": 0.2532069176435471,
      "num_tokens": 63772226.0,
      "step": 27810
    },
    {
      "entropy": 5.378662633895874,
      "epoch": 2.6719500480307397,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042897010852984004,
      "loss": 4.8029,
      "mean_token_accuracy": 0.2441457539796829,
      "num_tokens": 63782918.0,
      "step": 27815
    },
    {
      "entropy": 5.331945514678955,
      "epoch": 2.6724303554274735,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042894510752645586,
      "loss": 4.7001,
      "mean_token_accuracy": 0.24428645819425582,
      "num_tokens": 63793461.0,
      "step": 27820
    },
    {
      "entropy": 5.305151081085205,
      "epoch": 2.6729106628242074,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004289201029489235,
      "loss": 4.6765,
      "mean_token_accuracy": 0.2512824460864067,
      "num_tokens": 63804648.0,
      "step": 27825
    },
    {
      "entropy": 5.3448230743408205,
      "epoch": 2.6733909702209413,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004288950947978234,
      "loss": 4.726,
      "mean_token_accuracy": 0.24623325914144517,
      "num_tokens": 63815809.0,
      "step": 27830
    },
    {
      "entropy": 5.294317007064819,
      "epoch": 2.673871277617675,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004288700830737361,
      "loss": 4.7035,
      "mean_token_accuracy": 0.25182729959487915,
      "num_tokens": 63828310.0,
      "step": 27835
    },
    {
      "entropy": 5.291182136535644,
      "epoch": 2.674351585014409,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042884506777724244,
      "loss": 4.6939,
      "mean_token_accuracy": 0.24383069425821305,
      "num_tokens": 63840546.0,
      "step": 27840
    },
    {
      "entropy": 5.276570510864258,
      "epoch": 2.6748318924111434,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004288200489089231,
      "loss": 4.6109,
      "mean_token_accuracy": 0.2547426253557205,
      "num_tokens": 63852049.0,
      "step": 27845
    },
    {
      "entropy": 5.304743099212646,
      "epoch": 2.675312199807877,
      "grad_norm": 1.09375,
      "learning_rate": 0.000428795026469359,
      "loss": 4.7055,
      "mean_token_accuracy": 0.24670605957508088,
      "num_tokens": 63863474.0,
      "step": 27850
    },
    {
      "entropy": 5.315161037445068,
      "epoch": 2.675792507204611,
      "grad_norm": 1.21875,
      "learning_rate": 0.000428770000459131,
      "loss": 4.7061,
      "mean_token_accuracy": 0.24508919417858124,
      "num_tokens": 63874906.0,
      "step": 27855
    },
    {
      "entropy": 5.330074977874756,
      "epoch": 2.6762728146013446,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004287449708788202,
      "loss": 4.7086,
      "mean_token_accuracy": 0.24065853506326676,
      "num_tokens": 63885443.0,
      "step": 27860
    },
    {
      "entropy": 5.3186671257019045,
      "epoch": 2.676753121998079,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004287199377290077,
      "loss": 4.7569,
      "mean_token_accuracy": 0.24611937552690505,
      "num_tokens": 63897772.0,
      "step": 27865
    },
    {
      "entropy": 5.427214765548706,
      "epoch": 2.677233429394813,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004286949010102748,
      "loss": 4.7706,
      "mean_token_accuracy": 0.2417786806821823,
      "num_tokens": 63910409.0,
      "step": 27870
    },
    {
      "entropy": 5.306625843048096,
      "epoch": 2.6777137367915467,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004286698607232026,
      "loss": 4.7128,
      "mean_token_accuracy": 0.2468500778079033,
      "num_tokens": 63922671.0,
      "step": 27875
    },
    {
      "entropy": 5.385310983657837,
      "epoch": 2.6781940441882806,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042864481686837253,
      "loss": 4.7503,
      "mean_token_accuracy": 0.24743371158838273,
      "num_tokens": 63933964.0,
      "step": 27880
    },
    {
      "entropy": 5.420191049575806,
      "epoch": 2.6786743515850144,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042861976944636604,
      "loss": 4.7218,
      "mean_token_accuracy": 0.2500866040587425,
      "num_tokens": 63945347.0,
      "step": 27885
    },
    {
      "entropy": 5.328283882141113,
      "epoch": 2.6791546589817483,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004285947184577647,
      "loss": 4.7553,
      "mean_token_accuracy": 0.24079434126615523,
      "num_tokens": 63957573.0,
      "step": 27890
    },
    {
      "entropy": 5.350186777114868,
      "epoch": 2.679634966378482,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042856966390315013,
      "loss": 4.7143,
      "mean_token_accuracy": 0.251377934217453,
      "num_tokens": 63968666.0,
      "step": 27895
    },
    {
      "entropy": 5.302673053741455,
      "epoch": 2.680115273775216,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004285446057831039,
      "loss": 4.698,
      "mean_token_accuracy": 0.24393858760595322,
      "num_tokens": 63981147.0,
      "step": 27900
    },
    {
      "entropy": 5.394190931320191,
      "epoch": 2.68059558117195,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004285195440982078,
      "loss": 4.8324,
      "mean_token_accuracy": 0.23535267263650894,
      "num_tokens": 63992474.0,
      "step": 27905
    },
    {
      "entropy": 5.350878381729126,
      "epoch": 2.681075888568684,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004284944788490439,
      "loss": 4.7132,
      "mean_token_accuracy": 0.24447631537914277,
      "num_tokens": 64003804.0,
      "step": 27910
    },
    {
      "entropy": 5.2817154884338375,
      "epoch": 2.6815561959654177,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004284694100361938,
      "loss": 4.7354,
      "mean_token_accuracy": 0.24755799025297165,
      "num_tokens": 64015587.0,
      "step": 27915
    },
    {
      "entropy": 5.405901336669922,
      "epoch": 2.682036503362152,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004284443376602398,
      "loss": 4.8281,
      "mean_token_accuracy": 0.2459500715136528,
      "num_tokens": 64027088.0,
      "step": 27920
    },
    {
      "entropy": 5.365456533432007,
      "epoch": 2.6825168107588855,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004284192617217639,
      "loss": 4.7234,
      "mean_token_accuracy": 0.24639266580343247,
      "num_tokens": 64037050.0,
      "step": 27925
    },
    {
      "entropy": 5.27004451751709,
      "epoch": 2.68299711815562,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004283941822213484,
      "loss": 4.7359,
      "mean_token_accuracy": 0.24401412010192872,
      "num_tokens": 64049287.0,
      "step": 27930
    },
    {
      "entropy": 5.346831846237182,
      "epoch": 2.6834774255523532,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004283690991595754,
      "loss": 4.7087,
      "mean_token_accuracy": 0.25041641145944593,
      "num_tokens": 64060578.0,
      "step": 27935
    },
    {
      "entropy": 5.374514532089234,
      "epoch": 2.6839577329490876,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042834401253702734,
      "loss": 4.8017,
      "mean_token_accuracy": 0.23739527612924577,
      "num_tokens": 64071540.0,
      "step": 27940
    },
    {
      "entropy": 5.3596296310424805,
      "epoch": 2.6844380403458215,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004283189223542866,
      "loss": 4.7491,
      "mean_token_accuracy": 0.2424227252602577,
      "num_tokens": 64084014.0,
      "step": 27945
    },
    {
      "entropy": 5.33119797706604,
      "epoch": 2.6849183477425553,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042829382861193585,
      "loss": 4.7786,
      "mean_token_accuracy": 0.2463359147310257,
      "num_tokens": 64096735.0,
      "step": 27950
    },
    {
      "entropy": 5.402970790863037,
      "epoch": 2.685398655139289,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004282687313105575,
      "loss": 4.823,
      "mean_token_accuracy": 0.23498864024877547,
      "num_tokens": 64107548.0,
      "step": 27955
    },
    {
      "entropy": 5.322093772888183,
      "epoch": 2.685878962536023,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042824363045073434,
      "loss": 4.7472,
      "mean_token_accuracy": 0.24114889800548553,
      "num_tokens": 64118433.0,
      "step": 27960
    },
    {
      "entropy": 5.338133001327515,
      "epoch": 2.686359269932757,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004282185260330491,
      "loss": 4.7422,
      "mean_token_accuracy": 0.24835185408592225,
      "num_tokens": 64129883.0,
      "step": 27965
    },
    {
      "entropy": 5.388800621032715,
      "epoch": 2.686839577329491,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042819341805808473,
      "loss": 4.7753,
      "mean_token_accuracy": 0.24765200167894363,
      "num_tokens": 64141827.0,
      "step": 27970
    },
    {
      "entropy": 5.331389331817627,
      "epoch": 2.6873198847262247,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042816830652642396,
      "loss": 4.7046,
      "mean_token_accuracy": 0.2471479892730713,
      "num_tokens": 64152957.0,
      "step": 27975
    },
    {
      "entropy": 5.3013652801513675,
      "epoch": 2.6878001921229586,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004281431914386501,
      "loss": 4.6768,
      "mean_token_accuracy": 0.24858610481023788,
      "num_tokens": 64164800.0,
      "step": 27980
    },
    {
      "entropy": 5.374055528640747,
      "epoch": 2.6882804995196925,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000428118072795346,
      "loss": 4.755,
      "mean_token_accuracy": 0.24467783868312837,
      "num_tokens": 64175941.0,
      "step": 27985
    },
    {
      "entropy": 5.308044290542602,
      "epoch": 2.6887608069164264,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042809295059709483,
      "loss": 4.6985,
      "mean_token_accuracy": 0.24977784156799315,
      "num_tokens": 64187229.0,
      "step": 27990
    },
    {
      "entropy": 5.343579435348511,
      "epoch": 2.6892411143131603,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00042806782484448,
      "loss": 4.743,
      "mean_token_accuracy": 0.24891913086175918,
      "num_tokens": 64198328.0,
      "step": 27995
    },
    {
      "entropy": 5.347207021713257,
      "epoch": 2.689721421709894,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004280426955380848,
      "loss": 4.6663,
      "mean_token_accuracy": 0.2510742530226707,
      "num_tokens": 64207434.0,
      "step": 28000
    },
    {
      "entropy": 5.310294961929321,
      "epoch": 2.6902017291066285,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042801756267849266,
      "loss": 4.6957,
      "mean_token_accuracy": 0.2502310246229172,
      "num_tokens": 64219128.0,
      "step": 28005
    },
    {
      "entropy": 5.332197046279907,
      "epoch": 2.690682036503362,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004279924262662871,
      "loss": 4.7191,
      "mean_token_accuracy": 0.24494616985321044,
      "num_tokens": 64230664.0,
      "step": 28010
    },
    {
      "entropy": 5.3306561470031735,
      "epoch": 2.6911623439000962,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004279672863020517,
      "loss": 4.6909,
      "mean_token_accuracy": 0.25370009541511535,
      "num_tokens": 64242339.0,
      "step": 28015
    },
    {
      "entropy": 5.327930641174317,
      "epoch": 2.69164265129683,
      "grad_norm": 1.125,
      "learning_rate": 0.00042794214278637013,
      "loss": 4.7506,
      "mean_token_accuracy": 0.23906636536121367,
      "num_tokens": 64253005.0,
      "step": 28020
    },
    {
      "entropy": 5.323345518112182,
      "epoch": 2.692122958693564,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042791699571982606,
      "loss": 4.7586,
      "mean_token_accuracy": 0.24295444637537003,
      "num_tokens": 64264791.0,
      "step": 28025
    },
    {
      "entropy": 5.3165855884552,
      "epoch": 2.692603266090298,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004278918451030035,
      "loss": 4.6757,
      "mean_token_accuracy": 0.2468927651643753,
      "num_tokens": 64274501.0,
      "step": 28030
    },
    {
      "entropy": 5.318749332427979,
      "epoch": 2.6930835734870318,
      "grad_norm": 1.125,
      "learning_rate": 0.0004278666909364863,
      "loss": 4.6574,
      "mean_token_accuracy": 0.24791471064090728,
      "num_tokens": 64284501.0,
      "step": 28035
    },
    {
      "entropy": 5.2764880657196045,
      "epoch": 2.6935638808837656,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004278415332208584,
      "loss": 4.6719,
      "mean_token_accuracy": 0.24855909496545792,
      "num_tokens": 64295627.0,
      "step": 28040
    },
    {
      "entropy": 5.368828678131104,
      "epoch": 2.6940441882804995,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042781637195670396,
      "loss": 4.7859,
      "mean_token_accuracy": 0.23888262808322908,
      "num_tokens": 64307060.0,
      "step": 28045
    },
    {
      "entropy": 5.360940027236938,
      "epoch": 2.6945244956772334,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004277912071446072,
      "loss": 4.8112,
      "mean_token_accuracy": 0.24336155652999877,
      "num_tokens": 64318329.0,
      "step": 28050
    },
    {
      "entropy": 5.366869020462036,
      "epoch": 2.6950048030739673,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004277660387851522,
      "loss": 4.7372,
      "mean_token_accuracy": 0.24669528007507324,
      "num_tokens": 64329871.0,
      "step": 28055
    },
    {
      "entropy": 5.259217357635498,
      "epoch": 2.695485110470701,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004277408668789235,
      "loss": 4.7024,
      "mean_token_accuracy": 0.25340892374515533,
      "num_tokens": 64341950.0,
      "step": 28060
    },
    {
      "entropy": 5.341985034942627,
      "epoch": 2.695965417867435,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004277156914265054,
      "loss": 4.7806,
      "mean_token_accuracy": 0.24575734585523606,
      "num_tokens": 64353841.0,
      "step": 28065
    },
    {
      "entropy": 5.355032110214234,
      "epoch": 2.696445725264169,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004276905124284824,
      "loss": 4.6873,
      "mean_token_accuracy": 0.25155325829982755,
      "num_tokens": 64365914.0,
      "step": 28070
    },
    {
      "entropy": 5.26514368057251,
      "epoch": 2.696926032660903,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004276653298854391,
      "loss": 4.6437,
      "mean_token_accuracy": 0.25197167992591857,
      "num_tokens": 64376321.0,
      "step": 28075
    },
    {
      "entropy": 5.330988454818725,
      "epoch": 2.697406340057637,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004276401437979601,
      "loss": 4.7719,
      "mean_token_accuracy": 0.2455425277352333,
      "num_tokens": 64387344.0,
      "step": 28080
    },
    {
      "entropy": 5.300676965713501,
      "epoch": 2.6978866474543706,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004276149541666303,
      "loss": 4.6793,
      "mean_token_accuracy": 0.2532040074467659,
      "num_tokens": 64398553.0,
      "step": 28085
    },
    {
      "entropy": 5.380024194717407,
      "epoch": 2.698366954851105,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042758976099203444,
      "loss": 4.8269,
      "mean_token_accuracy": 0.2388285607099533,
      "num_tokens": 64409174.0,
      "step": 28090
    },
    {
      "entropy": 5.295743942260742,
      "epoch": 2.6988472622478388,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042756456427475736,
      "loss": 4.6384,
      "mean_token_accuracy": 0.2531014934182167,
      "num_tokens": 64420840.0,
      "step": 28095
    },
    {
      "entropy": 5.322788238525391,
      "epoch": 2.6993275696445727,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004275393640153842,
      "loss": 4.6404,
      "mean_token_accuracy": 0.25123409628868104,
      "num_tokens": 64431796.0,
      "step": 28100
    },
    {
      "entropy": 5.310386323928833,
      "epoch": 2.6998078770413065,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042751416021449986,
      "loss": 4.7096,
      "mean_token_accuracy": 0.24195995777845383,
      "num_tokens": 64443675.0,
      "step": 28105
    },
    {
      "entropy": 5.328419208526611,
      "epoch": 2.7002881844380404,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004274889528726896,
      "loss": 4.7076,
      "mean_token_accuracy": 0.2575435429811478,
      "num_tokens": 64455556.0,
      "step": 28110
    },
    {
      "entropy": 5.332771921157837,
      "epoch": 2.7007684918347743,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004274637419905388,
      "loss": 4.7052,
      "mean_token_accuracy": 0.24875894635915757,
      "num_tokens": 64466873.0,
      "step": 28115
    },
    {
      "entropy": 5.31856336593628,
      "epoch": 2.701248799231508,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042743852756863253,
      "loss": 4.7575,
      "mean_token_accuracy": 0.24655157923698426,
      "num_tokens": 64478530.0,
      "step": 28120
    },
    {
      "entropy": 5.334121942520142,
      "epoch": 2.701729106628242,
      "grad_norm": 1.125,
      "learning_rate": 0.0004274133096075563,
      "loss": 4.7285,
      "mean_token_accuracy": 0.24618494659662246,
      "num_tokens": 64488941.0,
      "step": 28125
    },
    {
      "entropy": 5.303606843948364,
      "epoch": 2.702209414024976,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004273880881078956,
      "loss": 4.6865,
      "mean_token_accuracy": 0.24622514098882675,
      "num_tokens": 64500727.0,
      "step": 28130
    },
    {
      "entropy": 5.299788522720337,
      "epoch": 2.70268972142171,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004273628630702359,
      "loss": 4.7046,
      "mean_token_accuracy": 0.24846078902482988,
      "num_tokens": 64513381.0,
      "step": 28135
    },
    {
      "entropy": 5.356375980377197,
      "epoch": 2.7031700288184437,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042733763449516313,
      "loss": 4.753,
      "mean_token_accuracy": 0.24410779923200607,
      "num_tokens": 64525442.0,
      "step": 28140
    },
    {
      "entropy": 5.415639448165893,
      "epoch": 2.7036503362151776,
      "grad_norm": 1.125,
      "learning_rate": 0.00042731240238326273,
      "loss": 4.6897,
      "mean_token_accuracy": 0.24260349273681642,
      "num_tokens": 64536008.0,
      "step": 28145
    },
    {
      "entropy": 5.379747915267944,
      "epoch": 2.7041306436119115,
      "grad_norm": 1.125,
      "learning_rate": 0.00042728716673512065,
      "loss": 4.8267,
      "mean_token_accuracy": 0.23886943012475967,
      "num_tokens": 64547161.0,
      "step": 28150
    },
    {
      "entropy": 5.317096710205078,
      "epoch": 2.704610951008646,
      "grad_norm": 1.328125,
      "learning_rate": 0.00042726192755132276,
      "loss": 4.7316,
      "mean_token_accuracy": 0.24251939356327057,
      "num_tokens": 64558152.0,
      "step": 28155
    },
    {
      "entropy": 5.41666808128357,
      "epoch": 2.7050912584053792,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042723668483245496,
      "loss": 4.878,
      "mean_token_accuracy": 0.23282308876514435,
      "num_tokens": 64569150.0,
      "step": 28160
    },
    {
      "entropy": 5.402124881744385,
      "epoch": 2.7055715658021136,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004272114385791035,
      "loss": 4.7162,
      "mean_token_accuracy": 0.2519373133778572,
      "num_tokens": 64580417.0,
      "step": 28165
    },
    {
      "entropy": 5.343264627456665,
      "epoch": 2.706051873198847,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042718618879185435,
      "loss": 4.681,
      "mean_token_accuracy": 0.24830176532268525,
      "num_tokens": 64591628.0,
      "step": 28170
    },
    {
      "entropy": 5.3377190113067625,
      "epoch": 2.7065321805955813,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004271609354712938,
      "loss": 4.7082,
      "mean_token_accuracy": 0.24730242490768434,
      "num_tokens": 64602801.0,
      "step": 28175
    },
    {
      "entropy": 5.344158172607422,
      "epoch": 2.707012487992315,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004271356786180082,
      "loss": 4.7443,
      "mean_token_accuracy": 0.2453673467040062,
      "num_tokens": 64614327.0,
      "step": 28180
    },
    {
      "entropy": 5.343890237808227,
      "epoch": 2.707492795389049,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004271104182325838,
      "loss": 4.7677,
      "mean_token_accuracy": 0.24300477504730225,
      "num_tokens": 64625508.0,
      "step": 28185
    },
    {
      "entropy": 5.319051074981689,
      "epoch": 2.707973102785783,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042708515431560723,
      "loss": 4.7702,
      "mean_token_accuracy": 0.24510857462882996,
      "num_tokens": 64637428.0,
      "step": 28190
    },
    {
      "entropy": 5.256767749786377,
      "epoch": 2.708453410182517,
      "grad_norm": 1.078125,
      "learning_rate": 0.000427059886867665,
      "loss": 4.7385,
      "mean_token_accuracy": 0.24567970782518386,
      "num_tokens": 64648189.0,
      "step": 28195
    },
    {
      "entropy": 5.291081953048706,
      "epoch": 2.7089337175792507,
      "grad_norm": 1.125,
      "learning_rate": 0.0004270346158893436,
      "loss": 4.6781,
      "mean_token_accuracy": 0.2501142919063568,
      "num_tokens": 64659250.0,
      "step": 28200
    },
    {
      "entropy": 5.29158091545105,
      "epoch": 2.7094140249759846,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042700934138123004,
      "loss": 4.6796,
      "mean_token_accuracy": 0.25422645807266236,
      "num_tokens": 64669746.0,
      "step": 28205
    },
    {
      "entropy": 5.330658674240112,
      "epoch": 2.7098943323727185,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042698406334391084,
      "loss": 4.7124,
      "mean_token_accuracy": 0.24680943936109542,
      "num_tokens": 64681758.0,
      "step": 28210
    },
    {
      "entropy": 5.388662528991699,
      "epoch": 2.7103746397694524,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000426958781777973,
      "loss": 4.7604,
      "mean_token_accuracy": 0.24649139046669005,
      "num_tokens": 64692944.0,
      "step": 28215
    },
    {
      "entropy": 5.437770223617553,
      "epoch": 2.7108549471661862,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004269334966840035,
      "loss": 4.8676,
      "mean_token_accuracy": 0.2368044972419739,
      "num_tokens": 64705436.0,
      "step": 28220
    },
    {
      "entropy": 5.407470655441284,
      "epoch": 2.71133525456292,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042690820806258933,
      "loss": 4.7381,
      "mean_token_accuracy": 0.24245813339948655,
      "num_tokens": 64716542.0,
      "step": 28225
    },
    {
      "entropy": 5.294199371337891,
      "epoch": 2.7118155619596545,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004268829159143176,
      "loss": 4.6616,
      "mean_token_accuracy": 0.24909802228212358,
      "num_tokens": 64727800.0,
      "step": 28230
    },
    {
      "entropy": 5.432557153701782,
      "epoch": 2.712295869356388,
      "grad_norm": 1.125,
      "learning_rate": 0.0004268576202397757,
      "loss": 4.8444,
      "mean_token_accuracy": 0.23729638904333114,
      "num_tokens": 64739790.0,
      "step": 28235
    },
    {
      "entropy": 5.3226995944976805,
      "epoch": 2.712776176753122,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004268323210395506,
      "loss": 4.7131,
      "mean_token_accuracy": 0.24494094401597977,
      "num_tokens": 64751031.0,
      "step": 28240
    },
    {
      "entropy": 5.306326913833618,
      "epoch": 2.7132564841498557,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042680701831423004,
      "loss": 4.7367,
      "mean_token_accuracy": 0.2432014226913452,
      "num_tokens": 64761751.0,
      "step": 28245
    },
    {
      "entropy": 5.435659456253052,
      "epoch": 2.71373679154659,
      "grad_norm": 1.125,
      "learning_rate": 0.0004267817120644012,
      "loss": 4.7787,
      "mean_token_accuracy": 0.24778858274221421,
      "num_tokens": 64773257.0,
      "step": 28250
    },
    {
      "entropy": 5.4222368717193605,
      "epoch": 2.714217098943324,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042675640229065167,
      "loss": 4.7307,
      "mean_token_accuracy": 0.2380976051092148,
      "num_tokens": 64783409.0,
      "step": 28255
    },
    {
      "entropy": 5.336287975311279,
      "epoch": 2.7146974063400577,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042673108899356915,
      "loss": 4.8106,
      "mean_token_accuracy": 0.23678110986948014,
      "num_tokens": 64794392.0,
      "step": 28260
    },
    {
      "entropy": 5.320166206359863,
      "epoch": 2.7151777137367916,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004267057721737413,
      "loss": 4.7908,
      "mean_token_accuracy": 0.2405206263065338,
      "num_tokens": 64806289.0,
      "step": 28265
    },
    {
      "entropy": 5.328478527069092,
      "epoch": 2.7156580211335255,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004266804518317559,
      "loss": 4.7661,
      "mean_token_accuracy": 0.2412852853536606,
      "num_tokens": 64817278.0,
      "step": 28270
    },
    {
      "entropy": 5.339891242980957,
      "epoch": 2.7161383285302594,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004266551279682008,
      "loss": 4.698,
      "mean_token_accuracy": 0.24301531612873079,
      "num_tokens": 64828889.0,
      "step": 28275
    },
    {
      "entropy": 5.29971661567688,
      "epoch": 2.7166186359269933,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004266298005836639,
      "loss": 4.6777,
      "mean_token_accuracy": 0.24758182764053344,
      "num_tokens": 64840260.0,
      "step": 28280
    },
    {
      "entropy": 5.365533256530762,
      "epoch": 2.717098943323727,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042660446967873327,
      "loss": 4.7866,
      "mean_token_accuracy": 0.2404222533106804,
      "num_tokens": 64853099.0,
      "step": 28285
    },
    {
      "entropy": 5.360078382492065,
      "epoch": 2.717579250720461,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042657913525399703,
      "loss": 4.7017,
      "mean_token_accuracy": 0.25267804265022276,
      "num_tokens": 64863501.0,
      "step": 28290
    },
    {
      "entropy": 5.297384405136109,
      "epoch": 2.718059558117195,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004265537973100435,
      "loss": 4.7075,
      "mean_token_accuracy": 0.25038125514984133,
      "num_tokens": 64874071.0,
      "step": 28295
    },
    {
      "entropy": 5.417505979537964,
      "epoch": 2.718539865513929,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004265284558474607,
      "loss": 4.7653,
      "mean_token_accuracy": 0.24008936882019044,
      "num_tokens": 64885238.0,
      "step": 28300
    },
    {
      "entropy": 5.398816156387329,
      "epoch": 2.7190201729106627,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042650311086683715,
      "loss": 4.825,
      "mean_token_accuracy": 0.2426741525530815,
      "num_tokens": 64896384.0,
      "step": 28305
    },
    {
      "entropy": 5.392080879211425,
      "epoch": 2.7195004803073966,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004264777623687612,
      "loss": 4.749,
      "mean_token_accuracy": 0.24483936727046968,
      "num_tokens": 64907219.0,
      "step": 28310
    },
    {
      "entropy": 5.35737943649292,
      "epoch": 2.719980787704131,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004264524103538214,
      "loss": 4.6815,
      "mean_token_accuracy": 0.24759158194065095,
      "num_tokens": 64918637.0,
      "step": 28315
    },
    {
      "entropy": 5.323477935791016,
      "epoch": 2.7204610951008643,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004264270548226064,
      "loss": 4.777,
      "mean_token_accuracy": 0.24223656505346297,
      "num_tokens": 64931532.0,
      "step": 28320
    },
    {
      "entropy": 5.354473876953125,
      "epoch": 2.7209414024975986,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004264016957757048,
      "loss": 4.7307,
      "mean_token_accuracy": 0.2445736825466156,
      "num_tokens": 64943214.0,
      "step": 28325
    },
    {
      "entropy": 5.352134799957275,
      "epoch": 2.7214217098943325,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042637633321370545,
      "loss": 4.791,
      "mean_token_accuracy": 0.24210274666547776,
      "num_tokens": 64954726.0,
      "step": 28330
    },
    {
      "entropy": 5.3514642238616945,
      "epoch": 2.7219020172910664,
      "grad_norm": 1.0,
      "learning_rate": 0.0004263509671371971,
      "loss": 4.7743,
      "mean_token_accuracy": 0.2370154082775116,
      "num_tokens": 64967759.0,
      "step": 28335
    },
    {
      "entropy": 5.42636137008667,
      "epoch": 2.7223823246878003,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042632559754676865,
      "loss": 4.8772,
      "mean_token_accuracy": 0.23138341754674913,
      "num_tokens": 64979431.0,
      "step": 28340
    },
    {
      "entropy": 5.323632907867432,
      "epoch": 2.722862632084534,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004263002244430092,
      "loss": 4.7171,
      "mean_token_accuracy": 0.2533188536763191,
      "num_tokens": 64992148.0,
      "step": 28345
    },
    {
      "entropy": 5.296235275268555,
      "epoch": 2.723342939481268,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004262748478265078,
      "loss": 4.6496,
      "mean_token_accuracy": 0.2535421386361122,
      "num_tokens": 65003750.0,
      "step": 28350
    },
    {
      "entropy": 5.262677717208862,
      "epoch": 2.723823246878002,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004262494676978537,
      "loss": 4.7086,
      "mean_token_accuracy": 0.2517756298184395,
      "num_tokens": 65013921.0,
      "step": 28355
    },
    {
      "entropy": 5.278525447845459,
      "epoch": 2.724303554274736,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042622408405763607,
      "loss": 4.6934,
      "mean_token_accuracy": 0.24837974309921265,
      "num_tokens": 65025043.0,
      "step": 28360
    },
    {
      "entropy": 5.3435484886169435,
      "epoch": 2.7247838616714697,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004261986969064442,
      "loss": 4.6473,
      "mean_token_accuracy": 0.25647803395986557,
      "num_tokens": 65037155.0,
      "step": 28365
    },
    {
      "entropy": 5.29276852607727,
      "epoch": 2.7252641690682036,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042617330624486753,
      "loss": 4.7343,
      "mean_token_accuracy": 0.241153547167778,
      "num_tokens": 65048876.0,
      "step": 28370
    },
    {
      "entropy": 5.37612829208374,
      "epoch": 2.7257444764649374,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004261479120734956,
      "loss": 4.7736,
      "mean_token_accuracy": 0.2422764778137207,
      "num_tokens": 65060321.0,
      "step": 28375
    },
    {
      "entropy": 5.405725526809692,
      "epoch": 2.7262247838616713,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000426122514392918,
      "loss": 4.7747,
      "mean_token_accuracy": 0.24464660435914992,
      "num_tokens": 65073019.0,
      "step": 28380
    },
    {
      "entropy": 5.3627103805542,
      "epoch": 2.726705091258405,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042609711320372435,
      "loss": 4.7296,
      "mean_token_accuracy": 0.24441371113061905,
      "num_tokens": 65084597.0,
      "step": 28385
    },
    {
      "entropy": 5.385254669189453,
      "epoch": 2.7271853986551395,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004260717085065045,
      "loss": 4.7241,
      "mean_token_accuracy": 0.24548205882310867,
      "num_tokens": 65097137.0,
      "step": 28390
    },
    {
      "entropy": 5.372491693496704,
      "epoch": 2.727665706051873,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042604630030184797,
      "loss": 4.8701,
      "mean_token_accuracy": 0.23410627841949463,
      "num_tokens": 65110229.0,
      "step": 28395
    },
    {
      "entropy": 5.373446130752564,
      "epoch": 2.7281460134486073,
      "grad_norm": 1.03125,
      "learning_rate": 0.000426020888590345,
      "loss": 4.8206,
      "mean_token_accuracy": 0.23574705570936202,
      "num_tokens": 65121767.0,
      "step": 28400
    },
    {
      "entropy": 5.292734003067016,
      "epoch": 2.728626320845341,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042599547337258536,
      "loss": 4.6366,
      "mean_token_accuracy": 0.2602911561727524,
      "num_tokens": 65132918.0,
      "step": 28405
    },
    {
      "entropy": 5.308487129211426,
      "epoch": 2.729106628242075,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042597005464915924,
      "loss": 4.7103,
      "mean_token_accuracy": 0.24950263351202012,
      "num_tokens": 65144384.0,
      "step": 28410
    },
    {
      "entropy": 5.429462337493897,
      "epoch": 2.729586935638809,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042594463242065674,
      "loss": 4.8441,
      "mean_token_accuracy": 0.23531299084424973,
      "num_tokens": 65157495.0,
      "step": 28415
    },
    {
      "entropy": 5.339891195297241,
      "epoch": 2.730067243035543,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004259192066876681,
      "loss": 4.6343,
      "mean_token_accuracy": 0.25900115966796877,
      "num_tokens": 65168583.0,
      "step": 28420
    },
    {
      "entropy": 5.324780368804932,
      "epoch": 2.7305475504322767,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042589377745078354,
      "loss": 4.744,
      "mean_token_accuracy": 0.24087867140769958,
      "num_tokens": 65179909.0,
      "step": 28425
    },
    {
      "entropy": 5.377510643005371,
      "epoch": 2.7310278578290106,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042586834471059366,
      "loss": 4.794,
      "mean_token_accuracy": 0.24247771948575975,
      "num_tokens": 65192239.0,
      "step": 28430
    },
    {
      "entropy": 5.297750997543335,
      "epoch": 2.7315081652257445,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042584290846768867,
      "loss": 4.7113,
      "mean_token_accuracy": 0.24646169692277908,
      "num_tokens": 65203454.0,
      "step": 28435
    },
    {
      "entropy": 5.258848428726196,
      "epoch": 2.7319884726224783,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004258174687226593,
      "loss": 4.6694,
      "mean_token_accuracy": 0.24589411467313765,
      "num_tokens": 65216265.0,
      "step": 28440
    },
    {
      "entropy": 5.370613479614258,
      "epoch": 2.7324687800192122,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004257920254760962,
      "loss": 4.6954,
      "mean_token_accuracy": 0.25114975273609164,
      "num_tokens": 65227742.0,
      "step": 28445
    },
    {
      "entropy": 5.327229309082031,
      "epoch": 2.732949087415946,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004257665787285899,
      "loss": 4.7555,
      "mean_token_accuracy": 0.2525359407067299,
      "num_tokens": 65239485.0,
      "step": 28450
    },
    {
      "entropy": 5.3002845287323,
      "epoch": 2.73342939481268,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042574112848073147,
      "loss": 4.6873,
      "mean_token_accuracy": 0.248943892121315,
      "num_tokens": 65252818.0,
      "step": 28455
    },
    {
      "entropy": 5.35421667098999,
      "epoch": 2.733909702209414,
      "grad_norm": 1.125,
      "learning_rate": 0.00042571567473311157,
      "loss": 4.7711,
      "mean_token_accuracy": 0.2436019405722618,
      "num_tokens": 65264636.0,
