{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.7540928833945872,
  "eval_steps": 3000,
  "global_step": 21000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 6.7180821895599365,
      "epoch": 0.000417641162712997,
      "grad_norm": 1.234375,
      "learning_rate": 2e-06,
      "loss": 6.5123,
      "mean_token_accuracy": 0.10666800066828727,
      "num_tokens": 9371.0,
      "step": 5
    },
    {
      "entropy": 6.808059024810791,
      "epoch": 0.000835282325425994,
      "grad_norm": 1.2109375,
      "learning_rate": 4.5e-06,
      "loss": 6.5956,
      "mean_token_accuracy": 0.10533942952752114,
      "num_tokens": 18155.0,
      "step": 10
    },
    {
      "entropy": 6.759247636795044,
      "epoch": 0.001252923488138991,
      "grad_norm": 1.140625,
      "learning_rate": 7e-06,
      "loss": 6.5801,
      "mean_token_accuracy": 0.10018297508358956,
      "num_tokens": 27776.0,
      "step": 15
    },
    {
      "entropy": 6.797110748291016,
      "epoch": 0.001670564650851988,
      "grad_norm": 1.203125,
      "learning_rate": 9.5e-06,
      "loss": 6.5797,
      "mean_token_accuracy": 0.10600945577025414,
      "num_tokens": 37166.0,
      "step": 20
    },
    {
      "entropy": 6.759796857833862,
      "epoch": 0.002088205813564985,
      "grad_norm": 1.0546875,
      "learning_rate": 1.2e-05,
      "loss": 6.5305,
      "mean_token_accuracy": 0.11286764591932297,
      "num_tokens": 46667.0,
      "step": 25
    },
    {
      "entropy": 6.808503246307373,
      "epoch": 0.002505846976277982,
      "grad_norm": 1.078125,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 6.5297,
      "mean_token_accuracy": 0.10954836457967758,
      "num_tokens": 55738.0,
      "step": 30
    },
    {
      "entropy": 6.905849361419678,
      "epoch": 0.002923488138990979,
      "grad_norm": 1.0390625,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 6.6363,
      "mean_token_accuracy": 0.10485906973481178,
      "num_tokens": 65303.0,
      "step": 35
    },
    {
      "entropy": 6.810987138748169,
      "epoch": 0.003341129301703976,
      "grad_norm": 1.1875,
      "learning_rate": 1.95e-05,
      "loss": 6.4128,
      "mean_token_accuracy": 0.11579142212867737,
      "num_tokens": 74876.0,
      "step": 40
    },
    {
      "entropy": 6.819403839111328,
      "epoch": 0.003758770464416973,
      "grad_norm": 1.1796875,
      "learning_rate": 2.2e-05,
      "loss": 6.5539,
      "mean_token_accuracy": 0.11837585121393204,
      "num_tokens": 84683.0,
      "step": 45
    },
    {
      "entropy": 6.842317581176758,
      "epoch": 0.00417641162712997,
      "grad_norm": 1.109375,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 6.5476,
      "mean_token_accuracy": 0.1076658695936203,
      "num_tokens": 94511.0,
      "step": 50
    },
    {
      "entropy": 6.806653213500977,
      "epoch": 0.004594052789842967,
      "grad_norm": 1.078125,
      "learning_rate": 2.7e-05,
      "loss": 6.529,
      "mean_token_accuracy": 0.12048889845609664,
      "num_tokens": 103820.0,
      "step": 55
    },
    {
      "entropy": 6.758758592605591,
      "epoch": 0.005011693952555964,
      "grad_norm": 1.0390625,
      "learning_rate": 2.95e-05,
      "loss": 6.5035,
      "mean_token_accuracy": 0.1163383036851883,
      "num_tokens": 112398.0,
      "step": 60
    },
    {
      "entropy": 6.7997699737548825,
      "epoch": 0.005429335115268961,
      "grad_norm": 1.1484375,
      "learning_rate": 3.2e-05,
      "loss": 6.4622,
      "mean_token_accuracy": 0.119983372092247,
      "num_tokens": 121167.0,
      "step": 65
    },
    {
      "entropy": 6.826302528381348,
      "epoch": 0.005846976277981958,
      "grad_norm": 1.171875,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 6.4857,
      "mean_token_accuracy": 0.11032287627458573,
      "num_tokens": 129399.0,
      "step": 70
    },
    {
      "entropy": 6.845746660232544,
      "epoch": 0.006264617440694955,
      "grad_norm": 1.1796875,
      "learning_rate": 3.7e-05,
      "loss": 6.4687,
      "mean_token_accuracy": 0.1163133703172207,
      "num_tokens": 138512.0,
      "step": 75
    },
    {
      "entropy": 6.800804281234742,
      "epoch": 0.006682258603407952,
      "grad_norm": 1.140625,
      "learning_rate": 3.95e-05,
      "loss": 6.5206,
      "mean_token_accuracy": 0.10875637605786323,
      "num_tokens": 147712.0,
      "step": 80
    },
    {
      "entropy": 6.811766719818115,
      "epoch": 0.007099899766120949,
      "grad_norm": 1.28125,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 6.4852,
      "mean_token_accuracy": 0.1177453450858593,
      "num_tokens": 156759.0,
      "step": 85
    },
    {
      "entropy": 6.773466730117798,
      "epoch": 0.007517540928833946,
      "grad_norm": 1.2265625,
      "learning_rate": 4.45e-05,
      "loss": 6.533,
      "mean_token_accuracy": 0.11238256990909576,
