{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0023387905111927,
  "eval_steps": 3000,
  "global_step": 12000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 4.944228982925415,
      "epoch": 0.000417641162712997,
      "grad_norm": 14.75,
      "learning_rate": 2e-06,
      "loss": 14.2398,
      "mean_token_accuracy": 0.0,
      "num_tokens": 9371.0,
      "step": 5
    },
    {
      "entropy": 4.94694414138794,
      "epoch": 0.000835282325425994,
      "grad_norm": 14.8125,
      "learning_rate": 4.5e-06,
      "loss": 14.1961,
      "mean_token_accuracy": 0.0,
      "num_tokens": 18155.0,
      "step": 10
    },
    {
      "entropy": 4.961578607559204,
      "epoch": 0.001252923488138991,
      "grad_norm": 15.625,
      "learning_rate": 7e-06,
      "loss": 14.1818,
      "mean_token_accuracy": 9.380863048136234e-05,
      "num_tokens": 27776.0,
      "step": 15
    },
    {
      "entropy": 4.998337459564209,
      "epoch": 0.001670564650851988,
      "grad_norm": 17.5,
      "learning_rate": 9.5e-06,
      "loss": 13.9811,
      "mean_token_accuracy": 0.0,
      "num_tokens": 37166.0,
      "step": 20
    },
    {
      "entropy": 5.118108034133911,
      "epoch": 0.002088205813564985,
      "grad_norm": 22.125,
      "learning_rate": 1.2e-05,
      "loss": 13.6488,
      "mean_token_accuracy": 0.0,
      "num_tokens": 46667.0,
      "step": 25
    },
    {
      "entropy": 5.413705348968506,
      "epoch": 0.002505846976277982,
      "grad_norm": 27.5,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 13.0874,
      "mean_token_accuracy": 0.0,
      "num_tokens": 55738.0,
      "step": 30
    },
    {
      "entropy": 6.7129909038543705,
      "epoch": 0.002923488138990979,
      "grad_norm": 29.25,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 12.0634,
      "mean_token_accuracy": 0.0,
      "num_tokens": 65303.0,
      "step": 35
    },
    {
      "entropy": 9.943841934204102,
      "epoch": 0.003341129301703976,
      "grad_norm": 3.921875,
      "learning_rate": 1.95e-05,
      "loss": 10.9276,
      "mean_token_accuracy": 0.00012578616151586174,
      "num_tokens": 74876.0,
      "step": 40
    },
    {
      "entropy": 10.703345203399659,
      "epoch": 0.003758770464416973,
      "grad_norm": 3.21875,
      "learning_rate": 2.2e-05,
      "loss": 10.6802,
      "mean_token_accuracy": 0.002392532047815621,
      "num_tokens": 84683.0,
      "step": 45
    },
    {
      "entropy": 10.740305995941162,
      "epoch": 0.00417641162712997,
      "grad_norm": 3.0,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 10.5254,
      "mean_token_accuracy": 0.013313664309680461,
      "num_tokens": 94511.0,
      "step": 50
    },
    {
      "entropy": 10.734064483642578,
      "epoch": 0.004594052789842967,
      "grad_norm": 2.15625,
      "learning_rate": 2.7e-05,
      "loss": 10.2867,
      "mean_token_accuracy": 0.018075952026993036,
      "num_tokens": 103820.0,
      "step": 55
    },
    {
      "entropy": 10.681211948394775,
      "epoch": 0.005011693952555964,
      "grad_norm": 2.65625,
      "learning_rate": 2.95e-05,
      "loss": 10.1047,
      "mean_token_accuracy": 0.019883409794420003,
      "num_tokens": 112398.0,
      "step": 60
    },
    {
      "entropy": 10.702549839019776,
      "epoch": 0.005429335115268961,
      "grad_norm": 2.015625,
      "learning_rate": 3.2e-05,
      "loss": 9.9646,
      "mean_token_accuracy": 0.01945118773728609,
      "num_tokens": 121167.0,
      "step": 65
    },
    {
      "entropy": 10.730839729309082,
      "epoch": 0.005846976277981958,
      "grad_norm": 2.0,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 9.8414,
      "mean_token_accuracy": 0.025720258615911006,
      "num_tokens": 129399.0,
      "step": 70
    },
    {
      "entropy": 10.728861236572266,
      "epoch": 0.006264617440694955,
      "grad_norm": 1.96875,
      "learning_rate": 3.7e-05,
      "loss": 9.7822,
      "mean_token_accuracy": 0.032033701054751874,
      "num_tokens": 138512.0,
      "step": 75
    },
    {
      "entropy": 10.722623062133788,
      "epoch": 0.006682258603407952,
      "grad_norm": 1.78125,
      "learning_rate": 3.95e-05,
      "loss": 9.7341,
      "mean_token_accuracy": 0.032669741101562975,
      "num_tokens": 147712.0,
      "step": 80
    },
    {
      "entropy": 10.71749963760376,
      "epoch": 0.007099899766120949,
      "grad_norm": 1.8515625,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 9.6255,
      "mean_token_accuracy": 0.03035105634480715,
      "num_tokens": 156759.0,
      "step": 85
    },
    {
      "entropy": 10.71173620223999,
      "epoch": 0.007517540928833946,
      "grad_norm": 1.9140625,
      "learning_rate": 4.45e-05,
      "loss": 9.5187,
      "mean_token_accuracy": 0.035255284421145915,
      "num_tokens": 166395.0,
      "step": 90
    },
    {
      "entropy": 10.703879737854004,
      "epoch": 0.007935182091546943,
      "grad_norm": 1.9140625,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 9.3862,
      "mean_token_accuracy": 0.044158879294991495,
      "num_tokens": 174739.0,
      "step": 95
    },
    {
      "entropy": 10.680936431884765,
      "epoch": 0.00835282325425994,
      "grad_norm": 1.8125,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 9.2512,
      "mean_token_accuracy": 0.03875316996127367,
      "num_tokens": 183010.0,
      "step": 100
    },
    {
      "entropy": 10.65852222442627,
      "epoch": 0.008770464416972937,
      "grad_norm": 1.796875,
      "learning_rate": 5.2e-05,
      "loss": 9.2045,
      "mean_token_accuracy": 0.040588128939270975,
      "num_tokens": 192589.0,
      "step": 105
    },
    {
      "entropy": 10.639169883728027,
      "epoch": 0.009188105579685933,
      "grad_norm": 1.8359375,
      "learning_rate": 5.45e-05,
      "loss": 9.0572,
      "mean_token_accuracy": 0.0537865225225687,
      "num_tokens": 201638.0,
      "step": 110
    },
    {
      "entropy": 10.59452133178711,
      "epoch": 0.00960574674239893,
      "grad_norm": 1.8359375,
      "learning_rate": 5.7e-05,
      "loss": 8.9186,
      "mean_token_accuracy": 0.05064392797648907,
      "num_tokens": 209862.0,
      "step": 115
    },
    {
      "entropy": 10.544370651245117,
      "epoch": 0.010023387905111928,
      "grad_norm": 1.6640625,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 8.8996,
      "mean_token_accuracy": 0.043613025173544887,
      "num_tokens": 218897.0,
      "step": 120
    },
    {
      "entropy": 10.49176197052002,
      "epoch": 0.010441029067824925,
      "grad_norm": 1.6484375,
      "learning_rate": 6.2e-05,
      "loss": 8.7626,
      "mean_token_accuracy": 0.05363074168562889,
      "num_tokens": 227682.0,
      "step": 125
    },
    {
      "entropy": 10.402692413330078,
      "epoch": 0.010858670230537921,
      "grad_norm": 1.7421875,
      "learning_rate": 6.450000000000001e-05,
      "loss": 8.586,
      "mean_token_accuracy": 0.053992220759391786,
      "num_tokens": 236170.0,
      "step": 130
    },
    {
      "entropy": 10.307563400268554,
      "epoch": 0.011276311393250919,
      "grad_norm": 1.4921875,
      "learning_rate": 6.7e-05,
      "loss": 8.5381,
      "mean_token_accuracy": 0.0543564785271883,
      "num_tokens": 245710.0,
      "step": 135
    },
    {
      "entropy": 10.217035388946533,
      "epoch": 0.011693952555963916,
      "grad_norm": 1.7109375,
      "learning_rate": 6.950000000000001e-05,
      "loss": 8.368,
      "mean_token_accuracy": 0.05643460154533386,
      "num_tokens": 254751.0,
      "step": 140
    },
    {
      "entropy": 10.128512001037597,
      "epoch": 0.012111593718676913,
      "grad_norm": 1.53125,
      "learning_rate": 7.2e-05,
      "loss": 8.241,
      "mean_token_accuracy": 0.05597202591598034,
      "num_tokens": 264210.0,
      "step": 145
    },
    {
      "entropy": 10.00536880493164,
      "epoch": 0.01252923488138991,
      "grad_norm": 1.5078125,
      "learning_rate": 7.45e-05,
      "loss": 8.1832,
      "mean_token_accuracy": 0.0563384834676981,
      "num_tokens": 272944.0,
      "step": 150
    },
    {
      "entropy": 9.889774703979493,
      "epoch": 0.012946876044102907,
      "grad_norm": 1.609375,
      "learning_rate": 7.7e-05,
      "loss": 8.0239,
      "mean_token_accuracy": 0.055541551858186725,
      "num_tokens": 281311.0,
      "step": 155
    },
    {
      "entropy": 9.758612728118896,
      "epoch": 0.013364517206815904,
      "grad_norm": 1.390625,
      "learning_rate": 7.950000000000001e-05,
      "loss": 7.9127,
      "mean_token_accuracy": 0.059244291856884956,
      "num_tokens": 289701.0,
      "step": 160
    },
    {
      "entropy": 9.522632217407226,
      "epoch": 0.013782158369528901,
      "grad_norm": 1.15625,
      "learning_rate": 8.2e-05,
      "loss": 7.9089,
      "mean_token_accuracy": 0.05390802845358848,
      "num_tokens": 299459.0,
      "step": 165
    },
    {
      "entropy": 9.387851905822753,
      "epoch": 0.014199799532241897,
      "grad_norm": 1.1484375,
      "learning_rate": 8.450000000000001e-05,
      "loss": 7.8314,
      "mean_token_accuracy": 0.05658908933401108,
      "num_tokens": 309343.0,
      "step": 170
    },
    {
      "entropy": 9.159295749664306,
      "epoch": 0.014617440694954895,
      "grad_norm": 0.9921875,
      "learning_rate": 8.7e-05,
      "loss": 7.7869,
      "mean_token_accuracy": 0.0535435926169157,
      "num_tokens": 318840.0,
      "step": 175
    },
    {
      "entropy": 8.917513465881347,
      "epoch": 0.015035081857667892,
      "grad_norm": 0.8671875,
      "learning_rate": 8.95e-05,
      "loss": 7.5732,
      "mean_token_accuracy": 0.056347102671861646,
      "num_tokens": 328689.0,
      "step": 180
    },
    {
      "entropy": 8.685413837432861,
      "epoch": 0.01545272302038089,
      "grad_norm": 0.9375,
      "learning_rate": 9.2e-05,
      "loss": 7.658,
      "mean_token_accuracy": 0.05352019853889942,
      "num_tokens": 338209.0,
      "step": 185
    },
    {
      "entropy": 8.488839721679687,
      "epoch": 0.015870364183093885,
      "grad_norm": 0.796875,
      "learning_rate": 9.45e-05,
      "loss": 7.4968,
      "mean_token_accuracy": 0.056180522218346594,
      "num_tokens": 347270.0,
      "step": 190
    },
    {
      "entropy": 8.256251239776612,
      "epoch": 0.016288005345806884,
      "grad_norm": 1.40625,
      "learning_rate": 9.7e-05,
      "loss": 7.5217,
      "mean_token_accuracy": 0.057517141476273535,
      "num_tokens": 356366.0,
      "step": 195
    },
    {
      "entropy": 8.176424980163574,
      "epoch": 0.01670564650851988,
      "grad_norm": 0.96484375,
      "learning_rate": 9.95e-05,
      "loss": 7.4756,
      "mean_token_accuracy": 0.059065777063369754,
      "num_tokens": 365980.0,
      "step": 200
    },
    {
      "entropy": 8.109168100357056,
      "epoch": 0.017123287671232876,
      "grad_norm": 1.09375,
      "learning_rate": 0.000102,
      "loss": 7.4352,
      "mean_token_accuracy": 0.06301593109965324,
      "num_tokens": 374591.0,
      "step": 205
    },
    {
      "entropy": 8.09407901763916,
      "epoch": 0.017540928833945875,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00010449999999999999,
      "loss": 7.4288,
      "mean_token_accuracy": 0.06486969366669655,
      "num_tokens": 383250.0,
      "step": 210
    },
    {
      "entropy": 8.025069379806519,
      "epoch": 0.01795856999665887,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000107,
      "loss": 7.3621,
      "mean_token_accuracy": 0.06629893258213997,
      "num_tokens": 391502.0,
      "step": 215
    },
    {
      "entropy": 7.9847424030303955,
      "epoch": 0.018376211159371866,
      "grad_norm": 1.390625,
      "learning_rate": 0.0001095,
      "loss": 7.2782,
      "mean_token_accuracy": 0.06931861937046051,
      "num_tokens": 400225.0,
      "step": 220
    },
    {
      "entropy": 8.029051160812378,
      "epoch": 0.018793852322084866,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000112,
      "loss": 7.4779,
      "mean_token_accuracy": 0.060866087675094604,
      "num_tokens": 410008.0,
      "step": 225
    },
    {
      "entropy": 8.031083345413208,
      "epoch": 0.01921149348479786,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0001145,
      "loss": 7.381,
      "mean_token_accuracy": 0.0702624622732401,
      "num_tokens": 418979.0,
      "step": 230
    },
    {
      "entropy": 7.950044965744018,
      "epoch": 0.01962913464751086,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00011700000000000001,
      "loss": 7.349,
      "mean_token_accuracy": 0.06797430105507374,
      "num_tokens": 428139.0,
      "step": 235
    },
    {
      "entropy": 7.96306037902832,
      "epoch": 0.020046775810223856,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00011949999999999999,
      "loss": 7.2651,
      "mean_token_accuracy": 0.07591628283262253,
      "num_tokens": 437113.0,
      "step": 240
    },
    {
      "entropy": 7.929630136489868,
      "epoch": 0.020464416972936852,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000122,
      "loss": 7.4462,
      "mean_token_accuracy": 0.06979386620223522,
      "num_tokens": 446879.0,
      "step": 245
    },
    {
      "entropy": 7.925119638442993,
      "epoch": 0.02088205813564985,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0001245,
      "loss": 7.33,
      "mean_token_accuracy": 0.07119030728936196,
      "num_tokens": 456362.0,
      "step": 250
    },
    {
      "entropy": 7.896007394790649,
      "epoch": 0.021299699298362847,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000127,
      "loss": 7.157,
      "mean_token_accuracy": 0.07790606319904328,
      "num_tokens": 464930.0,
      "step": 255
    },
    {
      "entropy": 7.96244444847107,
      "epoch": 0.021717340461075842,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0001295,
      "loss": 7.3203,
      "mean_token_accuracy": 0.07410374954342842,
      "num_tokens": 474303.0,
      "step": 260
    },
    {
      "entropy": 7.917203140258789,
      "epoch": 0.02213498162378884,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000132,
      "loss": 7.2993,
      "mean_token_accuracy": 0.0734116319566965,
      "num_tokens": 483750.0,
      "step": 265
    },
    {
      "entropy": 7.86039023399353,
      "epoch": 0.022552622786501837,
      "grad_norm": 1.234375,
      "learning_rate": 0.00013450000000000002,
      "loss": 7.199,
      "mean_token_accuracy": 0.07768494263291359,
      "num_tokens": 493325.0,
      "step": 270
    },
    {
      "entropy": 7.840238094329834,
      "epoch": 0.022970263949214833,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00013700000000000002,
      "loss": 7.1762,
      "mean_token_accuracy": 0.07519763223826885,
      "num_tokens": 501907.0,
      "step": 275
    },
    {
      "entropy": 7.810028219223023,
      "epoch": 0.023387905111927832,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0001395,
      "loss": 7.1942,
      "mean_token_accuracy": 0.0795279961079359,
      "num_tokens": 511522.0,
      "step": 280
    },
    {
      "entropy": 7.749888896942139,
      "epoch": 0.023805546274640828,
      "grad_norm": 0.890625,
      "learning_rate": 0.00014199999999999998,
      "loss": 7.1636,
      "mean_token_accuracy": 0.07597449868917465,
      "num_tokens": 521287.0,
      "step": 285
    },
    {
      "entropy": 7.849759721755982,
      "epoch": 0.024223187437353827,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0001445,
      "loss": 7.2325,
      "mean_token_accuracy": 0.07516792006790637,
      "num_tokens": 530631.0,
      "step": 290
    },
    {
      "entropy": 7.650192070007324,
      "epoch": 0.024640828600066823,
      "grad_norm": 1.046875,
      "learning_rate": 0.000147,
      "loss": 7.1083,
      "mean_token_accuracy": 0.08667415156960487,
      "num_tokens": 540386.0,
      "step": 295
    },
    {
      "entropy": 7.79657678604126,
      "epoch": 0.02505846976277982,
      "grad_norm": 1.125,
      "learning_rate": 0.0001495,
      "loss": 7.191,
      "mean_token_accuracy": 0.07905804216861725,
      "num_tokens": 550121.0,
      "step": 300
    },
    {
      "entropy": 7.630661725997925,
      "epoch": 0.025476110925492818,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000152,
      "loss": 7.048,
      "mean_token_accuracy": 0.08168927729129791,
      "num_tokens": 559460.0,
      "step": 305
    },
    {
      "entropy": 7.653521871566772,
      "epoch": 0.025893752088205813,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00015450000000000001,
      "loss": 7.0115,
      "mean_token_accuracy": 0.08255561292171479,
      "num_tokens": 568392.0,
      "step": 310
    },
    {
      "entropy": 7.603034496307373,
      "epoch": 0.02631139325091881,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000157,
      "loss": 7.0908,
      "mean_token_accuracy": 0.07983561903238297,
      "num_tokens": 578248.0,
      "step": 315
    },
    {
      "entropy": 7.682858037948608,
      "epoch": 0.026729034413631808,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0001595,
      "loss": 7.157,
      "mean_token_accuracy": 0.07829831764101983,
      "num_tokens": 588164.0,
      "step": 320
    },
    {
      "entropy": 7.5797460079193115,
      "epoch": 0.027146675576344804,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000162,
      "loss": 7.0789,
      "mean_token_accuracy": 0.08381804451346397,
      "num_tokens": 598715.0,
      "step": 325
    },
    {
      "entropy": 7.650176906585694,
      "epoch": 0.027564316739057803,
      "grad_norm": 1.09375,
      "learning_rate": 0.00016450000000000001,
      "loss": 7.0161,
      "mean_token_accuracy": 0.09236654937267304,
      "num_tokens": 606939.0,
      "step": 330
    },
    {
      "entropy": 7.4912495613098145,
      "epoch": 0.0279819579017708,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00016700000000000002,
      "loss": 7.1029,
      "mean_token_accuracy": 0.0873373556882143,
      "num_tokens": 615803.0,
      "step": 335
    },
    {
      "entropy": 7.667604541778564,
      "epoch": 0.028399599064483794,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00016950000000000003,
      "loss": 6.9664,
      "mean_token_accuracy": 0.08756924346089363,
      "num_tokens": 624730.0,
      "step": 340
    },
    {
      "entropy": 7.603380632400513,
      "epoch": 0.028817240227196794,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00017199999999999998,
      "loss": 7.0542,
      "mean_token_accuracy": 0.08866709396243096,
      "num_tokens": 634082.0,
      "step": 345
    },
    {
      "entropy": 7.52916407585144,
      "epoch": 0.02923488138990979,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00017449999999999999,
      "loss": 7.0332,
      "mean_token_accuracy": 0.08419890776276588,
      "num_tokens": 643367.0,
      "step": 350
    },
    {
      "entropy": 7.499692583084107,
      "epoch": 0.029652522552622785,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000177,
      "loss": 6.8535,
      "mean_token_accuracy": 0.08740218952298165,
      "num_tokens": 651861.0,
      "step": 355
    },
    {
      "entropy": 7.492964124679565,
      "epoch": 0.030070163715335784,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0001795,
      "loss": 6.9752,
      "mean_token_accuracy": 0.08579764813184738,
      "num_tokens": 661064.0,
      "step": 360
    },
    {
      "entropy": 7.502609062194824,
      "epoch": 0.03048780487804878,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000182,
      "loss": 6.9488,
      "mean_token_accuracy": 0.0932283952832222,
      "num_tokens": 669299.0,
      "step": 365
    },
    {
      "entropy": 7.401155567169189,
      "epoch": 0.03090544604076178,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0001845,
      "loss": 7.0171,
      "mean_token_accuracy": 0.08307098634541035,
      "num_tokens": 678631.0,
      "step": 370
    },
    {
      "entropy": 7.472736454010009,
      "epoch": 0.03132308720347477,
      "grad_norm": 0.875,
      "learning_rate": 0.000187,
      "loss": 7.0131,
      "mean_token_accuracy": 0.08543515279889106,
      "num_tokens": 688511.0,
      "step": 375
    },
    {
      "entropy": 7.527131605148315,
      "epoch": 0.03174072836618777,
      "grad_norm": 1.109375,
      "learning_rate": 0.0001895,
      "loss": 6.9657,
      "mean_token_accuracy": 0.08108578622341156,
      "num_tokens": 697956.0,
      "step": 380
    },
    {
      "entropy": 7.438411712646484,
      "epoch": 0.03215836952890077,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000192,
      "loss": 6.8724,
      "mean_token_accuracy": 0.08787695616483689,
      "num_tokens": 706779.0,
      "step": 385
    },
    {
      "entropy": 7.325165557861328,
      "epoch": 0.03257601069161377,
      "grad_norm": 1.0,
      "learning_rate": 0.0001945,
      "loss": 6.8617,
      "mean_token_accuracy": 0.09062610566616058,
      "num_tokens": 715749.0,
      "step": 390
    },
    {
      "entropy": 7.425607204437256,
      "epoch": 0.03299365185432676,
      "grad_norm": 1.8046875,
      "learning_rate": 0.00019700000000000002,
      "loss": 6.8375,
      "mean_token_accuracy": 0.0908269003033638,
      "num_tokens": 724756.0,
      "step": 395
    },
    {
      "entropy": 7.373869371414185,
      "epoch": 0.03341129301703976,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00019950000000000002,
      "loss": 6.8659,
      "mean_token_accuracy": 0.09110177978873253,
      "num_tokens": 733288.0,
      "step": 400
    },
    {
      "entropy": 7.444917821884156,
      "epoch": 0.03382893417975276,
      "grad_norm": 1.765625,
      "learning_rate": 0.000202,
      "loss": 6.9685,
      "mean_token_accuracy": 0.08536023795604705,
      "num_tokens": 741629.0,
      "step": 405
    },
    {
      "entropy": 7.3436695575714115,
      "epoch": 0.03424657534246575,
      "grad_norm": 1.265625,
      "learning_rate": 0.00020449999999999998,
      "loss": 6.9068,
      "mean_token_accuracy": 0.08802182823419571,
      "num_tokens": 751336.0,
      "step": 410
    },
    {
      "entropy": 7.426488780975342,
      "epoch": 0.03466421650517875,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000207,
      "loss": 6.9733,
      "mean_token_accuracy": 0.08576110526919364,
      "num_tokens": 759725.0,
      "step": 415
    },
    {
      "entropy": 7.417571353912353,
      "epoch": 0.03508185766789175,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002095,
      "loss": 6.9582,
      "mean_token_accuracy": 0.08887427300214767,
      "num_tokens": 769752.0,
      "step": 420
    },
    {
      "entropy": 7.324219417572022,
      "epoch": 0.03549949883060474,
      "grad_norm": 1.03125,
      "learning_rate": 0.000212,
      "loss": 6.8629,
      "mean_token_accuracy": 0.08838045299053192,
      "num_tokens": 779124.0,
      "step": 425
    },
    {
      "entropy": 7.344010305404663,
      "epoch": 0.03591713999331774,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002145,
      "loss": 6.8149,
      "mean_token_accuracy": 0.09638451412320137,
      "num_tokens": 788892.0,
      "step": 430
    },
    {
      "entropy": 7.214268350601197,
      "epoch": 0.03633478115603074,
      "grad_norm": 1.046875,
      "learning_rate": 0.00021700000000000002,
      "loss": 6.7784,
      "mean_token_accuracy": 0.09481111764907837,
      "num_tokens": 798265.0,
      "step": 435
    },
    {
      "entropy": 7.317943811416626,
      "epoch": 0.03675242231874373,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002195,
      "loss": 6.7438,
      "mean_token_accuracy": 0.09948835447430611,
      "num_tokens": 806896.0,
      "step": 440
    },
    {
      "entropy": 7.3143962860107425,
      "epoch": 0.03717006348145673,
      "grad_norm": 1.171875,
      "learning_rate": 0.000222,
      "loss": 6.8058,
      "mean_token_accuracy": 0.09371909499168396,
      "num_tokens": 815540.0,
      "step": 445
    },
    {
      "entropy": 7.221705627441406,
      "epoch": 0.03758770464416973,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002245,
      "loss": 6.8558,
      "mean_token_accuracy": 0.08694771379232406,
      "num_tokens": 824555.0,
      "step": 450
    },
    {
      "entropy": 7.346810579299927,
      "epoch": 0.03800534580688272,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00022700000000000002,
      "loss": 6.8838,
      "mean_token_accuracy": 0.09503390341997146,
      "num_tokens": 833414.0,
      "step": 455
    },
    {
      "entropy": 7.228864622116089,
      "epoch": 0.03842298696959572,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00022950000000000002,
      "loss": 6.745,
      "mean_token_accuracy": 0.09891379103064538,
      "num_tokens": 842644.0,
      "step": 460
    },
    {
      "entropy": 7.261135673522949,
      "epoch": 0.03884062813230872,
      "grad_norm": 1.15625,
      "learning_rate": 0.00023200000000000003,
      "loss": 6.7796,
      "mean_token_accuracy": 0.09051957502961158,
      "num_tokens": 851786.0,
      "step": 465
    },
    {
      "entropy": 7.189091730117798,
      "epoch": 0.03925826929502172,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00023449999999999998,
      "loss": 6.8274,
      "mean_token_accuracy": 0.09392543509602547,
      "num_tokens": 861508.0,
      "step": 470
    },
    {
      "entropy": 7.204404687881469,
      "epoch": 0.03967591045773471,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000237,
      "loss": 6.7046,
      "mean_token_accuracy": 0.09566405564546585,
      "num_tokens": 870433.0,
      "step": 475
    },
    {
      "entropy": 7.245524454116821,
      "epoch": 0.04009355162044771,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002395,
      "loss": 6.8033,
      "mean_token_accuracy": 0.09623450264334679,
      "num_tokens": 879059.0,
      "step": 480
    },
    {
      "entropy": 7.304554224014282,
      "epoch": 0.04051119278316071,
      "grad_norm": 1.140625,
      "learning_rate": 0.000242,
      "loss": 6.8465,
      "mean_token_accuracy": 0.09020947441458702,
      "num_tokens": 888009.0,
      "step": 485
    },
    {
      "entropy": 7.111552095413208,
      "epoch": 0.040928833945873704,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002445,
      "loss": 6.7091,
      "mean_token_accuracy": 0.09530647322535515,
      "num_tokens": 897335.0,
      "step": 490
    },
    {
      "entropy": 7.2552978515625,
      "epoch": 0.0413464751085867,
      "grad_norm": 1.21875,
      "learning_rate": 0.000247,
      "loss": 6.7299,
      "mean_token_accuracy": 0.09859334379434585,
      "num_tokens": 905802.0,
      "step": 495
    },
    {
      "entropy": 7.216646766662597,
      "epoch": 0.0417641162712997,
      "grad_norm": 1.578125,
      "learning_rate": 0.0002495,
      "loss": 6.7167,
      "mean_token_accuracy": 0.09858945608139039,
      "num_tokens": 914836.0,
      "step": 500
    },
    {
      "entropy": 7.100709247589111,
      "epoch": 0.042181757434012694,
      "grad_norm": 1.734375,
      "learning_rate": 0.000252,
      "loss": 6.6645,
      "mean_token_accuracy": 0.09969677850604057,
      "num_tokens": 924254.0,
      "step": 505
    },
    {
      "entropy": 7.183725643157959,
      "epoch": 0.04259939859672569,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002545,
      "loss": 6.7576,
      "mean_token_accuracy": 0.09866931736469269,
      "num_tokens": 933841.0,
      "step": 510
    },
    {
      "entropy": 7.210815715789795,
      "epoch": 0.04301703975943869,
      "grad_norm": 1.265625,
      "learning_rate": 0.000257,
      "loss": 6.8149,
      "mean_token_accuracy": 0.0937366984784603,
      "num_tokens": 943304.0,
      "step": 515
    },
    {
      "entropy": 7.095678329467773,
      "epoch": 0.043434680922151685,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002595,
      "loss": 6.6426,
      "mean_token_accuracy": 0.09861454665660858,
      "num_tokens": 952915.0,
      "step": 520
    },
    {
      "entropy": 7.156992769241333,
      "epoch": 0.043852322084864684,
      "grad_norm": 1.09375,
      "learning_rate": 0.000262,
      "loss": 6.6932,
      "mean_token_accuracy": 0.09782055467367172,
      "num_tokens": 962752.0,
      "step": 525
    },
    {
      "entropy": 7.15832552909851,
      "epoch": 0.04426996324757768,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00026450000000000003,
      "loss": 6.7124,
      "mean_token_accuracy": 0.09217726364731789,
      "num_tokens": 971240.0,
      "step": 530
    },
    {
      "entropy": 7.168659973144531,
      "epoch": 0.044687604410290675,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00026700000000000004,
      "loss": 6.7519,
      "mean_token_accuracy": 0.09481079205870628,
      "num_tokens": 981391.0,
      "step": 535
    },
    {
      "entropy": 7.195070505142212,
      "epoch": 0.045105245573003674,
      "grad_norm": 0.984375,
      "learning_rate": 0.00026950000000000005,
      "loss": 6.7498,
      "mean_token_accuracy": 0.09649497717618942,
      "num_tokens": 990551.0,
      "step": 540
    },
    {
      "entropy": 7.086696815490723,
      "epoch": 0.045522886735716674,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00027200000000000005,
      "loss": 6.6753,
      "mean_token_accuracy": 0.10147869288921356,
      "num_tokens": 999593.0,
      "step": 545
    },
    {
      "entropy": 7.119489288330078,
      "epoch": 0.045940527898429666,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002745,
      "loss": 6.7394,
      "mean_token_accuracy": 0.09457549154758453,
      "num_tokens": 1008785.0,
      "step": 550
    },
    {
      "entropy": 7.081824779510498,
      "epoch": 0.046358169061142665,
      "grad_norm": 1.1875,
      "learning_rate": 0.000277,
      "loss": 6.7256,
      "mean_token_accuracy": 0.0977080836892128,
      "num_tokens": 1016815.0,
      "step": 555
    },
    {
      "entropy": 7.124262952804566,
      "epoch": 0.046775810223855664,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002795,
      "loss": 6.6273,
      "mean_token_accuracy": 0.09984842911362649,
      "num_tokens": 1026182.0,
      "step": 560
    },
    {
      "entropy": 7.061550188064575,
      "epoch": 0.04719345138656866,
      "grad_norm": 1.59375,
      "learning_rate": 0.00028199999999999997,
      "loss": 6.5498,
      "mean_token_accuracy": 0.10488305762410163,
      "num_tokens": 1034032.0,
      "step": 565
    },
    {
      "entropy": 7.058889532089234,
      "epoch": 0.047611092549281656,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002845,
      "loss": 6.7444,
      "mean_token_accuracy": 0.09858610928058624,
      "num_tokens": 1043493.0,
      "step": 570
    },
    {
      "entropy": 7.081225299835205,
      "epoch": 0.048028733711994655,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000287,
      "loss": 6.6467,
      "mean_token_accuracy": 0.10488652363419533,
      "num_tokens": 1052989.0,
      "step": 575
    },
    {
      "entropy": 7.103153085708618,
      "epoch": 0.048446374874707654,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002895,
      "loss": 6.6193,
      "mean_token_accuracy": 0.100984375923872,
      "num_tokens": 1061803.0,
      "step": 580
    },
    {
      "entropy": 7.087803888320923,
      "epoch": 0.048864016037420646,
      "grad_norm": 1.203125,
      "learning_rate": 0.000292,
      "loss": 6.7207,
      "mean_token_accuracy": 0.09783147796988487,
      "num_tokens": 1071227.0,
      "step": 585
    },
    {
      "entropy": 7.093145227432251,
      "epoch": 0.049281657200133645,
      "grad_norm": 1.546875,
      "learning_rate": 0.0002945,
      "loss": 6.7154,
      "mean_token_accuracy": 0.10207850188016891,
      "num_tokens": 1080082.0,
      "step": 590
    },
    {
      "entropy": 6.990021800994873,
      "epoch": 0.049699298362846644,
      "grad_norm": 1.28125,
      "learning_rate": 0.000297,
      "loss": 6.5626,
      "mean_token_accuracy": 0.10718206688761711,
      "num_tokens": 1089750.0,
      "step": 595
    },
    {
      "entropy": 6.936739349365235,
      "epoch": 0.05011693952555964,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002995,
      "loss": 6.5546,
      "mean_token_accuracy": 0.10501954108476638,
      "num_tokens": 1098691.0,
      "step": 600
    },
    {
      "entropy": 7.053757810592652,
      "epoch": 0.050534580688272636,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000302,
      "loss": 6.5961,
      "mean_token_accuracy": 0.10269566178321839,
      "num_tokens": 1107602.0,
      "step": 605
    },
    {
      "entropy": 6.988081264495849,
      "epoch": 0.050952221850985635,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003045,
      "loss": 6.5094,
      "mean_token_accuracy": 0.10492742285132409,
      "num_tokens": 1116426.0,
      "step": 610
    },
    {
      "entropy": 7.099151086807251,
      "epoch": 0.05136986301369863,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000307,
      "loss": 6.6837,
      "mean_token_accuracy": 0.10082755610346794,
      "num_tokens": 1126221.0,
      "step": 615
    },
    {
      "entropy": 6.922530031204223,
      "epoch": 0.051787504176411626,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003095,
      "loss": 6.5676,
      "mean_token_accuracy": 0.10399623438715935,
      "num_tokens": 1135662.0,
      "step": 620
    },
    {
      "entropy": 7.000867414474487,
      "epoch": 0.052205145339124626,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000312,
      "loss": 6.6522,
      "mean_token_accuracy": 0.10379877239465714,
      "num_tokens": 1143963.0,
      "step": 625
    },
    {
      "entropy": 6.910499048233032,
      "epoch": 0.05262278650183762,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003145,
      "loss": 6.5471,
      "mean_token_accuracy": 0.1115996465086937,
      "num_tokens": 1152470.0,
      "step": 630
    },
    {
      "entropy": 7.002635765075683,
      "epoch": 0.05304042766455062,
      "grad_norm": 1.15625,
      "learning_rate": 0.000317,
      "loss": 6.5138,
      "mean_token_accuracy": 0.10561447665095329,
      "num_tokens": 1161257.0,
      "step": 635
    },
    {
      "entropy": 7.011125802993774,
      "epoch": 0.053458068827263616,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003195,
      "loss": 6.6179,
      "mean_token_accuracy": 0.10046471804380416,
      "num_tokens": 1170969.0,
      "step": 640
    },
    {
      "entropy": 7.024777126312256,
      "epoch": 0.053875709989976615,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000322,
      "loss": 6.7546,
      "mean_token_accuracy": 0.10059637054800988,
      "num_tokens": 1181127.0,
      "step": 645
    },
    {
      "entropy": 6.979152011871338,
      "epoch": 0.05429335115268961,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.5445,
      "mean_token_accuracy": 0.10627811849117279,
      "num_tokens": 1190260.0,
      "step": 650
    },
    {
      "entropy": 6.932449388504028,
      "epoch": 0.05471099231540261,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.6163,
      "mean_token_accuracy": 0.10629051998257637,
      "num_tokens": 1200013.0,
      "step": 655
    },
    {
      "entropy": 6.959189987182617,
      "epoch": 0.055128633478115606,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00032950000000000004,
      "loss": 6.5578,
      "mean_token_accuracy": 0.1079456403851509,
      "num_tokens": 1209239.0,
      "step": 660
    },
    {
      "entropy": 6.908866500854492,
      "epoch": 0.0555462746408286,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.4939,
      "mean_token_accuracy": 0.10315626636147499,
      "num_tokens": 1218552.0,
      "step": 665
    },
    {
      "entropy": 6.846392631530762,
      "epoch": 0.0559639158035416,
      "grad_norm": 1.390625,
      "learning_rate": 0.00033450000000000005,
      "loss": 6.557,
      "mean_token_accuracy": 0.10559443160891532,
      "num_tokens": 1227632.0,
      "step": 670
    },
    {
      "entropy": 6.971360492706299,
      "epoch": 0.056381556966254597,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000337,
      "loss": 6.59,
      "mean_token_accuracy": 0.10625778958201408,
      "num_tokens": 1236433.0,
      "step": 675
    },
    {
      "entropy": 6.848383378982544,
      "epoch": 0.05679919812896759,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003395,
      "loss": 6.6224,
      "mean_token_accuracy": 0.09530000388622284,
      "num_tokens": 1245700.0,
      "step": 680
    },
    {
      "entropy": 6.936568307876587,
      "epoch": 0.05721683929168059,
      "grad_norm": 1.125,
      "learning_rate": 0.000342,
      "loss": 6.451,
      "mean_token_accuracy": 0.10879322364926339,
      "num_tokens": 1254206.0,
      "step": 685
    },
    {
      "entropy": 6.84785647392273,
      "epoch": 0.05763448045439359,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.5067,
      "mean_token_accuracy": 0.11103252843022346,
      "num_tokens": 1263302.0,
      "step": 690
    },
    {
      "entropy": 6.9186759948730465,
      "epoch": 0.05805212161710658,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000347,
      "loss": 6.555,
      "mean_token_accuracy": 0.11069672629237175,
      "num_tokens": 1272998.0,
      "step": 695
    },
    {
      "entropy": 6.80730710029602,
      "epoch": 0.05846976277981958,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003495,
      "loss": 6.5205,
      "mean_token_accuracy": 0.10760543420910836,
      "num_tokens": 1281955.0,
      "step": 700
    },
    {
      "entropy": 6.894368362426758,
      "epoch": 0.05888740394253258,
      "grad_norm": 1.25,
      "learning_rate": 0.000352,
      "loss": 6.5413,
      "mean_token_accuracy": 0.1049070693552494,
      "num_tokens": 1291219.0,
      "step": 705
    },
    {
      "entropy": 6.867767763137818,
      "epoch": 0.05930504510524557,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003545,
      "loss": 6.5563,
      "mean_token_accuracy": 0.10610216856002808,
      "num_tokens": 1300963.0,
      "step": 710
    },
    {
      "entropy": 6.923541641235351,
      "epoch": 0.05972268626795857,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000357,
      "loss": 6.5669,
      "mean_token_accuracy": 0.10349256247282028,
      "num_tokens": 1310112.0,
      "step": 715
    },
    {
      "entropy": 6.8491448879241945,
      "epoch": 0.06014032743067157,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003595,
      "loss": 6.501,
      "mean_token_accuracy": 0.11320272982120513,
      "num_tokens": 1319309.0,
      "step": 720
    },
    {
      "entropy": 6.741468000411987,
      "epoch": 0.06055796859338457,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000362,
      "loss": 6.4675,
      "mean_token_accuracy": 0.10969400852918625,
      "num_tokens": 1329049.0,
      "step": 725
    },
    {
      "entropy": 6.756487131118774,
      "epoch": 0.06097560975609756,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003645,
      "loss": 6.3257,
      "mean_token_accuracy": 0.11157802790403366,
      "num_tokens": 1337770.0,
      "step": 730
    },
    {
      "entropy": 6.703513860702515,
      "epoch": 0.06139325091881056,
      "grad_norm": 1.046875,
      "learning_rate": 0.000367,
      "loss": 6.4064,
      "mean_token_accuracy": 0.11474933475255966,
      "num_tokens": 1346705.0,
      "step": 735
    },
    {
      "entropy": 6.771881055831909,
      "epoch": 0.06181089208152356,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003695,
      "loss": 6.4697,
      "mean_token_accuracy": 0.106501704454422,
      "num_tokens": 1354860.0,
      "step": 740
    },
    {
      "entropy": 6.815874862670898,
      "epoch": 0.06222853324423655,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000372,
      "loss": 6.4437,
      "mean_token_accuracy": 0.11245461031794549,
      "num_tokens": 1364140.0,
      "step": 745
    },
    {
      "entropy": 6.8553221225738525,
      "epoch": 0.06264617440694954,
      "grad_norm": 1.125,
      "learning_rate": 0.0003745,
      "loss": 6.5912,
      "mean_token_accuracy": 0.10352180451154709,
      "num_tokens": 1373645.0,
      "step": 750
    },
    {
      "entropy": 6.804048633575439,
      "epoch": 0.06306381556966255,
      "grad_norm": 1.109375,
      "learning_rate": 0.000377,
      "loss": 6.4578,
      "mean_token_accuracy": 0.11486915424466133,
      "num_tokens": 1383291.0,
      "step": 755
    },
    {
      "entropy": 6.777978563308716,
      "epoch": 0.06348145673237554,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003795,
      "loss": 6.4021,
      "mean_token_accuracy": 0.11272331476211547,
      "num_tokens": 1391172.0,
      "step": 760
    },
    {
      "entropy": 6.718217420578003,
      "epoch": 0.06389909789508853,
      "grad_norm": 1.125,
      "learning_rate": 0.000382,
      "loss": 6.3903,
      "mean_token_accuracy": 0.11387034431099892,
      "num_tokens": 1400864.0,
      "step": 765
    },
    {
      "entropy": 6.734544563293457,
      "epoch": 0.06431673905780154,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003845,
      "loss": 6.4421,
      "mean_token_accuracy": 0.11549821272492408,
      "num_tokens": 1410570.0,
      "step": 770
    },
    {
      "entropy": 6.7772174835205075,
      "epoch": 0.06473438022051453,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.4169,
      "mean_token_accuracy": 0.11346649378538132,
      "num_tokens": 1419536.0,
      "step": 775
    },
    {
      "entropy": 6.61550006866455,
      "epoch": 0.06515202138322754,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.3284,
      "mean_token_accuracy": 0.1189441479742527,
      "num_tokens": 1428000.0,
      "step": 780
    },
    {
      "entropy": 6.729035377502441,
      "epoch": 0.06556966254594053,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.5179,
      "mean_token_accuracy": 0.11673909649252892,
      "num_tokens": 1437789.0,
      "step": 785
    },
    {
      "entropy": 6.7100263118743895,
      "epoch": 0.06598730370865352,
      "grad_norm": 1.25,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.3385,
      "mean_token_accuracy": 0.11243527829647064,
      "num_tokens": 1446848.0,
      "step": 790
    },
    {
      "entropy": 6.658129739761352,
      "epoch": 0.06640494487136653,
      "grad_norm": 1.125,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.3964,
      "mean_token_accuracy": 0.11076711565256118,
      "num_tokens": 1455749.0,
      "step": 795
    },
    {
      "entropy": 6.6773820400238035,
      "epoch": 0.06682258603407952,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003995,
      "loss": 6.318,
      "mean_token_accuracy": 0.11754591464996338,
      "num_tokens": 1464556.0,
      "step": 800
    },
    {
      "entropy": 6.685335445404053,
      "epoch": 0.06724022719679251,
      "grad_norm": 1.140625,
      "learning_rate": 0.000402,
      "loss": 6.4215,
      "mean_token_accuracy": 0.11340176910161973,
      "num_tokens": 1473617.0,
      "step": 805
    },
    {
      "entropy": 6.737184619903564,
      "epoch": 0.06765786835950552,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004045,
      "loss": 6.4597,
      "mean_token_accuracy": 0.11323800832033157,
      "num_tokens": 1483006.0,
      "step": 810
    },
    {
      "entropy": 6.7530170440673825,
      "epoch": 0.06807550952221851,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.4705,
      "mean_token_accuracy": 0.10952020063996315,
      "num_tokens": 1493009.0,
      "step": 815
    },
    {
      "entropy": 6.672597074508667,
      "epoch": 0.0684931506849315,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004095,
      "loss": 6.3574,
      "mean_token_accuracy": 0.11739298403263092,
      "num_tokens": 1501261.0,
      "step": 820
    },
    {
      "entropy": 6.6765588283538815,
      "epoch": 0.06891079184764451,
      "grad_norm": 1.109375,
      "learning_rate": 0.000412,
      "loss": 6.4492,
      "mean_token_accuracy": 0.11198581531643867,
      "num_tokens": 1511623.0,
      "step": 825
    },
    {
      "entropy": 6.677420949935913,
      "epoch": 0.0693284330103575,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004145,
      "loss": 6.4494,
      "mean_token_accuracy": 0.10618537515401841,
      "num_tokens": 1521741.0,
      "step": 830
    },
    {
      "entropy": 6.656988143920898,
      "epoch": 0.0697460741730705,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000417,
      "loss": 6.3866,
      "mean_token_accuracy": 0.11220621392130851,
      "num_tokens": 1530988.0,
      "step": 835
    },
    {
      "entropy": 6.581060457229614,
      "epoch": 0.0701637153357835,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004195,
      "loss": 6.2383,
      "mean_token_accuracy": 0.1149817444384098,
      "num_tokens": 1539415.0,
      "step": 840
    },
    {
      "entropy": 6.736028957366943,
      "epoch": 0.07058135649849649,
      "grad_norm": 1.171875,
      "learning_rate": 0.000422,
      "loss": 6.4412,
      "mean_token_accuracy": 0.1137430302798748,
      "num_tokens": 1547969.0,
      "step": 845
    },
    {
      "entropy": 6.564602851867676,
      "epoch": 0.07099899766120948,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004245,
      "loss": 6.2574,
      "mean_token_accuracy": 0.11980320662260055,
      "num_tokens": 1557526.0,
      "step": 850
    },
    {
      "entropy": 6.555064153671265,
      "epoch": 0.07141663882392249,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000427,
      "loss": 6.2604,
      "mean_token_accuracy": 0.12150657400488854,
      "num_tokens": 1565982.0,
      "step": 855
    },
    {
      "entropy": 6.5531515121459964,
      "epoch": 0.07183427998663548,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004295,
      "loss": 6.3223,
      "mean_token_accuracy": 0.12056252211332322,
      "num_tokens": 1574492.0,
      "step": 860
    },
    {
      "entropy": 6.53031177520752,
      "epoch": 0.07225192114934847,
      "grad_norm": 1.3125,
      "learning_rate": 0.000432,
      "loss": 6.2788,
      "mean_token_accuracy": 0.11661688834428788,
      "num_tokens": 1583726.0,
      "step": 865
    },
    {
      "entropy": 6.618681621551514,
      "epoch": 0.07266956231206148,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004345,
      "loss": 6.3043,
      "mean_token_accuracy": 0.11185830906033516,
      "num_tokens": 1593780.0,
      "step": 870
    },
    {
      "entropy": 6.555753469467163,
      "epoch": 0.07308720347477447,
      "grad_norm": 1.109375,
      "learning_rate": 0.000437,
      "loss": 6.2861,
      "mean_token_accuracy": 0.11970926225185394,
      "num_tokens": 1602746.0,
      "step": 875
    },
    {
      "entropy": 6.615863084793091,
      "epoch": 0.07350484463748747,
      "grad_norm": 1.25,
      "learning_rate": 0.0004395,
      "loss": 6.3353,
      "mean_token_accuracy": 0.1214637279510498,
      "num_tokens": 1611767.0,
      "step": 880
    },
    {
      "entropy": 6.630537700653076,
      "epoch": 0.07392248580020047,
      "grad_norm": 1.0,
      "learning_rate": 0.000442,
      "loss": 6.4031,
      "mean_token_accuracy": 0.12145772501826287,
      "num_tokens": 1621568.0,
      "step": 885
    },
    {
      "entropy": 6.681979560852051,
      "epoch": 0.07434012696291346,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004445,
      "loss": 6.3366,
      "mean_token_accuracy": 0.12460207119584084,
      "num_tokens": 1630650.0,
      "step": 890
    },
    {
      "entropy": 6.63215217590332,
      "epoch": 0.07475776812562646,
      "grad_norm": 1.078125,
      "learning_rate": 0.000447,
      "loss": 6.3467,
      "mean_token_accuracy": 0.11655501201748848,
      "num_tokens": 1640482.0,
      "step": 895
    },
    {
      "entropy": 6.57461371421814,
      "epoch": 0.07517540928833946,
      "grad_norm": 1.25,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.3777,
      "mean_token_accuracy": 0.11941920891404152,
      "num_tokens": 1649344.0,
      "step": 900
    },
    {
      "entropy": 6.6421778202056885,
      "epoch": 0.07559305045105245,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.3096,
      "mean_token_accuracy": 0.12080781757831574,
      "num_tokens": 1658962.0,
      "step": 905
    },
    {
      "entropy": 6.565168476104736,
      "epoch": 0.07601069161376545,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.3148,
      "mean_token_accuracy": 0.12183948084712029,
      "num_tokens": 1668615.0,
      "step": 910
    },
    {
      "entropy": 6.573621845245361,
      "epoch": 0.07642833277647845,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.2734,
      "mean_token_accuracy": 0.12021931409835815,
      "num_tokens": 1679282.0,
      "step": 915
    },
    {
      "entropy": 6.5133349895477295,
      "epoch": 0.07684597393919144,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.3597,
      "mean_token_accuracy": 0.12050894126296044,
      "num_tokens": 1688362.0,
      "step": 920
    },
    {
      "entropy": 6.551096725463867,
      "epoch": 0.07726361510190444,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000462,
      "loss": 6.2584,
      "mean_token_accuracy": 0.11896604299545288,
      "num_tokens": 1697593.0,
      "step": 925
    },
    {
      "entropy": 6.589361953735351,
      "epoch": 0.07768125626461744,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004645,
      "loss": 6.3227,
      "mean_token_accuracy": 0.12139411941170693,
      "num_tokens": 1706961.0,
      "step": 930
    },
    {
      "entropy": 6.611653614044189,
      "epoch": 0.07809889742733044,
      "grad_norm": 1.078125,
      "learning_rate": 0.000467,
      "loss": 6.3515,
      "mean_token_accuracy": 0.1260267861187458,
      "num_tokens": 1716222.0,
      "step": 935
    },
    {
      "entropy": 6.465685606002808,
      "epoch": 0.07851653859004344,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004695,
      "loss": 6.2154,
      "mean_token_accuracy": 0.11723333075642586,
      "num_tokens": 1726750.0,
      "step": 940
    },
    {
      "entropy": 6.595513200759887,
      "epoch": 0.07893417975275643,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000472,
      "loss": 6.3318,
      "mean_token_accuracy": 0.11234179660677909,
      "num_tokens": 1736262.0,
      "step": 945
    },
    {
      "entropy": 6.523676586151123,
      "epoch": 0.07935182091546943,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004745,
      "loss": 6.342,
      "mean_token_accuracy": 0.12143515646457673,
      "num_tokens": 1745555.0,
      "step": 950
    },
    {
      "entropy": 6.409244871139526,
      "epoch": 0.07976946207818243,
      "grad_norm": 1.390625,
      "learning_rate": 0.000477,
      "loss": 6.2088,
      "mean_token_accuracy": 0.1247611589729786,
      "num_tokens": 1754107.0,
      "step": 955
    },
    {
      "entropy": 6.638109350204468,
      "epoch": 0.08018710324089542,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004795,
      "loss": 6.389,
      "mean_token_accuracy": 0.11902901157736778,
      "num_tokens": 1762982.0,
      "step": 960
    },
    {
      "entropy": 6.475300741195679,
      "epoch": 0.08060474440360842,
      "grad_norm": 1.15625,
      "learning_rate": 0.000482,
      "loss": 6.2843,
      "mean_token_accuracy": 0.12456510663032531,
      "num_tokens": 1772545.0,
      "step": 965
    },
    {
      "entropy": 6.663013553619384,
      "epoch": 0.08102238556632142,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004845,
      "loss": 6.4068,
      "mean_token_accuracy": 0.11605887189507484,
      "num_tokens": 1782052.0,
      "step": 970
    },
    {
      "entropy": 6.443525791168213,
      "epoch": 0.08144002672903441,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000487,
      "loss": 6.2043,
      "mean_token_accuracy": 0.12791636884212493,
      "num_tokens": 1791682.0,
      "step": 975
    },
    {
      "entropy": 6.452620172500611,
      "epoch": 0.08185766789174741,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004895,
      "loss": 6.2055,
      "mean_token_accuracy": 0.1278427518904209,
      "num_tokens": 1800526.0,
      "step": 980
    },
    {
      "entropy": 6.595154523849487,
      "epoch": 0.08227530905446041,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000492,
      "loss": 6.3582,
      "mean_token_accuracy": 0.12327518835663795,
      "num_tokens": 1810062.0,
      "step": 985
    },
    {
      "entropy": 6.507117128372192,
      "epoch": 0.0826929502171734,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004945,
      "loss": 6.3231,
      "mean_token_accuracy": 0.1312553495168686,
      "num_tokens": 1818909.0,
      "step": 990
    },
    {
      "entropy": 6.483727598190308,
      "epoch": 0.0831105913798864,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000497,
      "loss": 6.282,
      "mean_token_accuracy": 0.12424813508987427,
      "num_tokens": 1827978.0,
      "step": 995
    },
    {
      "entropy": 6.446569299697876,
      "epoch": 0.0835282325425994,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995,
      "loss": 6.2372,
      "mean_token_accuracy": 0.12770728915929794,
      "num_tokens": 1837955.0,
      "step": 1000
    },
    {
      "entropy": 6.493737745285034,
      "epoch": 0.0839458737053124,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004999999987395534,
      "loss": 6.3469,
      "mean_token_accuracy": 0.11529407426714897,
      "num_tokens": 1847699.0,
      "step": 1005
    },
    {
      "entropy": 6.606112098693847,
      "epoch": 0.08436351486802539,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999999936189888,
      "loss": 6.2756,
      "mean_token_accuracy": 0.12251131236553192,
      "num_tokens": 1856490.0,
      "step": 1010
    },
    {
      "entropy": 6.418572616577149,
      "epoch": 0.0847811560307384,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999999845595285,
      "loss": 6.2846,
      "mean_token_accuracy": 0.12404887601733208,
      "num_tokens": 1865644.0,
      "step": 1015
    },
    {
      "entropy": 6.5010144233703615,
      "epoch": 0.08519879719345139,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004999999715611727,
      "loss": 6.2203,
      "mean_token_accuracy": 0.1302325703203678,
      "num_tokens": 1874450.0,
      "step": 1020
    },
    {
      "entropy": 6.401221656799317,
      "epoch": 0.08561643835616438,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999999546239215,
      "loss": 6.2216,
      "mean_token_accuracy": 0.1253162756562233,
      "num_tokens": 1883259.0,
      "step": 1025
    },
    {
      "entropy": 6.503570652008056,
      "epoch": 0.08603407951887738,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999999337477753,
      "loss": 6.3041,
      "mean_token_accuracy": 0.11902791410684585,
      "num_tokens": 1891897.0,
      "step": 1030
    },
    {
      "entropy": 6.500441360473633,
      "epoch": 0.08645172068159038,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999999089327344,
      "loss": 6.2862,
      "mean_token_accuracy": 0.1212472915649414,
      "num_tokens": 1901011.0,
      "step": 1035
    },
    {
      "entropy": 6.40726375579834,
      "epoch": 0.08686936184430337,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999998801787993,
      "loss": 6.1837,
      "mean_token_accuracy": 0.13080133870244026,
      "num_tokens": 1910132.0,
      "step": 1040
    },
    {
      "entropy": 6.537573909759521,
      "epoch": 0.08728700300701638,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999998474859704,
      "loss": 6.3178,
      "mean_token_accuracy": 0.12691309154033661,
      "num_tokens": 1919938.0,
      "step": 1045
    },
    {
      "entropy": 6.34345645904541,
      "epoch": 0.08770464416972937,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999998108542483,
      "loss": 6.1791,
      "mean_token_accuracy": 0.12733794674277304,
      "num_tokens": 1928199.0,
      "step": 1050
    },
    {
      "entropy": 6.4469976902008055,
      "epoch": 0.08812228533244236,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004999997702836339,
      "loss": 6.1611,
      "mean_token_accuracy": 0.12794166281819344,
      "num_tokens": 1937133.0,
      "step": 1055
    },
    {
      "entropy": 6.510084438323974,
      "epoch": 0.08853992649515537,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999997257741274,
      "loss": 6.3356,
      "mean_token_accuracy": 0.12451440691947938,
      "num_tokens": 1946580.0,
      "step": 1060
    },
    {
      "entropy": 6.4161351203918455,
      "epoch": 0.08895756765786836,
      "grad_norm": 1.078125,
      "learning_rate": 0.00049999967732573,
      "loss": 6.2346,
      "mean_token_accuracy": 0.12813472747802734,
      "num_tokens": 1955413.0,
      "step": 1065
    },
    {
      "entropy": 6.572496318817139,
      "epoch": 0.08937520882058135,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999996249384425,
      "loss": 6.3134,
      "mean_token_accuracy": 0.12880465313792228,
      "num_tokens": 1964716.0,
      "step": 1070
    },
    {
      "entropy": 6.361472511291504,
      "epoch": 0.08979284998329436,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999995686122657,
      "loss": 6.1731,
      "mean_token_accuracy": 0.13375702425837516,
      "num_tokens": 1974081.0,
      "step": 1075
    },
    {
      "entropy": 6.319162797927857,
      "epoch": 0.09021049114600735,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004999995083472006,
      "loss": 6.1591,
      "mean_token_accuracy": 0.12707530632615088,
      "num_tokens": 1983035.0,
      "step": 1080
    },
    {
      "entropy": 6.4332092761993405,
      "epoch": 0.09062813230872034,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999994441432481,
      "loss": 6.1721,
      "mean_token_accuracy": 0.13475777059793473,
      "num_tokens": 1992475.0,
      "step": 1085
    },
    {
      "entropy": 6.325397729873657,
      "epoch": 0.09104577347143335,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999993760004097,
      "loss": 6.2343,
      "mean_token_accuracy": 0.12572435140609742,
      "num_tokens": 2002729.0,
      "step": 1090
    },
    {
      "entropy": 6.269658946990967,
      "epoch": 0.09146341463414634,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999993039186864,
      "loss": 6.0865,
      "mean_token_accuracy": 0.13097318336367608,
      "num_tokens": 2011319.0,
      "step": 1095
    },
    {
      "entropy": 6.446812629699707,
      "epoch": 0.09188105579685933,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999992278980794,
      "loss": 6.13,
      "mean_token_accuracy": 0.12895696312189103,
      "num_tokens": 2020456.0,
      "step": 1100
    },
    {
      "entropy": 6.271724033355713,
      "epoch": 0.09229869695957234,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004999991479385901,
      "loss": 6.2195,
      "mean_token_accuracy": 0.1341036967933178,
      "num_tokens": 2029417.0,
      "step": 1105
    },
    {
      "entropy": 6.4861917972564695,
      "epoch": 0.09271633812228533,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004999990640402198,
      "loss": 6.2979,
      "mean_token_accuracy": 0.12562437877058982,
      "num_tokens": 2039290.0,
      "step": 1110
    },
    {
      "entropy": 6.4251697063446045,
      "epoch": 0.09313397928499834,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999989762029701,
      "loss": 6.1128,
      "mean_token_accuracy": 0.13544357791543007,
      "num_tokens": 2048485.0,
      "step": 1115
    },
    {
      "entropy": 6.2083420753479,
      "epoch": 0.09355162044771133,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999988844268425,
      "loss": 6.096,
      "mean_token_accuracy": 0.13170384764671325,
      "num_tokens": 2057131.0,
      "step": 1120
    },
    {
      "entropy": 6.312107372283935,
      "epoch": 0.09396926161042432,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999987887118387,
      "loss": 6.0797,
      "mean_token_accuracy": 0.13444808647036552,
      "num_tokens": 2066295.0,
      "step": 1125
    },
    {
      "entropy": 6.347876739501953,
      "epoch": 0.09438690277313733,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999986890579602,
      "loss": 6.0873,
      "mean_token_accuracy": 0.1415345251560211,
      "num_tokens": 2075656.0,
      "step": 1130
    },
    {
      "entropy": 6.335701656341553,
      "epoch": 0.09480454393585032,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999985854652087,
      "loss": 6.2225,
      "mean_token_accuracy": 0.12682175189256667,
      "num_tokens": 2086046.0,
      "step": 1135
    },
    {
      "entropy": 6.416574192047119,
      "epoch": 0.09522218509856331,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999984779335862,
      "loss": 6.2349,
      "mean_token_accuracy": 0.12994211688637733,
      "num_tokens": 2094808.0,
      "step": 1140
    },
    {
      "entropy": 6.290297937393189,
      "epoch": 0.09563982626127632,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999983664630945,
      "loss": 6.1605,
      "mean_token_accuracy": 0.1330721542239189,
      "num_tokens": 2104155.0,
      "step": 1145
    },
    {
      "entropy": 6.382285833358765,
      "epoch": 0.09605746742398931,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999982510537357,
      "loss": 6.0692,
      "mean_token_accuracy": 0.12856426909565927,
      "num_tokens": 2112517.0,
      "step": 1150
    },
    {
      "entropy": 6.208939743041992,
      "epoch": 0.0964751085867023,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999981317055116,
      "loss": 6.0236,
      "mean_token_accuracy": 0.13144809231162072,
      "num_tokens": 2121355.0,
      "step": 1155
    },
    {
      "entropy": 6.3527082920074465,
      "epoch": 0.09689274974941531,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999980084184244,
      "loss": 6.1534,
      "mean_token_accuracy": 0.13644161820411682,
      "num_tokens": 2129278.0,
      "step": 1160
    },
    {
      "entropy": 6.3879131317138675,
      "epoch": 0.0973103909121283,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999978811924762,
      "loss": 6.2951,
      "mean_token_accuracy": 0.12582162991166115,
      "num_tokens": 2139548.0,
      "step": 1165
    },
    {
      "entropy": 6.277176761627198,
      "epoch": 0.09772803207484129,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999977500276693,
      "loss": 6.2004,
      "mean_token_accuracy": 0.12778721749782562,
      "num_tokens": 2148765.0,
      "step": 1170
    },
    {
      "entropy": 6.430830717086792,
      "epoch": 0.0981456732375543,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999976149240058,
      "loss": 6.1291,
      "mean_token_accuracy": 0.1366495445370674,
      "num_tokens": 2157482.0,
      "step": 1175
    },
    {
      "entropy": 6.255828189849853,
      "epoch": 0.09856331440026729,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999974758814884,
      "loss": 6.2152,
      "mean_token_accuracy": 0.1265973374247551,
      "num_tokens": 2167320.0,
      "step": 1180
    },
    {
      "entropy": 6.394571590423584,
      "epoch": 0.09898095556298028,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999973329001193,
      "loss": 6.1659,
      "mean_token_accuracy": 0.12668881118297576,
      "num_tokens": 2175454.0,
      "step": 1185
    },
    {
      "entropy": 6.309787034988403,
      "epoch": 0.09939859672569329,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999971859799009,
      "loss": 6.1164,
      "mean_token_accuracy": 0.12462705001235008,
      "num_tokens": 2184391.0,
      "step": 1190
    },
    {
      "entropy": 6.220657396316528,
      "epoch": 0.09981623788840628,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999970351208361,
      "loss": 5.9863,
      "mean_token_accuracy": 0.13656294345855713,
      "num_tokens": 2193309.0,
      "step": 1195
    },
    {
      "entropy": 6.33141303062439,
      "epoch": 0.10023387905111927,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999968803229275,
      "loss": 6.132,
      "mean_token_accuracy": 0.12762896567583085,
      "num_tokens": 2202057.0,
      "step": 1200
    },
    {
      "entropy": 6.345908117294312,
      "epoch": 0.10065152021383228,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999967215861775,
      "loss": 6.2193,
      "mean_token_accuracy": 0.12862889021635054,
      "num_tokens": 2212384.0,
      "step": 1205
    },
    {
      "entropy": 6.33440523147583,
      "epoch": 0.10106916137654527,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499996558910589,
      "loss": 6.129,
      "mean_token_accuracy": 0.13246983289718628,
      "num_tokens": 2221451.0,
      "step": 1210
    },
    {
      "entropy": 6.197815370559693,
      "epoch": 0.10148680253925826,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499996392296165,
      "loss": 6.033,
      "mean_token_accuracy": 0.13185914009809493,
      "num_tokens": 2230440.0,
      "step": 1215
    },
    {
      "entropy": 6.355322027206421,
      "epoch": 0.10190444370197127,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999962217429083,
      "loss": 6.214,
      "mean_token_accuracy": 0.12665950432419776,
      "num_tokens": 2239985.0,
      "step": 1220
    },
    {
      "entropy": 6.3212282180786135,
      "epoch": 0.10232208486468426,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999960472508219,
      "loss": 6.0842,
      "mean_token_accuracy": 0.13362730666995049,
      "num_tokens": 2250391.0,
      "step": 1225
    },
    {
      "entropy": 6.305384016036987,
      "epoch": 0.10273972602739725,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499995868819909,
      "loss": 6.1996,
      "mean_token_accuracy": 0.1324283190071583,
      "num_tokens": 2260620.0,
      "step": 1230
    },
    {
      "entropy": 6.320593690872192,
      "epoch": 0.10315736719011026,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999956864501725,
      "loss": 6.16,
      "mean_token_accuracy": 0.12899876162409782,
      "num_tokens": 2271176.0,
      "step": 1235
    },
    {
      "entropy": 6.355190849304199,
      "epoch": 0.10357500835282325,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999955001416156,
      "loss": 6.0567,
      "mean_token_accuracy": 0.1392730712890625,
      "num_tokens": 2280371.0,
      "step": 1240
    },
    {
      "entropy": 6.190901041030884,
      "epoch": 0.10399264951553625,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999953098942417,
      "loss": 6.0273,
      "mean_token_accuracy": 0.13370997533202172,
      "num_tokens": 2289015.0,
      "step": 1245
    },
    {
      "entropy": 6.2859964847564695,
      "epoch": 0.10441029067824925,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999951157080541,
      "loss": 6.1068,
      "mean_token_accuracy": 0.13385709300637244,
      "num_tokens": 2297409.0,
      "step": 1250
    },
    {
      "entropy": 6.135534811019897,
      "epoch": 0.10482793184096224,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999949175830561,
      "loss": 6.1245,
      "mean_token_accuracy": 0.12929951921105384,
      "num_tokens": 2307017.0,
      "step": 1255
    },
    {
      "entropy": 6.280382251739502,
      "epoch": 0.10524557300367524,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999947155192514,
      "loss": 5.9706,
      "mean_token_accuracy": 0.1417563498020172,
      "num_tokens": 2316014.0,
      "step": 1260
    },
    {
      "entropy": 6.253998756408691,
      "epoch": 0.10566321416638824,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999945095166433,
      "loss": 6.0745,
      "mean_token_accuracy": 0.13147094696760178,
      "num_tokens": 2326326.0,
      "step": 1265
    },
    {
      "entropy": 6.383511114120483,
      "epoch": 0.10608085532910123,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999942995752354,
      "loss": 6.1693,
      "mean_token_accuracy": 0.1279841721057892,
      "num_tokens": 2335108.0,
      "step": 1270
    },
    {
      "entropy": 6.264583396911621,
      "epoch": 0.10649849649181423,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999940856950315,
      "loss": 6.0791,
      "mean_token_accuracy": 0.14134034141898155,
      "num_tokens": 2343971.0,
      "step": 1275
    },
    {
      "entropy": 6.292309951782227,
      "epoch": 0.10691613765452723,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999938678760354,
      "loss": 6.1318,
      "mean_token_accuracy": 0.13392335250973703,
      "num_tokens": 2353742.0,
      "step": 1280
    },
    {
      "entropy": 6.289759016036987,
      "epoch": 0.10733377881724022,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999936461182507,
      "loss": 6.0966,
      "mean_token_accuracy": 0.13741502091288565,
      "num_tokens": 2363629.0,
      "step": 1285
    },
    {
      "entropy": 6.082193899154663,
      "epoch": 0.10775141997995323,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999934204216814,
      "loss": 5.9501,
      "mean_token_accuracy": 0.1403780050575733,
      "num_tokens": 2372705.0,
      "step": 1290
    },
    {
      "entropy": 6.2366783618927,
      "epoch": 0.10816906114266622,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999931907863315,
      "loss": 6.0658,
      "mean_token_accuracy": 0.13578577116131782,
      "num_tokens": 2382136.0,
      "step": 1295
    },
    {
      "entropy": 6.113440036773682,
      "epoch": 0.10858670230537922,
      "grad_norm": 1.078125,
      "learning_rate": 0.000499992957212205,
      "loss": 6.0356,
      "mean_token_accuracy": 0.1353696718811989,
      "num_tokens": 2391060.0,
      "step": 1300
    },
    {
      "entropy": 6.3979072093963625,
      "epoch": 0.10900434346809222,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999927196993059,
      "loss": 6.1711,
      "mean_token_accuracy": 0.13628476485610008,
      "num_tokens": 2399885.0,
      "step": 1305
    },
    {
      "entropy": 6.263845205307007,
      "epoch": 0.10942198463080521,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999924782476384,
      "loss": 6.0141,
      "mean_token_accuracy": 0.1464387968182564,
      "num_tokens": 2409654.0,
      "step": 1310
    },
    {
      "entropy": 6.226789665222168,
      "epoch": 0.1098396257935182,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999922328572068,
      "loss": 6.0925,
      "mean_token_accuracy": 0.13425977975130082,
      "num_tokens": 2419342.0,
      "step": 1315
    },
    {
      "entropy": 6.150572347640991,
      "epoch": 0.11025726695623121,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999919835280154,
      "loss": 5.9804,
      "mean_token_accuracy": 0.1443939872086048,
      "num_tokens": 2428845.0,
      "step": 1320
    },
    {
      "entropy": 6.273833227157593,
      "epoch": 0.1106749081189442,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999917302600684,
      "loss": 5.9739,
      "mean_token_accuracy": 0.14326085448265075,
      "num_tokens": 2437134.0,
      "step": 1325
    },
    {
      "entropy": 6.0786871910095215,
      "epoch": 0.1110925492816572,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999914730533704,
      "loss": 5.819,
      "mean_token_accuracy": 0.14700671806931495,
      "num_tokens": 2445438.0,
      "step": 1330
    },
    {
      "entropy": 6.22350287437439,
      "epoch": 0.1115101904443702,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999912119079258,
      "loss": 6.1506,
      "mean_token_accuracy": 0.1359838679432869,
      "num_tokens": 2455039.0,
      "step": 1335
    },
    {
      "entropy": 6.229891204833985,
      "epoch": 0.1119278316070832,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999909468237393,
      "loss": 6.0765,
      "mean_token_accuracy": 0.13616930097341537,
      "num_tokens": 2465170.0,
      "step": 1340
    },
    {
      "entropy": 6.1581672668457035,
      "epoch": 0.11234547276979619,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999906778008153,
      "loss": 5.8308,
      "mean_token_accuracy": 0.14951538890600205,
      "num_tokens": 2473688.0,
      "step": 1345
    },
    {
      "entropy": 6.1050516128540036,
      "epoch": 0.11276311393250919,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999904048391588,
      "loss": 6.0805,
      "mean_token_accuracy": 0.13373313397169112,
      "num_tokens": 2482335.0,
      "step": 1350
    },
    {
      "entropy": 6.243770885467529,
      "epoch": 0.11318075509522219,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999901279387743,
      "loss": 6.0296,
      "mean_token_accuracy": 0.14334997087717055,
      "num_tokens": 2491808.0,
      "step": 1355
    },
    {
      "entropy": 6.295271968841552,
      "epoch": 0.11359839625793518,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999898470996669,
      "loss": 6.1569,
      "mean_token_accuracy": 0.12944946438074112,
      "num_tokens": 2501422.0,
      "step": 1360
    },
    {
      "entropy": 6.184634590148926,
      "epoch": 0.11401603742064818,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999895623218415,
      "loss": 5.9515,
      "mean_token_accuracy": 0.1408703289926052,
      "num_tokens": 2511201.0,
      "step": 1365
    },
    {
      "entropy": 6.188165283203125,
      "epoch": 0.11443367858336118,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999892736053027,
      "loss": 5.9929,
      "mean_token_accuracy": 0.13730614483356476,
      "num_tokens": 2519798.0,
      "step": 1370
    },
    {
      "entropy": 6.183836936950684,
      "epoch": 0.11485131974607417,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999889809500561,
      "loss": 5.968,
      "mean_token_accuracy": 0.14003934487700462,
      "num_tokens": 2529177.0,
      "step": 1375
    },
    {
      "entropy": 6.234930992126465,
      "epoch": 0.11526896090878717,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999886843561065,
      "loss": 5.9642,
      "mean_token_accuracy": 0.14338480234146117,
      "num_tokens": 2537571.0,
      "step": 1380
    },
    {
      "entropy": 6.02715220451355,
      "epoch": 0.11568660207150017,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999883838234592,
      "loss": 5.881,
      "mean_token_accuracy": 0.14644728004932403,
      "num_tokens": 2546258.0,
      "step": 1385
    },
    {
      "entropy": 6.149686002731324,
      "epoch": 0.11610424323421316,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999880793521194,
      "loss": 5.8893,
      "mean_token_accuracy": 0.1373589113354683,
      "num_tokens": 2555862.0,
      "step": 1390
    },
    {
      "entropy": 6.014897727966309,
      "epoch": 0.11652188439692616,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999877709420924,
      "loss": 5.9175,
      "mean_token_accuracy": 0.14411035925149918,
      "num_tokens": 2564583.0,
      "step": 1395
    },
    {
      "entropy": 6.16389479637146,
      "epoch": 0.11693952555963916,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999874585933837,
      "loss": 6.0099,
      "mean_token_accuracy": 0.1459191247820854,
      "num_tokens": 2574217.0,
      "step": 1400
    },
    {
      "entropy": 6.194282054901123,
      "epoch": 0.11735716672235215,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999871423059987,
      "loss": 5.9063,
      "mean_token_accuracy": 0.1440393716096878,
      "num_tokens": 2582821.0,
      "step": 1405
    },
    {
      "entropy": 6.116659116744995,
      "epoch": 0.11777480788506516,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499986822079943,
      "loss": 6.0746,
      "mean_token_accuracy": 0.13738386183977128,
      "num_tokens": 2592748.0,
      "step": 1410
    },
    {
      "entropy": 6.222877502441406,
      "epoch": 0.11819244904777815,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999864979152221,
      "loss": 5.9736,
      "mean_token_accuracy": 0.1457105740904808,
      "num_tokens": 2602053.0,
      "step": 1415
    },
    {
      "entropy": 6.020002746582032,
      "epoch": 0.11861009021049114,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999861698118417,
      "loss": 6.0455,
      "mean_token_accuracy": 0.13822002336382866,
      "num_tokens": 2610557.0,
      "step": 1420
    },
    {
      "entropy": 6.204063892364502,
      "epoch": 0.11902773137320415,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999858377698076,
      "loss": 5.9384,
      "mean_token_accuracy": 0.13854485303163527,
      "num_tokens": 2619628.0,
      "step": 1425
    },
    {
      "entropy": 6.294986391067505,
      "epoch": 0.11944537253591714,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999855017891256,
      "loss": 6.0939,
      "mean_token_accuracy": 0.13476530238986015,
      "num_tokens": 2629284.0,
      "step": 1430
    },
    {
      "entropy": 6.102528619766235,
      "epoch": 0.11986301369863013,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999851618698017,
      "loss": 6.0164,
      "mean_token_accuracy": 0.14004518166184426,
      "num_tokens": 2638001.0,
      "step": 1435
    },
    {
      "entropy": 6.149356412887573,
      "epoch": 0.12028065486134314,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999848180118416,
      "loss": 6.0075,
      "mean_token_accuracy": 0.1504116952419281,
      "num_tokens": 2647071.0,
      "step": 1440
    },
    {
      "entropy": 6.189749431610108,
      "epoch": 0.12069829602405613,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999844702152516,
      "loss": 5.9865,
      "mean_token_accuracy": 0.14875612705945968,
      "num_tokens": 2655826.0,
      "step": 1445
    },
    {
      "entropy": 6.086092138290406,
      "epoch": 0.12111593718676913,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999841184800374,
      "loss": 5.8555,
      "mean_token_accuracy": 0.15078829377889633,
      "num_tokens": 2664431.0,
      "step": 1450
    },
    {
      "entropy": 6.010515975952148,
      "epoch": 0.12153357834948213,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999837628062055,
      "loss": 5.7622,
      "mean_token_accuracy": 0.15104235708713531,
      "num_tokens": 2673631.0,
      "step": 1455
    },
    {
      "entropy": 6.052291202545166,
      "epoch": 0.12195121951219512,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999834031937621,
      "loss": 6.047,
      "mean_token_accuracy": 0.14088533371686934,
      "num_tokens": 2683363.0,
      "step": 1460
    },
    {
      "entropy": 6.159712791442871,
      "epoch": 0.12236886067490813,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999830396427133,
      "loss": 5.9934,
      "mean_token_accuracy": 0.13705805838108062,
      "num_tokens": 2692533.0,
      "step": 1465
    },
    {
      "entropy": 6.13315601348877,
      "epoch": 0.12278650183762112,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999826721530657,
      "loss": 5.9982,
      "mean_token_accuracy": 0.14548553973436357,
      "num_tokens": 2701211.0,
      "step": 1470
    },
    {
      "entropy": 6.203657341003418,
      "epoch": 0.12320414300033411,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999823007248255,
      "loss": 5.9974,
      "mean_token_accuracy": 0.1427378311753273,
      "num_tokens": 2709543.0,
      "step": 1475
    },
    {
      "entropy": 6.1870471954345705,
      "epoch": 0.12362178416304712,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999819253579992,
      "loss": 6.0064,
      "mean_token_accuracy": 0.13959514498710632,
      "num_tokens": 2718336.0,
      "step": 1480
    },
    {
      "entropy": 6.127546596527099,
      "epoch": 0.12403942532576011,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999815460525938,
      "loss": 5.9441,
      "mean_token_accuracy": 0.14745968729257583,
      "num_tokens": 2726504.0,
      "step": 1485
    },
    {
      "entropy": 6.105073881149292,
      "epoch": 0.1244570664884731,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999811628086155,
      "loss": 5.9567,
      "mean_token_accuracy": 0.1415959879755974,
      "num_tokens": 2736076.0,
      "step": 1490
    },
    {
      "entropy": 6.055798721313477,
      "epoch": 0.1248747076511861,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499980775626071,
      "loss": 5.8843,
      "mean_token_accuracy": 0.14771346226334572,
      "num_tokens": 2745497.0,
      "step": 1495
    },
    {
      "entropy": 6.116729402542115,
      "epoch": 0.12529234881389908,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999803845049674,
      "loss": 6.0194,
      "mean_token_accuracy": 0.14392556473612786,
      "num_tokens": 2754455.0,
      "step": 1500
    },
    {
      "entropy": 6.213052797317505,
      "epoch": 0.1257099899766121,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999799894453112,
      "loss": 6.0521,
      "mean_token_accuracy": 0.14362442642450332,
      "num_tokens": 2763603.0,
      "step": 1505
    },
    {
      "entropy": 6.095961046218872,
      "epoch": 0.1261276311393251,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999795904471095,
      "loss": 5.9026,
      "mean_token_accuracy": 0.14689221382141113,
      "num_tokens": 2772152.0,
      "step": 1510
    },
    {
      "entropy": 6.245121812820434,
      "epoch": 0.12654527230203808,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999791875103694,
      "loss": 6.0004,
      "mean_token_accuracy": 0.1482790566980839,
      "num_tokens": 2781459.0,
      "step": 1515
    },
    {
      "entropy": 6.167483615875244,
      "epoch": 0.12696291346475108,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999787806350977,
      "loss": 5.9445,
      "mean_token_accuracy": 0.14450416788458825,
      "num_tokens": 2791177.0,
      "step": 1520
    },
    {
      "entropy": 5.988239908218384,
      "epoch": 0.1273805546274641,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999783698213016,
      "loss": 5.9481,
      "mean_token_accuracy": 0.15092730671167373,
      "num_tokens": 2800543.0,
      "step": 1525
    },
    {
      "entropy": 6.051485967636109,
      "epoch": 0.12779819579017707,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999779550689885,
      "loss": 5.7947,
      "mean_token_accuracy": 0.1509787768125534,
      "num_tokens": 2809301.0,
      "step": 1530
    },
    {
      "entropy": 6.191277027130127,
      "epoch": 0.12821583695289007,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999775363781654,
      "loss": 5.9932,
      "mean_token_accuracy": 0.14468340426683426,
      "num_tokens": 2818269.0,
      "step": 1535
    },
    {
      "entropy": 6.0712157726287845,
      "epoch": 0.12863347811560308,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999771137488395,
      "loss": 5.8629,
      "mean_token_accuracy": 0.14732339829206467,
      "num_tokens": 2827170.0,
      "step": 1540
    },
    {
      "entropy": 5.983927059173584,