      "step": 28460
    },
    {
      "entropy": 5.308528184890747,
      "epoch": 2.734390009606148,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004256902174863213,
      "loss": 4.7176,
      "mean_token_accuracy": 0.2484564557671547,
      "num_tokens": 65275426.0,
      "step": 28465
    },
    {
      "entropy": 5.334276485443115,
      "epoch": 2.7348703170028816,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042566475674095155,
      "loss": 4.763,
      "mean_token_accuracy": 0.24492740035057067,
      "num_tokens": 65286241.0,
      "step": 28470
    },
    {
      "entropy": 5.349634265899658,
      "epoch": 2.735350624399616,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004256392924975936,
      "loss": 4.7153,
      "mean_token_accuracy": 0.24878531098365783,
      "num_tokens": 65298307.0,
      "step": 28475
    },
    {
      "entropy": 5.423646688461304,
      "epoch": 2.7358309317963494,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042561382475683854,
      "loss": 4.813,
      "mean_token_accuracy": 0.23944899290800095,
      "num_tokens": 65310529.0,
      "step": 28480
    },
    {
      "entropy": 5.32802414894104,
      "epoch": 2.7363112391930837,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004255883535192777,
      "loss": 4.725,
      "mean_token_accuracy": 0.2492177352309227,
      "num_tokens": 65321061.0,
      "step": 28485
    },
    {
      "entropy": 5.345407724380493,
      "epoch": 2.7367915465898176,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004255628787855025,
      "loss": 4.7138,
      "mean_token_accuracy": 0.2471391126513481,
      "num_tokens": 65332068.0,
      "step": 28490
    },
    {
      "entropy": 5.441101598739624,
      "epoch": 2.7372718539865515,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004255374005561043,
      "loss": 4.9146,
      "mean_token_accuracy": 0.23064900040626526,
      "num_tokens": 65344484.0,
      "step": 28495
    },
    {
      "entropy": 5.3642419338226315,
      "epoch": 2.7377521613832854,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042551191883167464,
      "loss": 4.7938,
      "mean_token_accuracy": 0.24408969432115554,
      "num_tokens": 65355021.0,
      "step": 28500
    },
    {
      "entropy": 5.298466157913208,
      "epoch": 2.7382324687800192,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004254864336128052,
      "loss": 4.6509,
      "mean_token_accuracy": 0.2461075633764267,
      "num_tokens": 65367024.0,
      "step": 28505
    },
    {
      "entropy": 5.335089540481567,
      "epoch": 2.738712776176753,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042546094490008765,
      "loss": 4.678,
      "mean_token_accuracy": 0.2489225074648857,
      "num_tokens": 65377894.0,
      "step": 28510
    },
    {
      "entropy": 5.359706211090088,
      "epoch": 2.739193083573487,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004254354526941136,
      "loss": 4.7113,
      "mean_token_accuracy": 0.24332261979579925,
      "num_tokens": 65389627.0,
      "step": 28515
    },
    {
      "entropy": 5.3342814445495605,
      "epoch": 2.739673390970221,
      "grad_norm": 1.125,
      "learning_rate": 0.0004254099569954751,
      "loss": 4.6424,
      "mean_token_accuracy": 0.2526775896549225,
      "num_tokens": 65399715.0,
      "step": 28520
    },
    {
      "entropy": 5.292548322677613,
      "epoch": 2.7401536983669548,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004253844578047641,
      "loss": 4.7482,
      "mean_token_accuracy": 0.250483027100563,
      "num_tokens": 65410459.0,
      "step": 28525
    },
    {
      "entropy": 5.27071738243103,
      "epoch": 2.7406340057636887,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004253589551225725,
      "loss": 4.6804,
      "mean_token_accuracy": 0.24960433095693588,
      "num_tokens": 65422412.0,
      "step": 28530
    },
    {
      "entropy": 5.352057933807373,
      "epoch": 2.7411143131604225,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042533344894949245,
      "loss": 4.7439,
      "mean_token_accuracy": 0.2448047697544098,
      "num_tokens": 65433982.0,
      "step": 28535
    },
    {
      "entropy": 5.2933165550231935,
      "epoch": 2.7415946205571564,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042530793928611605,
      "loss": 4.6889,
      "mean_token_accuracy": 0.2488004207611084,
      "num_tokens": 65445828.0,
      "step": 28540
    },
    {
      "entropy": 5.351819896697998,
      "epoch": 2.7420749279538903,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004252824261330357,
      "loss": 4.7065,
      "mean_token_accuracy": 0.2472889930009842,
      "num_tokens": 65457271.0,
      "step": 28545
    },
    {
      "entropy": 5.365768384933472,
      "epoch": 2.7425552353506246,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042525690949084364,
      "loss": 4.8014,
      "mean_token_accuracy": 0.24621607810258866,
      "num_tokens": 65468414.0,
      "step": 28550
    },
    {
      "entropy": 5.298802995681763,
      "epoch": 2.743035542747358,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042523138936013233,
      "loss": 4.6951,
      "mean_token_accuracy": 0.24702955782413483,
      "num_tokens": 65479454.0,
      "step": 28555
    },
    {
      "entropy": 5.367850971221924,
      "epoch": 2.7435158501440924,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042520586574149423,
      "loss": 4.797,
      "mean_token_accuracy": 0.24919554740190505,
      "num_tokens": 65490028.0,
      "step": 28560
    },
    {
      "entropy": 5.2853234767913815,
      "epoch": 2.7439961575408263,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042518033863552185,
      "loss": 4.6778,
      "mean_token_accuracy": 0.2512817353010178,
      "num_tokens": 65501929.0,
      "step": 28565
    },
    {
      "entropy": 5.3439820289611815,
      "epoch": 2.74447646493756,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004251548080428081,
      "loss": 4.6978,
      "mean_token_accuracy": 0.2497142732143402,
      "num_tokens": 65513654.0,
      "step": 28570
    },
    {
      "entropy": 5.278007698059082,
      "epoch": 2.744956772334294,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004251292739639455,
      "loss": 4.6763,
      "mean_token_accuracy": 0.2519522160291672,
      "num_tokens": 65525219.0,
      "step": 28575
    },
    {
      "entropy": 5.288584613800049,
      "epoch": 2.745437079731028,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042510373639952694,
      "loss": 4.6813,
      "mean_token_accuracy": 0.24749501943588256,
      "num_tokens": 65536255.0,
      "step": 28580
    },
    {
      "entropy": 5.422650241851807,
      "epoch": 2.745917387127762,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042507819535014547,
      "loss": 4.8182,
      "mean_token_accuracy": 0.23311312794685363,
      "num_tokens": 65547489.0,
      "step": 28585
    },
    {
      "entropy": 5.3523963451385494,
      "epoch": 2.7463976945244957,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042505265081639376,
      "loss": 4.6983,
      "mean_token_accuracy": 0.25043206512928007,
      "num_tokens": 65559892.0,
      "step": 28590
    },
    {
      "entropy": 5.462583255767822,
      "epoch": 2.7468780019212296,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004250271027988652,
      "loss": 4.8351,
      "mean_token_accuracy": 0.23671056926250458,
      "num_tokens": 65570742.0,
      "step": 28595
    },
    {
      "entropy": 5.366917848587036,
      "epoch": 2.7473583093179634,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042500155129815274,
      "loss": 4.8024,
      "mean_token_accuracy": 0.24088198095560073,
      "num_tokens": 65584202.0,
      "step": 28600
    },
    {
      "entropy": 5.328596687316894,
      "epoch": 2.7478386167146973,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042497599631484965,
      "loss": 4.6718,
      "mean_token_accuracy": 0.2489815890789032,
      "num_tokens": 65595744.0,
      "step": 28605
    },
    {
      "entropy": 5.267159366607666,
      "epoch": 2.748318924111431,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042495043784954926,
      "loss": 4.6792,
      "mean_token_accuracy": 0.24837034940719604,
      "num_tokens": 65607914.0,
      "step": 28610
    },
    {
      "entropy": 5.319956636428833,
      "epoch": 2.748799231508165,
      "grad_norm": 1.03125,
      "learning_rate": 0.000424924875902845,
      "loss": 4.7073,
      "mean_token_accuracy": 0.24572682678699492,
      "num_tokens": 65620134.0,
      "step": 28615
    },
    {
      "entropy": 5.367625522613525,
      "epoch": 2.749279538904899,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004248993104753303,
      "loss": 4.6995,
      "mean_token_accuracy": 0.2449987128376961,
      "num_tokens": 65631804.0,
      "step": 28620
    },
    {
      "entropy": 5.362853240966797,
      "epoch": 2.7497598463016333,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004248737415675987,
      "loss": 4.7717,
      "mean_token_accuracy": 0.24973810762166976,
      "num_tokens": 65644067.0,
      "step": 28625
    },
    {
      "entropy": 5.347786283493042,
      "epoch": 2.7502401536983667,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004248481691802439,
      "loss": 4.7795,
      "mean_token_accuracy": 0.24431742280721663,
      "num_tokens": 65655036.0,
      "step": 28630
    },
    {
      "entropy": 5.375054836273193,
      "epoch": 2.750720461095101,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004248225933138595,
      "loss": 4.7149,
      "mean_token_accuracy": 0.24131773710250853,
      "num_tokens": 65666827.0,
      "step": 28635
    },
    {
      "entropy": 5.426890182495117,
      "epoch": 2.751200768491835,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042479701396903945,
      "loss": 4.806,
      "mean_token_accuracy": 0.23702596873044968,
      "num_tokens": 65677635.0,
      "step": 28640
    },
    {
      "entropy": 5.315655755996704,
      "epoch": 2.751681075888569,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004247714311463775,
      "loss": 4.726,
      "mean_token_accuracy": 0.24141821265220642,
      "num_tokens": 65689695.0,
      "step": 28645
    },
    {
      "entropy": 5.339170837402344,
      "epoch": 2.7521613832853027,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042474584484646766,
      "loss": 4.676,
      "mean_token_accuracy": 0.2424202173948288,
      "num_tokens": 65700969.0,
      "step": 28650
    },
    {
      "entropy": 5.391905689239502,
      "epoch": 2.7526416906820366,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004247202550699039,
      "loss": 4.7686,
      "mean_token_accuracy": 0.23931529521942138,
      "num_tokens": 65712743.0,
      "step": 28655
    },
    {
      "entropy": 5.219114351272583,
      "epoch": 2.7531219980787704,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004246946618172805,
      "loss": 4.5855,
      "mean_token_accuracy": 0.2576104924082756,
      "num_tokens": 65724912.0,
      "step": 28660
    },
    {
      "entropy": 5.305458641052246,
      "epoch": 2.7536023054755043,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004246690650891915,
      "loss": 4.6873,
      "mean_token_accuracy": 0.24462884664535522,
      "num_tokens": 65736295.0,
      "step": 28665
    },
    {
      "entropy": 5.376247119903565,
      "epoch": 2.754082612872238,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004246434648862312,
      "loss": 4.7297,
      "mean_token_accuracy": 0.24312980175018312,
      "num_tokens": 65747032.0,
      "step": 28670
    },
    {
      "entropy": 5.382432699203491,
      "epoch": 2.754562920268972,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004246178612089941,
      "loss": 4.7523,
      "mean_token_accuracy": 0.245663845539093,
      "num_tokens": 65759001.0,
      "step": 28675
    },
    {
      "entropy": 5.311346626281738,
      "epoch": 2.755043227665706,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004245922540580744,
      "loss": 4.7324,
      "mean_token_accuracy": 0.2418140411376953,
      "num_tokens": 65769781.0,
      "step": 28680
    },
    {
      "entropy": 5.33437762260437,
      "epoch": 2.75552353506244,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004245666434340668,
      "loss": 4.7386,
      "mean_token_accuracy": 0.24597300142049788,
      "num_tokens": 65781281.0,
      "step": 28685
    },
    {
      "entropy": 5.389808702468872,
      "epoch": 2.7560038424591737,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004245410293375659,
      "loss": 4.8545,
      "mean_token_accuracy": 0.2326107457280159,
      "num_tokens": 65792884.0,
      "step": 28690
    },
    {
      "entropy": 5.352551221847534,
      "epoch": 2.7564841498559076,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004245154117691664,
      "loss": 4.6917,
      "mean_token_accuracy": 0.25172089487314225,
      "num_tokens": 65804616.0,
      "step": 28695
    },
    {
      "entropy": 5.333279228210449,
      "epoch": 2.756964457252642,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004244897907294628,
      "loss": 4.6993,
      "mean_token_accuracy": 0.24972668588161467,
      "num_tokens": 65814859.0,
      "step": 28700
    },
    {
      "entropy": 5.341541576385498,
      "epoch": 2.7574447646493754,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004244641662190504,
      "loss": 4.7011,
      "mean_token_accuracy": 0.24587175101041794,
      "num_tokens": 65824416.0,
      "step": 28705
    },
    {
      "entropy": 5.376867818832397,
      "epoch": 2.7579250720461097,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042443853823852376,
      "loss": 4.778,
      "mean_token_accuracy": 0.24316465109586716,
      "num_tokens": 65835428.0,
      "step": 28710
    },
    {
      "entropy": 5.366439533233643,
      "epoch": 2.7584053794428436,
      "grad_norm": 1.03125,
      "learning_rate": 0.000424412906788478,
      "loss": 4.7641,
      "mean_token_accuracy": 0.2385518193244934,
      "num_tokens": 65846683.0,
      "step": 28715
    },
    {
      "entropy": 5.386359119415284,
      "epoch": 2.7588856868395775,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004243872718695082,
      "loss": 4.8277,
      "mean_token_accuracy": 0.24143849909305573,
      "num_tokens": 65857850.0,
      "step": 28720
    },
    {
      "entropy": 5.254974222183227,
      "epoch": 2.7593659942363113,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042436163348220956,
      "loss": 4.6277,
      "mean_token_accuracy": 0.26173876971006393,
      "num_tokens": 65869688.0,
      "step": 28725
    },
    {
      "entropy": 5.355910348892212,
      "epoch": 2.7598463016330452,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004243359916271773,
      "loss": 4.6882,
      "mean_token_accuracy": 0.2503521054983139,
      "num_tokens": 65880705.0,
      "step": 28730
    },
    {
      "entropy": 5.305455255508423,
      "epoch": 2.760326609029779,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004243103463050067,
      "loss": 4.6656,
      "mean_token_accuracy": 0.24822425544261933,
      "num_tokens": 65894070.0,
      "step": 28735
    },
    {
      "entropy": 5.367801141738892,
      "epoch": 2.760806916426513,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004242846975162933,
      "loss": 4.8041,
      "mean_token_accuracy": 0.23859828859567642,
      "num_tokens": 65905899.0,
      "step": 28740
    },
    {
      "entropy": 5.299457168579101,
      "epoch": 2.761287223823247,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042425904526163246,
      "loss": 4.7519,
      "mean_token_accuracy": 0.24410845190286637,
      "num_tokens": 65917173.0,
      "step": 28745
    },
    {
      "entropy": 5.374064874649048,
      "epoch": 2.7617675312199808,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004242333895416198,
      "loss": 4.8117,
      "mean_token_accuracy": 0.23854345828294754,
      "num_tokens": 65930125.0,
      "step": 28750
    },
    {
      "entropy": 5.360675382614136,
      "epoch": 2.7622478386167146,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000424207730356851,
      "loss": 4.714,
      "mean_token_accuracy": 0.2462862104177475,
      "num_tokens": 65941569.0,
      "step": 28755
    },
    {
      "entropy": 5.31685881614685,
      "epoch": 2.7627281460134485,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004241820677079218,
      "loss": 4.6927,
      "mean_token_accuracy": 0.2460198298096657,
      "num_tokens": 65953970.0,
      "step": 28760
    },
    {
      "entropy": 5.318710660934448,
      "epoch": 2.7632084534101824,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042415640159542783,
      "loss": 4.716,
      "mean_token_accuracy": 0.2517242804169655,
      "num_tokens": 65965740.0,
      "step": 28765
    },
    {
      "entropy": 5.311764192581177,
      "epoch": 2.7636887608069163,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004241307320199652,
      "loss": 4.7343,
      "mean_token_accuracy": 0.24574301540851592,
      "num_tokens": 65975986.0,
      "step": 28770
    },
    {
      "entropy": 5.367891788482666,
      "epoch": 2.7641690682036506,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004241050589821298,
      "loss": 4.7924,
      "mean_token_accuracy": 0.24168431758880615,
      "num_tokens": 65988356.0,
      "step": 28775
    },
    {
      "entropy": 5.358522224426269,
      "epoch": 2.764649375600384,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004240793824825177,
      "loss": 4.6994,
      "mean_token_accuracy": 0.24372514337301254,
      "num_tokens": 65998526.0,
      "step": 28780
    },
    {
      "entropy": 5.290151929855346,
      "epoch": 2.7651296829971184,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042405370252172496,
      "loss": 4.6255,
      "mean_token_accuracy": 0.2602264106273651,
      "num_tokens": 66009605.0,
      "step": 28785
    },
    {
      "entropy": 5.273860740661621,
      "epoch": 2.765609990393852,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004240280191003479,
      "loss": 4.6434,
      "mean_token_accuracy": 0.25081934183835985,
      "num_tokens": 66020794.0,
      "step": 28790
    },
    {
      "entropy": 5.336876344680786,
      "epoch": 2.766090297790586,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004240023322189828,
      "loss": 4.7859,
      "mean_token_accuracy": 0.2500520944595337,
      "num_tokens": 66032174.0,
      "step": 28795
    },
    {
      "entropy": 5.3778526306152346,
      "epoch": 2.76657060518732,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004239766418782258,
      "loss": 4.6985,
      "mean_token_accuracy": 0.25255347341299056,
      "num_tokens": 66042552.0,
      "step": 28800
    },
    {
      "entropy": 5.271438694000244,
      "epoch": 2.767050912584054,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004239509480786737,
      "loss": 4.7009,
      "mean_token_accuracy": 0.2498173952102661,
      "num_tokens": 66053935.0,
      "step": 28805
    },
    {
      "entropy": 5.251417350769043,
      "epoch": 2.7675312199807878,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042392525082092286,
      "loss": 4.5745,
      "mean_token_accuracy": 0.2496820345520973,
      "num_tokens": 66064957.0,
      "step": 28810
    },
    {
      "entropy": 5.379936170578003,
      "epoch": 2.7680115273775217,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004238995501055699,
      "loss": 4.7813,
      "mean_token_accuracy": 0.24251348972320558,
      "num_tokens": 66076270.0,
      "step": 28815
    },
    {
      "entropy": 5.421066570281982,
      "epoch": 2.7684918347742555,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004238738459332115,
      "loss": 4.8177,
      "mean_token_accuracy": 0.24127038419246674,
      "num_tokens": 66088458.0,
      "step": 28820
    },
    {
      "entropy": 5.302821254730224,
      "epoch": 2.7689721421709894,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004238481383044445,
      "loss": 4.683,
      "mean_token_accuracy": 0.24802623093128204,
      "num_tokens": 66100114.0,
      "step": 28825
    },
    {
      "entropy": 5.348134565353393,
      "epoch": 2.7694524495677233,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042382242721986573,
      "loss": 4.7598,
      "mean_token_accuracy": 0.24235756695270538,
      "num_tokens": 66112448.0,
      "step": 28830
    },
    {
      "entropy": 5.351592206954956,
      "epoch": 2.769932756964457,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042379671268007207,
      "loss": 4.7512,
      "mean_token_accuracy": 0.24623372554779052,
      "num_tokens": 66124122.0,
      "step": 28835
    },
    {
      "entropy": 5.308913326263427,
      "epoch": 2.770413064361191,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004237709946856607,
      "loss": 4.6903,
      "mean_token_accuracy": 0.24356915503740312,
      "num_tokens": 66135375.0,
      "step": 28840
    },
    {
      "entropy": 5.364663887023926,
      "epoch": 2.770893371757925,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042374527323722836,
      "loss": 4.7713,
      "mean_token_accuracy": 0.2410118982195854,
      "num_tokens": 66148850.0,
      "step": 28845
    },
    {
      "entropy": 5.29100604057312,
      "epoch": 2.771373679154659,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042371954833537263,
      "loss": 4.6671,
      "mean_token_accuracy": 0.2551876962184906,
      "num_tokens": 66160040.0,
      "step": 28850
    },
    {
      "entropy": 5.340890407562256,
      "epoch": 2.7718539865513927,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042369381998069055,
      "loss": 4.7117,
      "mean_token_accuracy": 0.24425951093435289,
      "num_tokens": 66172857.0,
      "step": 28855
    },
    {
      "entropy": 5.283665561676026,
      "epoch": 2.772334293948127,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004236680881737795,
      "loss": 4.7132,
      "mean_token_accuracy": 0.24089475721120834,
      "num_tokens": 66184546.0,
      "step": 28860
    },
    {
      "entropy": 5.340497732162476,
      "epoch": 2.7728146013448605,
      "grad_norm": 1.125,
      "learning_rate": 0.0004236423529152369,
      "loss": 4.6913,
      "mean_token_accuracy": 0.24997561872005464,
      "num_tokens": 66195270.0,
      "step": 28865
    },
    {
      "entropy": 5.3328622341156,
      "epoch": 2.773294908741595,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004236166142056602,
      "loss": 4.7087,
      "mean_token_accuracy": 0.24731903225183488,
      "num_tokens": 66206998.0,
      "step": 28870
    },
    {
      "entropy": 5.385538482666016,
      "epoch": 2.7737752161383287,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004235908720456471,
      "loss": 4.7826,
      "mean_token_accuracy": 0.24168385416269303,
      "num_tokens": 66217645.0,
      "step": 28875
    },
    {
      "entropy": 5.3030847072601315,
      "epoch": 2.7742555235350626,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004235651264357951,
      "loss": 4.7071,
      "mean_token_accuracy": 0.2540661245584488,
      "num_tokens": 66228740.0,
      "step": 28880
    },
    {
      "entropy": 5.258869695663452,
      "epoch": 2.7747358309317964,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042353937737670206,
      "loss": 4.6001,
      "mean_token_accuracy": 0.2550571620464325,
      "num_tokens": 66239459.0,
      "step": 28885
    },
    {
      "entropy": 5.329424524307251,
      "epoch": 2.7752161383285303,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004235136248689658,
      "loss": 4.749,
      "mean_token_accuracy": 0.25173594504594804,
      "num_tokens": 66251470.0,
      "step": 28890
    },
    {
      "entropy": 5.379359531402588,
      "epoch": 2.775696445725264,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004234878689131841,
      "loss": 4.7443,
      "mean_token_accuracy": 0.2402291253209114,
      "num_tokens": 66262045.0,
      "step": 28895
    },
    {
      "entropy": 5.30416054725647,
      "epoch": 2.776176753121998,
      "grad_norm": 1.140625,
      "learning_rate": 0.000423462109509955,
      "loss": 4.6754,
      "mean_token_accuracy": 0.2508550316095352,
      "num_tokens": 66272077.0,
      "step": 28900
    },
    {
      "entropy": 5.306685876846314,
      "epoch": 2.776657060518732,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004234363466598765,
      "loss": 4.705,
      "mean_token_accuracy": 0.2491367980837822,
      "num_tokens": 66283091.0,
      "step": 28905
    },
    {
      "entropy": 5.377762365341186,
      "epoch": 2.777137367915466,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042341058036354687,
      "loss": 4.7655,
      "mean_token_accuracy": 0.24589539170265198,
      "num_tokens": 66293770.0,
      "step": 28910
    },
    {
      "entropy": 5.325264167785645,
      "epoch": 2.7776176753121997,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042338481062156424,
      "loss": 4.7078,
      "mean_token_accuracy": 0.2518277674913406,
      "num_tokens": 66305288.0,
      "step": 28915
    },
    {
      "entropy": 5.385660171508789,
      "epoch": 2.7780979827089336,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042335903743452694,
      "loss": 4.7678,
      "mean_token_accuracy": 0.246068274974823,
      "num_tokens": 66316414.0,
      "step": 28920
    },
    {
      "entropy": 5.401999568939209,
      "epoch": 2.7785782901056675,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004233332608030333,
      "loss": 4.7934,
      "mean_token_accuracy": 0.23625959157943727,
      "num_tokens": 66327573.0,
      "step": 28925
    },
    {
      "entropy": 5.37809386253357,
      "epoch": 2.7790585975024014,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042330748072768183,
      "loss": 4.7708,
      "mean_token_accuracy": 0.24848437309265137,
      "num_tokens": 66338950.0,
      "step": 28930
    },
    {
      "entropy": 5.281873798370361,
      "epoch": 2.7795389048991357,
      "grad_norm": 1.203125,
      "learning_rate": 0.000423281697209071,
      "loss": 4.6416,
      "mean_token_accuracy": 0.25449823290109636,
      "num_tokens": 66348814.0,
      "step": 28935
    },
    {
      "entropy": 5.282920503616333,
      "epoch": 2.780019212295869,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004232559102477995,
      "loss": 4.6517,
      "mean_token_accuracy": 0.2515536561608315,
      "num_tokens": 66360119.0,
      "step": 28940
    },
    {
      "entropy": 5.359606838226318,
      "epoch": 2.7804995196926034,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000423230119844466,
      "loss": 4.723,
      "mean_token_accuracy": 0.24631162583827973,
      "num_tokens": 66370843.0,
      "step": 28945
    },
    {
      "entropy": 5.383068323135376,
      "epoch": 2.7809798270893373,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004232043259996692,
      "loss": 4.6961,
      "mean_token_accuracy": 0.24702351540327072,
      "num_tokens": 66382420.0,
      "step": 28950
    },
    {
      "entropy": 5.278239965438843,
      "epoch": 2.781460134486071,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004231785287140081,
      "loss": 4.7047,
      "mean_token_accuracy": 0.24938927739858627,
      "num_tokens": 66394327.0,
      "step": 28955
    },
    {
      "entropy": 5.331783056259155,
      "epoch": 2.781940441882805,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004231527279880816,
      "loss": 4.7753,
      "mean_token_accuracy": 0.2407500848174095,
      "num_tokens": 66407927.0,
      "step": 28960
    },
    {
      "entropy": 5.298979187011719,
      "epoch": 2.782420749279539,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004231269238224885,
      "loss": 4.6153,
      "mean_token_accuracy": 0.2509156480431557,
      "num_tokens": 66419248.0,
      "step": 28965
    },
    {
      "entropy": 5.264204597473144,
      "epoch": 2.782901056676273,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004231011162178282,
      "loss": 4.6614,
      "mean_token_accuracy": 0.2519888311624527,
      "num_tokens": 66430750.0,
      "step": 28970
    },
    {
      "entropy": 5.345128679275513,
      "epoch": 2.7833813640730067,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004230753051746998,
      "loss": 4.7283,
      "mean_token_accuracy": 0.24299123734235764,
      "num_tokens": 66443792.0,
      "step": 28975
    },
    {
      "entropy": 5.338004207611084,
      "epoch": 2.7838616714697406,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042304949069370246,
      "loss": 4.6673,
      "mean_token_accuracy": 0.24895552098751067,
      "num_tokens": 66455091.0,
      "step": 28980
    },
    {
      "entropy": 5.335790205001831,
      "epoch": 2.7843419788664745,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042302367277543553,
      "loss": 4.7494,
      "mean_token_accuracy": 0.2522065654397011,
      "num_tokens": 66466269.0,
      "step": 28985
    },
    {
      "entropy": 5.323958206176758,
      "epoch": 2.7848222862632084,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042299785142049855,
      "loss": 4.6987,
      "mean_token_accuracy": 0.25008429288864137,
      "num_tokens": 66476032.0,
      "step": 28990
    },
    {
      "entropy": 5.307012414932251,
      "epoch": 2.7853025936599423,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004229720266294908,
      "loss": 4.7492,
      "mean_token_accuracy": 0.2477290987968445,
      "num_tokens": 66487265.0,
      "step": 28995
    },
    {
      "entropy": 5.436753416061402,
      "epoch": 2.785782901056676,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000422946198403012,
      "loss": 4.827,
      "mean_token_accuracy": 0.2437721908092499,
      "num_tokens": 66498762.0,
      "step": 29000
    },
    {
      "entropy": 5.437411069869995,
      "epoch": 2.78626320845341,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004229203667416619,
      "loss": 4.8377,
      "mean_token_accuracy": 0.2365734189748764,
      "num_tokens": 66509455.0,
      "step": 29005
    },
    {
      "entropy": 5.387892770767212,
      "epoch": 2.7867435158501443,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042289453164604,
      "loss": 4.8246,
      "mean_token_accuracy": 0.24378793090581893,
      "num_tokens": 66520858.0,
      "step": 29010
    },
    {
      "entropy": 5.335113286972046,
      "epoch": 2.787223823246878,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004228686931167463,
      "loss": 4.7476,
      "mean_token_accuracy": 0.23903078138828276,
      "num_tokens": 66532680.0,
      "step": 29015
    },
    {
      "entropy": 5.385481691360473,
      "epoch": 2.787704130643612,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004228428511543806,
      "loss": 4.7625,
      "mean_token_accuracy": 0.2464073896408081,
      "num_tokens": 66543554.0,
      "step": 29020
    },
    {
      "entropy": 5.3225099563598635,
      "epoch": 2.7881844380403455,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042281700575954283,
      "loss": 4.6476,
      "mean_token_accuracy": 0.2509810015559196,
      "num_tokens": 66554041.0,
      "step": 29025
    },
    {
      "entropy": 5.325544214248657,
      "epoch": 2.78866474543708,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004227911569328332,
      "loss": 4.6871,
      "mean_token_accuracy": 0.25097827464342115,
      "num_tokens": 66566216.0,
      "step": 29030
    },
    {
      "entropy": 5.331598234176636,
      "epoch": 2.7891450528338138,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004227653046748517,
      "loss": 4.7266,
      "mean_token_accuracy": 0.2451488584280014,
      "num_tokens": 66577964.0,
      "step": 29035
    },
    {
      "entropy": 5.350780248641968,
      "epoch": 2.7896253602305476,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042273944898619864,
      "loss": 4.7587,
      "mean_token_accuracy": 0.24923342019319533,
      "num_tokens": 66590172.0,
      "step": 29040
    },
    {
      "entropy": 5.2739002227783205,
      "epoch": 2.7901056676272815,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042271358986747427,
      "loss": 4.679,
      "mean_token_accuracy": 0.24976609498262406,
      "num_tokens": 66601277.0,
      "step": 29045
    },
    {
      "entropy": 5.328644418716431,
      "epoch": 2.7905859750240154,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042268772731927895,
      "loss": 4.6546,
      "mean_token_accuracy": 0.2576287806034088,
      "num_tokens": 66612080.0,
      "step": 29050
    },
    {
      "entropy": 5.305047130584716,
      "epoch": 2.7910662824207493,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042266186134221317,
      "loss": 4.712,
      "mean_token_accuracy": 0.24649282097816466,
      "num_tokens": 66623080.0,
      "step": 29055
    },
    {
      "entropy": 5.323803234100342,
      "epoch": 2.791546589817483,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004226359919368774,
      "loss": 4.7249,
      "mean_token_accuracy": 0.24626699537038804,
      "num_tokens": 66635486.0,
      "step": 29060
    },
    {
      "entropy": 5.290660285949707,
      "epoch": 2.792026897214217,
      "grad_norm": 1.34375,
      "learning_rate": 0.00042261011910387224,
      "loss": 4.7,
      "mean_token_accuracy": 0.2492271199822426,
      "num_tokens": 66647137.0,
      "step": 29065
    },
    {
      "entropy": 5.436800479888916,
      "epoch": 2.792507204610951,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004225842428437985,
      "loss": 4.822,
      "mean_token_accuracy": 0.23891291469335557,
      "num_tokens": 66658890.0,
      "step": 29070
    },
    {
      "entropy": 5.290129709243774,
      "epoch": 2.792987512007685,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042255836315725694,
      "loss": 4.695,
      "mean_token_accuracy": 0.24883728474378586,
      "num_tokens": 66670706.0,
      "step": 29075
    },
    {
      "entropy": 5.37441668510437,
      "epoch": 2.7934678194044187,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004225324800448483,
      "loss": 4.7576,
      "mean_token_accuracy": 0.23962825387716294,
      "num_tokens": 66681930.0,
      "step": 29080
    },
    {
      "entropy": 5.30542516708374,
      "epoch": 2.793948126801153,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042250659350717343,
      "loss": 4.6582,
      "mean_token_accuracy": 0.2530030757188797,
      "num_tokens": 66694224.0,
      "step": 29085
    },
    {
      "entropy": 5.399457883834839,
      "epoch": 2.7944284341978864,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042248070354483354,
      "loss": 4.7653,
      "mean_token_accuracy": 0.25249743461608887,
      "num_tokens": 66705905.0,
      "step": 29090
    },
    {
      "entropy": 5.346906948089599,
      "epoch": 2.7949087415946208,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004224548101584297,
      "loss": 4.7547,
      "mean_token_accuracy": 0.24231753200292588,
      "num_tokens": 66717189.0,
      "step": 29095
    },
    {
      "entropy": 5.317743873596191,
      "epoch": 2.795389048991354,
      "grad_norm": 1.09375,
      "learning_rate": 0.000422428913348563,
      "loss": 4.6623,
      "mean_token_accuracy": 0.2480239599943161,
      "num_tokens": 66728066.0,
      "step": 29100
    },
    {
      "entropy": 5.36374797821045,
      "epoch": 2.7958693563880885,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004224030131158346,
      "loss": 4.7511,
      "mean_token_accuracy": 0.2506965085864067,
      "num_tokens": 66738530.0,
      "step": 29105
    },
    {
      "entropy": 5.3938416004180905,
      "epoch": 2.7963496637848224,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004223771094608461,
      "loss": 4.8022,
      "mean_token_accuracy": 0.24283582121133804,
      "num_tokens": 66749563.0,
      "step": 29110
    },
    {