      "num_tokens": 166395.0,
      "step": 90
    },
    {
      "entropy": 6.842059373855591,
      "epoch": 0.007935182091546943,
      "grad_norm": 1.15625,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 6.4534,
      "mean_token_accuracy": 0.11896919012069702,
      "num_tokens": 174739.0,
      "step": 95
    },
    {
      "entropy": 6.710035228729248,
      "epoch": 0.00835282325425994,
      "grad_norm": 1.1640625,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 6.3439,
      "mean_token_accuracy": 0.12140651717782021,
      "num_tokens": 183010.0,
      "step": 100
    },
    {
      "entropy": 6.729191780090332,
      "epoch": 0.008770464416972937,
      "grad_norm": 1.03125,
      "learning_rate": 5.2e-05,
      "loss": 6.5163,
      "mean_token_accuracy": 0.11011387556791305,
      "num_tokens": 192589.0,
      "step": 105
    },
    {
      "entropy": 6.83894910812378,
      "epoch": 0.009188105579685933,
      "grad_norm": 1.1953125,
      "learning_rate": 5.45e-05,
      "loss": 6.45,
      "mean_token_accuracy": 0.11901598572731018,
      "num_tokens": 201638.0,
      "step": 110
    },
    {
      "entropy": 6.744243049621582,
      "epoch": 0.00960574674239893,
      "grad_norm": 1.3125,
      "learning_rate": 5.7e-05,
      "loss": 6.3679,
      "mean_token_accuracy": 0.12338117063045502,
      "num_tokens": 209862.0,
      "step": 115
    },
    {
      "entropy": 6.756110143661499,
      "epoch": 0.010023387905111928,
      "grad_norm": 1.1875,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 6.5095,
      "mean_token_accuracy": 0.11191049665212631,
      "num_tokens": 218897.0,
      "step": 120
    },
    {
      "entropy": 6.908449125289917,
      "epoch": 0.010441029067824925,
      "grad_norm": 1.265625,
      "learning_rate": 6.2e-05,
      "loss": 6.5859,
      "mean_token_accuracy": 0.11279568374156952,
      "num_tokens": 227682.0,
      "step": 125
    },
    {
      "entropy": 6.8623613834381105,
      "epoch": 0.010858670230537921,
      "grad_norm": 1.3046875,
      "learning_rate": 6.450000000000001e-05,
      "loss": 6.4542,
      "mean_token_accuracy": 0.11863475143909455,
      "num_tokens": 236170.0,
      "step": 130
    },
    {
      "entropy": 6.790626573562622,
      "epoch": 0.011276311393250919,
      "grad_norm": 1.2109375,
      "learning_rate": 6.7e-05,
      "loss": 6.5522,
      "mean_token_accuracy": 0.11070087552070618,
      "num_tokens": 245710.0,
      "step": 135
    },
    {
      "entropy": 6.78648738861084,
      "epoch": 0.011693952555963916,
      "grad_norm": 1.3125,
      "learning_rate": 6.950000000000001e-05,
      "loss": 6.5029,
      "mean_token_accuracy": 0.11403205692768097,
      "num_tokens": 254751.0,
      "step": 140
    },
    {
      "entropy": 6.7859738826751705,
      "epoch": 0.012111593718676913,
      "grad_norm": 1.1875,
      "learning_rate": 7.2e-05,
      "loss": 6.433,
      "mean_token_accuracy": 0.1188641369342804,
      "num_tokens": 264210.0,
      "step": 145
    },
    {
      "entropy": 6.8096400737762455,
      "epoch": 0.01252923488138991,
      "grad_norm": 1.265625,
      "learning_rate": 7.45e-05,
      "loss": 6.4861,
      "mean_token_accuracy": 0.12114087268710136,
      "num_tokens": 272944.0,
      "step": 150
    },
    {
      "entropy": 6.791913175582886,
      "epoch": 0.012946876044102907,
      "grad_norm": 1.328125,
      "learning_rate": 7.7e-05,
      "loss": 6.3829,
      "mean_token_accuracy": 0.12244183719158172,
      "num_tokens": 281311.0,
      "step": 155
    },
    {
      "entropy": 6.694759273529053,
      "epoch": 0.013364517206815904,
      "grad_norm": 1.2421875,
      "learning_rate": 7.950000000000001e-05,
      "loss": 6.3769,
      "mean_token_accuracy": 0.12260683327913284,
      "num_tokens": 289701.0,
      "step": 160
    },
    {
      "entropy": 6.748257827758789,
      "epoch": 0.013782158369528901,
      "grad_norm": 1.0859375,
      "learning_rate": 8.2e-05,
      "loss": 6.5215,
      "mean_token_accuracy": 0.1143926203250885,
      "num_tokens": 299459.0,
      "step": 165
    },
    {
      "entropy": 6.883009147644043,
      "epoch": 0.014199799532241897,
      "grad_norm": 1.3515625,
      "learning_rate": 8.450000000000001e-05,
      "loss": 6.5421,
      "mean_token_accuracy": 0.11752694174647331,
      "num_tokens": 309343.0,
      "step": 170
    },
    {
      "entropy": 6.830853223800659,
      "epoch": 0.014617440694954895,
      "grad_norm": 1.421875,
      "learning_rate": 8.7e-05,
      "loss": 6.5302,
      "mean_token_accuracy": 0.11090763658285141,
      "num_tokens": 318840.0,
      "step": 175
    },
    {
      "entropy": 6.675867223739624,
      "epoch": 0.015035081857667892,
      "grad_norm": 1.15625,
      "learning_rate": 8.95e-05,
      "loss": 6.3784,
      "mean_token_accuracy": 0.11448234245181084,
      "num_tokens": 328689.0,
      "step": 180