      "epoch": 0.12905111927831608,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999766871810188,
      "loss": 5.8646,
      "mean_token_accuracy": 0.14839889854192734,
      "num_tokens": 2835320.0,
      "step": 1545
    },
    {
      "entropy": 6.071327495574951,
      "epoch": 0.12946876044102906,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999762566747101,
      "loss": 5.9244,
      "mean_token_accuracy": 0.14501172602176665,
      "num_tokens": 2843755.0,
      "step": 1550
    },
    {
      "entropy": 6.151789903640747,
      "epoch": 0.12988640160374207,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999758222299214,
      "loss": 5.9715,
      "mean_token_accuracy": 0.13980228453874588,
      "num_tokens": 2853462.0,
      "step": 1555
    },
    {
      "entropy": 6.041283750534058,
      "epoch": 0.13030404276645507,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999753838466601,
      "loss": 5.8737,
      "mean_token_accuracy": 0.1441491112112999,
      "num_tokens": 2863041.0,
      "step": 1560
    },
    {
      "entropy": 6.223154211044312,
      "epoch": 0.13072168392916805,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999749415249338,
      "loss": 6.0883,
      "mean_token_accuracy": 0.14627306610345842,
      "num_tokens": 2873019.0,
      "step": 1565
    },
    {
      "entropy": 6.021166467666626,
      "epoch": 0.13113932509188106,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999744952647505,
      "loss": 5.8394,
      "mean_token_accuracy": 0.14528503715991975,
      "num_tokens": 2881467.0,
      "step": 1570
    },
    {
      "entropy": 6.078176975250244,
      "epoch": 0.13155696625459407,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999740450661176,
      "loss": 5.9283,
      "mean_token_accuracy": 0.14148435667157172,
      "num_tokens": 2890914.0,
      "step": 1575
    },
    {
      "entropy": 5.97394208908081,
      "epoch": 0.13197460741730704,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999735909290436,
      "loss": 5.8474,
      "mean_token_accuracy": 0.15130585730075835,
      "num_tokens": 2901164.0,
      "step": 1580
    },
    {
      "entropy": 6.17360486984253,
      "epoch": 0.13239224858002005,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999731328535358,
      "loss": 5.9837,
      "mean_token_accuracy": 0.14389178678393363,
      "num_tokens": 2910560.0,
      "step": 1585
    },
    {
      "entropy": 6.00739221572876,
      "epoch": 0.13280988974273306,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999726708396026,
      "loss": 5.8006,
      "mean_token_accuracy": 0.15138354450464248,
      "num_tokens": 2918980.0,
      "step": 1590
    },
    {
      "entropy": 6.089150905609131,
      "epoch": 0.13322753090544603,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499972204887252,
      "loss": 5.9897,
      "mean_token_accuracy": 0.13849136978387833,
      "num_tokens": 2929516.0,
      "step": 1595
    },
    {
      "entropy": 6.0801748752594,
      "epoch": 0.13364517206815904,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999717349964922,
      "loss": 6.068,
      "mean_token_accuracy": 0.13731011226773263,
      "num_tokens": 2940603.0,
      "step": 1600
    },
    {
      "entropy": 6.169479703903198,
      "epoch": 0.13406281323087205,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999712611673313,
      "loss": 5.9216,
      "mean_token_accuracy": 0.14724995344877242,
      "num_tokens": 2950082.0,
      "step": 1605
    },
    {
      "entropy": 5.934827041625977,
      "epoch": 0.13448045439358502,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999707833997777,
      "loss": 5.7908,
      "mean_token_accuracy": 0.1533707395195961,
      "num_tokens": 2959098.0,
      "step": 1610
    },
    {
      "entropy": 6.161834812164306,
      "epoch": 0.13489809555629803,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999703016938398,
      "loss": 5.9323,
      "mean_token_accuracy": 0.14561702385544778,
      "num_tokens": 2968585.0,
      "step": 1615
    },
    {
      "entropy": 6.053872299194336,
      "epoch": 0.13531573671901104,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000499969816049526,
      "loss": 5.9497,
      "mean_token_accuracy": 0.14551280438899994,
      "num_tokens": 2978430.0,
      "step": 1620
    },
    {
      "entropy": 6.212032508850098,
      "epoch": 0.13573337788172402,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999693264668446,
      "loss": 5.9962,
      "mean_token_accuracy": 0.13857592716813089,
      "num_tokens": 2988398.0,
      "step": 1625
    },
    {
      "entropy": 6.085241079330444,
      "epoch": 0.13615101904443702,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999688329458045,
      "loss": 6.0168,
      "mean_token_accuracy": 0.14099103808403016,
      "num_tokens": 2999255.0,
      "step": 1630
    },
    {
      "entropy": 6.169172143936157,
      "epoch": 0.13656866020715003,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999683354864141,
      "loss": 5.9099,
      "mean_token_accuracy": 0.14426834136247635,
      "num_tokens": 3008638.0,
      "step": 1635
    },
    {
      "entropy": 6.051702165603638,
      "epoch": 0.136986301369863,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999678340886823,
      "loss": 5.9137,
      "mean_token_accuracy": 0.14842754304409028,
      "num_tokens": 3018070.0,
      "step": 1640
    },
    {
      "entropy": 5.950930213928222,
      "epoch": 0.137403942532576,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999673287526176,
      "loss": 5.8784,
      "mean_token_accuracy": 0.15201137959957123,
      "num_tokens": 3027878.0,
      "step": 1645
    },
    {
      "entropy": 6.174559926986694,
      "epoch": 0.13782158369528902,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999668194782291,
      "loss": 6.0703,
      "mean_token_accuracy": 0.13952327594161035,
      "num_tokens": 3037031.0,
      "step": 1650
    },
    {
      "entropy": 6.233021545410156,
      "epoch": 0.138239224858002,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999663062655257,
      "loss": 6.0016,
      "mean_token_accuracy": 0.14014986380934716,
      "num_tokens": 3046311.0,
      "step": 1655
    },
    {
      "entropy": 6.069745588302612,
      "epoch": 0.138656866020715,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999657891145162,
      "loss": 5.8249,
      "mean_token_accuracy": 0.15104526281356812,
      "num_tokens": 3055108.0,
      "step": 1660
    },
    {
      "entropy": 5.9705235958099365,
      "epoch": 0.139074507183428,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999652680252099,
      "loss": 5.8508,
      "mean_token_accuracy": 0.14930724799633027,
      "num_tokens": 3063492.0,
      "step": 1665
    },
    {
      "entropy": 6.149300765991211,
      "epoch": 0.139492148346141,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999647429976157,
      "loss": 5.9054,
      "mean_token_accuracy": 0.15113156288862228,
      "num_tokens": 3072572.0,
      "step": 1670
    },
    {
      "entropy": 6.122020435333252,
      "epoch": 0.139909789508854,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999642140317429,
      "loss": 5.9546,
      "mean_token_accuracy": 0.14124443158507347,
      "num_tokens": 3082942.0,
      "step": 1675
    },
    {
      "entropy": 6.042447471618653,
      "epoch": 0.140327430671567,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999636811276007,
      "loss": 5.919,
      "mean_token_accuracy": 0.14785890504717827,
      "num_tokens": 3091410.0,
      "step": 1680
    },
    {
      "entropy": 6.153117609024048,
      "epoch": 0.14074507183427998,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999631442851986,
      "loss": 5.9736,
      "mean_token_accuracy": 0.1465502806007862,
      "num_tokens": 3100626.0,
      "step": 1685
    },
    {
      "entropy": 6.026203060150147,
      "epoch": 0.14116271299699298,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999626035045458,
      "loss": 5.7808,
      "mean_token_accuracy": 0.1571854069828987,
      "num_tokens": 3109339.0,
      "step": 1690
    },
    {
      "entropy": 6.103426456451416,
      "epoch": 0.141580354159706,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999620587856519,
      "loss": 5.914,
      "mean_token_accuracy": 0.14427188262343407,
      "num_tokens": 3118636.0,
      "step": 1695
    },
    {
      "entropy": 5.9886900901794435,
      "epoch": 0.14199799532241897,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999615101285262,
      "loss": 5.8792,
      "mean_token_accuracy": 0.14374638795852662,
      "num_tokens": 3128162.0,
      "step": 1700
    },
    {
      "entropy": 6.051160144805908,
      "epoch": 0.14241563648513197,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999609575331786,
      "loss": 5.8891,
      "mean_token_accuracy": 0.1519470438361168,
      "num_tokens": 3136451.0,
      "step": 1705
    },
    {
      "entropy": 6.093698120117187,
      "epoch": 0.14283327764784498,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999604009996186,
      "loss": 5.8842,
      "mean_token_accuracy": 0.1466047689318657,
      "num_tokens": 3146599.0,
      "step": 1710
    },
    {
      "entropy": 6.099306440353393,
      "epoch": 0.14325091881055796,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499959840527856,
      "loss": 5.9379,
      "mean_token_accuracy": 0.14325512945652008,
      "num_tokens": 3155941.0,
      "step": 1715
    },
    {
      "entropy": 6.005642986297607,
      "epoch": 0.14366855997327097,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999592761179007,
      "loss": 5.8656,
      "mean_token_accuracy": 0.15379103869199753,
      "num_tokens": 3164641.0,
      "step": 1720
    },
    {
      "entropy": 6.069437789916992,
      "epoch": 0.14408620113598397,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999587077697624,
      "loss": 5.8973,
      "mean_token_accuracy": 0.15470442548394203,
      "num_tokens": 3173701.0,
      "step": 1725
    },
    {
      "entropy": 5.939715385437012,
      "epoch": 0.14450384229869695,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999581354834512,
      "loss": 5.7458,
      "mean_token_accuracy": 0.14850927889347076,
      "num_tokens": 3182268.0,
      "step": 1730
    },
    {
      "entropy": 5.929523706436157,
      "epoch": 0.14492148346140996,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499957559258977,
      "loss": 5.8259,
      "mean_token_accuracy": 0.15093481987714769,
      "num_tokens": 3191513.0,
      "step": 1735
    },
    {
      "entropy": 6.022497940063476,
      "epoch": 0.14533912462412296,
      "grad_norm": 1.0,
      "learning_rate": 0.00049995697909635,
      "loss": 5.7883,
      "mean_token_accuracy": 0.14721936583518982,
      "num_tokens": 3200580.0,
      "step": 1740
    },
    {
      "entropy": 5.952256393432617,
      "epoch": 0.14575676578683594,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999563949955803,
      "loss": 5.8709,
      "mean_token_accuracy": 0.1540740728378296,
      "num_tokens": 3209053.0,
      "step": 1745
    },
    {
      "entropy": 6.015556478500367,
      "epoch": 0.14617440694954895,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499955806956678,
      "loss": 5.8598,
      "mean_token_accuracy": 0.14972457140684128,
      "num_tokens": 3218241.0,
      "step": 1750
    },
    {
      "entropy": 6.084406089782715,
      "epoch": 0.14659204811226195,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999552149796536,
      "loss": 5.9255,
      "mean_token_accuracy": 0.14822182059288025,
      "num_tokens": 3227887.0,
      "step": 1755
    },
    {
      "entropy": 5.975216627120972,
      "epoch": 0.14700968927497493,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999546190645175,
      "loss": 5.7331,
      "mean_token_accuracy": 0.15878990590572356,
      "num_tokens": 3237281.0,
      "step": 1760
    },
    {
      "entropy": 5.977021408081055,
      "epoch": 0.14742733043768794,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999540192112799,
      "loss": 5.7757,
      "mean_token_accuracy": 0.15504969358444215,
      "num_tokens": 3246907.0,
      "step": 1765
    },
    {
      "entropy": 5.893853664398193,
      "epoch": 0.14784497160040094,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999534154199514,
      "loss": 5.7749,
      "mean_token_accuracy": 0.14786117523908615,
      "num_tokens": 3256312.0,
      "step": 1770
    },
    {
      "entropy": 6.171048641204834,
      "epoch": 0.14826261276311392,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999528076905426,
      "loss": 5.8781,
      "mean_token_accuracy": 0.15021531730890275,
      "num_tokens": 3265554.0,
      "step": 1775
    },
    {
      "entropy": 5.7882325649261475,
      "epoch": 0.14868025392582693,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999521960230641,
      "loss": 5.7149,
      "mean_token_accuracy": 0.1591077446937561,
      "num_tokens": 3274731.0,
      "step": 1780
    },
    {
      "entropy": 6.024976682662964,
      "epoch": 0.14909789508853993,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999515804175268,
      "loss": 5.8542,
      "mean_token_accuracy": 0.14886621683835982,
      "num_tokens": 3283518.0,
      "step": 1785
    },
    {
      "entropy": 6.0305921077728275,
      "epoch": 0.1495155362512529,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000499950960873941,
      "loss": 5.8717,
      "mean_token_accuracy": 0.1539292372763157,
      "num_tokens": 3292244.0,
      "step": 1790
    },
    {
      "entropy": 6.053695440292358,
      "epoch": 0.14993317741396592,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999503373923182,
      "loss": 5.8294,
      "mean_token_accuracy": 0.15408268123865126,
      "num_tokens": 3301198.0,
      "step": 1795
    },
    {
      "entropy": 5.911142015457154,
      "epoch": 0.15035081857667892,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999497099726687,
      "loss": 5.8257,
      "mean_token_accuracy": 0.15286746770143508,
      "num_tokens": 3309983.0,
      "step": 1800
    },
    {
      "entropy": 6.056093406677246,
      "epoch": 0.1507684597393919,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999490786150039,
      "loss": 5.8947,
      "mean_token_accuracy": 0.14988958984613418,
      "num_tokens": 3319313.0,
      "step": 1805
    },
    {
      "entropy": 5.970861577987671,
      "epoch": 0.1511861009021049,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999484433193348,
      "loss": 5.7466,
      "mean_token_accuracy": 0.15159421861171724,
      "num_tokens": 3327497.0,
      "step": 1810
    },
    {
      "entropy": 5.922145032882691,
      "epoch": 0.15160374206481791,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999478040856722,
      "loss": 5.7622,
      "mean_token_accuracy": 0.15839407294988633,
      "num_tokens": 3337029.0,
      "step": 1815
    },
    {
      "entropy": 5.838753938674927,
      "epoch": 0.1520213832275309,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999471609140276,
      "loss": 5.7499,
      "mean_token_accuracy": 0.15191098004579545,
      "num_tokens": 3346243.0,
      "step": 1820
    },
    {
      "entropy": 6.137970495223999,
      "epoch": 0.1524390243902439,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999465138044122,
      "loss": 5.8839,
      "mean_token_accuracy": 0.14960521906614305,
      "num_tokens": 3354846.0,
      "step": 1825
    },
    {
      "entropy": 5.996419382095337,
      "epoch": 0.1528566655529569,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999458627568373,
      "loss": 5.8623,
      "mean_token_accuracy": 0.1439855642616749,
      "num_tokens": 3364683.0,
      "step": 1830
    },
    {
      "entropy": 6.020885181427002,
      "epoch": 0.15327430671566988,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999452077713144,
      "loss": 5.7839,
      "mean_token_accuracy": 0.15447492599487306,
      "num_tokens": 3374652.0,
      "step": 1835
    },
    {
      "entropy": 5.986085510253906,
      "epoch": 0.1536919478783829,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999445488478547,
      "loss": 5.8018,
      "mean_token_accuracy": 0.1501589760184288,
      "num_tokens": 3384040.0,
      "step": 1840
    },
    {
      "entropy": 5.89569411277771,
      "epoch": 0.1541095890410959,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00049994388598647,
      "loss": 5.8626,
      "mean_token_accuracy": 0.15379735231399536,
      "num_tokens": 3393185.0,
      "step": 1845
    },
    {
      "entropy": 5.9624189853668215,
      "epoch": 0.15452723020380887,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999432191871719,
      "loss": 5.8107,
      "mean_token_accuracy": 0.15432377755641938,
      "num_tokens": 3403077.0,
      "step": 1850
    },
    {
      "entropy": 5.947211456298828,
      "epoch": 0.15494487136652188,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999425484499718,
      "loss": 5.7819,
      "mean_token_accuracy": 0.16086574792861938,
      "num_tokens": 3412165.0,
      "step": 1855
    },
    {
      "entropy": 6.029496955871582,
      "epoch": 0.1553625125292349,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999418737748817,
      "loss": 5.8519,
      "mean_token_accuracy": 0.14368949756026267,
      "num_tokens": 3421162.0,
      "step": 1860
    },
    {
      "entropy": 6.066939353942871,
      "epoch": 0.15578015369194786,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999411951619134,
      "loss": 5.9092,
      "mean_token_accuracy": 0.146198970079422,
      "num_tokens": 3430363.0,
      "step": 1865
    },
    {
      "entropy": 6.006467866897583,
      "epoch": 0.15619779485466087,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999405126110786,
      "loss": 5.8189,
      "mean_token_accuracy": 0.14844847470521927,
      "num_tokens": 3439199.0,
      "step": 1870
    },
    {
      "entropy": 6.101349639892578,
      "epoch": 0.15661543601737388,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999398261223895,
      "loss": 5.9561,
      "mean_token_accuracy": 0.1441134199500084,
      "num_tokens": 3448720.0,
      "step": 1875
    },
    {
      "entropy": 6.027431869506836,
      "epoch": 0.15703307718008688,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999391356958579,
      "loss": 5.9127,
      "mean_token_accuracy": 0.1493662789463997,
      "num_tokens": 3458472.0,
      "step": 1880
    },
    {
      "entropy": 5.976059818267823,
      "epoch": 0.15745071834279986,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499938441331496,
      "loss": 5.8454,
      "mean_token_accuracy": 0.14676398932933807,
      "num_tokens": 3467140.0,
      "step": 1885
    },
    {
      "entropy": 6.097150707244873,
      "epoch": 0.15786835950551287,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999377430293159,
      "loss": 5.8452,
      "mean_token_accuracy": 0.15335095673799515,
      "num_tokens": 3476431.0,
      "step": 1890
    },
    {
      "entropy": 5.8885892868042,
      "epoch": 0.15828600066822587,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999370407893299,
      "loss": 5.7525,
      "mean_token_accuracy": 0.14893790632486342,
      "num_tokens": 3485768.0,
      "step": 1895
    },
    {
      "entropy": 6.066452074050903,
      "epoch": 0.15870364183093885,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999363346115501,
      "loss": 5.9003,
      "mean_token_accuracy": 0.14735231101512908,
      "num_tokens": 3495678.0,
      "step": 1900
    },
    {
      "entropy": 5.97677001953125,
      "epoch": 0.15912128299365186,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999356244959893,
      "loss": 5.881,
      "mean_token_accuracy": 0.1467648006975651,
      "num_tokens": 3505168.0,
      "step": 1905
    },
    {
      "entropy": 5.975177431106568,
      "epoch": 0.15953892415636486,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999349104426594,
      "loss": 5.8312,
      "mean_token_accuracy": 0.1447260484099388,
      "num_tokens": 3514498.0,
      "step": 1910
    },
    {
      "entropy": 6.0214598178863525,
      "epoch": 0.15995656531907784,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999341924515732,
      "loss": 5.7596,
      "mean_token_accuracy": 0.1468587689101696,
      "num_tokens": 3524617.0,
      "step": 1915
    },
    {
      "entropy": 5.928234100341797,
      "epoch": 0.16037420648179085,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999334705227431,
      "loss": 5.7413,
      "mean_token_accuracy": 0.15519107580184938,
      "num_tokens": 3533523.0,
      "step": 1920
    },
    {
      "entropy": 5.961508655548096,
      "epoch": 0.16079184764450385,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499932744656182,
      "loss": 5.8162,
      "mean_token_accuracy": 0.15407586544752122,
      "num_tokens": 3542922.0,
      "step": 1925
    },
    {
      "entropy": 6.024217510223389,
      "epoch": 0.16120948880721683,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999320148519024,
      "loss": 5.8079,
      "mean_token_accuracy": 0.1524025820195675,
      "num_tokens": 3551827.0,
      "step": 1930
    },
    {
      "entropy": 5.900309419631958,
      "epoch": 0.16162712996992984,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999312811099171,
      "loss": 5.7693,
      "mean_token_accuracy": 0.1520832046866417,
      "num_tokens": 3560911.0,
      "step": 1935
    },
    {
      "entropy": 5.964160490036011,
      "epoch": 0.16204477113264285,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999305434302391,
      "loss": 5.8033,
      "mean_token_accuracy": 0.14884971603751182,
      "num_tokens": 3569473.0,
      "step": 1940
    },
    {
      "entropy": 5.934193325042725,
      "epoch": 0.16246241229535582,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499929801812881,
      "loss": 5.7947,
      "mean_token_accuracy": 0.15742168873548507,
      "num_tokens": 3578843.0,
      "step": 1945
    },
    {
      "entropy": 5.9684545516967775,
      "epoch": 0.16288005345806883,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999290562578561,
      "loss": 5.8019,
      "mean_token_accuracy": 0.15691341012716292,
      "num_tokens": 3588304.0,
      "step": 1950
    },
    {
      "entropy": 5.96470685005188,
      "epoch": 0.16329769462078184,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999283067651773,
      "loss": 5.8696,
      "mean_token_accuracy": 0.14991875737905502,
      "num_tokens": 3598354.0,
      "step": 1955
    },
    {
      "entropy": 6.10321011543274,
      "epoch": 0.16371533578349481,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999275533348577,
      "loss": 5.9036,
      "mean_token_accuracy": 0.14993757382035255,
      "num_tokens": 3607503.0,
      "step": 1960
    },
    {
      "entropy": 5.818184232711792,
      "epoch": 0.16413297694620782,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999267959669106,
      "loss": 5.7108,
      "mean_token_accuracy": 0.1614347442984581,
      "num_tokens": 3617467.0,
      "step": 1965
    },
    {
      "entropy": 5.997748470306396,
      "epoch": 0.16455061810892083,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999260346613491,
      "loss": 5.7274,
      "mean_token_accuracy": 0.1477570079267025,
      "num_tokens": 3627048.0,
      "step": 1970
    },
    {
      "entropy": 5.856850242614746,
      "epoch": 0.1649682592716338,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999252694181867,
      "loss": 5.7536,
      "mean_token_accuracy": 0.15221063643693925,
      "num_tokens": 3636683.0,
      "step": 1975
    },
    {
      "entropy": 5.8793130874633786,
      "epoch": 0.1653859004343468,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999245002374367,
      "loss": 5.6541,
      "mean_token_accuracy": 0.16297252625226974,
      "num_tokens": 3645231.0,
      "step": 1980
    },
    {
      "entropy": 5.910471391677857,
      "epoch": 0.16580354159705982,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999237271191125,
      "loss": 5.7481,
      "mean_token_accuracy": 0.1537959724664688,
      "num_tokens": 3653851.0,
      "step": 1985
    },
    {
      "entropy": 5.851402044296265,
      "epoch": 0.1662211827597728,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999229500632279,
      "loss": 5.7256,
      "mean_token_accuracy": 0.15754058510065078,
      "num_tokens": 3662720.0,
      "step": 1990
    },
    {
      "entropy": 5.811029100418091,
      "epoch": 0.1666388239224858,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999221690697961,
      "loss": 5.6604,
      "mean_token_accuracy": 0.15830501317977905,
      "num_tokens": 3671777.0,
      "step": 1995
    },
    {
      "entropy": 5.900216579437256,
      "epoch": 0.1670564650851988,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999213841388312,
      "loss": 5.7487,
      "mean_token_accuracy": 0.16085124611854554,
      "num_tokens": 3681390.0,
      "step": 2000
    },
    {
      "entropy": 6.014535903930664,
      "epoch": 0.16747410624791179,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999205952703466,
      "loss": 5.8095,
      "mean_token_accuracy": 0.15208923071622849,
      "num_tokens": 3690899.0,
      "step": 2005
    },
    {
      "entropy": 5.994240045547485,
      "epoch": 0.1678917474106248,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999198024643563,
      "loss": 5.9006,
      "mean_token_accuracy": 0.14157435148954392,
      "num_tokens": 3699891.0,
      "step": 2010
    },
    {
      "entropy": 5.851044797897339,
      "epoch": 0.1683093885733378,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999190057208741,
      "loss": 5.682,
      "mean_token_accuracy": 0.1597720503807068,
      "num_tokens": 3709860.0,
      "step": 2015
    },
    {
      "entropy": 5.947671413421631,
      "epoch": 0.16872702973605078,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499918205039914,
      "loss": 5.8256,
      "mean_token_accuracy": 0.14998155385255812,
      "num_tokens": 3718896.0,
      "step": 2020
    },
    {
      "entropy": 5.87712082862854,
      "epoch": 0.16914467089876378,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999174004214899,
      "loss": 5.6315,
      "mean_token_accuracy": 0.157985882461071,
      "num_tokens": 3727355.0,
      "step": 2025
    },
    {
      "entropy": 5.853571224212646,
      "epoch": 0.1695623120614768,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999165918656161,
      "loss": 5.7242,
      "mean_token_accuracy": 0.1521446391940117,
      "num_tokens": 3736191.0,
      "step": 2030
    },
    {
      "entropy": 5.971621370315551,
      "epoch": 0.16997995322418977,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999157793723065,
      "loss": 5.7667,
      "mean_token_accuracy": 0.15560206323862075,
      "num_tokens": 3744832.0,
      "step": 2035
    },
    {
      "entropy": 5.861223936080933,
      "epoch": 0.17039759438690277,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999149629415755,
      "loss": 5.7308,
      "mean_token_accuracy": 0.152314256131649,
      "num_tokens": 3753835.0,
      "step": 2040
    },
    {
      "entropy": 6.030371236801147,
      "epoch": 0.17081523554961578,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999141425734374,
      "loss": 5.8026,
      "mean_token_accuracy": 0.15111032724380494,
      "num_tokens": 3762816.0,
      "step": 2045
    },
    {
      "entropy": 5.8637120723724365,
      "epoch": 0.17123287671232876,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999133182679064,
      "loss": 5.8427,
      "mean_token_accuracy": 0.14801040440797805,
      "num_tokens": 3773244.0,
      "step": 2050
    },
    {
      "entropy": 5.971320104598999,
      "epoch": 0.17165051787504176,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999124900249971,
      "loss": 5.7059,
      "mean_token_accuracy": 0.15795618146657944,
      "num_tokens": 3782122.0,
      "step": 2055
    },
    {
      "entropy": 5.919811820983886,
      "epoch": 0.17206815903775477,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999116578447239,
      "loss": 5.7008,
      "mean_token_accuracy": 0.16247085183858873,
      "num_tokens": 3791303.0,
      "step": 2060
    },
    {
      "entropy": 5.881378173828125,
      "epoch": 0.17248580020046775,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999108217271015,
      "loss": 5.7661,
      "mean_token_accuracy": 0.15498722046613694,
      "num_tokens": 3800991.0,
      "step": 2065
    },
    {
      "entropy": 5.8623046875,
      "epoch": 0.17290344136318075,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999099816721444,
      "loss": 5.6712,
      "mean_token_accuracy": 0.15415250658988952,
      "num_tokens": 3811293.0,
      "step": 2070
    },
    {
      "entropy": 6.0197525978088375,
      "epoch": 0.17332108252589376,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999091376798673,
      "loss": 5.7834,
      "mean_token_accuracy": 0.15429357066750526,
      "num_tokens": 3820013.0,
      "step": 2075
    },
    {
      "entropy": 5.867264175415039,
      "epoch": 0.17373872368860674,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999082897502849,
      "loss": 5.6794,
      "mean_token_accuracy": 0.15733524262905121,
      "num_tokens": 3829099.0,
      "step": 2080
    },
    {
      "entropy": 5.878523349761963,
      "epoch": 0.17415636485131974,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999074378834123,
      "loss": 5.7719,
      "mean_token_accuracy": 0.1538541316986084,
      "num_tokens": 3838308.0,
      "step": 2085
    },
    {
      "entropy": 5.929077816009522,
      "epoch": 0.17457400601403275,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999065820792644,
      "loss": 5.7144,
      "mean_token_accuracy": 0.16047724187374116,
      "num_tokens": 3847650.0,
      "step": 2090
    },
    {
      "entropy": 5.942487955093384,
      "epoch": 0.17499164717674573,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999057223378559,
      "loss": 5.9009,
      "mean_token_accuracy": 0.1422325260937214,
      "num_tokens": 3857252.0,
      "step": 2095
    },
    {
      "entropy": 5.95904016494751,
      "epoch": 0.17540928833945874,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499904858659202,
      "loss": 5.7534,
      "mean_token_accuracy": 0.1525118112564087,
      "num_tokens": 3865747.0,
      "step": 2100
    },
    {
      "entropy": 5.90725622177124,
      "epoch": 0.17582692950217174,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999039910433179,
      "loss": 5.7229,
      "mean_token_accuracy": 0.1620723456144333,
      "num_tokens": 3874958.0,
      "step": 2105
    },
    {
      "entropy": 5.7633425235748295,
      "epoch": 0.17624457066488472,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999031194902187,
      "loss": 5.661,
      "mean_token_accuracy": 0.16675298810005187,
      "num_tokens": 3885216.0,
      "step": 2110
    },
    {
      "entropy": 5.845773363113404,
      "epoch": 0.17666221182759773,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999022439999198,
      "loss": 5.7095,
      "mean_token_accuracy": 0.16534354239702226,
      "num_tokens": 3893973.0,
      "step": 2115
    },
    {
      "entropy": 5.968949317932129,
      "epoch": 0.17707985299031073,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999013645724362,
      "loss": 5.7955,
      "mean_token_accuracy": 0.15663563907146455,
      "num_tokens": 3903532.0,
      "step": 2120
    },
    {
      "entropy": 5.935855197906494,
      "epoch": 0.1774974941530237,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999004812077836,
      "loss": 5.723,
      "mean_token_accuracy": 0.1600687876343727,
      "num_tokens": 3913427.0,
      "step": 2125
    },
    {
      "entropy": 5.915048122406006,
      "epoch": 0.17791513531573672,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004998995939059773,
      "loss": 5.8404,
      "mean_token_accuracy": 0.14371922314167024,
      "num_tokens": 3923870.0,
      "step": 2130
    },
    {
      "entropy": 5.892444753646851,
      "epoch": 0.17833277647844972,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499898702667033,
      "loss": 5.6634,
      "mean_token_accuracy": 0.16210781186819076,
      "num_tokens": 3933469.0,
      "step": 2135
    },
    {
      "entropy": 5.829478549957275,
      "epoch": 0.1787504176411627,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998978074909661,
      "loss": 5.6856,
      "mean_token_accuracy": 0.16462984681129456,
      "num_tokens": 3942427.0,
      "step": 2140
    },
    {
      "entropy": 5.888026475906372,
      "epoch": 0.1791680588038757,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998969083777925,
      "loss": 5.731,
      "mean_token_accuracy": 0.16469569504261017,
      "num_tokens": 3951250.0,
      "step": 2145
    },
    {
      "entropy": 5.877253484725952,
      "epoch": 0.1795856999665887,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998960053275277,
      "loss": 5.7122,
      "mean_token_accuracy": 0.15940521210432052,
      "num_tokens": 3959969.0,
      "step": 2150
    },
    {
      "entropy": 5.904099845886231,
      "epoch": 0.1800033411293017,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998950983401875,
      "loss": 5.7457,
      "mean_token_accuracy": 0.15853157490491868,
      "num_tokens": 3968171.0,
      "step": 2155
    },
    {
      "entropy": 5.909996271133423,
      "epoch": 0.1804209822920147,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998941874157882,
      "loss": 5.652,
      "mean_token_accuracy": 0.16395194977521896,
      "num_tokens": 3976686.0,
      "step": 2160
    },
    {
      "entropy": 5.807675743103028,
      "epoch": 0.1808386234547277,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004998932725543453,
      "loss": 5.7738,
      "mean_token_accuracy": 0.15376742482185363,
      "num_tokens": 3986552.0,
      "step": 2165
    },
    {
      "entropy": 5.893441867828369,
      "epoch": 0.18125626461744068,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998923537558749,
      "loss": 5.6375,
      "mean_token_accuracy": 0.15640538930892944,
      "num_tokens": 3995501.0,
      "step": 2170
    },
    {
      "entropy": 5.737188863754272,
      "epoch": 0.1816739057801537,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998914310203932,
      "loss": 5.6072,
      "mean_token_accuracy": 0.16420529335737227,
      "num_tokens": 4004423.0,
      "step": 2175
    },
    {
      "entropy": 5.930015420913696,
      "epoch": 0.1820915469428667,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998905043479163,
      "loss": 5.8099,
      "mean_token_accuracy": 0.153107500821352,
      "num_tokens": 4013324.0,
      "step": 2180
    },
    {
      "entropy": 5.915607213973999,
      "epoch": 0.18250918810557967,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998895737384605,
      "loss": 5.7165,
      "mean_token_accuracy": 0.15703777819871903,
      "num_tokens": 4021896.0,
      "step": 2185
    },
    {
      "entropy": 5.81768684387207,
      "epoch": 0.18292682926829268,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998886391920418,
      "loss": 5.6943,
      "mean_token_accuracy": 0.1572074055671692,
      "num_tokens": 4030590.0,
      "step": 2190
    },
    {
      "entropy": 5.833494710922241,
      "epoch": 0.18334447043100568,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998877007086769,
      "loss": 5.5793,
      "mean_token_accuracy": 0.16950051933526994,
      "num_tokens": 4040041.0,
      "step": 2195
    },
    {
      "entropy": 5.832726621627808,
      "epoch": 0.18376211159371866,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998867582883821,
      "loss": 5.7421,
      "mean_token_accuracy": 0.15105979070067405,
      "num_tokens": 4049119.0,
      "step": 2200
    },
    {
      "entropy": 5.889599370956421,
      "epoch": 0.18417975275643167,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998858119311739,
      "loss": 5.6816,
      "mean_token_accuracy": 0.15757090598344803,
      "num_tokens": 4059265.0,
      "step": 2205
    },
    {
      "entropy": 5.868921375274658,
      "epoch": 0.18459739391914468,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998848616370689,
      "loss": 5.7303,
      "mean_token_accuracy": 0.15591793954372407,
      "num_tokens": 4068363.0,
      "step": 2210
    },
    {
      "entropy": 5.834519529342652,
      "epoch": 0.18501503508185768,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998839074060837,
      "loss": 5.6149,
      "mean_token_accuracy": 0.1670417383313179,
      "num_tokens": 4077390.0,
      "step": 2215
    },
    {
      "entropy": 5.817612791061402,
      "epoch": 0.18543267624457066,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499882949238235,
      "loss": 5.6919,
      "mean_token_accuracy": 0.15247385650873185,
      "num_tokens": 4087331.0,
      "step": 2220
    },
    {
      "entropy": 5.97557053565979,
      "epoch": 0.18585031740728367,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998819871335396,
      "loss": 5.8788,
      "mean_token_accuracy": 0.14540984779596328,
      "num_tokens": 4096425.0,
      "step": 2225
    },
    {
      "entropy": 5.819278335571289,
      "epoch": 0.18626795856999667,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998810210920142,
      "loss": 5.6527,
      "mean_token_accuracy": 0.157076658308506,
      "num_tokens": 4106056.0,
      "step": 2230
    },
    {
      "entropy": 5.835240602493286,
      "epoch": 0.18668559973270965,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998800511136759,
      "loss": 5.7259,
      "mean_token_accuracy": 0.15631824880838394,
      "num_tokens": 4115133.0,
      "step": 2235
    },
    {
      "entropy": 5.864358139038086,
      "epoch": 0.18710324089542266,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998790771985417,
      "loss": 5.646,
      "mean_token_accuracy": 0.15951664745807648,
      "num_tokens": 4123630.0,
      "step": 2240
    },
    {
      "entropy": 5.7766501903533936,
      "epoch": 0.18752088205813566,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998780993466284,
      "loss": 5.7808,
      "mean_token_accuracy": 0.15915831923484802,
      "num_tokens": 4133337.0,
      "step": 2245
    },
    {
      "entropy": 5.940265655517578,
      "epoch": 0.18793852322084864,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998771175579534,
      "loss": 5.6988,
      "mean_token_accuracy": 0.16006659865379333,
      "num_tokens": 4142240.0,
      "step": 2250
    },
    {
      "entropy": 5.763757801055908,
      "epoch": 0.18835616438356165,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004998761318325337,
      "loss": 5.6857,
      "mean_token_accuracy": 0.15862759053707123,
      "num_tokens": 4151297.0,
      "step": 2255
    },
    {
      "entropy": 5.8735432624816895,
      "epoch": 0.18877380554627465,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998751421703868,
      "loss": 5.6468,
      "mean_token_accuracy": 0.16166311353445054,
      "num_tokens": 4160193.0,
      "step": 2260
    },
    {
      "entropy": 5.8956348419189455,
      "epoch": 0.18919144670898763,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998741485715297,
      "loss": 5.7878,
      "mean_token_accuracy": 0.15754115879535674,
      "num_tokens": 4169948.0,
      "step": 2265
    },
    {
      "entropy": 6.02105131149292,
      "epoch": 0.18960908787170064,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00049987315103598,
      "loss": 5.8139,
      "mean_token_accuracy": 0.1528298608958721,
      "num_tokens": 4179981.0,
      "step": 2270
    },
    {
      "entropy": 5.867388772964477,
      "epoch": 0.19002672903441364,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998721495637551,
      "loss": 5.8346,
      "mean_token_accuracy": 0.1540336161851883,
      "num_tokens": 4189172.0,
      "step": 2275
    },
    {
      "entropy": 5.91773419380188,
      "epoch": 0.19044437019712662,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998711441548726,
      "loss": 5.6361,
      "mean_token_accuracy": 0.15736496299505234,
      "num_tokens": 4197312.0,
      "step": 2280
    },
    {
      "entropy": 5.889427852630615,
      "epoch": 0.19086201135983963,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998701348093499,
      "loss": 5.6907,
      "mean_token_accuracy": 0.16210058331489563,
      "num_tokens": 4206828.0,
      "step": 2285
    },
    {
      "entropy": 5.964065837860107,
      "epoch": 0.19127965252255263,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499869121527205,
      "loss": 5.6598,
      "mean_token_accuracy": 0.15945481210947038,
      "num_tokens": 4214835.0,
      "step": 2290
    },
    {
      "entropy": 5.850011205673217,
      "epoch": 0.1916972936852656,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998681043084553,
      "loss": 5.8345,
      "mean_token_accuracy": 0.15231008231639862,
      "num_tokens": 4225119.0,
      "step": 2295
    },
    {
      "entropy": 5.758636331558227,
      "epoch": 0.19211493484797862,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998670831531188,
      "loss": 5.6729,
      "mean_token_accuracy": 0.1581309109926224,
      "num_tokens": 4235257.0,
      "step": 2300
    },
    {
      "entropy": 5.886775064468384,
      "epoch": 0.19253257601069163,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004998660580612135,
      "loss": 5.7484,
      "mean_token_accuracy": 0.1589869663119316,
      "num_tokens": 4244211.0,
      "step": 2305
    },
    {
      "entropy": 5.924040460586548,
      "epoch": 0.1929502171734046,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998650290327569,
      "loss": 5.7565,
      "mean_token_accuracy": 0.1577833667397499,
      "num_tokens": 4253149.0,
      "step": 2310
    },
    {
      "entropy": 5.840446710586548,
      "epoch": 0.1933678583361176,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998639960677675,
      "loss": 5.6477,
      "mean_token_accuracy": 0.1637775108218193,
      "num_tokens": 4262557.0,
      "step": 2315
    },
    {
      "entropy": 5.783983182907105,
      "epoch": 0.19378549949883062,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998629591662631,
      "loss": 5.6643,
      "mean_token_accuracy": 0.16009251475334169,
      "num_tokens": 4272881.0,
      "step": 2320
    },
    {
      "entropy": 5.816336631774902,
      "epoch": 0.1942031406615436,
      "grad_norm": 1.09375,
      "learning_rate": 0.000499861918328262,
      "loss": 5.6766,
      "mean_token_accuracy": 0.16216119825839997,
      "num_tokens": 4281158.0,
      "step": 2325
    },
    {
      "entropy": 5.818071556091309,
      "epoch": 0.1946207818242566,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998608735537822,
      "loss": 5.6633,
      "mean_token_accuracy": 0.15752760022878648,
      "num_tokens": 4290319.0,
      "step": 2330
    },
    {
      "entropy": 5.862519454956055,
      "epoch": 0.1950384229869696,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998598248428423,
      "loss": 5.7436,
      "mean_token_accuracy": 0.16369648277759552,
      "num_tokens": 4299276.0,
      "step": 2335
    },
    {
      "entropy": 5.846491718292237,
      "epoch": 0.19545606414968258,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998587721954604,
      "loss": 5.7291,
      "mean_token_accuracy": 0.15878876075148582,
      "num_tokens": 4308291.0,
      "step": 2340
    },
    {
      "entropy": 5.825134420394898,
      "epoch": 0.1958737053123956,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998577156116551,
      "loss": 5.6619,
      "mean_token_accuracy": 0.1588639385998249,
      "num_tokens": 4316654.0,
      "step": 2345
    },
    {
      "entropy": 5.863021326065064,
      "epoch": 0.1962913464751086,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998566550914448,
      "loss": 5.6168,
      "mean_token_accuracy": 0.1598825752735138,
      "num_tokens": 4325911.0,
      "step": 2350
    },
    {
      "entropy": 5.84814043045044,
      "epoch": 0.19670898763782158,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998555906348481,
      "loss": 5.8069,
      "mean_token_accuracy": 0.1560124486684799,
      "num_tokens": 4334846.0,
      "step": 2355
    },
    {
      "entropy": 5.906886148452759,
      "epoch": 0.19712662880053458,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998545222418836,
      "loss": 5.6736,
      "mean_token_accuracy": 0.15957642048597337,
      "num_tokens": 4343728.0,
      "step": 2360
    },
    {
      "entropy": 5.83021650314331,
      "epoch": 0.1975442699632476,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998534499125701,
      "loss": 5.6578,
      "mean_token_accuracy": 0.1660468950867653,
      "num_tokens": 4352774.0,
      "step": 2365
    },
    {
      "entropy": 5.949106073379516,
      "epoch": 0.19796191112596057,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998523736469263,
      "loss": 5.8009,
      "mean_token_accuracy": 0.1570994645357132,
      "num_tokens": 4361663.0,
      "step": 2370
    },
    {
      "entropy": 5.815302324295044,
      "epoch": 0.19837955228867357,
      "grad_norm": 1.0,
      "learning_rate": 0.000499851293444971,
      "loss": 5.6353,
      "mean_token_accuracy": 0.16739457994699478,
      "num_tokens": 4370278.0,
      "step": 2375
    },
    {
      "entropy": 5.9134697914123535,
      "epoch": 0.19879719345138658,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998502093067231,
      "loss": 5.6912,
      "mean_token_accuracy": 0.16134303212165832,
      "num_tokens": 4379010.0,
      "step": 2380
    },
    {
      "entropy": 5.838262128829956,
      "epoch": 0.19921483461409956,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998491212322017,
      "loss": 5.663,
      "mean_token_accuracy": 0.16272303760051726,
      "num_tokens": 4388145.0,
      "step": 2385
    },
    {
      "entropy": 5.780917119979859,
      "epoch": 0.19963247577681256,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998480292214258,
      "loss": 5.806,
      "mean_token_accuracy": 0.14910988062620162,
      "num_tokens": 4396321.0,
      "step": 2390
    },
    {
      "entropy": 5.903149604797363,
      "epoch": 0.20005011693952557,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998469332744144,
      "loss": 5.5749,
      "mean_token_accuracy": 0.16502940952777861,
      "num_tokens": 4404854.0,
      "step": 2395
    },
    {
      "entropy": 5.642773008346557,
      "epoch": 0.20046775810223855,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998458333911869,
      "loss": 5.5776,
      "mean_token_accuracy": 0.1670069932937622,
      "num_tokens": 4413890.0,
      "step": 2400
    },
    {
      "entropy": 5.862590408325195,
      "epoch": 0.20088539926495155,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998447295717624,
      "loss": 5.7323,
      "mean_token_accuracy": 0.157737997174263,
      "num_tokens": 4423019.0,
      "step": 2405
    },
    {
      "entropy": 5.81047739982605,
      "epoch": 0.20130304042766456,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998436218161602,
      "loss": 5.6565,
      "mean_token_accuracy": 0.16053846180438996,
      "num_tokens": 4432340.0,
      "step": 2410
    },
    {
      "entropy": 5.810161399841308,
      "epoch": 0.20172068159037754,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998425101243999,
      "loss": 5.6375,
      "mean_token_accuracy": 0.16139356642961503,
      "num_tokens": 4441834.0,
      "step": 2415
    },
    {
      "entropy": 5.797228527069092,
      "epoch": 0.20213832275309054,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998413944965007,
      "loss": 5.6622,
      "mean_token_accuracy": 0.1592136040329933,
      "num_tokens": 4450864.0,
      "step": 2420
    },
    {
      "entropy": 5.749743509292602,
      "epoch": 0.20255596391580355,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998402749324823,
      "loss": 5.7345,
      "mean_token_accuracy": 0.1552075505256653,
      "num_tokens": 4460697.0,
      "step": 2425
    },
    {
      "entropy": 6.035622787475586,
      "epoch": 0.20297360507851653,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998391514323642,
      "loss": 5.7444,
      "mean_token_accuracy": 0.16117378175258637,
      "num_tokens": 4469926.0,
      "step": 2430
    },
    {
      "entropy": 5.734647989273071,
      "epoch": 0.20339124624122953,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998380239961661,
      "loss": 5.6838,
      "mean_token_accuracy": 0.16252014935016632,
      "num_tokens": 4478818.0,
      "step": 2435
    },
    {
      "entropy": 5.8350136280059814,
      "epoch": 0.20380888740394254,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998368926239078,
      "loss": 5.6176,
      "mean_token_accuracy": 0.1623082473874092,
      "num_tokens": 4487451.0,
      "step": 2440
    },
    {
      "entropy": 5.804310703277588,
      "epoch": 0.20422652856665552,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499835757315609,
      "loss": 5.6839,
      "mean_token_accuracy": 0.15567411631345748,
      "num_tokens": 4497674.0,
      "step": 2445
    },
    {
      "entropy": 5.778557586669922,
      "epoch": 0.20464416972936852,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998346180712898,
      "loss": 5.6422,
      "mean_token_accuracy": 0.1650843396782875,
      "num_tokens": 4507666.0,
      "step": 2450
    },
    {
      "entropy": 5.834744691848755,
      "epoch": 0.20506181089208153,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998334748909698,
      "loss": 5.743,
      "mean_token_accuracy": 0.15409338921308519,
      "num_tokens": 4517409.0,
      "step": 2455
    },
    {
      "entropy": 5.8095001697540285,
      "epoch": 0.2054794520547945,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998323277746694,
      "loss": 5.5504,
      "mean_token_accuracy": 0.16385659277439119,
      "num_tokens": 4526495.0,
      "step": 2460
    },
    {
      "entropy": 5.77939510345459,
      "epoch": 0.20589709321750752,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998311767224081,
      "loss": 5.6284,
      "mean_token_accuracy": 0.16991629004478453,
      "num_tokens": 4535240.0,
      "step": 2465
    },
    {
      "entropy": 5.838644981384277,
      "epoch": 0.20631473438022052,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004998300217342067,
      "loss": 5.724,
      "mean_token_accuracy": 0.15681551843881608,
      "num_tokens": 4544716.0,
      "step": 2470
    },
    {
      "entropy": 5.715267181396484,
      "epoch": 0.2067323755429335,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998288628100851,
      "loss": 5.6474,
      "mean_token_accuracy": 0.162817020714283,
      "num_tokens": 4553867.0,
      "step": 2475
    },
    {
      "entropy": 5.868233108520508,
      "epoch": 0.2071500167056465,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998276999500636,
      "loss": 5.7037,
      "mean_token_accuracy": 0.15431201457977295,
      "num_tokens": 4562529.0,
      "step": 2480
    },
    {
      "entropy": 5.715342092514038,
      "epoch": 0.2075676578683595,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998265331541627,
      "loss": 5.4993,
      "mean_token_accuracy": 0.17513344287872315,
      "num_tokens": 4571041.0,
      "step": 2485
    },
    {
      "entropy": 5.711450433731079,
      "epoch": 0.2079852990310725,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998253624224026,
      "loss": 5.6309,
      "mean_token_accuracy": 0.16020862385630608,
      "num_tokens": 4579914.0,
      "step": 2490
    },
    {
      "entropy": 5.813286972045899,
      "epoch": 0.2084029401937855,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000499824187754804,
      "loss": 5.7104,
      "mean_token_accuracy": 0.1612517148256302,
      "num_tokens": 4590277.0,
      "step": 2495
    },
    {
      "entropy": 5.822319984436035,
      "epoch": 0.2088205813564985,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998230091513872,
      "loss": 5.6216,
      "mean_token_accuracy": 0.1658724918961525,
      "num_tokens": 4598719.0,
      "step": 2500
    },
    {
      "entropy": 5.7216949462890625,
      "epoch": 0.20923822251921148,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998218266121731,
      "loss": 5.5871,
      "mean_token_accuracy": 0.16493814885616304,
      "num_tokens": 4607804.0,
      "step": 2505
    },
    {
      "entropy": 5.74538516998291,
      "epoch": 0.2096558636819245,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998206401371823,
      "loss": 5.573,
      "mean_token_accuracy": 0.16702073365449904,
      "num_tokens": 4616529.0,
      "step": 2510
    },
    {
      "entropy": 5.838421297073364,
      "epoch": 0.2100735048446375,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998194497264357,
      "loss": 5.6378,
      "mean_token_accuracy": 0.15869868099689483,
      "num_tokens": 4624545.0,
      "step": 2515
    },
    {
      "entropy": 5.782481956481933,
      "epoch": 0.21049114600735047,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998182553799538,
      "loss": 5.6006,
      "mean_token_accuracy": 0.15949173867702485,
      "num_tokens": 4633316.0,
      "step": 2520
    },
    {
      "entropy": 5.748195505142212,
      "epoch": 0.21090878717006348,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998170570977579,
      "loss": 5.5587,
      "mean_token_accuracy": 0.1721911072731018,
      "num_tokens": 4643353.0,
      "step": 2525
    },
    {
      "entropy": 5.747497320175171,
      "epoch": 0.21132642833277648,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998158548798687,
      "loss": 5.6494,
      "mean_token_accuracy": 0.15524984747171403,
      "num_tokens": 4651998.0,
      "step": 2530
    },
    {
      "entropy": 5.837891101837158,
      "epoch": 0.21174406949548946,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998146487263074,
      "loss": 5.6515,
      "mean_token_accuracy": 0.16123034954071044,
      "num_tokens": 4659937.0,
      "step": 2535
    },
    {
      "entropy": 5.730093622207642,
      "epoch": 0.21216171065820247,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998134386370951,
      "loss": 5.6277,
      "mean_token_accuracy": 0.16407201513648034,
      "num_tokens": 4669396.0,
      "step": 2540
    },
    {
      "entropy": 5.881675386428833,
      "epoch": 0.21257935182091547,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004998122246122529,
      "loss": 5.7015,
      "mean_token_accuracy": 0.16837172955274582,
      "num_tokens": 4679320.0,
      "step": 2545
    },
    {
      "entropy": 5.7952516078948975,
      "epoch": 0.21299699298362845,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004998110066518021,
      "loss": 5.6161,
      "mean_token_accuracy": 0.16251329183578492,
      "num_tokens": 4688807.0,
      "step": 2550
    },
    {
      "entropy": 5.831697654724121,
      "epoch": 0.21341463414634146,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499809784755764,
      "loss": 5.62,
      "mean_token_accuracy": 0.17055489122867584,
      "num_tokens": 4697819.0,
      "step": 2555
    },
    {
      "entropy": 5.700474548339844,
      "epoch": 0.21383227530905446,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00049980855892416,
      "loss": 5.6051,
      "mean_token_accuracy": 0.16278748512268065,
      "num_tokens": 4706931.0,
      "step": 2560
    },
    {
      "entropy": 5.756153774261475,
      "epoch": 0.21424991647176747,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998073291570116,
      "loss": 5.5802,
      "mean_token_accuracy": 0.16360058337450029,
      "num_tokens": 4716262.0,
      "step": 2565
    },
    {
      "entropy": 5.81894998550415,
      "epoch": 0.21466755763448045,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998060954543404,
      "loss": 5.6063,
      "mean_token_accuracy": 0.16175924837589264,
      "num_tokens": 4725415.0,
      "step": 2570
    },
    {
      "entropy": 5.707076120376587,
      "epoch": 0.21508519879719346,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998048578161678,
      "loss": 5.6729,
      "mean_token_accuracy": 0.16455809623003007,
      "num_tokens": 4734580.0,
      "step": 2575
    },
    {
      "entropy": 5.7181299209594725,
      "epoch": 0.21550283995990646,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998036162425155,
      "loss": 5.4732,
      "mean_token_accuracy": 0.1736983448266983,
      "num_tokens": 4742966.0,
      "step": 2580
    },
    {
      "entropy": 5.801785516738891,
      "epoch": 0.21592048112261944,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998023707334055,
      "loss": 5.6369,
      "mean_token_accuracy": 0.1645279809832573,
      "num_tokens": 4751696.0,
      "step": 2585
    },
    {
      "entropy": 5.846903324127197,
      "epoch": 0.21633812228533245,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998011212888592,
      "loss": 5.6879,
      "mean_token_accuracy": 0.16777165681123735,
      "num_tokens": 4760460.0,
      "step": 2590
    },
    {
      "entropy": 5.780740785598755,
      "epoch": 0.21675576344804545,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004997998679088988,
      "loss": 5.6102,
      "mean_token_accuracy": 0.1645229384303093,
      "num_tokens": 4769004.0,
      "step": 2595
    },
    {
      "entropy": 5.757664060592651,
      "epoch": 0.21717340461075843,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997986105935461,
      "loss": 5.5898,
      "mean_token_accuracy": 0.16781664937734603,
      "num_tokens": 4777204.0,
      "step": 2600
    },
    {
      "entropy": 5.864018487930298,
      "epoch": 0.21759104577347144,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997973493428231,
      "loss": 5.7145,
      "mean_token_accuracy": 0.1602879449725151,
      "num_tokens": 4786766.0,
      "step": 2605
    },
    {
      "entropy": 5.663840627670288,
      "epoch": 0.21800868693618444,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997960841567519,
      "loss": 5.6135,
      "mean_token_accuracy": 0.16582195907831193,
      "num_tokens": 4796676.0,
      "step": 2610
    },
    {
      "entropy": 5.9396082878112795,
      "epoch": 0.21842632809889742,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004997948150353548,
      "loss": 5.6034,
      "mean_token_accuracy": 0.16275097280740738,
      "num_tokens": 4805514.0,
      "step": 2615
    },
    {
      "entropy": 5.718913602828979,
      "epoch": 0.21884396926161043,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997935419786537,
      "loss": 5.696,
      "mean_token_accuracy": 0.16018566042184829,
      "num_tokens": 4815487.0,
      "step": 2620
    },
    {
      "entropy": 5.88177261352539,
      "epoch": 0.21926161042432343,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004997922649866712,
      "loss": 5.6489,
      "mean_token_accuracy": 0.16259622126817702,
      "num_tokens": 4825044.0,
      "step": 2625
    },
    {
      "entropy": 5.689206600189209,
      "epoch": 0.2196792515870364,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997909840594294,
      "loss": 5.4412,
      "mean_token_accuracy": 0.17392415851354598,
      "num_tokens": 4833875.0,
      "step": 2630
    },
    {
      "entropy": 5.6562834739685055,
      "epoch": 0.22009689274974942,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997896991969509,
      "loss": 5.6608,
      "mean_token_accuracy": 0.15934335738420485,
      "num_tokens": 4843317.0,
      "step": 2635
    },
    {
      "entropy": 5.805779695510864,
      "epoch": 0.22051453391246242,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997884103992581,
      "loss": 5.6515,
      "mean_token_accuracy": 0.15986357778310775,
      "num_tokens": 4852514.0,
      "step": 2640
    },
    {
      "entropy": 5.787460565567017,
      "epoch": 0.2209321750751754,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997871176663736,
      "loss": 5.6123,
      "mean_token_accuracy": 0.16665420085191726,
      "num_tokens": 4861260.0,
      "step": 2645
    },
    {
      "entropy": 5.70442008972168,
      "epoch": 0.2213498162378884,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997858209983201,
      "loss": 5.5749,
      "mean_token_accuracy": 0.15970537588000297,
      "num_tokens": 4869654.0,
      "step": 2650
    },
    {
      "entropy": 5.780343818664551,
      "epoch": 0.22176745740060141,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997845203951202,
      "loss": 5.6071,
      "mean_token_accuracy": 0.15780386179685593,
      "num_tokens": 4878311.0,
      "step": 2655
    },
    {
      "entropy": 5.839015626907349,
      "epoch": 0.2221850985633144,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004997832158567967,
      "loss": 5.6874,
      "mean_token_accuracy": 0.1592348925769329,
      "num_tokens": 4887251.0,
      "step": 2660
    },
    {
      "entropy": 5.693773031234741,
      "epoch": 0.2226027397260274,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997819073833724,
      "loss": 5.5377,
      "mean_token_accuracy": 0.1633484482765198,
      "num_tokens": 4896528.0,
      "step": 2665
    },
    {
      "entropy": 5.7157878398895265,
      "epoch": 0.2230203808887404,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997805949748702,
      "loss": 5.603,
      "mean_token_accuracy": 0.16120654642581939,
      "num_tokens": 4904962.0,
      "step": 2670
    },
    {
      "entropy": 5.8058465957641605,
      "epoch": 0.22343802205145338,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997792786313132,
      "loss": 5.6042,
      "mean_token_accuracy": 0.15915837287902831,
      "num_tokens": 4914735.0,
      "step": 2675
    },
    {
      "entropy": 5.616079473495484,
      "epoch": 0.2238556632141664,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997779583527243,
      "loss": 5.4757,
      "mean_token_accuracy": 0.16390765011310576,
      "num_tokens": 4924271.0,
      "step": 2680
    },
    {
      "entropy": 5.869078683853149,
      "epoch": 0.2242733043768794,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997766341391268,
      "loss": 5.716,
      "mean_token_accuracy": 0.15845021605491638,
      "num_tokens": 4933928.0,
      "step": 2685
    },
    {
      "entropy": 5.7307792663574215,
      "epoch": 0.22469094553959237,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997753059905437,
      "loss": 5.4338,
      "mean_token_accuracy": 0.17364991903305055,
      "num_tokens": 4942714.0,
      "step": 2690
    },
    {
      "entropy": 5.661212921142578,
      "epoch": 0.22510858670230538,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997739739069981,
      "loss": 5.5491,
      "mean_token_accuracy": 0.16816597133874894,
      "num_tokens": 4952297.0,
      "step": 2695
    },
    {
      "entropy": 5.714229440689087,
      "epoch": 0.22552622786501839,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997726378885138,
      "loss": 5.4968,
      "mean_token_accuracy": 0.17623727917671203,
      "num_tokens": 4960947.0,
      "step": 2700
    },
    {
      "entropy": 5.757331514358521,
      "epoch": 0.22594386902773136,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004997712979351137,
      "loss": 5.6602,
      "mean_token_accuracy": 0.161906411498785,
      "num_tokens": 4969705.0,
      "step": 2705
    },
    {
      "entropy": 5.852696895599365,
      "epoch": 0.22636151019044437,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997699540468216,
      "loss": 5.6251,
      "mean_token_accuracy": 0.16277607381343842,
      "num_tokens": 4979112.0,
      "step": 2710
    },
    {
      "entropy": 5.718151617050171,
      "epoch": 0.22677915135315738,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997686062236609,
      "loss": 5.5491,
      "mean_token_accuracy": 0.16444555073976516,
      "num_tokens": 4987894.0,
      "step": 2715
    },
    {
      "entropy": 5.755270385742188,
      "epoch": 0.22719679251587035,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004997672544656553,
      "loss": 5.6945,
      "mean_token_accuracy": 0.15452539920806885,
      "num_tokens": 4997774.0,
      "step": 2720
    },
    {
      "entropy": 5.837758636474609,
      "epoch": 0.22761443367858336,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997658987728283,
      "loss": 5.6182,
      "mean_token_accuracy": 0.16429695934057237,
      "num_tokens": 5006183.0,
      "step": 2725
    },
    {
      "entropy": 5.711544799804687,
      "epoch": 0.22803207484129637,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997645391452035,
      "loss": 5.5984,
      "mean_token_accuracy": 0.1637036234140396,
      "num_tokens": 5016909.0,
      "step": 2730
    },
    {
      "entropy": 5.844042348861694,
      "epoch": 0.22844971600400935,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499763175582805,
      "loss": 5.6299,
      "mean_token_accuracy": 0.16498590111732483,
      "num_tokens": 5027326.0,
      "step": 2735
    },
    {
      "entropy": 5.746332120895386,
      "epoch": 0.22886735716672235,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004997618080856566,
      "loss": 5.597,
      "mean_token_accuracy": 0.1639745995402336,
      "num_tokens": 5037043.0,
      "step": 2740
    },
    {
      "entropy": 5.687271451950073,
      "epoch": 0.22928499832943536,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997604366537823,
      "loss": 5.5803,
      "mean_token_accuracy": 0.16922535002231598,
      "num_tokens": 5045992.0,
      "step": 2745
    },
    {
      "entropy": 5.6921813011169435,
      "epoch": 0.22970263949214834,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997590612872058,
      "loss": 5.5552,
      "mean_token_accuracy": 0.17040622532367705,
      "num_tokens": 5055568.0,
      "step": 2750
    },
    {
      "entropy": 5.785310697555542,
      "epoch": 0.23012028065486134,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004997576819859515,
      "loss": 5.634,
      "mean_token_accuracy": 0.16597387343645095,
      "num_tokens": 5065512.0,
      "step": 2755
    },
    {
      "entropy": 5.613603973388672,
      "epoch": 0.23053792181757435,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004997562987500434,
      "loss": 5.504,
      "mean_token_accuracy": 0.17271397709846498,
      "num_tokens": 5074776.0,
      "step": 2760
    },
    {
      "entropy": 5.682473802566529,
      "epoch": 0.23095556298028733,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997549115795058,
      "loss": 5.5364,
      "mean_token_accuracy": 0.16364094167947768,
      "num_tokens": 5084743.0,
      "step": 2765
    },
    {
      "entropy": 5.821541166305542,
      "epoch": 0.23137320414300033,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499753520474363,
      "loss": 5.698,
      "mean_token_accuracy": 0.16082392036914825,
      "num_tokens": 5095387.0,
      "step": 2770
    },
    {
      "entropy": 5.723036479949951,
      "epoch": 0.23179084530571334,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997521254346391,
      "loss": 5.621,
      "mean_token_accuracy": 0.16801968365907669,
      "num_tokens": 5104628.0,
      "step": 2775
    },
    {
      "entropy": 5.748641872406006,
      "epoch": 0.23220848646842632,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997507264603589,
      "loss": 5.5579,
      "mean_token_accuracy": 0.16903598308563234,
      "num_tokens": 5113169.0,
      "step": 2780
    },
    {
      "entropy": 5.735997152328491,
      "epoch": 0.23262612763113932,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997493235515466,
      "loss": 5.616,
      "mean_token_accuracy": 0.1675027921795845,
      "num_tokens": 5122294.0,
      "step": 2785
    },
    {
      "entropy": 5.68292818069458,
      "epoch": 0.23304376879385233,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997479167082269,
      "loss": 5.4565,
      "mean_token_accuracy": 0.17553396224975587,
      "num_tokens": 5131191.0,
      "step": 2790
    },
    {
      "entropy": 5.684044027328492,
      "epoch": 0.2334614099565653,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997465059304243,
      "loss": 5.6454,
      "mean_token_accuracy": 0.1675591215491295,
      "num_tokens": 5141974.0,
      "step": 2795
    },
    {
      "entropy": 5.787416267395019,
      "epoch": 0.23387905111927831,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004997450912181637,
      "loss": 5.593,
      "mean_token_accuracy": 0.16815385073423386,
      "num_tokens": 5151412.0,
      "step": 2800
    },
    {
      "entropy": 5.694646167755127,
      "epoch": 0.23429669228199132,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997436725714696,
      "loss": 5.5224,
      "mean_token_accuracy": 0.17085290551185608,
      "num_tokens": 5160989.0,
      "step": 2805
    },
    {
      "entropy": 5.750547647476196,
      "epoch": 0.2347143334447043,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997422499903671,
      "loss": 5.5962,
      "mean_token_accuracy": 0.16256070286035537,
      "num_tokens": 5170241.0,
      "step": 2810
    },
    {
      "entropy": 5.86253604888916,
      "epoch": 0.2351319746074173,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499740823474881,
      "loss": 5.7114,
      "mean_token_accuracy": 0.1583871603012085,
      "num_tokens": 5180247.0,
      "step": 2815
    },
    {
      "entropy": 5.691003131866455,
      "epoch": 0.2355496157701303,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997393930250363,
      "loss": 5.5975,
      "mean_token_accuracy": 0.16719801127910613,
      "num_tokens": 5189144.0,
      "step": 2820
    },
    {
      "entropy": 5.7188694953918455,
      "epoch": 0.2359672569328433,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499737958640858,
      "loss": 5.6503,
      "mean_token_accuracy": 0.163092702627182,
      "num_tokens": 5199410.0,
      "step": 2825
    },
    {
      "entropy": 5.776434516906738,
      "epoch": 0.2363848980955563,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004997365203223711,
      "loss": 5.5338,
      "mean_token_accuracy": 0.1717270389199257,
      "num_tokens": 5209032.0,
      "step": 2830
    },
    {
      "entropy": 5.60194149017334,
      "epoch": 0.2368025392582693,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499735078069601,
      "loss": 5.4409,
      "mean_token_accuracy": 0.17052695155143738,
      "num_tokens": 5217787.0,
      "step": 2835
    },
    {
      "entropy": 5.668485689163208,
      "epoch": 0.23722018042098228,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997336318825728,
      "loss": 5.5446,
      "mean_token_accuracy": 0.17174894362688065,
      "num_tokens": 5226962.0,
      "step": 2840
    },
    {
      "entropy": 5.678308296203613,
      "epoch": 0.23763782158369529,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004997321817613119,
      "loss": 5.5173,
      "mean_token_accuracy": 0.1678185984492302,
      "num_tokens": 5236111.0,
      "step": 2845
    },
    {
      "entropy": 5.770784091949463,
      "epoch": 0.2380554627464083,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004997307277058436,
      "loss": 5.6541,
      "mean_token_accuracy": 0.16095943748950958,
      "num_tokens": 5246628.0,
      "step": 2850
    },
    {
      "entropy": 5.723996591567993,
      "epoch": 0.23847310390912127,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997292697161935,
      "loss": 5.5896,
      "mean_token_accuracy": 0.16508384943008422,
      "num_tokens": 5255543.0,
      "step": 2855
    },
    {
      "entropy": 5.646565294265747,
      "epoch": 0.23889074507183428,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499727807792387,
      "loss": 5.4457,
      "mean_token_accuracy": 0.17628723680973052,
      "num_tokens": 5264559.0,
      "step": 2860
    },
    {
      "entropy": 5.608687114715576,
      "epoch": 0.23930838623454728,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004997263419344495,
      "loss": 5.5094,
      "mean_token_accuracy": 0.1655178412795067,
      "num_tokens": 5274599.0,
      "step": 2865
    },
    {
      "entropy": 5.7203874588012695,
      "epoch": 0.23972602739726026,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997248721424071,
      "loss": 5.4905,
      "mean_token_accuracy": 0.17523384392261504,
      "num_tokens": 5283495.0,
      "step": 2870
    },
    {
      "entropy": 5.695050477981567,
      "epoch": 0.24014366855997327,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997233984162853,
      "loss": 5.5477,
      "mean_token_accuracy": 0.17063757181167602,
      "num_tokens": 5292064.0,
      "step": 2875
    },
    {
      "entropy": 5.682350397109985,
      "epoch": 0.24056130972268627,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004997219207561099,
      "loss": 5.5498,
      "mean_token_accuracy": 0.16449775397777558,
      "num_tokens": 5301696.0,
      "step": 2880
    },
    {
      "entropy": 5.779582738876343,
      "epoch": 0.24097895088539925,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997204391619068,
      "loss": 5.6756,
      "mean_token_accuracy": 0.15998486280441285,
      "num_tokens": 5311427.0,
      "step": 2885
    },
    {
      "entropy": 5.861279916763306,
      "epoch": 0.24139659204811226,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997189536337018,
      "loss": 5.5943,
      "mean_token_accuracy": 0.165325590968132,
      "num_tokens": 5321406.0,
      "step": 2890
    },
    {
      "entropy": 5.737252998352051,
      "epoch": 0.24181423321082526,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004997174641715211,
      "loss": 5.6747,
      "mean_token_accuracy": 0.16726240664720535,
      "num_tokens": 5332257.0,
      "step": 2895
    },
    {
      "entropy": 5.6061920642852785,
      "epoch": 0.24223187437353827,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997159707753906,
      "loss": 5.4147,
      "mean_token_accuracy": 0.17394435107707978,
      "num_tokens": 5341857.0,
      "step": 2900
    },
    {
      "entropy": 5.674947690963745,
      "epoch": 0.24264951553625125,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997144734453367,
      "loss": 5.5556,
      "mean_token_accuracy": 0.1720260888338089,
      "num_tokens": 5351367.0,
      "step": 2905
    },
    {
      "entropy": 5.615740251541138,
      "epoch": 0.24306715669896425,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997129721813853,
      "loss": 5.409,
      "mean_token_accuracy": 0.1757777988910675,
      "num_tokens": 5360627.0,
      "step": 2910
    },
    {
      "entropy": 5.6760059833526615,
      "epoch": 0.24348479786167726,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997114669835629,
      "loss": 5.5008,
      "mean_token_accuracy": 0.1785655990242958,
      "num_tokens": 5369703.0,
      "step": 2915
    },
    {
      "entropy": 5.673794412612915,
      "epoch": 0.24390243902439024,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004997099578518959,
      "loss": 5.562,
      "mean_token_accuracy": 0.16899080127477645,
      "num_tokens": 5378432.0,
      "step": 2920
    },
    {
      "entropy": 5.56113076210022,
      "epoch": 0.24432008018710324,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004997084447864105,
      "loss": 5.4234,
      "mean_token_accuracy": 0.1710927039384842,
      "num_tokens": 5386489.0,
      "step": 2925
    },
    {
      "entropy": 5.833731555938721,
      "epoch": 0.24473772134981625,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997069277871333,
      "loss": 5.6423,
      "mean_token_accuracy": 0.16405942440032958,
      "num_tokens": 5395908.0,
      "step": 2930
    },
    {
      "entropy": 5.775952911376953,
      "epoch": 0.24515536251252923,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997054068540908,
      "loss": 5.5536,
      "mean_token_accuracy": 0.16982482224702836,
      "num_tokens": 5404822.0,
      "step": 2935
    },
    {
      "entropy": 5.635651350021362,
      "epoch": 0.24557300367524224,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004997038819873098,
      "loss": 5.4707,
      "mean_token_accuracy": 0.17467086464166642,
      "num_tokens": 5413980.0,
      "step": 2940
    },
    {
      "entropy": 5.684035778045654,
      "epoch": 0.24599064483795524,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997023531868167,
      "loss": 5.5357,
      "mean_token_accuracy": 0.16595565229654313,
      "num_tokens": 5423538.0,
      "step": 2945
    },
    {
      "entropy": 5.622634029388427,
      "epoch": 0.24640828600066822,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997008204526386,
      "loss": 5.4702,
      "mean_token_accuracy": 0.1707819879055023,
      "num_tokens": 5431930.0,
      "step": 2950
    },
    {
      "entropy": 5.666269826889038,
      "epoch": 0.24682592716338123,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004996992837848021,
      "loss": 5.4754,
      "mean_token_accuracy": 0.17090596705675126,
      "num_tokens": 5440721.0,
      "step": 2955
    },
    {
      "entropy": 5.596457624435425,
      "epoch": 0.24724356832609423,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004996977431833342,
      "loss": 5.5369,
      "mean_token_accuracy": 0.16840151995420455,
      "num_tokens": 5449617.0,
      "step": 2960
    },
    {
      "entropy": 5.668986415863037,
      "epoch": 0.2476612094888072,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996961986482618,
      "loss": 5.5755,
      "mean_token_accuracy": 0.16662422716617584,
      "num_tokens": 5458827.0,
      "step": 2965
    },
    {
      "entropy": 5.7311032772064205,
      "epoch": 0.24807885065152022,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996946501796119,
      "loss": 5.4595,
      "mean_token_accuracy": 0.17484788447618485,
      "num_tokens": 5467348.0,
      "step": 2970
    },
    {
      "entropy": 5.71124677658081,
      "epoch": 0.24849649181423322,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996930977774118,
      "loss": 5.5207,
      "mean_token_accuracy": 0.15785626620054244,
      "num_tokens": 5476645.0,
      "step": 2975
    },
    {
      "entropy": 5.715850973129273,
      "epoch": 0.2489141329769462,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004996915414416885,
      "loss": 5.57,
      "mean_token_accuracy": 0.16464744806289672,
      "num_tokens": 5486120.0,
      "step": 2980
    },
    {
      "entropy": 5.771138858795166,
      "epoch": 0.2493317741396592,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004996899811724693,
      "loss": 5.5541,
      "mean_token_accuracy": 0.16355062425136566,
      "num_tokens": 5495313.0,
      "step": 2985
    },
    {
      "entropy": 5.665059328079224,
      "epoch": 0.2497494153023722,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004996884169697816,
      "loss": 5.5356,
      "mean_token_accuracy": 0.17020975649356843,
      "num_tokens": 5504587.0,
      "step": 2990
    },
    {
      "entropy": 5.6389179706573485,
      "epoch": 0.2501670564650852,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004996868488336526,
      "loss": 5.5605,
      "mean_token_accuracy": 0.16468265056610107,
      "num_tokens": 5514188.0,
      "step": 2995
    },
    {
      "entropy": 5.653175640106201,
      "epoch": 0.25058469762779817,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996852767641099,
      "loss": 5.6088,
      "mean_token_accuracy": 0.16206363141536712,
      "num_tokens": 5524075.0,
      "step": 3000
    },
    {
      "epoch": 0.25058469762779817,
      "eval_entropy": 5.512934588497156,
      "eval_loss": 5.5222487449646,
      "eval_mean_token_accuracy": 0.17784448021585295,
      "eval_num_tokens": 5524075.0,
      "eval_runtime": 21.1176,
      "eval_samples_per_second": 1788.224,
      "eval_steps_per_second": 223.558,
      "step": 3000
    },
    {
      "entropy": 5.6670747756958,
      "epoch": 0.2510023387905112,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004996837007611811,
      "loss": 5.3743,
      "mean_token_accuracy": 0.17157501429319383,
      "num_tokens": 5533323.0,
      "step": 3005
    },
    {
      "entropy": 5.623773765563965,
      "epoch": 0.2514199799532242,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004996821208248935,
      "loss": 5.4096,
      "mean_token_accuracy": 0.17664626240730286,
      "num_tokens": 5542288.0,
      "step": 3010
    },
    {
      "entropy": 5.706943416595459,
      "epoch": 0.25183762111593716,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004996805369552749,
      "loss": 5.5772,
      "mean_token_accuracy": 0.15779562294483185,
      "num_tokens": 5550869.0,
      "step": 3015
    },
    {
      "entropy": 5.663392162322998,
      "epoch": 0.2522552622786502,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996789491523532,
      "loss": 5.4893,
      "mean_token_accuracy": 0.1784336343407631,
      "num_tokens": 5560147.0,
      "step": 3020
    },
    {
      "entropy": 5.582896566390991,
      "epoch": 0.2526729034413632,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004996773574161559,
      "loss": 5.3962,
      "mean_token_accuracy": 0.17424482554197313,
      "num_tokens": 5568358.0,
      "step": 3025
    },
    {
      "entropy": 5.678020191192627,
      "epoch": 0.25309054460407615,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004996757617467112,
      "loss": 5.4874,
      "mean_token_accuracy": 0.16868007332086563,
      "num_tokens": 5577755.0,
      "step": 3030
    },
    {
      "entropy": 5.69749083518982,
      "epoch": 0.2535081857667892,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004996741621440466,
      "loss": 5.5425,
      "mean_token_accuracy": 0.15935412645339966,
      "num_tokens": 5587515.0,
      "step": 3035
    },
    {
      "entropy": 5.710214376449585,
      "epoch": 0.25392582692950216,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004996725586081906,
      "loss": 5.6747,
      "mean_token_accuracy": 0.15729862898588182,
      "num_tokens": 5597860.0,
      "step": 3040
    },
    {
      "entropy": 5.751253366470337,
      "epoch": 0.25434346809221514,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004996709511391709,
      "loss": 5.5439,
      "mean_token_accuracy": 0.16929943710565568,
      "num_tokens": 5607302.0,
      "step": 3045
    },
    {