      "entropy": 5.367386531829834,
      "epoch": 2.7968299711815563,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004223512023841986,
      "loss": 4.7846,
      "mean_token_accuracy": 0.24295885860919952,
      "num_tokens": 66761376.0,
      "step": 29115
    },
    {
      "entropy": 5.315048694610596,
      "epoch": 2.79731027857829,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042232529188649374,
      "loss": 4.6627,
      "mean_token_accuracy": 0.24617561101913452,
      "num_tokens": 66774193.0,
      "step": 29120
    },
    {
      "entropy": 5.326666307449341,
      "epoch": 2.797790585975024,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004222993779683331,
      "loss": 4.7159,
      "mean_token_accuracy": 0.24216060042381288,
      "num_tokens": 66787007.0,
      "step": 29125
    },
    {
      "entropy": 5.329649353027344,
      "epoch": 2.798270893371758,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042227346063031837,
      "loss": 4.6721,
      "mean_token_accuracy": 0.2478941410779953,
      "num_tokens": 66798426.0,
      "step": 29130
    },
    {
      "entropy": 5.332892799377442,
      "epoch": 2.798751200768492,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004222475398730511,
      "loss": 4.7468,
      "mean_token_accuracy": 0.24660037606954574,
      "num_tokens": 66810843.0,
      "step": 29135
    },
    {
      "entropy": 5.276748895645142,
      "epoch": 2.7992315081652257,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004222216156971332,
      "loss": 4.6402,
      "mean_token_accuracy": 0.2551211193203926,
      "num_tokens": 66823350.0,
      "step": 29140
    },
    {
      "entropy": 5.327723979949951,
      "epoch": 2.7997118155619596,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042219568810316656,
      "loss": 4.7547,
      "mean_token_accuracy": 0.2505632698535919,
      "num_tokens": 66836830.0,
      "step": 29145
    },
    {
      "entropy": 5.424122428894043,
      "epoch": 2.8001921229586935,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004221697570917531,
      "loss": 4.821,
      "mean_token_accuracy": 0.23268102258443832,
      "num_tokens": 66848843.0,
      "step": 29150
    },
    {
      "entropy": 5.400762128829956,
      "epoch": 2.8006724303554273,
      "grad_norm": 1.171875,
      "learning_rate": 0.000422143822663495,
      "loss": 4.7481,
      "mean_token_accuracy": 0.24352713227272033,
      "num_tokens": 66858598.0,
      "step": 29155
    },
    {
      "entropy": 5.298258352279663,
      "epoch": 2.8011527377521612,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004221178848189941,
      "loss": 4.6996,
      "mean_token_accuracy": 0.247042915225029,
      "num_tokens": 66869633.0,
      "step": 29160
    },
    {
      "entropy": 5.3076887130737305,
      "epoch": 2.801633045148895,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042209194355885283,
      "loss": 4.7281,
      "mean_token_accuracy": 0.24867709279060363,
      "num_tokens": 66880501.0,
      "step": 29165
    },
    {
      "entropy": 5.355508089065552,
      "epoch": 2.8021133525456294,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004220659988836734,
      "loss": 4.7889,
      "mean_token_accuracy": 0.23721372485160827,
      "num_tokens": 66892488.0,
      "step": 29170
    },
    {
      "entropy": 5.283464574813843,
      "epoch": 2.802593659942363,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004220400507940582,
      "loss": 4.5989,
      "mean_token_accuracy": 0.25815806835889815,
      "num_tokens": 66903861.0,
      "step": 29175
    },
    {
      "entropy": 5.340424346923828,
      "epoch": 2.803073967339097,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042201409929060955,
      "loss": 4.6929,
      "mean_token_accuracy": 0.24503919929265977,
      "num_tokens": 66915526.0,
      "step": 29180
    },
    {
      "entropy": 5.274062204360962,
      "epoch": 2.803554274735831,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004219881443739301,
      "loss": 4.6836,
      "mean_token_accuracy": 0.25418801307678224,
      "num_tokens": 66926798.0,
      "step": 29185
    },
    {
      "entropy": 5.384524250030518,
      "epoch": 2.804034582132565,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004219621860446225,
      "loss": 4.8353,
      "mean_token_accuracy": 0.2333313435316086,
      "num_tokens": 66938185.0,
      "step": 29190
    },
    {
      "entropy": 5.339869785308838,
      "epoch": 2.804514889529299,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004219362243032892,
      "loss": 4.7387,
      "mean_token_accuracy": 0.25038043558597567,
      "num_tokens": 66950423.0,
      "step": 29195
    },
    {
      "entropy": 5.33274621963501,
      "epoch": 2.8049951969260327,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042191025915053323,
      "loss": 4.6991,
      "mean_token_accuracy": 0.2450547158718109,
      "num_tokens": 66961545.0,
      "step": 29200
    },
    {
      "entropy": 5.408564758300781,
      "epoch": 2.8054755043227666,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042188429058695714,
      "loss": 4.7538,
      "mean_token_accuracy": 0.240926156938076,
      "num_tokens": 66971932.0,
      "step": 29205
    },
    {
      "entropy": 5.2841963291168215,
      "epoch": 2.8059558117195005,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042185831861316406,
      "loss": 4.6887,
      "mean_token_accuracy": 0.25084990859031675,
      "num_tokens": 66983830.0,
      "step": 29210
    },
    {
      "entropy": 5.299205636978149,
      "epoch": 2.8064361191162344,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004218323432297568,
      "loss": 4.7464,
      "mean_token_accuracy": 0.24579211473464965,
      "num_tokens": 66995587.0,
      "step": 29215
    },
    {
      "entropy": 5.387013959884643,
      "epoch": 2.8069164265129682,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042180636443733864,
      "loss": 4.7728,
      "mean_token_accuracy": 0.24020078033208847,
      "num_tokens": 67007947.0,
      "step": 29220
    },
    {
      "entropy": 5.291222190856933,
      "epoch": 2.807396733909702,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042178038223651253,
      "loss": 4.6752,
      "mean_token_accuracy": 0.24531911611557006,
      "num_tokens": 67019708.0,
      "step": 29225
    },
    {
      "entropy": 5.309346342086792,
      "epoch": 2.807877041306436,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042175439662788195,
      "loss": 4.6777,
      "mean_token_accuracy": 0.25043275356292727,
      "num_tokens": 67030372.0,
      "step": 29230
    },
    {
      "entropy": 5.489475345611572,
      "epoch": 2.80835734870317,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042172840761204986,
      "loss": 4.8521,
      "mean_token_accuracy": 0.23985618203878403,
      "num_tokens": 67040614.0,
      "step": 29235
    },
    {
      "entropy": 5.404358053207398,
      "epoch": 2.8088376560999038,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004217024151896199,
      "loss": 4.7817,
      "mean_token_accuracy": 0.24855649918317796,
      "num_tokens": 67053409.0,
      "step": 29240
    },
    {
      "entropy": 5.401741027832031,
      "epoch": 2.809317963496638,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042167641936119557,
      "loss": 4.7652,
      "mean_token_accuracy": 0.2445250779390335,
      "num_tokens": 67065649.0,
      "step": 29245
    },
    {
      "entropy": 5.370772218704223,
      "epoch": 2.8097982708933715,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004216504201273802,
      "loss": 4.7868,
      "mean_token_accuracy": 0.2372426226735115,
      "num_tokens": 67077714.0,
      "step": 29250
    },
    {
      "entropy": 5.405280542373657,
      "epoch": 2.810278578290106,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004216244174887776,
      "loss": 4.7826,
      "mean_token_accuracy": 0.241496305167675,
      "num_tokens": 67089674.0,
      "step": 29255
    },
    {
      "entropy": 5.281787014007568,
      "epoch": 2.8107588856868397,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042159841144599145,
      "loss": 4.6792,
      "mean_token_accuracy": 0.2468097060918808,
      "num_tokens": 67101843.0,
      "step": 29260
    },
    {
      "entropy": 5.370062303543091,
      "epoch": 2.8112391930835736,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042157240199962537,
      "loss": 4.7541,
      "mean_token_accuracy": 0.2423197016119957,
      "num_tokens": 67113529.0,
      "step": 29265
    },
    {
      "entropy": 5.353477144241333,
      "epoch": 2.8117195004803075,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004215463891502834,
      "loss": 4.714,
      "mean_token_accuracy": 0.24595750123262405,
      "num_tokens": 67125067.0,
      "step": 29270
    },
    {
      "entropy": 5.298383188247681,
      "epoch": 2.8121998078770414,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042152037289856954,
      "loss": 4.6841,
      "mean_token_accuracy": 0.24831785559654235,
      "num_tokens": 67137550.0,
      "step": 29275
    },
    {
      "entropy": 5.298932695388794,
      "epoch": 2.8126801152737753,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042149435324508755,
      "loss": 4.6449,
      "mean_token_accuracy": 0.2530412614345551,
      "num_tokens": 67149730.0,
      "step": 29280
    },
    {
      "entropy": 5.382327651977539,
      "epoch": 2.813160422670509,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004214683301904417,
      "loss": 4.7554,
      "mean_token_accuracy": 0.24484000951051713,
      "num_tokens": 67161050.0,
      "step": 29285
    },
    {
      "entropy": 5.301955080032348,
      "epoch": 2.813640730067243,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042144230373523624,
      "loss": 4.717,
      "mean_token_accuracy": 0.2424033358693123,
      "num_tokens": 67173131.0,
      "step": 29290
    },
    {
      "entropy": 5.31696515083313,
      "epoch": 2.814121037463977,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004214162738800753,
      "loss": 4.7404,
      "mean_token_accuracy": 0.23956208676099777,
      "num_tokens": 67183723.0,
      "step": 29295
    },
    {
      "entropy": 5.386362504959107,
      "epoch": 2.814601344860711,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004213902406255632,
      "loss": 4.7688,
      "mean_token_accuracy": 0.24503443986177445,
      "num_tokens": 67196622.0,
      "step": 29300
    },
    {
      "entropy": 5.331689548492432,
      "epoch": 2.8150816522574447,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004213642039723044,
      "loss": 4.7212,
      "mean_token_accuracy": 0.24667247682809829,
      "num_tokens": 67208581.0,
      "step": 29305
    },
    {
      "entropy": 5.389964437484741,
      "epoch": 2.8155619596541785,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042133816392090343,
      "loss": 4.7444,
      "mean_token_accuracy": 0.24194575995206832,
      "num_tokens": 67220451.0,
      "step": 29310
    },
    {
      "entropy": 5.326496362686157,
      "epoch": 2.8160422670509124,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042131212047196484,
      "loss": 4.6945,
      "mean_token_accuracy": 0.2507534459233284,
      "num_tokens": 67232191.0,
      "step": 29315
    },
    {
      "entropy": 5.266912460327148,
      "epoch": 2.8165225744476468,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042128607362609317,
      "loss": 4.6038,
      "mean_token_accuracy": 0.2593701481819153,
      "num_tokens": 67242518.0,
      "step": 29320
    },
    {
      "entropy": 5.320117425918579,
      "epoch": 2.81700288184438,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042126002338389336,
      "loss": 4.8069,
      "mean_token_accuracy": 0.2390848994255066,
      "num_tokens": 67254598.0,
      "step": 29325
    },
    {
      "entropy": 5.291589784622192,
      "epoch": 2.8174831892411145,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042123396974597007,
      "loss": 4.6809,
      "mean_token_accuracy": 0.25385428965091705,
      "num_tokens": 67266054.0,
      "step": 29330
    },
    {
      "entropy": 5.404545497894287,
      "epoch": 2.817963496637848,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042120791271292823,
      "loss": 4.8396,
      "mean_token_accuracy": 0.2359533503651619,
      "num_tokens": 67277703.0,
      "step": 29335
    },
    {
      "entropy": 5.345236492156983,
      "epoch": 2.8184438040345823,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042118185228537283,
      "loss": 4.6955,
      "mean_token_accuracy": 0.25264197736978533,
      "num_tokens": 67288992.0,
      "step": 29340
    },
    {
      "entropy": 5.316795921325683,
      "epoch": 2.818924111431316,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042115578846390884,
      "loss": 4.6725,
      "mean_token_accuracy": 0.2533732160925865,
      "num_tokens": 67299728.0,
      "step": 29345
    },
    {
      "entropy": 5.378533124923706,
      "epoch": 2.81940441882805,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004211297212491414,
      "loss": 4.8278,
      "mean_token_accuracy": 0.2340795800089836,
      "num_tokens": 67310836.0,
      "step": 29350
    },
    {
      "entropy": 5.333154201507568,
      "epoch": 2.819884726224784,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004211036506416759,
      "loss": 4.7723,
      "mean_token_accuracy": 0.24330590516328812,
      "num_tokens": 67323426.0,
      "step": 29355
    },
    {
      "entropy": 5.354513216018677,
      "epoch": 2.820365033621518,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004210775766421173,
      "loss": 4.7524,
      "mean_token_accuracy": 0.2450658604502678,
      "num_tokens": 67334909.0,
      "step": 29360
    },
    {
      "entropy": 5.274520397186279,
      "epoch": 2.8208453410182517,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004210514992510713,
      "loss": 4.6715,
      "mean_token_accuracy": 0.24706070274114608,
      "num_tokens": 67347315.0,
      "step": 29365
    },
    {
      "entropy": 5.310046195983887,
      "epoch": 2.8213256484149856,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004210254184691431,
      "loss": 4.6691,
      "mean_token_accuracy": 0.2569255143404007,
      "num_tokens": 67358092.0,
      "step": 29370
    },
    {
      "entropy": 5.3700727939605715,
      "epoch": 2.8218059558117194,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042099933429693814,
      "loss": 4.8025,
      "mean_token_accuracy": 0.2493741810321808,
      "num_tokens": 67369318.0,
      "step": 29375
    },
    {
      "entropy": 5.369178581237793,
      "epoch": 2.8222862632084533,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004209732467350624,
      "loss": 4.7343,
      "mean_token_accuracy": 0.2468223437666893,
      "num_tokens": 67379604.0,
      "step": 29380
    },
    {
      "entropy": 5.364450979232788,
      "epoch": 2.822766570605187,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004209471557841212,
      "loss": 4.6762,
      "mean_token_accuracy": 0.256191785633564,
      "num_tokens": 67391134.0,
      "step": 29385
    },
    {
      "entropy": 5.351017761230469,
      "epoch": 2.823246878001921,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004209210614447204,
      "loss": 4.6633,
      "mean_token_accuracy": 0.25627495646476744,
      "num_tokens": 67400868.0,
      "step": 29390
    },
    {
      "entropy": 5.343967628479004,
      "epoch": 2.8237271853986554,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000420894963717466,
      "loss": 4.7204,
      "mean_token_accuracy": 0.23869194388389586,
      "num_tokens": 67412172.0,
      "step": 29395
    },
    {
      "entropy": 5.332471752166748,
      "epoch": 2.824207492795389,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004208688626029636,
      "loss": 4.6787,
      "mean_token_accuracy": 0.24337587654590606,
      "num_tokens": 67423700.0,
      "step": 29400
    },
    {
      "entropy": 5.376784610748291,
      "epoch": 2.824687800192123,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004208427581018194,
      "loss": 4.7935,
      "mean_token_accuracy": 0.2443583071231842,
      "num_tokens": 67436471.0,
      "step": 29405
    },
    {
      "entropy": 5.435949277877808,
      "epoch": 2.8251681075888566,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004208166502146394,
      "loss": 4.8442,
      "mean_token_accuracy": 0.23022285103797913,
      "num_tokens": 67447832.0,
      "step": 29410
    },
    {
      "entropy": 5.336432981491089,
      "epoch": 2.825648414985591,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042079053894202977,
      "loss": 4.7586,
      "mean_token_accuracy": 0.24914563745260238,
      "num_tokens": 67460467.0,
      "step": 29415
    },
    {
      "entropy": 5.310596895217896,
      "epoch": 2.826128722382325,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004207644242845968,
      "loss": 4.7136,
      "mean_token_accuracy": 0.2473523885011673,
      "num_tokens": 67471979.0,
      "step": 29420
    },
    {
      "entropy": 5.339407825469971,
      "epoch": 2.8266090297790587,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004207383062429467,
      "loss": 4.7331,
      "mean_token_accuracy": 0.24244083911180497,
      "num_tokens": 67483441.0,
      "step": 29425
    },
    {
      "entropy": 5.4403266429901125,
      "epoch": 2.8270893371757926,
      "grad_norm": 1.0,
      "learning_rate": 0.0004207121848176858,
      "loss": 4.8295,
      "mean_token_accuracy": 0.24310430586338044,
      "num_tokens": 67496691.0,
      "step": 29430
    },
    {
      "entropy": 5.384951877593994,
      "epoch": 2.8275696445725265,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042068606000942075,
      "loss": 4.7709,
      "mean_token_accuracy": 0.24035773128271104,
      "num_tokens": 67508120.0,
      "step": 29435
    },
    {
      "entropy": 5.366170024871826,
      "epoch": 2.8280499519692603,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042065993181875794,
      "loss": 4.8061,
      "mean_token_accuracy": 0.23378439843654633,
      "num_tokens": 67519699.0,
      "step": 29440
    },
    {
      "entropy": 5.343707752227783,
      "epoch": 2.8285302593659942,
      "grad_norm": 1.0625,
      "learning_rate": 0.000420633800246304,
      "loss": 4.8092,
      "mean_token_accuracy": 0.23850771486759187,
      "num_tokens": 67530842.0,
      "step": 29445
    },
    {
      "entropy": 5.392846155166626,
      "epoch": 2.829010566762728,
      "grad_norm": 1.125,
      "learning_rate": 0.00042060766529266577,
      "loss": 4.7703,
      "mean_token_accuracy": 0.24173525124788284,
      "num_tokens": 67543447.0,
      "step": 29450
    },
    {
      "entropy": 5.386161184310913,
      "epoch": 2.829490874159462,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042058152695844986,
      "loss": 4.7293,
      "mean_token_accuracy": 0.2516179233789444,
      "num_tokens": 67555339.0,
      "step": 29455
    },
    {
      "entropy": 5.316379690170288,
      "epoch": 2.829971181556196,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042055538524426317,
      "loss": 4.7082,
      "mean_token_accuracy": 0.24878898411989211,
      "num_tokens": 67566747.0,
      "step": 29460
    },
    {
      "entropy": 5.289395523071289,
      "epoch": 2.8304514889529298,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004205292401507127,
      "loss": 4.6649,
      "mean_token_accuracy": 0.25562499910593034,
      "num_tokens": 67577690.0,
      "step": 29465
    },
    {
      "entropy": 5.428337430953979,
      "epoch": 2.8309317963496636,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004205030916784053,
      "loss": 4.8931,
      "mean_token_accuracy": 0.23309322744607924,
      "num_tokens": 67589287.0,
      "step": 29470
    },
    {
      "entropy": 5.345920562744141,
      "epoch": 2.8314121037463975,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042047693982794824,
      "loss": 4.7189,
      "mean_token_accuracy": 0.2440275877714157,
      "num_tokens": 67602233.0,
      "step": 29475
    },
    {
      "entropy": 5.345448017120361,
      "epoch": 2.831892411143132,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042045078459994854,
      "loss": 4.6715,
      "mean_token_accuracy": 0.24947133362293245,
      "num_tokens": 67613076.0,
      "step": 29480
    },
    {
      "entropy": 5.341232252120972,
      "epoch": 2.8323727185398653,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004204246259950136,
      "loss": 4.7267,
      "mean_token_accuracy": 0.24360605627298354,
      "num_tokens": 67623517.0,
      "step": 29485
    },
    {
      "entropy": 5.378477907180786,
      "epoch": 2.8328530259365996,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042039846401375065,
      "loss": 4.7357,
      "mean_token_accuracy": 0.24323469400405884,
      "num_tokens": 67635982.0,
      "step": 29490
    },
    {
      "entropy": 5.3211455821990965,
      "epoch": 2.8333333333333335,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042037229865676714,
      "loss": 4.6395,
      "mean_token_accuracy": 0.24933079183101653,
      "num_tokens": 67646436.0,
      "step": 29495
    },
    {
      "entropy": 5.368142986297608,
      "epoch": 2.8338136407300674,
      "grad_norm": 1.125,
      "learning_rate": 0.00042034612992467046,
      "loss": 4.7147,
      "mean_token_accuracy": 0.25532507598400117,
      "num_tokens": 67657436.0,
      "step": 29500
    },
    {
      "entropy": 5.327057123184204,
      "epoch": 2.8342939481268012,
      "grad_norm": 1.125,
      "learning_rate": 0.0004203199578180683,
      "loss": 4.7261,
      "mean_token_accuracy": 0.24001633524894714,
      "num_tokens": 67669542.0,
      "step": 29505
    },
    {
      "entropy": 5.289640188217163,
      "epoch": 2.834774255523535,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004202937823375682,
      "loss": 4.622,
      "mean_token_accuracy": 0.2581586167216301,
      "num_tokens": 67680512.0,
      "step": 29510
    },
    {
      "entropy": 5.395693445205689,
      "epoch": 2.835254562920269,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004202676034837779,
      "loss": 4.6881,
      "mean_token_accuracy": 0.24808421730995178,
      "num_tokens": 67691392.0,
      "step": 29515
    },
    {
      "entropy": 5.35817289352417,
      "epoch": 2.835734870317003,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004202414212573052,
      "loss": 4.7293,
      "mean_token_accuracy": 0.24917559176683426,
      "num_tokens": 67701674.0,
      "step": 29520
    },
    {
      "entropy": 5.307111597061157,
      "epoch": 2.8362151777137368,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042021523565875796,
      "loss": 4.774,
      "mean_token_accuracy": 0.23978398442268373,
      "num_tokens": 67713960.0,
      "step": 29525
    },
    {
      "entropy": 5.318181324005127,
      "epoch": 2.8366954851104706,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004201890466887442,
      "loss": 4.6758,
      "mean_token_accuracy": 0.2514350041747093,
      "num_tokens": 67725644.0,
      "step": 29530
    },
    {
      "entropy": 5.316012763977051,
      "epoch": 2.8371757925072045,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004201628543478718,
      "loss": 4.6926,
      "mean_token_accuracy": 0.2515620902180672,
      "num_tokens": 67737997.0,
      "step": 29535
    },
    {
      "entropy": 5.334788799285889,
      "epoch": 2.8376560999039384,
      "grad_norm": 1.03125,
      "learning_rate": 0.000420136658636749,
      "loss": 4.7492,
      "mean_token_accuracy": 0.24489209800958633,
      "num_tokens": 67750379.0,
      "step": 29540
    },
    {
      "entropy": 5.357128667831421,
      "epoch": 2.8381364073006723,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004201104595559841,
      "loss": 4.7968,
      "mean_token_accuracy": 0.24101360738277436,
      "num_tokens": 67762567.0,
      "step": 29545
    },
    {
      "entropy": 5.343801212310791,
      "epoch": 2.838616714697406,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042008425710618507,
      "loss": 4.7434,
      "mean_token_accuracy": 0.24800380319356918,
      "num_tokens": 67774660.0,
      "step": 29550
    },
    {
      "entropy": 5.403322219848633,
      "epoch": 2.8390970220941405,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042005805128796043,
      "loss": 4.6998,
      "mean_token_accuracy": 0.25083619356155396,
      "num_tokens": 67784757.0,
      "step": 29555
    },
    {
      "entropy": 5.350348043441772,
      "epoch": 2.839577329490874,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004200318421019186,
      "loss": 4.7031,
      "mean_token_accuracy": 0.2461518257856369,
      "num_tokens": 67795723.0,
      "step": 29560
    },
    {
      "entropy": 5.4478801727294925,
      "epoch": 2.8400576368876083,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000420005629548668,
      "loss": 4.88,
      "mean_token_accuracy": 0.23753004670143127,
      "num_tokens": 67807685.0,
      "step": 29565
    },
    {
      "entropy": 5.333388423919677,
      "epoch": 2.840537944284342,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041997941362881735,
      "loss": 4.6685,
      "mean_token_accuracy": 0.25184024721384046,
      "num_tokens": 67818887.0,
      "step": 29570
    },
    {
      "entropy": 5.372644662857056,
      "epoch": 2.841018251681076,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004199531943429752,
      "loss": 4.7145,
      "mean_token_accuracy": 0.25363257974386216,
      "num_tokens": 67830210.0,
      "step": 29575
    },
    {
      "entropy": 5.306720495223999,
      "epoch": 2.84149855907781,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004199269716917502,
      "loss": 4.7509,
      "mean_token_accuracy": 0.24439150989055633,
      "num_tokens": 67841948.0,
      "step": 29580
    },
    {
      "entropy": 5.337637186050415,
      "epoch": 2.841978866474544,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004199007456757513,
      "loss": 4.7139,
      "mean_token_accuracy": 0.24692326188087463,
      "num_tokens": 67854117.0,
      "step": 29585
    },
    {
      "entropy": 5.384342336654663,
      "epoch": 2.8424591738712777,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041987451629558743,
      "loss": 4.7809,
      "mean_token_accuracy": 0.24113975018262862,
      "num_tokens": 67866004.0,
      "step": 29590
    },
    {
      "entropy": 5.36110291481018,
      "epoch": 2.8429394812680115,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004198482835518674,
      "loss": 4.6893,
      "mean_token_accuracy": 0.25678893029689787,
      "num_tokens": 67877178.0,
      "step": 29595
    },
    {
      "entropy": 5.287784433364868,
      "epoch": 2.8434197886647454,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004198220474452004,
      "loss": 4.6696,
      "mean_token_accuracy": 0.25563161373138427,
      "num_tokens": 67888183.0,
      "step": 29600
    },
    {
      "entropy": 5.409580564498901,
      "epoch": 2.8439000960614793,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004197958079761954,
      "loss": 4.7984,
      "mean_token_accuracy": 0.23482998907566072,
      "num_tokens": 67899300.0,
      "step": 29605
    },
    {
      "entropy": 5.346108722686767,
      "epoch": 2.844380403458213,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041976956514546185,
      "loss": 4.7007,
      "mean_token_accuracy": 0.25120815485715864,
      "num_tokens": 67909494.0,
      "step": 29610
    },
    {
      "entropy": 5.349529981613159,
      "epoch": 2.844860710854947,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041974331895360873,
      "loss": 4.7821,
      "mean_token_accuracy": 0.23994128853082658,
      "num_tokens": 67921241.0,
      "step": 29615
    },
    {
      "entropy": 5.297420310974121,
      "epoch": 2.845341018251681,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004197170694012456,
      "loss": 4.6582,
      "mean_token_accuracy": 0.2477112367749214,
      "num_tokens": 67933571.0,
      "step": 29620
    },
    {
      "entropy": 5.400738096237182,
      "epoch": 2.845821325648415,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004196908164889818,
      "loss": 4.772,
      "mean_token_accuracy": 0.24533444494009018,
      "num_tokens": 67944100.0,
      "step": 29625
    },
    {
      "entropy": 5.320158863067627,
      "epoch": 2.846301633045149,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004196645602174269,
      "loss": 4.683,
      "mean_token_accuracy": 0.2543098568916321,
      "num_tokens": 67955908.0,
      "step": 29630
    },
    {
      "entropy": 5.391142654418945,
      "epoch": 2.8467819404418826,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041963830058719046,
      "loss": 4.7678,
      "mean_token_accuracy": 0.23747845143079757,
      "num_tokens": 67966763.0,
      "step": 29635
    },
    {
      "entropy": 5.309642028808594,
      "epoch": 2.847262247838617,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004196120375988822,
      "loss": 4.6762,
      "mean_token_accuracy": 0.24341261237859727,
      "num_tokens": 67977668.0,
      "step": 29640
    },
    {
      "entropy": 5.304222488403321,
      "epoch": 2.8477425552353504,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004195857712531119,
      "loss": 4.7345,
      "mean_token_accuracy": 0.2426365628838539,
      "num_tokens": 67989429.0,
      "step": 29645
    },
    {
      "entropy": 5.3976672172546385,
      "epoch": 2.8482228626320847,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004195595015504892,
      "loss": 4.8056,
      "mean_token_accuracy": 0.24540923088788985,
      "num_tokens": 68000911.0,
      "step": 29650
    },
    {
      "entropy": 5.352900695800781,
      "epoch": 2.8487031700288186,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041953322849162415,
      "loss": 4.6762,
      "mean_token_accuracy": 0.24961501508951187,
      "num_tokens": 68012804.0,
      "step": 29655
    },
    {
      "entropy": 5.426133489608764,
      "epoch": 2.8491834774255524,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004195069520771268,
      "loss": 4.7699,
      "mean_token_accuracy": 0.24025824666023254,
      "num_tokens": 68023508.0,
      "step": 29660
    },
    {
      "entropy": 5.36446738243103,
      "epoch": 2.8496637848222863,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041948067230760706,
      "loss": 4.7802,
      "mean_token_accuracy": 0.2432662397623062,
      "num_tokens": 68035383.0,
      "step": 29665
    },
    {
      "entropy": 5.434145736694336,
      "epoch": 2.85014409221902,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041945438918367513,
      "loss": 4.7734,
      "mean_token_accuracy": 0.24042617082595824,
      "num_tokens": 68045922.0,
      "step": 29670
    },
    {
      "entropy": 5.235366678237915,
      "epoch": 2.850624399615754,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041942810270594115,
      "loss": 4.7014,
      "mean_token_accuracy": 0.24622589349746704,
      "num_tokens": 68057291.0,
      "step": 29675
    },
    {
      "entropy": 5.378088283538818,
      "epoch": 2.851104707012488,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004194018128750157,
      "loss": 4.8024,
      "mean_token_accuracy": 0.24333362877368928,
      "num_tokens": 68069693.0,
      "step": 29680
    },
    {
      "entropy": 5.410233354568481,
      "epoch": 2.851585014409222,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041937551969150873,
      "loss": 4.7649,
      "mean_token_accuracy": 0.24355264157056808,
      "num_tokens": 68079696.0,
      "step": 29685
    },
    {
      "entropy": 5.293911600112915,
      "epoch": 2.8520653218059557,
      "grad_norm": 1.0,
      "learning_rate": 0.000419349223156031,
      "loss": 4.7128,
      "mean_token_accuracy": 0.2475861743092537,
      "num_tokens": 68091330.0,
      "step": 29690
    },
    {
      "entropy": 5.312848377227783,
      "epoch": 2.8525456292026896,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004193229232691929,
      "loss": 4.6706,
      "mean_token_accuracy": 0.2515350252389908,
      "num_tokens": 68102499.0,
      "step": 29695
    },
    {
      "entropy": 5.371563911437988,
      "epoch": 2.8530259365994235,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041929662003160504,
      "loss": 4.7139,
      "mean_token_accuracy": 0.24351563602685927,
      "num_tokens": 68114563.0,
      "step": 29700
    },
    {
      "entropy": 5.346712303161621,
      "epoch": 2.8535062439961574,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041927031344387824,
      "loss": 4.757,
      "mean_token_accuracy": 0.24769661873579024,
      "num_tokens": 68124798.0,
      "step": 29705
    },
    {
      "entropy": 5.411809301376342,
      "epoch": 2.8539865513928913,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041924400350662304,
      "loss": 4.8383,
      "mean_token_accuracy": 0.2337000325322151,
      "num_tokens": 68136190.0,
      "step": 29710
    },
    {
      "entropy": 5.3970945358276365,
      "epoch": 2.8544668587896256,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041921769022045045,
      "loss": 4.7298,
      "mean_token_accuracy": 0.24725134372711183,
      "num_tokens": 68146368.0,
      "step": 29715
    },
    {
      "entropy": 5.407712841033936,
      "epoch": 2.854947166186359,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041919137358597137,
      "loss": 4.7995,
      "mean_token_accuracy": 0.24129583239555358,
      "num_tokens": 68159338.0,
      "step": 29720
    },
    {
      "entropy": 5.412119197845459,
      "epoch": 2.8554274735830933,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004191650536037967,
      "loss": 4.757,
      "mean_token_accuracy": 0.23718305230140685,
      "num_tokens": 68170906.0,
      "step": 29725
    },
    {
      "entropy": 5.4179364204406735,
      "epoch": 2.8559077809798272,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041913873027453756,
      "loss": 4.721,
      "mean_token_accuracy": 0.24703067243099214,
      "num_tokens": 68181963.0,
      "step": 29730
    },
    {
      "entropy": 5.27937502861023,
      "epoch": 2.856388088376561,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041911240359880517,
      "loss": 4.6618,
      "mean_token_accuracy": 0.25325340777635574,
      "num_tokens": 68193184.0,
      "step": 29735
    },
    {
      "entropy": 5.347622394561768,
      "epoch": 2.856868395773295,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041908607357721067,
      "loss": 4.7746,
      "mean_token_accuracy": 0.25005810260772704,
      "num_tokens": 68203984.0,
      "step": 29740
    },
    {
      "entropy": 5.357220029830932,
      "epoch": 2.857348703170029,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041905974021036533,
      "loss": 4.6796,
      "mean_token_accuracy": 0.25240590125322343,
      "num_tokens": 68215294.0,
      "step": 29745
    },
    {
      "entropy": 5.35086669921875,
      "epoch": 2.8578290105667628,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041903340349888065,
      "loss": 4.7629,
      "mean_token_accuracy": 0.24392421096563338,
      "num_tokens": 68226800.0,
      "step": 29750
    },
    {
      "entropy": 5.321911859512329,
      "epoch": 2.8583093179634966,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000419007063443368,
      "loss": 4.7049,
      "mean_token_accuracy": 0.253336301445961,
      "num_tokens": 68238868.0,
      "step": 29755
    },
    {
      "entropy": 5.36569037437439,
      "epoch": 2.8587896253602305,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041898072004443906,
      "loss": 4.746,
      "mean_token_accuracy": 0.24844890385866164,
      "num_tokens": 68250133.0,
      "step": 29760
    },
    {
      "entropy": 5.347324657440185,