    },
    {
      "entropy": 6.797796392440796,
      "epoch": 0.01545272302038089,
      "grad_norm": 1.375,
      "learning_rate": 9.2e-05,
      "loss": 6.4889,
      "mean_token_accuracy": 0.11072958707809448,
      "num_tokens": 338209.0,
      "step": 185
    },
    {
      "entropy": 6.7780595302581785,
      "epoch": 0.015870364183093885,
      "grad_norm": 1.2421875,
      "learning_rate": 9.45e-05,
      "loss": 6.417,
      "mean_token_accuracy": 0.11931143030524254,
      "num_tokens": 347270.0,
      "step": 190
    },
    {
      "entropy": 6.799556255340576,
      "epoch": 0.016288005345806884,
      "grad_norm": 1.3203125,
      "learning_rate": 9.7e-05,
      "loss": 6.4735,
      "mean_token_accuracy": 0.11163464188575745,
      "num_tokens": 356366.0,
      "step": 195
    },
    {
      "entropy": 6.763126516342163,
      "epoch": 0.01670564650851988,
      "grad_norm": 1.28125,
      "learning_rate": 9.95e-05,
      "loss": 6.4167,
      "mean_token_accuracy": 0.11596830189228058,
      "num_tokens": 365980.0,
      "step": 200
    },
    {
      "entropy": 6.702800369262695,
      "epoch": 0.017123287671232876,
      "grad_norm": 1.21875,
      "learning_rate": 0.000102,
      "loss": 6.3882,
      "mean_token_accuracy": 0.11993374973535538,
      "num_tokens": 374591.0,
      "step": 205
    },
    {
      "entropy": 6.849011707305908,
      "epoch": 0.017540928833945875,
      "grad_norm": 1.40625,
      "learning_rate": 0.00010449999999999999,
      "loss": 6.4294,
      "mean_token_accuracy": 0.11508302688598633,
      "num_tokens": 383250.0,
      "step": 210
    },
    {
      "entropy": 6.693080568313599,
      "epoch": 0.01795856999665887,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000107,
      "loss": 6.3306,
      "mean_token_accuracy": 0.11563051715493203,
      "num_tokens": 391502.0,
      "step": 215
    },
    {
      "entropy": 6.630117034912109,
      "epoch": 0.018376211159371866,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0001095,
      "loss": 6.2712,
      "mean_token_accuracy": 0.11645801290869713,
      "num_tokens": 400225.0,
      "step": 220
    },
    {
      "entropy": 6.8181227684021,
      "epoch": 0.018793852322084866,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000112,
      "loss": 6.4937,
      "mean_token_accuracy": 0.1128729097545147,
      "num_tokens": 410008.0,
      "step": 225
    },
    {
      "entropy": 6.736743688583374,
      "epoch": 0.01921149348479786,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0001145,
      "loss": 6.4332,
      "mean_token_accuracy": 0.11869536191225052,
      "num_tokens": 418979.0,
      "step": 230
    },
    {
      "entropy": 6.718912887573242,
      "epoch": 0.01962913464751086,
      "grad_norm": 1.28125,
      "learning_rate": 0.00011700000000000001,
      "loss": 6.4259,
      "mean_token_accuracy": 0.11591459438204765,
      "num_tokens": 428139.0,
      "step": 235
    },
    {
      "entropy": 6.7471612930297855,
      "epoch": 0.020046775810223856,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00011949999999999999,
      "loss": 6.3418,
      "mean_token_accuracy": 0.12110752388834953,
      "num_tokens": 437113.0,
      "step": 240
    },
    {
      "entropy": 6.818037843704223,
      "epoch": 0.020464416972936852,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000122,
      "loss": 6.6019,
      "mean_token_accuracy": 0.11445037350058555,
      "num_tokens": 446879.0,
      "step": 245
    },
    {
      "entropy": 6.695638751983642,
      "epoch": 0.02088205813564985,
      "grad_norm": 1.328125,
      "learning_rate": 0.0001245,
      "loss": 6.5041,
      "mean_token_accuracy": 0.11364466175436974,
      "num_tokens": 456362.0,
      "step": 250
    },
    {
      "entropy": 6.799724435806274,
      "epoch": 0.021299699298362847,
      "grad_norm": 1.3125,
      "learning_rate": 0.000127,
      "loss": 6.2963,
      "mean_token_accuracy": 0.12356706708669662,
      "num_tokens": 464930.0,
      "step": 255
    },
    {
      "entropy": 6.682189035415649,
      "epoch": 0.021717340461075842,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0001295,
      "loss": 6.469,
      "mean_token_accuracy": 0.11262325197458267,
      "num_tokens": 474303.0,
      "step": 260
    },
    {
      "entropy": 6.784652948379517,
      "epoch": 0.02213498162378884,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000132,
      "loss": 6.4998,
      "mean_token_accuracy": 0.11313469260931015,
      "num_tokens": 483750.0,
      "step": 265
    },
    {
      "entropy": 6.743142223358154,
      "epoch": 0.022552622786501837,
      "grad_norm": 1.0625,
      "learning_rate": 0.00013450000000000002,
      "loss": 6.3944,
      "mean_token_accuracy": 0.1151240773499012,
      "num_tokens": 493325.0,
      "step": 270
    },
    {
      "entropy": 6.729674768447876,