      "entropy": 5.693378543853759,
      "epoch": 0.2547611092549282,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004996693397370158,
      "loss": 5.6059,
      "mean_token_accuracy": 0.17057730555534362,
      "num_tokens": 5616345.0,
      "step": 3050
    },
    {
      "entropy": 5.669316959381104,
      "epoch": 0.25517875041764115,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004996677244017534,
      "loss": 5.4216,
      "mean_token_accuracy": 0.17369939684867858,
      "num_tokens": 5624745.0,
      "step": 3055
    },
    {
      "entropy": 5.658686542510987,
      "epoch": 0.25559639158035413,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996661051334121,
      "loss": 5.5359,
      "mean_token_accuracy": 0.16642606854438782,
      "num_tokens": 5634022.0,
      "step": 3060
    },
    {
      "entropy": 5.676845598220825,
      "epoch": 0.25601403274306717,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004996644819320201,
      "loss": 5.477,
      "mean_token_accuracy": 0.17220388054847718,
      "num_tokens": 5643631.0,
      "step": 3065
    },
    {
      "entropy": 5.69556360244751,
      "epoch": 0.25643167390578014,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004996628547976061,
      "loss": 5.5774,
      "mean_token_accuracy": 0.1674746036529541,
      "num_tokens": 5653155.0,
      "step": 3070
    },
    {
      "entropy": 5.675973224639892,
      "epoch": 0.2568493150684932,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004996612237301983,
      "loss": 5.5595,
      "mean_token_accuracy": 0.17027886658906938,
      "num_tokens": 5663091.0,
      "step": 3075
    },
    {
      "entropy": 5.5919482707977295,
      "epoch": 0.25726695623120616,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004996595887298254,
      "loss": 5.4637,
      "mean_token_accuracy": 0.1782810017466545,
      "num_tokens": 5671885.0,
      "step": 3080
    },
    {
      "entropy": 5.6924995422363285,
      "epoch": 0.25768459739391913,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499657949796516,
      "loss": 5.4859,
      "mean_token_accuracy": 0.1723874107003212,
      "num_tokens": 5680225.0,
      "step": 3085
    },
    {
      "entropy": 5.531622838973999,
      "epoch": 0.25810223855663217,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004996563069302986,
      "loss": 5.3463,
      "mean_token_accuracy": 0.1835720121860504,
      "num_tokens": 5689456.0,
      "step": 3090
    },
    {
      "entropy": 5.613016796112061,
      "epoch": 0.25851987971934515,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004996546601312023,
      "loss": 5.4715,
      "mean_token_accuracy": 0.17324957102537156,
      "num_tokens": 5698127.0,
      "step": 3095
    },
    {
      "entropy": 5.575536108016967,
      "epoch": 0.2589375208820581,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996530093992558,
      "loss": 5.5148,
      "mean_token_accuracy": 0.16883110553026198,
      "num_tokens": 5706825.0,
      "step": 3100
    },
    {
      "entropy": 5.6777756690979,
      "epoch": 0.25935516204477116,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004996513547344877,
      "loss": 5.4516,
      "mean_token_accuracy": 0.17284267097711564,
      "num_tokens": 5715677.0,
      "step": 3105
    },
    {
      "entropy": 5.585661268234253,
      "epoch": 0.25977280320748414,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004996496961369274,
      "loss": 5.3991,
      "mean_token_accuracy": 0.173941034078598,
      "num_tokens": 5725371.0,
      "step": 3110
    },
    {
      "entropy": 5.6625477313995365,
      "epoch": 0.2601904443701971,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996480336066037,
      "loss": 5.477,
      "mean_token_accuracy": 0.17416553050279618,
      "num_tokens": 5734664.0,
      "step": 3115
    },
    {
      "entropy": 5.634637260437012,
      "epoch": 0.26060808553291015,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996463671435458,
      "loss": 5.4855,
      "mean_token_accuracy": 0.17505145072937012,
      "num_tokens": 5743870.0,
      "step": 3120
    },
    {
      "entropy": 5.520635080337525,
      "epoch": 0.26102572669562313,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004996446967477829,
      "loss": 5.5082,
      "mean_token_accuracy": 0.17191022932529448,
      "num_tokens": 5752855.0,
      "step": 3125
    },
    {
      "entropy": 5.769512271881103,
      "epoch": 0.2614433678583361,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004996430224193441,
      "loss": 5.5609,
      "mean_token_accuracy": 0.16556276828050615,
      "num_tokens": 5762435.0,
      "step": 3130
    },
    {
      "entropy": 5.700036287307739,
      "epoch": 0.26186100902104914,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996413441582589,
      "loss": 5.5485,
      "mean_token_accuracy": 0.1690956547856331,
      "num_tokens": 5771463.0,
      "step": 3135
    },
    {
      "entropy": 5.670450019836426,
      "epoch": 0.2622786501837621,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996396619645564,
      "loss": 5.5492,
      "mean_token_accuracy": 0.1639704242348671,
      "num_tokens": 5780530.0,
      "step": 3140
    },
    {
      "entropy": 5.647158193588257,
      "epoch": 0.2626962913464751,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004996379758382662,
      "loss": 5.4251,
      "mean_token_accuracy": 0.17983635514974594,
      "num_tokens": 5789199.0,
      "step": 3145
    },
    {
      "entropy": 5.751897525787354,
      "epoch": 0.26311393250918813,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000499636285779418,
      "loss": 5.603,
      "mean_token_accuracy": 0.16786201298236847,
      "num_tokens": 5799170.0,
      "step": 3150
    },
    {
      "entropy": 5.688197231292724,
      "epoch": 0.2635315736719011,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004996345917880412,
      "loss": 5.5348,
      "mean_token_accuracy": 0.1685117945075035,
      "num_tokens": 5807662.0,
      "step": 3155
    },
    {
      "entropy": 5.563128089904785,
      "epoch": 0.2639492148346141,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996328938641654,
      "loss": 5.4455,
      "mean_token_accuracy": 0.1695580303668976,
      "num_tokens": 5815790.0,
      "step": 3160
    },
    {
      "entropy": 5.651099061965942,
      "epoch": 0.2643668559973271,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996311920078204,
      "loss": 5.4506,
      "mean_token_accuracy": 0.17824542075395583,
      "num_tokens": 5824746.0,
      "step": 3165
    },
    {
      "entropy": 5.578860282897949,
      "epoch": 0.2647844971600401,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004996294862190361,
      "loss": 5.4676,
      "mean_token_accuracy": 0.17490578442811966,
      "num_tokens": 5833678.0,
      "step": 3170
    },
    {
      "entropy": 5.641597127914428,
      "epoch": 0.2652021383227531,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499627776497842,
      "loss": 5.506,
      "mean_token_accuracy": 0.17095759958028794,
      "num_tokens": 5842218.0,
      "step": 3175
    },
    {
      "entropy": 5.803808689117432,
      "epoch": 0.2656197794854661,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004996260628442686,
      "loss": 5.6728,
      "mean_token_accuracy": 0.15886926651000977,
      "num_tokens": 5853250.0,
      "step": 3180
    },
    {
      "entropy": 5.659803962707519,
      "epoch": 0.2660374206481791,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996243452583454,
      "loss": 5.4128,
      "mean_token_accuracy": 0.17672599405050277,
      "num_tokens": 5862661.0,
      "step": 3185
    },
    {
      "entropy": 5.631625938415527,
      "epoch": 0.26645506181089207,
      "grad_norm": 0.875,
      "learning_rate": 0.0004996226237401028,
      "loss": 5.5717,
      "mean_token_accuracy": 0.166826468706131,
      "num_tokens": 5872327.0,
      "step": 3190
    },
    {
      "entropy": 5.5299359321594235,
      "epoch": 0.2668727029736051,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004996208982895706,
      "loss": 5.4184,
      "mean_token_accuracy": 0.17263006567955017,
      "num_tokens": 5881539.0,
      "step": 3195
    },
    {
      "entropy": 5.717694091796875,
      "epoch": 0.2672903441363181,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004996191689067794,
      "loss": 5.5266,
      "mean_token_accuracy": 0.17140432745218276,
      "num_tokens": 5891602.0,
      "step": 3200
    },
    {
      "entropy": 5.647522211074829,
      "epoch": 0.26770798529903106,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996174355917591,
      "loss": 5.5255,
      "mean_token_accuracy": 0.16782172918319702,
      "num_tokens": 5899831.0,
      "step": 3205
    },
    {
      "entropy": 5.723945808410645,
      "epoch": 0.2681256264617441,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004996156983445404,
      "loss": 5.6214,
      "mean_token_accuracy": 0.16156760454177857,
      "num_tokens": 5909193.0,
      "step": 3210
    },
    {
      "entropy": 5.71587610244751,
      "epoch": 0.26854326762445707,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004996139571651534,
      "loss": 5.5222,
      "mean_token_accuracy": 0.17045873999595643,
      "num_tokens": 5919353.0,
      "step": 3215
    },
    {
      "entropy": 5.618978881835938,
      "epoch": 0.26896090878717005,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996122120536288,
      "loss": 5.4652,
      "mean_token_accuracy": 0.17441712468862533,
      "num_tokens": 5927853.0,
      "step": 3220
    },
    {
      "entropy": 5.591433715820313,
      "epoch": 0.2693785499498831,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996104630099969,
      "loss": 5.5718,
      "mean_token_accuracy": 0.16816680431365966,
      "num_tokens": 5937943.0,
      "step": 3225
    },
    {
      "entropy": 5.687422418594361,
      "epoch": 0.26979619111259606,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004996087100342885,
      "loss": 5.4781,
      "mean_token_accuracy": 0.16925078332424165,
      "num_tokens": 5946213.0,
      "step": 3230
    },
    {
      "entropy": 5.607429122924804,
      "epoch": 0.27021383227530904,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004996069531265345,
      "loss": 5.3814,
      "mean_token_accuracy": 0.17218527644872667,
      "num_tokens": 5954997.0,
      "step": 3235
    },
    {
      "entropy": 5.5736700057983395,
      "epoch": 0.2706314734380221,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996051922867653,
      "loss": 5.4225,
      "mean_token_accuracy": 0.17648832201957704,
      "num_tokens": 5963947.0,
      "step": 3240
    },
    {
      "entropy": 5.572920799255371,
      "epoch": 0.27104911460073505,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996034275150118,
      "loss": 5.4528,
      "mean_token_accuracy": 0.16782740950584413,
      "num_tokens": 5972077.0,
      "step": 3245
    },
    {
      "entropy": 5.604636096954346,
      "epoch": 0.27146675576344803,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000499601658811305,
      "loss": 5.4698,
      "mean_token_accuracy": 0.1688433527946472,
      "num_tokens": 5982649.0,
      "step": 3250
    },
    {
      "entropy": 5.635434293746949,
      "epoch": 0.27188439692616106,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004995998861756758,
      "loss": 5.4808,
      "mean_token_accuracy": 0.17200591415166855,
      "num_tokens": 5991349.0,
      "step": 3255
    },
    {
      "entropy": 5.579713439941406,
      "epoch": 0.27230203808887404,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004995981096081552,
      "loss": 5.4557,
      "mean_token_accuracy": 0.1713761046528816,
      "num_tokens": 6000555.0,
      "step": 3260
    },
    {
      "entropy": 5.774811124801635,
      "epoch": 0.272719679251587,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004995963291087744,
      "loss": 5.5699,
      "mean_token_accuracy": 0.16197201758623123,
      "num_tokens": 6009554.0,
      "step": 3265
    },
    {
      "entropy": 5.698873901367188,
      "epoch": 0.27313732041430006,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004995945446775645,
      "loss": 5.6716,
      "mean_token_accuracy": 0.1660139337182045,
      "num_tokens": 6019330.0,
      "step": 3270
    },
    {
      "entropy": 5.700035476684571,
      "epoch": 0.27355496157701303,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004995927563145566,
      "loss": 5.6101,
      "mean_token_accuracy": 0.1618780493736267,
      "num_tokens": 6028527.0,
      "step": 3275
    },
    {
      "entropy": 5.673366355895996,
      "epoch": 0.273972602739726,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995909640197825,
      "loss": 5.4725,
      "mean_token_accuracy": 0.1742170959711075,
      "num_tokens": 6038223.0,
      "step": 3280
    },
    {
      "entropy": 5.579315853118897,
      "epoch": 0.27439024390243905,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004995891677932729,
      "loss": 5.4028,
      "mean_token_accuracy": 0.17360525131225585,
      "num_tokens": 6047459.0,
      "step": 3285
    },
    {
      "entropy": 5.698335456848144,
      "epoch": 0.274807885065152,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995873676350597,
      "loss": 5.5953,
      "mean_token_accuracy": 0.16751037091016768,
      "num_tokens": 6057322.0,
      "step": 3290
    },
    {
      "entropy": 5.7665680885314945,
      "epoch": 0.275225526227865,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995855635451743,
      "loss": 5.7047,
      "mean_token_accuracy": 0.16124847531318665,
      "num_tokens": 6067520.0,
      "step": 3295
    },
    {
      "entropy": 5.641880464553833,
      "epoch": 0.27564316739057804,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995837555236484,
      "loss": 5.5851,
      "mean_token_accuracy": 0.16289951056241989,
      "num_tokens": 6076729.0,
      "step": 3300
    },
    {
      "entropy": 5.714790391921997,
      "epoch": 0.276060808553291,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004995819435705133,
      "loss": 5.5182,
      "mean_token_accuracy": 0.16744791120290756,
      "num_tokens": 6086529.0,
      "step": 3305
    },
    {
      "entropy": 5.677750301361084,
      "epoch": 0.276478449716004,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995801276858011,
      "loss": 5.5337,
      "mean_token_accuracy": 0.1721680462360382,
      "num_tokens": 6095784.0,
      "step": 3310
    },
    {
      "entropy": 5.748060846328736,
      "epoch": 0.276896090878717,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004995783078695434,
      "loss": 5.5648,
      "mean_token_accuracy": 0.1712627023458481,
      "num_tokens": 6104626.0,
      "step": 3315
    },
    {
      "entropy": 5.665297269821167,
      "epoch": 0.27731373204143,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499576484121772,
      "loss": 5.4757,
      "mean_token_accuracy": 0.17615763247013091,
      "num_tokens": 6113758.0,
      "step": 3320
    },
    {
      "entropy": 5.744389915466309,
      "epoch": 0.277731373204143,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995746564425191,
      "loss": 5.4852,
      "mean_token_accuracy": 0.17800086438655854,
      "num_tokens": 6123265.0,
      "step": 3325
    },
    {
      "entropy": 5.687781858444214,
      "epoch": 0.278149014366856,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995728248318163,
      "loss": 5.5508,
      "mean_token_accuracy": 0.16848849207162858,
      "num_tokens": 6132855.0,
      "step": 3330
    },
    {
      "entropy": 5.5548903465271,
      "epoch": 0.278566655529569,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004995709892896961,
      "loss": 5.4233,
      "mean_token_accuracy": 0.17018820494413375,
      "num_tokens": 6142739.0,
      "step": 3335
    },
    {
      "entropy": 5.585405588150024,
      "epoch": 0.278984296692282,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995691498161903,
      "loss": 5.3956,
      "mean_token_accuracy": 0.17602959722280503,
      "num_tokens": 6152169.0,
      "step": 3340
    },
    {
      "entropy": 5.495415258407593,
      "epoch": 0.279401937854995,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995673064113313,
      "loss": 5.4168,
      "mean_token_accuracy": 0.17533375918865204,
      "num_tokens": 6160235.0,
      "step": 3345
    },
    {
      "entropy": 5.647177648544312,
      "epoch": 0.279819579017708,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004995654590751513,
      "loss": 5.4098,
      "mean_token_accuracy": 0.17150053828954698,
      "num_tokens": 6169440.0,
      "step": 3350
    },
    {
      "entropy": 5.576392459869385,
      "epoch": 0.28023722018042097,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004995636078076825,
      "loss": 5.4206,
      "mean_token_accuracy": 0.17606925815343857,
      "num_tokens": 6178456.0,
      "step": 3355
    },
    {
      "entropy": 5.555343532562256,
      "epoch": 0.280654861343134,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004995617526089575,
      "loss": 5.5021,
      "mean_token_accuracy": 0.16990070939064025,
      "num_tokens": 6186844.0,
      "step": 3360
    },
    {
      "entropy": 5.629121732711792,
      "epoch": 0.281072502505847,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004995598934790087,
      "loss": 5.4163,
      "mean_token_accuracy": 0.1740152806043625,
      "num_tokens": 6196117.0,
      "step": 3365
    },
    {
      "entropy": 5.60472526550293,
      "epoch": 0.28149014366855996,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004995580304178687,
      "loss": 5.4128,
      "mean_token_accuracy": 0.17387232184410095,
      "num_tokens": 6204636.0,
      "step": 3370
    },
    {
      "entropy": 5.57462706565857,
      "epoch": 0.281907784831273,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004995561634255701,
      "loss": 5.4332,
      "mean_token_accuracy": 0.1764407992362976,
      "num_tokens": 6214537.0,
      "step": 3375
    },
    {
      "entropy": 5.614592170715332,
      "epoch": 0.28232542599398597,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004995542925021455,
      "loss": 5.4936,
      "mean_token_accuracy": 0.1737373188138008,
      "num_tokens": 6224753.0,
      "step": 3380
    },
    {
      "entropy": 5.837904787063598,
      "epoch": 0.28274306715669895,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004995524176476277,
      "loss": 5.6749,
      "mean_token_accuracy": 0.16878011524677278,
      "num_tokens": 6233842.0,
      "step": 3385
    },
    {
      "entropy": 5.717729902267456,
      "epoch": 0.283160708319412,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004995505388620497,
      "loss": 5.5252,
      "mean_token_accuracy": 0.17395778745412827,
      "num_tokens": 6243467.0,
      "step": 3390
    },
    {
      "entropy": 5.555162382125855,
      "epoch": 0.28357834948212496,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499548656145444,
      "loss": 5.5569,
      "mean_token_accuracy": 0.16584165692329406,
      "num_tokens": 6253263.0,
      "step": 3395
    },
    {
      "entropy": 5.693689489364624,
      "epoch": 0.28399599064483794,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004995467694978439,
      "loss": 5.4,
      "mean_token_accuracy": 0.18335435092449187,
      "num_tokens": 6262249.0,
      "step": 3400
    },
    {
      "entropy": 5.573141384124756,
      "epoch": 0.28441363180755097,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995448789192823,
      "loss": 5.422,
      "mean_token_accuracy": 0.17384118884801864,
      "num_tokens": 6271154.0,
      "step": 3405
    },
    {
      "entropy": 5.59210467338562,
      "epoch": 0.28483127297026395,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004995429844097922,
      "loss": 5.5069,
      "mean_token_accuracy": 0.17582175135612488,
      "num_tokens": 6280377.0,
      "step": 3410
    },
    {
      "entropy": 5.646082258224487,
      "epoch": 0.2852489141329769,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004995410859694071,
      "loss": 5.4808,
      "mean_token_accuracy": 0.17655370235443116,
      "num_tokens": 6289941.0,
      "step": 3415
    },
    {
      "entropy": 5.637767171859741,
      "epoch": 0.28566655529568996,
      "grad_norm": 0.875,
      "learning_rate": 0.0004995391835981599,
      "loss": 5.5495,
      "mean_token_accuracy": 0.16371782571077348,
      "num_tokens": 6299281.0,
      "step": 3420
    },
    {
      "entropy": 5.677474927902222,
      "epoch": 0.28608419645840294,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499537277296084,
      "loss": 5.5084,
      "mean_token_accuracy": 0.16698610335588454,
      "num_tokens": 6309408.0,
      "step": 3425
    },
    {
      "entropy": 5.62541093826294,
      "epoch": 0.2865018376211159,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004995353670632127,
      "loss": 5.5236,
      "mean_token_accuracy": 0.17414335906505585,
      "num_tokens": 6318979.0,
      "step": 3430
    },
    {
      "entropy": 5.595382642745972,
      "epoch": 0.28691947878382895,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004995334528995797,
      "loss": 5.3714,
      "mean_token_accuracy": 0.1781489819288254,
      "num_tokens": 6327616.0,
      "step": 3435
    },
    {
      "entropy": 5.535250473022461,
      "epoch": 0.28733711994654193,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995315348052182,
      "loss": 5.3939,
      "mean_token_accuracy": 0.1786615163087845,
      "num_tokens": 6336486.0,
      "step": 3440
    },
    {
      "entropy": 5.610561084747315,
      "epoch": 0.2877547611092549,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499529612780162,
      "loss": 5.4429,
      "mean_token_accuracy": 0.17057020962238312,
      "num_tokens": 6345671.0,
      "step": 3445
    },
    {
      "entropy": 5.666832304000854,
      "epoch": 0.28817240227196794,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995276868244446,
      "loss": 5.5339,
      "mean_token_accuracy": 0.17009024173021317,
      "num_tokens": 6354534.0,
      "step": 3450
    },
    {
      "entropy": 5.604935932159424,
      "epoch": 0.2885900434346809,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004995257569380998,
      "loss": 5.3777,
      "mean_token_accuracy": 0.17774803787469864,
      "num_tokens": 6364052.0,
      "step": 3455
    },
    {
      "entropy": 5.6663824081420895,
      "epoch": 0.2890076845973939,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004995238231211614,
      "loss": 5.5152,
      "mean_token_accuracy": 0.17114375531673431,
      "num_tokens": 6373305.0,
      "step": 3460
    },
    {
      "entropy": 5.575594902038574,
      "epoch": 0.28942532576010693,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004995218853736631,
      "loss": 5.4307,
      "mean_token_accuracy": 0.17053433507680893,
      "num_tokens": 6382673.0,
      "step": 3465
    },
    {
      "entropy": 5.640283870697021,
      "epoch": 0.2898429669228199,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995199436956391,
      "loss": 5.5353,
      "mean_token_accuracy": 0.16578124612569808,
      "num_tokens": 6392122.0,
      "step": 3470
    },
    {
      "entropy": 5.708892965316773,
      "epoch": 0.2902606080855329,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995179980871231,
      "loss": 5.5571,
      "mean_token_accuracy": 0.16502658128738404,
      "num_tokens": 6401130.0,
      "step": 3475
    },
    {
      "entropy": 5.625602769851684,
      "epoch": 0.2906782492482459,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995160485481493,
      "loss": 5.4269,
      "mean_token_accuracy": 0.17211932986974715,
      "num_tokens": 6409471.0,
      "step": 3480
    },
    {
      "entropy": 5.576659202575684,
      "epoch": 0.2910958904109589,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000499514095078752,
      "loss": 5.4322,
      "mean_token_accuracy": 0.17299148291349412,
      "num_tokens": 6418902.0,
      "step": 3485
    },
    {
      "entropy": 5.612267351150512,
      "epoch": 0.2915135315736719,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499512137678965,
      "loss": 5.517,
      "mean_token_accuracy": 0.17158575505018234,
      "num_tokens": 6428529.0,
      "step": 3490
    },
    {
      "entropy": 5.630302476882934,
      "epoch": 0.2919311727363849,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995101763488229,
      "loss": 5.3775,
      "mean_token_accuracy": 0.17603698819875718,
      "num_tokens": 6437460.0,
      "step": 3495
    },
    {
      "entropy": 5.596737051010132,
      "epoch": 0.2923488138990979,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004995082110883599,
      "loss": 5.4227,
      "mean_token_accuracy": 0.1792820855975151,
      "num_tokens": 6446005.0,
      "step": 3500
    },
    {
      "entropy": 5.583709478378296,
      "epoch": 0.29276645506181087,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004995062418976104,
      "loss": 5.4648,
      "mean_token_accuracy": 0.16964059174060822,
      "num_tokens": 6455241.0,
      "step": 3505
    },
    {
      "entropy": 5.609921503067016,
      "epoch": 0.2931840962245239,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004995042687766089,
      "loss": 5.445,
      "mean_token_accuracy": 0.1753866121172905,
      "num_tokens": 6464748.0,
      "step": 3510
    },
    {
      "entropy": 5.594579076766967,
      "epoch": 0.2936017373872369,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004995022917253899,
      "loss": 5.492,
      "mean_token_accuracy": 0.1760092243552208,
      "num_tokens": 6473834.0,
      "step": 3515
    },
    {
      "entropy": 5.635177898406982,
      "epoch": 0.29401937854994986,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004995003107439882,
      "loss": 5.4482,
      "mean_token_accuracy": 0.1760343998670578,
      "num_tokens": 6482169.0,
      "step": 3520
    },
    {
      "entropy": 5.509727525711059,
      "epoch": 0.2944370197126629,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004994983258324383,
      "loss": 5.3935,
      "mean_token_accuracy": 0.17419851422309876,
      "num_tokens": 6491798.0,
      "step": 3525
    },
    {
      "entropy": 5.533378934860229,
      "epoch": 0.2948546608753759,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004994963369907748,
      "loss": 5.3488,
      "mean_token_accuracy": 0.1736716225743294,
      "num_tokens": 6500143.0,
      "step": 3530
    },
    {
      "entropy": 5.518638277053833,
      "epoch": 0.29527230203808885,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004994943442190328,
      "loss": 5.4781,
      "mean_token_accuracy": 0.17285873889923095,
      "num_tokens": 6509745.0,
      "step": 3535
    },
    {
      "entropy": 5.696667146682739,
      "epoch": 0.2956899432008019,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004994923475172471,
      "loss": 5.5376,
      "mean_token_accuracy": 0.1681595653295517,
      "num_tokens": 6520931.0,
      "step": 3540
    },
    {
      "entropy": 5.607496070861816,
      "epoch": 0.29610758436351486,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004994903468854527,
      "loss": 5.4744,
      "mean_token_accuracy": 0.1733265310525894,
      "num_tokens": 6529774.0,
      "step": 3545
    },
    {
      "entropy": 5.551492595672608,
      "epoch": 0.29652522552622784,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004994883423236843,
      "loss": 5.3844,
      "mean_token_accuracy": 0.17458131313323974,
      "num_tokens": 6538922.0,
      "step": 3550
    },
    {
      "entropy": 5.519823980331421,
      "epoch": 0.2969428666889409,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004994863338319775,
      "loss": 5.4342,
      "mean_token_accuracy": 0.17835731208324432,
      "num_tokens": 6548891.0,
      "step": 3555
    },
    {
      "entropy": 5.4987424373626705,
      "epoch": 0.29736050785165385,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004994843214103672,
      "loss": 5.2739,
      "mean_token_accuracy": 0.17617563158273697,
      "num_tokens": 6557025.0,
      "step": 3560
    },
    {
      "entropy": 5.628918981552124,
      "epoch": 0.29777814901436683,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004994823050588886,
      "loss": 5.4464,
      "mean_token_accuracy": 0.17203134596347808,
      "num_tokens": 6567037.0,
      "step": 3565
    },
    {
      "entropy": 5.5824846744537355,
      "epoch": 0.29819579017707987,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499480284777577,
      "loss": 5.4834,
      "mean_token_accuracy": 0.16981624513864518,
      "num_tokens": 6576766.0,
      "step": 3570
    },
    {
      "entropy": 5.602394247055054,
      "epoch": 0.29861343133979285,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004994782605664677,
      "loss": 5.4469,
      "mean_token_accuracy": 0.17235205918550492,
      "num_tokens": 6585576.0,
      "step": 3575
    },
    {
      "entropy": 5.659743213653565,
      "epoch": 0.2990310725025058,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004994762324255965,
      "loss": 5.5025,
      "mean_token_accuracy": 0.17509230673313142,
      "num_tokens": 6594463.0,
      "step": 3580
    },
    {
      "entropy": 5.590396499633789,
      "epoch": 0.29944871366521886,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994742003549986,
      "loss": 5.389,
      "mean_token_accuracy": 0.17503619939088821,
      "num_tokens": 6603651.0,
      "step": 3585
    },
    {
      "entropy": 5.613985061645508,
      "epoch": 0.29986635482793184,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004994721643547095,
      "loss": 5.4582,
      "mean_token_accuracy": 0.17667034417390823,
      "num_tokens": 6612950.0,
      "step": 3590
    },
    {
      "entropy": 5.655868196487427,
      "epoch": 0.3002839959906448,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499470124424765,
      "loss": 5.3922,
      "mean_token_accuracy": 0.1811322718858719,
      "num_tokens": 6622013.0,
      "step": 3595
    },
    {
      "entropy": 5.586564588546753,
      "epoch": 0.30070163715335785,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004994680805652008,
      "loss": 5.4748,
      "mean_token_accuracy": 0.17597898244857788,
      "num_tokens": 6632046.0,
      "step": 3600
    },
    {
      "entropy": 5.5558055400848385,
      "epoch": 0.3011192783160708,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004994660327760527,
      "loss": 5.5244,
      "mean_token_accuracy": 0.17655322402715684,
      "num_tokens": 6641169.0,
      "step": 3605
    },
    {
      "entropy": 5.568960666656494,
      "epoch": 0.3015369194787838,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004994639810573565,
      "loss": 5.3343,
      "mean_token_accuracy": 0.18257040977478028,
      "num_tokens": 6649663.0,
      "step": 3610
    },
    {
      "entropy": 5.693808078765869,
      "epoch": 0.30195456064149684,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499461925409148,
      "loss": 5.5233,
      "mean_token_accuracy": 0.16796292513608932,
      "num_tokens": 6659448.0,
      "step": 3615
    },
    {
      "entropy": 5.640341663360596,
      "epoch": 0.3023722018042098,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004994598658314634,
      "loss": 5.4704,
      "mean_token_accuracy": 0.17719862312078477,
      "num_tokens": 6670314.0,
      "step": 3620
    },
    {
      "entropy": 5.5831200122833256,
      "epoch": 0.3027898429669228,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994578023243386,
      "loss": 5.4244,
      "mean_token_accuracy": 0.17784592658281326,
      "num_tokens": 6679413.0,
      "step": 3625
    },
    {
      "entropy": 5.574049806594848,
      "epoch": 0.30320748412963583,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004994557348878098,
      "loss": 5.4179,
      "mean_token_accuracy": 0.17593526989221572,
      "num_tokens": 6688439.0,
      "step": 3630
    },
    {
      "entropy": 5.582159566879272,
      "epoch": 0.3036251252923488,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004994536635219133,
      "loss": 5.3971,
      "mean_token_accuracy": 0.17586517035961152,
      "num_tokens": 6697471.0,
      "step": 3635
    },
    {
      "entropy": 5.631478309631348,
      "epoch": 0.3040427664550618,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004994515882266852,
      "loss": 5.4558,
      "mean_token_accuracy": 0.17909279018640517,
      "num_tokens": 6706698.0,
      "step": 3640
    },
    {
      "entropy": 5.527060413360596,
      "epoch": 0.3044604076177748,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994495090021618,
      "loss": 5.368,
      "mean_token_accuracy": 0.18144972026348113,
      "num_tokens": 6715676.0,
      "step": 3645
    },
    {
      "entropy": 5.492686986923218,
      "epoch": 0.3048780487804878,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004994474258483796,
      "loss": 5.4178,
      "mean_token_accuracy": 0.1791821986436844,
      "num_tokens": 6723818.0,
      "step": 3650
    },
    {
      "entropy": 5.566056585311889,
      "epoch": 0.3052956899432008,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004994453387653751,
      "loss": 5.4289,
      "mean_token_accuracy": 0.17524709850549697,
      "num_tokens": 6733450.0,
      "step": 3655
    },
    {
      "entropy": 5.6221230030059814,
      "epoch": 0.3057133311059138,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004994432477531847,
      "loss": 5.3651,
      "mean_token_accuracy": 0.1746278703212738,
      "num_tokens": 6743142.0,
      "step": 3660
    },
    {
      "entropy": 5.5026391506195065,
      "epoch": 0.3061309722686268,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004994411528118452,
      "loss": 5.3328,
      "mean_token_accuracy": 0.1746651664376259,
      "num_tokens": 6752185.0,
      "step": 3665
    },
    {
      "entropy": 5.532116985321045,
      "epoch": 0.30654861343133977,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004994390539413931,
      "loss": 5.4232,
      "mean_token_accuracy": 0.17844787687063218,
      "num_tokens": 6761549.0,
      "step": 3670
    },
    {
      "entropy": 5.5767427444458,
      "epoch": 0.3069662545940528,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994369511418652,
      "loss": 5.4279,
      "mean_token_accuracy": 0.17511649280786515,
      "num_tokens": 6770395.0,
      "step": 3675
    },
    {
      "entropy": 5.646317052841186,
      "epoch": 0.3073838957567658,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004994348444132983,
      "loss": 5.5171,
      "mean_token_accuracy": 0.17111391127109526,
      "num_tokens": 6779154.0,
      "step": 3680
    },
    {
      "entropy": 5.533228874206543,
      "epoch": 0.30780153691947876,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004994327337557293,
      "loss": 5.4046,
      "mean_token_accuracy": 0.1810103803873062,
      "num_tokens": 6789263.0,
      "step": 3685
    },
    {
      "entropy": 5.6153871536254885,
      "epoch": 0.3082191780821918,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004994306191691952,
      "loss": 5.4921,
      "mean_token_accuracy": 0.17272201478481292,
      "num_tokens": 6797985.0,
      "step": 3690
    },
    {
      "entropy": 5.569174194335938,
      "epoch": 0.30863681924490477,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004994285006537329,
      "loss": 5.3692,
      "mean_token_accuracy": 0.17812356054782869,
      "num_tokens": 6807566.0,
      "step": 3695
    },
    {
      "entropy": 5.507190179824829,
      "epoch": 0.30905446040761775,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004994263782093796,
      "loss": 5.4156,
      "mean_token_accuracy": 0.17545130997896194,
      "num_tokens": 6815637.0,
      "step": 3700
    },
    {
      "entropy": 5.549441623687744,
      "epoch": 0.3094721015703308,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004994242518361723,
      "loss": 5.3038,
      "mean_token_accuracy": 0.1856495276093483,
      "num_tokens": 6823721.0,
      "step": 3705
    },
    {
      "entropy": 5.660196971893311,
      "epoch": 0.30988974273304376,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004994221215341484,
      "loss": 5.478,
      "mean_token_accuracy": 0.17654901295900344,
      "num_tokens": 6833437.0,
      "step": 3710
    },
    {
      "entropy": 5.587020826339722,
      "epoch": 0.31030738389575674,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004994199873033452,
      "loss": 5.5247,
      "mean_token_accuracy": 0.17180875837802886,
      "num_tokens": 6842287.0,
      "step": 3715
    },
    {
      "entropy": 5.6327098369598385,
      "epoch": 0.3107250250584698,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004994178491437999,
      "loss": 5.3624,
      "mean_token_accuracy": 0.173313145339489,
      "num_tokens": 6851520.0,
      "step": 3720
    },
    {
      "entropy": 5.53466100692749,
      "epoch": 0.31114266622118275,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004994157070555501,
      "loss": 5.4252,
      "mean_token_accuracy": 0.1777387723326683,
      "num_tokens": 6861219.0,
      "step": 3725
    },
    {
      "entropy": 5.6607451915740965,
      "epoch": 0.31156030738389573,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004994135610386331,
      "loss": 5.5468,
      "mean_token_accuracy": 0.17208370715379714,
      "num_tokens": 6870525.0,
      "step": 3730
    },
    {
      "entropy": 5.653220367431641,
      "epoch": 0.31197794854660876,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004994114110930867,
      "loss": 5.4669,
      "mean_token_accuracy": 0.1724211037158966,
      "num_tokens": 6880183.0,
      "step": 3735
    },
    {
      "entropy": 5.5725923538208,
      "epoch": 0.31239558970932174,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004994092572189483,
      "loss": 5.5096,
      "mean_token_accuracy": 0.16694111824035646,
      "num_tokens": 6890442.0,
      "step": 3740
    },
    {
      "entropy": 5.611990356445313,
      "epoch": 0.3128132308720348,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994070994162558,
      "loss": 5.4245,
      "mean_token_accuracy": 0.17552689462900162,
      "num_tokens": 6900091.0,
      "step": 3745
    },
    {
      "entropy": 5.548181390762329,
      "epoch": 0.31323087203474775,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004994049376850471,
      "loss": 5.4349,
      "mean_token_accuracy": 0.17823719084262848,
      "num_tokens": 6910489.0,
      "step": 3750
    },
    {
      "entropy": 5.600250673294068,
      "epoch": 0.31364851319746073,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994027720253596,
      "loss": 5.4354,
      "mean_token_accuracy": 0.17544099390506745,
      "num_tokens": 6919895.0,
      "step": 3755
    },
    {
      "entropy": 5.729648590087891,
      "epoch": 0.31406615436017377,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004994006024372315,
      "loss": 5.5846,
      "mean_token_accuracy": 0.16913123428821564,
      "num_tokens": 6929511.0,
      "step": 3760
    },
    {
      "entropy": 5.553057909011841,
      "epoch": 0.31448379552288674,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004993984289207007,
      "loss": 5.3369,
      "mean_token_accuracy": 0.18162955939769745,
      "num_tokens": 6938059.0,
      "step": 3765
    },
    {
      "entropy": 5.5744287967681885,
      "epoch": 0.3149014366855997,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004993962514758054,
      "loss": 5.3731,
      "mean_token_accuracy": 0.17309387475252153,
      "num_tokens": 6946894.0,
      "step": 3770
    },
    {
      "entropy": 5.613360500335693,
      "epoch": 0.31531907784831276,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004993940701025836,
      "loss": 5.3745,
      "mean_token_accuracy": 0.17189164310693741,
      "num_tokens": 6955494.0,
      "step": 3775
    },
    {
      "entropy": 5.477834796905517,
      "epoch": 0.31573671901102573,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004993918848010734,
      "loss": 5.4932,
      "mean_token_accuracy": 0.1744875878095627,
      "num_tokens": 6964347.0,
      "step": 3780
    },
    {
      "entropy": 5.638558101654053,
      "epoch": 0.3161543601737387,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004993896955713133,
      "loss": 5.4434,
      "mean_token_accuracy": 0.16530326753854752,
      "num_tokens": 6974750.0,
      "step": 3785
    },
    {
      "entropy": 5.606763648986816,
      "epoch": 0.31657200133645175,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004993875024133414,
      "loss": 5.4378,
      "mean_token_accuracy": 0.17337581068277358,
      "num_tokens": 6984228.0,
      "step": 3790
    },
    {
      "entropy": 5.537848043441772,
      "epoch": 0.3169896424991647,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993853053271962,
      "loss": 5.4246,
      "mean_token_accuracy": 0.17393614947795868,
      "num_tokens": 6993173.0,
      "step": 3795
    },
    {
      "entropy": 5.713309001922608,
      "epoch": 0.3174072836618777,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004993831043129161,
      "loss": 5.6243,
      "mean_token_accuracy": 0.16073523610830306,
      "num_tokens": 7003486.0,
      "step": 3800
    },
    {
      "entropy": 5.521305418014526,
      "epoch": 0.31782492482459074,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004993808993705397,
      "loss": 5.3674,
      "mean_token_accuracy": 0.18047840744256974,
      "num_tokens": 7011985.0,
      "step": 3805
    },
    {
      "entropy": 5.524029731750488,
      "epoch": 0.3182425659873037,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004993786905001055,
      "loss": 5.3361,
      "mean_token_accuracy": 0.1789662793278694,
      "num_tokens": 7021148.0,
      "step": 3810
    },
    {
      "entropy": 5.51288070678711,
      "epoch": 0.3186602071500167,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993764777016524,
      "loss": 5.3763,
      "mean_token_accuracy": 0.17878497391939163,
      "num_tokens": 7030110.0,
      "step": 3815
    },
    {
      "entropy": 5.618573570251465,
      "epoch": 0.31907784831272973,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993742609752188,
      "loss": 5.4678,
      "mean_token_accuracy": 0.18152561783790588,
      "num_tokens": 7039428.0,
      "step": 3820
    },
    {
      "entropy": 5.663624382019043,
      "epoch": 0.3194954894754427,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004993720403208438,
      "loss": 5.5172,
      "mean_token_accuracy": 0.17332621663808823,
      "num_tokens": 7048703.0,
      "step": 3825
    },
    {
      "entropy": 5.571570301055909,
      "epoch": 0.3199131306381557,
      "grad_norm": 1.0,
      "learning_rate": 0.000499369815738566,
      "loss": 5.4035,
      "mean_token_accuracy": 0.1688519835472107,
      "num_tokens": 7057612.0,
      "step": 3830
    },
    {
      "entropy": 5.535891962051392,
      "epoch": 0.3203307718008687,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004993675872284248,
      "loss": 5.3706,
      "mean_token_accuracy": 0.18030135184526444,
      "num_tokens": 7066660.0,
      "step": 3835
    },
    {
      "entropy": 5.50783224105835,
      "epoch": 0.3207484129635817,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004993653547904586,
      "loss": 5.3625,
      "mean_token_accuracy": 0.1667173057794571,
      "num_tokens": 7075368.0,
      "step": 3840
    },
    {
      "entropy": 5.576121377944946,
      "epoch": 0.3211660541262947,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499363118424707,
      "loss": 5.4491,
      "mean_token_accuracy": 0.17960931062698365,
      "num_tokens": 7085276.0,
      "step": 3845
    },
    {
      "entropy": 5.578032970428467,
      "epoch": 0.3215836952890077,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004993608781312088,
      "loss": 5.4212,
      "mean_token_accuracy": 0.16872155666351318,
      "num_tokens": 7094292.0,
      "step": 3850
    },
    {
      "entropy": 5.697376775741577,
      "epoch": 0.3220013364517207,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004993586339100035,
      "loss": 5.6413,
      "mean_token_accuracy": 0.16545760929584502,
      "num_tokens": 7104712.0,
      "step": 3855
    },
    {
      "entropy": 5.66398983001709,
      "epoch": 0.32241897761443367,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004993563857611301,
      "loss": 5.433,
      "mean_token_accuracy": 0.17558406442403793,
      "num_tokens": 7113297.0,
      "step": 3860
    },
    {
      "entropy": 5.4992107391357425,
      "epoch": 0.3228366187771467,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004993541336846283,
      "loss": 5.4331,
      "mean_token_accuracy": 0.16872737258672715,
      "num_tokens": 7122706.0,
      "step": 3865
    },
    {
      "entropy": 5.5710976123809814,
      "epoch": 0.3232542599398597,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004993518776805372,
      "loss": 5.4093,
      "mean_token_accuracy": 0.18045277893543243,
      "num_tokens": 7131939.0,
      "step": 3870
    },
    {
      "entropy": 5.491407155990601,
      "epoch": 0.32367190110257266,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004993496177488965,
      "loss": 5.2809,
      "mean_token_accuracy": 0.18518373519182205,
      "num_tokens": 7141016.0,
      "step": 3875
    },
    {
      "entropy": 5.479959201812744,
      "epoch": 0.3240895422652857,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004993473538897457,
      "loss": 5.2594,
      "mean_token_accuracy": 0.18782902956008912,
      "num_tokens": 7149177.0,
      "step": 3880
    },
    {
      "entropy": 5.541644525527954,
      "epoch": 0.32450718342799867,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004993450861031243,
      "loss": 5.3747,
      "mean_token_accuracy": 0.18481706231832504,
      "num_tokens": 7157839.0,
      "step": 3885
    },
    {
      "entropy": 5.630143499374389,
      "epoch": 0.32492482459071165,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004993428143890722,
      "loss": 5.5146,
      "mean_token_accuracy": 0.16957764476537704,
      "num_tokens": 7167759.0,
      "step": 3890
    },
    {
      "entropy": 5.669652509689331,
      "epoch": 0.3253424657534247,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004993405387476291,
      "loss": 5.459,
      "mean_token_accuracy": 0.1642366275191307,
      "num_tokens": 7177141.0,
      "step": 3895
    },
    {
      "entropy": 5.51043667793274,
      "epoch": 0.32576010691613766,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004993382591788349,
      "loss": 5.3715,
      "mean_token_accuracy": 0.182450732588768,
      "num_tokens": 7185350.0,
      "step": 3900
    },
    {
      "entropy": 5.479080486297607,
      "epoch": 0.32617774807885064,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004993359756827293,
      "loss": 5.305,
      "mean_token_accuracy": 0.17941731065511704,
      "num_tokens": 7194073.0,
      "step": 3905
    },
    {
      "entropy": 5.560730886459351,
      "epoch": 0.32659538924156367,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004993336882593524,
      "loss": 5.4841,
      "mean_token_accuracy": 0.16640772223472594,
      "num_tokens": 7203421.0,
      "step": 3910
    },
    {
      "entropy": 5.533968257904053,
      "epoch": 0.32701303040427665,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004993313969087444,
      "loss": 5.2458,
      "mean_token_accuracy": 0.17937704026699067,
      "num_tokens": 7212548.0,
      "step": 3915
    },
    {
      "entropy": 5.55244460105896,
      "epoch": 0.32743067156698963,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004993291016309452,
      "loss": 5.4967,
      "mean_token_accuracy": 0.1723368287086487,
      "num_tokens": 7221668.0,
      "step": 3920
    },
    {
      "entropy": 5.537858486175537,
      "epoch": 0.32784831272970266,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000499326802425995,
      "loss": 5.3871,
      "mean_token_accuracy": 0.17311936169862746,
      "num_tokens": 7231775.0,
      "step": 3925
    },
    {
      "entropy": 5.626779317855835,
      "epoch": 0.32826595389241564,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004993244992939342,
      "loss": 5.4126,
      "mean_token_accuracy": 0.1729939341545105,
      "num_tokens": 7241264.0,
      "step": 3930
    },
    {
      "entropy": 5.576012229919433,
      "epoch": 0.3286835950551286,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499322192234803,
      "loss": 5.4435,
      "mean_token_accuracy": 0.17053294330835342,
      "num_tokens": 7250519.0,
      "step": 3935
    },
    {
      "entropy": 5.690303087234497,
      "epoch": 0.32910123621784165,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004993198812486418,
      "loss": 5.5006,
      "mean_token_accuracy": 0.17200805842876435,
      "num_tokens": 7259533.0,
      "step": 3940
    },
    {
      "entropy": 5.534047651290893,
      "epoch": 0.32951887738055463,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000499317566335491,
      "loss": 5.3108,
      "mean_token_accuracy": 0.17637526243925095,
      "num_tokens": 7268980.0,
      "step": 3945
    },
    {
      "entropy": 5.550717115402222,
      "epoch": 0.3299365185432676,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004993152474953913,
      "loss": 5.4076,
      "mean_token_accuracy": 0.17732179611921312,
      "num_tokens": 7277126.0,
      "step": 3950
    },
    {
      "entropy": 5.575921821594238,
      "epoch": 0.33035415970598064,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004993129247283831,
      "loss": 5.38,
      "mean_token_accuracy": 0.1864038050174713,
      "num_tokens": 7286223.0,
      "step": 3955
    },
    {
      "entropy": 5.54976716041565,
      "epoch": 0.3307718008686936,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004993105980345071,
      "loss": 5.3832,
      "mean_token_accuracy": 0.1787816286087036,
      "num_tokens": 7295054.0,
      "step": 3960
    },
    {
      "entropy": 5.55245418548584,
      "epoch": 0.3311894420314066,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004993082674138042,
      "loss": 5.4102,
      "mean_token_accuracy": 0.17833967059850692,
      "num_tokens": 7304502.0,
      "step": 3965
    },
    {
      "entropy": 5.590841054916382,
      "epoch": 0.33160708319411963,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000499305932866315,
      "loss": 5.3731,
      "mean_token_accuracy": 0.1795377776026726,
      "num_tokens": 7314093.0,
      "step": 3970
    },
    {
      "entropy": 5.607157611846924,
      "epoch": 0.3320247243568326,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004993035943920805,
      "loss": 5.4079,
      "mean_token_accuracy": 0.17507885843515397,
      "num_tokens": 7323059.0,
      "step": 3975
    },
    {
      "entropy": 5.636029767990112,
      "epoch": 0.3324423655195456,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004993012519911416,
      "loss": 5.5475,
      "mean_token_accuracy": 0.16774446070194243,
      "num_tokens": 7332596.0,
      "step": 3980
    },
    {
      "entropy": 5.547656583786011,
      "epoch": 0.3328600066822586,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004992989056635392,
      "loss": 5.3501,
      "mean_token_accuracy": 0.17981956601142884,
      "num_tokens": 7341671.0,
      "step": 3985
    },
    {
      "entropy": 5.529853487014771,
      "epoch": 0.3332776478449716,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004992965554093146,
      "loss": 5.4497,
      "mean_token_accuracy": 0.17468183636665344,
      "num_tokens": 7350344.0,
      "step": 3990
    },
    {
      "entropy": 5.607574319839477,
      "epoch": 0.3336952890076846,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004992942012285086,
      "loss": 5.4638,
      "mean_token_accuracy": 0.1764901638031006,
      "num_tokens": 7359844.0,
      "step": 3995
    },
    {
      "entropy": 5.538512563705444,
      "epoch": 0.3341129301703976,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004992918431211629,
      "loss": 5.3973,
      "mean_token_accuracy": 0.17774405628442763,
      "num_tokens": 7369377.0,
      "step": 4000
    },
    {
      "entropy": 5.569111490249634,
      "epoch": 0.3345305713331106,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992894810873183,
      "loss": 5.3726,
      "mean_token_accuracy": 0.1755051791667938,
      "num_tokens": 7378814.0,
      "step": 4005
    },
    {
      "entropy": 5.536145162582398,
      "epoch": 0.33494821249582357,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004992871151270165,
      "loss": 5.3726,
      "mean_token_accuracy": 0.1799241617321968,
      "num_tokens": 7387333.0,
      "step": 4010
    },
    {
      "entropy": 5.485998868942261,
      "epoch": 0.3353658536585366,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004992847452402986,
      "loss": 5.4088,
      "mean_token_accuracy": 0.17984551936388016,
      "num_tokens": 7396895.0,
      "step": 4015
    },
    {
      "entropy": 5.539252376556396,
      "epoch": 0.3357834948212496,
      "grad_norm": 1.0,
      "learning_rate": 0.0004992823714272064,
      "loss": 5.3991,
      "mean_token_accuracy": 0.1765997514128685,
      "num_tokens": 7406430.0,
      "step": 4020
    },
    {
      "entropy": 5.543996572494507,
      "epoch": 0.33620113598396256,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004992799936877813,
      "loss": 5.2429,
      "mean_token_accuracy": 0.18930246978998183,
      "num_tokens": 7415224.0,
      "step": 4025
    },
    {
      "entropy": 5.600168132781983,
      "epoch": 0.3366187771466756,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004992776120220649,
      "loss": 5.5703,
      "mean_token_accuracy": 0.16221928745508193,
      "num_tokens": 7424250.0,
      "step": 4030
    },
    {
      "entropy": 5.644545841217041,
      "epoch": 0.3370364183093886,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004992752264300989,
      "loss": 5.3963,
      "mean_token_accuracy": 0.18497857600450515,
      "num_tokens": 7433579.0,
      "step": 4035
    },
    {
      "entropy": 5.4642468929290775,
      "epoch": 0.33745405947210155,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004992728369119252,
      "loss": 5.3512,
      "mean_token_accuracy": 0.1829281285405159,
      "num_tokens": 7442621.0,
      "step": 4040
    },
    {
      "entropy": 5.557367420196533,
      "epoch": 0.3378717006348146,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004992704434675855,
      "loss": 5.4493,
      "mean_token_accuracy": 0.17626117467880248,
      "num_tokens": 7452077.0,
      "step": 4045
    },
    {
      "entropy": 5.5196576595306395,
      "epoch": 0.33828934179752757,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004992680460971218,
      "loss": 5.3824,
      "mean_token_accuracy": 0.17632665485143661,
      "num_tokens": 7461146.0,
      "step": 4050
    },
    {
      "entropy": 5.454443550109863,
      "epoch": 0.33870698296024054,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004992656448005758,
      "loss": 5.3614,
      "mean_token_accuracy": 0.17880332618951797,
      "num_tokens": 7470335.0,
      "step": 4055
    },
    {
      "entropy": 5.605902624130249,
      "epoch": 0.3391246241229536,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004992632395779899,
      "loss": 5.3603,
      "mean_token_accuracy": 0.17632118463516236,
      "num_tokens": 7479458.0,
      "step": 4060
    },
    {
      "entropy": 5.424135828018189,
      "epoch": 0.33954226528566656,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004992608304294061,
      "loss": 5.3186,
      "mean_token_accuracy": 0.18735262602567673,
      "num_tokens": 7488005.0,
      "step": 4065
    },
    {
      "entropy": 5.453942441940308,
      "epoch": 0.33995990644837953,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004992584173548664,
      "loss": 5.3216,
      "mean_token_accuracy": 0.1802588865160942,
      "num_tokens": 7496807.0,
      "step": 4070
    },
    {
      "entropy": 5.533922147750855,
      "epoch": 0.34037754761109257,
      "grad_norm": 0.875,
      "learning_rate": 0.0004992560003544131,
      "loss": 5.3529,
      "mean_token_accuracy": 0.181650410592556,
      "num_tokens": 7505383.0,
      "step": 4075
    },
    {
      "entropy": 5.604040288925171,
      "epoch": 0.34079518877380555,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004992535794280887,
      "loss": 5.4234,
      "mean_token_accuracy": 0.17444170713424684,
      "num_tokens": 7515093.0,
      "step": 4080
    },
    {
      "entropy": 5.5768266201019285,
      "epoch": 0.3412128299365185,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004992511545759354,
      "loss": 5.404,
      "mean_token_accuracy": 0.17743877321481705,
      "num_tokens": 7524067.0,
      "step": 4085
    },
    {
      "entropy": 5.45261516571045,
      "epoch": 0.34163047109923156,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004992487257979958,
      "loss": 5.3395,
      "mean_token_accuracy": 0.17747515738010405,
      "num_tokens": 7533197.0,
      "step": 4090
    },
    {
      "entropy": 5.56703519821167,
      "epoch": 0.34204811226194454,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004992462930943123,
      "loss": 5.3613,
      "mean_token_accuracy": 0.1685240775346756,
      "num_tokens": 7543030.0,
      "step": 4095
    },
    {
      "entropy": 5.554940700531006,
      "epoch": 0.3424657534246575,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004992438564649275,
      "loss": 5.4004,
      "mean_token_accuracy": 0.17544247210025787,
      "num_tokens": 7551556.0,
      "step": 4100
    },
    {
      "entropy": 5.534994220733642,
      "epoch": 0.34288339458737055,
      "grad_norm": 1.0,
      "learning_rate": 0.0004992414159098841,
      "loss": 5.3341,
      "mean_token_accuracy": 0.1762341409921646,
      "num_tokens": 7559284.0,
      "step": 4105
    },
    {
      "entropy": 5.616180562973023,
      "epoch": 0.3433010357500835,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004992389714292247,
      "loss": 5.4462,
      "mean_token_accuracy": 0.1735177829861641,
      "num_tokens": 7568750.0,
      "step": 4110
    },
    {
      "entropy": 5.481297969818115,
      "epoch": 0.3437186769127965,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004992365230229922,
      "loss": 5.306,
      "mean_token_accuracy": 0.18254884332418442,
      "num_tokens": 7578309.0,
      "step": 4115
    },
    {
      "entropy": 5.452652168273926,
      "epoch": 0.34413631807550954,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004992340706912295,
      "loss": 5.3355,
      "mean_token_accuracy": 0.17476286590099335,
      "num_tokens": 7586760.0,
      "step": 4120
    },
    {
      "entropy": 5.4713013648986815,
      "epoch": 0.3445539592382225,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004992316144339794,
      "loss": 5.2942,
      "mean_token_accuracy": 0.177924844622612,
      "num_tokens": 7595472.0,
      "step": 4125
    },
    {
      "entropy": 5.571169805526734,
      "epoch": 0.3449716004009355,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004992291542512852,
      "loss": 5.4747,
      "mean_token_accuracy": 0.16837695986032486,
      "num_tokens": 7605663.0,
      "step": 4130
    },
    {
      "entropy": 5.601088666915894,
      "epoch": 0.34538924156364853,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004992266901431895,
      "loss": 5.3795,
      "mean_token_accuracy": 0.17392918914556504,
      "num_tokens": 7616110.0,
      "step": 4135
    },
    {
      "entropy": 5.511221885681152,
      "epoch": 0.3458068827263615,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004992242221097358,
      "loss": 5.3432,
      "mean_token_accuracy": 0.17719447314739228,
      "num_tokens": 7625311.0,
      "step": 4140
    },
    {
      "entropy": 5.505922746658325,
      "epoch": 0.3462245238890745,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004992217501509672,
      "loss": 5.3718,
      "mean_token_accuracy": 0.17909681648015977,
      "num_tokens": 7634562.0,
      "step": 4145
    },
    {
      "entropy": 5.503519344329834,
      "epoch": 0.3466421650517875,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499219274266927,
      "loss": 5.2993,
      "mean_token_accuracy": 0.18104137033224105,
      "num_tokens": 7643231.0,
      "step": 4150
    },
    {
      "entropy": 5.575137376785278,
      "epoch": 0.3470598062145005,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004992167944576584,
      "loss": 5.4305,
      "mean_token_accuracy": 0.1792694091796875,
      "num_tokens": 7653556.0,
      "step": 4155
    },
    {
      "entropy": 5.548227024078369,
      "epoch": 0.3474774473772135,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499214310723205,
      "loss": 5.358,
      "mean_token_accuracy": 0.1803708553314209,
      "num_tokens": 7662295.0,
      "step": 4160
    },
    {
      "entropy": 5.478659963607788,
      "epoch": 0.3478950885399265,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004992118230636102,
      "loss": 5.3701,
      "mean_token_accuracy": 0.1774037301540375,
      "num_tokens": 7671429.0,
      "step": 4165
    },
    {
      "entropy": 5.472581052780152,
      "epoch": 0.3483127297026395,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004992093314789176,
      "loss": 5.2754,
      "mean_token_accuracy": 0.17473143190145493,
      "num_tokens": 7680589.0,
      "step": 4170
    },
    {
      "entropy": 5.489585113525391,
      "epoch": 0.34873037086535247,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004992068359691707,
      "loss": 5.3204,
      "mean_token_accuracy": 0.1796663448214531,
      "num_tokens": 7689005.0,
      "step": 4175
    },
    {
      "entropy": 5.435379934310913,
      "epoch": 0.3491480120280655,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004992043365344132,
      "loss": 5.2733,
      "mean_token_accuracy": 0.17819844782352448,
      "num_tokens": 7698231.0,
      "step": 4180
    },
    {
      "entropy": 5.5927270412445065,
      "epoch": 0.3495656531907785,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004992018331746891,
      "loss": 5.375,
      "mean_token_accuracy": 0.1827683925628662,
      "num_tokens": 7706678.0,
      "step": 4185
    },
    {
      "entropy": 5.574276304244995,
      "epoch": 0.34998329435349146,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004991993258900418,
      "loss": 5.4317,
      "mean_token_accuracy": 0.17726176381111144,
      "num_tokens": 7715506.0,
      "step": 4190
    },
    {
      "entropy": 5.601906204223633,
      "epoch": 0.3504009355162045,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004991968146805156,
      "loss": 5.3859,
      "mean_token_accuracy": 0.17920258194208144,
      "num_tokens": 7726549.0,
      "step": 4195
    },
    {
      "entropy": 5.544230031967163,
      "epoch": 0.35081857667891747,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004991942995461542,
      "loss": 5.4791,
      "mean_token_accuracy": 0.17360853999853135,
      "num_tokens": 7735563.0,
      "step": 4200
    },
    {
      "entropy": 5.547840881347656,
      "epoch": 0.35123621784163045,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004991917804870018,
      "loss": 5.3981,
      "mean_token_accuracy": 0.1758334830403328,
      "num_tokens": 7746144.0,
      "step": 4205
    },
    {
      "entropy": 5.5478113174438475,
      "epoch": 0.3516538590043435,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004991892575031024,
      "loss": 5.4148,
      "mean_token_accuracy": 0.1782076984643936,
      "num_tokens": 7755394.0,
      "step": 4210
    },
    {
      "entropy": 5.453258275985718,
      "epoch": 0.35207150016705646,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004991867305945001,
      "loss": 5.3828,
      "mean_token_accuracy": 0.18133799433708192,
      "num_tokens": 7764855.0,
      "step": 4215
    },
    {
      "entropy": 5.482728910446167,
      "epoch": 0.35248914132976944,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004991841997612394,
      "loss": 5.2958,
      "mean_token_accuracy": 0.1761842519044876,
      "num_tokens": 7773394.0,
      "step": 4220
    },
    {
      "entropy": 5.574368858337403,
      "epoch": 0.3529067824924825,
      "grad_norm": 1.0,
      "learning_rate": 0.0004991816650033643,
      "loss": 5.3392,
      "mean_token_accuracy": 0.1783423200249672,
      "num_tokens": 7783111.0,
      "step": 4225
    },
    {
      "entropy": 5.590235376358033,
      "epoch": 0.35332442365519545,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004991791263209194,
      "loss": 5.4808,
      "mean_token_accuracy": 0.17077163010835647,
      "num_tokens": 7792569.0,
      "step": 4230
    },
    {
      "entropy": 5.551127338409424,
      "epoch": 0.35374206481790843,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004991765837139491,
      "loss": 5.3859,
      "mean_token_accuracy": 0.18391179144382477,
      "num_tokens": 7801483.0,
      "step": 4235
    },
    {
      "entropy": 5.587145709991455,
      "epoch": 0.35415970598062146,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004991740371824977,
      "loss": 5.4778,
      "mean_token_accuracy": 0.16704935878515242,
      "num_tokens": 7812274.0,
      "step": 4240
    },
    {
      "entropy": 5.458142280578613,
      "epoch": 0.35457734714333444,
      "grad_norm": 0.9375,
      "learning_rate": 0.00049917148672661,
      "loss": 5.317,
      "mean_token_accuracy": 0.18196004182100295,
      "num_tokens": 7821101.0,
      "step": 4245
    },
    {
      "entropy": 5.595104694366455,
      "epoch": 0.3549949883060474,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004991689323463306,
      "loss": 5.3043,
      "mean_token_accuracy": 0.18157126307487487,
      "num_tokens": 7829805.0,
      "step": 4250
    },
    {
      "entropy": 5.519512939453125,
      "epoch": 0.35541262946876045,
      "grad_norm": 0.875,
      "learning_rate": 0.0004991663740417042,
      "loss": 5.39,
      "mean_token_accuracy": 0.17301483899354936,
      "num_tokens": 7839772.0,
      "step": 4255
    },
    {
      "entropy": 5.502783203125,
      "epoch": 0.35583027063147343,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004991638118127757,
      "loss": 5.4796,
      "mean_token_accuracy": 0.16273352205753328,
      "num_tokens": 7850290.0,
      "step": 4260
    },
    {
      "entropy": 5.674912071228027,
      "epoch": 0.3562479117941864,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004991612456595896,
      "loss": 5.4473,
      "mean_token_accuracy": 0.17262645214796066,
      "num_tokens": 7860180.0,
      "step": 4265
    },
    {
      "entropy": 5.3785093307495115,
      "epoch": 0.35666555295689945,
      "grad_norm": 1.0,
      "learning_rate": 0.0004991586755821913,
      "loss": 5.314,
      "mean_token_accuracy": 0.1895632728934288,
      "num_tokens": 7869625.0,
      "step": 4270
    },
    {
      "entropy": 5.412072086334229,
      "epoch": 0.3570831941196124,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004991561015806254,
      "loss": 5.2757,
      "mean_token_accuracy": 0.17349614948034286,
      "num_tokens": 7878671.0,
      "step": 4275
    },
    {
      "entropy": 5.549697923660278,
      "epoch": 0.3575008352823254,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004991535236549372,
      "loss": 5.3936,
      "mean_token_accuracy": 0.1825173482298851,
      "num_tokens": 7887225.0,
      "step": 4280
    },
    {
      "entropy": 5.437959623336792,
      "epoch": 0.35791847644503844,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004991509418051718,
      "loss": 5.275,
      "mean_token_accuracy": 0.1859090059995651,
      "num_tokens": 7895504.0,
      "step": 4285
    },
    {
      "entropy": 5.597351455688477,
      "epoch": 0.3583361176077514,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004991483560313743,
      "loss": 5.3474,
      "mean_token_accuracy": 0.1786056265234947,
      "num_tokens": 7904875.0,
      "step": 4290
    },
    {
      "entropy": 5.547267007827759,
      "epoch": 0.3587537587704644,
      "grad_norm": 0.84375,
      "learning_rate": 0.00049914576633359,
      "loss": 5.4089,
      "mean_token_accuracy": 0.17962569296360015,
      "num_tokens": 7914747.0,
      "step": 4295
    },
    {
      "entropy": 5.550775718688965,
      "epoch": 0.3591713999331774,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004991431727118643,
      "loss": 5.3789,
      "mean_token_accuracy": 0.17607147842645646,
      "num_tokens": 7923858.0,
      "step": 4300
    },
    {
      "entropy": 5.456276893615723,
      "epoch": 0.3595890410958904,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004991405751662426,
      "loss": 5.3348,
      "mean_token_accuracy": 0.17600340992212296,
      "num_tokens": 7932396.0,
      "step": 4305
    },
    {
      "entropy": 5.554454755783081,
      "epoch": 0.3600066822586034,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004991379736967703,
      "loss": 5.3394,
      "mean_token_accuracy": 0.17855304926633836,
      "num_tokens": 7940763.0,
      "step": 4310
    },
    {
      "entropy": 5.432330799102783,
      "epoch": 0.3604243234213164,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004991353683034929,
      "loss": 5.3381,
      "mean_token_accuracy": 0.17469086796045302,
      "num_tokens": 7949424.0,
      "step": 4315
    },
    {
      "entropy": 5.48876166343689,
      "epoch": 0.3608419645840294,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004991327589864563,
      "loss": 5.3158,
      "mean_token_accuracy": 0.18421016335487367,
      "num_tokens": 7958805.0,
      "step": 4320
    },
    {
      "entropy": 5.6209570407867435,
      "epoch": 0.3612596057467424,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004991301457457057,
      "loss": 5.4406,
      "mean_token_accuracy": 0.18161950409412383,
      "num_tokens": 7967859.0,
      "step": 4325
    },
    {
      "entropy": 5.6464393615722654,
      "epoch": 0.3616772469094554,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004991275285812873,
      "loss": 5.5518,
      "mean_token_accuracy": 0.16830166578292846,
      "num_tokens": 7978357.0,
      "step": 4330
    },
    {
      "entropy": 5.612295866012573,
      "epoch": 0.3620948880721684,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004991249074932466,
      "loss": 5.3906,
      "mean_token_accuracy": 0.18021910786628723,
      "num_tokens": 7986637.0,
      "step": 4335
    },
    {
      "entropy": 5.457355737686157,
      "epoch": 0.36251252923488136,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004991222824816296,
      "loss": 5.3201,
      "mean_token_accuracy": 0.18055709600448608,
      "num_tokens": 7995954.0,
      "step": 4340
    },
    {
      "entropy": 5.544766855239868,
      "epoch": 0.3629301703975944,
      "grad_norm": 1.0,
      "learning_rate": 0.0004991196535464824,
      "loss": 5.3961,
      "mean_token_accuracy": 0.1681934341788292,
      "num_tokens": 8005193.0,
      "step": 4345
    },
    {
      "entropy": 5.667155790328979,
      "epoch": 0.3633478115603074,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004991170206878508,
      "loss": 5.4276,
      "mean_token_accuracy": 0.17900248616933823,
      "num_tokens": 8014419.0,
      "step": 4350
    },
    {
      "entropy": 5.507767391204834,
      "epoch": 0.36376545272302035,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499114383905781,
      "loss": 5.37,
      "mean_token_accuracy": 0.17622680068016053,
      "num_tokens": 8024133.0,
      "step": 4355
    },
    {
      "entropy": 5.471643590927124,
      "epoch": 0.3641830938857334,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004991117432003192,
      "loss": 5.3182,
      "mean_token_accuracy": 0.17904623448848725,
      "num_tokens": 8033059.0,
      "step": 4360
    },
    {
      "entropy": 5.583207654953003,
      "epoch": 0.36460073504844637,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004991090985715116,
      "loss": 5.3764,
      "mean_token_accuracy": 0.18116264194250106,
      "num_tokens": 8041866.0,
      "step": 4365
    },
    {
      "entropy": 5.4930966854095455,
      "epoch": 0.36501837621115935,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004991064500194043,
      "loss": 5.411,
      "mean_token_accuracy": 0.17382517755031585,
      "num_tokens": 8051541.0,
      "step": 4370
    },
    {
      "entropy": 5.569479703903198,
      "epoch": 0.3654360173738724,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004991037975440439,
      "loss": 5.4767,
      "mean_token_accuracy": 0.1752861112356186,
      "num_tokens": 8061964.0,
      "step": 4375
    },
    {
      "entropy": 5.4403942108154295,
      "epoch": 0.36585365853658536,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004991011411454769,
      "loss": 5.3101,
      "mean_token_accuracy": 0.17651308029890062,
      "num_tokens": 8071006.0,
      "step": 4380
    },
    {
      "entropy": 5.476379299163819,
      "epoch": 0.36627129969929834,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004990984808237495,
      "loss": 5.3072,
      "mean_token_accuracy": 0.18224988579750062,
      "num_tokens": 8080104.0,
      "step": 4385
    },
    {
      "entropy": 5.558085680007935,
      "epoch": 0.36668894086201137,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004990958165789086,
      "loss": 5.3518,
      "mean_token_accuracy": 0.17947374880313874,
      "num_tokens": 8089397.0,
      "step": 4390
    },
    {
      "entropy": 5.508536863327026,
      "epoch": 0.36710658202472435,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004990931484110005,
      "loss": 5.3851,
      "mean_token_accuracy": 0.17887150943279267,
      "num_tokens": 8099055.0,
      "step": 4395
    },
    {
      "entropy": 5.516378784179688,
      "epoch": 0.3675242231874373,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004990904763200722,
      "loss": 5.3377,
      "mean_token_accuracy": 0.18014725893735886,
      "num_tokens": 8107609.0,
      "step": 4400
    },
    {
      "entropy": 5.577397775650025,
      "epoch": 0.36794186435015036,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004990878003061705,
      "loss": 5.3474,
      "mean_token_accuracy": 0.18444286584854125,
      "num_tokens": 8116905.0,
      "step": 4405
    },
    {
      "entropy": 5.450827980041504,
      "epoch": 0.36835950551286334,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004990851203693419,
      "loss": 5.3419,
      "mean_token_accuracy": 0.18142086863517762,
      "num_tokens": 8125580.0,
      "step": 4410
    },
    {
      "entropy": 5.46176962852478,
      "epoch": 0.3687771466755763,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004990824365096336,
      "loss": 5.2505,
      "mean_token_accuracy": 0.17691653221845627,
      "num_tokens": 8134568.0,
      "step": 4415
    },
    {
      "entropy": 5.43082857131958,
      "epoch": 0.36919478783828935,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004990797487270926,
      "loss": 5.2515,
      "mean_token_accuracy": 0.1833440974354744,
      "num_tokens": 8143044.0,
      "step": 4420
    },
    {
      "entropy": 5.456305503845215,
      "epoch": 0.36961242900100233,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004990770570217659,
      "loss": 5.3589,
      "mean_token_accuracy": 0.17786064743995667,
      "num_tokens": 8153701.0,
      "step": 4425
    },
    {
      "entropy": 5.45921573638916,
      "epoch": 0.37003007016371536,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004990743613937005,
      "loss": 5.2591,
      "mean_token_accuracy": 0.1828637421131134,
      "num_tokens": 8163264.0,
      "step": 4430
    },
    {
      "entropy": 5.44285397529602,
      "epoch": 0.37044771132642834,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004990716618429436,
      "loss": 5.3695,
      "mean_token_accuracy": 0.1726057156920433,
      "num_tokens": 8173332.0,
      "step": 4435
    },
    {
      "entropy": 5.58035569190979,
      "epoch": 0.3708653524891413,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004990689583695428,
      "loss": 5.358,
      "mean_token_accuracy": 0.17852265238761902,
      "num_tokens": 8181881.0,
      "step": 4440
    },
    {
      "entropy": 5.60650897026062,
      "epoch": 0.37128299365185435,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499066250973545,
      "loss": 5.4015,
      "mean_token_accuracy": 0.1773255169391632,
      "num_tokens": 8190771.0,
      "step": 4445
    },
    {
      "entropy": 5.46733546257019,
      "epoch": 0.37170063481456733,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004990635396549978,
      "loss": 5.3369,
      "mean_token_accuracy": 0.1736823573708534,
      "num_tokens": 8200740.0,
      "step": 4450
    },
    {
      "entropy": 5.373090839385986,
      "epoch": 0.3721182759772803,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004990608244139486,
      "loss": 5.3298,
      "mean_token_accuracy": 0.1764607921242714,
      "num_tokens": 8209935.0,
      "step": 4455
    },
    {
      "entropy": 5.527897834777832,
      "epoch": 0.37253591713999334,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499058105250445,
      "loss": 5.3954,
      "mean_token_accuracy": 0.17625040262937547,
      "num_tokens": 8219137.0,
      "step": 4460
    },
    {
      "entropy": 5.578756952285767,
      "epoch": 0.3729535583027063,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004990553821645346,
      "loss": 5.3682,
      "mean_token_accuracy": 0.1824690580368042,
      "num_tokens": 8228076.0,
      "step": 4465
    },
    {
      "entropy": 5.459502077102661,
      "epoch": 0.3733711994654193,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499052655156265,
      "loss": 5.3303,
      "mean_token_accuracy": 0.17939531207084655,
      "num_tokens": 8237581.0,
      "step": 4470
    },
    {
      "entropy": 5.537692546844482,
      "epoch": 0.37378884062813234,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004990499242256839,
      "loss": 5.3731,
      "mean_token_accuracy": 0.17038365006446837,
      "num_tokens": 8247384.0,
      "step": 4475
    },
    {
      "entropy": 5.566897964477539,
      "epoch": 0.3742064817908453,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004990471893728394,
      "loss": 5.4421,
      "mean_token_accuracy": 0.17765509486198425,
      "num_tokens": 8256383.0,
      "step": 4480
    },
    {
      "entropy": 5.507769775390625,
      "epoch": 0.3746241229535583,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499044450597779,
      "loss": 5.2915,
      "mean_token_accuracy": 0.1839849293231964,
      "num_tokens": 8264848.0,
      "step": 4485
    },
    {
      "entropy": 5.4471183776855465,
      "epoch": 0.3750417641162713,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004990417079005508,
      "loss": 5.3257,
      "mean_token_accuracy": 0.176951003074646,
      "num_tokens": 8274403.0,
      "step": 4490
    },
    {
      "entropy": 5.457308053970337,
      "epoch": 0.3754594052789843,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499038961281203,
      "loss": 5.2523,
      "mean_token_accuracy": 0.18299420326948165,
      "num_tokens": 8282700.0,
      "step": 4495
    },
    {
      "entropy": 5.453111219406128,
      "epoch": 0.3758770464416973,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990362107397834,
      "loss": 5.2806,
      "mean_token_accuracy": 0.18726570159196854,
      "num_tokens": 8291718.0,
      "step": 4500
    },
    {
      "entropy": 5.469309854507446,
      "epoch": 0.3762946876044103,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004990334562763401,
      "loss": 5.3437,
      "mean_token_accuracy": 0.1799112230539322,
      "num_tokens": 8301113.0,
      "step": 4505
    },
    {
      "entropy": 5.592342662811279,
      "epoch": 0.3767123287671233,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004990306978909216,
      "loss": 5.4413,
      "mean_token_accuracy": 0.1775081366300583,
      "num_tokens": 8310357.0,
      "step": 4510
    },
    {
      "entropy": 5.545265483856201,
      "epoch": 0.3771299699298363,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004990279355835761,
      "loss": 5.3739,
      "mean_token_accuracy": 0.18223407715559006,
      "num_tokens": 8319680.0,
      "step": 4515
    },
    {
      "entropy": 5.543308401107788,
      "epoch": 0.3775476110925493,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004990251693543519,
      "loss": 5.3919,
      "mean_token_accuracy": 0.17246417850255966,
      "num_tokens": 8329127.0,
      "step": 4520
    },
    {
      "entropy": 5.413548469543457,
      "epoch": 0.3779652522552623,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004990223992032975,
      "loss": 5.2381,
      "mean_token_accuracy": 0.18571756780147552,
      "num_tokens": 8338308.0,
      "step": 4525
    },
    {
      "entropy": 5.495697927474976,
      "epoch": 0.37838289341797526,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990196251304612,
      "loss": 5.2682,
      "mean_token_accuracy": 0.18460340946912765,
      "num_tokens": 8347491.0,
      "step": 4530
    },
    {
      "entropy": 5.389685535430909,
      "epoch": 0.3788005345806883,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000499016847135892,
      "loss": 5.2748,
      "mean_token_accuracy": 0.18167904317378997,
      "num_tokens": 8357112.0,
      "step": 4535
    },
    {
      "entropy": 5.4584802150726315,
      "epoch": 0.3792181757434013,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499014065219638,
      "loss": 5.2835,
      "mean_token_accuracy": 0.18662642538547516,
      "num_tokens": 8366285.0,
      "step": 4540
    },
    {
      "entropy": 5.436573553085327,
      "epoch": 0.37963581690611425,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004990112793817481,
      "loss": 5.2019,
      "mean_token_accuracy": 0.19027214348316193,
      "num_tokens": 8374362.0,
      "step": 4545
    },
    {
      "entropy": 5.482431554794312,
      "epoch": 0.3800534580688273,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004990084896222713,