      "epoch": 2.8592699327569644,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004189543733027052,
      "loss": 4.7723,
      "mean_token_accuracy": 0.24246647357940673,
      "num_tokens": 68261338.0,
      "step": 29765
    },
    {
      "entropy": 5.377807998657227,
      "epoch": 2.8597502401536983,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004189280232187783,
      "loss": 4.7379,
      "mean_token_accuracy": 0.2525330126285553,
      "num_tokens": 68272369.0,
      "step": 29770
    },
    {
      "entropy": 5.34290828704834,
      "epoch": 2.860230547550432,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004189016697932701,
      "loss": 4.762,
      "mean_token_accuracy": 0.2504107177257538,
      "num_tokens": 68283997.0,
      "step": 29775
    },
    {
      "entropy": 5.299617099761963,
      "epoch": 2.860710854947166,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004188753130267924,
      "loss": 4.702,
      "mean_token_accuracy": 0.24070680588483812,
      "num_tokens": 68295719.0,
      "step": 29780
    },
    {
      "entropy": 5.3546491146087645,
      "epoch": 2.8611911623439,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004188489529199572,
      "loss": 4.7804,
      "mean_token_accuracy": 0.24526326805353166,
      "num_tokens": 68307352.0,
      "step": 29785
    },
    {
      "entropy": 5.3346131324768065,
      "epoch": 2.8616714697406342,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041882258947337637,
      "loss": 4.6342,
      "mean_token_accuracy": 0.24800150394439696,
      "num_tokens": 68318889.0,
      "step": 29790
    },
    {
      "entropy": 5.2352570533752445,
      "epoch": 2.8621517771373677,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041879622268766207,
      "loss": 4.625,
      "mean_token_accuracy": 0.2524045780301094,
      "num_tokens": 68329681.0,
      "step": 29795
    },
    {
      "entropy": 5.337472867965698,
      "epoch": 2.862632084534102,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004187698525634266,
      "loss": 4.7517,
      "mean_token_accuracy": 0.25040945410728455,
      "num_tokens": 68340323.0,
      "step": 29800
    },
    {
      "entropy": 5.296299171447754,
      "epoch": 2.863112391930836,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041874347910128193,
      "loss": 4.5947,
      "mean_token_accuracy": 0.25520395636558535,
      "num_tokens": 68351732.0,
      "step": 29805
    },
    {
      "entropy": 5.371278476715088,
      "epoch": 2.8635926993275698,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004187171023018406,
      "loss": 4.7673,
      "mean_token_accuracy": 0.2410984992980957,
      "num_tokens": 68362847.0,
      "step": 29810
    },
    {
      "entropy": 5.3325999736785885,
      "epoch": 2.8640730067243036,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041869072216571486,
      "loss": 4.6953,
      "mean_token_accuracy": 0.24832213073968887,
      "num_tokens": 68374368.0,
      "step": 29815
    },
    {
      "entropy": 5.264779329299927,
      "epoch": 2.8645533141210375,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041866433869351715,
      "loss": 4.7006,
      "mean_token_accuracy": 0.24781553000211715,
      "num_tokens": 68386463.0,
      "step": 29820
    },
    {
      "entropy": 5.305309867858886,
      "epoch": 2.8650336215177714,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004186379518858602,
      "loss": 4.6855,
      "mean_token_accuracy": 0.25145955979824064,
      "num_tokens": 68397873.0,
      "step": 29825
    },
    {
      "entropy": 5.3807720184326175,
      "epoch": 2.8655139289145053,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004186115617433565,
      "loss": 4.7483,
      "mean_token_accuracy": 0.2384913831949234,
      "num_tokens": 68408561.0,
      "step": 29830
    },
    {
      "entropy": 5.333583879470825,
      "epoch": 2.865994236311239,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041858516826661876,
      "loss": 4.6558,
      "mean_token_accuracy": 0.2588202327489853,
      "num_tokens": 68419878.0,
      "step": 29835
    },
    {
      "entropy": 5.374958181381226,
      "epoch": 2.866474543707973,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00041855877145625974,
      "loss": 4.7548,
      "mean_token_accuracy": 0.24749660342931748,
      "num_tokens": 68429913.0,
      "step": 29840
    },
    {
      "entropy": 5.323921394348145,
      "epoch": 2.866954851104707,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004185323713128924,
      "loss": 4.7831,
      "mean_token_accuracy": 0.2503242716193199,
      "num_tokens": 68441881.0,
      "step": 29845
    },
    {
      "entropy": 5.403569316864013,
      "epoch": 2.867435158501441,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041850596783712956,
      "loss": 4.7127,
      "mean_token_accuracy": 0.24442313462495804,
      "num_tokens": 68453191.0,
      "step": 29850
    },
    {
      "entropy": 5.351122236251831,
      "epoch": 2.8679154658981747,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004184795610295843,
      "loss": 4.7259,
      "mean_token_accuracy": 0.24858428388834,
      "num_tokens": 68464373.0,
      "step": 29855
    },
    {
      "entropy": 5.337408542633057,
      "epoch": 2.8683957732949086,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004184531508908697,
      "loss": 4.7491,
      "mean_token_accuracy": 0.24769836515188218,
      "num_tokens": 68474970.0,
      "step": 29860
    },
    {
      "entropy": 5.323741722106933,
      "epoch": 2.868876080691643,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004184267374215989,
      "loss": 4.7271,
      "mean_token_accuracy": 0.2500372603535652,
      "num_tokens": 68486587.0,
      "step": 29865
    },
    {
      "entropy": 5.3444013595581055,
      "epoch": 2.8693563880883763,
      "grad_norm": 1.09375,
      "learning_rate": 0.000418400320622385,
      "loss": 4.7639,
      "mean_token_accuracy": 0.2463758960366249,
      "num_tokens": 68498091.0,
      "step": 29870
    },
    {
      "entropy": 5.343894529342651,
      "epoch": 2.8698366954851107,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004183739004938416,
      "loss": 4.7191,
      "mean_token_accuracy": 0.24606311619281768,
      "num_tokens": 68510431.0,
      "step": 29875
    },
    {
      "entropy": 5.250717210769653,
      "epoch": 2.870317002881844,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000418347477036582,
      "loss": 4.6379,
      "mean_token_accuracy": 0.25189622938632966,
      "num_tokens": 68522085.0,
      "step": 29880
    },
    {
      "entropy": 5.300277805328369,
      "epoch": 2.8707973102785784,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041832105025121956,
      "loss": 4.7408,
      "mean_token_accuracy": 0.24014814645051957,
      "num_tokens": 68533251.0,
      "step": 29885
    },
    {
      "entropy": 5.348627710342408,
      "epoch": 2.8712776176753123,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004182946201383679,
      "loss": 4.6898,
      "mean_token_accuracy": 0.24764188081026078,
      "num_tokens": 68545131.0,
      "step": 29890
    },
    {
      "entropy": 5.321270561218261,
      "epoch": 2.871757925072046,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004182681866986407,
      "loss": 4.6807,
      "mean_token_accuracy": 0.25161130875349047,
      "num_tokens": 68556838.0,
      "step": 29895
    },
    {
      "entropy": 5.327554559707641,
      "epoch": 2.87223823246878,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041824174993265165,
      "loss": 4.7053,
      "mean_token_accuracy": 0.25019753277301787,
      "num_tokens": 68570506.0,
      "step": 29900
    },
    {
      "entropy": 5.335172033309936,
      "epoch": 2.872718539865514,
      "grad_norm": 1.125,
      "learning_rate": 0.00041821530984101444,
      "loss": 4.696,
      "mean_token_accuracy": 0.24864032864570618,
      "num_tokens": 68581364.0,
      "step": 29905
    },
    {
      "entropy": 5.360363960266113,
      "epoch": 2.873198847262248,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000418188866424343,
      "loss": 4.7121,
      "mean_token_accuracy": 0.23878676146268846,
      "num_tokens": 68593047.0,
      "step": 29910
    },
    {
      "entropy": 5.35821008682251,
      "epoch": 2.8736791546589817,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004181624196832513,
      "loss": 4.7268,
      "mean_token_accuracy": 0.24319228529930115,
      "num_tokens": 68604592.0,
      "step": 29915
    },
    {
      "entropy": 5.3488929748535154,
      "epoch": 2.8741594620557156,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041813596961835336,
      "loss": 4.7372,
      "mean_token_accuracy": 0.2511095941066742,
      "num_tokens": 68616566.0,
      "step": 29920
    },
    {
      "entropy": 5.248818445205688,
      "epoch": 2.8746397694524495,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041810951623026313,
      "loss": 4.6192,
      "mean_token_accuracy": 0.25187728106975554,
      "num_tokens": 68626507.0,
      "step": 29925
    },
    {
      "entropy": 5.234066390991211,
      "epoch": 2.8751200768491834,
      "grad_norm": 1.125,
      "learning_rate": 0.00041808305951959496,
      "loss": 4.5931,
      "mean_token_accuracy": 0.25668464601039886,
      "num_tokens": 68637748.0,
      "step": 29930
    },
    {
      "entropy": 5.3220850944519045,
      "epoch": 2.8756003842459172,
      "grad_norm": 1.109375,
      "learning_rate": 0.000418056599486963,
      "loss": 4.7224,
      "mean_token_accuracy": 0.2493916094303131,
      "num_tokens": 68649379.0,
      "step": 29935
    },
    {
      "entropy": 5.392708587646484,
      "epoch": 2.8760806916426516,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004180301361329816,
      "loss": 4.7259,
      "mean_token_accuracy": 0.238064906001091,
      "num_tokens": 68660834.0,
      "step": 29940
    },
    {
      "entropy": 5.3248528957366945,
      "epoch": 2.876560999039385,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004180036694582651,
      "loss": 4.7521,
      "mean_token_accuracy": 0.24002485871315002,
      "num_tokens": 68673748.0,
      "step": 29945
    },
    {
      "entropy": 5.339818286895752,
      "epoch": 2.8770413064361193,
      "grad_norm": 1.125,
      "learning_rate": 0.00041797719946342813,
      "loss": 4.7388,
      "mean_token_accuracy": 0.24173986166715622,
      "num_tokens": 68684594.0,
      "step": 29950
    },
    {
      "entropy": 5.432256078720092,
      "epoch": 2.8775216138328528,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041795072614908503,
      "loss": 4.8328,
      "mean_token_accuracy": 0.23106811344623565,
      "num_tokens": 68697954.0,
      "step": 29955
    },
    {
      "entropy": 5.46600079536438,
      "epoch": 2.878001921229587,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041792424951585055,
      "loss": 4.8742,
      "mean_token_accuracy": 0.2453203707933426,
      "num_tokens": 68708616.0,
      "step": 29960
    },
    {
      "entropy": 5.219512176513672,
      "epoch": 2.878482228626321,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041789776956433947,
      "loss": 4.5884,
      "mean_token_accuracy": 0.25869095623493193,
      "num_tokens": 68720226.0,
      "step": 29965
    },
    {
      "entropy": 5.383338356018067,
      "epoch": 2.878962536023055,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041787128629516645,
      "loss": 4.7282,
      "mean_token_accuracy": 0.2437780097126961,
      "num_tokens": 68731134.0,
      "step": 29970
    },
    {
      "entropy": 5.362998962402344,
      "epoch": 2.8794428434197887,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004178447997089464,
      "loss": 4.7803,
      "mean_token_accuracy": 0.2463746562600136,
      "num_tokens": 68743158.0,
      "step": 29975
    },
    {
      "entropy": 5.315081977844239,
      "epoch": 2.8799231508165226,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004178183098062943,
      "loss": 4.728,
      "mean_token_accuracy": 0.2485504165291786,
      "num_tokens": 68755411.0,
      "step": 29980
    },
    {
      "entropy": 5.332859802246094,
      "epoch": 2.8804034582132565,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004177918165878251,
      "loss": 4.7523,
      "mean_token_accuracy": 0.24754575788974761,
      "num_tokens": 68767621.0,
      "step": 29985
    },
    {
      "entropy": 5.301483297348023,
      "epoch": 2.8808837656099904,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004177653200541539,
      "loss": 4.6797,
      "mean_token_accuracy": 0.2515884771943092,
      "num_tokens": 68778861.0,
      "step": 29990
    },
    {
      "entropy": 5.3271317958831785,
      "epoch": 2.8813640730067243,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000417738820205896,
      "loss": 4.7231,
      "mean_token_accuracy": 0.24500856548547745,
      "num_tokens": 68790101.0,
      "step": 29995
    },
    {
      "entropy": 5.226854419708252,
      "epoch": 2.881844380403458,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004177123170436665,
      "loss": 4.6067,
      "mean_token_accuracy": 0.25926271826028824,
      "num_tokens": 68801612.0,
      "step": 30000
    },
    {
      "epoch": 2.881844380403458,
      "eval_entropy": 5.0945900651317055,
      "eval_loss": 4.790685653686523,
      "eval_mean_token_accuracy": 0.25274330055271194,
      "eval_num_tokens": 68801612.0,
      "eval_runtime": 20.5036,
      "eval_samples_per_second": 1600.449,
      "eval_steps_per_second": 200.062,
      "step": 30000
    },
    {
      "entropy": 5.367634010314942,
      "epoch": 2.882324687800192,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004176858105680807,
      "loss": 4.8164,
      "mean_token_accuracy": 0.2442636102437973,
      "num_tokens": 68814322.0,
      "step": 30005
    },
    {
      "entropy": 5.333733367919922,
      "epoch": 2.882804995196926,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041765930077975415,
      "loss": 4.6457,
      "mean_token_accuracy": 0.25319206565618513,
      "num_tokens": 68825024.0,
      "step": 30010
    },
    {
      "entropy": 5.385110759735108,
      "epoch": 2.88328530259366,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041763278767930213,
      "loss": 4.6852,
      "mean_token_accuracy": 0.24989440590143203,
      "num_tokens": 68834751.0,
      "step": 30015
    },
    {
      "entropy": 5.200782823562622,
      "epoch": 2.8837656099903937,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004176062712673404,
      "loss": 4.5869,
      "mean_token_accuracy": 0.2612914234399796,
      "num_tokens": 68845613.0,
      "step": 30020
    },
    {
      "entropy": 5.31419563293457,
      "epoch": 2.884245917387128,
      "grad_norm": 1.125,
      "learning_rate": 0.0004175797515444845,
      "loss": 4.6558,
      "mean_token_accuracy": 0.2508396416902542,
      "num_tokens": 68856902.0,
      "step": 30025
    },
    {
      "entropy": 5.3942869186401365,
      "epoch": 2.8847262247838614,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004175532285113501,
      "loss": 4.7699,
      "mean_token_accuracy": 0.24534392952919007,
      "num_tokens": 68868612.0,
      "step": 30030
    },
    {
      "entropy": 5.307479667663574,
      "epoch": 2.8852065321805958,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004175267021685531,
      "loss": 4.6618,
      "mean_token_accuracy": 0.2525428727269173,
      "num_tokens": 68880213.0,
      "step": 30035
    },
    {
      "entropy": 5.362073183059692,
      "epoch": 2.8856868395773296,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041750017251670926,
      "loss": 4.716,
      "mean_token_accuracy": 0.25248685777187346,
      "num_tokens": 68891844.0,
      "step": 30040
    },
    {
      "entropy": 5.256273984909058,
      "epoch": 2.8861671469740635,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004174736395564345,
      "loss": 4.5897,
      "mean_token_accuracy": 0.24981555938720704,
      "num_tokens": 68904566.0,
      "step": 30045
    },
    {
      "entropy": 5.329090881347656,
      "epoch": 2.8866474543707974,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041744710328834493,
      "loss": 4.7509,
      "mean_token_accuracy": 0.2430603951215744,
      "num_tokens": 68917233.0,
      "step": 30050
    },
    {
      "entropy": 5.352755212783814,
      "epoch": 2.8871277617675313,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041742056371305665,
      "loss": 4.7078,
      "mean_token_accuracy": 0.24710696786642075,
      "num_tokens": 68928065.0,
      "step": 30055
    },
    {
      "entropy": 5.461927604675293,
      "epoch": 2.887608069164265,
      "grad_norm": 1.296875,
      "learning_rate": 0.00041739402083118576,
      "loss": 4.8846,
      "mean_token_accuracy": 0.233016762137413,
      "num_tokens": 68940130.0,
      "step": 30060
    },
    {
      "entropy": 5.305572414398194,
      "epoch": 2.888088376560999,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004173674746433485,
      "loss": 4.6295,
      "mean_token_accuracy": 0.2475050151348114,
      "num_tokens": 68951486.0,
      "step": 30065
    },
    {
      "entropy": 5.298180198669433,
      "epoch": 2.888568683957733,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041734092515016127,
      "loss": 4.6097,
      "mean_token_accuracy": 0.2544050201773643,
      "num_tokens": 68961988.0,
      "step": 30070
    },
    {
      "entropy": 5.280521583557129,
      "epoch": 2.889048991354467,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041731437235224036,
      "loss": 4.6994,
      "mean_token_accuracy": 0.25085514038801193,
      "num_tokens": 68973922.0,
      "step": 30075
    },
    {
      "entropy": 5.360568952560425,
      "epoch": 2.8895292987512007,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004172878162502023,
      "loss": 4.7508,
      "mean_token_accuracy": 0.24694794714450835,
      "num_tokens": 68985609.0,
      "step": 30080
    },
    {
      "entropy": 5.434072637557984,
      "epoch": 2.8900096061479346,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041726125684466374,
      "loss": 4.7745,
      "mean_token_accuracy": 0.23682600855827332,
      "num_tokens": 68997087.0,
      "step": 30085
    },
    {
      "entropy": 5.374877309799194,
      "epoch": 2.8904899135446684,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004172346941362412,
      "loss": 4.7279,
      "mean_token_accuracy": 0.24716778993606567,
      "num_tokens": 69008668.0,
      "step": 30090
    },
    {
      "entropy": 5.327781009674072,
      "epoch": 2.8909702209414023,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041720812812555137,
      "loss": 4.7044,
      "mean_token_accuracy": 0.25125478357076647,
      "num_tokens": 69019834.0,
      "step": 30095
    },
    {
      "entropy": 5.336311340332031,
      "epoch": 2.8914505283381366,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004171815588132111,
      "loss": 4.6812,
      "mean_token_accuracy": 0.2519001916050911,
      "num_tokens": 69031072.0,
      "step": 30100
    },
    {
      "entropy": 5.413577270507813,
      "epoch": 2.89193083573487,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004171549861998372,
      "loss": 4.8009,
      "mean_token_accuracy": 0.2355630189180374,
      "num_tokens": 69043670.0,
      "step": 30105
    },
    {
      "entropy": 5.32275071144104,
      "epoch": 2.8924111431316044,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004171284102860467,
      "loss": 4.6937,
      "mean_token_accuracy": 0.25472147613763807,
      "num_tokens": 69056756.0,
      "step": 30110
    },
    {
      "entropy": 5.384601593017578,
      "epoch": 2.8928914505283383,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004171018310724565,
      "loss": 4.7358,
      "mean_token_accuracy": 0.24441213309764862,
      "num_tokens": 69069406.0,
      "step": 30115
    },
    {
      "entropy": 5.32967438697815,
      "epoch": 2.893371757925072,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004170752485596838,
      "loss": 4.7275,
      "mean_token_accuracy": 0.24871093779802322,
      "num_tokens": 69081887.0,
      "step": 30120
    },
    {
      "entropy": 5.308402061462402,
      "epoch": 2.893852065321806,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041704866274834557,
      "loss": 4.6665,
      "mean_token_accuracy": 0.25016736090183256,
      "num_tokens": 69092272.0,
      "step": 30125
    },
    {
      "entropy": 5.361314725875855,
      "epoch": 2.89433237271854,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041702207363905933,
      "loss": 4.7432,
      "mean_token_accuracy": 0.24468905478715897,
      "num_tokens": 69103677.0,
      "step": 30130
    },
    {
      "entropy": 5.311538791656494,
      "epoch": 2.894812680115274,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041699548123244216,
      "loss": 4.6944,
      "mean_token_accuracy": 0.2515221178531647,
      "num_tokens": 69114863.0,
      "step": 30135
    },
    {
      "entropy": 5.420682811737061,
      "epoch": 2.8952929875120077,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004169688855291116,
      "loss": 4.7883,
      "mean_token_accuracy": 0.2398746654391289,
      "num_tokens": 69125998.0,
      "step": 30140
    },
    {
      "entropy": 5.482408475875855,
      "epoch": 2.8957732949087416,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004169422865296851,
      "loss": 4.8892,
      "mean_token_accuracy": 0.23108418583869933,
      "num_tokens": 69136706.0,
      "step": 30145
    },
    {
      "entropy": 5.4170557975769045,
      "epoch": 2.8962536023054755,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004169156842347802,
      "loss": 4.8234,
      "mean_token_accuracy": 0.23827784806489943,
      "num_tokens": 69147360.0,
      "step": 30150
    },
    {
      "entropy": 5.37821683883667,
      "epoch": 2.8967339097022093,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004168890786450144,
      "loss": 4.8367,
      "mean_token_accuracy": 0.23888125866651536,
      "num_tokens": 69159230.0,
      "step": 30155
    },
    {
      "entropy": 5.34818754196167,
      "epoch": 2.8972142170989432,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004168624697610056,
      "loss": 4.7836,
      "mean_token_accuracy": 0.24549729377031326,
      "num_tokens": 69170469.0,
      "step": 30160
    },
    {
      "entropy": 5.489170360565185,
      "epoch": 2.897694524495677,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041683585758337156,
      "loss": 4.816,
      "mean_token_accuracy": 0.2410888761281967,
      "num_tokens": 69182576.0,
      "step": 30165
    },
    {
      "entropy": 5.343117713928223,
      "epoch": 2.898174831892411,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004168092421127299,
      "loss": 4.6559,
      "mean_token_accuracy": 0.2517757698893547,
      "num_tokens": 69193471.0,
      "step": 30170
    },
    {
      "entropy": 5.323098039627075,
      "epoch": 2.8986551392891453,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004167826233496989,
      "loss": 4.709,
      "mean_token_accuracy": 0.24304148852825164,
      "num_tokens": 69204416.0,
      "step": 30175
    },
    {
      "entropy": 5.343581438064575,
      "epoch": 2.8991354466858787,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004167560012948963,
      "loss": 4.7306,
      "mean_token_accuracy": 0.24815452992916107,
      "num_tokens": 69215198.0,
      "step": 30180
    },
    {
      "entropy": 5.426870346069336,
      "epoch": 2.899615754082613,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041672937594894034,
      "loss": 4.8596,
      "mean_token_accuracy": 0.2351148769259453,
      "num_tokens": 69229015.0,
      "step": 30185
    },
    {
      "entropy": 5.509494209289551,
      "epoch": 2.9000960614793465,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041670274731244903,
      "loss": 4.8843,
      "mean_token_accuracy": 0.23288621604442597,
      "num_tokens": 69240559.0,
      "step": 30190
    },
    {
      "entropy": 5.310254812240601,
      "epoch": 2.900576368876081,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004166761153860408,
      "loss": 4.6936,
      "mean_token_accuracy": 0.2506261348724365,
      "num_tokens": 69251741.0,
      "step": 30195
    },
    {
      "entropy": 5.321112108230591,
      "epoch": 2.9010566762728147,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041664948017033383,
      "loss": 4.7171,
      "mean_token_accuracy": 0.2473279744386673,
      "num_tokens": 69264086.0,
      "step": 30200
    },
    {
      "entropy": 5.387391853332519,
      "epoch": 2.9015369836695486,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004166228416659465,
      "loss": 4.7858,
      "mean_token_accuracy": 0.24032191783189774,
      "num_tokens": 69275886.0,
      "step": 30205
    },
    {
      "entropy": 5.378679513931274,
      "epoch": 2.9020172910662825,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041659619987349734,
      "loss": 4.7767,
      "mean_token_accuracy": 0.23861713260412215,
      "num_tokens": 69286649.0,
      "step": 30210
    },
    {
      "entropy": 5.3558532238006595,
      "epoch": 2.9024975984630164,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00041656955479360487,
      "loss": 4.6481,
      "mean_token_accuracy": 0.25403028875589373,
      "num_tokens": 69296945.0,
      "step": 30215
    },
    {
      "entropy": 5.36610860824585,
      "epoch": 2.9029779058597502,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004165429064268877,
      "loss": 4.7696,
      "mean_token_accuracy": 0.2433772310614586,
      "num_tokens": 69307338.0,
      "step": 30220
    },
    {
      "entropy": 5.267567205429077,
      "epoch": 2.903458213256484,
      "grad_norm": 1.125,
      "learning_rate": 0.0004165162547739646,
      "loss": 4.664,
      "mean_token_accuracy": 0.2511794865131378,
      "num_tokens": 69318406.0,
      "step": 30225
    },
    {
      "entropy": 5.3023521423339846,
      "epoch": 2.903938520653218,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004164895998354542,
      "loss": 4.6629,
      "mean_token_accuracy": 0.24998723566532136,
      "num_tokens": 69328702.0,
      "step": 30230
    },
    {
      "entropy": 5.3579083442687985,
      "epoch": 2.904418828049952,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004164629416119755,
      "loss": 4.7783,
      "mean_token_accuracy": 0.24763328284025193,
      "num_tokens": 69340953.0,
      "step": 30235
    },
    {
      "entropy": 5.303630208969116,
      "epoch": 2.9048991354466858,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041643628010414735,
      "loss": 4.6686,
      "mean_token_accuracy": 0.2517788171768188,
      "num_tokens": 69350924.0,
      "step": 30240
    },
    {
      "entropy": 5.392197036743164,
      "epoch": 2.9053794428434196,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041640961531258877,
      "loss": 4.8232,
      "mean_token_accuracy": 0.23699825555086135,
      "num_tokens": 69361100.0,
      "step": 30245
    },
    {
      "entropy": 5.339563274383545,
      "epoch": 2.905859750240154,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004163829472379187,
      "loss": 4.6299,
      "mean_token_accuracy": 0.25980590134859083,
      "num_tokens": 69372032.0,
      "step": 30250
    },
    {
      "entropy": 5.270361375808716,
      "epoch": 2.9063400576368874,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041635627588075655,
      "loss": 4.6357,
      "mean_token_accuracy": 0.25654005706310273,
      "num_tokens": 69382828.0,
      "step": 30255
    },
    {
      "entropy": 5.376888704299927,
      "epoch": 2.9068203650336217,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004163296012417213,
      "loss": 4.7666,
      "mean_token_accuracy": 0.24344106912612914,
      "num_tokens": 69393508.0,
      "step": 30260
    },
    {
      "entropy": 5.371595239639282,
      "epoch": 2.907300672430355,
      "grad_norm": 1.125,
      "learning_rate": 0.00041630292332143245,
      "loss": 4.7301,
      "mean_token_accuracy": 0.2444403737783432,
      "num_tokens": 69405278.0,
      "step": 30265
    },
    {
      "entropy": 5.313630628585815,
      "epoch": 2.9077809798270895,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004162762421205093,
      "loss": 4.7467,
      "mean_token_accuracy": 0.24666361957788469,
      "num_tokens": 69416617.0,
      "step": 30270
    },
    {
      "entropy": 5.325506734848022,
      "epoch": 2.9082612872238234,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041624955763957134,
      "loss": 4.7187,
      "mean_token_accuracy": 0.25325511395931244,
      "num_tokens": 69427667.0,
      "step": 30275
    },
    {
      "entropy": 5.43550705909729,
      "epoch": 2.9087415946205573,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004162228698792381,
      "loss": 4.9026,
      "mean_token_accuracy": 0.23124375343322753,
      "num_tokens": 69438280.0,
      "step": 30280
    },
    {
      "entropy": 5.375145196914673,
      "epoch": 2.909221902017291,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041619617884012904,
      "loss": 4.6466,
      "mean_token_accuracy": 0.24356479197740555,
      "num_tokens": 69449042.0,
      "step": 30285
    },
    {
      "entropy": 5.344279432296753,
      "epoch": 2.909702209414025,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004161694845228641,
      "loss": 4.7006,
      "mean_token_accuracy": 0.2550384372472763,
      "num_tokens": 69459710.0,
      "step": 30290
    },
    {
      "entropy": 5.30465669631958,
      "epoch": 2.910182516810759,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004161427869280628,
      "loss": 4.7034,
      "mean_token_accuracy": 0.24372361302375795,
      "num_tokens": 69472543.0,
      "step": 30295
    },
    {
      "entropy": 5.315557432174683,
      "epoch": 2.910662824207493,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041611608605634517,
      "loss": 4.7065,
      "mean_token_accuracy": 0.24718896001577378,
      "num_tokens": 69485197.0,
      "step": 30300
    },
    {
      "entropy": 5.351318407058716,
      "epoch": 2.9111431316042267,
      "grad_norm": 1.109375,
      "learning_rate": 0.000416089381908331,
      "loss": 4.7877,
      "mean_token_accuracy": 0.2361114040017128,
      "num_tokens": 69496287.0,
      "step": 30305
    },
    {
      "entropy": 5.327165317535401,
      "epoch": 2.9116234390009605,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004160626744846404,
      "loss": 4.6895,
      "mean_token_accuracy": 0.2481841638684273,
      "num_tokens": 69508142.0,
      "step": 30310
    },
    {
      "entropy": 5.353477048873901,
      "epoch": 2.9121037463976944,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004160359637858933,
      "loss": 4.7255,
      "mean_token_accuracy": 0.24515368789434433,
      "num_tokens": 69520070.0,
      "step": 30315
    },
    {
      "entropy": 5.380671739578247,
      "epoch": 2.9125840537944283,
      "grad_norm": 1.125,
      "learning_rate": 0.00041600924981270997,
      "loss": 4.7938,
      "mean_token_accuracy": 0.245683690905571,
      "num_tokens": 69532448.0,
      "step": 30320
    },
    {
      "entropy": 5.465792560577393,
      "epoch": 2.913064361191162,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041598253256571057,
      "loss": 4.8501,
      "mean_token_accuracy": 0.22900635302066802,
      "num_tokens": 69543074.0,
      "step": 30325
    },
    {
      "entropy": 5.445922374725342,
      "epoch": 2.913544668587896,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004159558120455154,
      "loss": 4.753,
      "mean_token_accuracy": 0.24445367008447647,
      "num_tokens": 69555078.0,
      "step": 30330
    },
    {
      "entropy": 5.407000780105591,
      "epoch": 2.9140249759846304,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004159290882527448,
      "loss": 4.7667,
      "mean_token_accuracy": 0.24697088003158568,
      "num_tokens": 69567654.0,
      "step": 30335
    },
    {
      "entropy": 5.3447239875793455,
      "epoch": 2.914505283381364,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004159023611880192,
      "loss": 4.7684,
      "mean_token_accuracy": 0.2439172700047493,
      "num_tokens": 69579024.0,
      "step": 30340
    },
    {
      "entropy": 5.385036849975586,
      "epoch": 2.914985590778098,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004158756308519592,
      "loss": 4.772,
      "mean_token_accuracy": 0.2422686368227005,
      "num_tokens": 69590793.0,
      "step": 30345
    },
    {
      "entropy": 5.382667303085327,
      "epoch": 2.915465898174832,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041584889724518545,
      "loss": 4.7431,
      "mean_token_accuracy": 0.24350041449069976,
      "num_tokens": 69602605.0,
      "step": 30350
    },
    {
      "entropy": 5.343983745574951,
      "epoch": 2.915946205571566,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041582216036831844,
      "loss": 4.7362,
      "mean_token_accuracy": 0.24461946934461593,
      "num_tokens": 69614523.0,
      "step": 30355
    },
    {
      "entropy": 5.274375247955322,
      "epoch": 2.9164265129683,
      "grad_norm": 1.09375,
      "learning_rate": 0.000415795420221979,
      "loss": 4.6855,
      "mean_token_accuracy": 0.25081157833337786,
      "num_tokens": 69624551.0,
      "step": 30360
    },
    {
      "entropy": 5.262297534942627,
      "epoch": 2.9169068203650337,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041576867680678803,
      "loss": 4.6685,
      "mean_token_accuracy": 0.2529284358024597,
      "num_tokens": 69636363.0,
      "step": 30365
    },
    {
      "entropy": 5.4445374488830565,
      "epoch": 2.9173871277617676,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004157419301233664,
      "loss": 4.7909,
      "mean_token_accuracy": 0.24274491369724274,
      "num_tokens": 69647841.0,
      "step": 30370
    },
    {
      "entropy": 5.38100266456604,
      "epoch": 2.9178674351585014,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041571518017233505,
      "loss": 4.6635,
      "mean_token_accuracy": 0.2517387345433235,
      "num_tokens": 69657649.0,
      "step": 30375
    },
    {
      "entropy": 5.2882122039794925,
      "epoch": 2.9183477425552353,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004156884269543151,
      "loss": 4.6245,
      "mean_token_accuracy": 0.263099679350853,
      "num_tokens": 69668580.0,
      "step": 30380
    },
    {
      "entropy": 5.316878128051758,
      "epoch": 2.918828049951969,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004156616704699275,
      "loss": 4.696,
      "mean_token_accuracy": 0.2508876249194145,
      "num_tokens": 69679258.0,
      "step": 30385
    },
    {
      "entropy": 5.3408431053161625,
      "epoch": 2.919308357348703,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041563491071979375,
      "loss": 4.7732,
      "mean_token_accuracy": 0.2476000428199768,
      "num_tokens": 69691119.0,
      "step": 30390
    },
    {
      "entropy": 5.377898025512695,
      "epoch": 2.919788664745437,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041560814770453495,
      "loss": 4.7269,
      "mean_token_accuracy": 0.24968319982290268,
      "num_tokens": 69702827.0,
      "step": 30395
    },
    {
      "entropy": 5.347004127502442,
      "epoch": 2.920268972142171,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041558138142477235,
      "loss": 4.6573,
      "mean_token_accuracy": 0.24859154373407363,
      "num_tokens": 69715008.0,
      "step": 30400
    },
    {
      "entropy": 5.368579149246216,
      "epoch": 2.9207492795389047,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041555461188112763,
      "loss": 4.6973,
      "mean_token_accuracy": 0.24787055253982543,
      "num_tokens": 69725603.0,
      "step": 30405
    },
    {