      "epoch": 0.022970263949214833,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00013700000000000002,
      "loss": 6.4083,
      "mean_token_accuracy": 0.11122140809893608,
      "num_tokens": 501907.0,
      "step": 275
    },
    {
      "entropy": 6.695761060714721,
      "epoch": 0.023387905111927832,
      "grad_norm": 1.21875,
      "learning_rate": 0.0001395,
      "loss": 6.4871,
      "mean_token_accuracy": 0.11502960547804833,
      "num_tokens": 511522.0,
      "step": 280
    },
    {
      "entropy": 6.770887994766236,
      "epoch": 0.023805546274640828,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00014199999999999998,
      "loss": 6.4136,
      "mean_token_accuracy": 0.1140117235481739,
      "num_tokens": 521287.0,
      "step": 285
    },
    {
      "entropy": 6.772694158554077,
      "epoch": 0.024223187437353827,
      "grad_norm": 1.28125,
      "learning_rate": 0.0001445,
      "loss": 6.5268,
      "mean_token_accuracy": 0.10702815651893616,
      "num_tokens": 530631.0,
      "step": 290
    },
    {
      "entropy": 6.726316118240357,
      "epoch": 0.024640828600066823,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000147,
      "loss": 6.3898,
      "mean_token_accuracy": 0.12561146542429924,
      "num_tokens": 540386.0,
      "step": 295
    },
    {
      "entropy": 6.716358995437622,
      "epoch": 0.02505846976277982,
      "grad_norm": 1.25,
      "learning_rate": 0.0001495,
      "loss": 6.455,
      "mean_token_accuracy": 0.11597420573234558,
      "num_tokens": 550121.0,
      "step": 300
    },
    {
      "entropy": 6.649577903747558,
      "epoch": 0.025476110925492818,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000152,
      "loss": 6.3557,
      "mean_token_accuracy": 0.11343038976192474,
      "num_tokens": 559460.0,
      "step": 305
    },
    {
      "entropy": 6.724728441238403,
      "epoch": 0.025893752088205813,
      "grad_norm": 1.265625,
      "learning_rate": 0.00015450000000000001,
      "loss": 6.3,
      "mean_token_accuracy": 0.12006132379174232,
      "num_tokens": 568392.0,
      "step": 310
    },
    {
      "entropy": 6.653720140457153,
      "epoch": 0.02631139325091881,
      "grad_norm": 1.171875,
      "learning_rate": 0.000157,
      "loss": 6.4208,
      "mean_token_accuracy": 0.11691968142986298,
      "num_tokens": 578248.0,
      "step": 315
    },
    {
      "entropy": 6.7646496295928955,
      "epoch": 0.026729034413631808,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0001595,
      "loss": 6.4952,
      "mean_token_accuracy": 0.11188015565276146,
      "num_tokens": 588164.0,
      "step": 320
    },
    {
      "entropy": 6.729947662353515,
      "epoch": 0.027146675576344804,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000162,
      "loss": 6.4382,
      "mean_token_accuracy": 0.11248505786061287,
      "num_tokens": 598715.0,
      "step": 325
    },
    {
      "entropy": 6.6656099319458,
      "epoch": 0.027564316739057803,
      "grad_norm": 1.390625,
      "learning_rate": 0.00016450000000000001,
      "loss": 6.3299,
      "mean_token_accuracy": 0.11870421469211578,
      "num_tokens": 606939.0,
      "step": 330
    },
    {
      "entropy": 6.688929510116577,
      "epoch": 0.0279819579017708,
      "grad_norm": 1.4375,
      "learning_rate": 0.00016700000000000002,
      "loss": 6.4408,
      "mean_token_accuracy": 0.12059589847922325,
      "num_tokens": 615803.0,
      "step": 335
    },
    {
      "entropy": 6.677912378311158,
      "epoch": 0.028399599064483794,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00016950000000000003,
      "loss": 6.3016,
      "mean_token_accuracy": 0.1152017168700695,
      "num_tokens": 624730.0,
      "step": 340
    },
    {
      "entropy": 6.668429660797119,
      "epoch": 0.028817240227196794,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00017199999999999998,
      "loss": 6.4014,
      "mean_token_accuracy": 0.11559123620390892,
      "num_tokens": 634082.0,
      "step": 345
    },
    {
      "entropy": 6.710170698165894,
      "epoch": 0.02923488138990979,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00017449999999999999,
      "loss": 6.3967,
      "mean_token_accuracy": 0.11764291599392891,
      "num_tokens": 643367.0,
      "step": 350
    },
    {
      "entropy": 6.57072377204895,
      "epoch": 0.029652522552622785,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000177,
      "loss": 6.2433,
      "mean_token_accuracy": 0.12115906029939652,
      "num_tokens": 651861.0,
      "step": 355
    },
    {
      "entropy": 6.699751520156861,
      "epoch": 0.030070163715335784,
      "grad_norm": 1.296875,
      "learning_rate": 0.0001795,
      "loss": 6.383,
      "mean_token_accuracy": 0.11497076004743575,
      "num_tokens": 661064.0,
      "step": 360
    },
    {
      "entropy": 6.6671435832977295,