      "loss": 5.4155,
      "mean_token_accuracy": 0.1810782879590988,
      "num_tokens": 8383297.0,
      "step": 4550
    },
    {
      "entropy": 5.507790613174438,
      "epoch": 0.38047109923154027,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004990056959412562,
      "loss": 5.3025,
      "mean_token_accuracy": 0.18850304037332535,
      "num_tokens": 8391926.0,
      "step": 4555
    },
    {
      "entropy": 5.538436031341552,
      "epoch": 0.38088874039425324,
      "grad_norm": 1.0,
      "learning_rate": 0.0004990028983387517,
      "loss": 5.2828,
      "mean_token_accuracy": 0.18198327869176864,
      "num_tokens": 8401226.0,
      "step": 4560
    },
    {
      "entropy": 5.5002540111541744,
      "epoch": 0.3813063815569663,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004990000968148068,
      "loss": 5.3135,
      "mean_token_accuracy": 0.18409028202295302,
      "num_tokens": 8409966.0,
      "step": 4565
    },
    {
      "entropy": 5.373306322097778,
      "epoch": 0.38172402271967926,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004989972913694707,
      "loss": 5.1614,
      "mean_token_accuracy": 0.19667479544878005,
      "num_tokens": 8418281.0,
      "step": 4570
    },
    {
      "entropy": 5.353603839874268,
      "epoch": 0.38214166388239224,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004989944820027923,
      "loss": 5.2698,
      "mean_token_accuracy": 0.18271706998348236,
      "num_tokens": 8427436.0,
      "step": 4575
    },
    {
      "entropy": 5.495364093780518,
      "epoch": 0.38255930504510527,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004989916687148209,
      "loss": 5.3205,
      "mean_token_accuracy": 0.18873778879642486,
      "num_tokens": 8436066.0,
      "step": 4580
    },
    {
      "entropy": 5.485512638092041,
      "epoch": 0.38297694620781825,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004989888515056057,
      "loss": 5.3186,
      "mean_token_accuracy": 0.1796857014298439,
      "num_tokens": 8445572.0,
      "step": 4585
    },
    {
      "entropy": 5.50971040725708,
      "epoch": 0.3833945873705312,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000498986030375196,
      "loss": 5.3549,
      "mean_token_accuracy": 0.17802461683750154,
      "num_tokens": 8454781.0,
      "step": 4590
    },
    {
      "entropy": 5.470432424545288,
      "epoch": 0.38381222853324426,
      "grad_norm": 1.0,
      "learning_rate": 0.0004989832053236413,
      "loss": 5.4422,
      "mean_token_accuracy": 0.17676590532064437,
      "num_tokens": 8463239.0,
      "step": 4595
    },
    {
      "entropy": 5.530746984481811,
      "epoch": 0.38422986969595724,
      "grad_norm": 1.03125,
      "learning_rate": 0.000498980376350991,
      "loss": 5.3357,
      "mean_token_accuracy": 0.18066398650407792,
      "num_tokens": 8472664.0,
      "step": 4600
    },
    {
      "entropy": 5.528481578826904,
      "epoch": 0.3846475108586702,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004989775434572945,
      "loss": 5.3998,
      "mean_token_accuracy": 0.17462089359760286,
      "num_tokens": 8483149.0,
      "step": 4605
    },
    {
      "entropy": 5.55580587387085,
      "epoch": 0.38506515202138325,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004989747066426015,
      "loss": 5.3298,
      "mean_token_accuracy": 0.18469495177268982,
      "num_tokens": 8492297.0,
      "step": 4610
    },
    {
      "entropy": 5.473330402374268,
      "epoch": 0.38548279318409623,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004989718659069618,
      "loss": 5.408,
      "mean_token_accuracy": 0.17273375242948533,
      "num_tokens": 8502092.0,
      "step": 4615
    },
    {
      "entropy": 5.5471116065979,
      "epoch": 0.3859004343468092,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000498969021250425,
      "loss": 5.3257,
      "mean_token_accuracy": 0.18129046261310577,
      "num_tokens": 8510816.0,
      "step": 4620
    },
    {
      "entropy": 5.526094007492065,
      "epoch": 0.38631807550952224,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004989661726730407,
      "loss": 5.3328,
      "mean_token_accuracy": 0.18268937319517137,
      "num_tokens": 8520055.0,
      "step": 4625
    },
    {
      "entropy": 5.496733140945435,
      "epoch": 0.3867357166722352,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004989633201748591,
      "loss": 5.3321,
      "mean_token_accuracy": 0.1819985553622246,
      "num_tokens": 8528271.0,
      "step": 4630
    },
    {
      "entropy": 5.4477317333221436,
      "epoch": 0.3871533578349482,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004989604637559299,
      "loss": 5.2482,
      "mean_token_accuracy": 0.19254219830036162,
      "num_tokens": 8537360.0,
      "step": 4635
    },
    {
      "entropy": 5.391276121139526,
      "epoch": 0.38757099899766123,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004989576034163033,
      "loss": 5.2078,
      "mean_token_accuracy": 0.1942689061164856,
      "num_tokens": 8546574.0,
      "step": 4640
    },
    {
      "entropy": 5.435660266876221,
      "epoch": 0.3879886401603742,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004989547391560293,
      "loss": 5.304,
      "mean_token_accuracy": 0.17881359308958053,
      "num_tokens": 8555581.0,
      "step": 4645
    },
    {
      "entropy": 5.424041223526001,
      "epoch": 0.3884062813230872,
      "grad_norm": 0.90625,
      "learning_rate": 0.000498951870975158,
      "loss": 5.3433,
      "mean_token_accuracy": 0.18788295835256577,
      "num_tokens": 8564539.0,
      "step": 4650
    },
    {
      "entropy": 5.526417779922485,
      "epoch": 0.3888239224858002,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004989489988737395,
      "loss": 5.3835,
      "mean_token_accuracy": 0.17857650816440582,
      "num_tokens": 8573804.0,
      "step": 4655
    },
    {
      "entropy": 5.497582626342774,
      "epoch": 0.3892415636485132,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004989461228518243,
      "loss": 5.3147,
      "mean_token_accuracy": 0.18248738050460817,
      "num_tokens": 8583262.0,
      "step": 4660
    },
    {
      "entropy": 5.461036443710327,
      "epoch": 0.3896592048112262,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004989432429094627,
      "loss": 5.3674,
      "mean_token_accuracy": 0.17988524287939073,
      "num_tokens": 8592763.0,
      "step": 4665
    },
    {
      "entropy": 5.541605806350708,
      "epoch": 0.3900768459739392,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004989403590467049,
      "loss": 5.241,
      "mean_token_accuracy": 0.1874819204211235,
      "num_tokens": 8601884.0,
      "step": 4670
    },
    {
      "entropy": 5.424175214767456,
      "epoch": 0.3904944871366522,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004989374712636017,
      "loss": 5.3771,
      "mean_token_accuracy": 0.17700473666191102,
      "num_tokens": 8611207.0,
      "step": 4675
    },
    {
      "entropy": 5.386784982681275,
      "epoch": 0.39091212829936517,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004989345795602036,
      "loss": 5.1197,
      "mean_token_accuracy": 0.20138658434152604,
      "num_tokens": 8620104.0,
      "step": 4680
    },
    {
      "entropy": 5.517288446426392,
      "epoch": 0.3913297694620782,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004989316839365609,
      "loss": 5.3427,
      "mean_token_accuracy": 0.17836138755083084,
      "num_tokens": 8629552.0,
      "step": 4685
    },
    {
      "entropy": 5.458013391494751,
      "epoch": 0.3917474106247912,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004989287843927246,
      "loss": 5.2737,
      "mean_token_accuracy": 0.18257849514484406,
      "num_tokens": 8638961.0,
      "step": 4690
    },
    {
      "entropy": 5.461923217773437,
      "epoch": 0.39216505178750416,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004989258809287455,
      "loss": 5.2857,
      "mean_token_accuracy": 0.18226059675216674,
      "num_tokens": 8648051.0,
      "step": 4695
    },
    {
      "entropy": 5.4944733619689945,
      "epoch": 0.3925826929502172,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004989229735446743,
      "loss": 5.2724,
      "mean_token_accuracy": 0.18453247100114822,
      "num_tokens": 8656863.0,
      "step": 4700
    },
    {
      "entropy": 5.319275617599487,
      "epoch": 0.39300033411293017,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004989200622405619,
      "loss": 5.1132,
      "mean_token_accuracy": 0.19016069620847703,
      "num_tokens": 8665455.0,
      "step": 4705
    },
    {
      "entropy": 5.358971881866455,
      "epoch": 0.39341797527564315,
      "grad_norm": 1.0,
      "learning_rate": 0.0004989171470164592,
      "loss": 5.2225,
      "mean_token_accuracy": 0.19097758531570436,
      "num_tokens": 8674119.0,
      "step": 4710
    },
    {
      "entropy": 5.506095552444458,
      "epoch": 0.3938356164383562,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004989142278724173,
      "loss": 5.3445,
      "mean_token_accuracy": 0.17946820110082626,
      "num_tokens": 8682979.0,
      "step": 4715
    },
    {
      "entropy": 5.4872089385986325,
      "epoch": 0.39425325760106916,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004989113048084874,
      "loss": 5.2247,
      "mean_token_accuracy": 0.18043747395277024,
      "num_tokens": 8692392.0,
      "step": 4720
    },
    {
      "entropy": 5.389395666122437,
      "epoch": 0.39467089876378214,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004989083778247206,
      "loss": 5.2939,
      "mean_token_accuracy": 0.18280829340219498,
      "num_tokens": 8700932.0,
      "step": 4725
    },
    {
      "entropy": 5.57949423789978,
      "epoch": 0.3950885399264952,
      "grad_norm": 1.0,
      "learning_rate": 0.000498905446921168,
      "loss": 5.4419,
      "mean_token_accuracy": 0.17536427080631256,
      "num_tokens": 8710683.0,
      "step": 4730
    },
    {
      "entropy": 5.397296047210693,
      "epoch": 0.39550618108920815,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004989025120978812,
      "loss": 5.294,
      "mean_token_accuracy": 0.18484196960926055,
      "num_tokens": 8719939.0,
      "step": 4735
    },
    {
      "entropy": 5.535792255401612,
      "epoch": 0.39592382225192113,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004988995733549113,
      "loss": 5.3488,
      "mean_token_accuracy": 0.18545869886875152,
      "num_tokens": 8729275.0,
      "step": 4740
    },
    {
      "entropy": 5.433511543273926,
      "epoch": 0.39634146341463417,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004988966306923099,
      "loss": 5.2334,
      "mean_token_accuracy": 0.18577802181243896,
      "num_tokens": 8738048.0,
      "step": 4745
    },
    {
      "entropy": 5.423494482040406,
      "epoch": 0.39675910457734714,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004988936841101284,
      "loss": 5.2932,
      "mean_token_accuracy": 0.1819676548242569,
      "num_tokens": 8746967.0,
      "step": 4750
    },
    {
      "entropy": 5.405490875244141,
      "epoch": 0.3971767457400601,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004988907336084186,
      "loss": 5.2492,
      "mean_token_accuracy": 0.18655565977096558,
      "num_tokens": 8754908.0,
      "step": 4755
    },
    {
      "entropy": 5.470695543289184,
      "epoch": 0.39759438690277316,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004988877791872319,
      "loss": 5.2764,
      "mean_token_accuracy": 0.19156011939048767,
      "num_tokens": 8764489.0,
      "step": 4760
    },
    {
      "entropy": 5.56093111038208,
      "epoch": 0.39801202806548613,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004988848208466202,
      "loss": 5.4406,
      "mean_token_accuracy": 0.1676798105239868,
      "num_tokens": 8774365.0,
      "step": 4765
    },
    {
      "entropy": 5.436862516403198,
      "epoch": 0.3984296692281991,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004988818585866351,
      "loss": 5.2434,
      "mean_token_accuracy": 0.18682124018669127,
      "num_tokens": 8783538.0,
      "step": 4770
    },
    {
      "entropy": 5.409594964981079,
      "epoch": 0.39884731039091215,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004988788924073288,
      "loss": 5.345,
      "mean_token_accuracy": 0.18234388828277587,
      "num_tokens": 8792171.0,
      "step": 4775
    },
    {
      "entropy": 5.494739103317261,
      "epoch": 0.3992649515536251,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004988759223087528,
      "loss": 5.3278,
      "mean_token_accuracy": 0.18156281113624573,
      "num_tokens": 8801325.0,
      "step": 4780
    },
    {
      "entropy": 5.475518465042114,
      "epoch": 0.3996825927163381,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004988729482909595,
      "loss": 5.2771,
      "mean_token_accuracy": 0.1823272153735161,
      "num_tokens": 8810566.0,
      "step": 4785
    },
    {
      "entropy": 5.383138942718506,
      "epoch": 0.40010023387905114,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004988699703540006,
      "loss": 5.1958,
      "mean_token_accuracy": 0.19508769363164902,
      "num_tokens": 8819735.0,
      "step": 4790
    },
    {
      "entropy": 5.4445481300354,
      "epoch": 0.4005178750417641,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004988669884979284,
      "loss": 5.3488,
      "mean_token_accuracy": 0.18017641752958297,
      "num_tokens": 8828820.0,
      "step": 4795
    },
    {
      "entropy": 5.4503051280975345,
      "epoch": 0.4009355162044771,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004988640027227951,
      "loss": 5.2479,
      "mean_token_accuracy": 0.17885902374982834,
      "num_tokens": 8837858.0,
      "step": 4800
    },
    {
      "entropy": 5.452023696899414,
      "epoch": 0.4013531573671901,
      "grad_norm": 0.90625,
      "learning_rate": 0.000498861013028653,
      "loss": 5.2272,
      "mean_token_accuracy": 0.18821717500686647,
      "num_tokens": 8847172.0,
      "step": 4805
    },
    {
      "entropy": 5.41439995765686,
      "epoch": 0.4017707985299031,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004988580194155544,
      "loss": 5.2373,
      "mean_token_accuracy": 0.18068647086620332,
      "num_tokens": 8856700.0,
      "step": 4810
    },
    {
      "entropy": 5.312618970870972,
      "epoch": 0.4021884396926161,
      "grad_norm": 1.0,
      "learning_rate": 0.0004988550218835517,
      "loss": 5.1438,
      "mean_token_accuracy": 0.1914048358798027,
      "num_tokens": 8866108.0,
      "step": 4815
    },
    {
      "entropy": 5.403158044815063,
      "epoch": 0.4026060808553291,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004988520204326975,
      "loss": 5.2304,
      "mean_token_accuracy": 0.19235294461250305,
      "num_tokens": 8874833.0,
      "step": 4820
    },
    {
      "entropy": 5.459530878067016,
      "epoch": 0.4030237220180421,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004988490150630441,
      "loss": 5.3115,
      "mean_token_accuracy": 0.18563491106033325,
      "num_tokens": 8884211.0,
      "step": 4825
    },
    {
      "entropy": 5.551653432846069,
      "epoch": 0.4034413631807551,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004988460057746443,
      "loss": 5.4189,
      "mean_token_accuracy": 0.1836405724287033,
      "num_tokens": 8893443.0,
      "step": 4830
    },
    {
      "entropy": 5.432301998138428,
      "epoch": 0.4038590043434681,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004988429925675507,
      "loss": 5.2873,
      "mean_token_accuracy": 0.18899856507778168,
      "num_tokens": 8903544.0,
      "step": 4835
    },
    {
      "entropy": 5.460377502441406,
      "epoch": 0.4042766455061811,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004988399754418161,
      "loss": 5.297,
      "mean_token_accuracy": 0.1854369044303894,
      "num_tokens": 8912270.0,
      "step": 4840
    },
    {
      "entropy": 5.5201929092407225,
      "epoch": 0.40469428666889407,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004988369543974932,
      "loss": 5.3989,
      "mean_token_accuracy": 0.17308679968118668,
      "num_tokens": 8922155.0,
      "step": 4845
    },
    {
      "entropy": 5.493629360198975,
      "epoch": 0.4051119278316071,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004988339294346351,
      "loss": 5.3926,
      "mean_token_accuracy": 0.1777765557169914,
      "num_tokens": 8931844.0,
      "step": 4850
    },
    {
      "entropy": 5.484465503692627,
      "epoch": 0.4055295689943201,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004988309005532946,
      "loss": 5.2513,
      "mean_token_accuracy": 0.18598968535661697,
      "num_tokens": 8940874.0,
      "step": 4855
    },
    {
      "entropy": 5.438623046875,
      "epoch": 0.40594721015703306,
      "grad_norm": 1.0,
      "learning_rate": 0.0004988278677535247,
      "loss": 5.3342,
      "mean_token_accuracy": 0.18418071120977403,
      "num_tokens": 8950108.0,
      "step": 4860
    },
    {
      "entropy": 5.481055879592896,
      "epoch": 0.4063648513197461,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004988248310353788,
      "loss": 5.3317,
      "mean_token_accuracy": 0.17852703332901002,
      "num_tokens": 8960610.0,
      "step": 4865
    },
    {
      "entropy": 5.394073104858398,
      "epoch": 0.40678249248245907,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004988217903989095,
      "loss": 5.2801,
      "mean_token_accuracy": 0.18238589465618132,
      "num_tokens": 8969853.0,
      "step": 4870
    },
    {
      "entropy": 5.5094287395477295,
      "epoch": 0.40720013364517205,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004988187458441704,
      "loss": 5.2421,
      "mean_token_accuracy": 0.18265690356492997,
      "num_tokens": 8979800.0,
      "step": 4875
    },
    {
      "entropy": 5.46052656173706,
      "epoch": 0.4076177748078851,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004988156973712149,
      "loss": 5.2954,
      "mean_token_accuracy": 0.1851310148835182,
      "num_tokens": 8988667.0,
      "step": 4880
    },
    {
      "entropy": 5.3278031826019285,
      "epoch": 0.40803541597059806,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004988126449800961,
      "loss": 5.1924,
      "mean_token_accuracy": 0.1896505370736122,
      "num_tokens": 8997385.0,
      "step": 4885
    },
    {
      "entropy": 5.415603303909302,
      "epoch": 0.40845305713331104,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004988095886708676,
      "loss": 5.2101,
      "mean_token_accuracy": 0.1889025777578354,
      "num_tokens": 9005999.0,
      "step": 4890
    },
    {
      "entropy": 5.468580627441407,
      "epoch": 0.40887069829602407,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004988065284435828,
      "loss": 5.3484,
      "mean_token_accuracy": 0.18623862713575362,
      "num_tokens": 9014837.0,
      "step": 4895
    },
    {
      "entropy": 5.472664737701416,
      "epoch": 0.40928833945873705,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004988034642982954,
      "loss": 5.3709,
      "mean_token_accuracy": 0.18317697942256927,
      "num_tokens": 9023956.0,
      "step": 4900
    },
    {
      "entropy": 5.5125658988952635,
      "epoch": 0.40970598062145,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004988003962350589,
      "loss": 5.3474,
      "mean_token_accuracy": 0.17757421284914016,
      "num_tokens": 9032945.0,
      "step": 4905
    },
    {
      "entropy": 5.483496046066284,
      "epoch": 0.41012362178416306,
      "grad_norm": 0.96875,
      "learning_rate": 0.000498797324253927,
      "loss": 5.2632,
      "mean_token_accuracy": 0.1916464239358902,
      "num_tokens": 9042109.0,
      "step": 4910
    },
    {
      "entropy": 5.38888111114502,
      "epoch": 0.41054126294687604,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004987942483549537,
      "loss": 5.2561,
      "mean_token_accuracy": 0.18333554714918138,
      "num_tokens": 9052210.0,
      "step": 4915
    },
    {
      "entropy": 5.451848936080933,
      "epoch": 0.410958904109589,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004987911685381926,
      "loss": 5.2162,
      "mean_token_accuracy": 0.19000166207551955,
      "num_tokens": 9060089.0,
      "step": 4920
    },
    {
      "entropy": 5.544635057449341,
      "epoch": 0.41137654527230205,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004987880848036977,
      "loss": 5.3231,
      "mean_token_accuracy": 0.17790356874465943,
      "num_tokens": 9069235.0,
      "step": 4925
    },
    {
      "entropy": 5.388187837600708,
      "epoch": 0.41179418643501503,
      "grad_norm": 1.0,
      "learning_rate": 0.0004987849971515232,
      "loss": 5.2783,
      "mean_token_accuracy": 0.18961505889892577,
      "num_tokens": 9077509.0,
      "step": 4930
    },
    {
      "entropy": 5.451679229736328,
      "epoch": 0.412211827597728,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004987819055817227,
      "loss": 5.191,
      "mean_token_accuracy": 0.19866640269756317,
      "num_tokens": 9085493.0,
      "step": 4935
    },
    {
      "entropy": 5.325146675109863,
      "epoch": 0.41262946876044104,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004987788100943507,
      "loss": 5.1671,
      "mean_token_accuracy": 0.18661801069974898,
      "num_tokens": 9094748.0,
      "step": 4940
    },
    {
      "entropy": 5.4416662693023685,
      "epoch": 0.413047109923154,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004987757106894613,
      "loss": 5.2297,
      "mean_token_accuracy": 0.18990323394536973,
      "num_tokens": 9103678.0,
      "step": 4945
    },
    {
      "entropy": 5.42334017753601,
      "epoch": 0.413464751085867,
      "grad_norm": 1.0,
      "learning_rate": 0.0004987726073671086,
      "loss": 5.2888,
      "mean_token_accuracy": 0.18152173459529877,
      "num_tokens": 9112851.0,
      "step": 4950
    },
    {
      "entropy": 5.536287927627564,
      "epoch": 0.41388239224858003,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004987695001273471,
      "loss": 5.5206,
      "mean_token_accuracy": 0.1700655922293663,
      "num_tokens": 9122106.0,
      "step": 4955
    },
    {
      "entropy": 5.483275032043457,
      "epoch": 0.414300033411293,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004987663889702313,
      "loss": 5.2262,
      "mean_token_accuracy": 0.1920120432972908,
      "num_tokens": 9131580.0,
      "step": 4960
    },
    {
      "entropy": 5.448633861541748,
      "epoch": 0.414717674574006,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004987632738958155,
      "loss": 5.3563,
      "mean_token_accuracy": 0.18088240474462508,
      "num_tokens": 9141642.0,
      "step": 4965
    },
    {
      "entropy": 5.572747993469238,
      "epoch": 0.415135315736719,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004987601549041541,
      "loss": 5.3976,
      "mean_token_accuracy": 0.17336764335632324,
      "num_tokens": 9151192.0,
      "step": 4970
    },
    {
      "entropy": 5.45492377281189,
      "epoch": 0.415552956899432,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004987570319953018,
      "loss": 5.303,
      "mean_token_accuracy": 0.1814391180872917,
      "num_tokens": 9161316.0,
      "step": 4975
    },
    {
      "entropy": 5.480196905136109,
      "epoch": 0.415970598062145,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004987539051693134,
      "loss": 5.2981,
      "mean_token_accuracy": 0.18222927004098893,
      "num_tokens": 9169535.0,
      "step": 4980
    },
    {
      "entropy": 5.461813735961914,
      "epoch": 0.416388239224858,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004987507744262436,
      "loss": 5.3266,
      "mean_token_accuracy": 0.17860324382781984,
      "num_tokens": 9178187.0,
      "step": 4985
    },
    {
      "entropy": 5.383362483978272,
      "epoch": 0.416805880387571,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004987476397661472,
      "loss": 5.2191,
      "mean_token_accuracy": 0.18512272834777832,
      "num_tokens": 9186342.0,
      "step": 4990
    },
    {
      "entropy": 5.465624713897705,
      "epoch": 0.41722352155028397,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004987445011890789,
      "loss": 5.3594,
      "mean_token_accuracy": 0.17655524462461472,
      "num_tokens": 9196470.0,
      "step": 4995
    },
    {
      "entropy": 5.40647554397583,
      "epoch": 0.417641162712997,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000498741358695094,
      "loss": 5.2688,
      "mean_token_accuracy": 0.18869243264198304,
      "num_tokens": 9206587.0,
      "step": 5000
    },
    {
      "entropy": 5.444352006912231,
      "epoch": 0.41805880387571,
      "grad_norm": 0.90625,
      "learning_rate": 0.000498738212284247,
      "loss": 5.258,
      "mean_token_accuracy": 0.18058222234249116,
      "num_tokens": 9216459.0,
      "step": 5005
    },
    {
      "entropy": 5.569244718551635,
      "epoch": 0.41847644503842296,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004987350619565934,
      "loss": 5.3704,
      "mean_token_accuracy": 0.18184609711170197,
      "num_tokens": 9227301.0,
      "step": 5010
    },
    {
      "entropy": 5.547540712356567,
      "epoch": 0.418894086201136,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004987319077121884,
      "loss": 5.285,
      "mean_token_accuracy": 0.17468804121017456,
      "num_tokens": 9236577.0,
      "step": 5015
    },
    {
      "entropy": 5.38317494392395,
      "epoch": 0.419311727363849,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004987287495510869,
      "loss": 5.2884,
      "mean_token_accuracy": 0.1898784816265106,
      "num_tokens": 9246451.0,
      "step": 5020
    },
    {
      "entropy": 5.401235675811767,
      "epoch": 0.41972936852656195,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004987255874733444,
      "loss": 5.1759,
      "mean_token_accuracy": 0.18578717559576036,
      "num_tokens": 9256108.0,
      "step": 5025
    },
    {
      "entropy": 5.381507396697998,
      "epoch": 0.420147009689275,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004987224214790161,
      "loss": 5.2368,
      "mean_token_accuracy": 0.18951483815908432,
      "num_tokens": 9265173.0,
      "step": 5030
    },
    {
      "entropy": 5.422999334335327,
      "epoch": 0.42056465085198796,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004987192515681575,
      "loss": 5.3285,
      "mean_token_accuracy": 0.18078091740608215,
      "num_tokens": 9273979.0,
      "step": 5035
    },
    {
      "entropy": 5.387002992630005,
      "epoch": 0.42098229201470094,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004987160777408243,
      "loss": 5.2754,
      "mean_token_accuracy": 0.18476943224668502,
      "num_tokens": 9281893.0,
      "step": 5040
    },
    {
      "entropy": 5.522434139251709,
      "epoch": 0.421399933177414,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004987128999970717,
      "loss": 5.3886,
      "mean_token_accuracy": 0.18378125131130219,
      "num_tokens": 9290799.0,
      "step": 5045
    },
    {
      "entropy": 5.5199041843414305,
      "epoch": 0.42181757434012696,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004987097183369555,
      "loss": 5.3528,
      "mean_token_accuracy": 0.1788174644112587,
      "num_tokens": 9300562.0,
      "step": 5050
    },
    {
      "entropy": 5.393524599075318,
      "epoch": 0.42223521550283993,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004987065327605315,
      "loss": 5.1192,
      "mean_token_accuracy": 0.19510368555784224,
      "num_tokens": 9308889.0,
      "step": 5055
    },
    {
      "entropy": 5.3221100807189945,
      "epoch": 0.42265285666555297,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004987033432678553,
      "loss": 5.2268,
      "mean_token_accuracy": 0.18440575003623963,
      "num_tokens": 9317957.0,
      "step": 5060
    },
    {
      "entropy": 5.53499174118042,
      "epoch": 0.42307049782826595,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004987001498589828,
      "loss": 5.4166,
      "mean_token_accuracy": 0.17679091393947602,
      "num_tokens": 9327849.0,
      "step": 5065
    },
    {
      "entropy": 5.475257110595703,
      "epoch": 0.4234881389909789,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00049869695253397,
      "loss": 5.2823,
      "mean_token_accuracy": 0.18098731338977814,
      "num_tokens": 9337165.0,
      "step": 5070
    },
    {
      "entropy": 5.372764301300049,
      "epoch": 0.42390578015369196,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004986937512928727,
      "loss": 5.1917,
      "mean_token_accuracy": 0.18649070113897323,
      "num_tokens": 9346817.0,
      "step": 5075
    },
    {
      "entropy": 5.548749589920044,
      "epoch": 0.42432342131640494,
      "grad_norm": 1.0,
      "learning_rate": 0.000498690546135747,
      "loss": 5.3055,
      "mean_token_accuracy": 0.18584760576486586,
      "num_tokens": 9356420.0,
      "step": 5080
    },
    {
      "entropy": 5.403382253646851,
      "epoch": 0.4247410624791179,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004986873370626491,
      "loss": 5.2443,
      "mean_token_accuracy": 0.18734410554170608,
      "num_tokens": 9365580.0,
      "step": 5085
    },
    {
      "entropy": 5.42599401473999,
      "epoch": 0.42515870364183095,
      "grad_norm": 0.96875,
      "learning_rate": 0.000498684124073635,
      "loss": 5.2845,
      "mean_token_accuracy": 0.17845439463853835,
      "num_tokens": 9375116.0,
      "step": 5090
    },
    {
      "entropy": 5.494064092636108,
      "epoch": 0.4255763448045439,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004986809071687612,
      "loss": 5.2568,
      "mean_token_accuracy": 0.18795115649700164,
      "num_tokens": 9384129.0,
      "step": 5095
    },
    {
      "entropy": 5.407192611694336,
      "epoch": 0.4259939859672569,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004986776863480838,
      "loss": 5.2182,
      "mean_token_accuracy": 0.18306127637624742,
      "num_tokens": 9392757.0,
      "step": 5100
    },
    {
      "entropy": 5.334506893157959,
      "epoch": 0.42641162712996994,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004986744616116593,
      "loss": 5.1896,
      "mean_token_accuracy": 0.1879316210746765,
      "num_tokens": 9401753.0,
      "step": 5105
    },
    {
      "entropy": 5.4469818592071535,
      "epoch": 0.4268292682926829,
      "grad_norm": 0.953125,
      "learning_rate": 0.000498671232959544,
      "loss": 5.3878,
      "mean_token_accuracy": 0.1787691369652748,
      "num_tokens": 9411400.0,
      "step": 5110
    },
    {
      "entropy": 5.494729232788086,
      "epoch": 0.42724690945539595,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004986680003917947,
      "loss": 5.2377,
      "mean_token_accuracy": 0.18792588412761688,
      "num_tokens": 9420206.0,
      "step": 5115
    },
    {
      "entropy": 5.431052398681641,
      "epoch": 0.42766455061810893,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004986647639084677,
      "loss": 5.2359,
      "mean_token_accuracy": 0.18809874057769777,
      "num_tokens": 9428418.0,
      "step": 5120
    },
    {
      "entropy": 5.46989803314209,
      "epoch": 0.4280821917808219,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00049866152350962,
      "loss": 5.2812,
      "mean_token_accuracy": 0.18487854152917862,
      "num_tokens": 9437624.0,
      "step": 5125
    },
    {
      "entropy": 5.396502113342285,
      "epoch": 0.42849983294353494,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004986582791953078,
      "loss": 5.187,
      "mean_token_accuracy": 0.1898106575012207,
      "num_tokens": 9446857.0,
      "step": 5130
    },
    {
      "entropy": 5.367677640914917,
      "epoch": 0.4289174741062479,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004986550309655884,
      "loss": 5.2919,
      "mean_token_accuracy": 0.18063514530658722,
      "num_tokens": 9456024.0,
      "step": 5135
    },
    {
      "entropy": 5.441341924667358,
      "epoch": 0.4293351152689609,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004986517788205184,
      "loss": 5.2083,
      "mean_token_accuracy": 0.19121423810720445,
      "num_tokens": 9464018.0,
      "step": 5140
    },
    {
      "entropy": 5.4211217880249025,
      "epoch": 0.42975275643167393,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004986485227601548,
      "loss": 5.3275,
      "mean_token_accuracy": 0.18502398878335952,
      "num_tokens": 9472430.0,
      "step": 5145
    },
    {
      "entropy": 5.420292901992798,
      "epoch": 0.4301703975943869,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004986452627845547,
      "loss": 5.2459,
      "mean_token_accuracy": 0.19181320518255235,
      "num_tokens": 9480592.0,
      "step": 5150
    },
    {
      "entropy": 5.365496301651001,
      "epoch": 0.4305880387570999,
      "grad_norm": 0.984375,
      "learning_rate": 0.000498641998893775,
      "loss": 5.1985,
      "mean_token_accuracy": 0.1923169270157814,
      "num_tokens": 9489162.0,
      "step": 5155
    },
    {
      "entropy": 5.442994213104248,
      "epoch": 0.4310056799198129,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004986387310878728,
      "loss": 5.2532,
      "mean_token_accuracy": 0.18618871420621871,
      "num_tokens": 9499210.0,
      "step": 5160
    },
    {
      "entropy": 5.447555732727051,
      "epoch": 0.4314233210825259,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004986354593669056,
      "loss": 5.2829,
      "mean_token_accuracy": 0.18904846757650376,
      "num_tokens": 9508062.0,
      "step": 5165
    },
    {
      "entropy": 5.485282611846924,
      "epoch": 0.4318409622452389,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004986321837309304,
      "loss": 5.2476,
      "mean_token_accuracy": 0.19065442383289338,
      "num_tokens": 9516756.0,
      "step": 5170
    },
    {
      "entropy": 5.371957015991211,
      "epoch": 0.4322586034079519,
      "grad_norm": 1.0,
      "learning_rate": 0.0004986289041800046,
      "loss": 5.2241,
      "mean_token_accuracy": 0.1923668712377548,
      "num_tokens": 9525305.0,
      "step": 5175
    },
    {
      "entropy": 5.36716742515564,
      "epoch": 0.4326762445706649,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004986256207141857,
      "loss": 5.2278,
      "mean_token_accuracy": 0.1915611892938614,
      "num_tokens": 9534310.0,
      "step": 5180
    },
    {
      "entropy": 5.533242177963257,
      "epoch": 0.43309388573337787,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004986223333335311,
      "loss": 5.3197,
      "mean_token_accuracy": 0.17931675910949707,
      "num_tokens": 9543648.0,
      "step": 5185
    },
    {
      "entropy": 5.386125230789185,
      "epoch": 0.4335115268960909,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004986190420380985,
      "loss": 5.2304,
      "mean_token_accuracy": 0.18825389593839645,
      "num_tokens": 9553081.0,
      "step": 5190
    },
    {
      "entropy": 5.473733282089233,
      "epoch": 0.4339291680588039,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004986157468279453,
      "loss": 5.2753,
      "mean_token_accuracy": 0.18827999234199524,
      "num_tokens": 9563184.0,
      "step": 5195
    },
    {
      "entropy": 5.438686418533325,
      "epoch": 0.43434680922151686,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004986124477031293,
      "loss": 5.2385,
      "mean_token_accuracy": 0.18429340422153473,
      "num_tokens": 9572550.0,
      "step": 5200
    },
    {
      "entropy": 5.358805131912232,
      "epoch": 0.4347644503842299,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004986091446637083,
      "loss": 5.2999,
      "mean_token_accuracy": 0.18634633123874664,
      "num_tokens": 9580900.0,
      "step": 5205
    },
    {
      "entropy": 5.4577717781066895,
      "epoch": 0.4351820915469429,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00049860583770974,
      "loss": 5.2677,
      "mean_token_accuracy": 0.19202205836772918,
      "num_tokens": 9589340.0,
      "step": 5210
    },
    {
      "entropy": 5.448734092712402,
      "epoch": 0.43559973270965585,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004986025268412823,
      "loss": 5.3874,
      "mean_token_accuracy": 0.18587642163038254,
      "num_tokens": 9599316.0,
      "step": 5215
    },
    {
      "entropy": 5.519431638717651,
      "epoch": 0.4360173738723689,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004985992120583933,
      "loss": 5.2782,
      "mean_token_accuracy": 0.1813632518053055,
      "num_tokens": 9609350.0,
      "step": 5220
    },
    {
      "entropy": 5.530301761627197,
      "epoch": 0.43643501503508186,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000498595893361131,
      "loss": 5.3343,
      "mean_token_accuracy": 0.18738499879837037,
      "num_tokens": 9618491.0,
      "step": 5225
    },
    {
      "entropy": 5.409563827514648,
      "epoch": 0.43685265619779484,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004985925707495533,
      "loss": 5.3936,
      "mean_token_accuracy": 0.17285268306732177,
      "num_tokens": 9629613.0,
      "step": 5230
    },
    {
      "entropy": 5.401585626602173,
      "epoch": 0.4372702973605079,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004985892442237186,
      "loss": 5.2665,
      "mean_token_accuracy": 0.19084574282169342,
      "num_tokens": 9638035.0,
      "step": 5235
    },
    {
      "entropy": 5.503404045104981,
      "epoch": 0.43768793852322085,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004985859137836851,
      "loss": 5.3366,
      "mean_token_accuracy": 0.1777690663933754,
      "num_tokens": 9647378.0,
      "step": 5240
    },
    {
      "entropy": 5.5066570281982425,
      "epoch": 0.43810557968593383,
      "grad_norm": 1.0,
      "learning_rate": 0.0004985825794295109,
      "loss": 5.248,
      "mean_token_accuracy": 0.18818107545375823,
      "num_tokens": 9655728.0,
      "step": 5245
    },
    {
      "entropy": 5.397564935684204,
      "epoch": 0.43852322084864687,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004985792411612546,
      "loss": 5.2838,
      "mean_token_accuracy": 0.18677612245082856,
      "num_tokens": 9665063.0,
      "step": 5250
    },
    {
      "entropy": 5.503706932067871,
      "epoch": 0.43894086201135984,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004985758989789746,
      "loss": 5.2784,
      "mean_token_accuracy": 0.184313502907753,
      "num_tokens": 9674406.0,
      "step": 5255
    },
    {
      "entropy": 5.423017120361328,
      "epoch": 0.4393585031740728,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004985725528827293,
      "loss": 5.2354,
      "mean_token_accuracy": 0.18182777911424636,
      "num_tokens": 9683303.0,
      "step": 5260
    },
    {
      "entropy": 5.409608983993531,
      "epoch": 0.43977614433678586,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004985692028725774,
      "loss": 5.2353,
      "mean_token_accuracy": 0.18337693810462952,
      "num_tokens": 9692938.0,
      "step": 5265
    },
    {
      "entropy": 5.490899848937988,
      "epoch": 0.44019378549949884,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004985658489485774,
      "loss": 5.299,
      "mean_token_accuracy": 0.18778119832277299,
      "num_tokens": 9702192.0,
      "step": 5270
    },
    {
      "entropy": 5.432302236557007,
      "epoch": 0.4406114266622118,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004985624911107882,
      "loss": 5.2747,
      "mean_token_accuracy": 0.17764863073825837,
      "num_tokens": 9711301.0,
      "step": 5275
    },
    {
      "entropy": 5.395459127426148,
      "epoch": 0.44102906782492485,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004985591293592685,
      "loss": 5.2777,
      "mean_token_accuracy": 0.17811281383037567,
      "num_tokens": 9720768.0,
      "step": 5280
    },
    {
      "entropy": 5.411895132064819,
      "epoch": 0.4414467089876378,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004985557636940771,
      "loss": 5.308,
      "mean_token_accuracy": 0.18872927725315095,
      "num_tokens": 9730007.0,
      "step": 5285
    },
    {
      "entropy": 5.490205764770508,
      "epoch": 0.4418643501503508,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000498552394115273,
      "loss": 5.3524,
      "mean_token_accuracy": 0.1783874064683914,
      "num_tokens": 9739546.0,
      "step": 5290
    },
    {
      "entropy": 5.4252622604370115,
      "epoch": 0.44228199131306384,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004985490206229151,
      "loss": 5.2352,
      "mean_token_accuracy": 0.1889718383550644,
      "num_tokens": 9749615.0,
      "step": 5295
    },
    {
      "entropy": 5.48985013961792,
      "epoch": 0.4426996324757768,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004985456432170626,
      "loss": 5.3304,
      "mean_token_accuracy": 0.17680660486221314,
      "num_tokens": 9759285.0,
      "step": 5300
    },
    {
      "entropy": 5.3831991195678714,
      "epoch": 0.4431172736384898,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004985422618977744,
      "loss": 5.2074,
      "mean_token_accuracy": 0.17996082603931426,
      "num_tokens": 9769396.0,
      "step": 5305
    },
    {
      "entropy": 5.430175733566284,
      "epoch": 0.44353491480120283,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00049853887666511,
      "loss": 5.3153,
      "mean_token_accuracy": 0.18187266886234282,
      "num_tokens": 9779632.0,
      "step": 5310
    },
    {
      "entropy": 5.499141836166382,
      "epoch": 0.4439525559639158,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004985354875191283,
      "loss": 5.2433,
      "mean_token_accuracy": 0.19233103543519975,
      "num_tokens": 9788090.0,
      "step": 5315
    },
    {
      "entropy": 5.282725429534912,
      "epoch": 0.4443701971266288,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004985320944598889,
      "loss": 5.1569,
      "mean_token_accuracy": 0.2003217563033104,
      "num_tokens": 9795844.0,
      "step": 5320
    },
    {
      "entropy": 5.3054807662963865,
      "epoch": 0.4447878382893418,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004985286974874512,
      "loss": 5.1588,
      "mean_token_accuracy": 0.1987578332424164,
      "num_tokens": 9805119.0,
      "step": 5325
    },
    {
      "entropy": 5.418432331085205,
      "epoch": 0.4452054794520548,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004985252966018745,
      "loss": 5.2804,
      "mean_token_accuracy": 0.18367889821529387,
      "num_tokens": 9813842.0,
      "step": 5330
    },
    {
      "entropy": 5.461290454864502,
      "epoch": 0.4456231206147678,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004985218918032184,
      "loss": 5.2801,
      "mean_token_accuracy": 0.17584045380353927,
      "num_tokens": 9822872.0,
      "step": 5335
    },
    {
      "entropy": 5.494472312927246,
      "epoch": 0.4460407617774808,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004985184830915426,
      "loss": 5.3933,
      "mean_token_accuracy": 0.17780580669641494,
      "num_tokens": 9832710.0,
      "step": 5340
    },
    {
      "entropy": 5.490469884872437,
      "epoch": 0.4464584029401938,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004985150704669067,
      "loss": 5.2868,
      "mean_token_accuracy": 0.18248433768749237,
      "num_tokens": 9841047.0,
      "step": 5345
    },
    {
      "entropy": 5.457961988449097,
      "epoch": 0.44687604410290677,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004985116539293704,
      "loss": 5.2577,
      "mean_token_accuracy": 0.18547517955303192,
      "num_tokens": 9850168.0,
      "step": 5350
    },
    {
      "entropy": 5.37015814781189,
      "epoch": 0.4472936852656198,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004985082334789936,
      "loss": 5.304,
      "mean_token_accuracy": 0.17832127213478088,
      "num_tokens": 9859550.0,
      "step": 5355
    },
    {
      "entropy": 5.544053411483764,
      "epoch": 0.4477113264283328,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004985048091158361,
      "loss": 5.2562,
      "mean_token_accuracy": 0.18138808608055115,
      "num_tokens": 9868317.0,
      "step": 5360
    },
    {
      "entropy": 5.5671820640563965,
      "epoch": 0.44812896759104576,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004985013808399578,
      "loss": 5.471,
      "mean_token_accuracy": 0.1696663334965706,
      "num_tokens": 9878421.0,
      "step": 5365
    },
    {
      "entropy": 5.433309841156006,
      "epoch": 0.4485466087537588,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004984979486514189,
      "loss": 5.2009,
      "mean_token_accuracy": 0.1938306972384453,
      "num_tokens": 9887213.0,
      "step": 5370
    },
    {
      "entropy": 5.374053573608398,
      "epoch": 0.44896424991647177,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004984945125502795,
      "loss": 5.3298,
      "mean_token_accuracy": 0.18225222676992417,
      "num_tokens": 9897274.0,
      "step": 5375
    },
    {
      "entropy": 5.384541845321655,
      "epoch": 0.44938189107918475,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004984910725365994,
      "loss": 5.2493,
      "mean_token_accuracy": 0.1819701835513115,
      "num_tokens": 9906765.0,
      "step": 5380
    },
    {
      "entropy": 5.526954174041748,
      "epoch": 0.4497995322418978,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004984876286104392,
      "loss": 5.2655,
      "mean_token_accuracy": 0.17805615514516832,
      "num_tokens": 9916149.0,
      "step": 5385
    },
    {
      "entropy": 5.45858244895935,
      "epoch": 0.45021717340461076,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004984841807718591,
      "loss": 5.2556,
      "mean_token_accuracy": 0.18223823308944703,
      "num_tokens": 9925657.0,
      "step": 5390
    },
    {
      "entropy": 5.381691980361938,
      "epoch": 0.45063481456732374,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004984807290209193,
      "loss": 5.2515,
      "mean_token_accuracy": 0.18012963235378265,
      "num_tokens": 9934740.0,
      "step": 5395
    },
    {
      "entropy": 5.555112695693969,
      "epoch": 0.45105245573003677,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004984772733576803,
      "loss": 5.3832,
      "mean_token_accuracy": 0.17855083495378493,
      "num_tokens": 9943516.0,
      "step": 5400
    },
    {
      "entropy": 5.464145469665527,
      "epoch": 0.45147009689274975,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004984738137822028,
      "loss": 5.3435,
      "mean_token_accuracy": 0.17752426713705063,
      "num_tokens": 9952998.0,
      "step": 5405
    },
    {
      "entropy": 5.383611345291138,
      "epoch": 0.45188773805546273,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000498470350294547,
      "loss": 5.1534,
      "mean_token_accuracy": 0.19169070869684218,
      "num_tokens": 9961698.0,
      "step": 5410
    },
    {
      "entropy": 5.42876706123352,
      "epoch": 0.45230537921817576,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004984668828947739,
      "loss": 5.2875,
      "mean_token_accuracy": 0.17814262211322784,
      "num_tokens": 9972015.0,
      "step": 5415
    },
    {
      "entropy": 5.519335079193115,
      "epoch": 0.45272302038088874,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004984634115829439,
      "loss": 5.3613,
      "mean_token_accuracy": 0.17557108104228974,
      "num_tokens": 9981834.0,
      "step": 5420
    },
    {
      "entropy": 5.377679967880249,
      "epoch": 0.4531406615436017,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000498459936359118,
      "loss": 5.2165,
      "mean_token_accuracy": 0.19437572807073594,
      "num_tokens": 9991113.0,
      "step": 5425
    },
    {
      "entropy": 5.4019910335540775,
      "epoch": 0.45355830270631475,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004984564572233568,
      "loss": 5.2071,
      "mean_token_accuracy": 0.18909807205200196,
      "num_tokens": 10000048.0,
      "step": 5430
    },
    {
      "entropy": 5.446583271026611,
      "epoch": 0.45397594386902773,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004984529741757216,
      "loss": 5.267,
      "mean_token_accuracy": 0.1838410437107086,
      "num_tokens": 10009389.0,
      "step": 5435
    },
    {
      "entropy": 5.39108624458313,
      "epoch": 0.4543935850317407,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004984494872162728,
      "loss": 5.1059,
      "mean_token_accuracy": 0.19814195334911347,
      "num_tokens": 10018342.0,
      "step": 5440
    },
    {
      "entropy": 5.462793588638306,
      "epoch": 0.45481122619445374,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000498445996345072,
      "loss": 5.2923,
      "mean_token_accuracy": 0.18794476687908174,
      "num_tokens": 10027219.0,
      "step": 5445
    },
    {
      "entropy": 5.362340688705444,
      "epoch": 0.4552288673571667,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004984425015621799,
      "loss": 5.1966,
      "mean_token_accuracy": 0.18359389752149582,
      "num_tokens": 10036635.0,
      "step": 5450
    },
    {
      "entropy": 5.345533943176269,
      "epoch": 0.4556465085198797,
      "grad_norm": 1.0625,
      "learning_rate": 0.000498439002867658,
      "loss": 5.2329,
      "mean_token_accuracy": 0.18499738425016404,
      "num_tokens": 10045655.0,
      "step": 5455
    },
    {
      "entropy": 5.5038693904876705,
      "epoch": 0.45606414968259273,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004984355002615673,
      "loss": 5.305,
      "mean_token_accuracy": 0.17853394895792007,
      "num_tokens": 10054366.0,
      "step": 5460
    },
    {
      "entropy": 5.4341929912567135,
      "epoch": 0.4564817908453057,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004984319937439693,
      "loss": 5.242,
      "mean_token_accuracy": 0.1883680686354637,
      "num_tokens": 10062548.0,
      "step": 5465
    },
    {
      "entropy": 5.384095144271851,
      "epoch": 0.4568994320080187,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004984284833149252,
      "loss": 5.2413,
      "mean_token_accuracy": 0.18972444236278535,
      "num_tokens": 10071333.0,
      "step": 5470
    },
    {
      "entropy": 5.321416711807251,
      "epoch": 0.4573170731707317,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004984249689744966,
      "loss": 5.134,
      "mean_token_accuracy": 0.19405237585306168,
      "num_tokens": 10079502.0,
      "step": 5475
    },
    {
      "entropy": 5.369308280944824,
      "epoch": 0.4577347143334447,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000498421450722745,
      "loss": 5.1726,
      "mean_token_accuracy": 0.19089630097150803,
      "num_tokens": 10088648.0,
      "step": 5480
    },
    {
      "entropy": 5.387316036224365,
      "epoch": 0.4581523554961577,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000498417928559732,
      "loss": 5.1778,
      "mean_token_accuracy": 0.1940348908305168,
      "num_tokens": 10097742.0,
      "step": 5485
    },
    {
      "entropy": 5.469959020614624,
      "epoch": 0.4585699966588707,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004984144024855191,
      "loss": 5.297,
      "mean_token_accuracy": 0.18087629228830338,
      "num_tokens": 10107228.0,
      "step": 5490
    },
    {
      "entropy": 5.444091796875,
      "epoch": 0.4589876378215837,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004984108725001683,
      "loss": 5.197,
      "mean_token_accuracy": 0.1878842830657959,
      "num_tokens": 10115488.0,
      "step": 5495
    },
    {
      "entropy": 5.328802871704101,
      "epoch": 0.45940527898429667,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004984073386037413,
      "loss": 5.1323,
      "mean_token_accuracy": 0.19642899483442305,
      "num_tokens": 10124235.0,
      "step": 5500
    },
    {
      "entropy": 5.3084447383880615,
      "epoch": 0.4598229201470097,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004984038007962997,
      "loss": 5.227,
      "mean_token_accuracy": 0.19213753193616867,
      "num_tokens": 10133208.0,
      "step": 5505
    },
    {
      "entropy": 5.378926944732666,
      "epoch": 0.4602405613097227,
      "grad_norm": 0.875,
      "learning_rate": 0.0004984002590779059,
      "loss": 5.1547,
      "mean_token_accuracy": 0.19189033210277556,
      "num_tokens": 10142754.0,
      "step": 5510
    },
    {
      "entropy": 5.456240606307984,
      "epoch": 0.46065820247243566,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004983967134486215,
      "loss": 5.2752,
      "mean_token_accuracy": 0.1869430348277092,
      "num_tokens": 10152114.0,
      "step": 5515
    },
    {
      "entropy": 5.397875595092773,
      "epoch": 0.4610758436351487,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004983931639085088,
      "loss": 5.1737,
      "mean_token_accuracy": 0.19056999534368516,
      "num_tokens": 10161562.0,
      "step": 5520
    },
    {
      "entropy": 5.456809949874878,
      "epoch": 0.4614934847978617,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004983896104576298,
      "loss": 5.2883,
      "mean_token_accuracy": 0.18473776280879975,
      "num_tokens": 10170461.0,
      "step": 5525
    },
    {
      "entropy": 5.370413875579834,
      "epoch": 0.46191112596057465,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000498386053096047,
      "loss": 5.2786,
      "mean_token_accuracy": 0.18490847647190095,
      "num_tokens": 10179243.0,
      "step": 5530
    },
    {
      "entropy": 5.405595159530639,
      "epoch": 0.4623287671232877,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004983824918238221,
      "loss": 5.2303,
      "mean_token_accuracy": 0.18954685777425767,
      "num_tokens": 10187935.0,
      "step": 5535
    },
    {
      "entropy": 5.433976554870606,
      "epoch": 0.46274640828600067,
      "grad_norm": 1.0,
      "learning_rate": 0.000498378926641018,
      "loss": 5.2983,
      "mean_token_accuracy": 0.18473003357648848,
      "num_tokens": 10196185.0,
      "step": 5540
    },
    {
      "entropy": 5.425156497955323,
      "epoch": 0.46316404944871364,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004983753575476969,
      "loss": 5.2798,
      "mean_token_accuracy": 0.18167456090450287,
      "num_tokens": 10205373.0,
      "step": 5545
    },
    {
      "entropy": 5.428696823120117,
      "epoch": 0.4635816906114267,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004983717845439212,
      "loss": 5.2625,
      "mean_token_accuracy": 0.19419895857572556,
      "num_tokens": 10214464.0,
      "step": 5550
    },
    {
      "entropy": 5.4097083568572994,
      "epoch": 0.46399933177413966,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004983682076297536,
      "loss": 5.3373,
      "mean_token_accuracy": 0.18244415819644927,
      "num_tokens": 10224462.0,
      "step": 5555
    },
    {
      "entropy": 5.489752006530762,
      "epoch": 0.46441697293685263,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004983646268052567,
      "loss": 5.244,
      "mean_token_accuracy": 0.18829308748245238,
      "num_tokens": 10233123.0,
      "step": 5560
    },
    {
      "entropy": 5.326920795440674,
      "epoch": 0.46483461409956567,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004983610420704931,
      "loss": 5.0926,
      "mean_token_accuracy": 0.19586901366710663,
      "num_tokens": 10241613.0,
      "step": 5565
    },
    {
      "entropy": 5.451515865325928,
      "epoch": 0.46525225526227865,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004983574534255256,
      "loss": 5.4433,
      "mean_token_accuracy": 0.17371949404478074,
      "num_tokens": 10251809.0,
      "step": 5570
    },
    {
      "entropy": 5.374123048782349,
      "epoch": 0.4656698964249916,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000498353860870417,
      "loss": 5.1295,
      "mean_token_accuracy": 0.19029677659273148,
      "num_tokens": 10260198.0,
      "step": 5575
    },
    {
      "entropy": 5.4094689846038815,
      "epoch": 0.46608753758770466,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004983502644052303,
      "loss": 5.1963,
      "mean_token_accuracy": 0.18829075396060943,
      "num_tokens": 10268906.0,
      "step": 5580
    },
    {
      "entropy": 5.375689077377319,
      "epoch": 0.46650517875041764,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004983466640300282,
      "loss": 5.1299,
      "mean_token_accuracy": 0.18463146388530732,
      "num_tokens": 10278526.0,
      "step": 5585
    },
    {
      "entropy": 5.38140640258789,
      "epoch": 0.4669228199131306,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004983430597448741,
      "loss": 5.2541,
      "mean_token_accuracy": 0.1882788583636284,
      "num_tokens": 10287428.0,
      "step": 5590
    },
    {
      "entropy": 5.54051947593689,
      "epoch": 0.46734046107584365,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004983394515498308,
      "loss": 5.2252,
      "mean_token_accuracy": 0.1844131425023079,
      "num_tokens": 10297136.0,
      "step": 5595
    },
    {
      "entropy": 5.464434385299683,
      "epoch": 0.46775810223855663,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004983358394449616,
      "loss": 5.2883,
      "mean_token_accuracy": 0.18291351646184922,
      "num_tokens": 10305678.0,
      "step": 5600
    },
    {
      "entropy": 5.303349256515503,
      "epoch": 0.4681757434012696,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004983322234303297,
      "loss": 5.2098,
      "mean_token_accuracy": 0.19036900103092194,
      "num_tokens": 10314690.0,
      "step": 5605
    },
    {
      "entropy": 5.375848722457886,
      "epoch": 0.46859338456398264,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004983286035059984,
      "loss": 5.1722,
      "mean_token_accuracy": 0.18854036033153534,
      "num_tokens": 10323521.0,
      "step": 5610
    },
    {
      "entropy": 5.408108854293824,
      "epoch": 0.4690110257266956,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004983249796720311,
      "loss": 5.2039,
      "mean_token_accuracy": 0.19393740147352218,
      "num_tokens": 10333077.0,
      "step": 5615
    },
    {
      "entropy": 5.441128396987915,
      "epoch": 0.4694286668894086,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004983213519284912,
      "loss": 5.2376,
      "mean_token_accuracy": 0.18737054914236068,
      "num_tokens": 10342194.0,
      "step": 5620
    },
    {
      "entropy": 5.33244047164917,
      "epoch": 0.46984630805212163,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004983177202754421,
      "loss": 5.0865,
      "mean_token_accuracy": 0.19418132305145264,
      "num_tokens": 10351009.0,
      "step": 5625
    },
    {
      "entropy": 5.406405019760132,
      "epoch": 0.4702639492148346,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004983140847129476,
      "loss": 5.2893,
      "mean_token_accuracy": 0.18420304358005524,
      "num_tokens": 10360872.0,
      "step": 5630
    },
    {
      "entropy": 5.464356231689453,
      "epoch": 0.4706815903775476,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004983104452410713,
      "loss": 5.2015,
      "mean_token_accuracy": 0.1856534168124199,
      "num_tokens": 10370159.0,
      "step": 5635
    },
    {
      "entropy": 5.38580813407898,
      "epoch": 0.4710992315402606,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004983068018598768,
      "loss": 5.1918,
      "mean_token_accuracy": 0.1921562895178795,
      "num_tokens": 10379219.0,
      "step": 5640
    },
    {
      "entropy": 5.388443422317505,
      "epoch": 0.4715168727029736,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000498303154569428,
      "loss": 5.2294,
      "mean_token_accuracy": 0.185270856320858,
      "num_tokens": 10389425.0,
      "step": 5645
    },
    {
      "entropy": 5.44331464767456,
      "epoch": 0.4719345138656866,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004982995033697887,
      "loss": 5.2486,
      "mean_token_accuracy": 0.18700142055749894,
      "num_tokens": 10398023.0,
      "step": 5650
    },
    {
      "entropy": 5.391577625274659,
      "epoch": 0.4723521550283996,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004982958482610226,
      "loss": 5.1826,
      "mean_token_accuracy": 0.19294381588697435,
      "num_tokens": 10406578.0,
      "step": 5655
    },
    {
      "entropy": 5.349926137924195,
      "epoch": 0.4727697961911126,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000498292189243194,
      "loss": 5.1882,
      "mean_token_accuracy": 0.18868072479963302,
      "num_tokens": 10416448.0,
      "step": 5660
    },
    {
      "entropy": 5.442293453216553,
      "epoch": 0.47318743735382557,
      "grad_norm": 1.015625,
      "learning_rate": 0.000498288526316367,
      "loss": 5.2091,
      "mean_token_accuracy": 0.1856846258044243,
      "num_tokens": 10425414.0,
      "step": 5665
    },
    {
      "entropy": 5.398424291610718,
      "epoch": 0.4736050785165386,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004982848594806054,
      "loss": 5.1995,
      "mean_token_accuracy": 0.20190430134534837,
      "num_tokens": 10433900.0,
      "step": 5670
    },
    {
      "entropy": 5.450738430023193,
      "epoch": 0.4740227196792516,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004982811887359736,
      "loss": 5.316,
      "mean_token_accuracy": 0.18356639593839646,
      "num_tokens": 10443077.0,
      "step": 5675
    },
    {
      "entropy": 5.436344718933105,
      "epoch": 0.47444036084196456,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004982775140825359,
      "loss": 5.2399,
      "mean_token_accuracy": 0.1881748616695404,
      "num_tokens": 10452791.0,
      "step": 5680
    },
    {
      "entropy": 5.510271692276001,
      "epoch": 0.4748580020046776,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004982738355203565,
      "loss": 5.4083,
      "mean_token_accuracy": 0.18155159801244736,
      "num_tokens": 10462087.0,
      "step": 5685
    },
    {
      "entropy": 5.529263114929199,
      "epoch": 0.47527564316739057,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004982701530494998,
      "loss": 5.393,
      "mean_token_accuracy": 0.17575380355119705,
      "num_tokens": 10472576.0,
      "step": 5690
    },
    {
      "entropy": 5.376764917373658,
      "epoch": 0.47569328433010355,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004982664666700304,
      "loss": 5.1241,
      "mean_token_accuracy": 0.19784934520721437,
      "num_tokens": 10481176.0,
      "step": 5695
    },
    {
      "entropy": 5.410578441619873,
      "epoch": 0.4761109254928166,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004982627763820128,
      "loss": 5.2208,
      "mean_token_accuracy": 0.18437573313713074,
      "num_tokens": 10490558.0,
      "step": 5700
    },
    {
      "entropy": 5.428354263305664,
      "epoch": 0.47652856665552956,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004982590821855115,
      "loss": 5.2917,
      "mean_token_accuracy": 0.18567686825990676,
      "num_tokens": 10499543.0,
      "step": 5705
    },
    {
      "entropy": 5.365946197509766,
      "epoch": 0.47694620781824254,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004982553840805912,
      "loss": 5.1689,
      "mean_token_accuracy": 0.187653249502182,
      "num_tokens": 10508162.0,
      "step": 5710
    },
    {
      "entropy": 5.385851955413818,
      "epoch": 0.4773638489809556,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004982516820673167,
      "loss": 5.0951,
      "mean_token_accuracy": 0.1896537199616432,
      "num_tokens": 10516837.0,
      "step": 5715
    },
    {
      "entropy": 5.344322633743286,
      "epoch": 0.47778149014366855,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004982479761457528,
      "loss": 5.2344,
      "mean_token_accuracy": 0.19141795337200165,
      "num_tokens": 10526208.0,
      "step": 5720
    },
    {
      "entropy": 5.443219184875488,
      "epoch": 0.47819913130638153,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004982442663159643,
      "loss": 5.2732,
      "mean_token_accuracy": 0.18132958561182022,
      "num_tokens": 10536002.0,
      "step": 5725
    },
    {
      "entropy": 5.45643424987793,
      "epoch": 0.47861677246909456,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004982405525780162,
      "loss": 5.2585,
      "mean_token_accuracy": 0.18747614771127702,
      "num_tokens": 10545861.0,
      "step": 5730
    },
    {
      "entropy": 5.387361335754394,
      "epoch": 0.47903441363180754,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004982368349319736,
      "loss": 5.1741,
      "mean_token_accuracy": 0.18652687221765518,
      "num_tokens": 10555357.0,
      "step": 5735
    },
    {
      "entropy": 5.366787385940552,
      "epoch": 0.4794520547945205,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004982331133779014,
      "loss": 5.2411,
      "mean_token_accuracy": 0.18766519278287888,
      "num_tokens": 10564744.0,
      "step": 5740
    },
    {
      "entropy": 5.381127405166626,
      "epoch": 0.47986969595723356,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004982293879158649,
      "loss": 5.2171,
      "mean_token_accuracy": 0.18826033771038056,
      "num_tokens": 10573736.0,
      "step": 5745
    },
    {
      "entropy": 5.3778234958648685,
      "epoch": 0.48028733711994653,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004982256585459293,
      "loss": 5.2744,
      "mean_token_accuracy": 0.18442912846803666,
      "num_tokens": 10583238.0,
      "step": 5750
    },
    {
      "entropy": 5.490876483917236,
      "epoch": 0.4807049782826595,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004982219252681597,
      "loss": 5.3089,
      "mean_token_accuracy": 0.1812056615948677,
      "num_tokens": 10592316.0,
      "step": 5755
    },
    {
      "entropy": 5.449782037734986,
      "epoch": 0.48112261944537255,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004982181880826218,
      "loss": 5.3241,
      "mean_token_accuracy": 0.18929636478424072,
      "num_tokens": 10601418.0,
      "step": 5760
    },
    {
      "entropy": 5.360013103485107,
      "epoch": 0.4815402606080855,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004982144469893807,
      "loss": 5.2132,
      "mean_token_accuracy": 0.1917594075202942,
      "num_tokens": 10610513.0,
      "step": 5765
    },
    {
      "entropy": 5.39288330078125,
      "epoch": 0.4819579017707985,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000498210701988502,
      "loss": 5.1488,
      "mean_token_accuracy": 0.1957583472132683,
      "num_tokens": 10619529.0,
      "step": 5770
    },
    {
      "entropy": 5.497353458404541,
      "epoch": 0.48237554293351154,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004982069530800514,
      "loss": 5.2473,
      "mean_token_accuracy": 0.19186618626117707,
      "num_tokens": 10628263.0,
      "step": 5775
    },
    {
      "entropy": 5.400147390365601,
      "epoch": 0.4827931840962245,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004982032002640943,
      "loss": 5.2645,
      "mean_token_accuracy": 0.19018780589103698,
      "num_tokens": 10637605.0,
      "step": 5780
    },
    {
      "entropy": 5.406908416748047,
      "epoch": 0.48321082525893755,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004981994435406965,
      "loss": 5.3261,
      "mean_token_accuracy": 0.18572087585926056,
      "num_tokens": 10647113.0,
      "step": 5785
    },
    {
      "entropy": 5.36240816116333,
      "epoch": 0.4836284664216505,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004981956829099239,
      "loss": 5.0913,
      "mean_token_accuracy": 0.19700256437063218,
      "num_tokens": 10655464.0,
      "step": 5790
    },
    {
      "entropy": 5.4342457294464115,
      "epoch": 0.4840461075843635,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004981919183718422,
      "loss": 5.203,
      "mean_token_accuracy": 0.19033023566007615,
      "num_tokens": 10664689.0,
      "step": 5795
    },
    {
      "entropy": 5.322588348388672,
      "epoch": 0.48446374874707654,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004981881499265173,
      "loss": 5.1117,
      "mean_token_accuracy": 0.18969281017780304,
      "num_tokens": 10673440.0,
      "step": 5800
    },
    {
      "entropy": 5.360542631149292,
      "epoch": 0.4848813899097895,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004981843775740151,
      "loss": 5.239,
      "mean_token_accuracy": 0.19217285960912706,
      "num_tokens": 10683075.0,
      "step": 5805
    },
    {
      "entropy": 5.410954332351684,
      "epoch": 0.4852990310725025,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004981806013144017,
      "loss": 5.2021,
      "mean_token_accuracy": 0.19118454456329345,
      "num_tokens": 10692663.0,
      "step": 5810
    },
    {
      "entropy": 5.36053581237793,
      "epoch": 0.48571667223521553,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004981768211477434,
      "loss": 5.1034,
      "mean_token_accuracy": 0.1941935122013092,
      "num_tokens": 10702068.0,
      "step": 5815
    },
    {
      "entropy": 5.362188148498535,
      "epoch": 0.4861343133979285,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000498173037074106,
      "loss": 5.2459,
      "mean_token_accuracy": 0.1780676066875458,
      "num_tokens": 10711230.0,
      "step": 5820
    },
    {
      "entropy": 5.336467313766479,
      "epoch": 0.4865519545606415,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004981692490935561,
      "loss": 5.2249,
      "mean_token_accuracy": 0.1963807985186577,
      "num_tokens": 10719651.0,
      "step": 5825
    },
    {
      "entropy": 5.384497213363647,
      "epoch": 0.4869695957233545,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004981654572061597,
      "loss": 5.2305,
      "mean_token_accuracy": 0.18997357040643692,
      "num_tokens": 10729342.0,
      "step": 5830
    },
    {
      "entropy": 5.449845170974731,
      "epoch": 0.4873872368860675,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004981616614119836,
      "loss": 5.2242,
      "mean_token_accuracy": 0.1907389283180237,
      "num_tokens": 10738617.0,
      "step": 5835
    },
    {
      "entropy": 5.4481383800506595,
      "epoch": 0.4878048780487805,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004981578617110938,
      "loss": 5.2658,
      "mean_token_accuracy": 0.19168426990509033,
      "num_tokens": 10746881.0,
      "step": 5840
    },
    {
      "entropy": 5.360753774642944,
      "epoch": 0.4882225192114935,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004981540581035571,
      "loss": 5.1168,
      "mean_token_accuracy": 0.19619991928339003,
      "num_tokens": 10755848.0,
      "step": 5845
    },
    {
      "entropy": 5.291107845306397,
      "epoch": 0.4886401603742065,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00049815025058944,
      "loss": 5.1341,
      "mean_token_accuracy": 0.189218907058239,
      "num_tokens": 10765519.0,
      "step": 5850
    },
    {
      "entropy": 5.321394586563111,
      "epoch": 0.48905780153691947,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004981464391688092,
      "loss": 5.2234,
      "mean_token_accuracy": 0.18370419144630432,
      "num_tokens": 10776408.0,
      "step": 5855
    },
    {
      "entropy": 5.384969902038574,
      "epoch": 0.4894754426996325,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004981426238417312,
      "loss": 5.2395,
      "mean_token_accuracy": 0.1924698382616043,
      "num_tokens": 10785499.0,
      "step": 5860
    },
    {
      "entropy": 5.3669140338897705,
      "epoch": 0.4898930838623455,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004981388046082731,
      "loss": 5.19,
      "mean_token_accuracy": 0.18840529769659042,
      "num_tokens": 10794736.0,
      "step": 5865
    },
    {
      "entropy": 5.38918628692627,
      "epoch": 0.49031072502505846,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004981349814685016,
      "loss": 5.266,
      "mean_token_accuracy": 0.1866907387971878,
      "num_tokens": 10804125.0,
      "step": 5870
    },
    {
      "entropy": 5.437146377563477,
      "epoch": 0.4907283661877715,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004981311544224838,
      "loss": 5.3084,
      "mean_token_accuracy": 0.18804036527872087,
      "num_tokens": 10813465.0,
      "step": 5875
    },
    {
      "entropy": 5.379113292694091,
      "epoch": 0.49114600735048447,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004981273234702864,
      "loss": 5.154,
      "mean_token_accuracy": 0.19134931415319442,
      "num_tokens": 10821550.0,
      "step": 5880
    },
    {
      "entropy": 5.350656270980835,
      "epoch": 0.49156364851319745,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004981234886119767,
      "loss": 5.1702,
      "mean_token_accuracy": 0.19213555455207826,
      "num_tokens": 10830683.0,
      "step": 5885
    },
    {
      "entropy": 5.382988452911377,
      "epoch": 0.4919812896759105,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004981196498476217,
      "loss": 5.187,
      "mean_token_accuracy": 0.1951914280653,
      "num_tokens": 10839492.0,
      "step": 5890
    },
    {
      "entropy": 5.3643683910369875,
      "epoch": 0.49239893083862346,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004981158071772887,
      "loss": 5.1076,
      "mean_token_accuracy": 0.19777220189571382,
      "num_tokens": 10847998.0,
      "step": 5895
    },
    {
      "entropy": 5.257833051681518,
      "epoch": 0.49281657200133644,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004981119606010449,
      "loss": 5.1145,
      "mean_token_accuracy": 0.19643002152442932,
      "num_tokens": 10856438.0,
      "step": 5900
    },
    {
      "entropy": 5.273256731033325,
      "epoch": 0.4932342131640495,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004981081101189577,
      "loss": 5.1159,
      "mean_token_accuracy": 0.19061104953289032,
      "num_tokens": 10865936.0,
      "step": 5905
    },
    {
      "entropy": 5.387513160705566,
      "epoch": 0.49365185432676245,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004981042557310945,
      "loss": 5.1318,
      "mean_token_accuracy": 0.19577510058879852,
      "num_tokens": 10874621.0,
      "step": 5910
    },
    {
      "entropy": 5.327988243103027,
      "epoch": 0.49406949548947543,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004981003974375227,
      "loss": 5.0968,
      "mean_token_accuracy": 0.19196219891309738,
      "num_tokens": 10883516.0,
      "step": 5915
    },
    {
      "entropy": 5.351433134078979,
      "epoch": 0.49448713665218846,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004980965352383099,
      "loss": 5.2275,
      "mean_token_accuracy": 0.18708742409944534,
      "num_tokens": 10892580.0,
      "step": 5920
    },
    {
      "entropy": 5.392227363586426,
      "epoch": 0.49490477781490144,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004980926691335238,
      "loss": 5.1572,
      "mean_token_accuracy": 0.19579843133687974,
      "num_tokens": 10901227.0,
      "step": 5925
    },
    {
      "entropy": 5.398783254623413,
      "epoch": 0.4953224189776144,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004980887991232318,
      "loss": 5.2183,
      "mean_token_accuracy": 0.1858757793903351,
      "num_tokens": 10909454.0,
      "step": 5930
    },
    {
      "entropy": 5.491789674758911,
      "epoch": 0.49574006014032745,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000498084925207502,
      "loss": 5.2851,
      "mean_token_accuracy": 0.18898377120494841,
      "num_tokens": 10919792.0,
      "step": 5935
    },
    {
      "entropy": 5.382387781143189,
      "epoch": 0.49615770130304043,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000498081047386402,
      "loss": 5.3157,
      "mean_token_accuracy": 0.18735468685626983,
      "num_tokens": 10930523.0,
      "step": 5940
    },
    {
      "entropy": 5.325424766540527,
      "epoch": 0.4965753424657534,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004980771656599998,
      "loss": 5.1858,
      "mean_token_accuracy": 0.19194009751081467,
      "num_tokens": 10940075.0,
      "step": 5945
    },
    {
      "entropy": 5.379294681549072,
      "epoch": 0.49699298362846644,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004980732800283633,
      "loss": 5.1993,
      "mean_token_accuracy": 0.18581117391586305,
      "num_tokens": 10949059.0,
      "step": 5950
    },
    {
      "entropy": 5.3666130065917965,
      "epoch": 0.4974106247911794,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004980693904915604,
      "loss": 5.171,
      "mean_token_accuracy": 0.19361854642629622,
      "num_tokens": 10958313.0,
      "step": 5955
    },
    {
      "entropy": 5.364543342590332,
      "epoch": 0.4978282659538924,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004980654970496595,
      "loss": 5.2149,
      "mean_token_accuracy": 0.18634458631277084,
      "num_tokens": 10967666.0,
      "step": 5960
    },
    {
      "entropy": 5.452500343322754,
      "epoch": 0.49824590711660544,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004980615997027284,
      "loss": 5.2619,
      "mean_token_accuracy": 0.1876556783914566,
      "num_tokens": 10976519.0,
      "step": 5965
    },
    {
      "entropy": 5.464095640182495,
      "epoch": 0.4986635482793184,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004980576984508356,
      "loss": 5.3474,
      "mean_token_accuracy": 0.17934405356645583,
      "num_tokens": 10987138.0,
      "step": 5970
    },
    {
      "entropy": 5.395865774154663,
      "epoch": 0.4990811894420314,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004980537932940492,
      "loss": 5.2233,
      "mean_token_accuracy": 0.1922852098941803,
      "num_tokens": 10996517.0,
      "step": 5975
    },
    {
      "entropy": 5.355505847930909,
      "epoch": 0.4994988306047444,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004980498842324377,
      "loss": 5.104,
      "mean_token_accuracy": 0.18747983425855635,
      "num_tokens": 11006312.0,
      "step": 5980
    },
    {
      "entropy": 5.37230372428894,
      "epoch": 0.4999164717674574,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004980459712660694,
      "loss": 5.2249,
      "mean_token_accuracy": 0.19582620710134507,
      "num_tokens": 11015535.0,
      "step": 5985
    },
    {
      "entropy": 5.384573888778687,
      "epoch": 0.5003341129301704,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004980420543950131,
      "loss": 5.1961,
      "mean_token_accuracy": 0.1938820958137512,
      "num_tokens": 11024887.0,
      "step": 5990
    },
    {
      "entropy": 5.296752977371216,
      "epoch": 0.5007517540928834,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004980381336193371,
      "loss": 5.0895,
      "mean_token_accuracy": 0.19957529455423356,
      "num_tokens": 11033147.0,
      "step": 5995
    },
    {
      "entropy": 5.320821905136109,
      "epoch": 0.5011693952555963,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004980342089391099,
      "loss": 5.2344,
      "mean_token_accuracy": 0.19215354472398757,
      "num_tokens": 11042726.0,
      "step": 6000
    },
    {
      "epoch": 0.5011693952555963,
      "eval_entropy": 5.088146142499019,
      "eval_loss": 5.181977272033691,
      "eval_mean_token_accuracy": 0.2012889325355188,
      "eval_num_tokens": 11042726.0,
      "eval_runtime": 20.9309,
      "eval_samples_per_second": 1804.175,
      "eval_steps_per_second": 225.552,
      "step": 6000
    },
    {
      "entropy": 5.375793361663819,
      "epoch": 0.5015870364183094,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004980302803544006,
      "loss": 5.2113,
      "mean_token_accuracy": 0.19045818597078323,
      "num_tokens": 11052181.0,
      "step": 6005
    },
    {
      "entropy": 5.4751685619354244,
      "epoch": 0.5020046775810224,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004980263478652778,
      "loss": 5.2618,
      "mean_token_accuracy": 0.18598974049091338,
      "num_tokens": 11061586.0,
      "step": 6010
    },
    {
      "entropy": 5.483823633193969,
      "epoch": 0.5024223187437353,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004980224114718102,
      "loss": 5.2643,
      "mean_token_accuracy": 0.1996988445520401,
      "num_tokens": 11070914.0,
      "step": 6015
    },
    {
      "entropy": 5.301267433166504,
      "epoch": 0.5028399599064484,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004980184711740669,
      "loss": 5.0983,
      "mean_token_accuracy": 0.19896316528320312,
      "num_tokens": 11079836.0,
      "step": 6020
    },
    {
      "entropy": 5.283141851425171,
      "epoch": 0.5032576010691614,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004980145269721168,
      "loss": 5.1185,
      "mean_token_accuracy": 0.1967083752155304,
      "num_tokens": 11088320.0,
      "step": 6025
    },
    {
      "entropy": 5.371067380905151,
      "epoch": 0.5036752422318743,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000498010578866029,
      "loss": 5.2872,
      "mean_token_accuracy": 0.1866062253713608,
      "num_tokens": 11097803.0,
      "step": 6030
    },
    {
      "entropy": 5.482487487792969,
      "epoch": 0.5040928833945874,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004980066268558726,
      "loss": 5.3074,
      "mean_token_accuracy": 0.18533538579940795,
      "num_tokens": 11107218.0,
      "step": 6035
    },
    {
      "entropy": 5.4608629703521725,
      "epoch": 0.5045105245573004,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004980026709417167,