      "entropy": 5.292243242263794,
      "epoch": 2.921229586935639,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041552783907422217,
      "loss": 4.6761,
      "mean_token_accuracy": 0.2496756359934807,
      "num_tokens": 69735733.0,
      "step": 30410
    },
    {
      "entropy": 5.248507499694824,
      "epoch": 2.9217098943323725,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004155010630046775,
      "loss": 4.6623,
      "mean_token_accuracy": 0.2503663063049316,
      "num_tokens": 69748078.0,
      "step": 30415
    },
    {
      "entropy": 5.281197929382325,
      "epoch": 2.922190201729107,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004154742836731155,
      "loss": 4.6847,
      "mean_token_accuracy": 0.2514675214886665,
      "num_tokens": 69759461.0,
      "step": 30420
    },
    {
      "entropy": 5.317628669738769,
      "epoch": 2.9226705091258407,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004154475010801576,
      "loss": 4.7091,
      "mean_token_accuracy": 0.2489780902862549,
      "num_tokens": 69770921.0,
      "step": 30425
    },
    {
      "entropy": 5.358846521377563,
      "epoch": 2.9231508165225746,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041542071522642583,
      "loss": 4.7604,
      "mean_token_accuracy": 0.24296114444732667,
      "num_tokens": 69782399.0,
      "step": 30430
    },
    {
      "entropy": 5.387794637680054,
      "epoch": 2.9236311239193085,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004153939261125421,
      "loss": 4.7204,
      "mean_token_accuracy": 0.251566170156002,
      "num_tokens": 69792584.0,
      "step": 30435
    },
    {
      "entropy": 5.3164294242858885,
      "epoch": 2.9241114313160423,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004153671337391281,
      "loss": 4.6613,
      "mean_token_accuracy": 0.2565651535987854,
      "num_tokens": 69803582.0,
      "step": 30440
    },
    {
      "entropy": 5.374818897247314,
      "epoch": 2.9245917387127762,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004153403381068062,
      "loss": 4.7189,
      "mean_token_accuracy": 0.2494154617190361,
      "num_tokens": 69815635.0,
      "step": 30445
    },
    {
      "entropy": 5.350349283218383,
      "epoch": 2.92507204610951,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041531353921619833,
      "loss": 4.7027,
      "mean_token_accuracy": 0.25143959522247317,
      "num_tokens": 69827260.0,
      "step": 30450
    },
    {
      "entropy": 5.226195001602173,
      "epoch": 2.925552353506244,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004152867370679267,
      "loss": 4.6518,
      "mean_token_accuracy": 0.25169351547956464,
      "num_tokens": 69839050.0,
      "step": 30455
    },
    {
      "entropy": 5.493021249771118,
      "epoch": 2.926032660902978,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041525993166261366,
      "loss": 4.906,
      "mean_token_accuracy": 0.23105596452951432,
      "num_tokens": 69849543.0,
      "step": 30460
    },
    {
      "entropy": 5.388900518417358,
      "epoch": 2.9265129682997117,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004152331230008814,
      "loss": 4.7343,
      "mean_token_accuracy": 0.2448473572731018,
      "num_tokens": 69861203.0,
      "step": 30465
    },
    {
      "entropy": 5.389406538009643,
      "epoch": 2.9269932756964456,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041520631108335254,
      "loss": 4.7485,
      "mean_token_accuracy": 0.2359006866812706,
      "num_tokens": 69873357.0,
      "step": 30470
    },
    {
      "entropy": 5.333707523345947,
      "epoch": 2.9274735830931795,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004151794959106494,
      "loss": 4.6933,
      "mean_token_accuracy": 0.25126419514417647,
      "num_tokens": 69885078.0,
      "step": 30475
    },
    {
      "entropy": 5.345600652694702,
      "epoch": 2.9279538904899134,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004151526774833947,
      "loss": 4.7292,
      "mean_token_accuracy": 0.2468865171074867,
      "num_tokens": 69897076.0,
      "step": 30480
    },
    {
      "entropy": 5.296048831939697,
      "epoch": 2.9284341978866477,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041512585580221086,
      "loss": 4.5994,
      "mean_token_accuracy": 0.2614844337105751,
      "num_tokens": 69907958.0,
      "step": 30485
    },
    {
      "entropy": 5.322109699249268,
      "epoch": 2.928914505283381,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004150990308677208,
      "loss": 4.7351,
      "mean_token_accuracy": 0.2463413432240486,
      "num_tokens": 69918684.0,
      "step": 30490
    },
    {
      "entropy": 5.414708423614502,
      "epoch": 2.9293948126801155,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041507220268054737,
      "loss": 4.7484,
      "mean_token_accuracy": 0.24819664061069488,
      "num_tokens": 69931458.0,
      "step": 30495
    },
    {
      "entropy": 5.390524101257324,
      "epoch": 2.929875120076849,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004150453712413131,
      "loss": 4.6887,
      "mean_token_accuracy": 0.24813174456357956,
      "num_tokens": 69941837.0,
      "step": 30500
    },
    {
      "entropy": 5.384254264831543,
      "epoch": 2.9303554274735832,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041501853655064134,
      "loss": 4.8079,
      "mean_token_accuracy": 0.24289435148239136,
      "num_tokens": 69954932.0,
      "step": 30505
    },
    {
      "entropy": 5.328463554382324,
      "epoch": 2.930835734870317,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004149916986091548,
      "loss": 4.7371,
      "mean_token_accuracy": 0.2521165147423744,
      "num_tokens": 69965990.0,
      "step": 30510
    },
    {
      "entropy": 5.323013973236084,
      "epoch": 2.931316042267051,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004149648574174768,
      "loss": 4.6492,
      "mean_token_accuracy": 0.2491701751947403,
      "num_tokens": 69977603.0,
      "step": 30515
    },
    {
      "entropy": 5.336713743209839,
      "epoch": 2.931796349663785,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004149380129762304,
      "loss": 4.6622,
      "mean_token_accuracy": 0.25322196036577227,
      "num_tokens": 69988043.0,
      "step": 30520
    },
    {
      "entropy": 5.289173412322998,
      "epoch": 2.9322766570605188,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004149111652860389,
      "loss": 4.6711,
      "mean_token_accuracy": 0.24953236132860185,
      "num_tokens": 69999397.0,
      "step": 30525
    },
    {
      "entropy": 5.351400327682495,
      "epoch": 2.9327569644572526,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004148843143475255,
      "loss": 4.7793,
      "mean_token_accuracy": 0.24530449956655503,
      "num_tokens": 70010673.0,
      "step": 30530
    },
    {
      "entropy": 5.2916947364807125,
      "epoch": 2.9332372718539865,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004148574601613137,
      "loss": 4.6744,
      "mean_token_accuracy": 0.25214419364929197,
      "num_tokens": 70022288.0,
      "step": 30535
    },
    {
      "entropy": 5.314777803421021,
      "epoch": 2.9337175792507204,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004148306027280271,
      "loss": 4.6604,
      "mean_token_accuracy": 0.2541650265455246,
      "num_tokens": 70033918.0,
      "step": 30540
    },
    {
      "entropy": 5.271858501434326,
      "epoch": 2.9341978866474543,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041480374204828896,
      "loss": 4.6619,
      "mean_token_accuracy": 0.2533440038561821,
      "num_tokens": 70046072.0,
      "step": 30545
    },
    {
      "entropy": 5.3774065494537355,
      "epoch": 2.934678194044188,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041477687812272314,
      "loss": 4.7619,
      "mean_token_accuracy": 0.24719048738479615,
      "num_tokens": 70057977.0,
      "step": 30550
    },
    {
      "entropy": 5.429800319671631,
      "epoch": 2.935158501440922,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041475001095195324,
      "loss": 4.7406,
      "mean_token_accuracy": 0.24676755368709563,
      "num_tokens": 70069000.0,
      "step": 30555
    },
    {
      "entropy": 5.351637983322144,
      "epoch": 2.9356388088376564,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004147231405366031,
      "loss": 4.7271,
      "mean_token_accuracy": 0.24752712547779082,
      "num_tokens": 70080648.0,
      "step": 30560
    },
    {
      "entropy": 5.304177856445312,
      "epoch": 2.93611911623439,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004146962668772965,
      "loss": 4.6856,
      "mean_token_accuracy": 0.25292500108480453,
      "num_tokens": 70092097.0,
      "step": 30565
    },
    {
      "entropy": 5.335848140716553,
      "epoch": 2.936599423631124,
      "grad_norm": 1.125,
      "learning_rate": 0.00041466938997465744,
      "loss": 4.6369,
      "mean_token_accuracy": 0.2537738710641861,
      "num_tokens": 70103665.0,
      "step": 30570
    },
    {
      "entropy": 5.311061716079712,
      "epoch": 2.9370797310278576,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041464250982930974,
      "loss": 4.6507,
      "mean_token_accuracy": 0.2572553977370262,
      "num_tokens": 70115587.0,
      "step": 30575
    },
    {
      "entropy": 5.292112922668457,
      "epoch": 2.937560038424592,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041461562644187777,
      "loss": 4.7413,
      "mean_token_accuracy": 0.24813725650310517,
      "num_tokens": 70126312.0,
      "step": 30580
    },
    {
      "entropy": 5.258976984024048,
      "epoch": 2.938040345821326,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041458873981298547,
      "loss": 4.5957,
      "mean_token_accuracy": 0.26464676707983015,
      "num_tokens": 70136688.0,
      "step": 30585
    },
    {
      "entropy": 5.337756156921387,
      "epoch": 2.9385206532180597,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041456184994325714,
      "loss": 4.7175,
      "mean_token_accuracy": 0.24647579044103624,
      "num_tokens": 70149169.0,
      "step": 30590
    },
    {
      "entropy": 5.460492563247681,
      "epoch": 2.9390009606147935,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041453495683331694,
      "loss": 4.8005,
      "mean_token_accuracy": 0.23614963591098787,
      "num_tokens": 70160246.0,
      "step": 30595
    },
    {
      "entropy": 5.375581216812134,
      "epoch": 2.9394812680115274,
      "grad_norm": 1.25,
      "learning_rate": 0.00041450806048378954,
      "loss": 4.67,
      "mean_token_accuracy": 0.25495648831129075,
      "num_tokens": 70171553.0,
      "step": 30600
    },
    {
      "entropy": 5.34877643585205,
      "epoch": 2.9399615754082613,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004144811608952992,
      "loss": 4.7283,
      "mean_token_accuracy": 0.24836386889219284,
      "num_tokens": 70182675.0,
      "step": 30605
    },
    {
      "entropy": 5.356368064880371,
      "epoch": 2.940441882804995,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041445425806847043,
      "loss": 4.7276,
      "mean_token_accuracy": 0.24901390224695205,
      "num_tokens": 70194222.0,
      "step": 30610
    },
    {
      "entropy": 5.3331787109375,
      "epoch": 2.940922190201729,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041442735200392783,
      "loss": 4.6216,
      "mean_token_accuracy": 0.25227846801280973,
      "num_tokens": 70204735.0,
      "step": 30615
    },
    {
      "entropy": 5.44020938873291,
      "epoch": 2.941402497598463,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004144004427022962,
      "loss": 4.8223,
      "mean_token_accuracy": 0.23912595063447953,
      "num_tokens": 70216368.0,
      "step": 30620
    },
    {
      "entropy": 5.350064468383789,
      "epoch": 2.941882804995197,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041437353016420025,
      "loss": 4.7654,
      "mean_token_accuracy": 0.239189276099205,
      "num_tokens": 70229497.0,
      "step": 30625
    },
    {
      "entropy": 5.345599794387818,
      "epoch": 2.9423631123919307,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004143466143902648,
      "loss": 4.6868,
      "mean_token_accuracy": 0.2489429995417595,
      "num_tokens": 70240826.0,
      "step": 30630
    },
    {
      "entropy": 5.352196311950683,
      "epoch": 2.9428434197886646,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041431969538111463,
      "loss": 4.7827,
      "mean_token_accuracy": 0.24051163643598555,
      "num_tokens": 70253337.0,
      "step": 30635
    },
    {
      "entropy": 5.366906595230103,
      "epoch": 2.9433237271853985,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004142927731373749,
      "loss": 4.6778,
      "mean_token_accuracy": 0.2517224371433258,
      "num_tokens": 70263881.0,
      "step": 30640
    },
    {
      "entropy": 5.332530117034912,
      "epoch": 2.943804034582133,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004142658476596706,
      "loss": 4.7185,
      "mean_token_accuracy": 0.2553650036454201,
      "num_tokens": 70276321.0,
      "step": 30645
    },
    {
      "entropy": 5.277769136428833,
      "epoch": 2.9442843419788662,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041423891894862687,
      "loss": 4.6979,
      "mean_token_accuracy": 0.24931949824094773,
      "num_tokens": 70288778.0,
      "step": 30650
    },
    {
      "entropy": 5.470089673995972,
      "epoch": 2.9447646493756006,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004142119870048689,
      "loss": 4.858,
      "mean_token_accuracy": 0.2338413640856743,
      "num_tokens": 70300338.0,
      "step": 30655
    },
    {
      "entropy": 5.462717199325562,
      "epoch": 2.9452449567723344,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004141850518290219,
      "loss": 4.7862,
      "mean_token_accuracy": 0.2391010418534279,
      "num_tokens": 70312682.0,
      "step": 30660
    },
    {
      "entropy": 5.403351879119873,
      "epoch": 2.9457252641690683,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041415811342171134,
      "loss": 4.7684,
      "mean_token_accuracy": 0.24511202573776245,
      "num_tokens": 70324408.0,
      "step": 30665
    },
    {
      "entropy": 5.263695096969604,
      "epoch": 2.946205571565802,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004141311717835625,
      "loss": 4.6632,
      "mean_token_accuracy": 0.25090006440877916,
      "num_tokens": 70335501.0,
      "step": 30670
    },
    {
      "entropy": 5.327083539962769,
      "epoch": 2.946685878962536,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041410422691520114,
      "loss": 4.6575,
      "mean_token_accuracy": 0.25249666571617124,
      "num_tokens": 70346848.0,
      "step": 30675
    },
    {
      "entropy": 5.397417259216309,
      "epoch": 2.94716618635927,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041407727881725265,
      "loss": 4.7623,
      "mean_token_accuracy": 0.24369217306375504,
      "num_tokens": 70356707.0,
      "step": 30680
    },
    {
      "entropy": 5.37922043800354,
      "epoch": 2.947646493756004,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004140503274903426,
      "loss": 4.7996,
      "mean_token_accuracy": 0.24128311276435851,
      "num_tokens": 70368115.0,
      "step": 30685
    },
    {
      "entropy": 5.348415374755859,
      "epoch": 2.9481268011527377,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000414023372935097,
      "loss": 4.68,
      "mean_token_accuracy": 0.25451764911413194,
      "num_tokens": 70379883.0,
      "step": 30690
    },
    {
      "entropy": 5.370724534988403,
      "epoch": 2.9486071085494716,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041399641515214137,
      "loss": 4.7489,
      "mean_token_accuracy": 0.24097025245428086,
      "num_tokens": 70390833.0,
      "step": 30695
    },
    {
      "entropy": 5.322110843658447,
      "epoch": 2.9490874159462055,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004139694541421018,
      "loss": 4.7651,
      "mean_token_accuracy": 0.24194575250148773,
      "num_tokens": 70402679.0,
      "step": 30700
    },
    {
      "entropy": 5.392458105087281,
      "epoch": 2.9495677233429394,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004139424899056042,
      "loss": 4.7859,
      "mean_token_accuracy": 0.23799656927585602,
      "num_tokens": 70415623.0,
      "step": 30705
    },
    {
      "entropy": 5.4351013660430905,
      "epoch": 2.9500480307396733,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041391552244327446,
      "loss": 4.8111,
      "mean_token_accuracy": 0.24727839529514312,
      "num_tokens": 70427192.0,
      "step": 30710
    },
    {
      "entropy": 5.33963942527771,
      "epoch": 2.950528338136407,
      "grad_norm": 1.0,
      "learning_rate": 0.0004138885517557387,
      "loss": 4.7428,
      "mean_token_accuracy": 0.2469615936279297,
      "num_tokens": 70439002.0,
      "step": 30715
    },
    {
      "entropy": 5.34798493385315,
      "epoch": 2.9510086455331415,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004138615778436234,
      "loss": 4.6803,
      "mean_token_accuracy": 0.25110501796007156,
      "num_tokens": 70449790.0,
      "step": 30720
    },
    {
      "entropy": 5.3636195182800295,
      "epoch": 2.951488952929875,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041383460070755447,
      "loss": 4.7876,
      "mean_token_accuracy": 0.23905376642942427,
      "num_tokens": 70460687.0,
      "step": 30725
    },
    {
      "entropy": 5.358436727523804,
      "epoch": 2.9519692603266092,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041380762034815834,
      "loss": 4.6433,
      "mean_token_accuracy": 0.2475594013929367,
      "num_tokens": 70470851.0,
      "step": 30730
    },
    {
      "entropy": 5.35434045791626,
      "epoch": 2.952449567723343,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041378063676606147,
      "loss": 4.7077,
      "mean_token_accuracy": 0.2502685561776161,
      "num_tokens": 70481979.0,
      "step": 30735
    },
    {
      "entropy": 5.31798152923584,
      "epoch": 2.952929875120077,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041375364996189035,
      "loss": 4.6626,
      "mean_token_accuracy": 0.2497686982154846,
      "num_tokens": 70493107.0,
      "step": 30740
    },
    {
      "entropy": 5.258630800247192,
      "epoch": 2.953410182516811,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041372665993627143,
      "loss": 4.6667,
      "mean_token_accuracy": 0.24592654556035995,
      "num_tokens": 70504385.0,
      "step": 30745
    },
    {
      "entropy": 5.290937852859497,
      "epoch": 2.9538904899135447,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041369966668983144,
      "loss": 4.6783,
      "mean_token_accuracy": 0.25059950798749925,
      "num_tokens": 70515906.0,
      "step": 30750
    },
    {
      "entropy": 5.422720670700073,
      "epoch": 2.9543707973102786,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041367267022319706,
      "loss": 4.7971,
      "mean_token_accuracy": 0.24043042957782745,
      "num_tokens": 70526927.0,
      "step": 30755
    },
    {
      "entropy": 5.4264391422271725,
      "epoch": 2.9548511047070125,
      "grad_norm": 1.078125,
      "learning_rate": 0.000413645670536995,
      "loss": 4.779,
      "mean_token_accuracy": 0.24170510917901994,
      "num_tokens": 70537521.0,
      "step": 30760
    },
    {
      "entropy": 5.4083176612854,
      "epoch": 2.9553314121037464,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004136186676318522,
      "loss": 4.7654,
      "mean_token_accuracy": 0.24449757635593414,
      "num_tokens": 70547851.0,
      "step": 30765
    },
    {
      "entropy": 5.306046772003174,
      "epoch": 2.9558117195004803,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004135916615083956,
      "loss": 4.6657,
      "mean_token_accuracy": 0.252124160528183,
      "num_tokens": 70559725.0,
      "step": 30770
    },
    {
      "entropy": 5.320335006713867,
      "epoch": 2.956292026897214,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041356465216725195,
      "loss": 4.697,
      "mean_token_accuracy": 0.2517889738082886,
      "num_tokens": 70571410.0,
      "step": 30775
    },
    {
      "entropy": 5.361250305175782,
      "epoch": 2.956772334293948,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041353763960904873,
      "loss": 4.7435,
      "mean_token_accuracy": 0.24528248757123947,
      "num_tokens": 70582665.0,
      "step": 30780
    },
    {
      "entropy": 5.340354299545288,
      "epoch": 2.957252641690682,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041351062383441286,
      "loss": 4.7226,
      "mean_token_accuracy": 0.2537853643298149,
      "num_tokens": 70593653.0,
      "step": 30785
    },
    {
      "entropy": 5.39779486656189,
      "epoch": 2.957732949087416,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004134836048439716,
      "loss": 4.7967,
      "mean_token_accuracy": 0.24253580421209336,
      "num_tokens": 70603706.0,
      "step": 30790
    },
    {
      "entropy": 5.331996774673462,
      "epoch": 2.95821325648415,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041345658263835215,
      "loss": 4.691,
      "mean_token_accuracy": 0.2510316506028175,
      "num_tokens": 70614940.0,
      "step": 30795
    },
    {
      "entropy": 5.4380566596984865,
      "epoch": 2.9586935638808836,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041342955721818207,
      "loss": 4.8024,
      "mean_token_accuracy": 0.24719777554273606,
      "num_tokens": 70626503.0,
      "step": 30800
    },
    {
      "entropy": 5.309935617446899,
      "epoch": 2.959173871277618,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041340252858408866,
      "loss": 4.6944,
      "mean_token_accuracy": 0.2572073459625244,
      "num_tokens": 70638483.0,
      "step": 30805
    },
    {
      "entropy": 5.318456840515137,
      "epoch": 2.9596541786743513,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041337549673669963,
      "loss": 4.7019,
      "mean_token_accuracy": 0.24817472398281099,
      "num_tokens": 70650609.0,
      "step": 30810
    },
    {
      "entropy": 5.330030536651611,
      "epoch": 2.9601344860710856,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004133484616766423,
      "loss": 4.7167,
      "mean_token_accuracy": 0.24880424588918687,
      "num_tokens": 70661368.0,
      "step": 30815
    },
    {
      "entropy": 5.4279358863830565,
      "epoch": 2.9606147934678195,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041332142340454463,
      "loss": 4.725,
      "mean_token_accuracy": 0.24652160853147506,
      "num_tokens": 70673324.0,
      "step": 30820
    },
    {
      "entropy": 5.32217116355896,
      "epoch": 2.9610951008645534,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004132943819210342,
      "loss": 4.7048,
      "mean_token_accuracy": 0.25561501681804655,
      "num_tokens": 70684639.0,
      "step": 30825
    },
    {
      "entropy": 5.268362379074096,
      "epoch": 2.9615754082612873,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041326733722673876,
      "loss": 4.6144,
      "mean_token_accuracy": 0.25555491000413894,
      "num_tokens": 70695742.0,
      "step": 30830
    },
    {
      "entropy": 5.381715154647827,
      "epoch": 2.962055715658021,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041324028932228645,
      "loss": 4.7302,
      "mean_token_accuracy": 0.24521666914224624,
      "num_tokens": 70706711.0,
      "step": 30835
    },
    {
      "entropy": 5.337053298950195,
      "epoch": 2.962536023054755,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004132132382083051,
      "loss": 4.7066,
      "mean_token_accuracy": 0.25174325704574585,
      "num_tokens": 70719090.0,
      "step": 30840
    },
    {
      "entropy": 5.316998815536499,
      "epoch": 2.963016330451489,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041318618388542274,
      "loss": 4.7491,
      "mean_token_accuracy": 0.2484269067645073,
      "num_tokens": 70730356.0,
      "step": 30845
    },
    {
      "entropy": 5.456576347351074,
      "epoch": 2.963496637848223,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004131591263542675,
      "loss": 4.8231,
      "mean_token_accuracy": 0.24189328253269196,
      "num_tokens": 70744864.0,
      "step": 30850
    },
    {
      "entropy": 5.46539077758789,
      "epoch": 2.9639769452449567,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004131320656154676,
      "loss": 4.8251,
      "mean_token_accuracy": 0.23867482841014862,
      "num_tokens": 70756882.0,
      "step": 30855
    },
    {
      "entropy": 5.3447315216064455,
      "epoch": 2.9644572526416906,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004131050016696514,
      "loss": 4.7056,
      "mean_token_accuracy": 0.24652895778417588,
      "num_tokens": 70768501.0,
      "step": 30860
    },
    {
      "entropy": 5.311851978302002,
      "epoch": 2.9649375600384245,
      "grad_norm": 1.203125,
      "learning_rate": 0.000413077934517447,
      "loss": 4.701,
      "mean_token_accuracy": 0.2509620890021324,
      "num_tokens": 70779545.0,
      "step": 30865
    },
    {
      "entropy": 5.306081485748291,
      "epoch": 2.9654178674351583,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004130508641594831,
      "loss": 4.7246,
      "mean_token_accuracy": 0.2493472069501877,
      "num_tokens": 70791423.0,
      "step": 30870
    },
    {
      "entropy": 5.303132963180542,
      "epoch": 2.965898174831892,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041302379059638794,
      "loss": 4.7322,
      "mean_token_accuracy": 0.2463749721646309,
      "num_tokens": 70802938.0,
      "step": 30875
    },
    {
      "entropy": 5.393779039382935,
      "epoch": 2.9663784822286265,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00041299671382879024,
      "loss": 4.7624,
      "mean_token_accuracy": 0.24239962697029113,
      "num_tokens": 70813285.0,
      "step": 30880
    },
    {
      "entropy": 5.31902756690979,
      "epoch": 2.96685878962536,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004129696338573187,
      "loss": 4.7106,
      "mean_token_accuracy": 0.24705754816532136,
      "num_tokens": 70826509.0,
      "step": 30885
    },
    {
      "entropy": 5.286365985870361,
      "epoch": 2.9673390970220943,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004129425506826019,
      "loss": 4.7177,
      "mean_token_accuracy": 0.24925039410591127,
      "num_tokens": 70837648.0,
      "step": 30890
    },
    {
      "entropy": 5.285222339630127,
      "epoch": 2.967819404418828,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041291546430526863,
      "loss": 4.6979,
      "mean_token_accuracy": 0.25606866776943205,
      "num_tokens": 70849702.0,
      "step": 30895
    },
    {
      "entropy": 5.374529552459717,
      "epoch": 2.968299711815562,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004128883747259478,
      "loss": 4.7679,
      "mean_token_accuracy": 0.24928651452064515,
      "num_tokens": 70861445.0,
      "step": 30900
    },
    {
      "entropy": 5.393187522888184,
      "epoch": 2.968780019212296,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004128612819452684,
      "loss": 4.7109,
      "mean_token_accuracy": 0.2516806721687317,
      "num_tokens": 70872448.0,
      "step": 30905
    },
    {
      "entropy": 5.3442747592926025,
      "epoch": 2.96926032660903,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041283418596385944,
      "loss": 4.6368,
      "mean_token_accuracy": 0.2516734555363655,
      "num_tokens": 70883777.0,
      "step": 30910
    },
    {
      "entropy": 5.310675048828125,
      "epoch": 2.9697406340057637,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004128070867823499,
      "loss": 4.7062,
      "mean_token_accuracy": 0.24575408548116684,
      "num_tokens": 70895084.0,
      "step": 30915
    },
    {
      "entropy": 5.310546398162842,
      "epoch": 2.9702209414024976,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000412779984401369,
      "loss": 4.7353,
      "mean_token_accuracy": 0.24674065560102462,
      "num_tokens": 70906304.0,
      "step": 30920
    },
    {
      "entropy": 5.3670261859893795,
      "epoch": 2.9707012487992315,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000412752878821546,
      "loss": 4.6705,
      "mean_token_accuracy": 0.25208906531333924,
      "num_tokens": 70917448.0,
      "step": 30925
    },
    {
      "entropy": 5.310788536071778,
      "epoch": 2.9711815561959654,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041272577004351026,
      "loss": 4.6707,
      "mean_token_accuracy": 0.2502093493938446,
      "num_tokens": 70927991.0,
      "step": 30930
    },
    {
      "entropy": 5.314220762252807,
      "epoch": 2.9716618635926992,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041269865806789095,
      "loss": 4.6762,
      "mean_token_accuracy": 0.24905722737312316,
      "num_tokens": 70939490.0,
      "step": 30935
    },
    {
      "entropy": 5.356766557693481,
      "epoch": 2.972142170989433,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004126715428953178,
      "loss": 4.6979,
      "mean_token_accuracy": 0.24512134194374086,
      "num_tokens": 70951611.0,
      "step": 30940
    },
    {
      "entropy": 5.381950092315674,
      "epoch": 2.972622478386167,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004126444245264202,
      "loss": 4.7384,
      "mean_token_accuracy": 0.2438073515892029,
      "num_tokens": 70962003.0,
      "step": 30945
    },
    {
      "entropy": 5.248794507980347,
      "epoch": 2.973102785782901,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004126173029618278,
      "loss": 4.5628,
      "mean_token_accuracy": 0.2599476531147957,
      "num_tokens": 70972022.0,
      "step": 30950
    },
    {
      "entropy": 5.327218675613404,
      "epoch": 2.973583093179635,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004125901782021702,
      "loss": 4.7549,
      "mean_token_accuracy": 0.24723852872848512,
      "num_tokens": 70984086.0,
      "step": 30955
    },
    {
      "entropy": 5.4143805503845215,
      "epoch": 2.9740634005763686,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004125630502480773,
      "loss": 4.816,
      "mean_token_accuracy": 0.23507488816976546,
      "num_tokens": 70995705.0,
      "step": 30960
    },
    {
      "entropy": 5.432412767410279,
      "epoch": 2.974543707973103,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004125359191001788,
      "loss": 4.7642,
      "mean_token_accuracy": 0.24783959984779358,
      "num_tokens": 71007812.0,
      "step": 30965
    },
    {
      "entropy": 5.35087251663208,
      "epoch": 2.975024015369837,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004125087847591047,
      "loss": 4.6988,
      "mean_token_accuracy": 0.2503730431199074,
      "num_tokens": 71018950.0,
      "step": 30970
    },
    {
      "entropy": 5.3044616222381595,
      "epoch": 2.9755043227665707,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041248164722548493,
      "loss": 4.6885,
      "mean_token_accuracy": 0.2502960816025734,
      "num_tokens": 71030391.0,
      "step": 30975
    },
    {
      "entropy": 5.4009318351745605,
      "epoch": 2.9759846301633046,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004124545064999495,
      "loss": 4.7264,
      "mean_token_accuracy": 0.24643306136131288,
      "num_tokens": 71041741.0,
      "step": 30980
    },
    {
      "entropy": 5.350307512283325,
      "epoch": 2.9764649375600385,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041242736258312866,
      "loss": 4.7132,
      "mean_token_accuracy": 0.253027506172657,
      "num_tokens": 71052062.0,
      "step": 30985
    },
    {
      "entropy": 5.315801382064819,
      "epoch": 2.9769452449567724,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004124002154756525,
      "loss": 4.7145,
      "mean_token_accuracy": 0.24751503765583038,
      "num_tokens": 71063608.0,
      "step": 30990
    },
    {
      "entropy": 5.392868137359619,
      "epoch": 2.9774255523535063,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041237306517815124,
      "loss": 4.7659,
      "mean_token_accuracy": 0.24717787355184556,
      "num_tokens": 71073526.0,
      "step": 30995
    },
    {
      "entropy": 5.278286933898926,
      "epoch": 2.97790585975024,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004123459116912554,
      "loss": 4.6221,
      "mean_token_accuracy": 0.25272875726222993,
      "num_tokens": 71085311.0,
      "step": 31000
    },
    {
      "entropy": 5.372762680053711,
      "epoch": 2.978386167146974,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041231875501559535,
      "loss": 4.7514,
      "mean_token_accuracy": 0.24657849669456483,
      "num_tokens": 71096438.0,
      "step": 31005
    },
    {
      "entropy": 5.334469604492187,
      "epoch": 2.978866474543708,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041229159515180155,
      "loss": 4.6825,
      "mean_token_accuracy": 0.24589990675449372,
      "num_tokens": 71108262.0,
      "step": 31010
    },
    {
      "entropy": 5.279165983200073,
      "epoch": 2.979346781940442,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004122644321005046,
      "loss": 4.7204,
      "mean_token_accuracy": 0.24643756449222565,
      "num_tokens": 71120905.0,
      "step": 31015
    },
    {
      "entropy": 5.36572642326355,
      "epoch": 2.9798270893371757,
      "grad_norm": 1.125,
      "learning_rate": 0.00041223726586233505,
      "loss": 4.7336,
      "mean_token_accuracy": 0.2539439097046852,
      "num_tokens": 71132522.0,
      "step": 31020
    },
    {
      "entropy": 5.341736364364624,
      "epoch": 2.9803073967339095,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041221009643792377,
      "loss": 4.6891,
      "mean_token_accuracy": 0.2569159761071205,
      "num_tokens": 71143406.0,
      "step": 31025
    },
    {
      "entropy": 5.37914662361145,
      "epoch": 2.980787704130644,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004121829238279014,
      "loss": 4.7735,
      "mean_token_accuracy": 0.2398884892463684,
      "num_tokens": 71153525.0,
      "step": 31030
    },
    {
      "entropy": 5.375209331512451,
      "epoch": 2.9812680115273773,
      "grad_norm": 1.125,
      "learning_rate": 0.00041215574803289896,
      "loss": 4.724,
      "mean_token_accuracy": 0.23985127061605455,
      "num_tokens": 71165672.0,
      "step": 31035
    },
    {
      "entropy": 5.339705467224121,
      "epoch": 2.9817483189241116,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004121285690535473,
      "loss": 4.6716,
      "mean_token_accuracy": 0.25114576518535614,
      "num_tokens": 71176411.0,
      "step": 31040
    },
    {
      "entropy": 5.342957448959351,
      "epoch": 2.982228626320845,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00041210138689047745,
      "loss": 4.714,
      "mean_token_accuracy": 0.24052476286888122,
      "num_tokens": 71189520.0,
      "step": 31045
    },
    {
      "entropy": 5.365981721878052,
      "epoch": 2.9827089337175794,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004120742015443206,
      "loss": 4.7089,
      "mean_token_accuracy": 0.2511673867702484,
      "num_tokens": 71200199.0,
      "step": 31050
    },
    {
      "entropy": 5.410485887527466,
      "epoch": 2.9831892411143133,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004120470130157077,
      "loss": 4.7645,
      "mean_token_accuracy": 0.23518042862415314,
      "num_tokens": 71211909.0,
      "step": 31055
    },
    {
      "entropy": 5.309051942825318,