      "epoch": 0.03048780487804878,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000182,
      "loss": 6.3574,
      "mean_token_accuracy": 0.12420301362872124,
      "num_tokens": 669299.0,
      "step": 365
    },
    {
      "entropy": 6.628156757354736,
      "epoch": 0.03090544604076178,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0001845,
      "loss": 6.4516,
      "mean_token_accuracy": 0.11220049262046813,
      "num_tokens": 678631.0,
      "step": 370
    },
    {
      "entropy": 6.7193104267120365,
      "epoch": 0.03132308720347477,
      "grad_norm": 1.296875,
      "learning_rate": 0.000187,
      "loss": 6.4496,
      "mean_token_accuracy": 0.1158321812748909,
      "num_tokens": 688511.0,
      "step": 375
    },
    {
      "entropy": 6.703116130828858,
      "epoch": 0.03174072836618777,
      "grad_norm": 1.25,
      "learning_rate": 0.0001895,
      "loss": 6.3797,
      "mean_token_accuracy": 0.11359225586056709,
      "num_tokens": 697956.0,
      "step": 380
    },
    {
      "entropy": 6.668367481231689,
      "epoch": 0.03215836952890077,
      "grad_norm": 1.4375,
      "learning_rate": 0.000192,
      "loss": 6.2828,
      "mean_token_accuracy": 0.12736818715929984,
      "num_tokens": 706779.0,
      "step": 385
    },
    {
      "entropy": 6.554675436019897,
      "epoch": 0.03257601069161377,
      "grad_norm": 1.25,
      "learning_rate": 0.0001945,
      "loss": 6.3132,
      "mean_token_accuracy": 0.11831648498773575,
      "num_tokens": 715749.0,
      "step": 390
    },
    {
      "entropy": 6.59646201133728,
      "epoch": 0.03299365185432676,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00019700000000000002,
      "loss": 6.2781,
      "mean_token_accuracy": 0.11623693332076072,
      "num_tokens": 724756.0,
      "step": 395
    },
    {
      "entropy": 6.562495231628418,
      "epoch": 0.03341129301703976,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00019950000000000002,
      "loss": 6.3001,
      "mean_token_accuracy": 0.1243792787194252,
      "num_tokens": 733288.0,
      "step": 400
    },
    {
      "entropy": 6.739299392700195,
      "epoch": 0.03382893417975276,
      "grad_norm": 1.5,
      "learning_rate": 0.000202,
      "loss": 6.4209,
      "mean_token_accuracy": 0.1129365734755993,
      "num_tokens": 741629.0,
      "step": 405
    },
    {
      "entropy": 6.62856707572937,
      "epoch": 0.03424657534246575,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00020449999999999998,
      "loss": 6.3605,
      "mean_token_accuracy": 0.12043223157525063,
      "num_tokens": 751336.0,
      "step": 410
    },
    {
      "entropy": 6.780359745025635,
      "epoch": 0.03466421650517875,
      "grad_norm": 1.28125,
      "learning_rate": 0.000207,
      "loss": 6.4408,
      "mean_token_accuracy": 0.11233260035514832,
      "num_tokens": 759725.0,
      "step": 415
    },
    {
      "entropy": 6.7300169467926025,
      "epoch": 0.03508185766789175,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002095,
      "loss": 6.4676,
      "mean_token_accuracy": 0.11555668562650681,
      "num_tokens": 769752.0,
      "step": 420
    },
    {
      "entropy": 6.61152400970459,
      "epoch": 0.03549949883060474,
      "grad_norm": 1.265625,
      "learning_rate": 0.000212,
      "loss": 6.3419,
      "mean_token_accuracy": 0.12118100821971893,
      "num_tokens": 779124.0,
      "step": 425
    },
    {
      "entropy": 6.572236108779907,
      "epoch": 0.03591713999331774,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002145,
      "loss": 6.3142,
      "mean_token_accuracy": 0.11852151229977607,
      "num_tokens": 788892.0,
      "step": 430
    },
    {
      "entropy": 6.499159049987793,
      "epoch": 0.03633478115603074,
      "grad_norm": 1.1875,
      "learning_rate": 0.00021700000000000002,
      "loss": 6.2789,
      "mean_token_accuracy": 0.12280359491705894,
      "num_tokens": 798265.0,
      "step": 435
    },
    {
      "entropy": 6.5266131401062015,
      "epoch": 0.03675242231874373,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002195,
      "loss": 6.2382,
      "mean_token_accuracy": 0.12434489950537682,
      "num_tokens": 806896.0,
      "step": 440
    },
    {
      "entropy": 6.646367168426513,
      "epoch": 0.03717006348145673,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000222,
      "loss": 6.313,
      "mean_token_accuracy": 0.1231755070388317,
      "num_tokens": 815540.0,
      "step": 445
    },
    {
      "entropy": 6.595507431030273,
      "epoch": 0.03758770464416973,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002245,
      "loss": 6.3785,
      "mean_token_accuracy": 0.11604066267609596,
      "num_tokens": 824555.0,
      "step": 450
    },
    {
      "entropy": 6.659368467330933,
      "epoch": 0.03800534580688272,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00022700000000000002,