      "loss": 5.2125,
      "mean_token_accuracy": 0.18697528839111327,
      "num_tokens": 11115599.0,
      "step": 6040
    },
    {
      "entropy": 5.285761404037475,
      "epoch": 0.5049281657200133,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004979987111236306,
      "loss": 5.0801,
      "mean_token_accuracy": 0.1982789993286133,
      "num_tokens": 11124312.0,
      "step": 6045
    },
    {
      "entropy": 5.291109561920166,
      "epoch": 0.5053458068827263,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004979947474016837,
      "loss": 5.2056,
      "mean_token_accuracy": 0.18848671168088912,
      "num_tokens": 11133117.0,
      "step": 6050
    },
    {
      "entropy": 5.505840301513672,
      "epoch": 0.5057634480454394,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004979907797759453,
      "loss": 5.2629,
      "mean_token_accuracy": 0.190187768638134,
      "num_tokens": 11141972.0,
      "step": 6055
    },
    {
      "entropy": 5.509038639068604,
      "epoch": 0.5061810892081523,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004979868082464849,
      "loss": 5.2648,
      "mean_token_accuracy": 0.1901376098394394,
      "num_tokens": 11151416.0,
      "step": 6060
    },
    {
      "entropy": 5.367153549194336,
      "epoch": 0.5065987303708653,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004979828328133719,
      "loss": 5.2277,
      "mean_token_accuracy": 0.18653842359781264,
      "num_tokens": 11160727.0,
      "step": 6065
    },
    {
      "entropy": 5.330143785476684,
      "epoch": 0.5070163715335784,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004979788534766762,
      "loss": 5.1692,
      "mean_token_accuracy": 0.1855221450328827,
      "num_tokens": 11171350.0,
      "step": 6070
    },
    {
      "entropy": 5.403621006011963,
      "epoch": 0.5074340126962913,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004979748702364672,
      "loss": 5.1612,
      "mean_token_accuracy": 0.188106532394886,
      "num_tokens": 11180233.0,
      "step": 6075
    },
    {
      "entropy": 5.382835340499878,
      "epoch": 0.5078516538590043,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004979708830928147,
      "loss": 5.1456,
      "mean_token_accuracy": 0.1849919930100441,
      "num_tokens": 11189269.0,
      "step": 6080
    },
    {
      "entropy": 5.375163269042969,
      "epoch": 0.5082692950217174,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004979668920457884,
      "loss": 5.2181,
      "mean_token_accuracy": 0.189941842854023,
      "num_tokens": 11199072.0,
      "step": 6085
    },
    {
      "entropy": 5.366530513763427,
      "epoch": 0.5086869361844303,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004979628970954584,
      "loss": 5.1296,
      "mean_token_accuracy": 0.19691332876682283,
      "num_tokens": 11207164.0,
      "step": 6090
    },
    {
      "entropy": 5.305984735488892,
      "epoch": 0.5091045773471433,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004979588982418944,
      "loss": 5.1687,
      "mean_token_accuracy": 0.18258725553750993,
      "num_tokens": 11216557.0,
      "step": 6095
    },
    {
      "entropy": 5.48675651550293,
      "epoch": 0.5095222185098564,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004979548954851666,
      "loss": 5.3105,
      "mean_token_accuracy": 0.18963435888290406,
      "num_tokens": 11227277.0,
      "step": 6100
    },
    {
      "entropy": 5.393910264968872,
      "epoch": 0.5099398596725693,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000497950888825345,
      "loss": 5.1272,
      "mean_token_accuracy": 0.1969315379858017,
      "num_tokens": 11236226.0,
      "step": 6105
    },
    {
      "entropy": 5.359139680862427,
      "epoch": 0.5103575008352823,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004979468782624996,
      "loss": 5.2195,
      "mean_token_accuracy": 0.19035741090774536,
      "num_tokens": 11245246.0,
      "step": 6110
    },
    {
      "entropy": 5.410820627212525,
      "epoch": 0.5107751419979953,
      "grad_norm": 1.0,
      "learning_rate": 0.0004979428637967009,
      "loss": 5.2801,
      "mean_token_accuracy": 0.1863731786608696,
      "num_tokens": 11254437.0,
      "step": 6115
    },
    {
      "entropy": 5.448827266693115,
      "epoch": 0.5111927831607083,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004979388454280189,
      "loss": 5.1287,
      "mean_token_accuracy": 0.19712733775377272,
      "num_tokens": 11262413.0,
      "step": 6120
    },
    {
      "entropy": 5.346326923370361,
      "epoch": 0.5116104243234213,
      "grad_norm": 1.046875,
      "learning_rate": 0.000497934823156524,
      "loss": 5.3235,
      "mean_token_accuracy": 0.1839967966079712,
      "num_tokens": 11272061.0,
      "step": 6125
    },
    {
      "entropy": 5.3600989818573,
      "epoch": 0.5120280654861343,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004979307969822869,
      "loss": 5.1424,
      "mean_token_accuracy": 0.19436496645212173,
      "num_tokens": 11281623.0,
      "step": 6130
    },
    {
      "entropy": 5.331154012680054,
      "epoch": 0.5124457066488474,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004979267669053776,
      "loss": 5.0208,
      "mean_token_accuracy": 0.20938571095466613,
      "num_tokens": 11290202.0,
      "step": 6135
    },
    {
      "entropy": 5.4022496223449705,
      "epoch": 0.5128633478115603,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004979227329258671,
      "loss": 5.2883,
      "mean_token_accuracy": 0.19044555872678756,
      "num_tokens": 11299040.0,
      "step": 6140
    },
    {
      "entropy": 5.4168483257293705,
      "epoch": 0.5132809889742733,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004979186950438256,
      "loss": 5.2098,
      "mean_token_accuracy": 0.18950813859701157,
      "num_tokens": 11307938.0,
      "step": 6145
    },
    {
      "entropy": 5.392462348937988,
      "epoch": 0.5136986301369864,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004979146532593243,
      "loss": 5.2109,
      "mean_token_accuracy": 0.19101665765047074,
      "num_tokens": 11316915.0,
      "step": 6150
    },
    {
      "entropy": 5.411058759689331,
      "epoch": 0.5141162712996993,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004979106075724334,
      "loss": 5.1761,
      "mean_token_accuracy": 0.1905566856265068,
      "num_tokens": 11326634.0,
      "step": 6155
    },
    {
      "entropy": 5.496545314788818,
      "epoch": 0.5145339124624123,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004979065579832242,
      "loss": 5.3332,
      "mean_token_accuracy": 0.18164477050304412,
      "num_tokens": 11336502.0,
      "step": 6160
    },
    {
      "entropy": 5.380741310119629,
      "epoch": 0.5149515536251253,
      "grad_norm": 1.0,
      "learning_rate": 0.0004979025044917673,
      "loss": 5.2367,
      "mean_token_accuracy": 0.18871651589870453,
      "num_tokens": 11346125.0,
      "step": 6165
    },
    {
      "entropy": 5.475302743911743,
      "epoch": 0.5153691947878383,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004978984470981339,
      "loss": 5.3422,
      "mean_token_accuracy": 0.17759273797273636,
      "num_tokens": 11355713.0,
      "step": 6170
    },
    {
      "entropy": 5.367629814147949,
      "epoch": 0.5157868359505513,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004978943858023946,
      "loss": 5.1533,
      "mean_token_accuracy": 0.19492397755384444,
      "num_tokens": 11364726.0,
      "step": 6175
    },
    {
      "entropy": 5.351145267486572,
      "epoch": 0.5162044771132643,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000497890320604621,
      "loss": 5.1445,
      "mean_token_accuracy": 0.18829655647277832,
      "num_tokens": 11373779.0,
      "step": 6180
    },
    {
      "entropy": 5.3211218357086185,
      "epoch": 0.5166221182759773,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000497886251504884,
      "loss": 5.1448,
      "mean_token_accuracy": 0.19589279294013978,
      "num_tokens": 11382641.0,
      "step": 6185
    },
    {
      "entropy": 5.341792631149292,
      "epoch": 0.5170397594386903,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004978821785032549,
      "loss": 5.1126,
      "mean_token_accuracy": 0.19628737568855287,
      "num_tokens": 11392200.0,
      "step": 6190
    },
    {
      "entropy": 5.398201513290405,
      "epoch": 0.5174574006014033,
      "grad_norm": 1.0,
      "learning_rate": 0.0004978781015998048,
      "loss": 5.2324,
      "mean_token_accuracy": 0.18350589275360107,
      "num_tokens": 11401884.0,
      "step": 6195
    },
    {
      "entropy": 5.393504047393799,
      "epoch": 0.5178750417641163,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004978740207946056,
      "loss": 5.1309,
      "mean_token_accuracy": 0.20106103867292405,
      "num_tokens": 11410710.0,
      "step": 6200
    },
    {
      "entropy": 5.296826171875,
      "epoch": 0.5182926829268293,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004978699360877281,
      "loss": 5.095,
      "mean_token_accuracy": 0.19616660624742507,
      "num_tokens": 11419622.0,
      "step": 6205
    },
    {
      "entropy": 5.3391142845153805,
      "epoch": 0.5187103240895423,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004978658474792442,
      "loss": 5.1005,
      "mean_token_accuracy": 0.1983638435602188,
      "num_tokens": 11427631.0,
      "step": 6210
    },
    {
      "entropy": 5.370728206634522,
      "epoch": 0.5191279652522552,
      "grad_norm": 1.0,
      "learning_rate": 0.0004978617549692253,
      "loss": 5.1661,
      "mean_token_accuracy": 0.19099049866199494,
      "num_tokens": 11437630.0,
      "step": 6215
    },
    {
      "entropy": 5.447141838073731,
      "epoch": 0.5195456064149683,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004978576585577432,
      "loss": 5.3721,
      "mean_token_accuracy": 0.1779620349407196,
      "num_tokens": 11447962.0,
      "step": 6220
    },
    {
      "entropy": 5.497256278991699,
      "epoch": 0.5199632475776813,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004978535582448695,
      "loss": 5.2427,
      "mean_token_accuracy": 0.1889750763773918,
      "num_tokens": 11458154.0,
      "step": 6225
    },
    {
      "entropy": 5.345706558227539,
      "epoch": 0.5203808887403942,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004978494540306761,
      "loss": 5.0865,
      "mean_token_accuracy": 0.19210751801729203,
      "num_tokens": 11467449.0,
      "step": 6230
    },
    {
      "entropy": 5.3936372756958,
      "epoch": 0.5207985299031073,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004978453459152346,
      "loss": 5.3511,
      "mean_token_accuracy": 0.1809331953525543,
      "num_tokens": 11477775.0,
      "step": 6235
    },
    {
      "entropy": 5.407290840148926,
      "epoch": 0.5212161710658203,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004978412338986172,
      "loss": 5.1233,
      "mean_token_accuracy": 0.19485079646110534,
      "num_tokens": 11486921.0,
      "step": 6240
    },
    {
      "entropy": 5.305521726608276,
      "epoch": 0.5216338122285332,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004978371179808958,
      "loss": 5.1481,
      "mean_token_accuracy": 0.19802139401435853,
      "num_tokens": 11496562.0,
      "step": 6245
    },
    {
      "entropy": 5.372973203659058,
      "epoch": 0.5220514533912463,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004978329981621424,
      "loss": 5.1172,
      "mean_token_accuracy": 0.18736779242753981,
      "num_tokens": 11504906.0,
      "step": 6250
    },
    {
      "entropy": 5.337685394287109,
      "epoch": 0.5224690945539593,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000497828874442429,
      "loss": 5.1586,
      "mean_token_accuracy": 0.19286476373672484,
      "num_tokens": 11513757.0,
      "step": 6255
    },
    {
      "entropy": 5.316742753982544,
      "epoch": 0.5228867357166722,
      "grad_norm": 0.90625,
      "learning_rate": 0.000497824746821828,
      "loss": 5.173,
      "mean_token_accuracy": 0.18971098959445953,
      "num_tokens": 11523195.0,
      "step": 6260
    },
    {
      "entropy": 5.4230451583862305,
      "epoch": 0.5233043768793852,
      "grad_norm": 1.0,
      "learning_rate": 0.0004978206153004117,
      "loss": 5.2103,
      "mean_token_accuracy": 0.1938289999961853,
      "num_tokens": 11532348.0,
      "step": 6265
    },
    {
      "entropy": 5.381652784347534,
      "epoch": 0.5237220180420983,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004978164798782523,
      "loss": 5.2103,
      "mean_token_accuracy": 0.19268518388271333,
      "num_tokens": 11542356.0,
      "step": 6270
    },
    {
      "entropy": 5.35362434387207,
      "epoch": 0.5241396592048112,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004978123405554221,
      "loss": 5.2214,
      "mean_token_accuracy": 0.19690311700105667,
      "num_tokens": 11550914.0,
      "step": 6275
    },
    {
      "entropy": 5.388011646270752,
      "epoch": 0.5245573003675242,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004978081973319938,
      "loss": 5.1608,
      "mean_token_accuracy": 0.19354240745306014,
      "num_tokens": 11559314.0,
      "step": 6280
    },
    {
      "entropy": 5.30581521987915,
      "epoch": 0.5249749415302373,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004978040502080398,
      "loss": 5.1348,
      "mean_token_accuracy": 0.19505325853824615,
      "num_tokens": 11567990.0,
      "step": 6285
    },
    {
      "entropy": 5.346914863586425,
      "epoch": 0.5253925826929502,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004977998991836327,
      "loss": 5.1537,
      "mean_token_accuracy": 0.19276527166366578,
      "num_tokens": 11577133.0,
      "step": 6290
    },
    {
      "entropy": 5.401226615905761,
      "epoch": 0.5258102238556632,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004977957442588451,
      "loss": 5.236,
      "mean_token_accuracy": 0.19031593054533005,
      "num_tokens": 11585330.0,
      "step": 6295
    },
    {
      "entropy": 5.425922060012818,
      "epoch": 0.5262278650183763,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004977915854337498,
      "loss": 5.1807,
      "mean_token_accuracy": 0.18993562906980516,
      "num_tokens": 11595005.0,
      "step": 6300
    },
    {
      "entropy": 5.314674520492554,
      "epoch": 0.5266455061810892,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004977874227084197,
      "loss": 5.0912,
      "mean_token_accuracy": 0.19887234568595885,
      "num_tokens": 11603969.0,
      "step": 6305
    },
    {
      "entropy": 5.307637023925781,
      "epoch": 0.5270631473438022,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004977832560829277,
      "loss": 5.1425,
      "mean_token_accuracy": 0.18859147727489473,
      "num_tokens": 11613118.0,
      "step": 6310
    },
    {
      "entropy": 5.348072671890259,
      "epoch": 0.5274807885065153,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004977790855573464,
      "loss": 5.0717,
      "mean_token_accuracy": 0.2019173637032509,
      "num_tokens": 11621799.0,
      "step": 6315
    },
    {
      "entropy": 5.326263856887818,
      "epoch": 0.5278984296692282,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004977749111317492,
      "loss": 5.1733,
      "mean_token_accuracy": 0.19220163077116012,
      "num_tokens": 11630557.0,
      "step": 6320
    },
    {
      "entropy": 5.344787120819092,
      "epoch": 0.5283160708319412,
      "grad_norm": 0.96875,
      "learning_rate": 0.000497770732806209,
      "loss": 5.1888,
      "mean_token_accuracy": 0.195334355533123,
      "num_tokens": 11639179.0,
      "step": 6325
    },
    {
      "entropy": 5.424606084823608,
      "epoch": 0.5287337119946542,
      "grad_norm": 1.109375,
      "learning_rate": 0.000497766550580799,
      "loss": 5.2794,
      "mean_token_accuracy": 0.18798788338899614,
      "num_tokens": 11648273.0,
      "step": 6330
    },
    {
      "entropy": 5.376404094696045,
      "epoch": 0.5291513531573672,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004977623644555923,
      "loss": 5.1824,
      "mean_token_accuracy": 0.18783495277166368,
      "num_tokens": 11656372.0,
      "step": 6335
    },
    {
      "entropy": 5.381846380233765,
      "epoch": 0.5295689943200802,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004977581744306624,
      "loss": 5.1344,
      "mean_token_accuracy": 0.19228546768426896,
      "num_tokens": 11665007.0,
      "step": 6340
    },
    {
      "entropy": 5.382721614837647,
      "epoch": 0.5299866354827932,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004977539805060824,
      "loss": 5.2376,
      "mean_token_accuracy": 0.19090797007083893,
      "num_tokens": 11674519.0,
      "step": 6345
    },
    {
      "entropy": 5.41424970626831,
      "epoch": 0.5304042766455062,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004977497826819259,
      "loss": 5.1658,
      "mean_token_accuracy": 0.19143140465021133,
      "num_tokens": 11683705.0,
      "step": 6350
    },
    {
      "entropy": 5.319455671310425,
      "epoch": 0.5308219178082192,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004977455809582664,
      "loss": 5.1405,
      "mean_token_accuracy": 0.19384844303131105,
      "num_tokens": 11693152.0,
      "step": 6355
    },
    {
      "entropy": 5.366359567642212,
      "epoch": 0.5312395589709322,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004977413753351772,
      "loss": 5.1623,
      "mean_token_accuracy": 0.19003622829914094,
      "num_tokens": 11702584.0,
      "step": 6360
    },
    {
      "entropy": 5.358088970184326,
      "epoch": 0.5316572001336451,
      "grad_norm": 1.0,
      "learning_rate": 0.0004977371658127323,
      "loss": 5.2123,
      "mean_token_accuracy": 0.1920382097363472,
      "num_tokens": 11711633.0,
      "step": 6365
    },
    {
      "entropy": 5.376801252365112,
      "epoch": 0.5320748412963582,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004977329523910051,
      "loss": 5.1553,
      "mean_token_accuracy": 0.19794583469629287,
      "num_tokens": 11720116.0,
      "step": 6370
    },
    {
      "entropy": 5.426304483413697,
      "epoch": 0.5324924824590712,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004977287350700695,
      "loss": 5.2394,
      "mean_token_accuracy": 0.18932390213012695,
      "num_tokens": 11729564.0,
      "step": 6375
    },
    {
      "entropy": 5.365380668640137,
      "epoch": 0.5329101236217841,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004977245138499993,
      "loss": 5.108,
      "mean_token_accuracy": 0.2013540968298912,
      "num_tokens": 11737935.0,
      "step": 6380
    },
    {
      "entropy": 5.2728128910064695,
      "epoch": 0.5333277647844972,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004977202887308684,
      "loss": 5.0925,
      "mean_token_accuracy": 0.19509465545415877,
      "num_tokens": 11745989.0,
      "step": 6385
    },
    {
      "entropy": 5.346738576889038,
      "epoch": 0.5337454059472102,
      "grad_norm": 1.125,
      "learning_rate": 0.0004977160597127507,
      "loss": 5.2605,
      "mean_token_accuracy": 0.18902895450592042,
      "num_tokens": 11755530.0,
      "step": 6390
    },
    {
      "entropy": 5.411718416213989,
      "epoch": 0.5341630471099231,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004977118267957204,
      "loss": 5.2125,
      "mean_token_accuracy": 0.18889476209878922,
      "num_tokens": 11764546.0,
      "step": 6395
    },
    {
      "entropy": 5.488783884048462,
      "epoch": 0.5345806882726362,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004977075899798513,
      "loss": 5.3533,
      "mean_token_accuracy": 0.18192564994096755,
      "num_tokens": 11774117.0,
      "step": 6400
    },
    {
      "entropy": 5.4040234088897705,
      "epoch": 0.5349983294353492,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000497703349265218,
      "loss": 5.1183,
      "mean_token_accuracy": 0.20040978044271468,
      "num_tokens": 11782952.0,
      "step": 6405
    },
    {
      "entropy": 5.390673732757568,
      "epoch": 0.5354159705980621,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004976991046518943,
      "loss": 5.2812,
      "mean_token_accuracy": 0.18780191838741303,
      "num_tokens": 11792369.0,
      "step": 6410
    },
    {
      "entropy": 5.359350299835205,
      "epoch": 0.5358336117607752,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004976948561399549,
      "loss": 5.1233,
      "mean_token_accuracy": 0.19984854608774186,
      "num_tokens": 11801365.0,
      "step": 6415
    },
    {
      "entropy": 5.400640964508057,
      "epoch": 0.5362512529234882,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004976906037294738,
      "loss": 5.1927,
      "mean_token_accuracy": 0.1966977521777153,
      "num_tokens": 11809785.0,
      "step": 6420
    },
    {
      "entropy": 5.327585029602051,
      "epoch": 0.5366688940862011,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004976863474205257,
      "loss": 5.0841,
      "mean_token_accuracy": 0.1984919399023056,
      "num_tokens": 11817978.0,
      "step": 6425
    },
    {
      "entropy": 5.378802824020386,
      "epoch": 0.5370865352489141,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004976820872131851,
      "loss": 5.202,
      "mean_token_accuracy": 0.19181052148342131,
      "num_tokens": 11827796.0,
      "step": 6430
    },
    {
      "entropy": 5.254422950744629,
      "epoch": 0.5375041764116272,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004976778231075266,
      "loss": 5.0739,
      "mean_token_accuracy": 0.19414258897304534,
      "num_tokens": 11836439.0,
      "step": 6435
    },
    {
      "entropy": 5.3741758346557615,
      "epoch": 0.5379218175743401,
      "grad_norm": 1.0,
      "learning_rate": 0.0004976735551036245,
      "loss": 5.2643,
      "mean_token_accuracy": 0.18768078535795213,
      "num_tokens": 11845268.0,
      "step": 6440
    },
    {
      "entropy": 5.435879802703857,
      "epoch": 0.5383394587370531,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000497669283201554,
      "loss": 5.1781,
      "mean_token_accuracy": 0.18884878754615783,
      "num_tokens": 11855185.0,
      "step": 6445
    },
    {
      "entropy": 5.364731931686402,
      "epoch": 0.5387570998997662,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004976650074013895,
      "loss": 5.2031,
      "mean_token_accuracy": 0.1893807753920555,
      "num_tokens": 11865073.0,
      "step": 6450
    },
    {
      "entropy": 5.375470399856567,
      "epoch": 0.5391747410624791,
      "grad_norm": 0.875,
      "learning_rate": 0.0004976607277032061,
      "loss": 5.2068,
      "mean_token_accuracy": 0.19191755205392838,
      "num_tokens": 11875004.0,
      "step": 6455
    },
    {
      "entropy": 5.421530246734619,
      "epoch": 0.5395923822251921,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004976564441070788,
      "loss": 5.1873,
      "mean_token_accuracy": 0.18632589727640153,
      "num_tokens": 11884905.0,
      "step": 6460
    },
    {
      "entropy": 5.436643600463867,
      "epoch": 0.5400100233879052,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004976521566130823,
      "loss": 5.1858,
      "mean_token_accuracy": 0.18591734021902084,
      "num_tokens": 11894669.0,
      "step": 6465
    },
    {
      "entropy": 5.486261653900146,
      "epoch": 0.5404276645506181,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004976478652212918,
      "loss": 5.3654,
      "mean_token_accuracy": 0.17533845603466033,
      "num_tokens": 11904127.0,
      "step": 6470
    },
    {
      "entropy": 5.4235564231872555,
      "epoch": 0.5408453057133311,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004976435699317824,
      "loss": 5.2026,
      "mean_token_accuracy": 0.18949461132287979,
      "num_tokens": 11912031.0,
      "step": 6475
    },
    {
      "entropy": 5.311607408523559,
      "epoch": 0.5412629468760441,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004976392707446293,
      "loss": 5.1104,
      "mean_token_accuracy": 0.19925084859132766,
      "num_tokens": 11920961.0,
      "step": 6480
    },
    {
      "entropy": 5.428405809402466,
      "epoch": 0.5416805880387571,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004976349676599078,
      "loss": 5.246,
      "mean_token_accuracy": 0.1853615939617157,
      "num_tokens": 11930454.0,
      "step": 6485
    },
    {
      "entropy": 5.309107303619385,
      "epoch": 0.5420982292014701,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004976306606776932,
      "loss": 5.0526,
      "mean_token_accuracy": 0.1942919448018074,
      "num_tokens": 11938610.0,
      "step": 6490
    },
    {
      "entropy": 5.341273021697998,
      "epoch": 0.5425158703641831,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004976263497980609,
      "loss": 5.2302,
      "mean_token_accuracy": 0.1937536284327507,
      "num_tokens": 11947882.0,
      "step": 6495
    },
    {
      "entropy": 5.404484367370605,
      "epoch": 0.5429335115268961,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004976220350210864,
      "loss": 5.2115,
      "mean_token_accuracy": 0.19132473170757294,
      "num_tokens": 11957552.0,
      "step": 6500
    },
    {
      "entropy": 5.356153774261474,
      "epoch": 0.5433511526896091,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004976177163468451,
      "loss": 5.1356,
      "mean_token_accuracy": 0.190580315887928,
      "num_tokens": 11967350.0,
      "step": 6505
    },
    {
      "entropy": 5.335919570922852,
      "epoch": 0.5437687938523221,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004976133937754128,
      "loss": 5.1517,
      "mean_token_accuracy": 0.18836811780929566,
      "num_tokens": 11976107.0,
      "step": 6510
    },
    {
      "entropy": 5.379608869552612,
      "epoch": 0.544186435015035,
      "grad_norm": 1.0,
      "learning_rate": 0.0004976090673068651,
      "loss": 5.143,
      "mean_token_accuracy": 0.19728839248418809,
      "num_tokens": 11985059.0,
      "step": 6515
    },
    {
      "entropy": 5.345880937576294,
      "epoch": 0.5446040761777481,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004976047369412778,
      "loss": 5.1207,
      "mean_token_accuracy": 0.19248676747083665,
      "num_tokens": 11993453.0,
      "step": 6520
    },
    {
      "entropy": 5.254969215393066,
      "epoch": 0.5450217173404611,
      "grad_norm": 1.0,
      "learning_rate": 0.0004976004026787266,
      "loss": 5.0891,
      "mean_token_accuracy": 0.20255645960569382,
      "num_tokens": 12002524.0,
      "step": 6525
    },
    {
      "entropy": 5.296452760696411,
      "epoch": 0.545439358503174,
      "grad_norm": 1.0,
      "learning_rate": 0.0004975960645192873,
      "loss": 5.1439,
      "mean_token_accuracy": 0.19440295845270156,
      "num_tokens": 12011516.0,
      "step": 6530
    },
    {
      "entropy": 5.349509859085083,
      "epoch": 0.5458569996658871,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004975917224630359,
      "loss": 5.179,
      "mean_token_accuracy": 0.18909975737333298,
      "num_tokens": 12020058.0,
      "step": 6535
    },
    {
      "entropy": 5.4346466064453125,
      "epoch": 0.5462746408286001,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004975873765100487,
      "loss": 5.286,
      "mean_token_accuracy": 0.18941357582807541,
      "num_tokens": 12029167.0,
      "step": 6540
    },
    {
      "entropy": 5.403171586990356,
      "epoch": 0.546692281991313,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004975830266604013,
      "loss": 5.2185,
      "mean_token_accuracy": 0.19409911036491395,
      "num_tokens": 12038298.0,
      "step": 6545
    },
    {
      "entropy": 5.4007080554962155,
      "epoch": 0.5471099231540261,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004975786729141702,
      "loss": 5.1572,
      "mean_token_accuracy": 0.19028274714946747,
      "num_tokens": 12047437.0,
      "step": 6550
    },
    {
      "entropy": 5.381250715255737,
      "epoch": 0.5475275643167391,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004975743152714314,
      "loss": 5.1948,
      "mean_token_accuracy": 0.197520150244236,
      "num_tokens": 12055821.0,
      "step": 6555
    },
    {
      "entropy": 5.363318347930909,
      "epoch": 0.547945205479452,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004975699537322616,
      "loss": 5.0839,
      "mean_token_accuracy": 0.20049280524253846,
      "num_tokens": 12064378.0,
      "step": 6560
    },
    {
      "entropy": 5.2858805656433105,
      "epoch": 0.5483628466421651,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004975655882967366,
      "loss": 5.1418,
      "mean_token_accuracy": 0.19583076238632202,
      "num_tokens": 12073110.0,
      "step": 6565
    },
    {
      "entropy": 5.363436841964722,
      "epoch": 0.5487804878048781,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004975612189649331,
      "loss": 5.1667,
      "mean_token_accuracy": 0.19316457360982894,
      "num_tokens": 12081224.0,
      "step": 6570
    },
    {
      "entropy": 5.37830286026001,
      "epoch": 0.549198128967591,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004975568457369276,
      "loss": 5.2264,
      "mean_token_accuracy": 0.19384944885969163,
      "num_tokens": 12090279.0,
      "step": 6575
    },
    {
      "entropy": 5.422812843322754,
      "epoch": 0.549615770130304,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004975524686127966,
      "loss": 5.205,
      "mean_token_accuracy": 0.19403305500745774,
      "num_tokens": 12098818.0,
      "step": 6580
    },
    {
      "entropy": 5.361269903182984,
      "epoch": 0.5500334112930171,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004975480875926168,
      "loss": 5.1603,
      "mean_token_accuracy": 0.19848973751068116,
      "num_tokens": 12107358.0,
      "step": 6585
    },
    {
      "entropy": 5.3718503475189205,
      "epoch": 0.55045105245573,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004975437026764648,
      "loss": 5.2129,
      "mean_token_accuracy": 0.18422484695911406,
      "num_tokens": 12116802.0,
      "step": 6590
    },
    {
      "entropy": 5.40802845954895,
      "epoch": 0.550868693618443,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004975393138644175,
      "loss": 5.1393,
      "mean_token_accuracy": 0.19678865522146224,
      "num_tokens": 12126347.0,
      "step": 6595
    },
    {
      "entropy": 5.264656686782837,
      "epoch": 0.5512863347811561,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004975349211565515,
      "loss": 5.1072,
      "mean_token_accuracy": 0.19755376726388932,
      "num_tokens": 12136138.0,
      "step": 6600
    },
    {
      "entropy": 5.27839412689209,
      "epoch": 0.551703975943869,
      "grad_norm": 0.890625,
      "learning_rate": 0.000497530524552944,
      "loss": 5.1811,
      "mean_token_accuracy": 0.19448596835136414,
      "num_tokens": 12146436.0,
      "step": 6605
    },
    {
      "entropy": 5.380583810806274,
      "epoch": 0.552121617106582,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004975261240536716,
      "loss": 5.177,
      "mean_token_accuracy": 0.18990160673856735,
      "num_tokens": 12157003.0,
      "step": 6610
    },
    {
      "entropy": 5.374546194076538,
      "epoch": 0.5525392582692951,
      "grad_norm": 0.875,
      "learning_rate": 0.0004975217196588117,
      "loss": 5.107,
      "mean_token_accuracy": 0.20234341472387313,
      "num_tokens": 12166671.0,
      "step": 6615
    },
    {
      "entropy": 5.307311868667602,
      "epoch": 0.552956899432008,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004975173113684413,
      "loss": 5.1669,
      "mean_token_accuracy": 0.1954772651195526,
      "num_tokens": 12176474.0,
      "step": 6620
    },
    {
      "entropy": 5.365773677825928,
      "epoch": 0.553374540594721,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004975128991826374,
      "loss": 5.1636,
      "mean_token_accuracy": 0.19454141855239868,
      "num_tokens": 12185307.0,
      "step": 6625
    },
    {
      "entropy": 5.399585247039795,
      "epoch": 0.553792181757434,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004975084831014774,
      "loss": 5.127,
      "mean_token_accuracy": 0.19482304602861406,
      "num_tokens": 12195179.0,
      "step": 6630
    },
    {
      "entropy": 5.339863967895508,
      "epoch": 0.554209822920147,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004975040631250386,
      "loss": 5.192,
      "mean_token_accuracy": 0.1984764739871025,
      "num_tokens": 12204083.0,
      "step": 6635
    },
    {
      "entropy": 5.299835681915283,
      "epoch": 0.55462746408286,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004974996392533983,
      "loss": 5.1191,
      "mean_token_accuracy": 0.1997724786400795,
      "num_tokens": 12213322.0,
      "step": 6640
    },
    {
      "entropy": 5.34976372718811,
      "epoch": 0.555045105245573,
      "grad_norm": 1.0,
      "learning_rate": 0.000497495211486634,
      "loss": 5.1145,
      "mean_token_accuracy": 0.1989807367324829,
      "num_tokens": 12222305.0,
      "step": 6645
    },
    {
      "entropy": 5.415833759307861,
      "epoch": 0.555462746408286,
      "grad_norm": 1.0,
      "learning_rate": 0.0004974907798248232,
      "loss": 5.2288,
      "mean_token_accuracy": 0.18770413845777512,
      "num_tokens": 12232196.0,
      "step": 6650
    },
    {
      "entropy": 5.393482208251953,
      "epoch": 0.555880387570999,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004974863442680434,
      "loss": 5.1702,
      "mean_token_accuracy": 0.20094767659902574,
      "num_tokens": 12241914.0,
      "step": 6655
    },
    {
      "entropy": 5.218241119384766,
      "epoch": 0.556298028733712,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004974819048163725,
      "loss": 5.0577,
      "mean_token_accuracy": 0.20477695167064666,
      "num_tokens": 12250185.0,
      "step": 6660
    },
    {
      "entropy": 5.351289749145508,
      "epoch": 0.556715669896425,
      "grad_norm": 0.9375,
      "learning_rate": 0.000497477461469888,
      "loss": 5.2277,
      "mean_token_accuracy": 0.1875116631388664,
      "num_tokens": 12258954.0,
      "step": 6665
    },
    {
      "entropy": 5.456501531600952,
      "epoch": 0.557133311059138,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004974730142286677,
      "loss": 5.2402,
      "mean_token_accuracy": 0.1866089791059494,
      "num_tokens": 12268320.0,
      "step": 6670
    },
    {
      "entropy": 5.3339526653289795,
      "epoch": 0.557550952221851,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004974685630927895,
      "loss": 5.115,
      "mean_token_accuracy": 0.1903128743171692,
      "num_tokens": 12276804.0,
      "step": 6675
    },
    {
      "entropy": 5.426059436798096,
      "epoch": 0.557968593384564,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004974641080623314,
      "loss": 5.1657,
      "mean_token_accuracy": 0.1880035474896431,
      "num_tokens": 12286024.0,
      "step": 6680
    },
    {
      "entropy": 5.406354999542236,
      "epoch": 0.558386234547277,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004974596491373711,
      "loss": 5.2066,
      "mean_token_accuracy": 0.1945355072617531,
      "num_tokens": 12295154.0,
      "step": 6685
    },
    {
      "entropy": 5.390251731872558,
      "epoch": 0.55880387570999,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004974551863179871,
      "loss": 5.2498,
      "mean_token_accuracy": 0.18046300262212753,
      "num_tokens": 12306349.0,
      "step": 6690
    },
    {
      "entropy": 5.395963144302368,
      "epoch": 0.5592215168727029,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004974507196042572,
      "loss": 5.2123,
      "mean_token_accuracy": 0.1926526978611946,
      "num_tokens": 12315496.0,
      "step": 6695
    },
    {
      "entropy": 5.259800004959106,
      "epoch": 0.559639158035416,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004974462489962597,
      "loss": 5.0452,
      "mean_token_accuracy": 0.20042645931243896,
      "num_tokens": 12325070.0,
      "step": 6700
    },
    {
      "entropy": 5.357366275787354,
      "epoch": 0.560056799198129,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004974417744940729,
      "loss": 5.1491,
      "mean_token_accuracy": 0.19077862203121185,
      "num_tokens": 12333509.0,
      "step": 6705
    },
    {
      "entropy": 5.3938392162322994,
      "epoch": 0.5604744403608419,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000497437296097775,
      "loss": 5.2314,
      "mean_token_accuracy": 0.1876072943210602,
      "num_tokens": 12342495.0,
      "step": 6710
    },
    {
      "entropy": 5.3260353088378904,
      "epoch": 0.560892081523555,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004974328138074445,
      "loss": 5.1182,
      "mean_token_accuracy": 0.19590144902467727,
      "num_tokens": 12351560.0,
      "step": 6715
    },
    {
      "entropy": 5.32634859085083,
      "epoch": 0.561309722686268,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004974283276231599,
      "loss": 5.2232,
      "mean_token_accuracy": 0.1886861354112625,
      "num_tokens": 12360773.0,
      "step": 6720
    },
    {
      "entropy": 5.399962520599365,
      "epoch": 0.5617273638489809,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004974238375449996,
      "loss": 5.2255,
      "mean_token_accuracy": 0.18719301968812943,
      "num_tokens": 12370519.0,
      "step": 6725
    },
    {
      "entropy": 5.431478691101074,
      "epoch": 0.562145005011694,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004974193435730424,
      "loss": 5.2092,
      "mean_token_accuracy": 0.1849524199962616,
      "num_tokens": 12380563.0,
      "step": 6730
    },
    {
      "entropy": 5.4053772449493405,
      "epoch": 0.562562646174407,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004974148457073668,
      "loss": 5.2282,
      "mean_token_accuracy": 0.19246326237916947,
      "num_tokens": 12389291.0,
      "step": 6735
    },
    {
      "entropy": 5.360533809661865,
      "epoch": 0.5629802873371199,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004974103439480517,
      "loss": 5.1966,
      "mean_token_accuracy": 0.1933261662721634,
      "num_tokens": 12399037.0,
      "step": 6740
    },
    {
      "entropy": 5.322629117965699,
      "epoch": 0.563397928499833,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004974058382951757,
      "loss": 5.0209,
      "mean_token_accuracy": 0.19869708269834518,
      "num_tokens": 12407553.0,
      "step": 6745
    },
    {
      "entropy": 5.2469738006591795,
      "epoch": 0.563815569662546,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004974013287488177,
      "loss": 5.0786,
      "mean_token_accuracy": 0.19238705635070802,
      "num_tokens": 12416298.0,
      "step": 6750
    },
    {
      "entropy": 5.3720509052276615,
      "epoch": 0.5642332108252589,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004973968153090569,
      "loss": 5.2461,
      "mean_token_accuracy": 0.18268570750951768,
      "num_tokens": 12425918.0,
      "step": 6755
    },
    {
      "entropy": 5.387443161010742,
      "epoch": 0.5646508519879719,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004973922979759722,
      "loss": 5.1674,
      "mean_token_accuracy": 0.19380963295698167,
      "num_tokens": 12434693.0,
      "step": 6760
    },
    {
      "entropy": 5.270232629776001,
      "epoch": 0.565068493150685,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004973877767496424,
      "loss": 5.0232,
      "mean_token_accuracy": 0.19575973004102706,
      "num_tokens": 12444055.0,
      "step": 6765
    },
    {
      "entropy": 5.316227436065674,
      "epoch": 0.5654861343133979,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004973832516301472,
      "loss": 5.1035,
      "mean_token_accuracy": 0.20128561556339264,
      "num_tokens": 12453012.0,
      "step": 6770
    },
    {
      "entropy": 5.302283716201782,
      "epoch": 0.5659037754761109,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004973787226175652,
      "loss": 5.1343,
      "mean_token_accuracy": 0.1996360793709755,
      "num_tokens": 12462292.0,
      "step": 6775
    },
    {
      "entropy": 5.283322763442993,
      "epoch": 0.566321416638824,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004973741897119761,
      "loss": 5.103,
      "mean_token_accuracy": 0.20736440569162368,
      "num_tokens": 12471178.0,
      "step": 6780
    },
    {
      "entropy": 5.310670518875122,
      "epoch": 0.5667390578015369,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004973696529134592,
      "loss": 5.1409,
      "mean_token_accuracy": 0.20046576708555222,
      "num_tokens": 12479977.0,
      "step": 6785
    },
    {
      "entropy": 5.298818588256836,
      "epoch": 0.5671566989642499,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004973651122220938,
      "loss": 5.1992,
      "mean_token_accuracy": 0.19250317960977553,
      "num_tokens": 12489004.0,
      "step": 6790
    },
    {
      "entropy": 5.404032611846924,
      "epoch": 0.567574340126963,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004973605676379595,
      "loss": 5.2087,
      "mean_token_accuracy": 0.18871997892856598,
      "num_tokens": 12498979.0,
      "step": 6795
    },
    {
      "entropy": 5.444263172149658,
      "epoch": 0.5679919812896759,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004973560191611357,
      "loss": 5.2648,
      "mean_token_accuracy": 0.18371393978595735,
      "num_tokens": 12508293.0,
      "step": 6800
    },
    {
      "entropy": 5.486499547958374,
      "epoch": 0.5684096224523889,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004973514667917021,
      "loss": 5.2833,
      "mean_token_accuracy": 0.18744366019964218,
      "num_tokens": 12517828.0,
      "step": 6805
    },
    {
      "entropy": 5.380046224594116,
      "epoch": 0.5688272636151019,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004973469105297386,
      "loss": 5.1314,
      "mean_token_accuracy": 0.1887920618057251,
      "num_tokens": 12526569.0,
      "step": 6810
    },
    {
      "entropy": 5.38340277671814,
      "epoch": 0.5692449047778149,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004973423503753247,
      "loss": 5.2684,
      "mean_token_accuracy": 0.18637744039297105,
      "num_tokens": 12535706.0,
      "step": 6815
    },
    {
      "entropy": 5.416188526153564,
      "epoch": 0.5696625459405279,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004973377863285403,
      "loss": 5.1311,
      "mean_token_accuracy": 0.18947291225194932,
      "num_tokens": 12544689.0,
      "step": 6820
    },
    {
      "entropy": 5.2991420269012455,
      "epoch": 0.5700801871032409,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004973332183894653,
      "loss": 5.1219,
      "mean_token_accuracy": 0.19626072645187378,
      "num_tokens": 12553318.0,
      "step": 6825
    },
    {
      "entropy": 5.310688781738281,
      "epoch": 0.5704978282659539,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004973286465581798,
      "loss": 5.0987,
      "mean_token_accuracy": 0.2027292102575302,
      "num_tokens": 12562009.0,
      "step": 6830
    },
    {
      "entropy": 5.4232518672943115,
      "epoch": 0.5709154694286669,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004973240708347637,
      "loss": 5.2339,
      "mean_token_accuracy": 0.19799666553735734,
      "num_tokens": 12571043.0,
      "step": 6835
    },
    {
      "entropy": 5.4503764629364015,
      "epoch": 0.5713331105913799,
      "grad_norm": 0.875,
      "learning_rate": 0.0004973194912192971,
      "loss": 5.1493,
      "mean_token_accuracy": 0.19454740583896638,
      "num_tokens": 12581151.0,
      "step": 6840
    },
    {
      "entropy": 5.36432991027832,
      "epoch": 0.5717507517540928,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004973149077118602,
      "loss": 5.184,
      "mean_token_accuracy": 0.19582328498363494,
      "num_tokens": 12590671.0,
      "step": 6845
    },
    {
      "entropy": 5.359883785247803,
      "epoch": 0.5721683929168059,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004973103203125333,
      "loss": 5.2323,
      "mean_token_accuracy": 0.19277480840682984,
      "num_tokens": 12599693.0,
      "step": 6850
    },
    {
      "entropy": 5.354971647262573,
      "epoch": 0.5725860340795189,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004973057290213965,
      "loss": 5.1331,
      "mean_token_accuracy": 0.19642451405525208,
      "num_tokens": 12609238.0,
      "step": 6855
    },
    {
      "entropy": 5.342548513412476,
      "epoch": 0.5730036752422318,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004973011338385304,
      "loss": 5.1362,
      "mean_token_accuracy": 0.19468239992856978,
      "num_tokens": 12618273.0,
      "step": 6860
    },
    {
      "entropy": 5.335744333267212,
      "epoch": 0.5734213164049449,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004972965347640154,
      "loss": 5.1943,
      "mean_token_accuracy": 0.18968237936496735,
      "num_tokens": 12628229.0,
      "step": 6865
    },
    {
      "entropy": 5.375525045394897,
      "epoch": 0.5738389575676579,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000497291931797932,
      "loss": 5.1051,
      "mean_token_accuracy": 0.19963084310293197,
      "num_tokens": 12636533.0,
      "step": 6870
    },
    {
      "entropy": 5.313108348846436,
      "epoch": 0.5742565987303708,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004972873249403608,
      "loss": 5.1344,
      "mean_token_accuracy": 0.19388388395309447,
      "num_tokens": 12645503.0,
      "step": 6875
    },
    {
      "entropy": 5.369278526306152,
      "epoch": 0.5746742398930839,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004972827141913822,
      "loss": 5.1469,
      "mean_token_accuracy": 0.1930328533053398,
      "num_tokens": 12654355.0,
      "step": 6880
    },
    {
      "entropy": 5.365603017807007,
      "epoch": 0.5750918810557969,
      "grad_norm": 1.0,
      "learning_rate": 0.0004972780995510773,
      "loss": 5.126,
      "mean_token_accuracy": 0.19479866474866867,
      "num_tokens": 12663774.0,
      "step": 6885
    },
    {
      "entropy": 5.273893117904663,
      "epoch": 0.5755095222185098,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004972734810195266,
      "loss": 5.1357,
      "mean_token_accuracy": 0.1953548327088356,
      "num_tokens": 12674436.0,
      "step": 6890
    },
    {
      "entropy": 5.421752548217773,
      "epoch": 0.5759271633812229,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004972688585968112,
      "loss": 5.1082,
      "mean_token_accuracy": 0.1914602115750313,
      "num_tokens": 12682891.0,
      "step": 6895
    },
    {
      "entropy": 5.270223712921142,
      "epoch": 0.5763448045439359,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004972642322830118,
      "loss": 5.0037,
      "mean_token_accuracy": 0.20601586401462554,
      "num_tokens": 12691170.0,
      "step": 6900
    },
    {
      "entropy": 5.255709409713745,
      "epoch": 0.5767624457066488,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004972596020782096,
      "loss": 5.0903,
      "mean_token_accuracy": 0.20099268853664398,
      "num_tokens": 12700314.0,
      "step": 6905
    },
    {
      "entropy": 5.342129325866699,
      "epoch": 0.5771800868693618,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004972549679824853,
      "loss": 5.1543,
      "mean_token_accuracy": 0.19703451693058013,
      "num_tokens": 12709418.0,
      "step": 6910
    },
    {
      "entropy": 5.3108916759490965,
      "epoch": 0.5775977280320749,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004972503299959204,
      "loss": 5.126,
      "mean_token_accuracy": 0.19411477148532869,
      "num_tokens": 12719287.0,
      "step": 6915
    },
    {
      "entropy": 5.331398010253906,
      "epoch": 0.5780153691947878,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000497245688118596,
      "loss": 5.0718,
      "mean_token_accuracy": 0.19871993660926818,
      "num_tokens": 12728836.0,
      "step": 6920
    },
    {
      "entropy": 5.403909826278687,
      "epoch": 0.5784330103575008,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004972410423505932,
      "loss": 5.1879,
      "mean_token_accuracy": 0.19033119529485704,
      "num_tokens": 12738173.0,
      "step": 6925
    },
    {
      "entropy": 5.338606262207032,
      "epoch": 0.5788506515202139,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004972363926919935,
      "loss": 5.1613,
      "mean_token_accuracy": 0.2002705842256546,
      "num_tokens": 12747428.0,
      "step": 6930
    },
    {
      "entropy": 5.339896488189697,
      "epoch": 0.5792682926829268,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004972317391428782,
      "loss": 5.2286,
      "mean_token_accuracy": 0.1873060017824173,
      "num_tokens": 12757148.0,
      "step": 6935
    },
    {
      "entropy": 5.4150477886199955,
      "epoch": 0.5796859338456398,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004972270817033288,
      "loss": 5.2363,
      "mean_token_accuracy": 0.1863723009824753,
      "num_tokens": 12766820.0,
      "step": 6940
    },
    {
      "entropy": 5.427341508865356,
      "epoch": 0.5801035750083529,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004972224203734269,
      "loss": 5.2367,
      "mean_token_accuracy": 0.18697526901960373,
      "num_tokens": 12776082.0,
      "step": 6945
    },
    {
      "entropy": 5.369603395462036,
      "epoch": 0.5805212161710658,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000497217755153254,
      "loss": 5.2048,
      "mean_token_accuracy": 0.19589358568191528,
      "num_tokens": 12785057.0,
      "step": 6950
    },
    {
      "entropy": 5.339285469055175,
      "epoch": 0.5809388573337788,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004972130860428918,
      "loss": 5.2049,
      "mean_token_accuracy": 0.19586018919944764,
      "num_tokens": 12793599.0,
      "step": 6955
    },
    {
      "entropy": 5.366975116729736,
      "epoch": 0.5813564984964918,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004972084130424222,
      "loss": 5.1689,
      "mean_token_accuracy": 0.19130999892950057,
      "num_tokens": 12803775.0,
      "step": 6960
    },
    {
      "entropy": 5.3479423999786375,
      "epoch": 0.5817741396592048,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004972037361519267,
      "loss": 5.1151,
      "mean_token_accuracy": 0.1975976273417473,
      "num_tokens": 12812327.0,
      "step": 6965
    },
    {
      "entropy": 5.351316595077515,
      "epoch": 0.5821917808219178,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004971990553714874,
      "loss": 5.0672,
      "mean_token_accuracy": 0.20321837365627288,
      "num_tokens": 12821218.0,
      "step": 6970
    },
    {
      "entropy": 5.346472120285034,
      "epoch": 0.5826094219846308,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004971943707011861,
      "loss": 5.1613,
      "mean_token_accuracy": 0.19154463857412338,
      "num_tokens": 12831579.0,
      "step": 6975
    },
    {
      "entropy": 5.4078680038452145,
      "epoch": 0.5830270631473438,
      "grad_norm": 0.9375,
      "learning_rate": 0.000497189682141105,
      "loss": 5.2151,
      "mean_token_accuracy": 0.19479777067899703,
      "num_tokens": 12841026.0,
      "step": 6980
    },
    {
      "entropy": 5.290831851959228,
      "epoch": 0.5834447043100568,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004971849896913259,
      "loss": 5.1685,
      "mean_token_accuracy": 0.18632251322269439,
      "num_tokens": 12850136.0,
      "step": 6985
    },
    {
      "entropy": 5.368047714233398,
      "epoch": 0.5838623454727698,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004971802933519312,
      "loss": 5.2304,
      "mean_token_accuracy": 0.19398770481348038,
      "num_tokens": 12860345.0,
      "step": 6990
    },
    {
      "entropy": 5.312956190109253,
      "epoch": 0.5842799866354828,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004971755931230031,
      "loss": 5.1719,
      "mean_token_accuracy": 0.19238100945949554,
      "num_tokens": 12868995.0,
      "step": 6995
    },
    {
      "entropy": 5.458567428588867,
      "epoch": 0.5846976277981958,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004971708890046236,
      "loss": 5.3176,
      "mean_token_accuracy": 0.18658459484577178,
      "num_tokens": 12878043.0,
      "step": 7000
    },
    {
      "entropy": 5.339959812164307,
      "epoch": 0.5851152689609088,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004971661809968755,
      "loss": 5.1416,
      "mean_token_accuracy": 0.19239504486322404,
      "num_tokens": 12887090.0,
      "step": 7005
    },
    {
      "entropy": 5.292165470123291,
      "epoch": 0.5855329101236217,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004971614690998409,
      "loss": 5.0789,
      "mean_token_accuracy": 0.2089754894375801,
      "num_tokens": 12896349.0,
      "step": 7010
    },
    {
      "entropy": 5.345342206954956,
      "epoch": 0.5859505512863348,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004971567533136024,
      "loss": 5.1331,
      "mean_token_accuracy": 0.1912553906440735,
      "num_tokens": 12905698.0,
      "step": 7015
    },
    {
      "entropy": 5.287728595733642,
      "epoch": 0.5863681924490478,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004971520336382424,
      "loss": 5.1102,
      "mean_token_accuracy": 0.20301002413034439,
      "num_tokens": 12915146.0,
      "step": 7020
    },
    {
      "entropy": 5.215633726119995,
      "epoch": 0.5867858336117607,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004971473100738437,
      "loss": 4.9937,
      "mean_token_accuracy": 0.2045297682285309,
      "num_tokens": 12924933.0,
      "step": 7025
    },
    {
      "entropy": 5.406471920013428,
      "epoch": 0.5872034747744738,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000497142582620489,
      "loss": 5.2233,
      "mean_token_accuracy": 0.18703412711620332,
      "num_tokens": 12934959.0,
      "step": 7030
    },
    {
      "entropy": 5.34332308769226,
      "epoch": 0.5876211159371868,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004971378512782609,
      "loss": 5.0697,
      "mean_token_accuracy": 0.20307493805885315,
      "num_tokens": 12943470.0,
      "step": 7035
    },
    {
      "entropy": 5.216269445419312,
      "epoch": 0.5880387570998997,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004971331160472424,
      "loss": 5.0929,
      "mean_token_accuracy": 0.2035591036081314,
      "num_tokens": 12953592.0,
      "step": 7040
    },
    {
      "entropy": 5.377045583724976,
      "epoch": 0.5884563982626128,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004971283769275163,
      "loss": 5.203,
      "mean_token_accuracy": 0.1881130665540695,
      "num_tokens": 12962893.0,
      "step": 7045
    },
    {
      "entropy": 5.309810924530029,
      "epoch": 0.5888740394253258,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004971236339191656,
      "loss": 5.0909,
      "mean_token_accuracy": 0.200022853910923,
      "num_tokens": 12971635.0,
      "step": 7050
    },
    {
      "entropy": 5.236034822463989,
      "epoch": 0.5892916805880387,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004971188870222734,
      "loss": 5.0725,
      "mean_token_accuracy": 0.20328239351511002,
      "num_tokens": 12981040.0,
      "step": 7055
    },
    {
      "entropy": 5.315973472595215,
      "epoch": 0.5897093217507517,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004971141362369226,
      "loss": 5.0945,
      "mean_token_accuracy": 0.19796462655067443,
      "num_tokens": 12989898.0,
      "step": 7060
    },
    {
      "entropy": 5.350116443634033,
      "epoch": 0.5901269629134648,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004971093815631965,
      "loss": 5.0794,
      "mean_token_accuracy": 0.1952229455113411,
      "num_tokens": 12998584.0,
      "step": 7065
    },
    {
      "entropy": 5.274097299575805,
      "epoch": 0.5905446040761777,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004971046230011783,
      "loss": 5.1175,
      "mean_token_accuracy": 0.19815860986709594,
      "num_tokens": 13008007.0,
      "step": 7070
    },
    {
      "entropy": 5.346667957305908,
      "epoch": 0.5909622452388907,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004970998605509514,
      "loss": 5.2458,
      "mean_token_accuracy": 0.18941018581390381,
      "num_tokens": 13017034.0,
      "step": 7075
    },
    {
      "entropy": 5.462687635421753,
      "epoch": 0.5913798864016038,
      "grad_norm": 0.984375,
      "learning_rate": 0.000497095094212599,
      "loss": 5.3166,
      "mean_token_accuracy": 0.1829977497458458,
      "num_tokens": 13025916.0,
      "step": 7080
    },
    {
      "entropy": 5.440711116790771,
      "epoch": 0.5917975275643167,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004970903239862047,
      "loss": 5.1919,
      "mean_token_accuracy": 0.18716439753770828,
      "num_tokens": 13035242.0,
      "step": 7085
    },
    {
      "entropy": 5.371887445449829,
      "epoch": 0.5922151687270297,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004970855498718519,
      "loss": 5.1092,
      "mean_token_accuracy": 0.19833823591470717,
      "num_tokens": 13044123.0,
      "step": 7090
    },
    {
      "entropy": 5.313987493515015,
      "epoch": 0.5926328098897428,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004970807718696243,
      "loss": 5.1762,
      "mean_token_accuracy": 0.1908554658293724,
      "num_tokens": 13053503.0,
      "step": 7095
    },
    {
      "entropy": 5.33303689956665,
      "epoch": 0.5930504510524557,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004970759899796055,
      "loss": 5.1617,
      "mean_token_accuracy": 0.19527328461408616,
      "num_tokens": 13062106.0,
      "step": 7100
    },
    {
      "entropy": 5.40173749923706,
      "epoch": 0.5934680922151687,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004970712042018791,
      "loss": 5.148,
      "mean_token_accuracy": 0.19486627876758575,
      "num_tokens": 13071509.0,
      "step": 7105
    },
    {
      "entropy": 5.306280899047851,
      "epoch": 0.5938857333778818,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004970664145365289,
      "loss": 5.0865,
      "mean_token_accuracy": 0.20155565887689592,
      "num_tokens": 13080469.0,
      "step": 7110
    },
    {
      "entropy": 5.3162613868713375,
      "epoch": 0.5943033745405947,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000497061620983639,
      "loss": 5.0525,
      "mean_token_accuracy": 0.20264907628297807,
      "num_tokens": 13089826.0,
      "step": 7115
    },
    {
      "entropy": 5.392253589630127,
      "epoch": 0.5947210157033077,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004970568235432931,
      "loss": 5.2439,
      "mean_token_accuracy": 0.19026910662651061,
      "num_tokens": 13098630.0,
      "step": 7120
    },
    {
      "entropy": 5.367432355880737,
      "epoch": 0.5951386568660207,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004970520222155751,
      "loss": 5.1246,
      "mean_token_accuracy": 0.19708813428878785,
      "num_tokens": 13106819.0,
      "step": 7125
    },
    {
      "entropy": 5.386378526687622,
      "epoch": 0.5955562980287337,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004970472170005692,
      "loss": 5.1438,
      "mean_token_accuracy": 0.19544464200735093,
      "num_tokens": 13115471.0,
      "step": 7130
    },
    {
      "entropy": 5.28109393119812,
      "epoch": 0.5959739391914467,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004970424078983596,
      "loss": 5.0209,
      "mean_token_accuracy": 0.20749326050281525,
      "num_tokens": 13124537.0,
      "step": 7135
    },
    {
      "entropy": 5.274632930755615,
      "epoch": 0.5963915803541597,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004970375949090302,
      "loss": 5.1531,
      "mean_token_accuracy": 0.19775061458349227,
      "num_tokens": 13135326.0,
      "step": 7140
    },
    {
      "entropy": 5.443521738052368,
      "epoch": 0.5968092215168727,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004970327780326655,
      "loss": 5.24,
      "mean_token_accuracy": 0.18989218175411224,
      "num_tokens": 13145502.0,
      "step": 7145
    },
    {
      "entropy": 5.408416700363159,
      "epoch": 0.5972268626795857,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004970279572693498,
      "loss": 5.1508,
      "mean_token_accuracy": 0.20208827257156373,
      "num_tokens": 13154808.0,
      "step": 7150
    },
    {
      "entropy": 5.291624689102173,
      "epoch": 0.5976445038422987,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004970231326191674,
      "loss": 5.257,
      "mean_token_accuracy": 0.18398074060678482,
      "num_tokens": 13163982.0,
      "step": 7155
    },
    {
      "entropy": 5.37433967590332,
      "epoch": 0.5980621450050116,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004970183040822029,
      "loss": 5.1543,
      "mean_token_accuracy": 0.19474094957113267,
      "num_tokens": 13172446.0,
      "step": 7160
    },
    {
      "entropy": 5.3652809143066404,
      "epoch": 0.5984797861677247,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004970134716585407,
      "loss": 5.1256,
      "mean_token_accuracy": 0.1945738360285759,
      "num_tokens": 13181671.0,
      "step": 7165
    },
    {
      "entropy": 5.320513582229614,
      "epoch": 0.5988974273304377,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004970086353482654,
      "loss": 5.0494,
      "mean_token_accuracy": 0.19915290027856827,
      "num_tokens": 13190207.0,
      "step": 7170
    },
    {
      "entropy": 5.329188632965088,
      "epoch": 0.5993150684931506,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004970037951514617,
      "loss": 5.0647,
      "mean_token_accuracy": 0.19937925785779953,
      "num_tokens": 13198866.0,
      "step": 7175
    },
    {
      "entropy": 5.310041666030884,
      "epoch": 0.5997327096558637,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004969989510682143,
      "loss": 5.0891,
      "mean_token_accuracy": 0.199345600605011,
      "num_tokens": 13207921.0,
      "step": 7180
    },
    {
      "entropy": 5.394176816940307,
      "epoch": 0.6001503508185767,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004969941030986082,
      "loss": 5.2389,
      "mean_token_accuracy": 0.19206036627292633,
      "num_tokens": 13217759.0,
      "step": 7185
    },
    {
      "entropy": 5.314651823043823,
      "epoch": 0.6005679919812896,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000496989251242728,
      "loss": 5.0365,
      "mean_token_accuracy": 0.2010976940393448,
      "num_tokens": 13226592.0,
      "step": 7190
    },
    {
      "entropy": 5.267102813720703,
      "epoch": 0.6009856331440027,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004969843955006588,
      "loss": 5.091,
      "mean_token_accuracy": 0.19863041788339614,
      "num_tokens": 13236123.0,
      "step": 7195
    },
    {
      "entropy": 5.355748844146729,
      "epoch": 0.6014032743067157,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004969795358724855,
      "loss": 5.1971,
      "mean_token_accuracy": 0.19445246905088426,
      "num_tokens": 13246169.0,
      "step": 7200
    },
    {
      "entropy": 5.272797584533691,
      "epoch": 0.6018209154694286,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004969746723582932,
      "loss": 5.0684,
      "mean_token_accuracy": 0.20237641781568527,
      "num_tokens": 13256141.0,
      "step": 7205
    },
    {
      "entropy": 5.354504156112671,
      "epoch": 0.6022385566321417,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004969698049581672,
      "loss": 5.2343,
      "mean_token_accuracy": 0.18991373777389525,
      "num_tokens": 13265491.0,
      "step": 7210
    },
    {
      "entropy": 5.48137059211731,
      "epoch": 0.6026561977948547,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004969649336721924,
      "loss": 5.2407,
      "mean_token_accuracy": 0.1982276752591133,
      "num_tokens": 13274805.0,
      "step": 7215
    },
    {
      "entropy": 5.422548913955689,
      "epoch": 0.6030738389575676,
      "grad_norm": 1.0,
      "learning_rate": 0.0004969600585004544,
      "loss": 5.193,
      "mean_token_accuracy": 0.19519280791282653,
      "num_tokens": 13284589.0,
      "step": 7220
    },
    {
      "entropy": 5.295168495178222,
      "epoch": 0.6034914801202806,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004969551794430383,
      "loss": 5.0245,
      "mean_token_accuracy": 0.20875733196735383,
      "num_tokens": 13293370.0,
      "step": 7225
    },
    {
      "entropy": 5.412866592407227,
      "epoch": 0.6039091212829937,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004969502965000297,
      "loss": 5.2119,
      "mean_token_accuracy": 0.19523725509643555,
      "num_tokens": 13302555.0,
      "step": 7230
    },
    {
      "entropy": 5.304563808441162,
      "epoch": 0.6043267624457066,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000496945409671514,
      "loss": 5.1358,
      "mean_token_accuracy": 0.19692928045988084,
      "num_tokens": 13311632.0,
      "step": 7235
    },
    {
      "entropy": 5.353261709213257,
      "epoch": 0.6047444036084196,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004969405189575768,
      "loss": 5.0871,
      "mean_token_accuracy": 0.20443372726440429,
      "num_tokens": 13320593.0,
      "step": 7240
    },
    {
      "entropy": 5.288142585754395,
      "epoch": 0.6051620447711327,
      "grad_norm": 1.0,
      "learning_rate": 0.0004969356243583036,
      "loss": 5.0502,
      "mean_token_accuracy": 0.21058019995689392,
      "num_tokens": 13329095.0,
      "step": 7245
    },
    {
      "entropy": 5.247852849960327,
      "epoch": 0.6055796859338456,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004969307258737802,
      "loss": 5.1077,
      "mean_token_accuracy": 0.19960125982761384,
      "num_tokens": 13338404.0,
      "step": 7250
    },
    {
      "entropy": 5.315798044204712,
      "epoch": 0.6059973270965586,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004969258235040921,
      "loss": 5.0806,
      "mean_token_accuracy": 0.2061529889702797,
      "num_tokens": 13347095.0,
      "step": 7255
    },
    {
      "entropy": 5.3056741714477536,
      "epoch": 0.6064149682592717,
      "grad_norm": 1.0,
      "learning_rate": 0.0004969209172493255,
      "loss": 5.139,
      "mean_token_accuracy": 0.19752053767442704,
      "num_tokens": 13356606.0,
      "step": 7260
    },
    {
      "entropy": 5.3308593273162845,
      "epoch": 0.6068326094219846,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004969160071095661,
      "loss": 5.1455,
      "mean_token_accuracy": 0.19146766662597656,
      "num_tokens": 13366889.0,
      "step": 7265
    },
    {
      "entropy": 5.359483051300049,
      "epoch": 0.6072502505846976,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004969110930848999,
      "loss": 5.1429,
      "mean_token_accuracy": 0.19468463510274886,
      "num_tokens": 13376761.0,
      "step": 7270
    },
    {
      "entropy": 5.324896669387817,
      "epoch": 0.6076678917474106,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004969061751754129,
      "loss": 5.0748,
      "mean_token_accuracy": 0.2029843211174011,
      "num_tokens": 13386226.0,
      "step": 7275
    },
    {
      "entropy": 5.257487773895264,
      "epoch": 0.6080855329101236,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004969012533811911,
      "loss": 5.0169,
      "mean_token_accuracy": 0.2024849086999893,
      "num_tokens": 13394243.0,
      "step": 7280
    },
    {
      "entropy": 5.316019582748413,
      "epoch": 0.6085031740728366,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004968963277023208,
      "loss": 5.2908,
      "mean_token_accuracy": 0.18212942779064178,
      "num_tokens": 13403708.0,
      "step": 7285
    },
    {
      "entropy": 5.348371934890747,
      "epoch": 0.6089208152355496,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004968913981388882,
      "loss": 5.1231,
      "mean_token_accuracy": 0.19981202632188796,
      "num_tokens": 13412911.0,
      "step": 7290
    },
    {
      "entropy": 5.41799840927124,
      "epoch": 0.6093384563982626,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004968864646909797,
      "loss": 5.2519,
      "mean_token_accuracy": 0.19189571738243102,
      "num_tokens": 13422423.0,
      "step": 7295
    },
    {
      "entropy": 5.328569459915161,
      "epoch": 0.6097560975609756,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004968815273586814,
      "loss": 5.1116,
      "mean_token_accuracy": 0.19493812769651414,
      "num_tokens": 13430984.0,
      "step": 7300
    },
    {
      "entropy": 5.299534702301026,
      "epoch": 0.6101737387236886,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004968765861420799,
      "loss": 5.1543,
      "mean_token_accuracy": 0.1970425844192505,
      "num_tokens": 13440492.0,
      "step": 7305
    },
    {
      "entropy": 5.399323844909668,
      "epoch": 0.6105913798864016,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004968716410412617,
      "loss": 5.1441,
      "mean_token_accuracy": 0.191926047205925,
      "num_tokens": 13449222.0,
      "step": 7310
    },
    {
      "entropy": 5.313930606842041,
      "epoch": 0.6110090210491146,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004968666920563134,
      "loss": 5.1071,
      "mean_token_accuracy": 0.19937815517187119,
      "num_tokens": 13459126.0,
      "step": 7315
    },
    {
      "entropy": 5.455713176727295,
      "epoch": 0.6114266622118276,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004968617391873216,
      "loss": 5.2554,
      "mean_token_accuracy": 0.1850447654724121,
      "num_tokens": 13468811.0,
      "step": 7320
    },
    {
      "entropy": 5.365213441848755,
      "epoch": 0.6118443033745405,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004968567824343729,
      "loss": 5.0446,
      "mean_token_accuracy": 0.19441951513290406,
      "num_tokens": 13477765.0,
      "step": 7325
    },
    {
      "entropy": 5.341405773162842,
      "epoch": 0.6122619445372536,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004968518217975542,
      "loss": 5.1895,
      "mean_token_accuracy": 0.18291649222373962,
      "num_tokens": 13487184.0,
      "step": 7330
    },
    {
      "entropy": 5.331346225738526,
      "epoch": 0.6126795856999666,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004968468572769524,
      "loss": 5.1277,
      "mean_token_accuracy": 0.19508163183927535,
      "num_tokens": 13496784.0,
      "step": 7335
    },
    {
      "entropy": 5.345685863494873,
      "epoch": 0.6130972268626795,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004968418888726543,
      "loss": 5.1652,
      "mean_token_accuracy": 0.19293158799409865,
      "num_tokens": 13505824.0,
      "step": 7340
    },
    {
      "entropy": 5.305212211608887,
      "epoch": 0.6135148680253926,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004968369165847469,
      "loss": 5.016,
      "mean_token_accuracy": 0.20005816519260405,
      "num_tokens": 13515111.0,
      "step": 7345
    },
    {
      "entropy": 5.281931924819946,
      "epoch": 0.6139325091881056,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004968319404133173,
      "loss": 5.0485,
      "mean_token_accuracy": 0.20362263768911362,
      "num_tokens": 13524163.0,
      "step": 7350
    },
    {
      "entropy": 5.291459321975708,
      "epoch": 0.6143501503508185,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004968269603584525,
      "loss": 5.0106,
      "mean_token_accuracy": 0.2093982681632042,
      "num_tokens": 13533368.0,
      "step": 7355
    },
    {
      "entropy": 5.342995500564575,
      "epoch": 0.6147677915135316,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004968219764202397,
      "loss": 5.2359,
      "mean_token_accuracy": 0.18994570523500443,
      "num_tokens": 13543708.0,
      "step": 7360
    },
    {
      "entropy": 5.32331485748291,
      "epoch": 0.6151854326762446,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004968169885987663,
      "loss": 5.1648,
      "mean_token_accuracy": 0.18977735787630082,
      "num_tokens": 13552561.0,
      "step": 7365
    },
    {
      "entropy": 5.458197689056396,
      "epoch": 0.6156030738389575,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004968119968941193,
      "loss": 5.2604,
      "mean_token_accuracy": 0.18444412499666213,
      "num_tokens": 13562050.0,
      "step": 7370
    },
    {
      "entropy": 5.277080965042114,
      "epoch": 0.6160207150016705,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004968070013063866,
      "loss": 5.0846,
      "mean_token_accuracy": 0.19239870011806487,
      "num_tokens": 13571356.0,
      "step": 7375
    },
    {
      "entropy": 5.331965637207031,
      "epoch": 0.6164383561643836,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004968020018356552,
      "loss": 5.0778,
      "mean_token_accuracy": 0.19358717948198317,
      "num_tokens": 13581357.0,
      "step": 7380
    },
    {
      "entropy": 5.314277219772339,
      "epoch": 0.6168559973270965,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004967969984820128,
      "loss": 5.0838,
      "mean_token_accuracy": 0.19463482797145842,
      "num_tokens": 13589392.0,
      "step": 7385
    },
    {
      "entropy": 5.329396057128906,
      "epoch": 0.6172736384898095,
      "grad_norm": 1.0,
      "learning_rate": 0.0004967919912455469,
      "loss": 5.0398,
      "mean_token_accuracy": 0.20953108817338945,
      "num_tokens": 13598376.0,
      "step": 7390
    },
    {
      "entropy": 5.185912084579468,
      "epoch": 0.6176912796525226,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004967869801263454,
      "loss": 4.9568,
      "mean_token_accuracy": 0.2108887955546379,
      "num_tokens": 13607014.0,
      "step": 7395
    },
    {
      "entropy": 5.269315958023071,
      "epoch": 0.6181089208152355,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004967819651244957,
      "loss": 5.1333,
      "mean_token_accuracy": 0.19486465752124787,
      "num_tokens": 13615700.0,
      "step": 7400
    },
    {
      "entropy": 5.385510110855103,
      "epoch": 0.6185265619779485,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004967769462400857,
      "loss": 5.1589,
      "mean_token_accuracy": 0.19661862254142762,
      "num_tokens": 13625230.0,
      "step": 7405
    },
    {
      "entropy": 5.4397111415863035,
      "epoch": 0.6189442031406616,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004967719234732033,
      "loss": 5.3219,
      "mean_token_accuracy": 0.18923029005527497,
      "num_tokens": 13635258.0,
      "step": 7410
    },
    {
      "entropy": 5.27698974609375,
      "epoch": 0.6193618443033745,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004967668968239365,
      "loss": 5.0213,
      "mean_token_accuracy": 0.20252898931503296,
      "num_tokens": 13643688.0,
      "step": 7415
    },
    {
      "entropy": 5.29644193649292,
      "epoch": 0.6197794854660875,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004967618662923732,
      "loss": 5.1722,
      "mean_token_accuracy": 0.19252877980470656,
      "num_tokens": 13652894.0,
      "step": 7420
    },
    {
      "entropy": 5.321287775039673,
      "epoch": 0.6201971266288006,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004967568318786016,
      "loss": 5.1053,
      "mean_token_accuracy": 0.19916468113660812,
      "num_tokens": 13662852.0,
      "step": 7425
    },
    {
      "entropy": 5.3512084007263185,
      "epoch": 0.6206147677915135,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004967517935827097,
      "loss": 5.0905,
      "mean_token_accuracy": 0.1965967059135437,
      "num_tokens": 13671457.0,
      "step": 7430
    },
    {
      "entropy": 5.340464162826538,
      "epoch": 0.6210324089542265,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004967467514047856,
      "loss": 5.1691,
      "mean_token_accuracy": 0.19662046581506729,
      "num_tokens": 13681002.0,
      "step": 7435
    },
    {
      "entropy": 5.330549192428589,
      "epoch": 0.6214500501169395,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004967417053449178,
      "loss": 5.1109,
      "mean_token_accuracy": 0.1979174107313156,
      "num_tokens": 13690677.0,
      "step": 7440
    },
    {
      "entropy": 5.277084732055664,
      "epoch": 0.6218676912796525,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004967366554031945,
      "loss": 4.937,
      "mean_token_accuracy": 0.2083348274230957,
      "num_tokens": 13699151.0,
      "step": 7445
    },
    {
      "entropy": 5.2770003318786625,
      "epoch": 0.6222853324423655,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004967316015797042,
      "loss": 5.1473,
      "mean_token_accuracy": 0.1948204219341278,
      "num_tokens": 13708105.0,
      "step": 7450
    },
    {
      "entropy": 5.354609441757202,
      "epoch": 0.6227029736050785,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004967265438745352,
      "loss": 5.2921,
      "mean_token_accuracy": 0.18965820521116256,
      "num_tokens": 13717094.0,
      "step": 7455
    },
    {
      "entropy": 5.391628360748291,
      "epoch": 0.6231206147677915,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004967214822877763,
      "loss": 5.1099,
      "mean_token_accuracy": 0.19859655797481537,
      "num_tokens": 13726639.0,
      "step": 7460
    },
    {
      "entropy": 5.335668325424194,
      "epoch": 0.6235382559305045,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004967164168195159,
      "loss": 5.0837,
      "mean_token_accuracy": 0.1948293849825859,
      "num_tokens": 13735220.0,
      "step": 7465
    },
    {
      "entropy": 5.186945009231567,
      "epoch": 0.6239558970932175,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004967113474698428,
      "loss": 5.0552,
      "mean_token_accuracy": 0.2078215643763542,
      "num_tokens": 13743277.0,
      "step": 7470
    },
    {
      "entropy": 5.272894716262817,
      "epoch": 0.6243735382559304,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004967062742388456,
      "loss": 5.1097,
      "mean_token_accuracy": 0.1974967047572136,
      "num_tokens": 13752946.0,
      "step": 7475
    },
    {
      "entropy": 5.379592800140381,
      "epoch": 0.6247911794186435,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004967011971266132,
      "loss": 5.1696,
      "mean_token_accuracy": 0.19510784447193147,
      "num_tokens": 13761297.0,
      "step": 7480
    },
    {
      "entropy": 5.342331743240356,
      "epoch": 0.6252088205813565,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004966961161332346,
      "loss": 5.0824,
      "mean_token_accuracy": 0.19345168620347977,
      "num_tokens": 13771238.0,
      "step": 7485
    },
    {
      "entropy": 5.331126356124878,
      "epoch": 0.6256264617440696,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004966910312587985,
      "loss": 5.1344,
      "mean_token_accuracy": 0.1943610966205597,
      "num_tokens": 13781420.0,
      "step": 7490
    },
    {
      "entropy": 5.245034646987915,
      "epoch": 0.6260441029067825,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004966859425033941,
      "loss": 5.1174,
      "mean_token_accuracy": 0.19917237758636475,
      "num_tokens": 13789833.0,
      "step": 7495
    },
    {
      "entropy": 5.368777465820313,
      "epoch": 0.6264617440694955,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004966808498671104,
      "loss": 5.2177,
      "mean_token_accuracy": 0.18571158051490783,
      "num_tokens": 13799202.0,
      "step": 7500
    },
    {
      "entropy": 5.402452230453491,
      "epoch": 0.6268793852322085,
      "grad_norm": 0.875,
      "learning_rate": 0.0004966757533500367,
      "loss": 5.1831,
      "mean_token_accuracy": 0.19444392174482344,
      "num_tokens": 13809097.0,
      "step": 7505
    },
    {
      "entropy": 5.331866836547851,
      "epoch": 0.6272970263949215,
      "grad_norm": 1.015625,
      "learning_rate": 0.000496670652952262,
      "loss": 5.1002,
      "mean_token_accuracy": 0.20072959661483764,
      "num_tokens": 13818437.0,
      "step": 7510
    },
    {
      "entropy": 5.252272510528565,
      "epoch": 0.6277146675576345,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004966655486738757,
      "loss": 5.0958,
      "mean_token_accuracy": 0.1987917646765709,
      "num_tokens": 13827083.0,
      "step": 7515
    },
    {
      "entropy": 5.242339038848877,
      "epoch": 0.6281323087203475,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004966604405149672,
      "loss": 5.0866,
      "mean_token_accuracy": 0.19696566462516785,
      "num_tokens": 13835905.0,
      "step": 7520
    },
    {
      "entropy": 5.300411796569824,
      "epoch": 0.6285499498830605,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004966553284756257,
      "loss": 5.0056,
      "mean_token_accuracy": 0.20877920389175414,
      "num_tokens": 13845136.0,
      "step": 7525
    },
    {
      "entropy": 5.311416721343994,
      "epoch": 0.6289675910457735,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004966502125559411,
      "loss": 5.0876,
      "mean_token_accuracy": 0.1963433012366295,
      "num_tokens": 13855087.0,
      "step": 7530
    },
    {
      "entropy": 5.324881458282471,
      "epoch": 0.6293852322084865,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004966450927560026,
      "loss": 5.1348,
      "mean_token_accuracy": 0.1960173800587654,