      "epoch": 2.983669548511047,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041201982130527006,
      "loss": 4.6708,
      "mean_token_accuracy": 0.25204419195652006,
      "num_tokens": 71222202.0,
      "step": 31060
    },
    {
      "entropy": 5.414158725738526,
      "epoch": 2.984149855907781,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041199262641363914,
      "loss": 4.8377,
      "mean_token_accuracy": 0.23482049703598024,
      "num_tokens": 71233681.0,
      "step": 31065
    },
    {
      "entropy": 5.382309627532959,
      "epoch": 2.984630163304515,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041196542834144617,
      "loss": 4.7321,
      "mean_token_accuracy": 0.25697826892137526,
      "num_tokens": 71244566.0,
      "step": 31070
    },
    {
      "entropy": 5.395444059371949,
      "epoch": 2.985110470701249,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041193822708932265,
      "loss": 4.73,
      "mean_token_accuracy": 0.2465959832072258,
      "num_tokens": 71254892.0,
      "step": 31075
    },
    {
      "entropy": 5.303956985473633,
      "epoch": 2.9855907780979827,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004119110226579002,
      "loss": 4.6738,
      "mean_token_accuracy": 0.2536274537444115,
      "num_tokens": 71266952.0,
      "step": 31080
    },
    {
      "entropy": 5.340962076187134,
      "epoch": 2.9860710854947166,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041188381504781026,
      "loss": 4.6645,
      "mean_token_accuracy": 0.249992835521698,
      "num_tokens": 71278039.0,
      "step": 31085
    },
    {
      "entropy": 5.390396356582642,
      "epoch": 2.9865513928914504,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004118566042596846,
      "loss": 4.74,
      "mean_token_accuracy": 0.24509156197309495,
      "num_tokens": 71289201.0,
      "step": 31090
    },
    {
      "entropy": 5.323314571380616,
      "epoch": 2.9870317002881843,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000411829390294155,
      "loss": 4.6624,
      "mean_token_accuracy": 0.25661278516054153,
      "num_tokens": 71301360.0,
      "step": 31095
    },
    {
      "entropy": 5.40875735282898,
      "epoch": 2.987512007684918,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041180217315185333,
      "loss": 4.7385,
      "mean_token_accuracy": 0.24627887457609177,
      "num_tokens": 71311846.0,
      "step": 31100
    },
    {
      "entropy": 5.363363790512085,
      "epoch": 2.9879923150816525,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041177495283341124,
      "loss": 4.7928,
      "mean_token_accuracy": 0.24592275619506837,
      "num_tokens": 71324420.0,
      "step": 31105
    },
    {
      "entropy": 5.265457963943481,
      "epoch": 2.988472622478386,
      "grad_norm": 1.25,
      "learning_rate": 0.000411747729339461,
      "loss": 4.5854,
      "mean_token_accuracy": 0.2594572171568871,
      "num_tokens": 71335641.0,
      "step": 31110
    },
    {
      "entropy": 5.2663520812988285,
      "epoch": 2.9889529298751203,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004117205026706345,
      "loss": 4.6827,
      "mean_token_accuracy": 0.25581717789173125,
      "num_tokens": 71346360.0,
      "step": 31115
    },
    {
      "entropy": 5.296803045272827,
      "epoch": 2.9894332372718537,
      "grad_norm": 1.125,
      "learning_rate": 0.00041169327282756396,
      "loss": 4.6468,
      "mean_token_accuracy": 0.2605433687567711,
      "num_tokens": 71357363.0,
      "step": 31120
    },
    {
      "entropy": 5.29811806678772,
      "epoch": 2.989913544668588,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004116660398108815,
      "loss": 4.6119,
      "mean_token_accuracy": 0.25107635259628297,
      "num_tokens": 71368731.0,
      "step": 31125
    },
    {
      "entropy": 5.310903263092041,
      "epoch": 2.990393852065322,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004116388036212194,
      "loss": 4.6944,
      "mean_token_accuracy": 0.2436070591211319,
      "num_tokens": 71379891.0,
      "step": 31130
    },
    {
      "entropy": 5.326344776153564,
      "epoch": 2.990874159462056,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041161156425921004,
      "loss": 4.6222,
      "mean_token_accuracy": 0.2534897029399872,
      "num_tokens": 71391353.0,
      "step": 31135
    },
    {
      "entropy": 5.306528902053833,
      "epoch": 2.9913544668587897,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041158432172548577,
      "loss": 4.6561,
      "mean_token_accuracy": 0.2545619636774063,
      "num_tokens": 71401968.0,
      "step": 31140
    },
    {
      "entropy": 5.27363133430481,
      "epoch": 2.9918347742555236,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041155707602067923,
      "loss": 4.7033,
      "mean_token_accuracy": 0.25090731829404833,
      "num_tokens": 71412056.0,
      "step": 31145
    },
    {
      "entropy": 5.3062104225158695,
      "epoch": 2.9923150816522575,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004115298271454227,
      "loss": 4.6859,
      "mean_token_accuracy": 0.252056360244751,
      "num_tokens": 71422182.0,
      "step": 31150
    },
    {
      "entropy": 5.37526421546936,
      "epoch": 2.9927953890489913,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004115025751003491,
      "loss": 4.7595,
      "mean_token_accuracy": 0.24392429292201995,
      "num_tokens": 71433796.0,
      "step": 31155
    },
    {
      "entropy": 5.361346292495727,
      "epoch": 2.993275696445725,
      "grad_norm": 1.078125,
      "learning_rate": 0.000411475319886091,
      "loss": 4.7252,
      "mean_token_accuracy": 0.2516699284315109,
      "num_tokens": 71445756.0,
      "step": 31160
    },
    {
      "entropy": 5.411783266067505,
      "epoch": 2.993756003842459,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041144806150328117,
      "loss": 4.7764,
      "mean_token_accuracy": 0.24522655606269836,
      "num_tokens": 71457124.0,
      "step": 31165
    },
    {
      "entropy": 5.287089443206787,
      "epoch": 2.994236311239193,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004114207999525527,
      "loss": 4.6683,
      "mean_token_accuracy": 0.2573847144842148,
      "num_tokens": 71469287.0,
      "step": 31170
    },
    {
      "entropy": 5.283834552764892,
      "epoch": 2.994716618635927,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041139353523453814,
      "loss": 4.6483,
      "mean_token_accuracy": 0.2487101674079895,
      "num_tokens": 71480403.0,
      "step": 31175
    },
    {
      "entropy": 5.386361122131348,
      "epoch": 2.9951969260326607,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004113662673498708,
      "loss": 4.7394,
      "mean_token_accuracy": 0.2540075644850731,
      "num_tokens": 71492431.0,
      "step": 31180
    },
    {
      "entropy": 5.32862811088562,
      "epoch": 2.9956772334293946,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041133899629918364,
      "loss": 4.6507,
      "mean_token_accuracy": 0.2537261128425598,
      "num_tokens": 71503588.0,
      "step": 31185
    },
    {
      "entropy": 5.259595823287964,
      "epoch": 2.996157540826129,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041131172208310986,
      "loss": 4.6133,
      "mean_token_accuracy": 0.2560037329792976,
      "num_tokens": 71514813.0,
      "step": 31190
    },
    {
      "entropy": 5.351832485198974,
      "epoch": 2.9966378482228624,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041128444470228253,
      "loss": 4.701,
      "mean_token_accuracy": 0.24409834444522857,
      "num_tokens": 71526961.0,
      "step": 31195
    },
    {
      "entropy": 5.312208223342895,
      "epoch": 2.9971181556195967,
      "grad_norm": 1.125,
      "learning_rate": 0.00041125716415733524,
      "loss": 4.6396,
      "mean_token_accuracy": 0.2595481649041176,
      "num_tokens": 71537793.0,
      "step": 31200
    },
    {
      "entropy": 5.296971035003662,
      "epoch": 2.9975984630163306,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004112298804489011,
      "loss": 4.7071,
      "mean_token_accuracy": 0.24418871700763703,
      "num_tokens": 71549959.0,
      "step": 31205
    },
    {
      "entropy": 5.3914408683776855,
      "epoch": 2.9980787704130645,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004112025935776137,
      "loss": 4.8181,
      "mean_token_accuracy": 0.23681433498859406,
      "num_tokens": 71560340.0,
      "step": 31210
    },
    {
      "entropy": 5.404863739013672,
      "epoch": 2.9985590778097984,
      "grad_norm": 1.125,
      "learning_rate": 0.00041117530354410647,
      "loss": 4.7625,
      "mean_token_accuracy": 0.23983166515827178,
      "num_tokens": 71571788.0,
      "step": 31215
    },
    {
      "entropy": 5.389599084854126,
      "epoch": 2.9990393852065322,
      "grad_norm": 1.25,
      "learning_rate": 0.0004111480103490131,
      "loss": 4.7373,
      "mean_token_accuracy": 0.25116525739431383,
      "num_tokens": 71582660.0,
      "step": 31220
    },
    {
      "entropy": 5.2823316097259525,
      "epoch": 2.999519692603266,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041112071399296724,
      "loss": 4.7316,
      "mean_token_accuracy": 0.24800180792808532,
      "num_tokens": 71594672.0,
      "step": 31225
    },
    {
      "entropy": 5.329966306686401,
      "epoch": 3.0,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004110934144766025,
      "loss": 4.7436,
      "mean_token_accuracy": 0.2465621277689934,
      "num_tokens": 71605608.0,
      "step": 31230
    },
    {
      "entropy": 5.367530632019043,
      "epoch": 3.000480307396734,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041106611180055284,
      "loss": 4.661,
      "mean_token_accuracy": 0.2525114223361015,
      "num_tokens": 71616917.0,
      "step": 31235
    },
    {
      "entropy": 5.360671091079712,
      "epoch": 3.0009606147934678,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041103880596545206,
      "loss": 4.6864,
      "mean_token_accuracy": 0.2543638214468956,
      "num_tokens": 71630190.0,
      "step": 31240
    },
    {
      "entropy": 5.330166292190552,
      "epoch": 3.0014409221902016,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004110114969719342,
      "loss": 4.6819,
      "mean_token_accuracy": 0.25917249619960786,
      "num_tokens": 71642305.0,
      "step": 31245
    },
    {
      "entropy": 5.331141471862793,
      "epoch": 3.0019212295869355,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004109841848206333,
      "loss": 4.6947,
      "mean_token_accuracy": 0.25061159729957583,
      "num_tokens": 71652735.0,
      "step": 31250
    },
    {
      "entropy": 5.295147275924682,
      "epoch": 3.0024015369836694,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004109568695121833,
      "loss": 4.5741,
      "mean_token_accuracy": 0.2586781531572342,
      "num_tokens": 71664051.0,
      "step": 31255
    },
    {
      "entropy": 5.353749179840088,
      "epoch": 3.0028818443804033,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004109295510472186,
      "loss": 4.686,
      "mean_token_accuracy": 0.2504920676350594,
      "num_tokens": 71675785.0,
      "step": 31260
    },
    {
      "entropy": 5.313034915924073,
      "epoch": 3.003362151777137,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041090222942637323,
      "loss": 4.6112,
      "mean_token_accuracy": 0.25818091481924055,
      "num_tokens": 71687322.0,
      "step": 31265
    },
    {
      "entropy": 5.373900890350342,
      "epoch": 3.0038424591738715,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041087490465028175,
      "loss": 4.6889,
      "mean_token_accuracy": 0.2520124077796936,
      "num_tokens": 71698269.0,
      "step": 31270
    },
    {
      "entropy": 5.390183830261231,
      "epoch": 3.0043227665706054,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041084757671957844,
      "loss": 4.7372,
      "mean_token_accuracy": 0.24695932418107985,
      "num_tokens": 71710078.0,
      "step": 31275
    },
    {
      "entropy": 5.403388357162475,
      "epoch": 3.0048030739673393,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041082024563489773,
      "loss": 4.7674,
      "mean_token_accuracy": 0.2503056228160858,
      "num_tokens": 71721882.0,
      "step": 31280
    },
    {
      "entropy": 5.373847389221192,
      "epoch": 3.005283381364073,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004107929113968743,
      "loss": 4.7056,
      "mean_token_accuracy": 0.2505654588341713,
      "num_tokens": 71733710.0,
      "step": 31285
    },
    {
      "entropy": 5.326976299285889,
      "epoch": 3.005763688760807,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004107655740061427,
      "loss": 4.6265,
      "mean_token_accuracy": 0.25305320620536803,
      "num_tokens": 71745294.0,
      "step": 31290
    },
    {
      "entropy": 5.34187126159668,
      "epoch": 3.006243996157541,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004107382334633375,
      "loss": 4.6617,
      "mean_token_accuracy": 0.25856254547834395,
      "num_tokens": 71757184.0,
      "step": 31295
    },
    {
      "entropy": 5.361162805557251,
      "epoch": 3.006724303554275,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004107108897690936,
      "loss": 4.7117,
      "mean_token_accuracy": 0.24659442752599717,
      "num_tokens": 71768618.0,
      "step": 31300
    },
    {
      "entropy": 5.414818620681762,
      "epoch": 3.0072046109510087,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004106835429240458,
      "loss": 4.7877,
      "mean_token_accuracy": 0.2527236670255661,
      "num_tokens": 71779741.0,
      "step": 31305
    },
    {
      "entropy": 5.316113376617432,
      "epoch": 3.0076849183477425,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004106561929288292,
      "loss": 4.5933,
      "mean_token_accuracy": 0.2528831347823143,
      "num_tokens": 71791085.0,
      "step": 31310
    },
    {
      "entropy": 5.300862836837768,
      "epoch": 3.0081652257444764,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041062883978407844,
      "loss": 4.6672,
      "mean_token_accuracy": 0.2559434473514557,
      "num_tokens": 71804061.0,
      "step": 31315
    },
    {
      "entropy": 5.418096256256104,
      "epoch": 3.0086455331412103,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041060148349042876,
      "loss": 4.6985,
      "mean_token_accuracy": 0.2488008663058281,
      "num_tokens": 71815303.0,
      "step": 31320
    },
    {
      "entropy": 5.397549486160278,
      "epoch": 3.009125840537944,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041057412404851536,
      "loss": 4.7714,
      "mean_token_accuracy": 0.24207072108983993,
      "num_tokens": 71826492.0,
      "step": 31325
    },
    {
      "entropy": 5.364451503753662,
      "epoch": 3.009606147934678,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004105467614589734,
      "loss": 4.6779,
      "mean_token_accuracy": 0.2524795442819595,
      "num_tokens": 71837245.0,
      "step": 31330
    },
    {
      "entropy": 5.383288955688476,
      "epoch": 3.010086455331412,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004105193957224381,
      "loss": 4.6785,
      "mean_token_accuracy": 0.24422649741172792,
      "num_tokens": 71848684.0,
      "step": 31335
    },
    {
      "entropy": 5.373349714279175,
      "epoch": 3.010566762728146,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041049202683954473,
      "loss": 4.6677,
      "mean_token_accuracy": 0.2603182464838028,
      "num_tokens": 71861472.0,
      "step": 31340
    },
    {
      "entropy": 5.32449746131897,
      "epoch": 3.0110470701248797,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041046465481092893,
      "loss": 4.6305,
      "mean_token_accuracy": 0.251866851747036,
      "num_tokens": 71873076.0,
      "step": 31345
    },
    {
      "entropy": 5.308641529083252,
      "epoch": 3.011527377521614,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041043727963722607,
      "loss": 4.6815,
      "mean_token_accuracy": 0.2486639440059662,
      "num_tokens": 71884416.0,
      "step": 31350
    },
    {
      "entropy": 5.3582898616790775,
      "epoch": 3.012007684918348,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004104099013190718,
      "loss": 4.6713,
      "mean_token_accuracy": 0.2522772416472435,
      "num_tokens": 71894113.0,
      "step": 31355
    },
    {
      "entropy": 5.391060829162598,
      "epoch": 3.012487992315082,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041038251985710164,
      "loss": 4.6774,
      "mean_token_accuracy": 0.2526190236210823,
      "num_tokens": 71904732.0,
      "step": 31360
    },
    {
      "entropy": 5.35156021118164,
      "epoch": 3.0129682997118157,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004103551352519514,
      "loss": 4.6613,
      "mean_token_accuracy": 0.25972236692905426,
      "num_tokens": 71915112.0,
      "step": 31365
    },
    {
      "entropy": 5.289450693130493,
      "epoch": 3.0134486071085496,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041032774750425683,
      "loss": 4.6307,
      "mean_token_accuracy": 0.25402156710624696,
      "num_tokens": 71927175.0,
      "step": 31370
    },
    {
      "entropy": 5.4385617733001705,
      "epoch": 3.0139289145052834,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004103003566146538,
      "loss": 4.7587,
      "mean_token_accuracy": 0.24087945520877838,
      "num_tokens": 71937596.0,
      "step": 31375
    },
    {
      "entropy": 5.280434894561767,
      "epoch": 3.0144092219020173,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004102729625837782,
      "loss": 4.5951,
      "mean_token_accuracy": 0.2630214512348175,
      "num_tokens": 71948594.0,
      "step": 31380
    },
    {
      "entropy": 5.393678140640259,
      "epoch": 3.014889529298751,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004102455654122662,
      "loss": 4.7314,
      "mean_token_accuracy": 0.25142778903245927,
      "num_tokens": 71961171.0,
      "step": 31385
    },
    {
      "entropy": 5.3706769943237305,
      "epoch": 3.015369836695485,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041021816510075366,
      "loss": 4.6286,
      "mean_token_accuracy": 0.2573365792632103,
      "num_tokens": 71972697.0,
      "step": 31390
    },
    {
      "entropy": 5.322516345977784,
      "epoch": 3.015850144092219,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041019076164987696,
      "loss": 4.6902,
      "mean_token_accuracy": 0.24632543623447417,
      "num_tokens": 71985431.0,
      "step": 31395
    },
    {
      "entropy": 5.259512090682984,
      "epoch": 3.016330451488953,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004101633550602721,
      "loss": 4.5746,
      "mean_token_accuracy": 0.2606945216655731,
      "num_tokens": 71996935.0,
      "step": 31400
    },
    {
      "entropy": 5.318569135665894,
      "epoch": 3.0168107588856867,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004101359453325755,
      "loss": 4.6739,
      "mean_token_accuracy": 0.24722499400377274,
      "num_tokens": 72008227.0,
      "step": 31405
    },
    {
      "entropy": 5.466337251663208,
      "epoch": 3.0172910662824206,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041010853246742357,
      "loss": 4.7325,
      "mean_token_accuracy": 0.24442485719919205,
      "num_tokens": 72020478.0,
      "step": 31410
    },
    {
      "entropy": 5.4444272994995115,
      "epoch": 3.0177713736791545,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004100811164654527,
      "loss": 4.7012,
      "mean_token_accuracy": 0.2466185510158539,
      "num_tokens": 72031542.0,
      "step": 31415
    },
    {
      "entropy": 5.402291011810303,
      "epoch": 3.0182516810758884,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004100536973272994,
      "loss": 4.73,
      "mean_token_accuracy": 0.24907872080802917,
      "num_tokens": 72043956.0,
      "step": 31420
    },
    {
      "entropy": 5.364996337890625,
      "epoch": 3.0187319884726227,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004100262750536003,
      "loss": 4.6708,
      "mean_token_accuracy": 0.25904034078121185,
      "num_tokens": 72055586.0,
      "step": 31425
    },
    {
      "entropy": 5.275601482391357,
      "epoch": 3.0192122958693566,
      "grad_norm": 1.125,
      "learning_rate": 0.00040999884964499196,
      "loss": 4.69,
      "mean_token_accuracy": 0.25312763303518293,
      "num_tokens": 72067889.0,
      "step": 31430
    },
    {
      "entropy": 5.309688472747803,
      "epoch": 3.0196926032660905,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040997142110211127,
      "loss": 4.6276,
      "mean_token_accuracy": 0.25375097095966337,
      "num_tokens": 72078948.0,
      "step": 31435
    },
    {
      "entropy": 5.372958946228027,
      "epoch": 3.0201729106628243,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040994398942559496,
      "loss": 4.7626,
      "mean_token_accuracy": 0.24288332164287568,
      "num_tokens": 72090238.0,
      "step": 31440
    },
    {
      "entropy": 5.327361440658569,
      "epoch": 3.020653218059558,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004099165546160799,
      "loss": 4.6791,
      "mean_token_accuracy": 0.25163933634757996,
      "num_tokens": 72102848.0,
      "step": 31445
    },
    {
      "entropy": 5.3699586391448975,
      "epoch": 3.021133525456292,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040988911667420305,
      "loss": 4.6811,
      "mean_token_accuracy": 0.24538961946964263,
      "num_tokens": 72115363.0,
      "step": 31450
    },
    {
      "entropy": 5.352334022521973,
      "epoch": 3.021613832853026,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004098616756006015,
      "loss": 4.677,
      "mean_token_accuracy": 0.2509564682841301,
      "num_tokens": 72126768.0,
      "step": 31455
    },
    {
      "entropy": 5.346184921264649,
      "epoch": 3.02209414024976,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004098342313959122,
      "loss": 4.6698,
      "mean_token_accuracy": 0.25589456409215927,
      "num_tokens": 72138421.0,
      "step": 31460
    },
    {
      "entropy": 5.3099753856658936,
      "epoch": 3.0225744476464937,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004098067840607725,
      "loss": 4.6824,
      "mean_token_accuracy": 0.2467314198613167,
      "num_tokens": 72149568.0,
      "step": 31465
    },
    {
      "entropy": 5.402155017852783,
      "epoch": 3.0230547550432276,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004097793335958195,
      "loss": 4.7426,
      "mean_token_accuracy": 0.24879728108644486,
      "num_tokens": 72161041.0,
      "step": 31470
    },
    {
      "entropy": 5.454194688796997,
      "epoch": 3.0235350624399615,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040975188000169074,
      "loss": 4.7356,
      "mean_token_accuracy": 0.254056079685688,
      "num_tokens": 72173104.0,
      "step": 31475
    },
    {
      "entropy": 5.399863624572754,
      "epoch": 3.0240153698366954,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040972442327902325,
      "loss": 4.7213,
      "mean_token_accuracy": 0.24485732167959212,
      "num_tokens": 72183549.0,
      "step": 31480
    },
    {
      "entropy": 5.188240385055542,
      "epoch": 3.0244956772334293,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004096969634284549,
      "loss": 4.58,
      "mean_token_accuracy": 0.26399900317192077,
      "num_tokens": 72195210.0,
      "step": 31485
    },
    {
      "entropy": 5.343024635314942,
      "epoch": 3.024975984630163,
      "grad_norm": 1.125,
      "learning_rate": 0.0004096695004506229,
      "loss": 4.6488,
      "mean_token_accuracy": 0.2519909977912903,
      "num_tokens": 72205003.0,
      "step": 31490
    },
    {
      "entropy": 5.347035932540893,
      "epoch": 3.025456292026897,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040964203434616496,
      "loss": 4.6372,
      "mean_token_accuracy": 0.2511625185608864,
      "num_tokens": 72217069.0,
      "step": 31495
    },
    {
      "entropy": 5.343995380401611,
      "epoch": 3.025936599423631,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004096145651157189,
      "loss": 4.7048,
      "mean_token_accuracy": 0.24987300634384155,
      "num_tokens": 72229634.0,
      "step": 31500
    },
    {
      "entropy": 5.335697793960572,
      "epoch": 3.0264169068203652,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004095870927599223,
      "loss": 4.622,
      "mean_token_accuracy": 0.2514357313513756,
      "num_tokens": 72241061.0,
      "step": 31505
    },
    {
      "entropy": 5.2831751823425295,
      "epoch": 3.026897214217099,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00040955961727941306,
      "loss": 4.6021,
      "mean_token_accuracy": 0.2565304100513458,
      "num_tokens": 72252506.0,
      "step": 31510
    },
    {
      "entropy": 5.270777893066406,
      "epoch": 3.027377521613833,
      "grad_norm": 1.125,
      "learning_rate": 0.0004095321386748291,
      "loss": 4.5753,
      "mean_token_accuracy": 0.25636431872844695,
      "num_tokens": 72264663.0,
      "step": 31515
    },
    {
      "entropy": 5.3267049312591555,
      "epoch": 3.027857829010567,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040950465694680825,
      "loss": 4.6276,
      "mean_token_accuracy": 0.25306264907121656,
      "num_tokens": 72276665.0,
      "step": 31520
    },
    {
      "entropy": 5.305922937393189,
      "epoch": 3.0283381364073008,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040947717209598877,
      "loss": 4.6416,
      "mean_token_accuracy": 0.25625288635492327,
      "num_tokens": 72287857.0,
      "step": 31525
    },
    {
      "entropy": 5.308625555038452,
      "epoch": 3.0288184438040346,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040944968412300867,
      "loss": 4.588,
      "mean_token_accuracy": 0.2585783302783966,
      "num_tokens": 72297775.0,
      "step": 31530
    },
    {
      "entropy": 5.4323218822479244,
      "epoch": 3.0292987512007685,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040942219302850605,
      "loss": 4.784,
      "mean_token_accuracy": 0.24889058470726014,
      "num_tokens": 72309350.0,
      "step": 31535
    },
    {
      "entropy": 5.307652568817138,
      "epoch": 3.0297790585975024,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004093946988131193,
      "loss": 4.6831,
      "mean_token_accuracy": 0.24711680859327317,
      "num_tokens": 72320037.0,
      "step": 31540
    },
    {
      "entropy": 5.298952436447143,
      "epoch": 3.0302593659942363,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004093672014774868,
      "loss": 4.6455,
      "mean_token_accuracy": 0.25341269820928575,
      "num_tokens": 72331871.0,
      "step": 31545
    },
    {
      "entropy": 5.344834184646606,
      "epoch": 3.03073967339097,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040933970102224675,
      "loss": 4.5885,
      "mean_token_accuracy": 0.2605839937925339,
      "num_tokens": 72342792.0,
      "step": 31550
    },
    {
      "entropy": 5.347465133666992,
      "epoch": 3.031219980787704,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040931219744803774,
      "loss": 4.689,
      "mean_token_accuracy": 0.24449041932821275,
      "num_tokens": 72354466.0,
      "step": 31555
    },
    {
      "entropy": 5.350706005096436,
      "epoch": 3.031700288184438,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004092846907554985,
      "loss": 4.6576,
      "mean_token_accuracy": 0.24902352094650268,
      "num_tokens": 72366161.0,
      "step": 31560
    },
    {
      "entropy": 5.342910575866699,
      "epoch": 3.032180595581172,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040925718094526724,
      "loss": 4.6482,
      "mean_token_accuracy": 0.2523476928472519,
      "num_tokens": 72377724.0,
      "step": 31565
    },
    {
      "entropy": 5.450038480758667,
      "epoch": 3.0326609029779057,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040922966801798305,
      "loss": 4.829,
      "mean_token_accuracy": 0.23501116186380386,
      "num_tokens": 72390498.0,
      "step": 31570
    },
    {
      "entropy": 5.361813735961914,
      "epoch": 3.0331412103746396,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040920215197428456,
      "loss": 4.7123,
      "mean_token_accuracy": 0.25028559267520906,
      "num_tokens": 72401841.0,
      "step": 31575
    },
    {
      "entropy": 5.396618795394898,
      "epoch": 3.0336215177713735,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040917463281481053,
      "loss": 4.7418,
      "mean_token_accuracy": 0.24998713582754134,
      "num_tokens": 72412755.0,
      "step": 31580
    },
    {
      "entropy": 5.258823204040527,
      "epoch": 3.034101825168108,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004091471105402,
      "loss": 4.639,
      "mean_token_accuracy": 0.2522189497947693,
      "num_tokens": 72424516.0,
      "step": 31585
    },
    {
      "entropy": 5.276626634597778,
      "epoch": 3.0345821325648417,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004091195851510918,
      "loss": 4.6143,
      "mean_token_accuracy": 0.25553967654705045,
      "num_tokens": 72435792.0,
      "step": 31590
    },
    {
      "entropy": 5.312246370315552,
      "epoch": 3.0350624399615755,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004090920566481252,
      "loss": 4.67,
      "mean_token_accuracy": 0.25185169875621793,
      "num_tokens": 72448262.0,
      "step": 31595
    },
    {
      "entropy": 5.385905838012695,
      "epoch": 3.0355427473583094,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004090645250319392,
      "loss": 4.6363,
      "mean_token_accuracy": 0.2570544555783272,
      "num_tokens": 72458329.0,
      "step": 31600
    },
    {
      "entropy": 5.389048910140991,
      "epoch": 3.0360230547550433,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000409036990303173,
      "loss": 4.6609,
      "mean_token_accuracy": 0.25559065490961075,
      "num_tokens": 72469774.0,
      "step": 31605
    },
    {
      "entropy": 5.3356041431427,
      "epoch": 3.036503362151777,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004090094524624658,
      "loss": 4.7176,
      "mean_token_accuracy": 0.2552242636680603,
      "num_tokens": 72482619.0,
      "step": 31610
    },
    {
      "entropy": 5.359411239624023,
      "epoch": 3.036983669548511,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040898191151045717,
      "loss": 4.7337,
      "mean_token_accuracy": 0.24355191141366958,
      "num_tokens": 72493553.0,
      "step": 31615
    },
    {
      "entropy": 5.300027084350586,
      "epoch": 3.037463976945245,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004089543674477864,
      "loss": 4.568,
      "mean_token_accuracy": 0.26417059302330015,
      "num_tokens": 72504853.0,
      "step": 31620
    },
    {
      "entropy": 5.358885860443115,
      "epoch": 3.037944284341979,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004089268202750929,
      "loss": 4.6941,
      "mean_token_accuracy": 0.2598932147026062,
      "num_tokens": 72515683.0,
      "step": 31625
    },
    {
      "entropy": 5.388095426559448,
      "epoch": 3.0384245917387127,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040889926999301634,
      "loss": 4.7287,
      "mean_token_accuracy": 0.2517446532845497,
      "num_tokens": 72525699.0,
      "step": 31630
    },
    {
      "entropy": 5.3168598175048825,
      "epoch": 3.0389048991354466,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004088717166021964,
      "loss": 4.6402,
      "mean_token_accuracy": 0.24873695522546768,
      "num_tokens": 72536742.0,
      "step": 31635
    },
    {
      "entropy": 5.316224002838135,
      "epoch": 3.0393852065321805,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004088441601032727,
      "loss": 4.6355,
      "mean_token_accuracy": 0.2571908816695213,
      "num_tokens": 72548645.0,
      "step": 31640
    },
    {
      "entropy": 5.349950361251831,
      "epoch": 3.0398655139289144,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004088166004968849,
      "loss": 4.7246,
      "mean_token_accuracy": 0.24645535051822662,
      "num_tokens": 72560332.0,
      "step": 31645
    },
    {
      "entropy": 5.340791320800781,
      "epoch": 3.0403458213256482,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040878903778367317,
      "loss": 4.6794,
      "mean_token_accuracy": 0.25432224571704865,
      "num_tokens": 72572714.0,
      "step": 31650
    },
    {
      "entropy": 5.298004055023194,
      "epoch": 3.040826128722382,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004087614719642772,
      "loss": 4.623,
      "mean_token_accuracy": 0.2601295918226242,
      "num_tokens": 72583956.0,
      "step": 31655
    },
    {
      "entropy": 5.40196967124939,
      "epoch": 3.0413064361191164,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040873390303933693,
      "loss": 4.6898,
      "mean_token_accuracy": 0.2466958850622177,
      "num_tokens": 72594429.0,
      "step": 31660
    },
    {
      "entropy": 5.376281356811523,
      "epoch": 3.0417867435158503,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040870633100949266,
      "loss": 4.7293,
      "mean_token_accuracy": 0.2481798157095909,
      "num_tokens": 72605841.0,
      "step": 31665
    },
    {
      "entropy": 5.371713018417358,
      "epoch": 3.042267050912584,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040867875587538436,
      "loss": 4.6189,
      "mean_token_accuracy": 0.2571859896183014,
      "num_tokens": 72616449.0,
      "step": 31670
    },
    {
      "entropy": 5.34609088897705,
      "epoch": 3.042747358309318,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004086511776376523,
      "loss": 4.6739,
      "mean_token_accuracy": 0.24219857007265091,
      "num_tokens": 72627540.0,
      "step": 31675
    },
    {
      "entropy": 5.3747804164886475,
      "epoch": 3.043227665706052,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040862359629693684,
      "loss": 4.7583,
      "mean_token_accuracy": 0.2552459672093391,
      "num_tokens": 72638788.0,
      "step": 31680
    },
    {
      "entropy": 5.357168865203858,
      "epoch": 3.043707973102786,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004085960118538781,
      "loss": 4.736,
      "mean_token_accuracy": 0.24225880950689316,
      "num_tokens": 72650370.0,
      "step": 31685
    },
    {
      "entropy": 5.381133699417115,
      "epoch": 3.0441882804995197,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004085684243091168,
      "loss": 4.7335,
      "mean_token_accuracy": 0.252097350358963,
      "num_tokens": 72662119.0,
      "step": 31690
    },
    {
      "entropy": 5.366591596603394,
      "epoch": 3.0446685878962536,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004085408336632933,
      "loss": 4.7084,
      "mean_token_accuracy": 0.24545688927173615,
      "num_tokens": 72673063.0,
      "step": 31695
    },
    {
      "entropy": 5.306812429428101,
      "epoch": 3.0451488952929875,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040851323991704803,
      "loss": 4.5888,
      "mean_token_accuracy": 0.25396682769060136,
      "num_tokens": 72684185.0,
      "step": 31700
    },
    {
      "entropy": 5.300999402999878,
      "epoch": 3.0456292026897214,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004084856430710219,
      "loss": 4.5987,
      "mean_token_accuracy": 0.2536487028002739,
      "num_tokens": 72695313.0,
      "step": 31705
    },
    {
      "entropy": 5.303580045700073,
      "epoch": 3.0461095100864553,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004084580431258555,