      "loss": 6.429,
      "mean_token_accuracy": 0.11681790426373481,
      "num_tokens": 833414.0,
      "step": 455
    },
    {
      "entropy": 6.521911573410034,
      "epoch": 0.03842298696959572,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00022950000000000002,
      "loss": 6.2667,
      "mean_token_accuracy": 0.1260523058474064,
      "num_tokens": 842644.0,
      "step": 460
    },
    {
      "entropy": 6.635482358932495,
      "epoch": 0.03884062813230872,
      "grad_norm": 1.375,
      "learning_rate": 0.00023200000000000003,
      "loss": 6.3016,
      "mean_token_accuracy": 0.11958117336034775,
      "num_tokens": 851786.0,
      "step": 465
    },
    {
      "entropy": 6.533578109741211,
      "epoch": 0.03925826929502172,
      "grad_norm": 1.25,
      "learning_rate": 0.00023449999999999998,
      "loss": 6.378,
      "mean_token_accuracy": 0.11788101494312286,
      "num_tokens": 861508.0,
      "step": 470
    },
    {
      "entropy": 6.512241077423096,
      "epoch": 0.03967591045773471,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000237,
      "loss": 6.24,
      "mean_token_accuracy": 0.1280357524752617,
      "num_tokens": 870433.0,
      "step": 475
    },
    {
      "entropy": 6.659976148605347,
      "epoch": 0.04009355162044771,
      "grad_norm": 1.25,
      "learning_rate": 0.0002395,
      "loss": 6.3483,
      "mean_token_accuracy": 0.12422682046890259,
      "num_tokens": 879059.0,
      "step": 480
    },
    {
      "entropy": 6.637609958648682,
      "epoch": 0.04051119278316071,
      "grad_norm": 1.265625,
      "learning_rate": 0.000242,
      "loss": 6.3906,
      "mean_token_accuracy": 0.11781356558203697,
      "num_tokens": 888009.0,
      "step": 485
    },
    {
      "entropy": 6.5305976390838625,
      "epoch": 0.040928833945873704,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002445,
      "loss": 6.263,
      "mean_token_accuracy": 0.11897468864917755,
      "num_tokens": 897335.0,
      "step": 490
    },
    {
      "entropy": 6.650092506408692,
      "epoch": 0.0413464751085867,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000247,
      "loss": 6.2553,
      "mean_token_accuracy": 0.12942158281803132,
      "num_tokens": 905802.0,
      "step": 495
    },
    {
      "entropy": 6.505962324142456,
      "epoch": 0.0417641162712997,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002495,
      "loss": 6.2894,
      "mean_token_accuracy": 0.1238292008638382,
      "num_tokens": 914836.0,
      "step": 500
    },
    {
      "entropy": 6.481074333190918,
      "epoch": 0.042181757434012694,
      "grad_norm": 1.359375,
      "learning_rate": 0.000252,
      "loss": 6.2294,
      "mean_token_accuracy": 0.12505172789096833,
      "num_tokens": 924254.0,
      "step": 505
    },
    {
      "entropy": 6.538729810714722,
      "epoch": 0.04259939859672569,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002545,
      "loss": 6.343,
      "mean_token_accuracy": 0.12079363018274307,
      "num_tokens": 933841.0,
      "step": 510
    },
    {
      "entropy": 6.602547264099121,
      "epoch": 0.04301703975943869,
      "grad_norm": 1.25,
      "learning_rate": 0.000257,
      "loss": 6.3927,
      "mean_token_accuracy": 0.11534046903252601,
      "num_tokens": 943304.0,
      "step": 515
    },
    {
      "entropy": 6.526961374282837,
      "epoch": 0.043434680922151685,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002595,
      "loss": 6.2249,
      "mean_token_accuracy": 0.12119961977005005,
      "num_tokens": 952915.0,
      "step": 520
    },
    {
      "entropy": 6.538762998580933,
      "epoch": 0.043852322084864684,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000262,
      "loss": 6.2935,
      "mean_token_accuracy": 0.12450704649090767,
      "num_tokens": 962752.0,
      "step": 525
    },
    {
      "entropy": 6.56287579536438,
      "epoch": 0.04426996324757768,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00026450000000000003,
      "loss": 6.309,
      "mean_token_accuracy": 0.1171677753329277,
      "num_tokens": 971240.0,
      "step": 530
    },
    {
      "entropy": 6.575090742111206,
      "epoch": 0.044687604410290675,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00026700000000000004,
      "loss": 6.3548,
      "mean_token_accuracy": 0.11371841728687286,
      "num_tokens": 981391.0,
      "step": 535
    },
    {
      "entropy": 6.543639373779297,
      "epoch": 0.045105245573003674,
      "grad_norm": 1.203125,
      "learning_rate": 0.00026950000000000005,
      "loss": 6.3565,
      "mean_token_accuracy": 0.12311091870069504,
      "num_tokens": 990551.0,
      "step": 540
    },
    {
      "entropy": 6.53157148361206,
      "epoch": 0.045522886735716674,
      "grad_norm": 1.4375,
      "learning_rate": 0.00027200000000000005,
      "loss": 6.2798,
      "mean_token_accuracy": 0.12509959563612938,