      "num_tokens": 13864077.0,
      "step": 7535
    },
    {
      "entropy": 5.3504749774932865,
      "epoch": 0.6298028733711994,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004966399690759,
      "loss": 5.1182,
      "mean_token_accuracy": 0.19554889500141143,
      "num_tokens": 13873404.0,
      "step": 7540
    },
    {
      "entropy": 5.375106430053711,
      "epoch": 0.6302205145339125,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000496634841515723,
      "loss": 5.1987,
      "mean_token_accuracy": 0.19144078940153123,
      "num_tokens": 13883231.0,
      "step": 7545
    },
    {
      "entropy": 5.22035231590271,
      "epoch": 0.6306381556966255,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004966297100755613,
      "loss": 4.9542,
      "mean_token_accuracy": 0.21137010604143142,
      "num_tokens": 13892314.0,
      "step": 7550
    },
    {
      "entropy": 5.267605400085449,
      "epoch": 0.6310557968593384,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004966245747555047,
      "loss": 5.0591,
      "mean_token_accuracy": 0.2032380074262619,
      "num_tokens": 13900771.0,
      "step": 7555
    },
    {
      "entropy": 5.259071922302246,
      "epoch": 0.6314734380220515,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004966194355556432,
      "loss": 5.0344,
      "mean_token_accuracy": 0.19890875667333602,
      "num_tokens": 13909615.0,
      "step": 7560
    },
    {
      "entropy": 5.308586120605469,
      "epoch": 0.6318910791847645,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004966142924760666,
      "loss": 5.1629,
      "mean_token_accuracy": 0.19826842546463014,
      "num_tokens": 13918491.0,
      "step": 7565
    },
    {
      "entropy": 5.28107328414917,
      "epoch": 0.6323087203474774,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004966091455168652,
      "loss": 5.1562,
      "mean_token_accuracy": 0.18997056186199188,
      "num_tokens": 13927935.0,
      "step": 7570
    },
    {
      "entropy": 5.313916635513306,
      "epoch": 0.6327263615101905,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004966039946781289,
      "loss": 4.9787,
      "mean_token_accuracy": 0.20492043644189833,
      "num_tokens": 13936524.0,
      "step": 7575
    },
    {
      "entropy": 5.249733257293701,
      "epoch": 0.6331440026729035,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000496598839959948,
      "loss": 5.0622,
      "mean_token_accuracy": 0.19564977437257766,
      "num_tokens": 13945544.0,
      "step": 7580
    },
    {
      "entropy": 5.29291033744812,
      "epoch": 0.6335616438356164,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004965936813624126,
      "loss": 5.0856,
      "mean_token_accuracy": 0.2052799180150032,
      "num_tokens": 13955016.0,
      "step": 7585
    },
    {
      "entropy": 5.279933929443359,
      "epoch": 0.6339792849983295,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000496588518885613,
      "loss": 5.0935,
      "mean_token_accuracy": 0.19492953419685363,
      "num_tokens": 13964045.0,
      "step": 7590
    },
    {
      "entropy": 5.323713159561157,
      "epoch": 0.6343969261610425,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004965833525296397,
      "loss": 5.0959,
      "mean_token_accuracy": 0.19961652904748917,
      "num_tokens": 13973554.0,
      "step": 7595
    },
    {
      "entropy": 5.31671371459961,
      "epoch": 0.6348145673237554,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004965781822945831,
      "loss": 5.0797,
      "mean_token_accuracy": 0.1947735995054245,
      "num_tokens": 13982669.0,
      "step": 7600
    },
    {
      "entropy": 5.379872894287109,
      "epoch": 0.6352322084864684,
      "grad_norm": 0.875,
      "learning_rate": 0.0004965730081805338,
      "loss": 5.1914,
      "mean_token_accuracy": 0.18872545659542084,
      "num_tokens": 13993392.0,
      "step": 7605
    },
    {
      "entropy": 5.394006824493408,
      "epoch": 0.6356498496491815,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004965678301875821,
      "loss": 5.1946,
      "mean_token_accuracy": 0.18972784876823426,
      "num_tokens": 14002659.0,
      "step": 7610
    },
    {
      "entropy": 5.337057876586914,
      "epoch": 0.6360674908118944,
      "grad_norm": 0.953125,
      "learning_rate": 0.000496562648315819,
      "loss": 5.1375,
      "mean_token_accuracy": 0.19336433112621307,
      "num_tokens": 14011561.0,
      "step": 7615
    },
    {
      "entropy": 5.3313178539276125,
      "epoch": 0.6364851319746074,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000496557462565335,
      "loss": 5.1163,
      "mean_token_accuracy": 0.1960323706269264,
      "num_tokens": 14021286.0,
      "step": 7620
    },
    {
      "entropy": 5.2872144222259525,
      "epoch": 0.6369027731373205,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004965522729362209,
      "loss": 5.117,
      "mean_token_accuracy": 0.1994802787899971,
      "num_tokens": 14031901.0,
      "step": 7625
    },
    {
      "entropy": 5.333450889587402,
      "epoch": 0.6373204143000334,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004965470794285676,
      "loss": 5.1767,
      "mean_token_accuracy": 0.1981829509139061,
      "num_tokens": 14041000.0,
      "step": 7630
    },
    {
      "entropy": 5.379024839401245,
      "epoch": 0.6377380554627464,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000496541882042466,
      "loss": 5.2031,
      "mean_token_accuracy": 0.19270453304052354,
      "num_tokens": 14050567.0,
      "step": 7635
    },
    {
      "entropy": 5.267486810684204,
      "epoch": 0.6381556966254595,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004965366807780071,
      "loss": 5.0374,
      "mean_token_accuracy": 0.19907334297895432,
      "num_tokens": 14059990.0,
      "step": 7640
    },
    {
      "entropy": 5.346701002120971,
      "epoch": 0.6385733377881724,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004965314756352818,
      "loss": 5.1643,
      "mean_token_accuracy": 0.19342345148324966,
      "num_tokens": 14069317.0,
      "step": 7645
    },
    {
      "entropy": 5.3036762237548825,
      "epoch": 0.6389909789508854,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004965262666143816,
      "loss": 5.1213,
      "mean_token_accuracy": 0.2005413517355919,
      "num_tokens": 14078809.0,
      "step": 7650
    },
    {
      "entropy": 5.325835466384888,
      "epoch": 0.6394086201135984,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004965210537153973,
      "loss": 5.148,
      "mean_token_accuracy": 0.19246143698692322,
      "num_tokens": 14087660.0,
      "step": 7655
    },
    {
      "entropy": 5.345402002334595,
      "epoch": 0.6398262612763114,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004965158369384204,
      "loss": 5.1475,
      "mean_token_accuracy": 0.19436953216791153,
      "num_tokens": 14096585.0,
      "step": 7660
    },
    {
      "entropy": 5.323912811279297,
      "epoch": 0.6402439024390244,
      "grad_norm": 0.9375,
      "learning_rate": 0.000496510616283542,
      "loss": 5.0626,
      "mean_token_accuracy": 0.20169996917247773,
      "num_tokens": 14106079.0,
      "step": 7665
    },
    {
      "entropy": 5.324014282226562,
      "epoch": 0.6406615436017374,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004965053917508537,
      "loss": 5.1762,
      "mean_token_accuracy": 0.1915215641260147,
      "num_tokens": 14115988.0,
      "step": 7670
    },
    {
      "entropy": 5.320602321624756,
      "epoch": 0.6410791847644504,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004965001633404469,
      "loss": 5.2354,
      "mean_token_accuracy": 0.18725120425224304,
      "num_tokens": 14126980.0,
      "step": 7675
    },
    {
      "entropy": 5.383184480667114,
      "epoch": 0.6414968259271634,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004964949310524131,
      "loss": 5.1549,
      "mean_token_accuracy": 0.19912902712821962,
      "num_tokens": 14135780.0,
      "step": 7680
    },
    {
      "entropy": 5.391537189483643,
      "epoch": 0.6419144670898764,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004964896948868439,
      "loss": 5.186,
      "mean_token_accuracy": 0.18771100044250488,
      "num_tokens": 14145655.0,
      "step": 7685
    },
    {
      "entropy": 5.330364465713501,
      "epoch": 0.6423321082525894,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004964844548438311,
      "loss": 5.082,
      "mean_token_accuracy": 0.20639275014400482,
      "num_tokens": 14154157.0,
      "step": 7690
    },
    {
      "entropy": 5.38147201538086,
      "epoch": 0.6427497494153024,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004964792109234662,
      "loss": 5.1944,
      "mean_token_accuracy": 0.19954712837934493,
      "num_tokens": 14163749.0,
      "step": 7695
    },
    {
      "entropy": 5.31485161781311,
      "epoch": 0.6431673905780154,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004964739631258411,
      "loss": 5.1169,
      "mean_token_accuracy": 0.19643400907516478,
      "num_tokens": 14174170.0,
      "step": 7700
    },
    {
      "entropy": 5.288280057907104,
      "epoch": 0.6435850317407283,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004964687114510478,
      "loss": 5.1149,
      "mean_token_accuracy": 0.1945824682712555,
      "num_tokens": 14182625.0,
      "step": 7705
    },
    {
      "entropy": 5.331317043304443,
      "epoch": 0.6440026729034414,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000496463455899178,
      "loss": 5.1128,
      "mean_token_accuracy": 0.19239243865013123,
      "num_tokens": 14192351.0,
      "step": 7710
    },
    {
      "entropy": 5.299941301345825,
      "epoch": 0.6444203140661544,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004964581964703239,
      "loss": 5.0365,
      "mean_token_accuracy": 0.2009262338280678,
      "num_tokens": 14201107.0,
      "step": 7715
    },
    {
      "entropy": 5.335964155197144,
      "epoch": 0.6448379552288673,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004964529331645776,
      "loss": 5.0795,
      "mean_token_accuracy": 0.2003314033150673,
      "num_tokens": 14209540.0,
      "step": 7720
    },
    {
      "entropy": 5.2357885360717775,
      "epoch": 0.6452555963915804,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000496447665982031,
      "loss": 4.987,
      "mean_token_accuracy": 0.20245920419692992,
      "num_tokens": 14217944.0,
      "step": 7725
    },
    {
      "entropy": 5.265289688110352,
      "epoch": 0.6456732375542934,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004964423949227765,
      "loss": 5.0456,
      "mean_token_accuracy": 0.20206111073493957,
      "num_tokens": 14226044.0,
      "step": 7730
    },
    {
      "entropy": 5.324347686767578,
      "epoch": 0.6460908787170063,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004964371199869062,
      "loss": 5.1132,
      "mean_token_accuracy": 0.19210731238126755,
      "num_tokens": 14235566.0,
      "step": 7735
    },
    {
      "entropy": 5.307280445098877,
      "epoch": 0.6465085198797194,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004964318411745127,
      "loss": 5.0423,
      "mean_token_accuracy": 0.20769945085048674,
      "num_tokens": 14244511.0,
      "step": 7740
    },
    {
      "entropy": 5.291833972930908,
      "epoch": 0.6469261610424324,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004964265584856883,
      "loss": 5.1242,
      "mean_token_accuracy": 0.1999090939760208,
      "num_tokens": 14254299.0,
      "step": 7745
    },
    {
      "entropy": 5.237125253677368,
      "epoch": 0.6473438022051453,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004964212719205254,
      "loss": 5.0576,
      "mean_token_accuracy": 0.19845819026231765,
      "num_tokens": 14262915.0,
      "step": 7750
    },
    {
      "entropy": 5.358282804489136,
      "epoch": 0.6477614433678583,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004964159814791167,
      "loss": 5.1052,
      "mean_token_accuracy": 0.2015347585082054,
      "num_tokens": 14272606.0,
      "step": 7755
    },
    {
      "entropy": 5.2958807945251465,
      "epoch": 0.6481790845305714,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004964106871615547,
      "loss": 5.1855,
      "mean_token_accuracy": 0.18664002418518066,
      "num_tokens": 14282117.0,
      "step": 7760
    },
    {
      "entropy": 5.207527017593383,
      "epoch": 0.6485967256932843,
      "grad_norm": 0.96875,
      "learning_rate": 0.000496405388967932,
      "loss": 5.0992,
      "mean_token_accuracy": 0.206837897002697,
      "num_tokens": 14290421.0,
      "step": 7765
    },
    {
      "entropy": 5.405949640274048,
      "epoch": 0.6490143668559973,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004964000868983416,
      "loss": 5.1998,
      "mean_token_accuracy": 0.19390546083450316,
      "num_tokens": 14299870.0,
      "step": 7770
    },
    {
      "entropy": 5.4170910835266115,
      "epoch": 0.6494320080187104,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004963947809528761,
      "loss": 5.118,
      "mean_token_accuracy": 0.19223859757184983,
      "num_tokens": 14309916.0,
      "step": 7775
    },
    {
      "entropy": 5.299497175216675,
      "epoch": 0.6498496491814233,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004963894711316285,
      "loss": 5.0992,
      "mean_token_accuracy": 0.1954879119992256,
      "num_tokens": 14319061.0,
      "step": 7780
    },
    {
      "entropy": 5.299534845352173,
      "epoch": 0.6502672903441363,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004963841574346917,
      "loss": 5.0685,
      "mean_token_accuracy": 0.2032836467027664,
      "num_tokens": 14327241.0,
      "step": 7785
    },
    {
      "entropy": 5.359089422225952,
      "epoch": 0.6506849315068494,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004963788398621587,
      "loss": 5.1771,
      "mean_token_accuracy": 0.1930241197347641,
      "num_tokens": 14336586.0,
      "step": 7790
    },
    {
      "entropy": 5.347798013687134,
      "epoch": 0.6511025726695623,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004963735184141227,
      "loss": 5.1484,
      "mean_token_accuracy": 0.18727537989616394,
      "num_tokens": 14346484.0,
      "step": 7795
    },
    {
      "entropy": 5.429263019561768,
      "epoch": 0.6515202138322753,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004963681930906768,
      "loss": 5.2144,
      "mean_token_accuracy": 0.18842751234769822,
      "num_tokens": 14356255.0,
      "step": 7800
    },
    {
      "entropy": 5.319133520126343,
      "epoch": 0.6519378549949884,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004963628638919141,
      "loss": 5.1427,
      "mean_token_accuracy": 0.19943154454231263,
      "num_tokens": 14365680.0,
      "step": 7805
    },
    {
      "entropy": 5.322173261642456,
      "epoch": 0.6523554961577013,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004963575308179281,
      "loss": 5.1345,
      "mean_token_accuracy": 0.19807028770446777,
      "num_tokens": 14374546.0,
      "step": 7810
    },
    {
      "entropy": 5.376398181915283,
      "epoch": 0.6527731373204143,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004963521938688122,
      "loss": 5.135,
      "mean_token_accuracy": 0.1999364286661148,
      "num_tokens": 14383624.0,
      "step": 7815
    },
    {
      "entropy": 5.370496511459351,
      "epoch": 0.6531907784831273,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004963468530446596,
      "loss": 5.2346,
      "mean_token_accuracy": 0.18958115577697754,
      "num_tokens": 14392482.0,
      "step": 7820
    },
    {
      "entropy": 5.356835842132568,
      "epoch": 0.6536084196458403,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004963415083455639,
      "loss": 5.1306,
      "mean_token_accuracy": 0.19529043287038803,
      "num_tokens": 14401940.0,
      "step": 7825
    },
    {
      "entropy": 5.340043735504151,
      "epoch": 0.6540260608085533,
      "grad_norm": 1.0,
      "learning_rate": 0.0004963361597716186,
      "loss": 5.1837,
      "mean_token_accuracy": 0.19353048950433732,
      "num_tokens": 14411127.0,
      "step": 7830
    },
    {
      "entropy": 5.279142284393311,
      "epoch": 0.6544437019712663,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004963308073229175,
      "loss": 5.0855,
      "mean_token_accuracy": 0.19947728514671326,
      "num_tokens": 14421358.0,
      "step": 7835
    },
    {
      "entropy": 5.331194972991943,
      "epoch": 0.6548613431339793,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004963254509995542,
      "loss": 5.1287,
      "mean_token_accuracy": 0.19109020084142686,
      "num_tokens": 14430507.0,
      "step": 7840
    },
    {
      "entropy": 5.320723915100098,
      "epoch": 0.6552789842966923,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004963200908016224,
      "loss": 5.1107,
      "mean_token_accuracy": 0.1978656381368637,
      "num_tokens": 14439723.0,
      "step": 7845
    },
    {
      "entropy": 5.314566659927368,
      "epoch": 0.6556966254594053,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000496314726729216,
      "loss": 5.0814,
      "mean_token_accuracy": 0.19859781563282014,
      "num_tokens": 14449011.0,
      "step": 7850
    },
    {
      "entropy": 5.274481296539307,
      "epoch": 0.6561142666221182,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004963093587824289,
      "loss": 5.067,
      "mean_token_accuracy": 0.2032626077532768,
      "num_tokens": 14457676.0,
      "step": 7855
    },
    {
      "entropy": 5.348059892654419,
      "epoch": 0.6565319077848313,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004963039869613551,
      "loss": 5.1418,
      "mean_token_accuracy": 0.19519722759723662,
      "num_tokens": 14466956.0,
      "step": 7860
    },
    {
      "entropy": 5.308815240859985,
      "epoch": 0.6569495489475443,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004962986112660887,
      "loss": 5.0388,
      "mean_token_accuracy": 0.20452324002981187,
      "num_tokens": 14476083.0,
      "step": 7865
    },
    {
      "entropy": 5.364091110229492,
      "epoch": 0.6573671901102572,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004962932316967237,
      "loss": 5.1464,
      "mean_token_accuracy": 0.19049460887908937,
      "num_tokens": 14484980.0,
      "step": 7870
    },
    {
      "entropy": 5.3212967872619625,
      "epoch": 0.6577848312729703,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004962878482533544,
      "loss": 5.1534,
      "mean_token_accuracy": 0.19653844237327575,
      "num_tokens": 14493965.0,
      "step": 7875
    },
    {
      "entropy": 5.3813779830932615,
      "epoch": 0.6582024724356833,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004962824609360748,
      "loss": 5.1395,
      "mean_token_accuracy": 0.1948006272315979,
      "num_tokens": 14501914.0,
      "step": 7880
    },
    {
      "entropy": 5.272090148925781,
      "epoch": 0.6586201135983962,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004962770697449794,
      "loss": 5.061,
      "mean_token_accuracy": 0.19928629398345948,
      "num_tokens": 14511447.0,
      "step": 7885
    },
    {
      "entropy": 5.240921974182129,
      "epoch": 0.6590377547611093,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004962716746801625,
      "loss": 5.0689,
      "mean_token_accuracy": 0.1970309719443321,
      "num_tokens": 14520558.0,
      "step": 7890
    },
    {
      "entropy": 5.281534624099732,
      "epoch": 0.6594553959238223,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004962662757417187,
      "loss": 5.0291,
      "mean_token_accuracy": 0.20256045758724212,
      "num_tokens": 14529995.0,
      "step": 7895
    },
    {
      "entropy": 5.307058525085449,
      "epoch": 0.6598730370865352,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004962608729297424,
      "loss": 5.184,
      "mean_token_accuracy": 0.18883098512887955,
      "num_tokens": 14538972.0,
      "step": 7900
    },
    {
      "entropy": 5.366240787506103,
      "epoch": 0.6602906782492483,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004962554662443281,
      "loss": 5.1132,
      "mean_token_accuracy": 0.1963333308696747,
      "num_tokens": 14548255.0,
      "step": 7905
    },
    {
      "entropy": 5.345869350433349,
      "epoch": 0.6607083194119613,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004962500556855706,
      "loss": 5.09,
      "mean_token_accuracy": 0.19700034856796264,
      "num_tokens": 14557265.0,
      "step": 7910
    },
    {
      "entropy": 5.401500225067139,
      "epoch": 0.6611259605746742,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004962446412535645,
      "loss": 5.1466,
      "mean_token_accuracy": 0.195510932803154,
      "num_tokens": 14566650.0,
      "step": 7915
    },
    {
      "entropy": 5.293809843063355,
      "epoch": 0.6615436017373872,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004962392229484046,
      "loss": 5.0393,
      "mean_token_accuracy": 0.20715844333171846,
      "num_tokens": 14575550.0,
      "step": 7920
    },
    {
      "entropy": 5.308446884155273,
      "epoch": 0.6619612429001003,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000496233800770186,
      "loss": 5.061,
      "mean_token_accuracy": 0.1985615849494934,
      "num_tokens": 14584459.0,
      "step": 7925
    },
    {
      "entropy": 5.387534332275391,
      "epoch": 0.6623788840628132,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004962283747190033,
      "loss": 5.1887,
      "mean_token_accuracy": 0.19417430758476256,
      "num_tokens": 14593709.0,
      "step": 7930
    },
    {
      "entropy": 5.2846475601196286,
      "epoch": 0.6627965252255262,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004962229447949515,
      "loss": 5.0428,
      "mean_token_accuracy": 0.1996523082256317,
      "num_tokens": 14602071.0,
      "step": 7935
    },
    {
      "entropy": 5.372671747207642,
      "epoch": 0.6632141663882393,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004962175109981259,
      "loss": 5.2266,
      "mean_token_accuracy": 0.19012019485235215,
      "num_tokens": 14612842.0,
      "step": 7940
    },
    {
      "entropy": 5.449300622940063,
      "epoch": 0.6636318075509522,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004962120733286215,
      "loss": 5.2345,
      "mean_token_accuracy": 0.191847363114357,
      "num_tokens": 14621645.0,
      "step": 7945
    },
    {
      "entropy": 5.263379764556885,
      "epoch": 0.6640494487136652,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004962066317865334,
      "loss": 5.0537,
      "mean_token_accuracy": 0.19839805811643602,
      "num_tokens": 14630447.0,
      "step": 7950
    },
    {
      "entropy": 5.2608662128448485,
      "epoch": 0.6644670898763783,
      "grad_norm": 1.0,
      "learning_rate": 0.000496201186371957,
      "loss": 5.1374,
      "mean_token_accuracy": 0.20101145058870315,
      "num_tokens": 14639371.0,
      "step": 7955
    },
    {
      "entropy": 5.324439620971679,
      "epoch": 0.6648847310390912,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004961957370849875,
      "loss": 5.0398,
      "mean_token_accuracy": 0.19604826122522354,
      "num_tokens": 14648540.0,
      "step": 7960
    },
    {
      "entropy": 5.353636455535889,
      "epoch": 0.6653023722018042,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004961902839257205,
      "loss": 5.2109,
      "mean_token_accuracy": 0.19647469222545624,
      "num_tokens": 14658130.0,
      "step": 7965
    },
    {
      "entropy": 5.384181880950928,
      "epoch": 0.6657200133645172,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004961848268942513,
      "loss": 5.1397,
      "mean_token_accuracy": 0.19351573884487153,
      "num_tokens": 14667538.0,
      "step": 7970
    },
    {
      "entropy": 5.350904273986816,
      "epoch": 0.6661376545272302,
      "grad_norm": 1.0,
      "learning_rate": 0.0004961793659906754,
      "loss": 5.1324,
      "mean_token_accuracy": 0.1939302459359169,
      "num_tokens": 14677115.0,
      "step": 7975
    },
    {
      "entropy": 5.347810077667236,
      "epoch": 0.6665552956899432,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004961739012150885,
      "loss": 5.0899,
      "mean_token_accuracy": 0.20003176778554915,
      "num_tokens": 14686444.0,
      "step": 7980
    },
    {
      "entropy": 5.308733081817627,
      "epoch": 0.6669729368526562,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004961684325675862,
      "loss": 5.0727,
      "mean_token_accuracy": 0.20306781977415084,
      "num_tokens": 14695951.0,
      "step": 7985
    },
    {
      "entropy": 5.293811941146851,
      "epoch": 0.6673905780153692,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004961629600482643,
      "loss": 5.1855,
      "mean_token_accuracy": 0.19224019944667817,
      "num_tokens": 14705415.0,
      "step": 7990
    },
    {
      "entropy": 5.397787237167359,
      "epoch": 0.6678082191780822,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004961574836572186,
      "loss": 5.1532,
      "mean_token_accuracy": 0.19240250289440156,
      "num_tokens": 14714793.0,
      "step": 7995
    },
    {
      "entropy": 5.242776250839233,
      "epoch": 0.6682258603407952,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004961520033945448,
      "loss": 4.9364,
      "mean_token_accuracy": 0.21116291880607604,
      "num_tokens": 14723820.0,
      "step": 8000
    },
    {
      "entropy": 5.274733686447144,
      "epoch": 0.6686435015035082,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004961465192603391,
      "loss": 5.0622,
      "mean_token_accuracy": 0.20018671303987504,
      "num_tokens": 14733055.0,
      "step": 8005
    },
    {
      "entropy": 5.335849189758301,
      "epoch": 0.6690611426662212,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004961410312546974,
      "loss": 5.0907,
      "mean_token_accuracy": 0.19245745688676835,
      "num_tokens": 14742416.0,
      "step": 8010
    },
    {
      "entropy": 5.267804050445557,
      "epoch": 0.6694787838289342,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004961355393777158,
      "loss": 5.0871,
      "mean_token_accuracy": 0.1978096544742584,
      "num_tokens": 14751384.0,
      "step": 8015
    },
    {
      "entropy": 5.276005458831787,
      "epoch": 0.6698964249916471,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004961300436294902,
      "loss": 4.9658,
      "mean_token_accuracy": 0.20420251935720443,
      "num_tokens": 14760525.0,
      "step": 8020
    },
    {
      "entropy": 5.3277098655700685,
      "epoch": 0.6703140661543602,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004961245440101171,
      "loss": 5.1424,
      "mean_token_accuracy": 0.1938997358083725,
      "num_tokens": 14769405.0,
      "step": 8025
    },
    {
      "entropy": 5.219655227661133,
      "epoch": 0.6707317073170732,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004961190405196928,
      "loss": 5.0174,
      "mean_token_accuracy": 0.1983474761247635,
      "num_tokens": 14778945.0,
      "step": 8030
    },
    {
      "entropy": 5.349035787582397,
      "epoch": 0.6711493484797861,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004961135331583135,
      "loss": 5.181,
      "mean_token_accuracy": 0.19320068359375,
      "num_tokens": 14788289.0,
      "step": 8035
    },
    {
      "entropy": 5.296178388595581,
      "epoch": 0.6715669896424992,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004961080219260755,
      "loss": 5.1268,
      "mean_token_accuracy": 0.19143992960453032,
      "num_tokens": 14797761.0,
      "step": 8040
    },
    {
      "entropy": 5.376413583755493,
      "epoch": 0.6719846308052122,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004961025068230757,
      "loss": 5.1467,
      "mean_token_accuracy": 0.19840621203184128,
      "num_tokens": 14806424.0,
      "step": 8045
    },
    {
      "entropy": 5.297316551208496,
      "epoch": 0.6724022719679251,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004960969878494103,
      "loss": 5.0865,
      "mean_token_accuracy": 0.19157626926898957,
      "num_tokens": 14815816.0,
      "step": 8050
    },
    {
      "entropy": 5.254986429214478,
      "epoch": 0.6728199131306382,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004960914650051758,
      "loss": 5.0914,
      "mean_token_accuracy": 0.19978946447372437,
      "num_tokens": 14824877.0,
      "step": 8055
    },
    {
      "entropy": 5.179501724243164,
      "epoch": 0.6732375542933512,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004960859382904691,
      "loss": 4.9977,
      "mean_token_accuracy": 0.2098226800560951,
      "num_tokens": 14833095.0,
      "step": 8060
    },
    {
      "entropy": 5.360788774490357,
      "epoch": 0.6736551954560641,
      "grad_norm": 0.90625,
      "learning_rate": 0.000496080407705387,
      "loss": 5.1312,
      "mean_token_accuracy": 0.19482934921979905,
      "num_tokens": 14842963.0,
      "step": 8065
    },
    {
      "entropy": 5.403423738479614,
      "epoch": 0.6740728366187771,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004960748732500263,
      "loss": 5.1975,
      "mean_token_accuracy": 0.19419121444225312,
      "num_tokens": 14851552.0,
      "step": 8070
    },
    {
      "entropy": 5.362990474700927,
      "epoch": 0.6744904777814902,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004960693349244838,
      "loss": 5.1094,
      "mean_token_accuracy": 0.1957293763756752,
      "num_tokens": 14859997.0,
      "step": 8075
    },
    {
      "entropy": 5.304352331161499,
      "epoch": 0.6749081189442031,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004960637927288566,
      "loss": 5.0228,
      "mean_token_accuracy": 0.20344861447811127,
      "num_tokens": 14868050.0,
      "step": 8080
    },
    {
      "entropy": 5.316249704360962,
      "epoch": 0.6753257601069161,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004960582466632414,
      "loss": 5.1474,
      "mean_token_accuracy": 0.18681503534317018,
      "num_tokens": 14877774.0,
      "step": 8085
    },
    {
      "entropy": 5.267063331604004,
      "epoch": 0.6757434012696292,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004960526967277356,
      "loss": 5.0683,
      "mean_token_accuracy": 0.20365924537181854,
      "num_tokens": 14886971.0,
      "step": 8090
    },
    {
      "entropy": 5.245394468307495,
      "epoch": 0.6761610424323421,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004960471429224363,
      "loss": 5.0194,
      "mean_token_accuracy": 0.1948849618434906,
      "num_tokens": 14895815.0,
      "step": 8095
    },
    {
      "entropy": 5.261016416549682,
      "epoch": 0.6765786835950551,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004960415852474406,
      "loss": 5.0472,
      "mean_token_accuracy": 0.2043766662478447,
      "num_tokens": 14904082.0,
      "step": 8100
    },
    {
      "entropy": 5.259164094924927,
      "epoch": 0.6769963247577682,
      "grad_norm": 0.921875,
      "learning_rate": 0.000496036023702846,
      "loss": 5.065,
      "mean_token_accuracy": 0.20216153860092162,
      "num_tokens": 14913533.0,
      "step": 8105
    },
    {
      "entropy": 5.269275569915772,
      "epoch": 0.6774139659204811,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004960304582887497,
      "loss": 5.0401,
      "mean_token_accuracy": 0.20138293504714966,
      "num_tokens": 14922063.0,
      "step": 8110
    },
    {
      "entropy": 5.308571529388428,
      "epoch": 0.6778316070831941,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004960248890052493,
      "loss": 5.0703,
      "mean_token_accuracy": 0.20289359390735626,
      "num_tokens": 14930219.0,
      "step": 8115
    },
    {
      "entropy": 5.3616663932800295,
      "epoch": 0.6782492482459072,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000496019315852442,
      "loss": 5.1796,
      "mean_token_accuracy": 0.19818756729364395,
      "num_tokens": 14939463.0,
      "step": 8120
    },
    {
      "entropy": 5.390101194381714,
      "epoch": 0.6786668894086201,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004960137388304255,
      "loss": 5.0749,
      "mean_token_accuracy": 0.19911114275455474,
      "num_tokens": 14948289.0,
      "step": 8125
    },
    {
      "entropy": 5.344555854797363,
      "epoch": 0.6790845305713331,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004960081579392975,
      "loss": 5.1728,
      "mean_token_accuracy": 0.1947576329112053,
      "num_tokens": 14958172.0,
      "step": 8130
    },
    {
      "entropy": 5.278376579284668,
      "epoch": 0.6795021717340461,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004960025731791558,
      "loss": 4.9567,
      "mean_token_accuracy": 0.20160674750804902,
      "num_tokens": 14967362.0,
      "step": 8135
    },
    {
      "entropy": 5.26883454322815,
      "epoch": 0.6799198128967591,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004959969845500979,
      "loss": 5.169,
      "mean_token_accuracy": 0.19465858191251756,
      "num_tokens": 14975942.0,
      "step": 8140
    },
    {
      "entropy": 5.275253057479858,
      "epoch": 0.6803374540594721,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004959913920522218,
      "loss": 5.0908,
      "mean_token_accuracy": 0.205180062353611,
      "num_tokens": 14985428.0,
      "step": 8145
    },
    {
      "entropy": 5.330047655105591,
      "epoch": 0.6807550952221851,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004959857956856254,
      "loss": 5.0311,
      "mean_token_accuracy": 0.19949153065681458,
      "num_tokens": 14994558.0,
      "step": 8150
    },
    {
      "entropy": 5.326516103744507,
      "epoch": 0.6811727363848981,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004959801954504067,
      "loss": 5.0493,
      "mean_token_accuracy": 0.20058966875076295,
      "num_tokens": 15003674.0,
      "step": 8155
    },
    {
      "entropy": 5.321355533599854,
      "epoch": 0.6815903775476111,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004959745913466635,
      "loss": 5.1133,
      "mean_token_accuracy": 0.1988556742668152,
      "num_tokens": 15012569.0,
      "step": 8160
    },
    {
      "entropy": 5.302389669418335,
      "epoch": 0.6820080187103241,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004959689833744942,
      "loss": 5.1222,
      "mean_token_accuracy": 0.19943516999483107,
      "num_tokens": 15021863.0,
      "step": 8165
    },
    {
      "entropy": 5.312733459472656,
      "epoch": 0.682425659873037,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004959633715339969,
      "loss": 5.1633,
      "mean_token_accuracy": 0.19152278006076812,
      "num_tokens": 15032272.0,
      "step": 8170
    },
    {
      "entropy": 5.328484678268433,
      "epoch": 0.6828433010357501,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004959577558252698,
      "loss": 5.0972,
      "mean_token_accuracy": 0.1956509009003639,
      "num_tokens": 15041066.0,
      "step": 8175
    },
    {
      "entropy": 5.4315088272094725,
      "epoch": 0.6832609421984631,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004959521362484112,
      "loss": 5.1809,
      "mean_token_accuracy": 0.19160880297422409,
      "num_tokens": 15050855.0,
      "step": 8180
    },
    {
      "entropy": 5.281504726409912,
      "epoch": 0.683678583361176,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004959465128035195,
      "loss": 5.0711,
      "mean_token_accuracy": 0.20465423315763473,
      "num_tokens": 15060810.0,
      "step": 8185
    },
    {
      "entropy": 5.3085390567779545,
      "epoch": 0.6840962245238891,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004959408854906932,
      "loss": 5.0849,
      "mean_token_accuracy": 0.19649745970964433,
      "num_tokens": 15069761.0,
      "step": 8190
    },
    {
      "entropy": 5.277175807952881,
      "epoch": 0.6845138656866021,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004959352543100306,
      "loss": 5.024,
      "mean_token_accuracy": 0.2015466645359993,
      "num_tokens": 15078691.0,
      "step": 8195
    },
    {
      "entropy": 5.28966646194458,
      "epoch": 0.684931506849315,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004959296192616305,
      "loss": 5.0283,
      "mean_token_accuracy": 0.20287408381700517,
      "num_tokens": 15087068.0,
      "step": 8200
    },
    {
      "entropy": 5.237844657897949,
      "epoch": 0.6853491480120281,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004959239803455916,
      "loss": 4.9343,
      "mean_token_accuracy": 0.2042916163802147,
      "num_tokens": 15095962.0,
      "step": 8205
    },
    {
      "entropy": 5.310767030715942,
      "epoch": 0.6857667891747411,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004959183375620123,
      "loss": 5.0925,
      "mean_token_accuracy": 0.18981428891420365,
      "num_tokens": 15105835.0,
      "step": 8210
    },
    {
      "entropy": 5.308083820343017,
      "epoch": 0.686184430337454,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004959126909109918,
      "loss": 5.0228,
      "mean_token_accuracy": 0.21236408054828643,
      "num_tokens": 15114140.0,
      "step": 8215
    },
    {
      "entropy": 5.302402210235596,
      "epoch": 0.686602071500167,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004959070403926286,
      "loss": 5.1201,
      "mean_token_accuracy": 0.19696976393461227,
      "num_tokens": 15123253.0,
      "step": 8220
    },
    {
      "entropy": 5.374145364761352,
      "epoch": 0.6870197126628801,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004959013860070217,
      "loss": 5.212,
      "mean_token_accuracy": 0.1917181760072708,
      "num_tokens": 15133699.0,
      "step": 8225
    },
    {
      "entropy": 5.31294264793396,
      "epoch": 0.687437353825593,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004958957277542701,
      "loss": 4.9948,
      "mean_token_accuracy": 0.20174918919801713,
      "num_tokens": 15142133.0,
      "step": 8230
    },
    {
      "entropy": 5.277373600006103,
      "epoch": 0.687854994988306,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004958900656344731,
      "loss": 5.0734,
      "mean_token_accuracy": 0.19658456593751908,
      "num_tokens": 15151367.0,
      "step": 8235
    },
    {
      "entropy": 5.343179035186767,
      "epoch": 0.6882726361510191,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004958843996477296,
      "loss": 5.0849,
      "mean_token_accuracy": 0.1937602087855339,
      "num_tokens": 15160938.0,
      "step": 8240
    },
    {
      "entropy": 5.359931182861328,
      "epoch": 0.688690277313732,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004958787297941386,
      "loss": 5.1358,
      "mean_token_accuracy": 0.19265848398208618,
      "num_tokens": 15170507.0,
      "step": 8245
    },
    {
      "entropy": 5.372533082962036,
      "epoch": 0.689107918476445,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004958730560737998,
      "loss": 5.1464,
      "mean_token_accuracy": 0.19587162882089615,
      "num_tokens": 15179822.0,
      "step": 8250
    },
    {
      "entropy": 5.282538318634034,
      "epoch": 0.6895255596391581,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004958673784868121,
      "loss": 5.0461,
      "mean_token_accuracy": 0.19758722186088562,
      "num_tokens": 15189353.0,
      "step": 8255
    },
    {
      "entropy": 5.3627111434936525,
      "epoch": 0.689943200801871,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004958616970332752,
      "loss": 5.1333,
      "mean_token_accuracy": 0.19421181082725525,
      "num_tokens": 15197772.0,
      "step": 8260
    },
    {
      "entropy": 5.330782890319824,
      "epoch": 0.690360841964584,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004958560117132884,
      "loss": 5.0792,
      "mean_token_accuracy": 0.20480501353740693,
      "num_tokens": 15207453.0,
      "step": 8265
    },
    {
      "entropy": 5.28204312324524,
      "epoch": 0.6907784831272971,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004958503225269513,
      "loss": 5.0013,
      "mean_token_accuracy": 0.20483452081680298,
      "num_tokens": 15215787.0,
      "step": 8270
    },
    {
      "entropy": 5.369385194778443,
      "epoch": 0.69119612429001,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004958446294743633,
      "loss": 5.122,
      "mean_token_accuracy": 0.20075528919696808,
      "num_tokens": 15225148.0,
      "step": 8275
    },
    {
      "entropy": 5.354285383224488,
      "epoch": 0.691613765452723,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004958389325556245,
      "loss": 5.1365,
      "mean_token_accuracy": 0.19636547565460205,
      "num_tokens": 15234533.0,
      "step": 8280
    },
    {
      "entropy": 5.308356809616089,
      "epoch": 0.692031406615436,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004958332317708341,
      "loss": 5.1112,
      "mean_token_accuracy": 0.18572145104408264,
      "num_tokens": 15243805.0,
      "step": 8285
    },
    {
      "entropy": 5.361861038208008,
      "epoch": 0.692449047778149,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004958275271200924,
      "loss": 5.1717,
      "mean_token_accuracy": 0.19840679466724395,
      "num_tokens": 15253298.0,
      "step": 8290
    },
    {
      "entropy": 5.276644897460938,
      "epoch": 0.692866688940862,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004958218186034988,
      "loss": 5.0946,
      "mean_token_accuracy": 0.20560214072465896,
      "num_tokens": 15261860.0,
      "step": 8295
    },
    {
      "entropy": 5.278173732757568,
      "epoch": 0.693284330103575,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004958161062211537,
      "loss": 5.0442,
      "mean_token_accuracy": 0.2027243822813034,
      "num_tokens": 15271021.0,
      "step": 8300
    },
    {
      "entropy": 5.268083238601685,
      "epoch": 0.693701971266288,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004958103899731566,
      "loss": 5.0707,
      "mean_token_accuracy": 0.19582585394382476,
      "num_tokens": 15280489.0,
      "step": 8305
    },
    {
      "entropy": 5.346872711181641,
      "epoch": 0.694119612429001,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004958046698596081,
      "loss": 5.097,
      "mean_token_accuracy": 0.2035209819674492,
      "num_tokens": 15289387.0,
      "step": 8310
    },
    {
      "entropy": 5.368372869491577,
      "epoch": 0.694537253591714,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004957989458806079,
      "loss": 5.1672,
      "mean_token_accuracy": 0.19401417523622513,
      "num_tokens": 15299334.0,
      "step": 8315
    },
    {
      "entropy": 5.321342134475708,
      "epoch": 0.694954894754427,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004957932180362564,
      "loss": 5.113,
      "mean_token_accuracy": 0.19733107089996338,
      "num_tokens": 15309174.0,
      "step": 8320
    },
    {
      "entropy": 5.340223217010498,
      "epoch": 0.69537253591714,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004957874863266538,
      "loss": 5.1728,
      "mean_token_accuracy": 0.19896778464317322,
      "num_tokens": 15318722.0,
      "step": 8325
    },
    {
      "entropy": 5.386840677261352,
      "epoch": 0.695790177079853,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004957817507519006,
      "loss": 5.1134,
      "mean_token_accuracy": 0.19858064949512483,
      "num_tokens": 15327606.0,
      "step": 8330
    },
    {
      "entropy": 5.256264734268188,
      "epoch": 0.696207818242566,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004957760113120971,
      "loss": 5.0889,
      "mean_token_accuracy": 0.20408816635608673,
      "num_tokens": 15336520.0,
      "step": 8335
    },
    {
      "entropy": 5.374856328964233,
      "epoch": 0.696625459405279,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004957702680073437,
      "loss": 5.1314,
      "mean_token_accuracy": 0.19796695411205292,
      "num_tokens": 15345437.0,
      "step": 8340
    },
    {
      "entropy": 5.364317846298218,
      "epoch": 0.697043100567992,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004957645208377411,
      "loss": 5.0885,
      "mean_token_accuracy": 0.19871389120817184,
      "num_tokens": 15354660.0,
      "step": 8345
    },
    {
      "entropy": 5.33889594078064,
      "epoch": 0.6974607417307049,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004957587698033898,
      "loss": 5.1129,
      "mean_token_accuracy": 0.20253059566020964,
      "num_tokens": 15363958.0,
      "step": 8350
    },
    {
      "entropy": 5.267718172073364,
      "epoch": 0.697878382893418,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004957530149043905,
      "loss": 5.0592,
      "mean_token_accuracy": 0.20933472514152526,
      "num_tokens": 15373351.0,
      "step": 8355
    },
    {
      "entropy": 5.344996595382691,
      "epoch": 0.698296024056131,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004957472561408438,
      "loss": 5.0778,
      "mean_token_accuracy": 0.19937078654766083,
      "num_tokens": 15383069.0,
      "step": 8360
    },
    {
      "entropy": 5.2971538543701175,
      "epoch": 0.6987136652188439,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004957414935128509,
      "loss": 5.0805,
      "mean_token_accuracy": 0.20104648619890214,
      "num_tokens": 15391394.0,
      "step": 8365
    },
    {
      "entropy": 5.328254127502442,
      "epoch": 0.699131306381557,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004957357270205123,
      "loss": 5.059,
      "mean_token_accuracy": 0.1952521026134491,
      "num_tokens": 15399993.0,
      "step": 8370
    },
    {
      "entropy": 5.257268571853638,
      "epoch": 0.69954894754427,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004957299566639292,
      "loss": 5.0719,
      "mean_token_accuracy": 0.19981346428394317,
      "num_tokens": 15409886.0,
      "step": 8375
    },
    {
      "entropy": 5.334570360183716,
      "epoch": 0.6999665887069829,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004957241824432025,
      "loss": 5.0545,
      "mean_token_accuracy": 0.205235655605793,
      "num_tokens": 15419772.0,
      "step": 8380
    },
    {
      "entropy": 5.339207744598388,
      "epoch": 0.700384229869696,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004957184043584333,
      "loss": 5.0743,
      "mean_token_accuracy": 0.1953941121697426,
      "num_tokens": 15429631.0,
      "step": 8385
    },
    {
      "entropy": 5.251938581466675,
      "epoch": 0.700801871032409,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004957126224097227,
      "loss": 5.0145,
      "mean_token_accuracy": 0.20826880186796187,
      "num_tokens": 15438411.0,
      "step": 8390
    },
    {
      "entropy": 5.18582706451416,
      "epoch": 0.7012195121951219,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000495706836597172,
      "loss": 5.0455,
      "mean_token_accuracy": 0.2070013701915741,
      "num_tokens": 15448943.0,
      "step": 8395
    },
    {
      "entropy": 5.272938394546509,
      "epoch": 0.7016371533578349,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004957010469208825,
      "loss": 5.0124,
      "mean_token_accuracy": 0.20000020712614058,
      "num_tokens": 15457908.0,
      "step": 8400
    },
    {
      "entropy": 5.365331935882568,
      "epoch": 0.702054794520548,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004956952533809554,
      "loss": 5.0891,
      "mean_token_accuracy": 0.19937152564525604,
      "num_tokens": 15467390.0,
      "step": 8405
    },
    {
      "entropy": 5.309229040145874,
      "epoch": 0.7024724356832609,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004956894559774923,
      "loss": 5.1005,
      "mean_token_accuracy": 0.1978869378566742,
      "num_tokens": 15476419.0,
      "step": 8410
    },
    {
      "entropy": 5.221187162399292,
      "epoch": 0.7028900768459739,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004956836547105947,
      "loss": 4.9641,
      "mean_token_accuracy": 0.21074582785367965,
      "num_tokens": 15485147.0,
      "step": 8415
    },
    {
      "entropy": 5.359545421600342,
      "epoch": 0.703307718008687,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004956778495803641,
      "loss": 5.0709,
      "mean_token_accuracy": 0.19837865829467774,
      "num_tokens": 15493816.0,
      "step": 8420
    },
    {
      "entropy": 5.3577385425567625,
      "epoch": 0.7037253591713999,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000495672040586902,
      "loss": 5.1845,
      "mean_token_accuracy": 0.19680008590221404,
      "num_tokens": 15503051.0,
      "step": 8425
    },
    {
      "entropy": 5.280332088470459,
      "epoch": 0.7041430003341129,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004956662277303103,
      "loss": 4.9888,
      "mean_token_accuracy": 0.20268148481845855,
      "num_tokens": 15511854.0,
      "step": 8430
    },
    {
      "entropy": 5.245299053192139,
      "epoch": 0.704560641496826,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004956604110106905,
      "loss": 5.1054,
      "mean_token_accuracy": 0.20257826447486876,
      "num_tokens": 15521492.0,
      "step": 8435
    },
    {
      "entropy": 5.3756246089935305,
      "epoch": 0.7049782826595389,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004956545904281447,
      "loss": 5.1866,
      "mean_token_accuracy": 0.18934756964445115,
      "num_tokens": 15530983.0,
      "step": 8440
    },
    {
      "entropy": 5.297769927978516,
      "epoch": 0.7053959238222519,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004956487659827748,
      "loss": 5.0242,
      "mean_token_accuracy": 0.20594093203544617,
      "num_tokens": 15539722.0,
      "step": 8445
    },
    {
      "entropy": 5.25525426864624,
      "epoch": 0.705813564984965,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004956429376746825,
      "loss": 5.0058,
      "mean_token_accuracy": 0.2023167446255684,
      "num_tokens": 15548738.0,
      "step": 8450
    },
    {
      "entropy": 5.225326871871948,
      "epoch": 0.7062312061476779,
      "grad_norm": 0.875,
      "learning_rate": 0.0004956371055039701,
      "loss": 5.0565,
      "mean_token_accuracy": 0.20497656464576722,
      "num_tokens": 15557520.0,
      "step": 8455
    },
    {
      "entropy": 5.366145086288452,
      "epoch": 0.7066488473103909,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004956312694707395,
      "loss": 5.1175,
      "mean_token_accuracy": 0.19009262919425965,
      "num_tokens": 15567860.0,
      "step": 8460
    },
    {
      "entropy": 5.378082370758056,
      "epoch": 0.7070664884731039,
      "grad_norm": 0.9375,
      "learning_rate": 0.000495625429575093,
      "loss": 5.1304,
      "mean_token_accuracy": 0.190438936650753,
      "num_tokens": 15577174.0,
      "step": 8465
    },
    {
      "entropy": 5.209462070465088,
      "epoch": 0.7074841296358169,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004956195858171327,
      "loss": 4.9258,
      "mean_token_accuracy": 0.21043189764022827,
      "num_tokens": 15585476.0,
      "step": 8470
    },
    {
      "entropy": 5.229205369949341,
      "epoch": 0.7079017707985299,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004956137381969611,
      "loss": 5.0624,
      "mean_token_accuracy": 0.2013426274061203,
      "num_tokens": 15595422.0,
      "step": 8475
    },
    {
      "entropy": 5.333587551116944,
      "epoch": 0.7083194119612429,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004956078867146805,
      "loss": 5.1775,
      "mean_token_accuracy": 0.19362486749887467,
      "num_tokens": 15604223.0,
      "step": 8480
    },
    {
      "entropy": 5.40556116104126,
      "epoch": 0.7087370531239559,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004956020313703933,
      "loss": 5.1428,
      "mean_token_accuracy": 0.19927604496479034,
      "num_tokens": 15613715.0,
      "step": 8485
    },
    {
      "entropy": 5.268238115310669,
      "epoch": 0.7091546942866689,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004955961721642018,
      "loss": 4.9569,
      "mean_token_accuracy": 0.2120129272341728,
      "num_tokens": 15623826.0,
      "step": 8490
    },
    {
      "entropy": 5.273080158233642,
      "epoch": 0.7095723354493819,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000495590309096209,
      "loss": 5.0942,
      "mean_token_accuracy": 0.19263549745082856,
      "num_tokens": 15633311.0,
      "step": 8495
    },
    {
      "entropy": 5.313785791397095,
      "epoch": 0.7099899766120948,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004955844421665171,
      "loss": 5.0564,
      "mean_token_accuracy": 0.1987652912735939,
      "num_tokens": 15642785.0,
      "step": 8500
    },
    {
      "entropy": 5.490389013290406,
      "epoch": 0.7104076177748079,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004955785713752292,
      "loss": 5.2624,
      "mean_token_accuracy": 0.18767114132642745,
      "num_tokens": 15653009.0,
      "step": 8505
    },
    {
      "entropy": 5.325741910934449,
      "epoch": 0.7108252589375209,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004955726967224479,
      "loss": 5.0543,
      "mean_token_accuracy": 0.2018526926636696,
      "num_tokens": 15661672.0,
      "step": 8510
    },
    {
      "entropy": 5.248639917373657,
      "epoch": 0.7112429001002338,
      "grad_norm": 0.953125,
      "learning_rate": 0.000495566818208276,
      "loss": 5.0441,
      "mean_token_accuracy": 0.1991819754242897,
      "num_tokens": 15672186.0,
      "step": 8515
    },
    {
      "entropy": 5.259713220596313,
      "epoch": 0.7116605412629469,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004955609358328163,
      "loss": 5.0697,
      "mean_token_accuracy": 0.2026923954486847,
      "num_tokens": 15680351.0,
      "step": 8520
    },
    {
      "entropy": 5.328412771224976,
      "epoch": 0.7120781824256599,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004955550495961722,
      "loss": 5.0475,
      "mean_token_accuracy": 0.1995689183473587,
      "num_tokens": 15690140.0,
      "step": 8525
    },
    {
      "entropy": 5.321934747695923,
      "epoch": 0.7124958235883728,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004955491594984462,
      "loss": 5.1517,
      "mean_token_accuracy": 0.19319183379411697,
      "num_tokens": 15699968.0,
      "step": 8530
    },
    {
      "entropy": 5.250706958770752,
      "epoch": 0.7129134647510859,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004955432655397419,
      "loss": 4.9996,
      "mean_token_accuracy": 0.20452168434858323,
      "num_tokens": 15709061.0,
      "step": 8535
    },
    {
      "entropy": 5.337929439544678,
      "epoch": 0.7133311059137989,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004955373677201621,
      "loss": 5.1319,
      "mean_token_accuracy": 0.19452804177999497,
      "num_tokens": 15718350.0,
      "step": 8540
    },
    {
      "entropy": 5.32964677810669,
      "epoch": 0.7137487470765118,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004955314660398105,
      "loss": 5.036,
      "mean_token_accuracy": 0.20184146761894226,
      "num_tokens": 15727120.0,
      "step": 8545
    },
    {
      "entropy": 5.273453235626221,
      "epoch": 0.7141663882392248,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004955255604987899,
      "loss": 5.1297,
      "mean_token_accuracy": 0.19439579546451569,
      "num_tokens": 15735865.0,
      "step": 8550
    },
    {
      "entropy": 5.260448694229126,
      "epoch": 0.7145840294019379,
      "grad_norm": 0.921875,
      "learning_rate": 0.000495519651097204,
      "loss": 5.0978,
      "mean_token_accuracy": 0.198029263317585,
      "num_tokens": 15745239.0,
      "step": 8555
    },
    {
      "entropy": 5.279319524765015,
      "epoch": 0.7150016705646508,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004955137378351561,
      "loss": 5.0157,
      "mean_token_accuracy": 0.20383731424808502,
      "num_tokens": 15754418.0,
      "step": 8560
    },
    {
      "entropy": 5.3021814823150635,
      "epoch": 0.7154193117273638,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004955078207127498,
      "loss": 5.0664,
      "mean_token_accuracy": 0.2053982675075531,
      "num_tokens": 15764175.0,
      "step": 8565
    },
    {
      "entropy": 5.319206619262696,
      "epoch": 0.7158369528900769,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004955018997300886,
      "loss": 5.0536,
      "mean_token_accuracy": 0.2022229850292206,
      "num_tokens": 15772766.0,
      "step": 8570
    },
    {
      "entropy": 5.226112174987793,
      "epoch": 0.7162545940527898,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004954959748872764,
      "loss": 5.0776,
      "mean_token_accuracy": 0.20469839125871658,
      "num_tokens": 15781768.0,
      "step": 8575
    },
    {
      "entropy": 5.341992139816284,
      "epoch": 0.7166722352155028,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004954900461844167,
      "loss": 5.1041,
      "mean_token_accuracy": 0.19976668655872346,
      "num_tokens": 15791706.0,
      "step": 8580
    },
    {
      "entropy": 5.376216840744019,
      "epoch": 0.7170898763782159,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004954841136216132,
      "loss": 5.1361,
      "mean_token_accuracy": 0.1960015058517456,
      "num_tokens": 15802015.0,
      "step": 8585
    },
    {
      "entropy": 5.338006258010864,
      "epoch": 0.7175075175409288,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004954781771989701,
      "loss": 5.0832,
      "mean_token_accuracy": 0.20067540258169175,
      "num_tokens": 15811116.0,
      "step": 8590
    },
    {
      "entropy": 5.370889806747437,
      "epoch": 0.7179251587036418,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000495472236916591,
      "loss": 5.0495,
      "mean_token_accuracy": 0.19519623816013337,
      "num_tokens": 15819240.0,
      "step": 8595
    },
    {
      "entropy": 5.283652830123901,
      "epoch": 0.7183427998663549,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004954662927745799,
      "loss": 5.0511,
      "mean_token_accuracy": 0.1975127026438713,
      "num_tokens": 15827740.0,
      "step": 8600
    },
    {
      "entropy": 5.234874963760376,
      "epoch": 0.7187604410290678,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004954603447730411,
      "loss": 5.1169,
      "mean_token_accuracy": 0.19945017248392105,
      "num_tokens": 15836763.0,
      "step": 8605
    },
    {
      "entropy": 5.427643156051635,
      "epoch": 0.7191780821917808,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004954543929120786,
      "loss": 5.1684,
      "mean_token_accuracy": 0.19788254648447037,
      "num_tokens": 15846624.0,
      "step": 8610
    },
    {
      "entropy": 5.3217181205749515,
      "epoch": 0.7195957233544938,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004954484371917967,
      "loss": 5.0892,
      "mean_token_accuracy": 0.1965802550315857,
      "num_tokens": 15855713.0,
      "step": 8615
    },
    {
      "entropy": 5.2196485042572025,
      "epoch": 0.7200133645172068,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004954424776122993,
      "loss": 4.9274,
      "mean_token_accuracy": 0.2101362645626068,
      "num_tokens": 15863577.0,
      "step": 8620
    },
    {
      "entropy": 5.320563840866089,
      "epoch": 0.7204310056799198,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004954365141736913,
      "loss": 5.0915,
      "mean_token_accuracy": 0.20157900154590608,
      "num_tokens": 15872479.0,
      "step": 8625
    },
    {
      "entropy": 5.371408367156983,
      "epoch": 0.7208486468426328,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004954305468760765,
      "loss": 5.178,
      "mean_token_accuracy": 0.1977343365550041,
      "num_tokens": 15882170.0,
      "step": 8630
    },
    {
      "entropy": 5.317372179031372,
      "epoch": 0.7212662880053458,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004954245757195598,
      "loss": 5.0664,
      "mean_token_accuracy": 0.2025775671005249,
      "num_tokens": 15891694.0,
      "step": 8635
    },
    {
      "entropy": 5.3052088737487795,
      "epoch": 0.7216839291680588,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004954186007042454,
      "loss": 5.1867,
      "mean_token_accuracy": 0.18571624755859376,
      "num_tokens": 15902232.0,
      "step": 8640
    },
    {
      "entropy": 5.33576397895813,
      "epoch": 0.7221015703307718,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004954126218302382,
      "loss": 5.0249,
      "mean_token_accuracy": 0.20079259425401688,
      "num_tokens": 15912218.0,
      "step": 8645
    },
    {
      "entropy": 5.244195413589478,
      "epoch": 0.7225192114934847,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004954066390976427,
      "loss": 5.0018,
      "mean_token_accuracy": 0.2123638942837715,
      "num_tokens": 15922011.0,
      "step": 8650
    },
    {
      "entropy": 5.1612958908081055,
      "epoch": 0.7229368526561978,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004954006525065639,
      "loss": 5.0085,
      "mean_token_accuracy": 0.20110675692558289,
      "num_tokens": 15930394.0,
      "step": 8655
    },
    {
      "entropy": 5.287024450302124,
      "epoch": 0.7233544938189108,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004953946620571061,
      "loss": 5.1032,
      "mean_token_accuracy": 0.19816453009843826,
      "num_tokens": 15939880.0,
      "step": 8660
    },
    {
      "entropy": 5.302048635482788,
      "epoch": 0.7237721349816237,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004953886677493746,
      "loss": 4.9785,
      "mean_token_accuracy": 0.21080075055360795,
      "num_tokens": 15948521.0,
      "step": 8665
    },
    {
      "entropy": 5.299206876754761,
      "epoch": 0.7241897761443368,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004953826695834741,
      "loss": 5.1017,
      "mean_token_accuracy": 0.19396685510873796,
      "num_tokens": 15956811.0,
      "step": 8670
    },
    {
      "entropy": 5.292146444320679,
      "epoch": 0.7246074173070498,
      "grad_norm": 0.875,
      "learning_rate": 0.0004953766675595098,
      "loss": 5.1188,
      "mean_token_accuracy": 0.20170308649539948,
      "num_tokens": 15967019.0,
      "step": 8675
    },
    {
      "entropy": 5.368972301483154,
      "epoch": 0.7250250584697627,
      "grad_norm": 1.0,
      "learning_rate": 0.0004953706616775866,
      "loss": 5.1211,
      "mean_token_accuracy": 0.19261504262685775,
      "num_tokens": 15976149.0,
      "step": 8680
    },
    {
      "entropy": 5.289343595504761,
      "epoch": 0.7254426996324758,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004953646519378097,
      "loss": 4.9292,
      "mean_token_accuracy": 0.2084263727068901,
      "num_tokens": 15986711.0,
      "step": 8685
    },
    {
      "entropy": 5.288632106781006,
      "epoch": 0.7258603407951888,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004953586383402844,
      "loss": 5.106,
      "mean_token_accuracy": 0.20032290518283843,
      "num_tokens": 15996482.0,
      "step": 8690
    },
    {
      "entropy": 5.179287385940552,
      "epoch": 0.7262779819579017,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004953526208851157,
      "loss": 4.978,
      "mean_token_accuracy": 0.19885029941797255,
      "num_tokens": 16005801.0,
      "step": 8695
    },
    {
      "entropy": 5.224212265014648,
      "epoch": 0.7266956231206148,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004953465995724092,
      "loss": 5.0122,
      "mean_token_accuracy": 0.19930606782436372,
      "num_tokens": 16014807.0,
      "step": 8700
    },
    {
      "entropy": 5.347987174987793,
      "epoch": 0.7271132642833278,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004953405744022705,
      "loss": 5.1568,
      "mean_token_accuracy": 0.1955742806196213,
      "num_tokens": 16023326.0,
      "step": 8705
    },
    {
      "entropy": 5.384998369216919,
      "epoch": 0.7275309054460407,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004953345453748045,
      "loss": 5.1066,
      "mean_token_accuracy": 0.19600083082914352,
      "num_tokens": 16031648.0,
      "step": 8710
    },
    {
      "entropy": 5.296690654754639,
      "epoch": 0.7279485466087537,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004953285124901171,
      "loss": 5.0296,
      "mean_token_accuracy": 0.2059485450387001,
      "num_tokens": 16040179.0,
      "step": 8715
    },
    {
      "entropy": 5.267956352233886,
      "epoch": 0.7283661877714668,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000495322475748314,
      "loss": 5.1826,
      "mean_token_accuracy": 0.189926415681839,
      "num_tokens": 16050770.0,
      "step": 8720
    },
    {
      "entropy": 5.351437759399414,
      "epoch": 0.7287838289341797,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004953164351495007,
      "loss": 5.0462,
      "mean_token_accuracy": 0.20160710513591767,
      "num_tokens": 16060504.0,
      "step": 8725
    },
    {
      "entropy": 5.258035802841187,
      "epoch": 0.7292014700968927,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004953103906937831,
      "loss": 4.8844,
      "mean_token_accuracy": 0.22275976687669755,
      "num_tokens": 16069240.0,
      "step": 8730
    },
    {
      "entropy": 5.205895042419433,
      "epoch": 0.7296191112596058,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004953043423812669,
      "loss": 5.1425,
      "mean_token_accuracy": 0.20010207146406173,
      "num_tokens": 16079098.0,
      "step": 8735
    },
    {
      "entropy": 5.20673394203186,
      "epoch": 0.7300367524223187,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004952982902120579,
      "loss": 4.9696,
      "mean_token_accuracy": 0.2101808026432991,
      "num_tokens": 16088454.0,
      "step": 8740
    },
    {
      "entropy": 5.339543390274048,
      "epoch": 0.7304543935850317,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004952922341862622,
      "loss": 5.1168,
      "mean_token_accuracy": 0.19888666719198228,
      "num_tokens": 16096985.0,
      "step": 8745
    },
    {
      "entropy": 5.413406562805176,
      "epoch": 0.7308720347477448,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004952861743039858,
      "loss": 5.2093,
      "mean_token_accuracy": 0.19003712385892868,
      "num_tokens": 16105434.0,
      "step": 8750
    },
    {
      "entropy": 5.202343511581421,
      "epoch": 0.7312896759104577,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004952801105653347,
      "loss": 4.9366,
      "mean_token_accuracy": 0.21177592873573303,
      "num_tokens": 16114933.0,
      "step": 8755
    },
    {
      "entropy": 5.362987947463989,
      "epoch": 0.7317073170731707,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004952740429704152,
      "loss": 5.1201,
      "mean_token_accuracy": 0.1924704924225807,
      "num_tokens": 16124954.0,
      "step": 8760
    },
    {
      "entropy": 5.288468885421753,
      "epoch": 0.7321249582358837,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004952679715193334,
      "loss": 4.9216,
      "mean_token_accuracy": 0.2091547891497612,
      "num_tokens": 16134442.0,
      "step": 8765
    },
    {
      "entropy": 5.317140245437622,
      "epoch": 0.7325425993985967,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004952618962121956,
      "loss": 5.1172,
      "mean_token_accuracy": 0.19841325730085374,
      "num_tokens": 16143994.0,
      "step": 8770
    },
    {
      "entropy": 5.278021574020386,
      "epoch": 0.7329602405613097,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004952558170491082,
      "loss": 5.0337,
      "mean_token_accuracy": 0.20188337564468384,
      "num_tokens": 16152931.0,
      "step": 8775
    },
    {
      "entropy": 5.26608567237854,
      "epoch": 0.7333778817240227,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004952497340301776,
      "loss": 5.0422,
      "mean_token_accuracy": 0.19102291911840438,
      "num_tokens": 16162716.0,
      "step": 8780
    },
    {
      "entropy": 5.326710271835327,
      "epoch": 0.7337955228867357,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004952436471555103,
      "loss": 5.0982,
      "mean_token_accuracy": 0.19720979928970336,
      "num_tokens": 16172371.0,
      "step": 8785
    },
    {
      "entropy": 5.333395624160767,
      "epoch": 0.7342131640494487,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004952375564252128,
      "loss": 5.1641,
      "mean_token_accuracy": 0.19591792523860932,
      "num_tokens": 16181823.0,
      "step": 8790
    },
    {
      "entropy": 5.274487066268921,
      "epoch": 0.7346308052121617,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004952314618393919,
      "loss": 5.0779,
      "mean_token_accuracy": 0.20523362159729003,
      "num_tokens": 16190720.0,
      "step": 8795
    },
    {
      "entropy": 5.275799369812011,
      "epoch": 0.7350484463748747,
      "grad_norm": 0.984375,
      "learning_rate": 0.000495225363398154,
      "loss": 5.0042,
      "mean_token_accuracy": 0.2035899981856346,
      "num_tokens": 16199589.0,
      "step": 8800
    },
    {
      "entropy": 5.398755693435669,
      "epoch": 0.7354660875375877,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004952192611016061,
      "loss": 5.187,
      "mean_token_accuracy": 0.19683736860752105,
      "num_tokens": 16208800.0,
      "step": 8805
    },
    {
      "entropy": 5.3511944770812985,
      "epoch": 0.7358837287003007,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000495213154949855,
      "loss": 5.1544,
      "mean_token_accuracy": 0.19367158859968187,
      "num_tokens": 16218835.0,
      "step": 8810
    },
    {
      "entropy": 5.361223125457764,
      "epoch": 0.7363013698630136,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004952070449430075,
      "loss": 5.1077,
      "mean_token_accuracy": 0.19308672994375228,
      "num_tokens": 16228441.0,
      "step": 8815
    },
    {
      "entropy": 5.337395524978637,
      "epoch": 0.7367190110257267,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004952009310811706,
      "loss": 5.0786,
      "mean_token_accuracy": 0.20750884115695953,
      "num_tokens": 16237877.0,
      "step": 8820
    },
    {
      "entropy": 5.266070127487183,
      "epoch": 0.7371366521884397,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004951948133644514,
      "loss": 5.0201,
      "mean_token_accuracy": 0.20008860826492308,
      "num_tokens": 16246809.0,
      "step": 8825
    },
    {
      "entropy": 5.344852924346924,
      "epoch": 0.7375542933511526,
      "grad_norm": 1.0,
      "learning_rate": 0.000495188691792957,
      "loss": 5.086,
      "mean_token_accuracy": 0.20195574760437013,
      "num_tokens": 16255754.0,
      "step": 8830
    },
    {
      "entropy": 5.336864757537842,
      "epoch": 0.7379719345138657,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004951825663667943,
      "loss": 5.1167,
      "mean_token_accuracy": 0.2024869978427887,
      "num_tokens": 16264861.0,
      "step": 8835
    },
    {
      "entropy": 5.286258506774902,
      "epoch": 0.7383895756765787,
      "grad_norm": 0.90625,
      "learning_rate": 0.000495176437086071,
      "loss": 5.0493,
      "mean_token_accuracy": 0.20388441085815429,
      "num_tokens": 16273628.0,
      "step": 8840
    },
    {
      "entropy": 5.28920955657959,
      "epoch": 0.7388072168392917,
      "grad_norm": 0.9375,
      "learning_rate": 0.000495170303950894,
      "loss": 5.0604,
      "mean_token_accuracy": 0.1995601996779442,
      "num_tokens": 16282958.0,
      "step": 8845
    },
    {
      "entropy": 5.349431657791138,
      "epoch": 0.7392248580020047,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004951641669613708,
      "loss": 5.082,
      "mean_token_accuracy": 0.19506461620330812,
      "num_tokens": 16292763.0,
      "step": 8850
    },
    {
      "entropy": 5.317784452438355,
      "epoch": 0.7396424991647177,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004951580261176088,
      "loss": 5.0746,
      "mean_token_accuracy": 0.20056016892194747,
      "num_tokens": 16302216.0,
      "step": 8855
    },
    {
      "entropy": 5.213425588607788,
      "epoch": 0.7400601403274307,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004951518814197156,
      "loss": 4.989,
      "mean_token_accuracy": 0.21325503140687943,
      "num_tokens": 16311672.0,
      "step": 8860
    },
    {
      "entropy": 5.281796836853028,
      "epoch": 0.7404777814901436,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004951457328677989,
      "loss": 5.0069,
      "mean_token_accuracy": 0.20630813091993333,
      "num_tokens": 16321027.0,
      "step": 8865
    },
    {
      "entropy": 5.322635841369629,
      "epoch": 0.7408954226528567,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004951395804619659,
      "loss": 5.1151,
      "mean_token_accuracy": 0.1965378686785698,
      "num_tokens": 16329471.0,
      "step": 8870
    },
    {
      "entropy": 5.25032114982605,
      "epoch": 0.7413130638155697,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004951334242023247,
      "loss": 5.0521,
      "mean_token_accuracy": 0.20165877193212509,
      "num_tokens": 16339085.0,
      "step": 8875
    },
    {
      "entropy": 5.382337045669556,
      "epoch": 0.7417307049782826,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004951272640889828,
      "loss": 5.1183,
      "mean_token_accuracy": 0.20480622053146363,
      "num_tokens": 16347950.0,
      "step": 8880
    },
    {
      "entropy": 5.291828632354736,
      "epoch": 0.7421483461409957,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004951211001220484,
      "loss": 4.9277,
      "mean_token_accuracy": 0.20828215330839156,
      "num_tokens": 16357129.0,
      "step": 8885
    },
    {
      "entropy": 5.222485065460205,
      "epoch": 0.7425659873037087,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000495114932301629,
      "loss": 5.0554,
      "mean_token_accuracy": 0.20285916626453399,
      "num_tokens": 16365768.0,
      "step": 8890
    },
    {
      "entropy": 5.263238191604614,
      "epoch": 0.7429836284664216,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004951087606278328,
      "loss": 5.0072,
      "mean_token_accuracy": 0.2080109566450119,
      "num_tokens": 16375548.0,
      "step": 8895
    },
    {
      "entropy": 5.323984146118164,
      "epoch": 0.7434012696291347,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004951025851007678,
      "loss": 5.0739,
      "mean_token_accuracy": 0.20259851962327957,
      "num_tokens": 16385472.0,
      "step": 8900
    },
    {
      "entropy": 5.30631012916565,
      "epoch": 0.7438189107918477,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004950964057205421,
      "loss": 4.949,
      "mean_token_accuracy": 0.20254386365413665,
      "num_tokens": 16395048.0,
      "step": 8905
    },
    {
      "entropy": 5.339939069747925,
      "epoch": 0.7442365519545606,
      "grad_norm": 0.875,
      "learning_rate": 0.000495090222487264,
      "loss": 5.1675,
      "mean_token_accuracy": 0.1944541350007057,
      "num_tokens": 16405378.0,
      "step": 8910
    },
    {
      "entropy": 5.244132709503174,
      "epoch": 0.7446541931172737,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004950840354010415,
      "loss": 5.0476,
      "mean_token_accuracy": 0.20192829370498658,
      "num_tokens": 16415206.0,
      "step": 8915
    },
    {
      "entropy": 5.3346442699432375,
      "epoch": 0.7450718342799867,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004950778444619831,
      "loss": 5.0574,
      "mean_token_accuracy": 0.2084740400314331,
      "num_tokens": 16424185.0,
      "step": 8920
    },
    {
      "entropy": 5.30698299407959,
      "epoch": 0.7454894754426996,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000495071649670197,
      "loss": 5.0351,
      "mean_token_accuracy": 0.19958733022212982,
      "num_tokens": 16433330.0,
      "step": 8925
    },
    {
      "entropy": 5.267160367965698,
      "epoch": 0.7459071166054126,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000495065451025792,
      "loss": 5.0495,
      "mean_token_accuracy": 0.19949113428592682,
      "num_tokens": 16442777.0,
      "step": 8930
    },
    {
      "entropy": 5.285257148742676,
      "epoch": 0.7463247577681257,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004950592485288761,
      "loss": 5.0361,
      "mean_token_accuracy": 0.2103156641125679,
      "num_tokens": 16451838.0,
      "step": 8935
    },
    {
      "entropy": 5.250347042083741,
      "epoch": 0.7467423989308386,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004950530421795584,
      "loss": 5.0128,
      "mean_token_accuracy": 0.2054109513759613,
      "num_tokens": 16460022.0,
      "step": 8940
    },
    {
      "entropy": 5.219787740707398,
      "epoch": 0.7471600400935516,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004950468319779472,
      "loss": 4.9052,
      "mean_token_accuracy": 0.21174386739730836,
      "num_tokens": 16468891.0,
      "step": 8945
    },
    {
      "entropy": 5.208614683151245,
      "epoch": 0.7475776812562647,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004950406179241515,
      "loss": 5.0011,
      "mean_token_accuracy": 0.20614684820175172,
      "num_tokens": 16478020.0,
      "step": 8950
    },
    {
      "entropy": 5.289824056625366,
      "epoch": 0.7479953224189776,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004950344000182798,
      "loss": 5.1508,
      "mean_token_accuracy": 0.19899638295173644,
      "num_tokens": 16487536.0,
      "step": 8955
    },
    {
      "entropy": 5.32879319190979,
      "epoch": 0.7484129635816906,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004950281782604411,
      "loss": 5.0736,
      "mean_token_accuracy": 0.19055721014738083,
      "num_tokens": 16495786.0,
      "step": 8960
    },
    {
      "entropy": 5.396314001083374,
      "epoch": 0.7488306047444037,
      "grad_norm": 1.0,
      "learning_rate": 0.0004950219526507442,
      "loss": 5.1475,
      "mean_token_accuracy": 0.19518198370933532,
      "num_tokens": 16505535.0,
      "step": 8965
    },
    {
      "entropy": 5.368054962158203,
      "epoch": 0.7492482459071166,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004950157231892983,
      "loss": 5.1537,
      "mean_token_accuracy": 0.1970652833580971,
      "num_tokens": 16515387.0,
      "step": 8970
    },
    {
      "entropy": 5.292455005645752,
      "epoch": 0.7496658870698296,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004950094898762123,
      "loss": 5.0942,
      "mean_token_accuracy": 0.20016653090715408,
      "num_tokens": 16524421.0,
      "step": 8975
    },
    {
      "entropy": 5.3481401920318605,
      "epoch": 0.7500835282325427,
      "grad_norm": 0.875,
      "learning_rate": 0.0004950032527115954,
      "loss": 5.0846,
      "mean_token_accuracy": 0.2012725993990898,
      "num_tokens": 16534533.0,
      "step": 8980
    },
    {
      "entropy": 5.306323385238647,
      "epoch": 0.7505011693952556,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004949970116955566,
      "loss": 5.0972,
      "mean_token_accuracy": 0.19048482775688172,
      "num_tokens": 16543712.0,
      "step": 8985
    },
    {
      "entropy": 5.315547323226928,
      "epoch": 0.7509188105579686,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004949907668282054,
      "loss": 5.0855,
      "mean_token_accuracy": 0.2044758290052414,
      "num_tokens": 16552475.0,
      "step": 8990
    },
    {
      "entropy": 5.351527690887451,
      "epoch": 0.7513364517206816,
      "grad_norm": 1.0,
      "learning_rate": 0.000494984518109651,
      "loss": 5.057,
      "mean_token_accuracy": 0.19992384761571885,
      "num_tokens": 16562237.0,
      "step": 8995
    },
    {
      "entropy": 5.289912939071655,
      "epoch": 0.7517540928833946,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004949782655400028,
      "loss": 5.0565,
      "mean_token_accuracy": 0.204191355407238,
      "num_tokens": 16570508.0,
      "step": 9000
    },
    {
      "epoch": 0.7517540928833946,
      "eval_entropy": 4.985084426526774,
      "eval_loss": 5.020449161529541,
      "eval_mean_token_accuracy": 0.2132798703614243,
      "eval_num_tokens": 16570508.0,
      "eval_runtime": 20.8884,
      "eval_samples_per_second": 1807.845,
      "eval_steps_per_second": 226.011,
      "step": 9000
    },
    {
      "entropy": 5.170556163787841,
      "epoch": 0.7521717340461076,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004949720091193704,
      "loss": 4.9773,
      "mean_token_accuracy": 0.21138520836830138,
      "num_tokens": 16578561.0,
      "step": 9005
    },
    {
      "entropy": 5.237123966217041,
      "epoch": 0.7525893752088206,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000494965748847863,
      "loss": 4.9459,
      "mean_token_accuracy": 0.20887226313352586,
      "num_tokens": 16587168.0,
      "step": 9010
    },
    {
      "entropy": 5.295994710922241,
      "epoch": 0.7530070163715336,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004949594847255905,
      "loss": 5.1062,
      "mean_token_accuracy": 0.19817576557397842,
      "num_tokens": 16597252.0,
      "step": 9015
    },
    {
      "entropy": 5.253061723709107,
      "epoch": 0.7534246575342466,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004949532167526625,
      "loss": 4.927,
      "mean_token_accuracy": 0.2133822336792946,
      "num_tokens": 16607558.0,
      "step": 9020
    },
    {
      "entropy": 5.3121092319488525,
      "epoch": 0.7538422986969596,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004949469449291886,
      "loss": 5.1424,
      "mean_token_accuracy": 0.19372957646846772,