      "loss": 4.6882,
      "mean_token_accuracy": 0.2522665485739708,
      "num_tokens": 72707052.0,
      "step": 31710
    },
    {
      "entropy": 5.275401830673218,
      "epoch": 3.046589817483189,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004084304400821896,
      "loss": 4.6427,
      "mean_token_accuracy": 0.25606767535209657,
      "num_tokens": 72717833.0,
      "step": 31715
    },
    {
      "entropy": 5.374927234649658,
      "epoch": 3.047070124879923,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004084028339406651,
      "loss": 4.6976,
      "mean_token_accuracy": 0.2399231895804405,
      "num_tokens": 72729591.0,
      "step": 31720
    },
    {
      "entropy": 5.3223062515258786,
      "epoch": 3.047550432276657,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040837522470192297,
      "loss": 4.6365,
      "mean_token_accuracy": 0.2560739487409592,
      "num_tokens": 72740777.0,
      "step": 31725
    },
    {
      "entropy": 5.314614915847779,
      "epoch": 3.0480307396733908,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004083476123666041,
      "loss": 4.6645,
      "mean_token_accuracy": 0.2500921130180359,
      "num_tokens": 72752093.0,
      "step": 31730
    },
    {
      "entropy": 5.397345781326294,
      "epoch": 3.048511047070125,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004083199969353496,
      "loss": 4.7886,
      "mean_token_accuracy": 0.24059210121631622,
      "num_tokens": 72763366.0,
      "step": 31735
    },
    {
      "entropy": 5.370598173141479,
      "epoch": 3.048991354466859,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040829237840880075,
      "loss": 4.6995,
      "mean_token_accuracy": 0.2521240308880806,
      "num_tokens": 72776460.0,
      "step": 31740
    },
    {
      "entropy": 5.355055904388427,
      "epoch": 3.049471661863593,
      "grad_norm": 1.125,
      "learning_rate": 0.00040826475678759855,
      "loss": 4.6893,
      "mean_token_accuracy": 0.24940920770168304,
      "num_tokens": 72788001.0,
      "step": 31745
    },
    {
      "entropy": 5.321230411529541,
      "epoch": 3.0499519692603267,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004082371320723845,
      "loss": 4.6375,
      "mean_token_accuracy": 0.25512031018733977,
      "num_tokens": 72799640.0,
      "step": 31750
    },
    {
      "entropy": 5.3442291736602785,
      "epoch": 3.0504322766570606,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040820950426379986,
      "loss": 4.672,
      "mean_token_accuracy": 0.2531282722949982,
      "num_tokens": 72810686.0,
      "step": 31755
    },
    {
      "entropy": 5.332836532592774,
      "epoch": 3.0509125840537945,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004081818733624861,
      "loss": 4.621,
      "mean_token_accuracy": 0.2527613312005997,
      "num_tokens": 72820997.0,
      "step": 31760
    },
    {
      "entropy": 5.312385749816895,
      "epoch": 3.0513928914505284,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004081542393690847,
      "loss": 4.7059,
      "mean_token_accuracy": 0.2564148619771004,
      "num_tokens": 72832044.0,
      "step": 31765
    },
    {
      "entropy": 5.339701318740845,
      "epoch": 3.0518731988472623,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004081266022842372,
      "loss": 4.6701,
      "mean_token_accuracy": 0.25347241908311846,
      "num_tokens": 72842204.0,
      "step": 31770
    },
    {
      "entropy": 5.274608325958252,
      "epoch": 3.052353506243996,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040809896210858537,
      "loss": 4.6292,
      "mean_token_accuracy": 0.2543463721871376,
      "num_tokens": 72853650.0,
      "step": 31775
    },
    {
      "entropy": 5.33727536201477,
      "epoch": 3.05283381364073,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040807131884277085,
      "loss": 4.69,
      "mean_token_accuracy": 0.2522790193557739,
      "num_tokens": 72864692.0,
      "step": 31780
    },
    {
      "entropy": 5.3455650329589846,
      "epoch": 3.053314121037464,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004080436724874354,
      "loss": 4.6681,
      "mean_token_accuracy": 0.25051446408033373,
      "num_tokens": 72875871.0,
      "step": 31785
    },
    {
      "entropy": 5.382707786560059,
      "epoch": 3.053794428434198,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040801602304322095,
      "loss": 4.6988,
      "mean_token_accuracy": 0.2508884832262993,
      "num_tokens": 72888620.0,
      "step": 31790
    },
    {
      "entropy": 5.352988243103027,
      "epoch": 3.0542747358309317,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040798837051076944,
      "loss": 4.6669,
      "mean_token_accuracy": 0.2527829632163048,
      "num_tokens": 72901022.0,
      "step": 31795
    },
    {
      "entropy": 5.409805536270142,
      "epoch": 3.0547550432276656,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040796071489072286,
      "loss": 4.7834,
      "mean_token_accuracy": 0.24368640631437302,
      "num_tokens": 72913588.0,
      "step": 31800
    },
    {
      "entropy": 5.367073345184326,
      "epoch": 3.0552353506243994,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004079330561837233,
      "loss": 4.7127,
      "mean_token_accuracy": 0.24571954607963561,
      "num_tokens": 72925397.0,
      "step": 31805
    },
    {
      "entropy": 5.340702772140503,
      "epoch": 3.0557156580211333,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040790539439041287,
      "loss": 4.636,
      "mean_token_accuracy": 0.25346615463495253,
      "num_tokens": 72936397.0,
      "step": 31810
    },
    {
      "entropy": 5.314151239395142,
      "epoch": 3.0561959654178676,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040787772951143386,
      "loss": 4.7299,
      "mean_token_accuracy": 0.2516863778233528,
      "num_tokens": 72948585.0,
      "step": 31815
    },
    {
      "entropy": 5.386921072006226,
      "epoch": 3.0566762728146015,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004078500615474285,
      "loss": 4.6948,
      "mean_token_accuracy": 0.24835402071475982,
      "num_tokens": 72961189.0,
      "step": 31820
    },
    {
      "entropy": 5.360252761840821,
      "epoch": 3.0571565802113354,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040782239049903926,
      "loss": 4.599,
      "mean_token_accuracy": 0.2579807624220848,
      "num_tokens": 72972845.0,
      "step": 31825
    },
    {
      "entropy": 5.426780319213867,
      "epoch": 3.0576368876080693,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040779471636690845,
      "loss": 4.6957,
      "mean_token_accuracy": 0.2505465283989906,
      "num_tokens": 72984401.0,
      "step": 31830
    },
    {
      "entropy": 5.336922311782837,
      "epoch": 3.058117195004803,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040776703915167866,
      "loss": 4.7046,
      "mean_token_accuracy": 0.25180926620960237,
      "num_tokens": 72995811.0,
      "step": 31835
    },
    {
      "entropy": 5.347846841812133,
      "epoch": 3.058597502401537,
      "grad_norm": 1.234375,
      "learning_rate": 0.00040773935885399254,
      "loss": 4.7124,
      "mean_token_accuracy": 0.25191009044647217,
      "num_tokens": 73006545.0,
      "step": 31840
    },
    {
      "entropy": 5.399473047256469,
      "epoch": 3.059077809798271,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004077116754744926,
      "loss": 4.6785,
      "mean_token_accuracy": 0.2561198309063911,
      "num_tokens": 73017144.0,
      "step": 31845
    },
    {
      "entropy": 5.414518880844116,
      "epoch": 3.059558117195005,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040768398901382157,
      "loss": 4.7599,
      "mean_token_accuracy": 0.24751498997211457,
      "num_tokens": 73027996.0,
      "step": 31850
    },
    {
      "entropy": 5.35634069442749,
      "epoch": 3.0600384245917387,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004076562994726223,
      "loss": 4.6791,
      "mean_token_accuracy": 0.2490548461675644,
      "num_tokens": 73038949.0,
      "step": 31855
    },
    {
      "entropy": 5.377555990219117,
      "epoch": 3.0605187319884726,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004076286068515378,
      "loss": 4.7149,
      "mean_token_accuracy": 0.2509446322917938,
      "num_tokens": 73050249.0,
      "step": 31860
    },
    {
      "entropy": 5.389694404602051,
      "epoch": 3.0609990393852065,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004076009111512108,
      "loss": 4.6492,
      "mean_token_accuracy": 0.2515803202986717,
      "num_tokens": 73061228.0,
      "step": 31865
    },
    {
      "entropy": 5.421904754638672,
      "epoch": 3.0614793467819403,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004075732123722844,
      "loss": 4.6553,
      "mean_token_accuracy": 0.25271360129117965,
      "num_tokens": 73070972.0,
      "step": 31870
    },
    {
      "entropy": 5.372074222564697,
      "epoch": 3.061959654178674,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004075455105154016,
      "loss": 4.7294,
      "mean_token_accuracy": 0.2461343824863434,
      "num_tokens": 73081569.0,
      "step": 31875
    },
    {
      "entropy": 5.35382227897644,
      "epoch": 3.062439961575408,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040751780558120573,
      "loss": 4.6911,
      "mean_token_accuracy": 0.2522904068231583,
      "num_tokens": 73092778.0,
      "step": 31880
    },
    {
      "entropy": 5.405173254013062,
      "epoch": 3.062920268972142,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004074900975703398,
      "loss": 4.7864,
      "mean_token_accuracy": 0.23980574011802674,
      "num_tokens": 73104166.0,
      "step": 31885
    },
    {
      "entropy": 5.429531097412109,
      "epoch": 3.063400576368876,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004074623864834473,
      "loss": 4.7618,
      "mean_token_accuracy": 0.24963502883911132,
      "num_tokens": 73115732.0,
      "step": 31890
    },
    {
      "entropy": 5.371484184265137,
      "epoch": 3.06388088376561,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004074346723211715,
      "loss": 4.6453,
      "mean_token_accuracy": 0.2523935839533806,
      "num_tokens": 73126305.0,
      "step": 31895
    },
    {
      "entropy": 5.306499052047729,
      "epoch": 3.064361191162344,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00040740695508415583,
      "loss": 4.6604,
      "mean_token_accuracy": 0.25280398428440093,
      "num_tokens": 73138255.0,
      "step": 31900
    },
    {
      "entropy": 5.3509925365447994,
      "epoch": 3.064841498559078,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040737923477304386,
      "loss": 4.5998,
      "mean_token_accuracy": 0.25956915616989135,
      "num_tokens": 73148359.0,
      "step": 31905
    },
    {
      "entropy": 5.347235918045044,
      "epoch": 3.065321805955812,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040735151138847917,
      "loss": 4.622,
      "mean_token_accuracy": 0.25112521797418597,
      "num_tokens": 73159781.0,
      "step": 31910
    },
    {
      "entropy": 5.267051792144775,
      "epoch": 3.0658021133525457,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004073237849311053,
      "loss": 4.6794,
      "mean_token_accuracy": 0.2550072267651558,
      "num_tokens": 73171930.0,
      "step": 31915
    },
    {
      "entropy": 5.413531923294068,
      "epoch": 3.0662824207492796,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004072960554015661,
      "loss": 4.7023,
      "mean_token_accuracy": 0.25343543142080305,
      "num_tokens": 73183842.0,
      "step": 31920
    },
    {
      "entropy": 5.390153694152832,
      "epoch": 3.0667627281460135,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004072683228005055,
      "loss": 4.7645,
      "mean_token_accuracy": 0.24879758805036545,
      "num_tokens": 73195204.0,
      "step": 31925
    },
    {
      "entropy": 5.409428644180298,
      "epoch": 3.0672430355427474,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040724058712856697,
      "loss": 4.7427,
      "mean_token_accuracy": 0.2533204048871994,
      "num_tokens": 73206829.0,
      "step": 31930
    },
    {
      "entropy": 5.319774007797241,
      "epoch": 3.0677233429394812,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004072128483863948,
      "loss": 4.6054,
      "mean_token_accuracy": 0.2673709556460381,
      "num_tokens": 73217188.0,
      "step": 31935
    },
    {
      "entropy": 5.310267162322998,
      "epoch": 3.068203650336215,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004071851065746328,
      "loss": 4.6152,
      "mean_token_accuracy": 0.25706933438777924,
      "num_tokens": 73228119.0,
      "step": 31940
    },
    {
      "entropy": 5.3446673393249515,
      "epoch": 3.068683957732949,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004071573616939252,
      "loss": 4.7128,
      "mean_token_accuracy": 0.24524996280670167,
      "num_tokens": 73239876.0,
      "step": 31945
    },
    {
      "entropy": 5.287390851974488,
      "epoch": 3.069164265129683,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004071296137449161,
      "loss": 4.6378,
      "mean_token_accuracy": 0.26135747581720353,
      "num_tokens": 73251834.0,
      "step": 31950
    },
    {
      "entropy": 5.370579767227173,
      "epoch": 3.0696445725264168,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040710186272824967,
      "loss": 4.6181,
      "mean_token_accuracy": 0.25482245087623595,
      "num_tokens": 73262949.0,
      "step": 31955
    },
    {
      "entropy": 5.326400375366211,
      "epoch": 3.0701248799231506,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004070741086445703,
      "loss": 4.6851,
      "mean_token_accuracy": 0.2557817667722702,
      "num_tokens": 73274964.0,
      "step": 31960
    },
    {
      "entropy": 5.33292818069458,
      "epoch": 3.0706051873198845,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040704635149452223,
      "loss": 4.6674,
      "mean_token_accuracy": 0.2483627900481224,
      "num_tokens": 73286527.0,
      "step": 31965
    },
    {
      "entropy": 5.316199207305909,
      "epoch": 3.071085494716619,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040701859127875,
      "loss": 4.5946,
      "mean_token_accuracy": 0.25501135885715487,
      "num_tokens": 73297197.0,
      "step": 31970
    },
    {
      "entropy": 5.376543951034546,
      "epoch": 3.0715658021133527,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040699082799789814,
      "loss": 4.705,
      "mean_token_accuracy": 0.25585541129112244,
      "num_tokens": 73308629.0,
      "step": 31975
    },
    {
      "entropy": 5.3309581756591795,
      "epoch": 3.0720461095100866,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040696306165261117,
      "loss": 4.7108,
      "mean_token_accuracy": 0.24979079216718675,
      "num_tokens": 73321420.0,
      "step": 31980
    },
    {
      "entropy": 5.406858682632446,
      "epoch": 3.0725264169068205,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004069352922435337,
      "loss": 4.7307,
      "mean_token_accuracy": 0.25020295977592466,
      "num_tokens": 73332617.0,
      "step": 31985
    },
    {
      "entropy": 5.281682682037354,
      "epoch": 3.0730067243035544,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004069075197713106,
      "loss": 4.64,
      "mean_token_accuracy": 0.2516136452555656,
      "num_tokens": 73344687.0,
      "step": 31990
    },
    {
      "entropy": 5.30743522644043,
      "epoch": 3.0734870317002883,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040687974423658655,
      "loss": 4.6243,
      "mean_token_accuracy": 0.2605572953820229,
      "num_tokens": 73355885.0,
      "step": 31995
    },
    {
      "entropy": 5.331444931030274,
      "epoch": 3.073967339097022,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040685196564000644,
      "loss": 4.658,
      "mean_token_accuracy": 0.2570805296301842,
      "num_tokens": 73367147.0,
      "step": 32000
    },
    {
      "entropy": 5.445551061630249,
      "epoch": 3.074447646493756,
      "grad_norm": 1.125,
      "learning_rate": 0.00040682418398221517,
      "loss": 4.7669,
      "mean_token_accuracy": 0.24212572276592254,
      "num_tokens": 73379780.0,
      "step": 32005
    },
    {
      "entropy": 5.307595443725586,
      "epoch": 3.07492795389049,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040679639926385783,
      "loss": 4.6578,
      "mean_token_accuracy": 0.2537778213620186,
      "num_tokens": 73392886.0,
      "step": 32010
    },
    {
      "entropy": 5.328367900848389,
      "epoch": 3.0754082612872238,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004067686114855794,
      "loss": 4.6554,
      "mean_token_accuracy": 0.2561619907617569,
      "num_tokens": 73404153.0,
      "step": 32015
    },
    {
      "entropy": 5.339376592636109,
      "epoch": 3.0758885686839577,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040674082064802507,
      "loss": 4.7104,
      "mean_token_accuracy": 0.24731936305761337,
      "num_tokens": 73415636.0,
      "step": 32020
    },
    {
      "entropy": 5.457571983337402,
      "epoch": 3.0763688760806915,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004067130267518401,
      "loss": 4.8746,
      "mean_token_accuracy": 0.24117017537355423,
      "num_tokens": 73426349.0,
      "step": 32025
    },
    {
      "entropy": 5.3902263164520265,
      "epoch": 3.0768491834774254,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004066852297976698,
      "loss": 4.6848,
      "mean_token_accuracy": 0.2449132099747658,
      "num_tokens": 73437936.0,
      "step": 32030
    },
    {
      "entropy": 5.40680341720581,
      "epoch": 3.0773294908741593,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004066574297861595,
      "loss": 4.6983,
      "mean_token_accuracy": 0.2528448760509491,
      "num_tokens": 73449818.0,
      "step": 32035
    },
    {
      "entropy": 5.261013412475586,
      "epoch": 3.077809798270893,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040662962671795454,
      "loss": 4.5509,
      "mean_token_accuracy": 0.25929124355316163,
      "num_tokens": 73459904.0,
      "step": 32040
    },
    {
      "entropy": 5.292002201080322,
      "epoch": 3.0782901056676275,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004066018205937006,
      "loss": 4.6738,
      "mean_token_accuracy": 0.24907797425985337,
      "num_tokens": 73470993.0,
      "step": 32045
    },
    {
      "entropy": 5.308251571655274,
      "epoch": 3.0787704130643614,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004065740114140431,
      "loss": 4.6079,
      "mean_token_accuracy": 0.2531385779380798,
      "num_tokens": 73482205.0,
      "step": 32050
    },
    {
      "entropy": 5.313709354400634,
      "epoch": 3.0792507204610953,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040654619917962774,
      "loss": 4.6259,
      "mean_token_accuracy": 0.2540906071662903,
      "num_tokens": 73493990.0,
      "step": 32055
    },
    {
      "entropy": 5.400446176528931,
      "epoch": 3.079731027857829,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004065183838911003,
      "loss": 4.7906,
      "mean_token_accuracy": 0.24296780228614806,
      "num_tokens": 73505738.0,
      "step": 32060
    },
    {
      "entropy": 5.3133862018585205,
      "epoch": 3.080211335254563,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004064905655491065,
      "loss": 4.5782,
      "mean_token_accuracy": 0.2614894941449165,
      "num_tokens": 73517248.0,
      "step": 32065
    },
    {
      "entropy": 5.4370402812957765,
      "epoch": 3.080691642651297,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040646274415429224,
      "loss": 4.7863,
      "mean_token_accuracy": 0.2481693759560585,
      "num_tokens": 73528496.0,
      "step": 32070
    },
    {
      "entropy": 5.408049392700195,
      "epoch": 3.081171950048031,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004064349197073033,
      "loss": 4.7072,
      "mean_token_accuracy": 0.24118769615888597,
      "num_tokens": 73540672.0,
      "step": 32075
    },
    {
      "entropy": 5.392041397094727,
      "epoch": 3.0816522574447647,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004064070922087859,
      "loss": 4.6809,
      "mean_token_accuracy": 0.25623157173395156,
      "num_tokens": 73551683.0,
      "step": 32080
    },
    {
      "entropy": 5.335136938095093,
      "epoch": 3.0821325648414986,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040637926165938606,
      "loss": 4.6121,
      "mean_token_accuracy": 0.25547813922166823,
      "num_tokens": 73562981.0,
      "step": 32085
    },
    {
      "entropy": 5.391943454742432,
      "epoch": 3.0826128722382324,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040635142805974986,
      "loss": 4.7592,
      "mean_token_accuracy": 0.25423442721366885,
      "num_tokens": 73574082.0,
      "step": 32090
    },
    {
      "entropy": 5.327897310256958,
      "epoch": 3.0830931796349663,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004063235914105235,
      "loss": 4.6083,
      "mean_token_accuracy": 0.25730431973934176,
      "num_tokens": 73586305.0,
      "step": 32095
    },
    {
      "entropy": 5.417872571945191,
      "epoch": 3.0835734870317,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040629575171235327,
      "loss": 4.7669,
      "mean_token_accuracy": 0.24906253963708877,
      "num_tokens": 73598036.0,
      "step": 32100
    },
    {
      "entropy": 5.361555767059326,
      "epoch": 3.084053794428434,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004062679089658856,
      "loss": 4.6814,
      "mean_token_accuracy": 0.25401071459054947,
      "num_tokens": 73608870.0,
      "step": 32105
    },
    {
      "entropy": 5.411904048919678,
      "epoch": 3.084534101825168,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040624006317176685,
      "loss": 4.6967,
      "mean_token_accuracy": 0.24250486493110657,
      "num_tokens": 73619965.0,
      "step": 32110
    },
    {
      "entropy": 5.327812957763672,
      "epoch": 3.085014409221902,
      "grad_norm": 1.28125,
      "learning_rate": 0.00040621221433064354,
      "loss": 4.6007,
      "mean_token_accuracy": 0.2575634986162186,
      "num_tokens": 73631183.0,
      "step": 32115
    },
    {
      "entropy": 5.330556011199951,
      "epoch": 3.0854947166186357,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004061843624431623,
      "loss": 4.6664,
      "mean_token_accuracy": 0.25625911355018616,
      "num_tokens": 73643550.0,
      "step": 32120
    },
    {
      "entropy": 5.381673812866211,
      "epoch": 3.08597502401537,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040615650750996956,
      "loss": 4.7359,
      "mean_token_accuracy": 0.2514713302254677,
      "num_tokens": 73654931.0,
      "step": 32125
    },
    {
      "entropy": 5.4047918796539305,
      "epoch": 3.086455331412104,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00040612864953171223,
      "loss": 4.7484,
      "mean_token_accuracy": 0.24646413177251816,
      "num_tokens": 73667771.0,
      "step": 32130
    },
    {
      "entropy": 5.401779317855835,
      "epoch": 3.086935638808838,
      "grad_norm": 1.25,
      "learning_rate": 0.00040610078850903715,
      "loss": 4.6509,
      "mean_token_accuracy": 0.2611093193292618,
      "num_tokens": 73678756.0,
      "step": 32135
    },
    {
      "entropy": 5.37713508605957,
      "epoch": 3.0874159462055717,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040607292444259094,
      "loss": 4.6589,
      "mean_token_accuracy": 0.2577410191297531,
      "num_tokens": 73689312.0,
      "step": 32140
    },
    {
      "entropy": 5.297479677200317,
      "epoch": 3.0878962536023056,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004060450573330206,
      "loss": 4.5936,
      "mean_token_accuracy": 0.2583267852663994,
      "num_tokens": 73702472.0,
      "step": 32145
    },
    {
      "entropy": 5.312840700149536,
      "epoch": 3.0883765609990395,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040601718718097325,
      "loss": 4.6798,
      "mean_token_accuracy": 0.2501051515340805,
      "num_tokens": 73714687.0,
      "step": 32150
    },
    {
      "entropy": 5.321476984024048,
      "epoch": 3.0888568683957733,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040598931398709576,
      "loss": 4.6507,
      "mean_token_accuracy": 0.23996489942073823,
      "num_tokens": 73726137.0,
      "step": 32155
    },
    {
      "entropy": 5.321951293945313,
      "epoch": 3.089337175792507,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040596143775203534,
      "loss": 4.6667,
      "mean_token_accuracy": 0.2534803792834282,
      "num_tokens": 73737409.0,
      "step": 32160
    },
    {
      "entropy": 5.307781839370728,
      "epoch": 3.089817483189241,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00040593355847643933,
      "loss": 4.648,
      "mean_token_accuracy": 0.25831172317266465,
      "num_tokens": 73750408.0,
      "step": 32165
    },
    {
      "entropy": 5.357502222061157,
      "epoch": 3.090297790585975,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004059056761609548,
      "loss": 4.6666,
      "mean_token_accuracy": 0.24247232377529143,
      "num_tokens": 73762417.0,
      "step": 32170
    },
    {
      "entropy": 5.345710134506225,
      "epoch": 3.090778097982709,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004058777908062292,
      "loss": 4.6594,
      "mean_token_accuracy": 0.25227912962436677,
      "num_tokens": 73774867.0,
      "step": 32175
    },
    {
      "entropy": 5.398387098312378,
      "epoch": 3.0912584053794427,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004058499024129099,
      "loss": 4.7364,
      "mean_token_accuracy": 0.24610422253608705,
      "num_tokens": 73785619.0,
      "step": 32180
    },
    {
      "entropy": 5.357243537902832,
      "epoch": 3.0917387127761766,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040582201098164443,
      "loss": 4.717,
      "mean_token_accuracy": 0.25089701265096664,
      "num_tokens": 73797520.0,
      "step": 32185
    },
    {
      "entropy": 5.292101287841797,
      "epoch": 3.0922190201729105,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040579411651308034,
      "loss": 4.6358,
      "mean_token_accuracy": 0.24919472485780716,
      "num_tokens": 73809135.0,
      "step": 32190
    },
    {
      "entropy": 5.276713848114014,
      "epoch": 3.0926993275696444,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040576621900786523,
      "loss": 4.5968,
      "mean_token_accuracy": 0.2618730187416077,
      "num_tokens": 73820126.0,
      "step": 32195
    },
    {
      "entropy": 5.459673023223877,
      "epoch": 3.0931796349663783,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004057383184666468,
      "loss": 4.8031,
      "mean_token_accuracy": 0.24291711747646333,
      "num_tokens": 73832788.0,
      "step": 32200
    },
    {
      "entropy": 5.36544942855835,
      "epoch": 3.0936599423631126,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040571041489007286,
      "loss": 4.6883,
      "mean_token_accuracy": 0.2562264710664749,
      "num_tokens": 73845322.0,
      "step": 32205
    },
    {
      "entropy": 5.34296350479126,
      "epoch": 3.0941402497598465,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040568250827879127,
      "loss": 4.7234,
      "mean_token_accuracy": 0.25495515316724776,
      "num_tokens": 73857818.0,
      "step": 32210
    },
    {
      "entropy": 5.239820957183838,
      "epoch": 3.0946205571565804,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004056545986334497,
      "loss": 4.535,
      "mean_token_accuracy": 0.26432245820760725,
      "num_tokens": 73869373.0,
      "step": 32215
    },
    {
      "entropy": 5.339227104187012,
      "epoch": 3.0951008645533142,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004056266859546965,
      "loss": 4.6198,
      "mean_token_accuracy": 0.2618795782327652,
      "num_tokens": 73879219.0,
      "step": 32220
    },
    {
      "entropy": 5.342383575439453,
      "epoch": 3.095581171950048,
      "grad_norm": 1.125,
      "learning_rate": 0.0004055987702431795,
      "loss": 4.6892,
      "mean_token_accuracy": 0.2535675376653671,
      "num_tokens": 73890287.0,
      "step": 32225
    },
    {
      "entropy": 5.30146017074585,
      "epoch": 3.096061479346782,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040557085149954677,
      "loss": 4.668,
      "mean_token_accuracy": 0.2482043132185936,
      "num_tokens": 73902798.0,
      "step": 32230
    },
    {
      "entropy": 5.358974409103394,
      "epoch": 3.096541786743516,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040554292972444663,
      "loss": 4.6922,
      "mean_token_accuracy": 0.2517870500683784,
      "num_tokens": 73914042.0,
      "step": 32235
    },
    {
      "entropy": 5.2597626686096195,
      "epoch": 3.0970220941402498,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040551500491852735,
      "loss": 4.5687,
      "mean_token_accuracy": 0.2613095104694366,
      "num_tokens": 73926682.0,
      "step": 32240
    },
    {
      "entropy": 5.339566421508789,
      "epoch": 3.0975024015369836,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004054870770824371,
      "loss": 4.6859,
      "mean_token_accuracy": 0.2526316732168198,
      "num_tokens": 73937657.0,
      "step": 32245
    },
    {
      "entropy": 5.370836067199707,
      "epoch": 3.0979827089337175,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040545914621682445,
      "loss": 4.6824,
      "mean_token_accuracy": 0.2519138678908348,
      "num_tokens": 73950234.0,
      "step": 32250
    },
    {
      "entropy": 5.412867784500122,
      "epoch": 3.0984630163304514,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004054312123223378,
      "loss": 4.7031,
      "mean_token_accuracy": 0.24845329970121383,
      "num_tokens": 73961404.0,
      "step": 32255
    },
    {
      "entropy": 5.341496372222901,
      "epoch": 3.0989433237271853,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040540327539962567,
      "loss": 4.6812,
      "mean_token_accuracy": 0.24872687458992004,
      "num_tokens": 73973189.0,
      "step": 32260
    },
    {
      "entropy": 5.346917581558228,
      "epoch": 3.099423631123919,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040537533544933674,
      "loss": 4.667,
      "mean_token_accuracy": 0.2540119543671608,
      "num_tokens": 73985303.0,
      "step": 32265
    },
    {
      "entropy": 5.3270592212677,
      "epoch": 3.099903938520653,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004053473924721197,
      "loss": 4.6188,
      "mean_token_accuracy": 0.2544162154197693,
      "num_tokens": 73997008.0,
      "step": 32270
    },
    {
      "entropy": 5.330216550827027,
      "epoch": 3.100384245917387,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004053194464686232,
      "loss": 4.662,
      "mean_token_accuracy": 0.25404727309942243,
      "num_tokens": 74007752.0,
      "step": 32275
    },
    {
      "entropy": 5.322950172424316,
      "epoch": 3.1008645533141213,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004052914974394961,
      "loss": 4.6795,
      "mean_token_accuracy": 0.2539056032896042,
      "num_tokens": 74020247.0,
      "step": 32280
    },
    {
      "entropy": 5.401501035690307,
      "epoch": 3.101344860710855,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040526354538538735,
      "loss": 4.706,
      "mean_token_accuracy": 0.2522758200764656,
      "num_tokens": 74030823.0,
      "step": 32285
    },
    {
      "entropy": 5.368149089813232,
      "epoch": 3.101825168107589,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004052355903069459,
      "loss": 4.6458,
      "mean_token_accuracy": 0.24978213459253312,
      "num_tokens": 74042157.0,
      "step": 32290
    },
    {
      "entropy": 5.32825231552124,
      "epoch": 3.102305475504323,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004052076322048207,
      "loss": 4.74,
      "mean_token_accuracy": 0.24678728282451629,
      "num_tokens": 74052772.0,
      "step": 32295
    },
    {
      "entropy": 5.343772315979004,
      "epoch": 3.1027857829010568,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040517967107966095,
      "loss": 4.6573,
      "mean_token_accuracy": 0.25774564743041994,
      "num_tokens": 74063694.0,
      "step": 32300
    },
    {
      "entropy": 5.295315074920654,
      "epoch": 3.1032660902977907,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040515170693211584,
      "loss": 4.542,
      "mean_token_accuracy": 0.25695219337940217,
      "num_tokens": 74075249.0,
      "step": 32305
    },
    {
      "entropy": 5.304211854934692,
      "epoch": 3.1037463976945245,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004051237397628345,
      "loss": 4.7009,
      "mean_token_accuracy": 0.2428532838821411,
      "num_tokens": 74086495.0,
      "step": 32310
    },
    {
      "entropy": 5.394620752334594,
      "epoch": 3.1042267050912584,
      "grad_norm": 1.125,
      "learning_rate": 0.0004050957695724663,
      "loss": 4.7498,
      "mean_token_accuracy": 0.24474676698446274,
      "num_tokens": 74097812.0,
      "step": 32315
    },
    {
      "entropy": 5.278182029724121,
      "epoch": 3.1047070124879923,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004050677963616607,
      "loss": 4.5553,
      "mean_token_accuracy": 0.262680584192276,
      "num_tokens": 74108680.0,
      "step": 32320
    },
    {
      "entropy": 5.308535194396972,
      "epoch": 3.105187319884726,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040503982013106706,
      "loss": 4.6645,
      "mean_token_accuracy": 0.2578253835439682,
      "num_tokens": 74120666.0,
      "step": 32325
    },
    {
      "entropy": 5.2831933975219725,
      "epoch": 3.10566762728146,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000405011840881335,
      "loss": 4.6422,
      "mean_token_accuracy": 0.2589506581425667,
      "num_tokens": 74133092.0,
      "step": 32330
    },
    {
      "entropy": 5.402908754348755,
      "epoch": 3.106147934678194,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004049838586131139,
      "loss": 4.7376,
      "mean_token_accuracy": 0.24244891554117204,
      "num_tokens": 74145109.0,
      "step": 32335
    },
    {
      "entropy": 5.3592408180236815,
      "epoch": 3.106628242074928,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004049558733270537,
      "loss": 4.6663,
      "mean_token_accuracy": 0.2519542172551155,
      "num_tokens": 74156505.0,
      "step": 32340
    },
    {
      "entropy": 5.3785051822662355,
      "epoch": 3.1071085494716617,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000404927885023804,
      "loss": 4.7034,
      "mean_token_accuracy": 0.25373289585113523,
      "num_tokens": 74168213.0,
      "step": 32345
    },
    {
      "entropy": 5.345128059387207,
      "epoch": 3.1075888568683956,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040489989370401456,
      "loss": 4.6672,
      "mean_token_accuracy": 0.2547969937324524,
      "num_tokens": 74178758.0,
      "step": 32350
    },
    {
      "entropy": 5.323075485229492,
      "epoch": 3.1080691642651295,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004048718993683353,
      "loss": 4.6574,
      "mean_token_accuracy": 0.263080695271492,
      "num_tokens": 74189716.0,
      "step": 32355
    },
    {
      "entropy": 5.34022889137268,
      "epoch": 3.108549471661864,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040484390201741627,
      "loss": 4.6747,
      "mean_token_accuracy": 0.25240445882081985,