      "num_tokens": 999593.0,
      "step": 545
    },
    {
      "entropy": 6.54977445602417,
      "epoch": 0.045940527898429666,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002745,
      "loss": 6.3563,
      "mean_token_accuracy": 0.11864383593201637,
      "num_tokens": 1008785.0,
      "step": 550
    },
    {
      "entropy": 6.53155255317688,
      "epoch": 0.046358169061142665,
      "grad_norm": 1.46875,
      "learning_rate": 0.000277,
      "loss": 6.3408,
      "mean_token_accuracy": 0.12224255055189133,
      "num_tokens": 1016815.0,
      "step": 555
    },
    {
      "entropy": 6.564304971694947,
      "epoch": 0.046775810223855664,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002795,
      "loss": 6.2135,
      "mean_token_accuracy": 0.13047643452882768,
      "num_tokens": 1026182.0,
      "step": 560
    },
    {
      "entropy": 6.417008876800537,
      "epoch": 0.04719345138656866,
      "grad_norm": 1.46875,
      "learning_rate": 0.00028199999999999997,
      "loss": 6.1538,
      "mean_token_accuracy": 0.1273562304675579,
      "num_tokens": 1034032.0,
      "step": 565
    },
    {
      "entropy": 6.540080308914185,
      "epoch": 0.047611092549281656,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002845,
      "loss": 6.3695,
      "mean_token_accuracy": 0.12380489185452462,
      "num_tokens": 1043493.0,
      "step": 570
    },
    {
      "entropy": 6.542786312103272,
      "epoch": 0.048028733711994655,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000287,
      "loss": 6.2786,
      "mean_token_accuracy": 0.12197929993271828,
      "num_tokens": 1052989.0,
      "step": 575
    },
    {
      "entropy": 6.446332406997681,
      "epoch": 0.048446374874707654,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002895,
      "loss": 6.2418,
      "mean_token_accuracy": 0.12374974638223649,
      "num_tokens": 1061803.0,
      "step": 580
    },
    {
      "entropy": 6.586713171005249,
      "epoch": 0.048864016037420646,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000292,
      "loss": 6.3667,
      "mean_token_accuracy": 0.12280836850404739,
      "num_tokens": 1071227.0,
      "step": 585
    },
    {
      "entropy": 6.523447036743164,
      "epoch": 0.049281657200133645,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002945,
      "loss": 6.3889,
      "mean_token_accuracy": 0.12325869053602219,
      "num_tokens": 1080082.0,
      "step": 590
    },
    {
      "entropy": 6.480297422409057,
      "epoch": 0.049699298362846644,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000297,
      "loss": 6.2021,
      "mean_token_accuracy": 0.12905326038599013,
      "num_tokens": 1089750.0,
      "step": 595
    },
    {
      "entropy": 6.378500175476074,
      "epoch": 0.05011693952555964,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002995,
      "loss": 6.2089,
      "mean_token_accuracy": 0.12159204185009002,
      "num_tokens": 1098691.0,
      "step": 600
    },
    {
      "entropy": 6.468329238891601,
      "epoch": 0.050534580688272636,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000302,
      "loss": 6.2107,
      "mean_token_accuracy": 0.1222094714641571,
      "num_tokens": 1107602.0,
      "step": 605
    },
    {
      "entropy": 6.37498927116394,
      "epoch": 0.050952221850985635,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003045,
      "loss": 6.1582,
      "mean_token_accuracy": 0.12626441866159438,
      "num_tokens": 1116426.0,
      "step": 610
    },
    {
      "entropy": 6.563107347488403,
      "epoch": 0.05136986301369863,
      "grad_norm": 1.109375,
      "learning_rate": 0.000307,
      "loss": 6.3579,
      "mean_token_accuracy": 0.12089578509330749,
      "num_tokens": 1126221.0,
      "step": 615
    },
    {
      "entropy": 6.436634206771851,
      "epoch": 0.051787504176411626,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003095,
      "loss": 6.2162,
      "mean_token_accuracy": 0.13165414333343506,
      "num_tokens": 1135662.0,
      "step": 620
    },
    {
      "entropy": 6.47764105796814,
      "epoch": 0.052205145339124626,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000312,
      "loss": 6.3415,
      "mean_token_accuracy": 0.1241353303194046,
      "num_tokens": 1143963.0,
      "step": 625
    },
    {
      "entropy": 6.503053283691406,
      "epoch": 0.05262278650183762,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003145,
      "loss": 6.2121,
      "mean_token_accuracy": 0.12860342785716056,
      "num_tokens": 1152470.0,
      "step": 630
    },
    {
      "entropy": 6.398306846618652,
      "epoch": 0.05304042766455062,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000317,
      "loss": 6.1715,
      "mean_token_accuracy": 0.12845860794186592,
      "num_tokens": 1161257.0,
      "step": 635
    },
    {