      "num_tokens": 16617019.0,
      "step": 9025
    },
    {
      "entropy": 5.254108858108521,
      "epoch": 0.7542599398596725,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004949406692552787,
      "loss": 4.9927,
      "mean_token_accuracy": 0.20554035604000093,
      "num_tokens": 16625815.0,
      "step": 9030
    },
    {
      "entropy": 5.350965118408203,
      "epoch": 0.7546775810223856,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004949343897310426,
      "loss": 5.2014,
      "mean_token_accuracy": 0.1915669783949852,
      "num_tokens": 16634876.0,
      "step": 9035
    },
    {
      "entropy": 5.272903823852539,
      "epoch": 0.7550952221850986,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004949281063565904,
      "loss": 5.0839,
      "mean_token_accuracy": 0.19184344708919526,
      "num_tokens": 16643770.0,
      "step": 9040
    },
    {
      "entropy": 5.329181957244873,
      "epoch": 0.7555128633478115,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004949218191320319,
      "loss": 5.0217,
      "mean_token_accuracy": 0.20560361742973327,
      "num_tokens": 16652845.0,
      "step": 9045
    },
    {
      "entropy": 5.250745344161987,
      "epoch": 0.7559305045105246,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004949155280574772,
      "loss": 4.9963,
      "mean_token_accuracy": 0.20292529165744783,
      "num_tokens": 16662224.0,
      "step": 9050
    },
    {
      "entropy": 5.269215154647827,
      "epoch": 0.7563481456732376,
      "grad_norm": 0.875,
      "learning_rate": 0.0004949092331330365,
      "loss": 5.0132,
      "mean_token_accuracy": 0.20422899723052979,
      "num_tokens": 16671114.0,
      "step": 9055
    },
    {
      "entropy": 5.31806583404541,
      "epoch": 0.7567657868359505,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00049490293435882,
      "loss": 5.0883,
      "mean_token_accuracy": 0.19502866715192796,
      "num_tokens": 16680395.0,
      "step": 9060
    },
    {
      "entropy": 5.306323289871216,
      "epoch": 0.7571834279986636,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004948966317349379,
      "loss": 5.0445,
      "mean_token_accuracy": 0.20500168204307556,
      "num_tokens": 16690251.0,
      "step": 9065
    },
    {
      "entropy": 5.218173170089722,
      "epoch": 0.7576010691613766,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004948903252615006,
      "loss": 4.9331,
      "mean_token_accuracy": 0.20970238298177718,
      "num_tokens": 16698945.0,
      "step": 9070
    },
    {
      "entropy": 5.30863585472107,
      "epoch": 0.7580187103240895,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004948840149386184,
      "loss": 5.124,
      "mean_token_accuracy": 0.19658341407775878,
      "num_tokens": 16708325.0,
      "step": 9075
    },
    {
      "entropy": 5.2907386302948,
      "epoch": 0.7584363514868026,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000494877700766402,
      "loss": 5.0291,
      "mean_token_accuracy": 0.19982946217060088,
      "num_tokens": 16716392.0,
      "step": 9080
    },
    {
      "entropy": 5.292391777038574,
      "epoch": 0.7588539926495156,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004948713827449616,
      "loss": 5.0485,
      "mean_token_accuracy": 0.19966347217559816,
      "num_tokens": 16725284.0,
      "step": 9085
    },
    {
      "entropy": 5.245575952529907,
      "epoch": 0.7592716338122285,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004948650608744081,
      "loss": 5.0675,
      "mean_token_accuracy": 0.20245892852544783,
      "num_tokens": 16734658.0,
      "step": 9090
    },
    {
      "entropy": 5.385666751861573,
      "epoch": 0.7596892749749415,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000494858735154852,
      "loss": 5.1584,
      "mean_token_accuracy": 0.19382836669683456,
      "num_tokens": 16744403.0,
      "step": 9095
    },
    {
      "entropy": 5.377468681335449,
      "epoch": 0.7601069161376546,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004948524055864041,
      "loss": 5.135,
      "mean_token_accuracy": 0.1971142277121544,
      "num_tokens": 16754254.0,
      "step": 9100
    },
    {
      "entropy": 5.312412214279175,
      "epoch": 0.7605245573003675,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004948460721691752,
      "loss": 5.1425,
      "mean_token_accuracy": 0.1937744736671448,
      "num_tokens": 16763941.0,
      "step": 9105
    },
    {
      "entropy": 5.284266614913941,
      "epoch": 0.7609421984630805,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004948397349032762,
      "loss": 5.0532,
      "mean_token_accuracy": 0.20246528089046478,
      "num_tokens": 16773181.0,
      "step": 9110
    },
    {
      "entropy": 5.214625024795533,
      "epoch": 0.7613598396257936,
      "grad_norm": 1.015625,
      "learning_rate": 0.000494833393788818,
      "loss": 4.8893,
      "mean_token_accuracy": 0.21349231898784637,
      "num_tokens": 16782622.0,
      "step": 9115
    },
    {
      "entropy": 5.319112777709961,
      "epoch": 0.7617774807885065,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004948270488259117,
      "loss": 5.1246,
      "mean_token_accuracy": 0.19989049136638642,
      "num_tokens": 16792805.0,
      "step": 9120
    },
    {
      "entropy": 5.259339237213135,
      "epoch": 0.7621951219512195,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004948207000146682,
      "loss": 5.0974,
      "mean_token_accuracy": 0.20535246282815933,
      "num_tokens": 16802641.0,
      "step": 9125
    },
    {
      "entropy": 5.330405044555664,
      "epoch": 0.7626127631139326,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004948143473551989,
      "loss": 5.0734,
      "mean_token_accuracy": 0.1947854146361351,
      "num_tokens": 16812037.0,
      "step": 9130
    },
    {
      "entropy": 5.293272066116333,
      "epoch": 0.7630304042766455,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004948079908476147,
      "loss": 5.0389,
      "mean_token_accuracy": 0.21257772743701936,
      "num_tokens": 16820279.0,
      "step": 9135
    },
    {
      "entropy": 5.241614532470703,
      "epoch": 0.7634480454393585,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004948016304920271,
      "loss": 5.031,
      "mean_token_accuracy": 0.20147226601839066,
      "num_tokens": 16829396.0,
      "step": 9140
    },
    {
      "entropy": 5.285134029388428,
      "epoch": 0.7638656866020715,
      "grad_norm": 1.0,
      "learning_rate": 0.0004947952662885473,
      "loss": 5.0276,
      "mean_token_accuracy": 0.20086823403835297,
      "num_tokens": 16838389.0,
      "step": 9145
    },
    {
      "entropy": 5.267738628387451,
      "epoch": 0.7642833277647845,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000494788898237287,
      "loss": 5.1257,
      "mean_token_accuracy": 0.1968758821487427,
      "num_tokens": 16848012.0,
      "step": 9150
    },
    {
      "entropy": 5.364105892181397,
      "epoch": 0.7647009689274975,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004947825263383573,
      "loss": 5.0901,
      "mean_token_accuracy": 0.20088518261909485,
      "num_tokens": 16857591.0,
      "step": 9155
    },
    {
      "entropy": 5.356039381027221,
      "epoch": 0.7651186100902105,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00049477615059187,
      "loss": 5.0875,
      "mean_token_accuracy": 0.19213848561048508,
      "num_tokens": 16866792.0,
      "step": 9160
    },
    {
      "entropy": 5.225041294097901,
      "epoch": 0.7655362512529235,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004947697709979367,
      "loss": 5.0984,
      "mean_token_accuracy": 0.20212144553661346,
      "num_tokens": 16876602.0,
      "step": 9165
    },
    {
      "entropy": 5.291879320144654,
      "epoch": 0.7659538924156365,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004947633875566688,
      "loss": 5.0801,
      "mean_token_accuracy": 0.1977691650390625,
      "num_tokens": 16886992.0,
      "step": 9170
    },
    {
      "entropy": 5.335131883621216,
      "epoch": 0.7663715335783495,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004947570002681785,
      "loss": 5.0184,
      "mean_token_accuracy": 0.20323084145784379,
      "num_tokens": 16895716.0,
      "step": 9175
    },
    {
      "entropy": 5.242017602920532,
      "epoch": 0.7667891747410625,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004947506091325772,
      "loss": 5.0209,
      "mean_token_accuracy": 0.2108732894062996,
      "num_tokens": 16904568.0,
      "step": 9180
    },
    {
      "entropy": 5.309365844726562,
      "epoch": 0.7672068159037755,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004947442141499771,
      "loss": 5.1177,
      "mean_token_accuracy": 0.20018309652805327,
      "num_tokens": 16913667.0,
      "step": 9185
    },
    {
      "entropy": 5.291602897644043,
      "epoch": 0.7676244570664885,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004947378153204899,
      "loss": 5.0441,
      "mean_token_accuracy": 0.1995931714773178,
      "num_tokens": 16922897.0,
      "step": 9190
    },
    {
      "entropy": 5.309848499298096,
      "epoch": 0.7680420982292014,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004947314126442279,
      "loss": 5.0437,
      "mean_token_accuracy": 0.20319053530693054,
      "num_tokens": 16931559.0,
      "step": 9195
    },
    {
      "entropy": 5.262634181976319,
      "epoch": 0.7684597393919145,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000494725006121303,
      "loss": 5.0492,
      "mean_token_accuracy": 0.19733755886554719,
      "num_tokens": 16940191.0,
      "step": 9200
    },
    {
      "entropy": 5.344131755828857,
      "epoch": 0.7688773805546275,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004947185957518273,
      "loss": 5.0972,
      "mean_token_accuracy": 0.19695994704961778,
      "num_tokens": 16948971.0,
      "step": 9205
    },
    {
      "entropy": 5.316513729095459,
      "epoch": 0.7692950217173404,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004947121815359133,
      "loss": 5.0304,
      "mean_token_accuracy": 0.20229455083608627,
      "num_tokens": 16957944.0,
      "step": 9210
    },
    {
      "entropy": 5.208393812179565,
      "epoch": 0.7697126628800535,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004947057634736729,
      "loss": 5.0059,
      "mean_token_accuracy": 0.199810491502285,
      "num_tokens": 16967418.0,
      "step": 9215
    },
    {
      "entropy": 5.263882160186768,
      "epoch": 0.7701303040427665,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004946993415652186,
      "loss": 4.9992,
      "mean_token_accuracy": 0.21057995557785034,
      "num_tokens": 16977233.0,
      "step": 9220
    },
    {
      "entropy": 5.272703981399536,
      "epoch": 0.7705479452054794,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000494692915810663,
      "loss": 5.0529,
      "mean_token_accuracy": 0.21028997600078583,
      "num_tokens": 16986023.0,
      "step": 9225
    },
    {
      "entropy": 5.244377565383911,
      "epoch": 0.7709655863681925,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004946864862101185,
      "loss": 4.9831,
      "mean_token_accuracy": 0.20163469016551971,
      "num_tokens": 16994987.0,
      "step": 9230
    },
    {
      "entropy": 5.258209609985352,
      "epoch": 0.7713832275309055,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004946800527636975,
      "loss": 4.9993,
      "mean_token_accuracy": 0.20105541497468948,
      "num_tokens": 17003902.0,
      "step": 9235
    },
    {
      "entropy": 5.315467262268067,
      "epoch": 0.7718008686936184,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004946736154715129,
      "loss": 5.11,
      "mean_token_accuracy": 0.19503421187400818,
      "num_tokens": 17013258.0,
      "step": 9240
    },
    {
      "entropy": 5.376742649078369,
      "epoch": 0.7722185098563314,
      "grad_norm": 1.0,
      "learning_rate": 0.0004946671743336771,
      "loss": 5.0986,
      "mean_token_accuracy": 0.19482320845127105,
      "num_tokens": 17022258.0,
      "step": 9245
    },
    {
      "entropy": 5.302328205108642,
      "epoch": 0.7726361510190445,
      "grad_norm": 1.0,
      "learning_rate": 0.0004946607293503031,
      "loss": 5.0356,
      "mean_token_accuracy": 0.1994878500699997,
      "num_tokens": 17031485.0,
      "step": 9250
    },
    {
      "entropy": 5.1247354507446286,
      "epoch": 0.7730537921817574,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004946542805215036,
      "loss": 4.9885,
      "mean_token_accuracy": 0.20981451869010925,
      "num_tokens": 17039916.0,
      "step": 9255
    },
    {
      "entropy": 5.192206907272339,
      "epoch": 0.7734714333444704,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004946478278473914,
      "loss": 5.0445,
      "mean_token_accuracy": 0.20091699212789535,
      "num_tokens": 17048153.0,
      "step": 9260
    },
    {
      "entropy": 5.278877878189087,
      "epoch": 0.7738890745071835,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004946413713280797,
      "loss": 5.0468,
      "mean_token_accuracy": 0.2015118271112442,
      "num_tokens": 17057473.0,
      "step": 9265
    },
    {
      "entropy": 5.322338199615478,
      "epoch": 0.7743067156698964,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004946349109636814,
      "loss": 5.067,
      "mean_token_accuracy": 0.20387126356363297,
      "num_tokens": 17066431.0,
      "step": 9270
    },
    {
      "entropy": 5.340468072891236,
      "epoch": 0.7747243568326094,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004946284467543096,
      "loss": 5.1462,
      "mean_token_accuracy": 0.19555752277374266,
      "num_tokens": 17076519.0,
      "step": 9275
    },
    {
      "entropy": 5.322460079193116,
      "epoch": 0.7751419979953225,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004946219787000776,
      "loss": 5.0741,
      "mean_token_accuracy": 0.19836290329694747,
      "num_tokens": 17084651.0,
      "step": 9280
    },
    {
      "entropy": 5.332851696014404,
      "epoch": 0.7755596391580354,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004946155068010984,
      "loss": 5.1086,
      "mean_token_accuracy": 0.1989627704024315,
      "num_tokens": 17094943.0,
      "step": 9285
    },
    {
      "entropy": 5.279789447784424,
      "epoch": 0.7759772803207484,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004946090310574854,
      "loss": 4.9896,
      "mean_token_accuracy": 0.2120742380619049,
      "num_tokens": 17103548.0,
      "step": 9290
    },
    {
      "entropy": 5.295510911941529,
      "epoch": 0.7763949214834615,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004946025514693519,
      "loss": 4.9535,
      "mean_token_accuracy": 0.20816740542650222,
      "num_tokens": 17112472.0,
      "step": 9295
    },
    {
      "entropy": 5.406742191314697,
      "epoch": 0.7768125626461744,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004945960680368115,
      "loss": 5.0966,
      "mean_token_accuracy": 0.19842805564403534,
      "num_tokens": 17121536.0,
      "step": 9300
    },
    {
      "entropy": 5.385002279281617,
      "epoch": 0.7772302038088874,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004945895807599775,
      "loss": 5.1612,
      "mean_token_accuracy": 0.19160933345556258,
      "num_tokens": 17130708.0,
      "step": 9305
    },
    {
      "entropy": 5.236834383010864,
      "epoch": 0.7776478449716004,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004945830896389636,
      "loss": 4.9347,
      "mean_token_accuracy": 0.20709720402956008,
      "num_tokens": 17139437.0,
      "step": 9310
    },
    {
      "entropy": 5.282104206085205,
      "epoch": 0.7780654861343134,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004945765946738835,
      "loss": 5.1022,
      "mean_token_accuracy": 0.1942005604505539,
      "num_tokens": 17149107.0,
      "step": 9315
    },
    {
      "entropy": 5.2932329177856445,
      "epoch": 0.7784831272970264,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004945700958648507,
      "loss": 5.0908,
      "mean_token_accuracy": 0.1964163213968277,
      "num_tokens": 17158866.0,
      "step": 9320
    },
    {
      "entropy": 5.264914512634277,
      "epoch": 0.7789007684597394,
      "grad_norm": 1.09375,
      "learning_rate": 0.000494563593211979,
      "loss": 4.9796,
      "mean_token_accuracy": 0.20714643895626067,
      "num_tokens": 17167909.0,
      "step": 9325
    },
    {
      "entropy": 5.27353801727295,
      "epoch": 0.7793184096224524,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004945570867153826,
      "loss": 5.0963,
      "mean_token_accuracy": 0.20186814218759536,
      "num_tokens": 17176746.0,
      "step": 9330
    },
    {
      "entropy": 5.297835445404052,
      "epoch": 0.7797360507851654,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004945505763751749,
      "loss": 5.0988,
      "mean_token_accuracy": 0.19718379378318787,
      "num_tokens": 17186219.0,
      "step": 9335
    },
    {
      "entropy": 5.29302659034729,
      "epoch": 0.7801536919478784,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004945440621914702,
      "loss": 4.9235,
      "mean_token_accuracy": 0.21178438812494277,
      "num_tokens": 17195811.0,
      "step": 9340
    },
    {
      "entropy": 5.194170331954956,
      "epoch": 0.7805713331105913,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004945375441643823,
      "loss": 5.0806,
      "mean_token_accuracy": 0.20125791430473328,
      "num_tokens": 17205333.0,
      "step": 9345
    },
    {
      "entropy": 5.310766172409058,
      "epoch": 0.7809889742733044,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004945310222940254,
      "loss": 5.0301,
      "mean_token_accuracy": 0.20145449489355088,
      "num_tokens": 17214493.0,
      "step": 9350
    },
    {
      "entropy": 5.3490368843078615,
      "epoch": 0.7814066154360174,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004945244965805138,
      "loss": 5.0567,
      "mean_token_accuracy": 0.20174674987792968,
      "num_tokens": 17223602.0,
      "step": 9355
    },
    {
      "entropy": 5.30027060508728,
      "epoch": 0.7818242565987303,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004945179670239616,
      "loss": 5.0837,
      "mean_token_accuracy": 0.19894535839557648,
      "num_tokens": 17232915.0,
      "step": 9360
    },
    {
      "entropy": 5.331206512451172,
      "epoch": 0.7822418977614434,
      "grad_norm": 1.125,
      "learning_rate": 0.0004945114336244832,
      "loss": 5.0555,
      "mean_token_accuracy": 0.19908300489187242,
      "num_tokens": 17242489.0,
      "step": 9365
    },
    {
      "entropy": 5.202166748046875,
      "epoch": 0.7826595389241564,
      "grad_norm": 0.953125,
      "learning_rate": 0.000494504896382193,
      "loss": 4.8248,
      "mean_token_accuracy": 0.21493325382471085,
      "num_tokens": 17251496.0,
      "step": 9370
    },
    {
      "entropy": 5.2160400390625,
      "epoch": 0.7830771800868693,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004944983552972053,
      "loss": 4.9989,
      "mean_token_accuracy": 0.20490347295999528,
      "num_tokens": 17260550.0,
      "step": 9375
    },
    {
      "entropy": 5.265153551101685,
      "epoch": 0.7834948212495824,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004944918103696347,
      "loss": 4.9848,
      "mean_token_accuracy": 0.2017824724316597,
      "num_tokens": 17269085.0,
      "step": 9380
    },
    {
      "entropy": 5.361507368087769,
      "epoch": 0.7839124624122954,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004944852615995956,
      "loss": 5.201,
      "mean_token_accuracy": 0.19807861298322677,
      "num_tokens": 17278766.0,
      "step": 9385
    },
    {
      "entropy": 5.3987144947052,
      "epoch": 0.7843301035750083,
      "grad_norm": 1.03125,
      "learning_rate": 0.000494478708987203,
      "loss": 5.0623,
      "mean_token_accuracy": 0.1967109724879265,
      "num_tokens": 17288316.0,
      "step": 9390
    },
    {
      "entropy": 5.220450687408447,
      "epoch": 0.7847477447377214,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004944721525325713,
      "loss": 4.9237,
      "mean_token_accuracy": 0.20765337198972703,
      "num_tokens": 17297829.0,
      "step": 9395
    },
    {
      "entropy": 5.267058086395264,
      "epoch": 0.7851653859004344,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004944655922358154,
      "loss": 5.0109,
      "mean_token_accuracy": 0.201911860704422,
      "num_tokens": 17306631.0,
      "step": 9400
    },
    {
      "entropy": 5.293751668930054,
      "epoch": 0.7855830270631473,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004944590280970503,
      "loss": 5.0055,
      "mean_token_accuracy": 0.2069832667708397,
      "num_tokens": 17316829.0,
      "step": 9405
    },
    {
      "entropy": 5.283893251419068,
      "epoch": 0.7860006682258603,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004944524601163905,
      "loss": 5.019,
      "mean_token_accuracy": 0.19803465902805328,
      "num_tokens": 17326568.0,
      "step": 9410
    },
    {
      "entropy": 5.292175579071045,
      "epoch": 0.7864183093885734,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004944458882939513,
      "loss": 5.0936,
      "mean_token_accuracy": 0.19795591980218888,
      "num_tokens": 17335524.0,
      "step": 9415
    },
    {
      "entropy": 5.338109827041626,
      "epoch": 0.7868359505512863,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004944393126298478,
      "loss": 5.1231,
      "mean_token_accuracy": 0.1944841593503952,
      "num_tokens": 17346198.0,
      "step": 9420
    },
    {
      "entropy": 5.294412899017334,
      "epoch": 0.7872535917139993,
      "grad_norm": 1.046875,
      "learning_rate": 0.000494432733124195,
      "loss": 4.9621,
      "mean_token_accuracy": 0.20951931476593016,
      "num_tokens": 17354696.0,
      "step": 9425
    },
    {
      "entropy": 5.2194891452789305,
      "epoch": 0.7876712328767124,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000494426149777108,
      "loss": 4.9448,
      "mean_token_accuracy": 0.20475293397903443,
      "num_tokens": 17363314.0,
      "step": 9430
    },
    {
      "entropy": 5.241823053359985,
      "epoch": 0.7880888740394253,
      "grad_norm": 0.90625,
      "learning_rate": 0.000494419562588702,
      "loss": 4.9988,
      "mean_token_accuracy": 0.2086746796965599,
      "num_tokens": 17372662.0,
      "step": 9435
    },
    {
      "entropy": 5.241186189651489,
      "epoch": 0.7885065152021383,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004944129715590927,
      "loss": 4.9881,
      "mean_token_accuracy": 0.20144128799438477,
      "num_tokens": 17381931.0,
      "step": 9440
    },
    {
      "entropy": 5.342121648788452,
      "epoch": 0.7889241563648514,
      "grad_norm": 1.0,
      "learning_rate": 0.000494406376688395,
      "loss": 5.0908,
      "mean_token_accuracy": 0.19401448220014572,
      "num_tokens": 17391210.0,
      "step": 9445
    },
    {
      "entropy": 5.240361928939819,
      "epoch": 0.7893417975275643,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004943997779767248,
      "loss": 4.9358,
      "mean_token_accuracy": 0.21305716931819915,
      "num_tokens": 17401102.0,
      "step": 9450
    },
    {
      "entropy": 5.241988706588745,
      "epoch": 0.7897594386902773,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004943931754241972,
      "loss": 4.9859,
      "mean_token_accuracy": 0.20448096245527267,
      "num_tokens": 17410036.0,
      "step": 9455
    },
    {
      "entropy": 5.27645902633667,
      "epoch": 0.7901770798529903,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000494386569030928,
      "loss": 5.0358,
      "mean_token_accuracy": 0.1969575732946396,
      "num_tokens": 17419348.0,
      "step": 9460
    },
    {
      "entropy": 5.267658853530884,
      "epoch": 0.7905947210157033,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004943799587970328,
      "loss": 5.0429,
      "mean_token_accuracy": 0.1980777457356453,
      "num_tokens": 17427937.0,
      "step": 9465
    },
    {
      "entropy": 5.241451072692871,
      "epoch": 0.7910123621784163,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004943733447226275,
      "loss": 4.9156,
      "mean_token_accuracy": 0.20798415690660477,
      "num_tokens": 17436259.0,
      "step": 9470
    },
    {
      "entropy": 5.357368612289429,
      "epoch": 0.7914300033411293,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004943667268078277,
      "loss": 5.1042,
      "mean_token_accuracy": 0.19595118314027787,
      "num_tokens": 17445202.0,
      "step": 9475
    },
    {
      "entropy": 5.244723510742188,
      "epoch": 0.7918476445038423,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004943601050527493,
      "loss": 4.9907,
      "mean_token_accuracy": 0.2045651078224182,
      "num_tokens": 17453928.0,
      "step": 9480
    },
    {
      "entropy": 5.287481784820557,
      "epoch": 0.7922652856665553,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004943534794575081,
      "loss": 5.0321,
      "mean_token_accuracy": 0.20597010999917983,
      "num_tokens": 17462609.0,
      "step": 9485
    },
    {
      "entropy": 5.339219617843628,
      "epoch": 0.7926829268292683,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004943468500222203,
      "loss": 5.0352,
      "mean_token_accuracy": 0.1990457609295845,
      "num_tokens": 17471326.0,
      "step": 9490
    },
    {
      "entropy": 5.2154700756073,
      "epoch": 0.7931005679919813,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004943402167470019,
      "loss": 4.9888,
      "mean_token_accuracy": 0.20640814155340195,
      "num_tokens": 17480296.0,
      "step": 9495
    },
    {
      "entropy": 5.23523178100586,
      "epoch": 0.7935182091546943,
      "grad_norm": 1.015625,
      "learning_rate": 0.000494333579631969,
      "loss": 4.9404,
      "mean_token_accuracy": 0.2033233940601349,
      "num_tokens": 17489779.0,
      "step": 9500
    },
    {
      "entropy": 5.272789430618286,
      "epoch": 0.7939358503174073,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004943269386772376,
      "loss": 5.0477,
      "mean_token_accuracy": 0.1986575484275818,
      "num_tokens": 17499796.0,
      "step": 9505
    },
    {
      "entropy": 5.308971214294433,
      "epoch": 0.7943534914801202,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004943202938829243,
      "loss": 5.139,
      "mean_token_accuracy": 0.19535888880491256,
      "num_tokens": 17509223.0,
      "step": 9510
    },
    {
      "entropy": 5.297016954421997,
      "epoch": 0.7947711326428333,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004943136452491452,
      "loss": 5.048,
      "mean_token_accuracy": 0.1978476330637932,
      "num_tokens": 17518712.0,
      "step": 9515
    },
    {
      "entropy": 5.2445885181427006,
      "epoch": 0.7951887738055463,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004943069927760168,
      "loss": 4.9888,
      "mean_token_accuracy": 0.2070267915725708,
      "num_tokens": 17528648.0,
      "step": 9520
    },
    {
      "entropy": 5.166000318527222,
      "epoch": 0.7956064149682592,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004943003364636554,
      "loss": 4.9918,
      "mean_token_accuracy": 0.19829148948192596,
      "num_tokens": 17537859.0,
      "step": 9525
    },
    {
      "entropy": 5.325325441360474,
      "epoch": 0.7960240561309723,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004942936763121777,
      "loss": 5.1503,
      "mean_token_accuracy": 0.19210512340068817,
      "num_tokens": 17546438.0,
      "step": 9530
    },
    {
      "entropy": 5.248593950271607,
      "epoch": 0.7964416972936853,
      "grad_norm": 1.125,
      "learning_rate": 0.0004942870123217002,
      "loss": 4.961,
      "mean_token_accuracy": 0.21220777332782745,
      "num_tokens": 17554537.0,
      "step": 9535
    },
    {
      "entropy": 5.380687284469604,
      "epoch": 0.7968593384563982,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004942803444923398,
      "loss": 5.1763,
      "mean_token_accuracy": 0.1917584091424942,
      "num_tokens": 17564239.0,
      "step": 9540
    },
    {
      "entropy": 5.317457151412964,
      "epoch": 0.7972769796191113,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004942736728242127,
      "loss": 5.0527,
      "mean_token_accuracy": 0.2041368305683136,
      "num_tokens": 17574437.0,
      "step": 9545
    },
    {
      "entropy": 5.23683009147644,
      "epoch": 0.7976946207818243,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004942669973174363,
      "loss": 4.9923,
      "mean_token_accuracy": 0.2037113845348358,
      "num_tokens": 17583173.0,
      "step": 9550
    },
    {
      "entropy": 5.273801803588867,
      "epoch": 0.7981122619445372,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000494260317972127,
      "loss": 5.0398,
      "mean_token_accuracy": 0.198993518948555,
      "num_tokens": 17591731.0,
      "step": 9555
    },
    {
      "entropy": 5.31051197052002,
      "epoch": 0.7985299031072502,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004942536347884019,
      "loss": 5.0664,
      "mean_token_accuracy": 0.1984912484884262,
      "num_tokens": 17600778.0,
      "step": 9560
    },
    {
      "entropy": 5.221505260467529,
      "epoch": 0.7989475442699633,
      "grad_norm": 1.0,
      "learning_rate": 0.0004942469477663782,
      "loss": 5.0404,
      "mean_token_accuracy": 0.21418730020523072,
      "num_tokens": 17609470.0,
      "step": 9565
    },
    {
      "entropy": 5.224123954772949,
      "epoch": 0.7993651854326762,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004942402569061726,
      "loss": 4.9266,
      "mean_token_accuracy": 0.20924785733222961,
      "num_tokens": 17618434.0,
      "step": 9570
    },
    {
      "entropy": 5.201136922836303,
      "epoch": 0.7997828265953892,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004942335622079025,
      "loss": 5.0605,
      "mean_token_accuracy": 0.19973597824573516,
      "num_tokens": 17628170.0,
      "step": 9575
    },
    {
      "entropy": 5.231966876983643,
      "epoch": 0.8002004677581023,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004942268636716848,
      "loss": 4.982,
      "mean_token_accuracy": 0.2037805736064911,
      "num_tokens": 17637391.0,
      "step": 9580
    },
    {
      "entropy": 5.278302097320557,
      "epoch": 0.8006181089208152,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004942201612976372,
      "loss": 5.0697,
      "mean_token_accuracy": 0.20564866662025452,
      "num_tokens": 17647466.0,
      "step": 9585
    },
    {
      "entropy": 5.251586961746216,
      "epoch": 0.8010357500835282,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004942134550858767,
      "loss": 4.9943,
      "mean_token_accuracy": 0.21021819561719896,
      "num_tokens": 17656155.0,
      "step": 9590
    },
    {
      "entropy": 5.388947629928589,
      "epoch": 0.8014533912462413,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004942067450365208,
      "loss": 5.0768,
      "mean_token_accuracy": 0.20041406750679017,
      "num_tokens": 17665242.0,
      "step": 9595
    },
    {
      "entropy": 5.303630685806274,
      "epoch": 0.8018710324089542,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004942000311496871,
      "loss": 5.1021,
      "mean_token_accuracy": 0.20105288922786713,
      "num_tokens": 17674964.0,
      "step": 9600
    },
    {
      "entropy": 5.333201885223389,
      "epoch": 0.8022886735716672,
      "grad_norm": 1.0,
      "learning_rate": 0.0004941933134254928,
      "loss": 5.0977,
      "mean_token_accuracy": 0.1999899372458458,
      "num_tokens": 17683275.0,
      "step": 9605
    },
    {
      "entropy": 5.254815244674683,
      "epoch": 0.8027063147343803,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004941865918640559,
      "loss": 5.0495,
      "mean_token_accuracy": 0.19754400104284286,
      "num_tokens": 17691841.0,
      "step": 9610
    },
    {
      "entropy": 5.305461883544922,
      "epoch": 0.8031239558970932,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004941798664654937,
      "loss": 5.0032,
      "mean_token_accuracy": 0.20951214283704758,
      "num_tokens": 17701258.0,
      "step": 9615
    },
    {
      "entropy": 5.321374130249024,
      "epoch": 0.8035415970598062,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004941731372299241,
      "loss": 5.0943,
      "mean_token_accuracy": 0.20163422971963882,
      "num_tokens": 17711028.0,
      "step": 9620
    },
    {
      "entropy": 5.272528886795044,
      "epoch": 0.8039592382225192,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004941664041574651,
      "loss": 5.0362,
      "mean_token_accuracy": 0.21170716881752014,
      "num_tokens": 17720175.0,
      "step": 9625
    },
    {
      "entropy": 5.283416604995727,
      "epoch": 0.8043768793852322,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004941596672482343,
      "loss": 5.0069,
      "mean_token_accuracy": 0.19935645610094072,
      "num_tokens": 17730420.0,
      "step": 9630
    },
    {
      "entropy": 5.321971082687378,
      "epoch": 0.8047945205479452,
      "grad_norm": 1.0,
      "learning_rate": 0.0004941529265023498,
      "loss": 5.0844,
      "mean_token_accuracy": 0.19544242173433304,
      "num_tokens": 17739802.0,
      "step": 9635
    },
    {
      "entropy": 5.3311573505401615,
      "epoch": 0.8052121617106582,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004941461819199293,
      "loss": 4.9846,
      "mean_token_accuracy": 0.21251770853996277,
      "num_tokens": 17749450.0,
      "step": 9640
    },
    {
      "entropy": 5.418461513519287,
      "epoch": 0.8056298028733712,
      "grad_norm": 1.0,
      "learning_rate": 0.0004941394335010912,
      "loss": 5.1837,
      "mean_token_accuracy": 0.19382675290107726,
      "num_tokens": 17758733.0,
      "step": 9645
    },
    {
      "entropy": 5.230041980743408,
      "epoch": 0.8060474440360842,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004941326812459535,
      "loss": 5.0899,
      "mean_token_accuracy": 0.1982561782002449,
      "num_tokens": 17769037.0,
      "step": 9650
    },
    {
      "entropy": 5.266266202926635,
      "epoch": 0.8064650851987972,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004941259251546346,
      "loss": 5.0368,
      "mean_token_accuracy": 0.19729335308074952,
      "num_tokens": 17777603.0,
      "step": 9655
    },
    {
      "entropy": 5.27465853691101,
      "epoch": 0.8068827263615101,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004941191652272526,
      "loss": 5.02,
      "mean_token_accuracy": 0.19491201639175415,
      "num_tokens": 17787513.0,
      "step": 9660
    },
    {
      "entropy": 5.255717611312866,
      "epoch": 0.8073003675242232,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004941124014639258,
      "loss": 5.0293,
      "mean_token_accuracy": 0.20482870191335678,
      "num_tokens": 17796720.0,
      "step": 9665
    },
    {
      "entropy": 5.262913465499878,
      "epoch": 0.8077180086869362,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004941056338647726,
      "loss": 4.9895,
      "mean_token_accuracy": 0.20260584503412246,
      "num_tokens": 17806024.0,
      "step": 9670
    },
    {
      "entropy": 5.196496868133545,
      "epoch": 0.8081356498496491,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004940988624299116,
      "loss": 4.985,
      "mean_token_accuracy": 0.20613298416137696,
      "num_tokens": 17815247.0,
      "step": 9675
    },
    {
      "entropy": 5.333111000061035,
      "epoch": 0.8085532910123622,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004940920871594613,
      "loss": 5.0292,
      "mean_token_accuracy": 0.20776699781417846,
      "num_tokens": 17824024.0,
      "step": 9680
    },
    {
      "entropy": 5.254497814178467,
      "epoch": 0.8089709321750752,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004940853080535403,
      "loss": 4.9978,
      "mean_token_accuracy": 0.19875686764717101,
      "num_tokens": 17832782.0,
      "step": 9685
    },
    {
      "entropy": 5.257366752624511,
      "epoch": 0.8093885733377881,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004940785251122673,
      "loss": 4.9828,
      "mean_token_accuracy": 0.20972108095884323,
      "num_tokens": 17841961.0,
      "step": 9690
    },
    {
      "entropy": 5.284517097473144,
      "epoch": 0.8098062145005012,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004940717383357608,
      "loss": 5.0772,
      "mean_token_accuracy": 0.21276423931121827,
      "num_tokens": 17851364.0,
      "step": 9695
    },
    {
      "entropy": 5.434416055679321,
      "epoch": 0.8102238556632142,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004940649477241401,
      "loss": 5.1307,
      "mean_token_accuracy": 0.19261317253112792,
      "num_tokens": 17860871.0,
      "step": 9700
    },
    {
      "entropy": 5.290762138366699,
      "epoch": 0.8106414968259271,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004940581532775237,
      "loss": 5.0616,
      "mean_token_accuracy": 0.20308563262224197,
      "num_tokens": 17869687.0,
      "step": 9705
    },
    {
      "entropy": 5.2350787162780765,
      "epoch": 0.8110591379886402,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004940513549960306,
      "loss": 5.018,
      "mean_token_accuracy": 0.2017708033323288,
      "num_tokens": 17878815.0,
      "step": 9710
    },
    {
      "entropy": 5.23651876449585,
      "epoch": 0.8114767791513532,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004940445528797798,
      "loss": 4.9504,
      "mean_token_accuracy": 0.20518733710050582,
      "num_tokens": 17888302.0,
      "step": 9715
    },
    {
      "entropy": 5.235987997055053,
      "epoch": 0.8118944203140661,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004940377469288906,
      "loss": 4.9035,
      "mean_token_accuracy": 0.21200938522815704,
      "num_tokens": 17896807.0,
      "step": 9720
    },
    {
      "entropy": 5.276819038391113,
      "epoch": 0.8123120614767791,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004940309371434819,
      "loss": 4.9692,
      "mean_token_accuracy": 0.209599868953228,
      "num_tokens": 17905988.0,
      "step": 9725
    },
    {
      "entropy": 5.326155805587769,
      "epoch": 0.8127297026394922,
      "grad_norm": 1.0,
      "learning_rate": 0.0004940241235236731,
      "loss": 5.1068,
      "mean_token_accuracy": 0.19877904951572417,
      "num_tokens": 17915917.0,
      "step": 9730
    },
    {
      "entropy": 5.3488874435424805,
      "epoch": 0.8131473438022051,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004940173060695833,
      "loss": 5.0837,
      "mean_token_accuracy": 0.19751158803701402,
      "num_tokens": 17926158.0,
      "step": 9735
    },
    {
      "entropy": 5.26666030883789,
      "epoch": 0.8135649849649181,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004940104847813319,
      "loss": 5.0444,
      "mean_token_accuracy": 0.2014840841293335,
      "num_tokens": 17935488.0,
      "step": 9740
    },
    {
      "entropy": 5.261229181289673,
      "epoch": 0.8139826261276312,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004940036596590384,
      "loss": 4.9708,
      "mean_token_accuracy": 0.19963575303554534,
      "num_tokens": 17944850.0,
      "step": 9745
    },
    {
      "entropy": 5.239579677581787,
      "epoch": 0.8144002672903441,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004939968307028221,
      "loss": 5.0003,
      "mean_token_accuracy": 0.20468050986528397,
      "num_tokens": 17954796.0,
      "step": 9750
    },
    {
      "entropy": 5.231853580474853,
      "epoch": 0.8148179084530571,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004939899979128028,
      "loss": 4.992,
      "mean_token_accuracy": 0.2128840684890747,
      "num_tokens": 17963875.0,
      "step": 9755
    },
    {
      "entropy": 5.243713617324829,
      "epoch": 0.8152355496157702,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004939831612890999,
      "loss": 5.0256,
      "mean_token_accuracy": 0.20398732721805574,
      "num_tokens": 17972789.0,
      "step": 9760
    },
    {
      "entropy": 5.253567934036255,
      "epoch": 0.8156531907784831,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004939763208318332,
      "loss": 5.0269,
      "mean_token_accuracy": 0.20555491149425506,
      "num_tokens": 17982355.0,
      "step": 9765
    },
    {
      "entropy": 5.333679676055908,
      "epoch": 0.8160708319411961,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004939694765411225,
      "loss": 5.0576,
      "mean_token_accuracy": 0.19933992028236389,
      "num_tokens": 17992046.0,
      "step": 9770
    },
    {
      "entropy": 5.239472818374634,
      "epoch": 0.8164884731039092,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004939626284170874,
      "loss": 4.9715,
      "mean_token_accuracy": 0.21828967481851577,
      "num_tokens": 18001440.0,
      "step": 9775
    },
    {
      "entropy": 5.370194625854492,
      "epoch": 0.8169061142666221,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000493955776459848,
      "loss": 5.1077,
      "mean_token_accuracy": 0.20007715225219727,
      "num_tokens": 18010817.0,
      "step": 9780
    },
    {
      "entropy": 5.2855517864227295,
      "epoch": 0.8173237554293351,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004939489206695241,
      "loss": 5.0412,
      "mean_token_accuracy": 0.20111846327781677,
      "num_tokens": 18019547.0,
      "step": 9785
    },
    {
      "entropy": 5.2308591365814205,
      "epoch": 0.8177413965920481,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004939420610462359,
      "loss": 5.0549,
      "mean_token_accuracy": 0.20276312828063964,
      "num_tokens": 18029191.0,
      "step": 9790
    },
    {
      "entropy": 5.319355249404907,
      "epoch": 0.8181590377547611,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004939351975901034,
      "loss": 5.086,
      "mean_token_accuracy": 0.19925022423267363,
      "num_tokens": 18038831.0,
      "step": 9795
    },
    {
      "entropy": 5.273286914825439,
      "epoch": 0.8185766789174741,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004939283303012467,
      "loss": 4.9981,
      "mean_token_accuracy": 0.20840023458003998,
      "num_tokens": 18049243.0,
      "step": 9800
    },
    {
      "entropy": 5.241224193572998,
      "epoch": 0.8189943200801871,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000493921459179786,
      "loss": 4.963,
      "mean_token_accuracy": 0.20449223667383193,
      "num_tokens": 18057960.0,
      "step": 9805
    },
    {
      "entropy": 5.286530780792236,
      "epoch": 0.8194119612429,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004939145842258418,
      "loss": 5.0853,
      "mean_token_accuracy": 0.20025518983602525,
      "num_tokens": 18066677.0,
      "step": 9810
    },
    {
      "entropy": 5.305978870391845,
      "epoch": 0.8198296024056131,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004939077054395341,
      "loss": 5.0394,
      "mean_token_accuracy": 0.1984740287065506,
      "num_tokens": 18075830.0,
      "step": 9815
    },
    {
      "entropy": 5.312576007843018,
      "epoch": 0.8202472435683261,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004939008228209836,
      "loss": 4.9507,
      "mean_token_accuracy": 0.19946401715278625,
      "num_tokens": 18085335.0,
      "step": 9820
    },
    {
      "entropy": 5.2644330024719235,
      "epoch": 0.820664884731039,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004938939363703106,
      "loss": 5.0171,
      "mean_token_accuracy": 0.20073747485876084,
      "num_tokens": 18094531.0,
      "step": 9825
    },
    {
      "entropy": 5.207372570037842,
      "epoch": 0.8210825258937521,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004938870460876359,
      "loss": 4.9892,
      "mean_token_accuracy": 0.20587844401597977,
      "num_tokens": 18103593.0,
      "step": 9830
    },
    {
      "entropy": 5.299793815612793,
      "epoch": 0.8215001670564651,
      "grad_norm": 0.953125,
      "learning_rate": 0.00049388015197308,
      "loss": 5.0554,
      "mean_token_accuracy": 0.20679358392953873,
      "num_tokens": 18112826.0,
      "step": 9835
    },
    {
      "entropy": 5.32416467666626,
      "epoch": 0.821917808219178,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004938732540267635,
      "loss": 5.1112,
      "mean_token_accuracy": 0.19724980741739273,
      "num_tokens": 18121687.0,
      "step": 9840
    },
    {
      "entropy": 5.2879071712493895,
      "epoch": 0.8223354493818911,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004938663522488072,
      "loss": 4.9722,
      "mean_token_accuracy": 0.20885904729366303,
      "num_tokens": 18130071.0,
      "step": 9845
    },
    {
      "entropy": 5.340536260604859,
      "epoch": 0.8227530905446041,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004938594466393321,
      "loss": 5.1883,
      "mean_token_accuracy": 0.19404108077287674,
      "num_tokens": 18139077.0,
      "step": 9850
    },
    {
      "entropy": 5.327898645401001,
      "epoch": 0.823170731707317,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004938525371984588,
      "loss": 5.0655,
      "mean_token_accuracy": 0.2043651297688484,
      "num_tokens": 18147173.0,
      "step": 9855
    },
    {
      "entropy": 5.274912977218628,
      "epoch": 0.8235883728700301,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004938456239263085,
      "loss": 5.0252,
      "mean_token_accuracy": 0.20748190432786942,
      "num_tokens": 18157061.0,
      "step": 9860
    },
    {
      "entropy": 5.310442733764648,
      "epoch": 0.8240060140327431,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004938387068230021,
      "loss": 4.9661,
      "mean_token_accuracy": 0.20626132190227509,
      "num_tokens": 18166115.0,
      "step": 9865
    },
    {
      "entropy": 5.280793237686157,
      "epoch": 0.824423655195456,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004938317858886606,
      "loss": 4.9957,
      "mean_token_accuracy": 0.2028559386730194,
      "num_tokens": 18174978.0,
      "step": 9870
    },
    {
      "entropy": 5.269491672515869,
      "epoch": 0.824841296358169,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004938248611234055,
      "loss": 5.0566,
      "mean_token_accuracy": 0.20465637892484664,
      "num_tokens": 18184948.0,
      "step": 9875
    },
    {
      "entropy": 5.293338203430176,
      "epoch": 0.8252589375208821,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004938179325273577,
      "loss": 4.9999,
      "mean_token_accuracy": 0.20436828583478928,
      "num_tokens": 18193443.0,
      "step": 9880
    },
    {
      "entropy": 5.335393714904785,
      "epoch": 0.825676578683595,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004938110001006385,
      "loss": 4.9993,
      "mean_token_accuracy": 0.20793971419334412,
      "num_tokens": 18202512.0,
      "step": 9885
    },
    {
      "entropy": 5.307349491119385,
      "epoch": 0.826094219846308,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004938040638433694,
      "loss": 5.0343,
      "mean_token_accuracy": 0.20211439877748488,
      "num_tokens": 18212328.0,
      "step": 9890
    },
    {
      "entropy": 5.272178363800049,
      "epoch": 0.8265118610090211,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004937971237556719,
      "loss": 4.9765,
      "mean_token_accuracy": 0.2093791261315346,
      "num_tokens": 18221155.0,
      "step": 9895
    },
    {
      "entropy": 5.278063011169434,
      "epoch": 0.826929502171734,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004937901798376672,
      "loss": 5.0351,
      "mean_token_accuracy": 0.2059718996286392,
      "num_tokens": 18230604.0,
      "step": 9900
    },
    {
      "entropy": 5.30039005279541,
      "epoch": 0.827347143334447,
      "grad_norm": 1.0,
      "learning_rate": 0.0004937832320894772,
      "loss": 4.9647,
      "mean_token_accuracy": 0.20549430698156357,
      "num_tokens": 18239126.0,
      "step": 9905
    },
    {
      "entropy": 5.250355577468872,
      "epoch": 0.8277647844971601,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004937762805112233,
      "loss": 5.0082,
      "mean_token_accuracy": 0.2015585035085678,
      "num_tokens": 18248634.0,
      "step": 9910
    },
    {
      "entropy": 5.272612905502319,
      "epoch": 0.828182425659873,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004937693251030273,
      "loss": 5.1308,
      "mean_token_accuracy": 0.1937979906797409,
      "num_tokens": 18258500.0,
      "step": 9915
    },
    {
      "entropy": 5.33410382270813,
      "epoch": 0.828600066822586,
      "grad_norm": 0.953125,
      "learning_rate": 0.000493762365865011,
      "loss": 4.9563,
      "mean_token_accuracy": 0.20484836846590043,
      "num_tokens": 18268374.0,
      "step": 9920
    },
    {
      "entropy": 5.339998483657837,
      "epoch": 0.8290177079852991,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004937554027972961,
      "loss": 5.0868,
      "mean_token_accuracy": 0.19745043218135833,
      "num_tokens": 18277667.0,
      "step": 9925
    },
    {
      "entropy": 5.307043981552124,
      "epoch": 0.829435349148012,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004937484359000046,
      "loss": 5.0941,
      "mean_token_accuracy": 0.20086869299411775,
      "num_tokens": 18287516.0,
      "step": 9930
    },
    {
      "entropy": 5.2328293800354,
      "epoch": 0.829852990310725,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004937414651732584,
      "loss": 4.9448,
      "mean_token_accuracy": 0.2047938585281372,
      "num_tokens": 18296534.0,
      "step": 9935
    },
    {
      "entropy": 5.294785499572754,
      "epoch": 0.830270631473438,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004937344906171795,
      "loss": 5.0786,
      "mean_token_accuracy": 0.20491352379322053,
      "num_tokens": 18305743.0,
      "step": 9940
    },
    {
      "entropy": 5.366404628753662,
      "epoch": 0.830688272636151,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004937275122318902,
      "loss": 5.0874,
      "mean_token_accuracy": 0.20139606297016144,
      "num_tokens": 18315524.0,
      "step": 9945
    },
    {
      "entropy": 5.272686195373535,
      "epoch": 0.831105913798864,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004937205300175123,
      "loss": 4.9474,
      "mean_token_accuracy": 0.21011454313993455,
      "num_tokens": 18324280.0,
      "step": 9950
    },
    {
      "entropy": 5.330067253112793,
      "epoch": 0.831523554961577,
      "grad_norm": 1.0,
      "learning_rate": 0.0004937135439741682,
      "loss": 5.042,
      "mean_token_accuracy": 0.20593102425336837,
      "num_tokens": 18332661.0,
      "step": 9955
    },
    {
      "entropy": 5.308542728424072,
      "epoch": 0.83194119612429,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004937065541019806,
      "loss": 5.1414,
      "mean_token_accuracy": 0.19236773252487183,
      "num_tokens": 18342986.0,
      "step": 9960
    },
    {
      "entropy": 5.358943128585816,
      "epoch": 0.832358837287003,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004936995604010711,
      "loss": 5.0963,
      "mean_token_accuracy": 0.2027817502617836,
      "num_tokens": 18351873.0,
      "step": 9965
    },
    {
      "entropy": 5.39743938446045,
      "epoch": 0.832776478449716,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004936925628715628,
      "loss": 5.1518,
      "mean_token_accuracy": 0.19456523209810256,
      "num_tokens": 18362174.0,
      "step": 9970
    },
    {
      "entropy": 5.286879777908325,
      "epoch": 0.833194119612429,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004936855615135779,
      "loss": 4.9705,
      "mean_token_accuracy": 0.20991315692663193,
      "num_tokens": 18371272.0,
      "step": 9975
    },
    {
      "entropy": 5.2962902069091795,
      "epoch": 0.833611760775142,
      "grad_norm": 1.09375,
      "learning_rate": 0.000493678556327239,
      "loss": 5.0786,
      "mean_token_accuracy": 0.20607325285673142,
      "num_tokens": 18380526.0,
      "step": 9980
    },
    {
      "entropy": 5.26721773147583,
      "epoch": 0.834029401937855,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004936715473126688,
      "loss": 4.9859,
      "mean_token_accuracy": 0.2070852905511856,
      "num_tokens": 18390027.0,
      "step": 9985
    },
    {
      "entropy": 5.230661058425904,
      "epoch": 0.8344470431005679,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00049366453446999,
      "loss": 4.9849,
      "mean_token_accuracy": 0.2114887997508049,
      "num_tokens": 18399012.0,
      "step": 9990
    },
    {
      "entropy": 5.188094568252564,
      "epoch": 0.834864684263281,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004936575177993251,
      "loss": 4.9333,
      "mean_token_accuracy": 0.20420264303684235,
      "num_tokens": 18408891.0,
      "step": 9995
    },
    {
      "entropy": 5.258261251449585,
      "epoch": 0.835282325425994,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004936504973007973,
      "loss": 4.9366,
      "mean_token_accuracy": 0.21137538701295852,
      "num_tokens": 18417909.0,
      "step": 10000
    },
    {
      "entropy": 5.241727209091186,
      "epoch": 0.8356999665887069,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004936434729745294,
      "loss": 5.0016,
      "mean_token_accuracy": 0.206212642788887,
      "num_tokens": 18427250.0,
      "step": 10005
    },
    {
      "entropy": 5.306929254531861,
      "epoch": 0.83611760775142,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004936364448206442,
      "loss": 4.9913,
      "mean_token_accuracy": 0.20443292558193207,
      "num_tokens": 18435851.0,
      "step": 10010
    },
    {
      "entropy": 5.225045156478882,
      "epoch": 0.836535248914133,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004936294128392649,
      "loss": 4.9991,
      "mean_token_accuracy": 0.1996292605996132,
      "num_tokens": 18444535.0,
      "step": 10015
    },
    {
      "entropy": 5.297957277297973,
      "epoch": 0.8369528900768459,
      "grad_norm": 0.875,
      "learning_rate": 0.0004936223770305146,
      "loss": 5.1042,
      "mean_token_accuracy": 0.20112494081258775,
      "num_tokens": 18454453.0,
      "step": 10020
    },
    {
      "entropy": 5.385451316833496,
      "epoch": 0.837370531239559,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004936153373945165,
      "loss": 5.1364,
      "mean_token_accuracy": 0.19662432223558426,
      "num_tokens": 18465033.0,
      "step": 10025
    },
    {
      "entropy": 5.339344882965088,
      "epoch": 0.837788172402272,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004936082939313936,
      "loss": 5.0251,
      "mean_token_accuracy": 0.20364271402359008,
      "num_tokens": 18473786.0,
      "step": 10030
    },
    {
      "entropy": 5.212394666671753,
      "epoch": 0.8382058135649849,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004936012466412695,
      "loss": 4.9873,
      "mean_token_accuracy": 0.20008451044559478,
      "num_tokens": 18483282.0,
      "step": 10035
    },
    {
      "entropy": 5.393288612365723,
      "epoch": 0.838623454727698,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004935941955242674,
      "loss": 5.1451,
      "mean_token_accuracy": 0.19586424976587297,
      "num_tokens": 18492612.0,
      "step": 10040
    },
    {
      "entropy": 5.275668573379517,
      "epoch": 0.839041095890411,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004935871405805109,
      "loss": 4.9666,
      "mean_token_accuracy": 0.21411871463060378,
      "num_tokens": 18501120.0,
      "step": 10045
    },
    {
      "entropy": 5.260760164260864,
      "epoch": 0.8394587370531239,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004935800818101232,
      "loss": 5.0776,
      "mean_token_accuracy": 0.20053426623344422,
      "num_tokens": 18510613.0,
      "step": 10050
    },
    {
      "entropy": 5.276189851760864,
      "epoch": 0.8398763782158369,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004935730192132282,
      "loss": 4.9811,
      "mean_token_accuracy": 0.2050245374441147,
      "num_tokens": 18518854.0,
      "step": 10055
    },
    {
      "entropy": 5.360947227478027,
      "epoch": 0.84029401937855,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004935659527899493,
      "loss": 5.0636,
      "mean_token_accuracy": 0.20014589428901672,
      "num_tokens": 18527592.0,
      "step": 10060
    },
    {
      "entropy": 5.21107029914856,
      "epoch": 0.8407116605412629,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004935588825404104,
      "loss": 4.9253,
      "mean_token_accuracy": 0.21830373853445054,
      "num_tokens": 18536396.0,
      "step": 10065
    },
    {
      "entropy": 5.195116806030273,
      "epoch": 0.8411293017039759,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004935518084647351,
      "loss": 4.8994,
      "mean_token_accuracy": 0.21492627412080764,
      "num_tokens": 18545201.0,
      "step": 10070
    },
    {
      "entropy": 5.270626354217529,
      "epoch": 0.841546942866689,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004935447305630474,
      "loss": 5.0715,
      "mean_token_accuracy": 0.20447982996702194,
      "num_tokens": 18554944.0,
      "step": 10075
    },
    {
      "entropy": 5.258065605163575,
      "epoch": 0.8419645840294019,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004935376488354711,
      "loss": 4.9751,
      "mean_token_accuracy": 0.20073315799236296,
      "num_tokens": 18564642.0,
      "step": 10080
    },
    {
      "entropy": 5.293374252319336,
      "epoch": 0.8423822251921149,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004935305632821303,
      "loss": 4.9832,
      "mean_token_accuracy": 0.20730831027030944,
      "num_tokens": 18573897.0,
      "step": 10085
    },
    {
      "entropy": 5.255651473999023,
      "epoch": 0.842799866354828,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004935234739031488,
      "loss": 5.0428,
      "mean_token_accuracy": 0.20551187098026275,
      "num_tokens": 18583174.0,
      "step": 10090
    },
    {
      "entropy": 5.2545829772949215,
      "epoch": 0.8432175075175409,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004935163806986508,
      "loss": 4.9421,
      "mean_token_accuracy": 0.2107224076986313,
      "num_tokens": 18592540.0,
      "step": 10095
    },
    {
      "entropy": 5.243731451034546,
      "epoch": 0.8436351486802539,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004935092836687607,
      "loss": 4.9522,
      "mean_token_accuracy": 0.20884281098842622,
      "num_tokens": 18602743.0,
      "step": 10100
    },
    {
      "entropy": 5.198603534698487,
      "epoch": 0.8440527898429669,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004935021828136025,
      "loss": 4.9353,
      "mean_token_accuracy": 0.20954020619392394,
      "num_tokens": 18611544.0,
      "step": 10105
    },
    {
      "entropy": 5.353778076171875,
      "epoch": 0.8444704310056799,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004934950781333005,
      "loss": 5.0432,
      "mean_token_accuracy": 0.19934172183275223,
      "num_tokens": 18620841.0,
      "step": 10110
    },
    {
      "entropy": 5.280141592025757,
      "epoch": 0.8448880721683929,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004934879696279793,
      "loss": 5.0771,
      "mean_token_accuracy": 0.20008258521556854,
      "num_tokens": 18630890.0,
      "step": 10115
    },
    {
      "entropy": 5.353300428390503,
      "epoch": 0.8453057133311059,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004934808572977632,
      "loss": 5.0737,
      "mean_token_accuracy": 0.19840428084135056,
      "num_tokens": 18639768.0,
      "step": 10120
    },
    {
      "entropy": 5.226650142669678,
      "epoch": 0.8457233544938189,
      "grad_norm": 1.0,
      "learning_rate": 0.0004934737411427765,
      "loss": 4.897,
      "mean_token_accuracy": 0.20902929902076722,
      "num_tokens": 18649578.0,
      "step": 10125
    },
    {
      "entropy": 5.24458065032959,
      "epoch": 0.8461409956565319,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004934666211631442,
      "loss": 4.9991,
      "mean_token_accuracy": 0.2054453745484352,
      "num_tokens": 18658470.0,
      "step": 10130
    },
    {
      "entropy": 5.2624781131744385,
      "epoch": 0.8465586368192449,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004934594973589906,
      "loss": 4.9631,
      "mean_token_accuracy": 0.1968930944800377,
      "num_tokens": 18668037.0,
      "step": 10135
    },
    {
      "entropy": 5.271680545806885,
      "epoch": 0.8469762779819578,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004934523697304406,
      "loss": 5.0741,
      "mean_token_accuracy": 0.20337503552436828,
      "num_tokens": 18677841.0,
      "step": 10140
    },
    {
      "entropy": 5.352348756790161,
      "epoch": 0.8473939191446709,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004934452382776189,
      "loss": 5.0365,
      "mean_token_accuracy": 0.2007971450686455,
      "num_tokens": 18686244.0,
      "step": 10145
    },
    {
      "entropy": 5.307337856292724,
      "epoch": 0.8478115603073839,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004934381030006504,
      "loss": 5.0619,
      "mean_token_accuracy": 0.20351528972387314,
      "num_tokens": 18695129.0,
      "step": 10150
    },
    {
      "entropy": 5.178586959838867,
      "epoch": 0.8482292014700968,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00049343096389966,
      "loss": 4.8905,
      "mean_token_accuracy": 0.21907635778188705,
      "num_tokens": 18703566.0,
      "step": 10155
    },
    {
      "entropy": 5.323379755020142,
      "epoch": 0.8486468426328099,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004934238209747724,
      "loss": 5.0864,
      "mean_token_accuracy": 0.19676927477121353,
      "num_tokens": 18713616.0,
      "step": 10160
    },
    {
      "entropy": 5.335180854797363,
      "epoch": 0.8490644837955229,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004934166742261132,
      "loss": 5.0122,
      "mean_token_accuracy": 0.20417239665985107,
      "num_tokens": 18722851.0,
      "step": 10165
    },
    {
      "entropy": 5.301330852508545,
      "epoch": 0.8494821249582358,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000493409523653807,
      "loss": 5.0064,
      "mean_token_accuracy": 0.20715524405241012,
      "num_tokens": 18731961.0,
      "step": 10170
    },
    {
      "entropy": 5.334019327163697,
      "epoch": 0.8498997661209489,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004934023692579793,
      "loss": 5.1873,
      "mean_token_accuracy": 0.2004974365234375,
      "num_tokens": 18741267.0,
      "step": 10175
    },
    {
      "entropy": 5.3129641056060795,
      "epoch": 0.8503174072836619,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004933952110387553,
      "loss": 5.0273,
      "mean_token_accuracy": 0.2029574766755104,
      "num_tokens": 18750266.0,
      "step": 10180
    },
    {
      "entropy": 5.26100525856018,
      "epoch": 0.8507350484463748,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004933880489962602,
      "loss": 5.0046,
      "mean_token_accuracy": 0.20313961207866668,
      "num_tokens": 18759910.0,
      "step": 10185
    },
    {
      "entropy": 5.379255390167236,
      "epoch": 0.8511526896090879,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004933808831306194,
      "loss": 5.1562,
      "mean_token_accuracy": 0.19180163890123367,
      "num_tokens": 18769510.0,
      "step": 10190
    },
    {
      "entropy": 5.232353210449219,
      "epoch": 0.8515703307718009,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004933737134419583,
      "loss": 4.9657,
      "mean_token_accuracy": 0.20214039832353592,
      "num_tokens": 18779893.0,
      "step": 10195
    },
    {
      "entropy": 5.268087768554688,
      "epoch": 0.8519879719345138,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004933665399304025,
      "loss": 5.0153,
      "mean_token_accuracy": 0.20948371440172195,
      "num_tokens": 18788802.0,
      "step": 10200
    },
    {
      "entropy": 5.2449723243713375,
      "epoch": 0.8524056130972268,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004933593625960777,
      "loss": 4.9237,
      "mean_token_accuracy": 0.20579939782619477,
      "num_tokens": 18797154.0,
      "step": 10205
    },
    {
      "entropy": 5.280235481262207,
      "epoch": 0.8528232542599399,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004933521814391094,
      "loss": 5.0482,
      "mean_token_accuracy": 0.20732472836971283,
      "num_tokens": 18806151.0,
      "step": 10210
    },
    {
      "entropy": 5.333980226516724,
      "epoch": 0.8532408954226529,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004933449964596232,
      "loss": 4.9444,
      "mean_token_accuracy": 0.21336816102266312,
      "num_tokens": 18815162.0,
      "step": 10215
    },
    {
      "entropy": 5.221742963790893,
      "epoch": 0.8536585365853658,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004933378076577451,
      "loss": 4.9388,
      "mean_token_accuracy": 0.20443326532840728,
      "num_tokens": 18823676.0,
      "step": 10220
    },
    {
      "entropy": 5.340446329116821,
      "epoch": 0.8540761777480789,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004933306150336009,
      "loss": 5.2442,
      "mean_token_accuracy": 0.18791325837373735,
      "num_tokens": 18832991.0,
      "step": 10225
    },
    {
      "entropy": 5.416504812240601,
      "epoch": 0.8544938189107919,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004933234185873164,
      "loss": 5.1,
      "mean_token_accuracy": 0.1986204817891121,
      "num_tokens": 18842614.0,
      "step": 10230
    },
    {
      "entropy": 5.392265892028808,
      "epoch": 0.8549114600735048,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004933162183190177,
      "loss": 5.1452,
      "mean_token_accuracy": 0.19194109439849855,
      "num_tokens": 18852747.0,
      "step": 10235
    },
    {
      "entropy": 5.24609522819519,
      "epoch": 0.8553291012362179,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004933090142288308,
      "loss": 5.0058,
      "mean_token_accuracy": 0.20486726313829423,
      "num_tokens": 18862005.0,
      "step": 10240
    },
    {
      "entropy": 5.272196435928345,
      "epoch": 0.8557467423989309,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004933018063168819,
      "loss": 4.9713,
      "mean_token_accuracy": 0.21595027148723603,
      "num_tokens": 18870172.0,
      "step": 10245
    },
    {
      "entropy": 5.303324937820435,
      "epoch": 0.8561643835616438,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000493294594583297,
      "loss": 5.0527,
      "mean_token_accuracy": 0.20630045086145402,
      "num_tokens": 18879428.0,
      "step": 10250
    },
    {
      "entropy": 5.308460092544555,
      "epoch": 0.8565820247243568,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004932873790282026,
      "loss": 4.9874,
      "mean_token_accuracy": 0.21532808989286423,
      "num_tokens": 18888694.0,
      "step": 10255
    },
    {
      "entropy": 5.318673324584961,
      "epoch": 0.8569996658870699,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004932801596517248,
      "loss": 5.0443,
      "mean_token_accuracy": 0.20069730281829834,
      "num_tokens": 18898317.0,
      "step": 10260
    },
    {
      "entropy": 5.207644653320313,
      "epoch": 0.8574173070497828,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004932729364539901,
      "loss": 5.0444,
      "mean_token_accuracy": 0.20067122727632522,
      "num_tokens": 18906749.0,
      "step": 10265
    },
    {
      "entropy": 5.2886285305023195,
      "epoch": 0.8578349482124958,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004932657094351248,
      "loss": 4.9306,
      "mean_token_accuracy": 0.207022587954998,
      "num_tokens": 18915404.0,
      "step": 10270
    },
    {
      "entropy": 5.2965826988220215,
      "epoch": 0.8582525893752089,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004932584785952555,
      "loss": 5.0163,
      "mean_token_accuracy": 0.20459155291318892,
      "num_tokens": 18925512.0,
      "step": 10275
    },
    {
      "entropy": 5.2212800025939945,
      "epoch": 0.8586702305379218,
      "grad_norm": 1.125,
      "learning_rate": 0.000493251243934509,
      "loss": 4.9622,
      "mean_token_accuracy": 0.21580297499895096,
      "num_tokens": 18934278.0,
      "step": 10280
    },
    {
      "entropy": 5.299727725982666,
      "epoch": 0.8590878717006348,
      "grad_norm": 1.0,
      "learning_rate": 0.0004932440054530116,
      "loss": 5.0523,
      "mean_token_accuracy": 0.20032110810279846,
      "num_tokens": 18943126.0,
      "step": 10285
    },
    {
      "entropy": 5.336291885375976,
      "epoch": 0.8595055128633479,
      "grad_norm": 1.125,
      "learning_rate": 0.0004932367631508903,
      "loss": 5.0273,
      "mean_token_accuracy": 0.2028697431087494,
      "num_tokens": 18951716.0,
      "step": 10290
    },
    {
      "entropy": 5.266165065765381,
      "epoch": 0.8599231540260608,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004932295170282717,
      "loss": 5.1126,
      "mean_token_accuracy": 0.19898587465286255,
      "num_tokens": 18960692.0,
      "step": 10295
    },
    {
      "entropy": 5.273776817321777,
      "epoch": 0.8603407951887738,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004932222670852828,
      "loss": 5.0511,
      "mean_token_accuracy": 0.20923716872930526,
      "num_tokens": 18969813.0,
      "step": 10300
    },
    {
      "entropy": 5.290828609466553,
      "epoch": 0.8607584363514869,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004932150133220504,
      "loss": 4.8943,
      "mean_token_accuracy": 0.21326633393764496,
      "num_tokens": 18978536.0,
      "step": 10305
    },
    {
      "entropy": 5.260556364059449,
      "epoch": 0.8611760775141998,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004932077557387015,
      "loss": 4.9974,
      "mean_token_accuracy": 0.20963544696569442,
      "num_tokens": 18987637.0,
      "step": 10310
    },
    {
      "entropy": 5.166515636444092,
      "epoch": 0.8615937186769128,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004932004943353633,
      "loss": 4.8374,
      "mean_token_accuracy": 0.2099147066473961,
      "num_tokens": 18997010.0,
      "step": 10315
    },
    {
      "entropy": 5.2591334819793705,
      "epoch": 0.8620113598396258,
      "grad_norm": 1.0,
      "learning_rate": 0.0004931932291121627,
      "loss": 4.9981,
      "mean_token_accuracy": 0.20627745985984802,
      "num_tokens": 19005685.0,
      "step": 10320
    },
    {
      "entropy": 5.24616551399231,
      "epoch": 0.8624290010023388,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004931859600692269,
      "loss": 5.0082,
      "mean_token_accuracy": 0.20305403023958207,
      "num_tokens": 19014836.0,
      "step": 10325
    },
    {
      "entropy": 5.306601905822754,
      "epoch": 0.8628466421650518,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004931786872066833,
      "loss": 5.0424,
      "mean_token_accuracy": 0.20136996507644653,
      "num_tokens": 19024442.0,
      "step": 10330
    },
    {
      "entropy": 5.294920110702515,
      "epoch": 0.8632642833277648,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004931714105246592,
      "loss": 4.9929,
      "mean_token_accuracy": 0.2095285251736641,
      "num_tokens": 19033730.0,
      "step": 10335
    },
    {
      "entropy": 5.271550512313842,
      "epoch": 0.8636819244904778,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004931641300232819,
      "loss": 4.9234,
      "mean_token_accuracy": 0.2196030080318451,
      "num_tokens": 19042765.0,
      "step": 10340
    },
    {
      "entropy": 5.314964962005615,
      "epoch": 0.8640995656531908,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004931568457026789,
      "loss": 4.9377,
      "mean_token_accuracy": 0.20492388308048248,
      "num_tokens": 19051408.0,
      "step": 10345
    },
    {
      "entropy": 5.311112594604492,
      "epoch": 0.8645172068159038,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004931495575629778,
      "loss": 5.0615,
      "mean_token_accuracy": 0.19700039625167848,
      "num_tokens": 19061145.0,
      "step": 10350
    },
    {
      "entropy": 5.299002313613892,
      "epoch": 0.8649348479786167,
      "grad_norm": 0.953125,
      "learning_rate": 0.000493142265604306,
      "loss": 5.0583,
      "mean_token_accuracy": 0.19705703407526015,
      "num_tokens": 19070157.0,
      "step": 10355
    },
    {
      "entropy": 5.252147054672241,
      "epoch": 0.8653524891413298,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004931349698267914,
      "loss": 4.8488,
      "mean_token_accuracy": 0.21993613839149476,
      "num_tokens": 19078532.0,
      "step": 10360
    },
    {
      "entropy": 5.23076114654541,
      "epoch": 0.8657701303040428,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004931276702305615,
      "loss": 4.9285,
      "mean_token_accuracy": 0.21286221742630004,
      "num_tokens": 19087283.0,
      "step": 10365
    },
    {
      "entropy": 5.295848989486695,
      "epoch": 0.8661877714667557,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004931203668157441,
      "loss": 5.0992,
      "mean_token_accuracy": 0.2058971866965294,
      "num_tokens": 19097568.0,
      "step": 10370
    },
    {
      "entropy": 5.3535929203033445,
      "epoch": 0.8666054126294688,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004931130595824672,
      "loss": 5.071,
      "mean_token_accuracy": 0.19715282022953035,
      "num_tokens": 19107218.0,
      "step": 10375
    },
    {
      "entropy": 5.317653751373291,
      "epoch": 0.8670230537921818,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004931057485308587,
      "loss": 5.0325,
      "mean_token_accuracy": 0.20822248011827468,
      "num_tokens": 19116128.0,
      "step": 10380
    },
    {
      "entropy": 5.233696603775025,
      "epoch": 0.8674406949548947,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004930984336610465,
      "loss": 4.9906,
      "mean_token_accuracy": 0.2111443892121315,
      "num_tokens": 19124885.0,
      "step": 10385
    },
    {
      "entropy": 5.336559152603149,
      "epoch": 0.8678583361176078,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004930911149731587,
      "loss": 5.0589,
      "mean_token_accuracy": 0.20646348148584365,
      "num_tokens": 19134125.0,
      "step": 10390
    },
    {
      "entropy": 5.263502979278565,
      "epoch": 0.8682759772803208,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004930837924673233,
      "loss": 4.9486,
      "mean_token_accuracy": 0.21311840415000916,
      "num_tokens": 19141925.0,
      "step": 10395
    },
    {
      "entropy": 5.3171943664550785,
      "epoch": 0.8686936184430337,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004930764661436688,
      "loss": 5.0311,
      "mean_token_accuracy": 0.2061261147260666,
      "num_tokens": 19151440.0,
      "step": 10400
    },
    {
      "entropy": 5.283300113677979,
      "epoch": 0.8691112596057468,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004930691360023231,
      "loss": 5.1113,
      "mean_token_accuracy": 0.20212262719869614,
      "num_tokens": 19161447.0,
      "step": 10405
    },
    {
      "entropy": 5.303946304321289,
      "epoch": 0.8695289007684598,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004930618020434147,
      "loss": 5.036,
      "mean_token_accuracy": 0.19985048323869706,
      "num_tokens": 19170804.0,
      "step": 10410
    },
    {
      "entropy": 5.393318748474121,
      "epoch": 0.8699465419311727,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004930544642670721,
      "loss": 5.1016,
      "mean_token_accuracy": 0.20270383656024932,
      "num_tokens": 19179898.0,
      "step": 10415
    },
    {
      "entropy": 5.2495049953460695,
      "epoch": 0.8703641830938857,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004930471226734235,
      "loss": 4.9663,
      "mean_token_accuracy": 0.2098995104432106,
      "num_tokens": 19188919.0,
      "step": 10420
    },
    {
      "entropy": 5.305373430252075,
      "epoch": 0.8707818242565988,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004930397772625976,
      "loss": 5.0331,
      "mean_token_accuracy": 0.20676478296518325,
      "num_tokens": 19197745.0,
      "step": 10425
    },
    {
      "entropy": 5.282780313491822,
      "epoch": 0.8711994654193117,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000493032428034723,
      "loss": 5.031,
      "mean_token_accuracy": 0.2116889163851738,
      "num_tokens": 19206681.0,
      "step": 10430
    },
    {
      "entropy": 5.363125562667847,
      "epoch": 0.8716171065820247,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004930250749899282,
      "loss": 5.0356,
      "mean_token_accuracy": 0.2038559913635254,
      "num_tokens": 19215806.0,
      "step": 10435
    },
    {
      "entropy": 5.285507392883301,
      "epoch": 0.8720347477447378,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004930177181283422,
      "loss": 5.0599,
      "mean_token_accuracy": 0.20010551065206528,
      "num_tokens": 19224681.0,
      "step": 10440
    },
    {
      "entropy": 5.256174373626709,
      "epoch": 0.8724523889074507,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004930103574500934,
      "loss": 5.0412,
      "mean_token_accuracy": 0.20514822602272034,
      "num_tokens": 19232684.0,
      "step": 10445
    },
    {
      "entropy": 5.291001605987549,
      "epoch": 0.8728700300701637,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004930029929553109,
      "loss": 4.9533,
      "mean_token_accuracy": 0.20501193255186081,
      "num_tokens": 19242538.0,
      "step": 10450
    },
    {
      "entropy": 5.269996118545532,
      "epoch": 0.8732876712328768,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004929956246441237,
      "loss": 5.0039,
      "mean_token_accuracy": 0.20045612752437592,
      "num_tokens": 19251146.0,
      "step": 10455
    },
    {
      "entropy": 5.374539852142334,
      "epoch": 0.8737053123955897,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004929882525166608,
      "loss": 5.0361,
      "mean_token_accuracy": 0.2038322478532791,
      "num_tokens": 19259931.0,
      "step": 10460
    },
    {
      "entropy": 5.307744455337525,
      "epoch": 0.8741229535583027,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004929808765730509,
      "loss": 5.0525,
      "mean_token_accuracy": 0.20045846253633498,
      "num_tokens": 19268796.0,
      "step": 10465
    },
    {
      "entropy": 5.25492672920227,
      "epoch": 0.8745405947210158,
      "grad_norm": 1.0,
      "learning_rate": 0.0004929734968134235,
      "loss": 5.071,
      "mean_token_accuracy": 0.20343612283468246,
      "num_tokens": 19278396.0,
      "step": 10470
    },
    {
      "entropy": 5.373476839065551,
      "epoch": 0.8749582358837287,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004929661132379076,
      "loss": 5.0676,
      "mean_token_accuracy": 0.19851451218128205,
      "num_tokens": 19287914.0,
      "step": 10475
    },
    {
      "entropy": 5.336198043823242,
      "epoch": 0.8753758770464417,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004929587258466326,
      "loss": 5.0465,
      "mean_token_accuracy": 0.2025327652692795,
      "num_tokens": 19297304.0,
      "step": 10480
    },
    {
      "entropy": 5.319211626052857,
      "epoch": 0.8757935182091547,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004929513346397278,
      "loss": 4.9931,
      "mean_token_accuracy": 0.21152844578027724,
      "num_tokens": 19306245.0,
      "step": 10485
    },
    {
      "entropy": 5.291439247131348,
      "epoch": 0.8762111593718677,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004929439396173223,
      "loss": 5.0145,
      "mean_token_accuracy": 0.19745678454637527,
      "num_tokens": 19316160.0,
      "step": 10490
    },
    {
      "entropy": 5.284034061431885,
      "epoch": 0.8766288005345807,
      "grad_norm": 1.0625,
      "learning_rate": 0.000492936540779546,
      "loss": 5.08,
      "mean_token_accuracy": 0.20020265281200408,
      "num_tokens": 19325411.0,
      "step": 10495
    },
    {
      "entropy": 5.347635793685913,
      "epoch": 0.8770464416972937,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004929291381265281,
      "loss": 5.014,
      "mean_token_accuracy": 0.2030327007174492,
      "num_tokens": 19334297.0,
      "step": 10500
    },
    {
      "entropy": 5.374361181259156,
      "epoch": 0.8774640828600067,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004929217316583985,
      "loss": 5.12,
      "mean_token_accuracy": 0.19446600824594498,
      "num_tokens": 19344121.0,
      "step": 10505
    },
    {
      "entropy": 5.300461483001709,
      "epoch": 0.8778817240227197,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004929143213752865,
      "loss": 5.0492,
      "mean_token_accuracy": 0.2016256719827652,
      "num_tokens": 19353428.0,
      "step": 10510
    },
    {
      "entropy": 5.330984878540039,
      "epoch": 0.8782993651854327,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000492906907277322,
      "loss": 5.0841,
      "mean_token_accuracy": 0.20332512855529786,
      "num_tokens": 19362661.0,
      "step": 10515
    },
    {
      "entropy": 5.225698471069336,
      "epoch": 0.8787170063481456,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000492899489364635,
      "loss": 4.9051,