      "num_tokens": 74202225.0,
      "step": 32360
    },
    {
      "entropy": 5.283945941925049,
      "epoch": 3.1090297790585977,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004048159016519073,
      "loss": 4.5992,
      "mean_token_accuracy": 0.2564329743385315,
      "num_tokens": 74212573.0,
      "step": 32365
    },
    {
      "entropy": 5.321036338806152,
      "epoch": 3.1095100864553316,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004047878982724586,
      "loss": 4.6798,
      "mean_token_accuracy": 0.253999562561512,
      "num_tokens": 74223415.0,
      "step": 32370
    },
    {
      "entropy": 5.299383354187012,
      "epoch": 3.1099903938520654,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040475989187972034,
      "loss": 4.6582,
      "mean_token_accuracy": 0.24798443913459778,
      "num_tokens": 74235046.0,
      "step": 32375
    },
    {
      "entropy": 5.291674184799194,
      "epoch": 3.1104707012487993,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040473188247434265,
      "loss": 4.6699,
      "mean_token_accuracy": 0.25622630268335345,
      "num_tokens": 74247286.0,
      "step": 32380
    },
    {
      "entropy": 5.292219543457032,
      "epoch": 3.110951008645533,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040470387005697587,
      "loss": 4.6337,
      "mean_token_accuracy": 0.2562601536512375,
      "num_tokens": 74258753.0,
      "step": 32385
    },
    {
      "entropy": 5.338306188583374,
      "epoch": 3.111431316042267,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004046758546282704,
      "loss": 4.7065,
      "mean_token_accuracy": 0.25471257865428926,
      "num_tokens": 74268878.0,
      "step": 32390
    },
    {
      "entropy": 5.369427490234375,
      "epoch": 3.111911623439001,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004046478361888766,
      "loss": 4.597,
      "mean_token_accuracy": 0.25268134027719497,
      "num_tokens": 74280912.0,
      "step": 32395
    },
    {
      "entropy": 5.314680528640747,
      "epoch": 3.112391930835735,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000404619814739445,
      "loss": 4.6531,
      "mean_token_accuracy": 0.25540547370910643,
      "num_tokens": 74291960.0,
      "step": 32400
    },
    {
      "entropy": 5.367692136764527,
      "epoch": 3.1128722382324687,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004045917902806263,
      "loss": 4.6062,
      "mean_token_accuracy": 0.24477706849575043,
      "num_tokens": 74302285.0,
      "step": 32405
    },
    {
      "entropy": 5.392345905303955,
      "epoch": 3.1133525456292026,
      "grad_norm": 1.34375,
      "learning_rate": 0.000404563762813071,
      "loss": 4.7455,
      "mean_token_accuracy": 0.24295269101858138,
      "num_tokens": 74313833.0,
      "step": 32410
    },
    {
      "entropy": 5.441923952102661,
      "epoch": 3.1138328530259365,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004045357323374298,
      "loss": 4.8081,
      "mean_token_accuracy": 0.2504452675580978,
      "num_tokens": 74324910.0,
      "step": 32415
    },
    {
      "entropy": 5.336488676071167,
      "epoch": 3.1143131604226704,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040450769885435364,
      "loss": 4.7054,
      "mean_token_accuracy": 0.24519082754850388,
      "num_tokens": 74337860.0,
      "step": 32420
    },
    {
      "entropy": 5.288605976104736,
      "epoch": 3.1147934678194042,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040447966236449313,
      "loss": 4.5802,
      "mean_token_accuracy": 0.2559446781873703,
      "num_tokens": 74348769.0,
      "step": 32425
    },
    {
      "entropy": 5.313462638854981,
      "epoch": 3.115273775216138,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00040445162286849935,
      "loss": 4.6609,
      "mean_token_accuracy": 0.25397994816303254,
      "num_tokens": 74360007.0,
      "step": 32430
    },
    {
      "entropy": 5.353163814544677,
      "epoch": 3.115754082612872,
      "grad_norm": 1.28125,
      "learning_rate": 0.00040442358036702343,
      "loss": 4.6717,
      "mean_token_accuracy": 0.2534180998802185,
      "num_tokens": 74369455.0,
      "step": 32435
    },
    {
      "entropy": 5.354326295852661,
      "epoch": 3.1162343900096063,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004043955348607161,
      "loss": 4.6365,
      "mean_token_accuracy": 0.25923206359148027,
      "num_tokens": 74381696.0,
      "step": 32440
    },
    {
      "entropy": 5.3828271389007565,
      "epoch": 3.11671469740634,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004043674863502288,
      "loss": 4.6993,
      "mean_token_accuracy": 0.24577438533306123,
      "num_tokens": 74392300.0,
      "step": 32445
    },
    {
      "entropy": 5.325166845321656,
      "epoch": 3.117195004803074,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040433943483621253,
      "loss": 4.6879,
      "mean_token_accuracy": 0.26010431349277496,
      "num_tokens": 74403953.0,
      "step": 32450
    },
    {
      "entropy": 5.395809030532837,
      "epoch": 3.117675312199808,
      "grad_norm": 1.125,
      "learning_rate": 0.0004043113803193187,
      "loss": 4.722,
      "mean_token_accuracy": 0.24852848798036575,
      "num_tokens": 74415789.0,
      "step": 32455
    },
    {
      "entropy": 5.4393229484558105,
      "epoch": 3.118155619596542,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040428332280019864,
      "loss": 4.7985,
      "mean_token_accuracy": 0.23630104660987855,
      "num_tokens": 74428173.0,
      "step": 32460
    },
    {
      "entropy": 5.299758195877075,
      "epoch": 3.1186359269932757,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004042552622795036,
      "loss": 4.5404,
      "mean_token_accuracy": 0.26593359410762785,
      "num_tokens": 74439623.0,
      "step": 32465
    },
    {
      "entropy": 5.3567667484283445,
      "epoch": 3.1191162343900096,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004042271987578852,
      "loss": 4.7444,
      "mean_token_accuracy": 0.24618625938892363,
      "num_tokens": 74451194.0,
      "step": 32470
    },
    {
      "entropy": 5.342185163497925,
      "epoch": 3.1195965417867435,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040419913223599505,
      "loss": 4.6272,
      "mean_token_accuracy": 0.25803307592868807,
      "num_tokens": 74462395.0,
      "step": 32475
    },
    {
      "entropy": 5.316565656661988,
      "epoch": 3.1200768491834774,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040417106271448464,
      "loss": 4.603,
      "mean_token_accuracy": 0.25949239283800124,
      "num_tokens": 74473332.0,
      "step": 32480
    },
    {
      "entropy": 5.443978595733642,
      "epoch": 3.1205571565802113,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004041429901940057,
      "loss": 4.8034,
      "mean_token_accuracy": 0.23874870240688323,
      "num_tokens": 74485605.0,
      "step": 32485
    },
    {
      "entropy": 5.282628536224365,
      "epoch": 3.121037463976945,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00040411491467521,
      "loss": 4.6015,
      "mean_token_accuracy": 0.26029004007577894,
      "num_tokens": 74496901.0,
      "step": 32490
    },
    {
      "entropy": 5.413909196853638,
      "epoch": 3.121517771373679,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004040868361587494,
      "loss": 4.8003,
      "mean_token_accuracy": 0.24271841943264008,
      "num_tokens": 74508067.0,
      "step": 32495
    },
    {
      "entropy": 5.417503786087036,
      "epoch": 3.121998078770413,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004040587546452758,
      "loss": 4.7006,
      "mean_token_accuracy": 0.24641806185245513,
      "num_tokens": 74519991.0,
      "step": 32500
    },
    {
      "entropy": 5.360054445266724,
      "epoch": 3.122478386167147,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040403067013544116,
      "loss": 4.686,
      "mean_token_accuracy": 0.2496445968747139,
      "num_tokens": 74532462.0,
      "step": 32505
    },
    {
      "entropy": 5.352707433700561,
      "epoch": 3.1229586935638807,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040400258262989744,
      "loss": 4.7208,
      "mean_token_accuracy": 0.24847717434167862,
      "num_tokens": 74543161.0,
      "step": 32510
    },
    {
      "entropy": 5.366953277587891,
      "epoch": 3.123439000960615,
      "grad_norm": 1.25,
      "learning_rate": 0.00040397449212929676,
      "loss": 4.7135,
      "mean_token_accuracy": 0.2462715670466423,
      "num_tokens": 74554904.0,
      "step": 32515
    },
    {
      "entropy": 5.384524726867676,
      "epoch": 3.123919308357349,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004039463986342914,
      "loss": 4.6327,
      "mean_token_accuracy": 0.2526039808988571,
      "num_tokens": 74566632.0,
      "step": 32520
    },
    {
      "entropy": 5.35510630607605,
      "epoch": 3.1243996157540828,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040391830214553365,
      "loss": 4.7079,
      "mean_token_accuracy": 0.2507174342870712,
      "num_tokens": 74578151.0,
      "step": 32525
    },
    {
      "entropy": 5.338244533538818,
      "epoch": 3.1248799231508166,
      "grad_norm": 1.125,
      "learning_rate": 0.0004038902026636756,
      "loss": 4.732,
      "mean_token_accuracy": 0.24862590283155442,
      "num_tokens": 74589615.0,
      "step": 32530
    },
    {
      "entropy": 5.371605014801025,
      "epoch": 3.1253602305475505,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004038621001893698,
      "loss": 4.6919,
      "mean_token_accuracy": 0.2509061902761459,
      "num_tokens": 74600549.0,
      "step": 32535
    },
    {
      "entropy": 5.328341817855835,
      "epoch": 3.1258405379442844,
      "grad_norm": 1.125,
      "learning_rate": 0.00040383399472326874,
      "loss": 4.6497,
      "mean_token_accuracy": 0.2600363284349442,
      "num_tokens": 74612449.0,
      "step": 32540
    },
    {
      "entropy": 5.261185979843139,
      "epoch": 3.1263208453410183,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00040380588626602484,
      "loss": 4.567,
      "mean_token_accuracy": 0.2580269619822502,
      "num_tokens": 74623804.0,
      "step": 32545
    },
    {
      "entropy": 5.296522569656372,
      "epoch": 3.126801152737752,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004037777748182907,
      "loss": 4.6714,
      "mean_token_accuracy": 0.25429490208625793,
      "num_tokens": 74636711.0,
      "step": 32550
    },
    {
      "entropy": 5.410401630401611,
      "epoch": 3.127281460134486,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004037496603807191,
      "loss": 4.7017,
      "mean_token_accuracy": 0.24863254129886628,
      "num_tokens": 74649037.0,
      "step": 32555
    },
    {
      "entropy": 5.375479984283447,
      "epoch": 3.12776176753122,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004037215429539626,
      "loss": 4.7001,
      "mean_token_accuracy": 0.24472954273223876,
      "num_tokens": 74660730.0,
      "step": 32560
    },
    {
      "entropy": 5.328767156600952,
      "epoch": 3.128242074927954,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040369342253867413,
      "loss": 4.6722,
      "mean_token_accuracy": 0.2537010610103607,
      "num_tokens": 74672558.0,
      "step": 32565
    },
    {
      "entropy": 5.295497083663941,
      "epoch": 3.1287223823246877,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004036652991355066,
      "loss": 4.6206,
      "mean_token_accuracy": 0.2606527492403984,
      "num_tokens": 74684777.0,
      "step": 32570
    },
    {
      "entropy": 5.419689559936524,
      "epoch": 3.1292026897214216,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004036371727451128,
      "loss": 4.7036,
      "mean_token_accuracy": 0.2497929647564888,
      "num_tokens": 74695706.0,
      "step": 32575
    },
    {
      "entropy": 5.365591239929199,
      "epoch": 3.1296829971181555,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040360904336814586,
      "loss": 4.7081,
      "mean_token_accuracy": 0.24394367635250092,
      "num_tokens": 74708277.0,
      "step": 32580
    },
    {
      "entropy": 5.355208253860473,
      "epoch": 3.1301633045148893,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004035809110052588,
      "loss": 4.6861,
      "mean_token_accuracy": 0.25324382036924364,
      "num_tokens": 74719188.0,
      "step": 32585
    },
    {
      "entropy": 5.372652959823609,
      "epoch": 3.1306436119116237,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004035527756571048,
      "loss": 4.7068,
      "mean_token_accuracy": 0.2506454885005951,
      "num_tokens": 74730170.0,
      "step": 32590
    },
    {
      "entropy": 5.312465381622315,
      "epoch": 3.1311239193083575,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040352463732433707,
      "loss": 4.6048,
      "mean_token_accuracy": 0.24756226688623428,
      "num_tokens": 74741212.0,
      "step": 32595
    },
    {
      "entropy": 5.365481662750244,
      "epoch": 3.1316042267050914,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040349649600760894,
      "loss": 4.6752,
      "mean_token_accuracy": 0.25049824714660646,
      "num_tokens": 74752323.0,
      "step": 32600
    },
    {
      "entropy": 5.381917667388916,
      "epoch": 3.1320845341018253,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004034683517075737,
      "loss": 4.7145,
      "mean_token_accuracy": 0.2516589298844337,
      "num_tokens": 74763462.0,
      "step": 32605
    },
    {
      "entropy": 5.412171316146851,
      "epoch": 3.132564841498559,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040344020442488476,
      "loss": 4.7001,
      "mean_token_accuracy": 0.25434879064559934,
      "num_tokens": 74774701.0,
      "step": 32610
    },
    {
      "entropy": 5.388439559936524,
      "epoch": 3.133045148895293,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040341205416019577,
      "loss": 4.7124,
      "mean_token_accuracy": 0.24899368286132811,
      "num_tokens": 74786378.0,
      "step": 32615
    },
    {
      "entropy": 5.369321489334107,
      "epoch": 3.133525456292027,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004033839009141601,
      "loss": 4.6738,
      "mean_token_accuracy": 0.2502829313278198,
      "num_tokens": 74798842.0,
      "step": 32620
    },
    {
      "entropy": 5.355684041976929,
      "epoch": 3.134005763688761,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040335574468743145,
      "loss": 4.6364,
      "mean_token_accuracy": 0.2477942481637001,
      "num_tokens": 74809707.0,
      "step": 32625
    },
    {
      "entropy": 5.339378356933594,
      "epoch": 3.1344860710854947,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004033275854806636,
      "loss": 4.679,
      "mean_token_accuracy": 0.25052040368318557,
      "num_tokens": 74821075.0,
      "step": 32630
    },
    {
      "entropy": 5.390213537216186,
      "epoch": 3.1349663784822286,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004032994232945103,
      "loss": 4.7493,
      "mean_token_accuracy": 0.24196498543024064,
      "num_tokens": 74833096.0,
      "step": 32635
    },
    {
      "entropy": 5.3934308052062985,
      "epoch": 3.1354466858789625,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004032712581296253,
      "loss": 4.7153,
      "mean_token_accuracy": 0.24887518733739852,
      "num_tokens": 74844512.0,
      "step": 32640
    },
    {
      "entropy": 5.408480310440064,
      "epoch": 3.1359269932756964,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040324308998666267,
      "loss": 4.7069,
      "mean_token_accuracy": 0.26000200808048246,
      "num_tokens": 74854253.0,
      "step": 32645
    },
    {
      "entropy": 5.34100980758667,
      "epoch": 3.1364073006724302,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040321491886627614,
      "loss": 4.7225,
      "mean_token_accuracy": 0.2479427143931389,
      "num_tokens": 74865772.0,
      "step": 32650
    },
    {
      "entropy": 5.385904169082641,
      "epoch": 3.136887608069164,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040318674476912006,
      "loss": 4.7007,
      "mean_token_accuracy": 0.25099838078022,
      "num_tokens": 74875773.0,
      "step": 32655
    },
    {
      "entropy": 5.38441915512085,
      "epoch": 3.137367915465898,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004031585676958483,
      "loss": 4.7708,
      "mean_token_accuracy": 0.24392875730991365,
      "num_tokens": 74887508.0,
      "step": 32660
    },
    {
      "entropy": 5.2990752220153805,
      "epoch": 3.1378482228626323,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040313038764711517,
      "loss": 4.5529,
      "mean_token_accuracy": 0.26582110673189163,
      "num_tokens": 74897350.0,
      "step": 32665
    },
    {
      "entropy": 5.263968086242675,
      "epoch": 3.138328530259366,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040310220462357494,
      "loss": 4.5867,
      "mean_token_accuracy": 0.2662956461310387,
      "num_tokens": 74908697.0,
      "step": 32670
    },
    {
      "entropy": 5.2703300476074215,
      "epoch": 3.1388088376561,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004030740186258819,
      "loss": 4.6397,
      "mean_token_accuracy": 0.2568092554807663,
      "num_tokens": 74919012.0,
      "step": 32675
    },
    {
      "entropy": 5.310075664520264,
      "epoch": 3.139289145052834,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004030458296546905,
      "loss": 4.6921,
      "mean_token_accuracy": 0.25398377180099485,
      "num_tokens": 74931266.0,
      "step": 32680
    },
    {
      "entropy": 5.3917149066925045,
      "epoch": 3.139769452449568,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040301763771065504,
      "loss": 4.7254,
      "mean_token_accuracy": 0.2552050784230232,
      "num_tokens": 74942860.0,
      "step": 32685
    },
    {
      "entropy": 5.342780256271363,
      "epoch": 3.1402497598463017,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004029894427944302,
      "loss": 4.6048,
      "mean_token_accuracy": 0.26088459342718123,
      "num_tokens": 74955441.0,
      "step": 32690
    },
    {
      "entropy": 5.275699186325073,
      "epoch": 3.1407300672430356,
      "grad_norm": 1.234375,
      "learning_rate": 0.00040296124490667065,
      "loss": 4.6132,
      "mean_token_accuracy": 0.2565402388572693,
      "num_tokens": 74966830.0,
      "step": 32695
    },
    {
      "entropy": 5.398275136947632,
      "epoch": 3.1412103746397695,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004029330440480308,
      "loss": 4.7072,
      "mean_token_accuracy": 0.24738182127475739,
      "num_tokens": 74978380.0,
      "step": 32700
    },
    {
      "entropy": 5.372025346755981,
      "epoch": 3.1416906820365034,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004029048402191656,
      "loss": 4.7349,
      "mean_token_accuracy": 0.24458655714988708,
      "num_tokens": 74990072.0,
      "step": 32705
    },
    {
      "entropy": 5.3488694190979,
      "epoch": 3.1421709894332372,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004028766334207299,
      "loss": 4.6846,
      "mean_token_accuracy": 0.24857844710350036,
      "num_tokens": 75000784.0,
      "step": 32710
    },
    {
      "entropy": 5.356149291992187,
      "epoch": 3.142651296829971,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004028484236533784,
      "loss": 4.685,
      "mean_token_accuracy": 0.24746823608875274,
      "num_tokens": 75012709.0,
      "step": 32715
    },
    {
      "entropy": 5.328042268753052,
      "epoch": 3.143131604226705,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040282021091776624,
      "loss": 4.6407,
      "mean_token_accuracy": 0.25305141508579254,
      "num_tokens": 75023644.0,
      "step": 32720
    },
    {
      "entropy": 5.318274307250976,
      "epoch": 3.143611911623439,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004027919952145482,
      "loss": 4.6412,
      "mean_token_accuracy": 0.2545661434531212,
      "num_tokens": 75035706.0,
      "step": 32725
    },
    {
      "entropy": 5.403039836883545,
      "epoch": 3.1440922190201728,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004027637765443795,
      "loss": 4.677,
      "mean_token_accuracy": 0.2539741948246956,
      "num_tokens": 75046447.0,
      "step": 32730
    },
    {
      "entropy": 5.302552700042725,
      "epoch": 3.1445725264169067,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040273555490791534,
      "loss": 4.6494,
      "mean_token_accuracy": 0.25725999027490615,
      "num_tokens": 75058515.0,
      "step": 32735
    },
    {
      "entropy": 5.349905490875244,
      "epoch": 3.1450528338136405,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004027073303058109,
      "loss": 4.6947,
      "mean_token_accuracy": 0.25641320943832396,
      "num_tokens": 75068638.0,
      "step": 32740
    },
    {
      "entropy": 5.408388900756836,
      "epoch": 3.1455331412103744,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004026791027387214,
      "loss": 4.7155,
      "mean_token_accuracy": 0.24416519254446029,
      "num_tokens": 75080711.0,
      "step": 32745
    },
    {
      "entropy": 5.503248405456543,
      "epoch": 3.1460134486071087,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004026508722073024,
      "loss": 4.7615,
      "mean_token_accuracy": 0.24567244201898575,
      "num_tokens": 75090337.0,
      "step": 32750
    },
    {
      "entropy": 5.370638656616211,
      "epoch": 3.1464937560038426,
      "grad_norm": 1.125,
      "learning_rate": 0.00040262263871220904,
      "loss": 4.6603,
      "mean_token_accuracy": 0.24958355128765106,
      "num_tokens": 75101446.0,
      "step": 32755
    },
    {
      "entropy": 5.275532817840576,
      "epoch": 3.1469740634005765,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004025944022540971,
      "loss": 4.6286,
      "mean_token_accuracy": 0.25548815429210664,
      "num_tokens": 75112070.0,
      "step": 32760
    },
    {
      "entropy": 5.326457262039185,
      "epoch": 3.1474543707973104,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040256616283362195,
      "loss": 4.6532,
      "mean_token_accuracy": 0.25453677773475647,
      "num_tokens": 75123635.0,
      "step": 32765
    },
    {
      "entropy": 5.43135814666748,
      "epoch": 3.1479346781940443,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040253792045143926,
      "loss": 4.6964,
      "mean_token_accuracy": 0.2537906751036644,
      "num_tokens": 75135118.0,
      "step": 32770
    },
    {
      "entropy": 5.370361375808716,
      "epoch": 3.148414985590778,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004025096751082048,
      "loss": 4.774,
      "mean_token_accuracy": 0.24585041105747224,
      "num_tokens": 75147319.0,
      "step": 32775
    },
    {
      "entropy": 5.279457759857178,
      "epoch": 3.148895292987512,
      "grad_norm": 1.125,
      "learning_rate": 0.0004024814268045743,
      "loss": 4.5858,
      "mean_token_accuracy": 0.26091828793287275,
      "num_tokens": 75158185.0,
      "step": 32780
    },
    {
      "entropy": 5.342339706420899,
      "epoch": 3.149375600384246,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040245317554120363,
      "loss": 4.6011,
      "mean_token_accuracy": 0.25473618507385254,
      "num_tokens": 75169502.0,
      "step": 32785
    },
    {
      "entropy": 5.366493463516235,
      "epoch": 3.14985590778098,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004024249213187487,
      "loss": 4.7305,
      "mean_token_accuracy": 0.2434501364827156,
      "num_tokens": 75180567.0,
      "step": 32790
    },
    {
      "entropy": 5.363834381103516,
      "epoch": 3.1503362151777137,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004023966641378655,
      "loss": 4.6736,
      "mean_token_accuracy": 0.2506408542394638,
      "num_tokens": 75191622.0,
      "step": 32795
    },
    {
      "entropy": 5.349266004562378,
      "epoch": 3.1508165225744476,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040236840399920996,
      "loss": 4.6986,
      "mean_token_accuracy": 0.24437156617641448,
      "num_tokens": 75203306.0,
      "step": 32800
    },
    {
      "entropy": 5.350957536697388,
      "epoch": 3.1512968299711814,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040234014090343833,
      "loss": 4.6295,
      "mean_token_accuracy": 0.2516651898622513,
      "num_tokens": 75214471.0,
      "step": 32805
    },
    {
      "entropy": 5.388721466064453,
      "epoch": 3.1517771373679153,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004023118748512068,
      "loss": 4.7109,
      "mean_token_accuracy": 0.25103927999734876,
      "num_tokens": 75226657.0,
      "step": 32810
    },
    {
      "entropy": 5.30912070274353,
      "epoch": 3.152257444764649,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004022836058431715,
      "loss": 4.6513,
      "mean_token_accuracy": 0.25998983830213546,
      "num_tokens": 75237850.0,
      "step": 32815
    },
    {
      "entropy": 5.368501996994018,
      "epoch": 3.152737752161383,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040225533387998883,
      "loss": 4.6208,
      "mean_token_accuracy": 0.25855754166841505,
      "num_tokens": 75248604.0,
      "step": 32820
    },
    {
      "entropy": 5.381751489639282,
      "epoch": 3.1532180595581174,
      "grad_norm": 1.25,
      "learning_rate": 0.0004022270589623152,
      "loss": 4.6185,
      "mean_token_accuracy": 0.25447334349155426,
      "num_tokens": 75259745.0,
      "step": 32825
    },
    {
      "entropy": 5.408913946151733,
      "epoch": 3.1536983669548513,
      "grad_norm": 1.15625,
      "learning_rate": 0.000402198781090807,
      "loss": 4.8146,
      "mean_token_accuracy": 0.24053041636943817,
      "num_tokens": 75272421.0,
      "step": 32830
    },
    {
      "entropy": 5.3324010372161865,
      "epoch": 3.154178674351585,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004021705002661208,
      "loss": 4.7139,
      "mean_token_accuracy": 0.23920404314994811,
      "num_tokens": 75283753.0,
      "step": 32835
    },
    {
      "entropy": 5.338586139678955,
      "epoch": 3.154658981748319,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004021422164889133,
      "loss": 4.6578,
      "mean_token_accuracy": 0.25054369419813155,
      "num_tokens": 75294732.0,
      "step": 32840
    },
    {
      "entropy": 5.323082780838012,
      "epoch": 3.155139289145053,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000402113929759841,
      "loss": 4.6747,
      "mean_token_accuracy": 0.2543557345867157,
      "num_tokens": 75306707.0,
      "step": 32845
    },
    {
      "entropy": 5.289838171005249,
      "epoch": 3.155619596541787,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040208564007956075,
      "loss": 4.6416,
      "mean_token_accuracy": 0.253476545214653,
      "num_tokens": 75317663.0,
      "step": 32850
    },
    {
      "entropy": 5.346141052246094,
      "epoch": 3.1560999039385207,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004020573474487293,
      "loss": 4.6717,
      "mean_token_accuracy": 0.2522124156355858,
      "num_tokens": 75328700.0,
      "step": 32855
    },
    {
      "entropy": 5.3855119228363035,
      "epoch": 3.1565802113352546,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040202905186800347,
      "loss": 4.6666,
      "mean_token_accuracy": 0.2581630989909172,
      "num_tokens": 75340233.0,
      "step": 32860
    },
    {
      "entropy": 5.318682861328125,
      "epoch": 3.1570605187319885,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004020007533380403,
      "loss": 4.5928,
      "mean_token_accuracy": 0.26552000641822815,
      "num_tokens": 75351595.0,
      "step": 32865
    },
    {
      "entropy": 5.330242443084717,
      "epoch": 3.1575408261287223,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004019724518594967,
      "loss": 4.7102,
      "mean_token_accuracy": 0.2489134982228279,
      "num_tokens": 75362240.0,
      "step": 32870
    },
    {
      "entropy": 5.282033920288086,
      "epoch": 3.158021133525456,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004019441474330298,
      "loss": 4.6303,
      "mean_token_accuracy": 0.25688171237707136,
      "num_tokens": 75373210.0,
      "step": 32875
    },
    {
      "entropy": 5.356266450881958,
      "epoch": 3.15850144092219,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040191584005929684,
      "loss": 4.6968,
      "mean_token_accuracy": 0.2504238814115524,
      "num_tokens": 75386320.0,
      "step": 32880
    },
    {
      "entropy": 5.361700534820557,
      "epoch": 3.158981748318924,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004018875297389549,
      "loss": 4.691,
      "mean_token_accuracy": 0.25130207985639574,
      "num_tokens": 75397729.0,
      "step": 32885
    },
    {
      "entropy": 5.358537435531616,
      "epoch": 3.159462055715658,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040185921647266126,
      "loss": 4.634,
      "mean_token_accuracy": 0.2579734593629837,
      "num_tokens": 75408508.0,
      "step": 32890
    },
    {
      "entropy": 5.330240631103516,
      "epoch": 3.1599423631123917,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040183090026107326,
      "loss": 4.6639,
      "mean_token_accuracy": 0.25335782021284103,
      "num_tokens": 75419960.0,
      "step": 32895
    },
    {
      "entropy": 5.322566270828247,
      "epoch": 3.160422670509126,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040180258110484847,
      "loss": 4.7108,
      "mean_token_accuracy": 0.24879374653100966,
      "num_tokens": 75431115.0,
      "step": 32900
    },
    {
      "entropy": 5.322994756698608,
      "epoch": 3.16090297790586,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004017742590046442,
      "loss": 4.594,
      "mean_token_accuracy": 0.2671569034457207,
      "num_tokens": 75441917.0,
      "step": 32905
    },
    {
      "entropy": 5.332314920425415,
      "epoch": 3.161383285302594,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040174593396111814,
      "loss": 4.7118,
      "mean_token_accuracy": 0.24826150983572007,
      "num_tokens": 75452514.0,
      "step": 32910
    },
    {
      "entropy": 5.457635593414307,
      "epoch": 3.1618635926993277,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040171760597492785,
      "loss": 4.7665,
      "mean_token_accuracy": 0.24897977262735366,
      "num_tokens": 75463511.0,
      "step": 32915
    },
    {
      "entropy": 5.3609375,
      "epoch": 3.1623439000960616,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040168927504673094,
      "loss": 4.6615,
      "mean_token_accuracy": 0.24864879995584488,
      "num_tokens": 75474595.0,
      "step": 32920
    },
    {
      "entropy": 5.384627962112427,
      "epoch": 3.1628242074927955,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004016609411771853,
      "loss": 4.6914,
      "mean_token_accuracy": 0.24679675549268723,
      "num_tokens": 75485639.0,
      "step": 32925
    },
    {
      "entropy": 5.331307697296142,
      "epoch": 3.1633045148895294,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040163260436694876,
      "loss": 4.7016,
      "mean_token_accuracy": 0.24251284897327424,
      "num_tokens": 75497280.0,
      "step": 32930
    },
    {
      "entropy": 5.329108762741089,
      "epoch": 3.1637848222862632,
      "grad_norm": 1.125,
      "learning_rate": 0.0004016042646166791,
      "loss": 4.6888,
      "mean_token_accuracy": 0.25469848811626433,
      "num_tokens": 75509134.0,
      "step": 32935
    },
    {
      "entropy": 5.313772106170655,
      "epoch": 3.164265129682997,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004015759219270344,
      "loss": 4.6238,
      "mean_token_accuracy": 0.25215612053871156,
      "num_tokens": 75520166.0,
      "step": 32940
    },
    {
      "entropy": 5.299851322174073,
      "epoch": 3.164745437079731,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004015475762986726,
      "loss": 4.6376,
      "mean_token_accuracy": 0.24765732139348984,
      "num_tokens": 75530698.0,
      "step": 32945
    },
    {
      "entropy": 5.313333463668823,
      "epoch": 3.165225744476465,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040151922773225187,
      "loss": 4.7218,
      "mean_token_accuracy": 0.2433375522494316,
      "num_tokens": 75542232.0,
      "step": 32950
    },
    {
      "entropy": 5.291018152236939,
      "epoch": 3.1657060518731988,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004014908762284303,
      "loss": 4.6369,
      "mean_token_accuracy": 0.2607827588915825,
      "num_tokens": 75553650.0,
      "step": 32955
    },
    {
      "entropy": 5.336763858795166,
      "epoch": 3.1661863592699326,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040146252178786633,
      "loss": 4.6587,
      "mean_token_accuracy": 0.2569544807076454,
      "num_tokens": 75565877.0,
      "step": 32960
    },
    {
      "entropy": 5.308238792419433,
      "epoch": 3.1666666666666665,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000401434164411218,
      "loss": 4.5419,
      "mean_token_accuracy": 0.2651386171579361,
      "num_tokens": 75578294.0,
      "step": 32965
    },
    {
      "entropy": 5.35093584060669,
      "epoch": 3.1671469740634004,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040140580409914385,
      "loss": 4.6992,
      "mean_token_accuracy": 0.2500303864479065,
      "num_tokens": 75589228.0,
      "step": 32970
    },
    {
      "entropy": 5.333810758590698,
      "epoch": 3.1676272814601343,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040137744085230227,
      "loss": 4.712,
      "mean_token_accuracy": 0.24860430806875228,
      "num_tokens": 75600443.0,
      "step": 32975
    },
    {
      "entropy": 5.377885103225708,
      "epoch": 3.168107588856868,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004013490746713518,
      "loss": 4.6399,
      "mean_token_accuracy": 0.25004616379737854,
      "num_tokens": 75612053.0,
      "step": 32980
    },
    {
      "entropy": 5.416900300979615,
      "epoch": 3.1685878962536025,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040132070555695096,
      "loss": 4.7016,
      "mean_token_accuracy": 0.24298408329486848,
      "num_tokens": 75624084.0,
      "step": 32985
    },
    {
      "entropy": 5.419850301742554,
      "epoch": 3.1690682036503364,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040129233350975847,
      "loss": 4.7143,
      "mean_token_accuracy": 0.24567925035953522,
      "num_tokens": 75634863.0,
      "step": 32990
    },
    {
      "entropy": 5.238531637191772,
      "epoch": 3.1695485110470702,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040126395853043293,
      "loss": 4.5389,
      "mean_token_accuracy": 0.26300257742404937,
      "num_tokens": 75645194.0,
      "step": 32995
    },
    {
      "entropy": 5.277801942825318,
      "epoch": 3.170028818443804,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004012355806196332,
      "loss": 4.6203,
      "mean_token_accuracy": 0.26074028164148333,
      "num_tokens": 75657444.0,
      "step": 33000
    },
    {
      "epoch": 3.170028818443804,
      "eval_entropy": 5.095041842067724,
      "eval_loss": 4.783809185028076,
      "eval_mean_token_accuracy": 0.2553269653515241,
      "eval_num_tokens": 75657444.0,
      "eval_runtime": 20.5873,
      "eval_samples_per_second": 1593.947,
      "eval_steps_per_second": 199.249,
      "step": 33000
    }
  ],
  "logging_steps": 5,
  "max_steps": 104090,
  "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": 1.716119487811584e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}