      "mean_token_accuracy": 0.2171364650130272,
      "num_tokens": 19371246.0,
      "step": 10520
    },
    {
      "entropy": 5.3087831974029545,
      "epoch": 0.8791346475108587,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004928920676373549,
      "loss": 4.9418,
      "mean_token_accuracy": 0.20454858243465424,
      "num_tokens": 19379777.0,
      "step": 10525
    },
    {
      "entropy": 5.2473431587219235,
      "epoch": 0.8795522886735717,
      "grad_norm": 0.953125,
      "learning_rate": 0.000492884642095612,
      "loss": 4.9946,
      "mean_token_accuracy": 0.2070219859480858,
      "num_tokens": 19388717.0,
      "step": 10530
    },
    {
      "entropy": 5.40709376335144,
      "epoch": 0.8799699298362846,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004928772127395361,
      "loss": 5.1077,
      "mean_token_accuracy": 0.19847869575023652,
      "num_tokens": 19397936.0,
      "step": 10535
    },
    {
      "entropy": 5.463192176818848,
      "epoch": 0.8803875709989977,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004928697795692575,
      "loss": 5.0899,
      "mean_token_accuracy": 0.1947287231683731,
      "num_tokens": 19406830.0,
      "step": 10540
    },
    {
      "entropy": 5.349688482284546,
      "epoch": 0.8808052121617107,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000492862342584906,
      "loss": 5.0911,
      "mean_token_accuracy": 0.20279779732227327,
      "num_tokens": 19415131.0,
      "step": 10545
    },
    {
      "entropy": 5.262214326858521,
      "epoch": 0.8812228533244236,
      "grad_norm": 1.0625,
      "learning_rate": 0.000492854901786612,
      "loss": 4.8713,
      "mean_token_accuracy": 0.21996891498565674,
      "num_tokens": 19423585.0,
      "step": 10550
    },
    {
      "entropy": 5.218292140960694,
      "epoch": 0.8816404944871367,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004928474571745058,
      "loss": 4.9192,
      "mean_token_accuracy": 0.21653331369161605,
      "num_tokens": 19431890.0,
      "step": 10555
    },
    {
      "entropy": 5.279499340057373,
      "epoch": 0.8820581356498497,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004928400087487175,
      "loss": 5.0696,
      "mean_token_accuracy": 0.1987017497420311,
      "num_tokens": 19441377.0,
      "step": 10560
    },
    {
      "entropy": 5.3249983310699465,
      "epoch": 0.8824757768125626,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004928325565093777,
      "loss": 5.0483,
      "mean_token_accuracy": 0.1995964840054512,
      "num_tokens": 19451003.0,
      "step": 10565
    },
    {
      "entropy": 5.2409998893737795,
      "epoch": 0.8828934179752757,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004928251004566167,
      "loss": 4.9563,
      "mean_token_accuracy": 0.207501183450222,
      "num_tokens": 19460166.0,
      "step": 10570
    },
    {
      "entropy": 5.2448389530181885,
      "epoch": 0.8833110591379887,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004928176405905653,
      "loss": 5.0313,
      "mean_token_accuracy": 0.20794627964496612,
      "num_tokens": 19469417.0,
      "step": 10575
    },
    {
      "entropy": 5.342432069778442,
      "epoch": 0.8837287003007016,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004928101769113537,
      "loss": 5.0801,
      "mean_token_accuracy": 0.21217771321535112,
      "num_tokens": 19479275.0,
      "step": 10580
    },
    {
      "entropy": 5.346053552627564,
      "epoch": 0.8841463414634146,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000492802709419113,
      "loss": 4.9882,
      "mean_token_accuracy": 0.2015756696462631,
      "num_tokens": 19488804.0,
      "step": 10585
    },
    {
      "entropy": 5.311376285552979,
      "epoch": 0.8845639826261277,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004927952381139735,
      "loss": 4.9751,
      "mean_token_accuracy": 0.20802096426486968,
      "num_tokens": 19497701.0,
      "step": 10590
    },
    {
      "entropy": 5.306021213531494,
      "epoch": 0.8849816237888406,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004927877629960663,
      "loss": 4.9882,
      "mean_token_accuracy": 0.20534731298685074,
      "num_tokens": 19508177.0,
      "step": 10595
    },
    {
      "entropy": 5.305554962158203,
      "epoch": 0.8853992649515536,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004927802840655221,
      "loss": 5.0492,
      "mean_token_accuracy": 0.2057444706559181,
      "num_tokens": 19516866.0,
      "step": 10600
    },
    {
      "entropy": 5.350468921661377,
      "epoch": 0.8858169061142667,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000492772801322472,
      "loss": 5.0558,
      "mean_token_accuracy": 0.19982658624649047,
      "num_tokens": 19525788.0,
      "step": 10605
    },
    {
      "entropy": 5.279517650604248,
      "epoch": 0.8862345472769796,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004927653147670468,
      "loss": 4.8855,
      "mean_token_accuracy": 0.20792837888002397,
      "num_tokens": 19534748.0,
      "step": 10610
    },
    {
      "entropy": 5.223420810699463,
      "epoch": 0.8866521884396926,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004927578243993777,
      "loss": 4.9598,
      "mean_token_accuracy": 0.2073904663324356,
      "num_tokens": 19544083.0,
      "step": 10615
    },
    {
      "entropy": 5.236402940750122,
      "epoch": 0.8870698296024057,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004927503302195957,
      "loss": 4.988,
      "mean_token_accuracy": 0.2053396910429001,
      "num_tokens": 19553978.0,
      "step": 10620
    },
    {
      "entropy": 5.360853099822998,
      "epoch": 0.8874874707651186,
      "grad_norm": 0.9375,
      "learning_rate": 0.000492742832227832,
      "loss": 5.0516,
      "mean_token_accuracy": 0.2062338411808014,
      "num_tokens": 19563265.0,
      "step": 10625
    },
    {
      "entropy": 5.390022563934326,
      "epoch": 0.8879051119278316,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000492735330424218,
      "loss": 5.0848,
      "mean_token_accuracy": 0.1915833294391632,
      "num_tokens": 19572904.0,
      "step": 10630
    },
    {
      "entropy": 5.341847324371338,
      "epoch": 0.8883227530905446,
      "grad_norm": 1.03125,
      "learning_rate": 0.000492727824808885,
      "loss": 5.0857,
      "mean_token_accuracy": 0.20433924496173858,
      "num_tokens": 19581680.0,
      "step": 10635
    },
    {
      "entropy": 5.205078268051148,
      "epoch": 0.8887403942532576,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004927203153819643,
      "loss": 4.9395,
      "mean_token_accuracy": 0.20329589992761612,
      "num_tokens": 19590885.0,
      "step": 10640
    },
    {
      "entropy": 5.213958740234375,
      "epoch": 0.8891580354159706,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004927128021435875,
      "loss": 4.9394,
      "mean_token_accuracy": 0.2143978789448738,
      "num_tokens": 19599901.0,
      "step": 10645
    },
    {
      "entropy": 5.353690052032471,
      "epoch": 0.8895756765786836,
      "grad_norm": 1.015625,
      "learning_rate": 0.000492705285093886,
      "loss": 5.0855,
      "mean_token_accuracy": 0.19590885192155838,
      "num_tokens": 19608745.0,
      "step": 10650
    },
    {
      "entropy": 5.174007701873779,
      "epoch": 0.8899933177413966,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004926977642329913,
      "loss": 4.8295,
      "mean_token_accuracy": 0.21452972441911697,
      "num_tokens": 19618418.0,
      "step": 10655
    },
    {
      "entropy": 5.186718845367432,
      "epoch": 0.8904109589041096,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004926902395610353,
      "loss": 4.9333,
      "mean_token_accuracy": 0.20177324265241622,
      "num_tokens": 19627118.0,
      "step": 10660
    },
    {
      "entropy": 5.335381174087525,
      "epoch": 0.8908286000668226,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004926827110781496,
      "loss": 5.093,
      "mean_token_accuracy": 0.2024550050497055,
      "num_tokens": 19636565.0,
      "step": 10665
    },
    {
      "entropy": 5.268762016296387,
      "epoch": 0.8912462412295356,
      "grad_norm": 1.0,
      "learning_rate": 0.000492675178784466,
      "loss": 4.9346,
      "mean_token_accuracy": 0.21084842085838318,
      "num_tokens": 19645038.0,
      "step": 10670
    },
    {
      "entropy": 5.224033355712891,
      "epoch": 0.8916638823922486,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004926676426801164,
      "loss": 4.9829,
      "mean_token_accuracy": 0.20760006904602052,
      "num_tokens": 19654088.0,
      "step": 10675
    },
    {
      "entropy": 5.282566213607788,
      "epoch": 0.8920815235549616,
      "grad_norm": 1.0,
      "learning_rate": 0.0004926601027652328,
      "loss": 5.0741,
      "mean_token_accuracy": 0.20213294625282288,
      "num_tokens": 19662551.0,
      "step": 10680
    },
    {
      "entropy": 5.362628698348999,
      "epoch": 0.8924991647176745,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000492652559039947,
      "loss": 5.0195,
      "mean_token_accuracy": 0.20249192416667938,
      "num_tokens": 19672405.0,
      "step": 10685
    },
    {
      "entropy": 5.269727277755737,
      "epoch": 0.8929168058803876,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004926450115043912,
      "loss": 5.007,
      "mean_token_accuracy": 0.2048602059483528,
      "num_tokens": 19681288.0,
      "step": 10690
    },
    {
      "entropy": 5.30151219367981,
      "epoch": 0.8933344470431006,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004926374601586975,
      "loss": 4.9237,
      "mean_token_accuracy": 0.2127156913280487,
      "num_tokens": 19690723.0,
      "step": 10695
    },
    {
      "entropy": 5.274267911911011,
      "epoch": 0.8937520882058135,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004926299050029981,
      "loss": 4.9776,
      "mean_token_accuracy": 0.20837395191192626,
      "num_tokens": 19699642.0,
      "step": 10700
    },
    {
      "entropy": 5.237767362594605,
      "epoch": 0.8941697293685266,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004926223460374251,
      "loss": 4.9385,
      "mean_token_accuracy": 0.21083255708217621,
      "num_tokens": 19707563.0,
      "step": 10705
    },
    {
      "entropy": 5.331913757324219,
      "epoch": 0.8945873705312396,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004926147832621111,
      "loss": 5.0665,
      "mean_token_accuracy": 0.20573477894067765,
      "num_tokens": 19717109.0,
      "step": 10710
    },
    {
      "entropy": 5.242582893371582,
      "epoch": 0.8950050116939525,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004926072166771884,
      "loss": 4.9059,
      "mean_token_accuracy": 0.2190503180027008,
      "num_tokens": 19725585.0,
      "step": 10715
    },
    {
      "entropy": 5.175770235061646,
      "epoch": 0.8954226528566656,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004925996462827895,
      "loss": 4.8592,
      "mean_token_accuracy": 0.2243387594819069,
      "num_tokens": 19733502.0,
      "step": 10720
    },
    {
      "entropy": 5.270907354354859,
      "epoch": 0.8958402940193786,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004925920720790466,
      "loss": 4.9973,
      "mean_token_accuracy": 0.20483084470033647,
      "num_tokens": 19742926.0,
      "step": 10725
    },
    {
      "entropy": 5.237758588790894,
      "epoch": 0.8962579351820915,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004925844940660928,
      "loss": 4.9647,
      "mean_token_accuracy": 0.21414320170879364,
      "num_tokens": 19752488.0,
      "step": 10730
    },
    {
      "entropy": 5.322755241394043,
      "epoch": 0.8966755763448045,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004925769122440606,
      "loss": 5.0992,
      "mean_token_accuracy": 0.19827325642108917,
      "num_tokens": 19762705.0,
      "step": 10735
    },
    {
      "entropy": 5.187228107452393,
      "epoch": 0.8970932175075176,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004925693266130824,
      "loss": 4.8871,
      "mean_token_accuracy": 0.21419924795627593,
      "num_tokens": 19771156.0,
      "step": 10740
    },
    {
      "entropy": 5.259825420379639,
      "epoch": 0.8975108586702305,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004925617371732914,
      "loss": 4.9431,
      "mean_token_accuracy": 0.20529670715332032,
      "num_tokens": 19780788.0,
      "step": 10745
    },
    {
      "entropy": 5.166040325164795,
      "epoch": 0.8979284998329435,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004925541439248203,
      "loss": 4.9225,
      "mean_token_accuracy": 0.21501104831695556,
      "num_tokens": 19790169.0,
      "step": 10750
    },
    {
      "entropy": 5.347001791000366,
      "epoch": 0.8983461409956566,
      "grad_norm": 1.0,
      "learning_rate": 0.000492546546867802,
      "loss": 5.1465,
      "mean_token_accuracy": 0.19041983932256698,
      "num_tokens": 19800249.0,
      "step": 10755
    },
    {
      "entropy": 5.291795492172241,
      "epoch": 0.8987637821583695,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004925389460023696,
      "loss": 4.9972,
      "mean_token_accuracy": 0.2081596866250038,
      "num_tokens": 19809797.0,
      "step": 10760
    },
    {
      "entropy": 5.205329132080078,
      "epoch": 0.8991814233210825,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004925313413286559,
      "loss": 4.8877,
      "mean_token_accuracy": 0.21147559881210326,
      "num_tokens": 19818826.0,
      "step": 10765
    },
    {
      "entropy": 5.2760485172271725,
      "epoch": 0.8995990644837956,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004925237328467945,
      "loss": 5.0444,
      "mean_token_accuracy": 0.20388191789388657,
      "num_tokens": 19827947.0,
      "step": 10770
    },
    {
      "entropy": 5.333606243133545,
      "epoch": 0.9000167056465085,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004925161205569182,
      "loss": 5.064,
      "mean_token_accuracy": 0.19826466143131255,
      "num_tokens": 19835734.0,
      "step": 10775
    },
    {
      "entropy": 5.370053768157959,
      "epoch": 0.9004343468092215,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004925085044591604,
      "loss": 5.1542,
      "mean_token_accuracy": 0.19305280596017838,
      "num_tokens": 19845175.0,
      "step": 10780
    },
    {
      "entropy": 5.310465002059937,
      "epoch": 0.9008519879719346,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004925008845536544,
      "loss": 4.9981,
      "mean_token_accuracy": 0.20881958901882172,
      "num_tokens": 19853684.0,
      "step": 10785
    },
    {
      "entropy": 5.289188432693481,
      "epoch": 0.9012696291346475,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004924932608405336,
      "loss": 4.9832,
      "mean_token_accuracy": 0.20111274421215058,
      "num_tokens": 19862813.0,
      "step": 10790
    },
    {
      "entropy": 5.274710369110108,
      "epoch": 0.9016872702973605,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004924856333199315,
      "loss": 5.0282,
      "mean_token_accuracy": 0.2073032170534134,
      "num_tokens": 19871545.0,
      "step": 10795
    },
    {
      "entropy": 5.298371839523315,
      "epoch": 0.9021049114600735,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004924780019919816,
      "loss": 4.9929,
      "mean_token_accuracy": 0.20241359919309615,
      "num_tokens": 19880060.0,
      "step": 10800
    },
    {
      "entropy": 5.269317197799682,
      "epoch": 0.9025225526227865,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004924703668568174,
      "loss": 5.032,
      "mean_token_accuracy": 0.20404421538114548,
      "num_tokens": 19889151.0,
      "step": 10805
    },
    {
      "entropy": 5.288854026794434,
      "epoch": 0.9029401937854995,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004924627279145728,
      "loss": 4.9762,
      "mean_token_accuracy": 0.20503151118755342,
      "num_tokens": 19898087.0,
      "step": 10810
    },
    {
      "entropy": 5.214474391937256,
      "epoch": 0.9033578349482125,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004924550851653814,
      "loss": 4.9851,
      "mean_token_accuracy": 0.20971933603286744,
      "num_tokens": 19906196.0,
      "step": 10815
    },
    {
      "entropy": 5.288983678817749,
      "epoch": 0.9037754761109255,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004924474386093769,
      "loss": 5.012,
      "mean_token_accuracy": 0.20628090500831603,
      "num_tokens": 19915837.0,
      "step": 10820
    },
    {
      "entropy": 5.234778690338135,
      "epoch": 0.9041931172736385,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004924397882466933,
      "loss": 4.9835,
      "mean_token_accuracy": 0.20756157636642455,
      "num_tokens": 19926012.0,
      "step": 10825
    },
    {
      "entropy": 5.2237465381622314,
      "epoch": 0.9046107584363515,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004924321340774644,
      "loss": 4.9641,
      "mean_token_accuracy": 0.2123892590403557,
      "num_tokens": 19935987.0,
      "step": 10830
    },
    {
      "entropy": 5.211829614639282,
      "epoch": 0.9050283995990644,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004924244761018243,
      "loss": 4.9785,
      "mean_token_accuracy": 0.2058514401316643,
      "num_tokens": 19945865.0,
      "step": 10835
    },
    {
      "entropy": 5.308677244186401,
      "epoch": 0.9054460407617775,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000492416814319907,
      "loss": 5.0841,
      "mean_token_accuracy": 0.2016024649143219,
      "num_tokens": 19954824.0,
      "step": 10840
    },
    {
      "entropy": 5.263192558288575,
      "epoch": 0.9058636819244905,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004924091487318468,
      "loss": 4.9871,
      "mean_token_accuracy": 0.2005952849984169,
      "num_tokens": 19964740.0,
      "step": 10845
    },
    {
      "entropy": 5.23516526222229,
      "epoch": 0.9062813230872034,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004924014793377776,
      "loss": 4.9429,
      "mean_token_accuracy": 0.20709327161312102,
      "num_tokens": 19974310.0,
      "step": 10850
    },
    {
      "entropy": 5.193000459671021,
      "epoch": 0.9066989642499165,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004923938061378339,
      "loss": 4.8737,
      "mean_token_accuracy": 0.2146912544965744,
      "num_tokens": 19982943.0,
      "step": 10855
    },
    {
      "entropy": 5.234940910339356,
      "epoch": 0.9071166054126295,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004923861291321499,
      "loss": 4.9702,
      "mean_token_accuracy": 0.2087484672665596,
      "num_tokens": 19991971.0,
      "step": 10860
    },
    {
      "entropy": 5.30410099029541,
      "epoch": 0.9075342465753424,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004923784483208601,
      "loss": 5.0361,
      "mean_token_accuracy": 0.20067485719919204,
      "num_tokens": 20001491.0,
      "step": 10865
    },
    {
      "entropy": 5.243232250213623,
      "epoch": 0.9079518877380555,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004923707637040989,
      "loss": 4.9083,
      "mean_token_accuracy": 0.2126132845878601,
      "num_tokens": 20010605.0,
      "step": 10870
    },
    {
      "entropy": 5.162050724029541,
      "epoch": 0.9083695289007685,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004923630752820008,
      "loss": 4.8012,
      "mean_token_accuracy": 0.2111753687262535,
      "num_tokens": 20019291.0,
      "step": 10875
    },
    {
      "entropy": 5.237993907928467,
      "epoch": 0.9087871700634814,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004923553830547004,
      "loss": 5.0799,
      "mean_token_accuracy": 0.1985035702586174,
      "num_tokens": 20029048.0,
      "step": 10880
    },
    {
      "entropy": 5.334658527374268,
      "epoch": 0.9092048112261945,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004923476870223324,
      "loss": 5.0329,
      "mean_token_accuracy": 0.1985561028122902,
      "num_tokens": 20037878.0,
      "step": 10885
    },
    {
      "entropy": 5.229171180725098,
      "epoch": 0.9096224523889075,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004923399871850315,
      "loss": 4.959,
      "mean_token_accuracy": 0.2098856970667839,
      "num_tokens": 20046662.0,
      "step": 10890
    },
    {
      "entropy": 5.182511234283448,
      "epoch": 0.9100400935516204,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004923322835429326,
      "loss": 4.9319,
      "mean_token_accuracy": 0.21091097295284272,
      "num_tokens": 20055329.0,
      "step": 10895
    },
    {
      "entropy": 5.1409111499786375,
      "epoch": 0.9104577347143334,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004923245760961704,
      "loss": 4.8402,
      "mean_token_accuracy": 0.2128286749124527,
      "num_tokens": 20064024.0,
      "step": 10900
    },
    {
      "entropy": 5.277817010879517,
      "epoch": 0.9108753758770465,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004923168648448799,
      "loss": 5.049,
      "mean_token_accuracy": 0.20034979432821273,
      "num_tokens": 20073022.0,
      "step": 10905
    },
    {
      "entropy": 5.321513700485229,
      "epoch": 0.9112930170397594,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004923091497891961,
      "loss": 5.0172,
      "mean_token_accuracy": 0.20084492564201356,
      "num_tokens": 20081495.0,
      "step": 10910
    },
    {
      "entropy": 5.301963186264038,
      "epoch": 0.9117106582024724,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004923014309292541,
      "loss": 4.9604,
      "mean_token_accuracy": 0.2124164879322052,
      "num_tokens": 20090545.0,
      "step": 10915
    },
    {
      "entropy": 5.267093801498413,
      "epoch": 0.9121282993651855,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000492293708265189,
      "loss": 5.0206,
      "mean_token_accuracy": 0.2065804421901703,
      "num_tokens": 20100236.0,
      "step": 10920
    },
    {
      "entropy": 5.188518762588501,
      "epoch": 0.9125459405278984,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004922859817971359,
      "loss": 4.9666,
      "mean_token_accuracy": 0.20653071999549866,
      "num_tokens": 20109522.0,
      "step": 10925
    },
    {
      "entropy": 5.229751443862915,
      "epoch": 0.9129635816906114,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004922782515252302,
      "loss": 4.9289,
      "mean_token_accuracy": 0.2103912800550461,
      "num_tokens": 20118248.0,
      "step": 10930
    },
    {
      "entropy": 5.188411521911621,
      "epoch": 0.9133812228533245,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004922705174496072,
      "loss": 4.9686,
      "mean_token_accuracy": 0.2095724582672119,
      "num_tokens": 20127096.0,
      "step": 10935
    },
    {
      "entropy": 5.223756551742554,
      "epoch": 0.9137988640160374,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004922627795704022,
      "loss": 4.959,
      "mean_token_accuracy": 0.20450431555509568,
      "num_tokens": 20136815.0,
      "step": 10940
    },
    {
      "entropy": 5.267538547515869,
      "epoch": 0.9142165051787504,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004922550378877507,
      "loss": 4.9547,
      "mean_token_accuracy": 0.2146147683262825,
      "num_tokens": 20145455.0,
      "step": 10945
    },
    {
      "entropy": 5.304619646072387,
      "epoch": 0.9146341463414634,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004922472924017882,
      "loss": 5.0454,
      "mean_token_accuracy": 0.20653569847345352,
      "num_tokens": 20156139.0,
      "step": 10950
    },
    {
      "entropy": 5.332718992233277,
      "epoch": 0.9150517875041764,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004922395431126506,
      "loss": 5.0321,
      "mean_token_accuracy": 0.20290612876415254,
      "num_tokens": 20166266.0,
      "step": 10955
    },
    {
      "entropy": 5.405002212524414,
      "epoch": 0.9154694286668894,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004922317900204732,
      "loss": 5.0697,
      "mean_token_accuracy": 0.20168160200119017,
      "num_tokens": 20175005.0,
      "step": 10960
    },
    {
      "entropy": 5.152327537536621,
      "epoch": 0.9158870698296024,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004922240331253916,
      "loss": 4.8899,
      "mean_token_accuracy": 0.21400400400161743,
      "num_tokens": 20183488.0,
      "step": 10965
    },
    {
      "entropy": 5.237144708633423,
      "epoch": 0.9163047109923154,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004922162724275421,
      "loss": 4.9845,
      "mean_token_accuracy": 0.21214784681797028,
      "num_tokens": 20193724.0,
      "step": 10970
    },
    {
      "entropy": 5.292639350891113,
      "epoch": 0.9167223521550284,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004922085079270602,
      "loss": 5.0531,
      "mean_token_accuracy": 0.2067160576581955,
      "num_tokens": 20203733.0,
      "step": 10975
    },
    {
      "entropy": 5.369188928604126,
      "epoch": 0.9171399933177414,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004922007396240819,
      "loss": 5.0516,
      "mean_token_accuracy": 0.20885550528764724,
      "num_tokens": 20213448.0,
      "step": 10980
    },
    {
      "entropy": 5.298404407501221,
      "epoch": 0.9175576344804544,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004921929675187433,
      "loss": 4.9697,
      "mean_token_accuracy": 0.20900634825229644,
      "num_tokens": 20222837.0,
      "step": 10985
    },
    {
      "entropy": 5.314679718017578,
      "epoch": 0.9179752756431674,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004921851916111802,
      "loss": 5.1505,
      "mean_token_accuracy": 0.19607130736112593,
      "num_tokens": 20232056.0,
      "step": 10990
    },
    {
      "entropy": 5.3548205375671385,
      "epoch": 0.9183929168058804,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004921774119015289,
      "loss": 5.0527,
      "mean_token_accuracy": 0.2025447055697441,
      "num_tokens": 20241052.0,
      "step": 10995
    },
    {
      "entropy": 5.355234575271607,
      "epoch": 0.9188105579685933,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004921696283899257,
      "loss": 5.0935,
      "mean_token_accuracy": 0.2025918334722519,
      "num_tokens": 20251015.0,
      "step": 11000
    },
    {
      "entropy": 5.249455356597901,
      "epoch": 0.9192281991313064,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004921618410765067,
      "loss": 4.9577,
      "mean_token_accuracy": 0.20590477138757707,
      "num_tokens": 20259388.0,
      "step": 11005
    },
    {
      "entropy": 5.277779054641724,
      "epoch": 0.9196458402940194,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004921540499614083,
      "loss": 4.9399,
      "mean_token_accuracy": 0.20603271275758744,
      "num_tokens": 20268146.0,
      "step": 11010
    },
    {
      "entropy": 5.2481752872467045,
      "epoch": 0.9200634814567323,
      "grad_norm": 1.0,
      "learning_rate": 0.0004921462550447669,
      "loss": 4.9156,
      "mean_token_accuracy": 0.21277791112661362,
      "num_tokens": 20277559.0,
      "step": 11015
    },
    {
      "entropy": 5.26139554977417,
      "epoch": 0.9204811226194454,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004921384563267189,
      "loss": 4.9533,
      "mean_token_accuracy": 0.2130708172917366,
      "num_tokens": 20285661.0,
      "step": 11020
    },
    {
      "entropy": 5.257532215118408,
      "epoch": 0.9208987637821584,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004921306538074008,
      "loss": 5.0513,
      "mean_token_accuracy": 0.2059824898838997,
      "num_tokens": 20294636.0,
      "step": 11025
    },
    {
      "entropy": 5.319442510604858,
      "epoch": 0.9213164049448713,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004921228474869493,
      "loss": 4.953,
      "mean_token_accuracy": 0.20831986218690873,
      "num_tokens": 20302372.0,
      "step": 11030
    },
    {
      "entropy": 5.273311376571655,
      "epoch": 0.9217340461075844,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000492115037365501,
      "loss": 4.9617,
      "mean_token_accuracy": 0.2055031180381775,
      "num_tokens": 20311974.0,
      "step": 11035
    },
    {
      "entropy": 5.249235343933106,
      "epoch": 0.9221516872702974,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004921072234431925,
      "loss": 4.9827,
      "mean_token_accuracy": 0.21112971156835555,
      "num_tokens": 20321720.0,
      "step": 11040
    },
    {
      "entropy": 5.2887818813323975,
      "epoch": 0.9225693284330103,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000492099405720161,
      "loss": 5.0391,
      "mean_token_accuracy": 0.20281946063041686,
      "num_tokens": 20331324.0,
      "step": 11045
    },
    {
      "entropy": 5.391857624053955,
      "epoch": 0.9229869695957233,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004920915841965429,
      "loss": 5.1151,
      "mean_token_accuracy": 0.19931060522794725,
      "num_tokens": 20340921.0,
      "step": 11050
    },
    {
      "entropy": 5.359329414367676,
      "epoch": 0.9234046107584364,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004920837588724753,
      "loss": 5.0069,
      "mean_token_accuracy": 0.2096709981560707,
      "num_tokens": 20349871.0,
      "step": 11055
    },
    {
      "entropy": 5.143516731262207,
      "epoch": 0.9238222519211493,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004920759297480953,
      "loss": 4.7915,
      "mean_token_accuracy": 0.21419969350099563,
      "num_tokens": 20359171.0,
      "step": 11060
    },
    {
      "entropy": 5.275776386260986,
      "epoch": 0.9242398930838623,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004920680968235397,
      "loss": 4.9928,
      "mean_token_accuracy": 0.21233939677476882,
      "num_tokens": 20367779.0,
      "step": 11065
    },
    {
      "entropy": 5.220272541046143,
      "epoch": 0.9246575342465754,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004920602600989459,
      "loss": 4.9241,
      "mean_token_accuracy": 0.21298634111881257,
      "num_tokens": 20377568.0,
      "step": 11070
    },
    {
      "entropy": 5.298672294616699,
      "epoch": 0.9250751754092883,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004920524195744509,
      "loss": 5.0542,
      "mean_token_accuracy": 0.2014471024274826,
      "num_tokens": 20385915.0,
      "step": 11075
    },
    {
      "entropy": 5.225722169876098,
      "epoch": 0.9254928165720013,
      "grad_norm": 1.0,
      "learning_rate": 0.0004920445752501921,
      "loss": 4.9039,
      "mean_token_accuracy": 0.2108586311340332,
      "num_tokens": 20394877.0,
      "step": 11080
    },
    {
      "entropy": 5.251455163955688,
      "epoch": 0.9259104577347144,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004920367271263066,
      "loss": 4.9299,
      "mean_token_accuracy": 0.20939255654811859,
      "num_tokens": 20403741.0,
      "step": 11085
    },
    {
      "entropy": 5.2638006687164305,
      "epoch": 0.9263280988974273,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004920288752029321,
      "loss": 4.9615,
      "mean_token_accuracy": 0.20568307489156723,
      "num_tokens": 20412702.0,
      "step": 11090
    },
    {
      "entropy": 5.280517959594727,
      "epoch": 0.9267457400601403,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004920210194802059,
      "loss": 5.0142,
      "mean_token_accuracy": 0.20845978409051896,
      "num_tokens": 20421776.0,
      "step": 11095
    },
    {
      "entropy": 5.311202049255371,
      "epoch": 0.9271633812228534,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004920131599582654,
      "loss": 5.1012,
      "mean_token_accuracy": 0.19238292872905732,
      "num_tokens": 20432204.0,
      "step": 11100
    },
    {
      "entropy": 5.296841669082641,
      "epoch": 0.9275810223855663,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004920052966372485,
      "loss": 4.9869,
      "mean_token_accuracy": 0.1969113454222679,
      "num_tokens": 20441344.0,
      "step": 11105
    },
    {
      "entropy": 5.251367712020874,
      "epoch": 0.9279986635482793,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004919974295172926,
      "loss": 4.9077,
      "mean_token_accuracy": 0.20976300537586212,
      "num_tokens": 20449800.0,
      "step": 11110
    },
    {
      "entropy": 5.215135765075684,
      "epoch": 0.9284163047109923,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004919895585985353,
      "loss": 4.863,
      "mean_token_accuracy": 0.2206546813249588,
      "num_tokens": 20458326.0,
      "step": 11115
    },
    {
      "entropy": 5.27689118385315,
      "epoch": 0.9288339458737053,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004919816838811149,
      "loss": 5.1288,
      "mean_token_accuracy": 0.1981542870402336,
      "num_tokens": 20468050.0,
      "step": 11120
    },
    {
      "entropy": 5.2480597496032715,
      "epoch": 0.9292515870364183,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004919738053651687,
      "loss": 4.8868,
      "mean_token_accuracy": 0.21272605061531066,
      "num_tokens": 20477046.0,
      "step": 11125
    },
    {
      "entropy": 5.259340715408325,
      "epoch": 0.9296692281991313,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000491965923050835,
      "loss": 4.9969,
      "mean_token_accuracy": 0.2060549408197403,
      "num_tokens": 20486228.0,
      "step": 11130
    },
    {
      "entropy": 5.291873550415039,
      "epoch": 0.9300868693618443,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004919580369382516,
      "loss": 5.009,
      "mean_token_accuracy": 0.2011019319295883,
      "num_tokens": 20494764.0,
      "step": 11135
    },
    {
      "entropy": 5.260359859466552,
      "epoch": 0.9305045105245573,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004919501470275567,
      "loss": 4.9213,
      "mean_token_accuracy": 0.20615689009428023,
      "num_tokens": 20502805.0,
      "step": 11140
    },
    {
      "entropy": 5.240499019622803,
      "epoch": 0.9309221516872703,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004919422533188882,
      "loss": 4.9792,
      "mean_token_accuracy": 0.20694670230150222,
      "num_tokens": 20512322.0,
      "step": 11145
    },
    {
      "entropy": 5.235571432113647,
      "epoch": 0.9313397928499833,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004919343558123845,
      "loss": 4.9638,
      "mean_token_accuracy": 0.2078326165676117,
      "num_tokens": 20520680.0,
      "step": 11150
    },
    {
      "entropy": 5.234818601608277,
      "epoch": 0.9317574340126963,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004919264545081838,
      "loss": 4.842,
      "mean_token_accuracy": 0.21795134246349335,
      "num_tokens": 20528864.0,
      "step": 11155
    },
    {
      "entropy": 5.2940545082092285,
      "epoch": 0.9321750751754093,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004919185494064244,
      "loss": 4.959,
      "mean_token_accuracy": 0.20560528934001923,
      "num_tokens": 20538195.0,
      "step": 11160
    },
    {
      "entropy": 5.22232494354248,
      "epoch": 0.9325927163381222,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004919106405072446,
      "loss": 4.9957,
      "mean_token_accuracy": 0.20476280748844147,
      "num_tokens": 20547312.0,
      "step": 11165
    },
    {
      "entropy": 5.264388179779052,
      "epoch": 0.9330103575008353,
      "grad_norm": 1.03125,
      "learning_rate": 0.000491902727810783,
      "loss": 5.0268,
      "mean_token_accuracy": 0.200617453455925,
      "num_tokens": 20557184.0,
      "step": 11170
    },
    {
      "entropy": 5.274028062820435,
      "epoch": 0.9334279986635483,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004918948113171781,
      "loss": 4.9695,
      "mean_token_accuracy": 0.20534661561250686,
      "num_tokens": 20565819.0,
      "step": 11175
    },
    {
      "entropy": 5.226150751113892,
      "epoch": 0.9338456398262612,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004918868910265684,
      "loss": 5.0176,
      "mean_token_accuracy": 0.20757012516260148,
      "num_tokens": 20574761.0,
      "step": 11180
    },
    {
      "entropy": 5.369115591049194,
      "epoch": 0.9342632809889743,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004918789669390927,
      "loss": 5.0795,
      "mean_token_accuracy": 0.1995453119277954,
      "num_tokens": 20584003.0,
      "step": 11185
    },
    {
      "entropy": 5.283097410202027,
      "epoch": 0.9346809221516873,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004918710390548895,
      "loss": 4.9224,
      "mean_token_accuracy": 0.20918429493904114,
      "num_tokens": 20593123.0,
      "step": 11190
    },
    {
      "entropy": 5.216720819473267,
      "epoch": 0.9350985633144002,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004918631073740977,
      "loss": 5.0013,
      "mean_token_accuracy": 0.20795248299837113,
      "num_tokens": 20602582.0,
      "step": 11195
    },
    {
      "entropy": 5.223920297622681,
      "epoch": 0.9355162044771133,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004918551718968562,
      "loss": 4.9187,
      "mean_token_accuracy": 0.21180672943592072,
      "num_tokens": 20611681.0,
      "step": 11200
    },
    {
      "entropy": 5.238240480422974,
      "epoch": 0.9359338456398263,
      "grad_norm": 1.015625,
      "learning_rate": 0.000491847232623304,
      "loss": 4.981,
      "mean_token_accuracy": 0.20732127428054808,
      "num_tokens": 20620671.0,
      "step": 11205
    },
    {
      "entropy": 5.2384175777435305,
      "epoch": 0.9363514868025392,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004918392895535798,
      "loss": 4.9615,
      "mean_token_accuracy": 0.20582193583250047,
      "num_tokens": 20629986.0,
      "step": 11210
    },
    {
      "entropy": 5.26754150390625,
      "epoch": 0.9367691279652522,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000491831342687823,
      "loss": 4.9497,
      "mean_token_accuracy": 0.20641645640134812,
      "num_tokens": 20639274.0,
      "step": 11215
    },
    {
      "entropy": 5.2580019474029545,
      "epoch": 0.9371867691279653,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004918233920261723,
      "loss": 4.9703,
      "mean_token_accuracy": 0.20302498787641526,
      "num_tokens": 20647753.0,
      "step": 11220
    },
    {
      "entropy": 5.182185363769531,
      "epoch": 0.9376044102906782,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004918154375687673,
      "loss": 4.8664,
      "mean_token_accuracy": 0.21534329801797866,
      "num_tokens": 20656226.0,
      "step": 11225
    },
    {
      "entropy": 5.176910543441773,
      "epoch": 0.9380220514533912,
      "grad_norm": 1.0,
      "learning_rate": 0.0004918074793157472,
      "loss": 4.8801,
      "mean_token_accuracy": 0.2116728663444519,
      "num_tokens": 20664682.0,
      "step": 11230
    },
    {
      "entropy": 5.321833705902099,
      "epoch": 0.9384396926161043,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004917995172672511,
      "loss": 4.9811,
      "mean_token_accuracy": 0.20719746947288514,
      "num_tokens": 20672909.0,
      "step": 11235
    },
    {
      "entropy": 5.205064058303833,
      "epoch": 0.9388573337788172,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004917915514234184,
      "loss": 4.821,
      "mean_token_accuracy": 0.2161825865507126,
      "num_tokens": 20681316.0,
      "step": 11240
    },
    {
      "entropy": 5.285470581054687,
      "epoch": 0.9392749749415302,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004917835817843887,
      "loss": 4.9256,
      "mean_token_accuracy": 0.20108587443828582,
      "num_tokens": 20690360.0,
      "step": 11245
    },
    {
      "entropy": 5.300791072845459,
      "epoch": 0.9396926161042433,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004917756083503014,
      "loss": 4.9965,
      "mean_token_accuracy": 0.20350388437509537,
      "num_tokens": 20700398.0,
      "step": 11250
    },
    {
      "entropy": 5.306829833984375,
      "epoch": 0.9401102572669562,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004917676311212962,
      "loss": 5.0788,
      "mean_token_accuracy": 0.20190684199333192,
      "num_tokens": 20709705.0,
      "step": 11255
    },
    {
      "entropy": 5.195009326934814,
      "epoch": 0.9405278984296692,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004917596500975127,
      "loss": 4.8957,
      "mean_token_accuracy": 0.21204507648944854,
      "num_tokens": 20717774.0,
      "step": 11260
    },
    {
      "entropy": 5.271906423568725,
      "epoch": 0.9409455395923823,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004917516652790907,
      "loss": 5.0433,
      "mean_token_accuracy": 0.2026720494031906,
      "num_tokens": 20725966.0,
      "step": 11265
    },
    {
      "entropy": 5.296416616439819,
      "epoch": 0.9413631807550952,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004917436766661697,
      "loss": 4.9413,
      "mean_token_accuracy": 0.20376110523939134,
      "num_tokens": 20735254.0,
      "step": 11270
    },
    {
      "entropy": 5.206459283828735,
      "epoch": 0.9417808219178082,
      "grad_norm": 1.109375,
      "learning_rate": 0.00049173568425889,
      "loss": 4.8569,
      "mean_token_accuracy": 0.2136695697903633,
      "num_tokens": 20743331.0,
      "step": 11275
    },
    {
      "entropy": 5.270145416259766,
      "epoch": 0.9421984630805212,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004917276880573911,
      "loss": 4.9709,
      "mean_token_accuracy": 0.1998789593577385,
      "num_tokens": 20753966.0,
      "step": 11280
    },
    {
      "entropy": 5.268353796005249,
      "epoch": 0.9426161042432342,
      "grad_norm": 1.0,
      "learning_rate": 0.0004917196880618132,
      "loss": 4.9215,
      "mean_token_accuracy": 0.21107981204986573,
      "num_tokens": 20762608.0,
      "step": 11285
    },
    {
      "entropy": 5.188342142105102,
      "epoch": 0.9430337454059472,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004917116842722962,
      "loss": 4.8839,
      "mean_token_accuracy": 0.20608496963977813,
      "num_tokens": 20771194.0,
      "step": 11290
    },
    {
      "entropy": 5.2690201759338375,
      "epoch": 0.9434513865686602,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004917036766889804,
      "loss": 4.9858,
      "mean_token_accuracy": 0.20828500837087632,
      "num_tokens": 20779700.0,
      "step": 11295
    },
    {
      "entropy": 5.312868213653564,
      "epoch": 0.9438690277313732,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004916956653120058,
      "loss": 5.0924,
      "mean_token_accuracy": 0.19759668856859208,
      "num_tokens": 20788117.0,
      "step": 11300
    },
    {
      "entropy": 5.278332710266113,
      "epoch": 0.9442866688940862,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004916876501415129,
      "loss": 5.0648,
      "mean_token_accuracy": 0.19761857092380525,
      "num_tokens": 20798041.0,
      "step": 11305
    },
    {
      "entropy": 5.265129566192627,
      "epoch": 0.9447043100567992,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004916796311776417,
      "loss": 5.0019,
      "mean_token_accuracy": 0.20152102112770082,
      "num_tokens": 20807423.0,
      "step": 11310
    },
    {
      "entropy": 5.241173171997071,
      "epoch": 0.9451219512195121,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004916716084205328,
      "loss": 4.9017,
      "mean_token_accuracy": 0.216687935590744,
      "num_tokens": 20816727.0,
      "step": 11315
    },
    {
      "entropy": 5.269484806060791,
      "epoch": 0.9455395923822252,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004916635818703266,
      "loss": 4.9118,
      "mean_token_accuracy": 0.20841801017522812,
      "num_tokens": 20824855.0,
      "step": 11320
    },
    {
      "entropy": 5.195207214355468,
      "epoch": 0.9459572335449382,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004916555515271635,
      "loss": 4.9396,
      "mean_token_accuracy": 0.2159525841474533,
      "num_tokens": 20833096.0,
      "step": 11325
    },
    {
      "entropy": 5.241073799133301,
      "epoch": 0.9463748747076511,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004916475173911844,
      "loss": 4.884,
      "mean_token_accuracy": 0.21156073063611985,
      "num_tokens": 20842398.0,
      "step": 11330
    },
    {
      "entropy": 5.382232284545898,
      "epoch": 0.9467925158703642,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004916394794625296,
      "loss": 5.0868,
      "mean_token_accuracy": 0.19999775439500808,
      "num_tokens": 20851745.0,
      "step": 11335
    },
    {
      "entropy": 5.411838865280151,
      "epoch": 0.9472101570330772,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004916314377413399,
      "loss": 4.9936,
      "mean_token_accuracy": 0.20086113214492798,
      "num_tokens": 20860935.0,
      "step": 11340
    },
    {
      "entropy": 5.293083906173706,
      "epoch": 0.9476277981957901,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004916233922277562,
      "loss": 4.9995,
      "mean_token_accuracy": 0.19883306473493575,
      "num_tokens": 20871088.0,
      "step": 11345
    },
    {
      "entropy": 5.235815095901489,
      "epoch": 0.9480454393585032,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004916153429219192,
      "loss": 5.0239,
      "mean_token_accuracy": 0.20588317960500718,
      "num_tokens": 20879313.0,
      "step": 11350
    },
    {
      "entropy": 5.261130094528198,
      "epoch": 0.9484630805212162,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004916072898239699,
      "loss": 4.9726,
      "mean_token_accuracy": 0.20892988443374633,
      "num_tokens": 20888392.0,
      "step": 11355
    },
    {
      "entropy": 5.350674629211426,
      "epoch": 0.9488807216839291,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004915992329340493,
      "loss": 5.0647,
      "mean_token_accuracy": 0.1929287150502205,
      "num_tokens": 20897456.0,
      "step": 11360
    },
    {
      "entropy": 5.2784641742706295,
      "epoch": 0.9492983628466422,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004915911722522984,
      "loss": 5.0061,
      "mean_token_accuracy": 0.20352528244256973,
      "num_tokens": 20906621.0,
      "step": 11365
    },
    {
      "entropy": 5.386155605316162,
      "epoch": 0.9497160040093552,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004915831077788583,
      "loss": 5.1602,
      "mean_token_accuracy": 0.20381270349025726,
      "num_tokens": 20916929.0,
      "step": 11370
    },
    {
      "entropy": 5.296943330764771,
      "epoch": 0.9501336451720681,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004915750395138702,
      "loss": 4.9869,
      "mean_token_accuracy": 0.20451260209083558,
      "num_tokens": 20926028.0,
      "step": 11375
    },
    {
      "entropy": 5.3175122261047365,
      "epoch": 0.9505512863347811,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004915669674574754,
      "loss": 5.0768,
      "mean_token_accuracy": 0.19539652168750762,
      "num_tokens": 20935516.0,
      "step": 11380
    },
    {
      "entropy": 5.229594945907593,
      "epoch": 0.9509689274974942,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004915588916098151,
      "loss": 4.8725,
      "mean_token_accuracy": 0.21338792145252228,
      "num_tokens": 20944100.0,
      "step": 11385
    },
    {
      "entropy": 5.223288345336914,
      "epoch": 0.9513865686602071,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004915508119710307,
      "loss": 5.0099,
      "mean_token_accuracy": 0.2048628196120262,
      "num_tokens": 20953733.0,
      "step": 11390
    },
    {
      "entropy": 5.3557572841644285,
      "epoch": 0.9518042098229201,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004915427285412636,
      "loss": 5.0193,
      "mean_token_accuracy": 0.1968952551484108,
      "num_tokens": 20962413.0,
      "step": 11395
    },
    {
      "entropy": 5.223087501525879,
      "epoch": 0.9522218509856332,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004915346413206555,
      "loss": 4.895,
      "mean_token_accuracy": 0.22014676332473754,
      "num_tokens": 20971172.0,
      "step": 11400
    },
    {
      "entropy": 5.230666017532348,
      "epoch": 0.9526394921483461,
      "grad_norm": 1.0,
      "learning_rate": 0.0004915265503093478,
      "loss": 4.9005,
      "mean_token_accuracy": 0.21012457609176635,
      "num_tokens": 20979450.0,
      "step": 11405
    },
    {
      "entropy": 5.236772918701172,
      "epoch": 0.9530571333110591,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004915184555074824,
      "loss": 4.9673,
      "mean_token_accuracy": 0.20498269200325012,
      "num_tokens": 20988518.0,
      "step": 11410
    },
    {
      "entropy": 5.366719436645508,
      "epoch": 0.9534747744737722,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004915103569152005,
      "loss": 5.1336,
      "mean_token_accuracy": 0.2005612999200821,
      "num_tokens": 20998721.0,
      "step": 11415
    },
    {
      "entropy": 5.361366081237793,
      "epoch": 0.9538924156364851,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004915022545326444,
      "loss": 5.0119,
      "mean_token_accuracy": 0.2097477599978447,
      "num_tokens": 21008007.0,
      "step": 11420
    },
    {
      "entropy": 5.248238801956177,
      "epoch": 0.9543100567991981,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004914941483599558,
      "loss": 5.1047,
      "mean_token_accuracy": 0.19974254369735717,
      "num_tokens": 21017199.0,
      "step": 11425
    },
    {
      "entropy": 5.174650478363037,
      "epoch": 0.9547276979619111,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004914860383972763,
      "loss": 4.8996,
      "mean_token_accuracy": 0.2149574264883995,
      "num_tokens": 21025571.0,
      "step": 11430
    },
    {
      "entropy": 5.309657096862793,
      "epoch": 0.9551453391246241,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004914779246447483,
      "loss": 4.9828,
      "mean_token_accuracy": 0.20705399364233018,
      "num_tokens": 21035040.0,
      "step": 11435
    },
    {
      "entropy": 5.299860334396362,
      "epoch": 0.9555629802873371,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004914698071025136,
      "loss": 5.0065,
      "mean_token_accuracy": 0.21097289472818376,
      "num_tokens": 21044018.0,
      "step": 11440
    },
    {
      "entropy": 5.208160161972046,
      "epoch": 0.9559806214500501,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004914616857707143,
      "loss": 4.91,
      "mean_token_accuracy": 0.2126847967505455,
      "num_tokens": 21052489.0,
      "step": 11445
    },
    {
      "entropy": 5.2636536121368405,
      "epoch": 0.9563982626127631,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004914535606494927,
      "loss": 4.9986,
      "mean_token_accuracy": 0.2124382182955742,
      "num_tokens": 21061153.0,
      "step": 11450
    },
    {
      "entropy": 5.214002513885498,
      "epoch": 0.9568159037754761,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000491445431738991,
      "loss": 4.8829,
      "mean_token_accuracy": 0.2135987475514412,
      "num_tokens": 21069671.0,
      "step": 11455
    },
    {
      "entropy": 5.3729626655578615,
      "epoch": 0.9572335449381891,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004914372990393515,
      "loss": 5.1634,
      "mean_token_accuracy": 0.19608964771032333,
      "num_tokens": 21079339.0,
      "step": 11460
    },
    {
      "entropy": 5.245329093933106,
      "epoch": 0.957651186100902,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004914291625507164,
      "loss": 4.8901,
      "mean_token_accuracy": 0.2120646744966507,
      "num_tokens": 21087868.0,
      "step": 11465
    },
    {
      "entropy": 5.287456369400024,
      "epoch": 0.9580688272636151,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004914210222732284,
      "loss": 4.9854,
      "mean_token_accuracy": 0.20944786369800567,
      "num_tokens": 21096258.0,
      "step": 11470
    },
    {
      "entropy": 5.272952890396118,
      "epoch": 0.9584864684263281,
      "grad_norm": 1.0,
      "learning_rate": 0.0004914128782070299,
      "loss": 4.9924,
      "mean_token_accuracy": 0.20144655257463456,
      "num_tokens": 21104504.0,
      "step": 11475
    },
    {
      "entropy": 5.263760757446289,
      "epoch": 0.958904109589041,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004914047303522636,
      "loss": 4.9356,
      "mean_token_accuracy": 0.20900096893310546,
      "num_tokens": 21114402.0,
      "step": 11480
    },
    {
      "entropy": 5.239157485961914,
      "epoch": 0.9593217507517541,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004913965787090718,
      "loss": 4.9367,
      "mean_token_accuracy": 0.21267976015806198,
      "num_tokens": 21123572.0,
      "step": 11485
    },
    {
      "entropy": 5.324149322509766,
      "epoch": 0.9597393919144671,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004913884232775974,
      "loss": 5.1544,
      "mean_token_accuracy": 0.1893838718533516,
      "num_tokens": 21132385.0,
      "step": 11490
    },
    {
      "entropy": 5.224599456787109,
      "epoch": 0.96015703307718,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004913802640579833,
      "loss": 4.9248,
      "mean_token_accuracy": 0.21170118302106858,
      "num_tokens": 21140383.0,
      "step": 11495
    },
    {
      "entropy": 5.310064029693604,
      "epoch": 0.9605746742398931,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004913721010503722,
      "loss": 5.0142,
      "mean_token_accuracy": 0.20528379529714585,
      "num_tokens": 21149793.0,
      "step": 11500
    },
    {
      "entropy": 5.30231466293335,
      "epoch": 0.9609923154026061,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000491363934254907,
      "loss": 4.9782,
      "mean_token_accuracy": 0.20120938122272491,
      "num_tokens": 21158955.0,
      "step": 11505
    },
    {
      "entropy": 5.312860536575317,
      "epoch": 0.961409956565319,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004913557636717307,
      "loss": 4.9761,
      "mean_token_accuracy": 0.20457708686590195,
      "num_tokens": 21167900.0,
      "step": 11510
    },
    {
      "entropy": 5.249720191955566,
      "epoch": 0.9618275977280321,
      "grad_norm": 1.0,
      "learning_rate": 0.0004913475893009864,
      "loss": 4.932,
      "mean_token_accuracy": 0.21308480501174926,
      "num_tokens": 21177243.0,
      "step": 11515
    },
    {
      "entropy": 5.2792277336120605,
      "epoch": 0.9622452388907451,
      "grad_norm": 1.109375,
      "learning_rate": 0.000491339411142817,
      "loss": 5.0145,
      "mean_token_accuracy": 0.2071506291627884,
      "num_tokens": 21186118.0,
      "step": 11520
    },
    {
      "entropy": 5.255590915679932,
      "epoch": 0.962662880053458,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004913312291973659,
      "loss": 5.0298,
      "mean_token_accuracy": 0.20723554790019988,
      "num_tokens": 21195032.0,
      "step": 11525
    },
    {
      "entropy": 5.250166654586792,
      "epoch": 0.963080521216171,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004913230434647762,
      "loss": 4.986,
      "mean_token_accuracy": 0.20382120609283447,
      "num_tokens": 21205388.0,
      "step": 11530
    },
    {
      "entropy": 5.220216703414917,
      "epoch": 0.9634981623788841,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004913148539451912,
      "loss": 4.8777,
      "mean_token_accuracy": 0.20997531712055206,
      "num_tokens": 21215181.0,
      "step": 11535
    },
    {
      "entropy": 5.163577222824097,
      "epoch": 0.963915803541597,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004913066606387544,
      "loss": 4.8171,
      "mean_token_accuracy": 0.21525585502386094,
      "num_tokens": 21223668.0,
      "step": 11540
    },
    {
      "entropy": 5.217168188095092,
      "epoch": 0.96433344470431,
      "grad_norm": 1.0625,
      "learning_rate": 0.000491298463545609,
      "loss": 4.9493,
      "mean_token_accuracy": 0.20610349923372268,
      "num_tokens": 21232448.0,
      "step": 11545
    },
    {
      "entropy": 5.230947875976563,
      "epoch": 0.9647510858670231,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004912902626658988,
      "loss": 4.926,
      "mean_token_accuracy": 0.20489526838064193,
      "num_tokens": 21240777.0,
      "step": 11550
    },
    {
      "entropy": 5.297998142242432,
      "epoch": 0.965168727029736,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004912820579997672,
      "loss": 5.0476,
      "mean_token_accuracy": 0.19644049406051636,
      "num_tokens": 21248803.0,
      "step": 11555
    },
    {
      "entropy": 5.352130460739136,
      "epoch": 0.965586368192449,
      "grad_norm": 1.125,
      "learning_rate": 0.0004912738495473577,
      "loss": 5.0071,
      "mean_token_accuracy": 0.20263514965772628,
      "num_tokens": 21257861.0,
      "step": 11560
    },
    {
      "entropy": 5.287946891784668,
      "epoch": 0.9660040093551621,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004912656373088142,
      "loss": 4.9234,
      "mean_token_accuracy": 0.21031851917505265,
      "num_tokens": 21267140.0,
      "step": 11565
    },
    {
      "entropy": 5.265135002136231,
      "epoch": 0.9664216505178751,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004912574212842804,
      "loss": 5.0342,
      "mean_token_accuracy": 0.20320686250925063,
      "num_tokens": 21276313.0,
      "step": 11570
    },
    {
      "entropy": 5.21587905883789,
      "epoch": 0.966839291680588,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004912492014739001,
      "loss": 4.9356,
      "mean_token_accuracy": 0.20771212130784988,
      "num_tokens": 21286156.0,
      "step": 11575
    },
    {
      "entropy": 5.3359276294708256,
      "epoch": 0.967256932843301,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004912409778778173,
      "loss": 5.0234,
      "mean_token_accuracy": 0.2112969547510147,
      "num_tokens": 21295420.0,
      "step": 11580
    },
    {
      "entropy": 5.350864028930664,
      "epoch": 0.9676745740060141,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004912327504961758,
      "loss": 5.1691,
      "mean_token_accuracy": 0.1923518806695938,
      "num_tokens": 21306292.0,
      "step": 11585
    },
    {
      "entropy": 5.335581636428833,
      "epoch": 0.968092215168727,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004912245193291198,
      "loss": 5.0213,
      "mean_token_accuracy": 0.2048722743988037,
      "num_tokens": 21315165.0,
      "step": 11590
    },
    {
      "entropy": 5.291277170181274,
      "epoch": 0.96850985633144,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004912162843767932,
      "loss": 5.0031,
      "mean_token_accuracy": 0.20899780988693237,
      "num_tokens": 21324608.0,
      "step": 11595
    },
    {
      "entropy": 5.307076168060303,
      "epoch": 0.9689274974941531,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004912080456393404,
      "loss": 4.9501,
      "mean_token_accuracy": 0.20400626808404923,
      "num_tokens": 21333609.0,
      "step": 11600
    },
    {
      "entropy": 5.332191705703735,
      "epoch": 0.969345138656866,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004911998031169056,
      "loss": 5.0634,
      "mean_token_accuracy": 0.1998114451766014,
      "num_tokens": 21344397.0,
      "step": 11605
    },
    {
      "entropy": 5.3374831199646,
      "epoch": 0.969762779819579,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004911915568096328,
      "loss": 5.0802,
      "mean_token_accuracy": 0.19303731471300126,
      "num_tokens": 21354111.0,
      "step": 11610
    },
    {
      "entropy": 5.190495920181275,
      "epoch": 0.9701804209822921,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004911833067176666,
      "loss": 4.8521,
      "mean_token_accuracy": 0.2119642823934555,
      "num_tokens": 21362807.0,
      "step": 11615
    },
    {
      "entropy": 5.308406639099121,
      "epoch": 0.970598062145005,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004911750528411515,
      "loss": 5.0432,
      "mean_token_accuracy": 0.2009720668196678,
      "num_tokens": 21372609.0,
      "step": 11620
    },
    {
      "entropy": 5.242914533615112,
      "epoch": 0.971015703307718,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004911667951802318,
      "loss": 4.9581,
      "mean_token_accuracy": 0.2107187867164612,
      "num_tokens": 21381182.0,
      "step": 11625
    },
    {
      "entropy": 5.30023078918457,
      "epoch": 0.9714333444704311,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004911585337350524,
      "loss": 4.9636,
      "mean_token_accuracy": 0.21112942695617676,
      "num_tokens": 21389930.0,
      "step": 11630
    },
    {
      "entropy": 5.277057695388794,
      "epoch": 0.971850985633144,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004911502685057573,
      "loss": 5.0151,
      "mean_token_accuracy": 0.20390140861272812,
      "num_tokens": 21399315.0,
      "step": 11635
    },
    {
      "entropy": 5.277867984771729,
      "epoch": 0.972268626795857,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004911419994924918,
      "loss": 5.0177,
      "mean_token_accuracy": 0.20502617210149765,
      "num_tokens": 21408857.0,
      "step": 11640
    },
    {
      "entropy": 5.330259799957275,
      "epoch": 0.97268626795857,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004911337266954004,
      "loss": 4.9454,
      "mean_token_accuracy": 0.21021246016025544,
      "num_tokens": 21417452.0,
      "step": 11645
    },
    {
      "entropy": 5.256307363510132,
      "epoch": 0.973103909121283,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004911254501146279,
      "loss": 4.9839,
      "mean_token_accuracy": 0.20891269892454148,
      "num_tokens": 21427315.0,
      "step": 11650
    },
    {
      "entropy": 5.271956968307495,
      "epoch": 0.973521550283996,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004911171697503193,
      "loss": 5.0376,
      "mean_token_accuracy": 0.20515555441379546,
      "num_tokens": 21437080.0,
      "step": 11655
    },
    {
      "entropy": 5.262950611114502,
      "epoch": 0.973939191446709,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004911088856026194,
      "loss": 4.8986,
      "mean_token_accuracy": 0.22233529537916183,
      "num_tokens": 21445421.0,
      "step": 11660
    },
    {
      "entropy": 5.262600421905518,
      "epoch": 0.974356832609422,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004911005976716734,
      "loss": 4.9122,
      "mean_token_accuracy": 0.2170039564371109,
      "num_tokens": 21454684.0,
      "step": 11665
    },
    {
      "entropy": 5.140411996841431,
      "epoch": 0.974774473772135,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004910923059576264,
      "loss": 4.8884,
      "mean_token_accuracy": 0.20958834737539292,
      "num_tokens": 21464287.0,
      "step": 11670
    },
    {
      "entropy": 5.236719369888306,
      "epoch": 0.975192114934848,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004910840104606234,
      "loss": 4.9993,
      "mean_token_accuracy": 0.2102099686861038,
      "num_tokens": 21472835.0,
      "step": 11675
    },
    {
      "entropy": 5.268878602981568,
      "epoch": 0.975609756097561,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004910757111808097,
      "loss": 4.9034,
      "mean_token_accuracy": 0.21437644064426423,
      "num_tokens": 21482327.0,
      "step": 11680
    },
    {
      "entropy": 5.255350828170776,
      "epoch": 0.976027397260274,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004910674081183305,
      "loss": 5.0051,
      "mean_token_accuracy": 0.20317334979772567,
      "num_tokens": 21492257.0,
      "step": 11685
    },
    {
      "entropy": 5.231694459915161,
      "epoch": 0.976445038422987,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004910591012733313,
      "loss": 4.9921,
      "mean_token_accuracy": 0.2080538749694824,
      "num_tokens": 21501930.0,
      "step": 11690
    },
    {
      "entropy": 5.336586809158325,
      "epoch": 0.9768626795857,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004910507906459575,
      "loss": 4.9808,
      "mean_token_accuracy": 0.20149205923080443,
      "num_tokens": 21511924.0,
      "step": 11695
    },
    {
      "entropy": 5.302269411087036,
      "epoch": 0.977280320748413,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004910424762363545,
      "loss": 5.0089,
      "mean_token_accuracy": 0.19976255893707276,
      "num_tokens": 21522279.0,
      "step": 11700
    },
    {
      "entropy": 5.226699876785278,
      "epoch": 0.977697961911126,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000491034158044668,
      "loss": 4.9908,
      "mean_token_accuracy": 0.20622302889823912,
      "num_tokens": 21532016.0,
      "step": 11705
    },
    {
      "entropy": 5.300827693939209,
      "epoch": 0.9781156030738389,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004910258360710434,
      "loss": 5.0415,
      "mean_token_accuracy": 0.19981015622615814,
      "num_tokens": 21541952.0,
      "step": 11710
    },
    {
      "entropy": 5.246514701843262,
      "epoch": 0.978533244236552,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004910175103156266,
      "loss": 4.8768,
      "mean_token_accuracy": 0.21614387780427932,
      "num_tokens": 21550185.0,
      "step": 11715
    },
    {
      "entropy": 5.294158792495727,
      "epoch": 0.978950885399265,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004910091807785632,
      "loss": 5.0132,
      "mean_token_accuracy": 0.20927384048700332,
      "num_tokens": 21558891.0,
      "step": 11720
    },
    {
      "entropy": 5.269440746307373,
      "epoch": 0.9793685265619779,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004910008474599991,
      "loss": 4.9783,
      "mean_token_accuracy": 0.2034325048327446,
      "num_tokens": 21568808.0,
      "step": 11725
    },
    {
      "entropy": 5.333498001098633,
      "epoch": 0.979786167724691,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004909925103600801,
      "loss": 5.0377,
      "mean_token_accuracy": 0.2012663796544075,
      "num_tokens": 21578838.0,
      "step": 11730
    },
    {
      "entropy": 5.1908571243286135,
      "epoch": 0.980203808887404,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004909841694789523,
      "loss": 4.9314,
      "mean_token_accuracy": 0.2071302816271782,
      "num_tokens": 21588008.0,
      "step": 11735
    },
    {
      "entropy": 5.270032215118408,
      "epoch": 0.9806214500501169,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004909758248167616,
      "loss": 5.0265,
      "mean_token_accuracy": 0.2073525533080101,
      "num_tokens": 21597036.0,
      "step": 11740
    },
    {
      "entropy": 5.294289016723633,
      "epoch": 0.98103909121283,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004909674763736542,
      "loss": 4.9731,
      "mean_token_accuracy": 0.2081419751048088,
      "num_tokens": 21606028.0,
      "step": 11745
    },
    {
      "entropy": 5.348976993560791,
      "epoch": 0.981456732375543,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004909591241497761,
      "loss": 5.0425,
      "mean_token_accuracy": 0.20958716720342635,
      "num_tokens": 21615377.0,
      "step": 11750
    },
    {
      "entropy": 5.367942953109742,
      "epoch": 0.9818743735382559,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004909507681452736,
      "loss": 5.1378,
      "mean_token_accuracy": 0.19741073995828629,
      "num_tokens": 21625751.0,
      "step": 11755
    },
    {
      "entropy": 5.389093446731567,
      "epoch": 0.9822920147009689,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004909424083602929,
      "loss": 5.0814,
      "mean_token_accuracy": 0.20286806374788285,
      "num_tokens": 21634825.0,
      "step": 11760
    },
    {
      "entropy": 5.315027570724487,
      "epoch": 0.982709655863682,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004909340447949805,
      "loss": 4.9359,
      "mean_token_accuracy": 0.2122182711958885,
      "num_tokens": 21644286.0,
      "step": 11765
    },
    {
      "entropy": 5.294584941864014,
      "epoch": 0.9831272970263949,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004909256774494827,
      "loss": 5.0434,
      "mean_token_accuracy": 0.20362290889024734,
      "num_tokens": 21653384.0,
      "step": 11770
    },
    {
      "entropy": 5.313011121749878,
      "epoch": 0.9835449381891079,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000490917306323946,
      "loss": 4.9565,
      "mean_token_accuracy": 0.2056444078683853,
      "num_tokens": 21662126.0,
      "step": 11775
    },
    {
      "entropy": 5.287021970748901,
      "epoch": 0.983962579351821,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004909089314185169,
      "loss": 4.9569,
      "mean_token_accuracy": 0.20453115105628966,
      "num_tokens": 21671940.0,
      "step": 11780
    },
    {
      "entropy": 5.181703805923462,
      "epoch": 0.9843802205145339,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004909005527333421,
      "loss": 4.8734,
      "mean_token_accuracy": 0.21516526639461517,
      "num_tokens": 21680928.0,
      "step": 11785
    },
    {
      "entropy": 5.207328605651855,
      "epoch": 0.9847978616772469,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004908921702685683,
      "loss": 4.9668,
      "mean_token_accuracy": 0.2089105486869812,
      "num_tokens": 21689712.0,
      "step": 11790
    },
    {
      "entropy": 5.283082485198975,
      "epoch": 0.98521550283996,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004908837840243421,
      "loss": 4.9339,
      "mean_token_accuracy": 0.20767928659915924,
      "num_tokens": 21698325.0,
      "step": 11795
    },
    {
      "entropy": 5.233205127716064,
      "epoch": 0.9856331440026729,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004908753940008105,
      "loss": 4.8961,
      "mean_token_accuracy": 0.21600585877895356,
      "num_tokens": 21707607.0,
      "step": 11800
    },
    {
      "entropy": 5.166473340988159,
      "epoch": 0.9860507851653859,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004908670001981201,
      "loss": 4.8954,
      "mean_token_accuracy": 0.21335337460041046,
      "num_tokens": 21715887.0,
      "step": 11805
    },
    {
      "entropy": 5.254672861099243,
      "epoch": 0.986468426328099,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004908586026164182,
      "loss": 5.0071,
      "mean_token_accuracy": 0.20970868766307832,
      "num_tokens": 21724992.0,
      "step": 11810
    },
    {
      "entropy": 5.317915964126587,
      "epoch": 0.9868860674908119,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004908502012558516,
      "loss": 4.9897,
      "mean_token_accuracy": 0.2065713718533516,
      "num_tokens": 21734317.0,
      "step": 11815
    },
    {
      "entropy": 5.323392629623413,
      "epoch": 0.9873037086535249,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004908417961165674,
      "loss": 5.0535,
      "mean_token_accuracy": 0.19670185595750808,
      "num_tokens": 21743014.0,
      "step": 11820
    },
    {
      "entropy": 5.324555921554565,
      "epoch": 0.9877213498162379,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004908333871987126,
      "loss": 4.9681,
      "mean_token_accuracy": 0.20581422001123428,
      "num_tokens": 21751058.0,
      "step": 11825
    },
    {
      "entropy": 5.335291385650635,
      "epoch": 0.9881389909789509,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004908249745024347,
      "loss": 5.0561,
      "mean_token_accuracy": 0.2071997717022896,
      "num_tokens": 21760549.0,
      "step": 11830
    },
    {
      "entropy": 5.271934127807617,
      "epoch": 0.9885566321416639,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000490816558027881,
      "loss": 4.9705,
      "mean_token_accuracy": 0.21397959291934968,
      "num_tokens": 21769873.0,
      "step": 11835
    },
    {
      "entropy": 5.311835861206054,
      "epoch": 0.9889742733043769,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004908081377751985,
      "loss": 5.0565,
      "mean_token_accuracy": 0.2081603780388832,
      "num_tokens": 21779036.0,
      "step": 11840
    },
    {
      "entropy": 5.383252859115601,
      "epoch": 0.9893919144670899,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004907997137445347,
      "loss": 5.0705,
      "mean_token_accuracy": 0.19567343592643738,
      "num_tokens": 21787671.0,
      "step": 11845
    },
    {
      "entropy": 5.31885142326355,
      "epoch": 0.9898095556298029,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004907912859360373,
      "loss": 5.0153,
      "mean_token_accuracy": 0.19948294013738632,
      "num_tokens": 21797269.0,
      "step": 11850
    },
    {
      "entropy": 5.352839565277099,
      "epoch": 0.9902271967925159,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004907828543498536,
      "loss": 5.0654,
      "mean_token_accuracy": 0.2057499498128891,
      "num_tokens": 21806738.0,
      "step": 11855
    },
    {
      "entropy": 5.307039546966553,
      "epoch": 0.9906448379552288,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004907744189861313,
      "loss": 4.9756,
      "mean_token_accuracy": 0.20881827175617218,
      "num_tokens": 21815612.0,
      "step": 11860
    },
    {
      "entropy": 5.317244911193848,
      "epoch": 0.9910624791179419,
      "grad_norm": 1.0625,
      "learning_rate": 0.000490765979845018,
      "loss": 5.0055,
      "mean_token_accuracy": 0.21223377138376237,
      "num_tokens": 21824222.0,
      "step": 11865
    },
    {
      "entropy": 5.291276216506958,
      "epoch": 0.9914801202806549,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004907575369266617,
      "loss": 4.982,
      "mean_token_accuracy": 0.2069606065750122,
      "num_tokens": 21833929.0,
      "step": 11870
    },
    {
      "entropy": 5.261525964736938,
      "epoch": 0.9918977614433678,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004907490902312099,
      "loss": 5.0209,
      "mean_token_accuracy": 0.20853301435708999,
      "num_tokens": 21843261.0,
      "step": 11875
    },
    {
      "entropy": 5.336874485015869,
      "epoch": 0.9923154026060809,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004907406397588106,
      "loss": 5.0811,
      "mean_token_accuracy": 0.20042007118463517,
      "num_tokens": 21853004.0,
      "step": 11880
    },
    {
      "entropy": 5.272183275222778,
      "epoch": 0.9927330437687939,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004907321855096117,
      "loss": 4.9167,
      "mean_token_accuracy": 0.2100757271051407,
      "num_tokens": 21862628.0,
      "step": 11885
    },
    {
      "entropy": 5.216411352157593,
      "epoch": 0.9931506849315068,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004907237274837613,
      "loss": 4.9193,
      "mean_token_accuracy": 0.20814323276281357,
      "num_tokens": 21872041.0,
      "step": 11890
    },
    {
      "entropy": 5.289367532730102,
      "epoch": 0.9935683260942199,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004907152656814072,
      "loss": 5.0174,
      "mean_token_accuracy": 0.20065023601055146,
      "num_tokens": 21881834.0,
      "step": 11895
    },
    {
      "entropy": 5.372476816177368,
      "epoch": 0.9939859672569329,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004907068001026978,
      "loss": 5.1003,
      "mean_token_accuracy": 0.2043833166360855,
      "num_tokens": 21892349.0,
      "step": 11900
    },
    {
      "entropy": 5.299916362762451,
      "epoch": 0.9944036084196458,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004906983307477813,
      "loss": 4.9822,
      "mean_token_accuracy": 0.20683847814798356,
      "num_tokens": 21901427.0,
      "step": 11905
    },
    {
      "entropy": 5.25177116394043,
      "epoch": 0.9948212495823588,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004906898576168057,
      "loss": 4.9173,
      "mean_token_accuracy": 0.20798718929290771,
      "num_tokens": 21910654.0,
      "step": 11910
    },
    {
      "entropy": 5.184196090698242,
      "epoch": 0.9952388907450719,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004906813807099197,
      "loss": 4.9545,
      "mean_token_accuracy": 0.21274185478687285,
      "num_tokens": 21919583.0,
      "step": 11915
    },
    {
      "entropy": 5.184610080718994,
      "epoch": 0.9956565319077848,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004906729000272714,
      "loss": 4.8604,
      "mean_token_accuracy": 0.21967073231935502,
      "num_tokens": 21927843.0,
      "step": 11920
    },
    {
      "entropy": 5.25602707862854,
      "epoch": 0.9960741730704978,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004906644155690095,
      "loss": 4.9513,
      "mean_token_accuracy": 0.20525966584682465,
      "num_tokens": 21937269.0,
      "step": 11925
    },
    {
      "entropy": 5.399289941787719,
      "epoch": 0.9964918142332109,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004906559273352823,
      "loss": 5.1073,
      "mean_token_accuracy": 0.19924258887767793,
      "num_tokens": 21946965.0,
      "step": 11930
    },
    {
      "entropy": 5.301539325714112,
      "epoch": 0.9969094553959238,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004906474353262386,
      "loss": 4.9556,
      "mean_token_accuracy": 0.2067214608192444,
      "num_tokens": 21955957.0,
      "step": 11935
    },
    {
      "entropy": 5.217137289047241,
      "epoch": 0.9973270965586368,
      "grad_norm": 1.0,
      "learning_rate": 0.0004906389395420269,
      "loss": 4.9119,
      "mean_token_accuracy": 0.21179112046957016,
      "num_tokens": 21965643.0,
      "step": 11940
    },
    {
      "entropy": 5.2958615779876705,
      "epoch": 0.9977447377213499,
      "grad_norm": 1.0625,
      "learning_rate": 0.000490630439982796,
      "loss": 4.9678,
      "mean_token_accuracy": 0.21330095529556276,
      "num_tokens": 21974789.0,
      "step": 11945
    },
    {
      "entropy": 5.208302307128906,
      "epoch": 0.9981623788840628,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004906219366486947,
      "loss": 4.8764,
      "mean_token_accuracy": 0.22183732688426971,
      "num_tokens": 21983411.0,
      "step": 11950
    },
    {
      "entropy": 5.312915325164795,
      "epoch": 0.9985800200467758,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004906134295398719,
      "loss": 4.9777,
      "mean_token_accuracy": 0.21182502955198287,
      "num_tokens": 21992570.0,
      "step": 11955
    },
    {
      "entropy": 5.13899507522583,
      "epoch": 0.9989976612094889,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004906049186564764,
      "loss": 4.8562,
      "mean_token_accuracy": 0.21910465359687806,
      "num_tokens": 22001411.0,
      "step": 11960
    },
    {
      "entropy": 5.361911869049072,
      "epoch": 0.9994153023722018,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004905964039986574,
      "loss": 5.0927,
      "mean_token_accuracy": 0.19569319486618042,
      "num_tokens": 22010882.0,
      "step": 11965
    },
    {
      "entropy": 5.315498733520508,
      "epoch": 0.9998329435349148,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004905878855665637,
      "loss": 5.0669,
      "mean_token_accuracy": 0.20154594928026198,
      "num_tokens": 22020885.0,
      "step": 11970
    },
    {
      "entropy": 5.298489713668824,
      "epoch": 1.0002505846976277,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004905793633603446,
      "loss": 4.8283,
      "mean_token_accuracy": 0.21429413557052612,
      "num_tokens": 22030278.0,
      "step": 11975
    },
    {
      "entropy": 5.329520654678345,
      "epoch": 1.0006682258603408,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004905708373801492,
      "loss": 5.0236,
      "mean_token_accuracy": 0.20447431206703187,
      "num_tokens": 22039601.0,
      "step": 11980
    },
    {
      "entropy": 5.225157594680786,
      "epoch": 1.0010858670230538,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004905623076261269,
      "loss": 4.856,
      "mean_token_accuracy": 0.2173604890704155,
      "num_tokens": 22049158.0,
      "step": 11985
    },
    {
      "entropy": 5.298311805725097,
      "epoch": 1.0015035081857668,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000490553774098427,
      "loss": 4.965,
      "mean_token_accuracy": 0.20221607983112336,
      "num_tokens": 22059278.0,
      "step": 11990
    },
    {
      "entropy": 5.351927375793457,
      "epoch": 1.0019211493484799,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004905452367971986,
      "loss": 4.9385,
      "mean_token_accuracy": 0.21039484888315202,
      "num_tokens": 22068415.0,
      "step": 11995
    },
    {
      "entropy": 5.29804539680481,
      "epoch": 1.0023387905111927,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004905366957225916,
      "loss": 4.9082,
      "mean_token_accuracy": 0.20734994858503342,
      "num_tokens": 22077815.0,
      "step": 12000
    },
    {
      "epoch": 1.0023387905111927,
      "eval_entropy": 4.986284167047931,
      "eval_loss": 4.924649238586426,
      "eval_mean_token_accuracy": 0.2198381631780519,
      "eval_num_tokens": 22077815.0,
      "eval_runtime": 20.8976,
      "eval_samples_per_second": 1807.045,
      "eval_steps_per_second": 225.911,
      "step": 12000
    }
  ],
  "logging_steps": 5,
  "max_steps": 119720,
  "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": 4844957378641920.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}