      "entropy": 6.530808067321777,
      "epoch": 0.053458068827263616,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003195,
      "loss": 6.3039,
      "mean_token_accuracy": 0.12109335586428642,
      "num_tokens": 1170969.0,
      "step": 640
    },
    {
      "entropy": 6.558892440795899,
      "epoch": 0.053875709989976615,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000322,
      "loss": 6.4402,
      "mean_token_accuracy": 0.12084705233573914,
      "num_tokens": 1181127.0,
      "step": 645
    },
    {
      "entropy": 6.407193994522094,
      "epoch": 0.05429335115268961,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.2425,
      "mean_token_accuracy": 0.12455282807350158,
      "num_tokens": 1190260.0,
      "step": 650
    },
    {
      "entropy": 6.455674695968628,
      "epoch": 0.05471099231540261,
      "grad_norm": 1.171875,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.316,
      "mean_token_accuracy": 0.12325879707932472,
      "num_tokens": 1200013.0,
      "step": 655
    },
    {
      "entropy": 6.5374016761779785,
      "epoch": 0.055128633478115606,
      "grad_norm": 1.484375,
      "learning_rate": 0.00032950000000000004,
      "loss": 6.2256,
      "mean_token_accuracy": 0.13025743141770363,
      "num_tokens": 1209239.0,
      "step": 660
    },
    {
      "entropy": 6.347896671295166,
      "epoch": 0.0555462746408286,
      "grad_norm": 1.3125,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.1814,
      "mean_token_accuracy": 0.12614438980817794,
      "num_tokens": 1218552.0,
      "step": 665
    },
    {
      "entropy": 6.401571559906006,
      "epoch": 0.0559639158035416,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00033450000000000005,
      "loss": 6.2656,
      "mean_token_accuracy": 0.12373785302042961,
      "num_tokens": 1227632.0,
      "step": 670
    },
    {
      "entropy": 6.45081524848938,
      "epoch": 0.056381556966254597,
      "grad_norm": 1.28125,
      "learning_rate": 0.000337,
      "loss": 6.2906,
      "mean_token_accuracy": 0.12531701177358628,
      "num_tokens": 1236433.0,
      "step": 675
    },
    {
      "entropy": 6.433235931396484,
      "epoch": 0.05679919812896759,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003395,
      "loss": 6.3248,
      "mean_token_accuracy": 0.11682864874601365,
      "num_tokens": 1245700.0,
      "step": 680
    },
    {
      "entropy": 6.424199199676513,
      "epoch": 0.05721683929168059,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000342,
      "loss": 6.1626,
      "mean_token_accuracy": 0.12845946177840234,
      "num_tokens": 1254206.0,
      "step": 685
    },
    {
      "entropy": 6.392566871643067,
      "epoch": 0.05763448045439359,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.2238,
      "mean_token_accuracy": 0.12663000896573068,
      "num_tokens": 1263302.0,
      "step": 690
    },
    {
      "entropy": 6.433683109283447,
      "epoch": 0.05805212161710658,
      "grad_norm": 1.21875,
      "learning_rate": 0.000347,
      "loss": 6.261,
      "mean_token_accuracy": 0.12566009759902955,
      "num_tokens": 1272998.0,
      "step": 695
    },
    {
      "entropy": 6.42986831665039,
      "epoch": 0.05846976277981958,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003495,
      "loss": 6.2316,
      "mean_token_accuracy": 0.12945592105388642,
      "num_tokens": 1281955.0,
      "step": 700
    },
    {
      "entropy": 6.384770917892456,
      "epoch": 0.05888740394253258,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000352,
      "loss": 6.2693,
      "mean_token_accuracy": 0.1220188558101654,
      "num_tokens": 1291219.0,
      "step": 705
    },
    {
      "entropy": 6.439512538909912,
      "epoch": 0.05930504510524557,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003545,
      "loss": 6.2807,
      "mean_token_accuracy": 0.11803281754255295,
      "num_tokens": 1300963.0,
      "step": 710
    },
    {
      "entropy": 6.434574270248413,
      "epoch": 0.05972268626795857,
      "grad_norm": 1.359375,
      "learning_rate": 0.000357,
      "loss": 6.315,
      "mean_token_accuracy": 0.11672014072537422,
      "num_tokens": 1310112.0,
      "step": 715
    },
    {
      "entropy": 6.382553958892823,
      "epoch": 0.06014032743067157,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003595,
      "loss": 6.2513,
      "mean_token_accuracy": 0.12916786298155786,
      "num_tokens": 1319309.0,
      "step": 720
    },
    {
      "entropy": 6.326040887832642,
      "epoch": 0.06055796859338457,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000362,
      "loss": 6.2324,
      "mean_token_accuracy": 0.12591121718287468,
      "num_tokens": 1329049.0,
      "step": 725
    },
    {
      "entropy": 6.351260471343994,
      "epoch": 0.06097560975609756,
      "grad_norm": 1.53125,
      "learning_rate": 0.0003645,
      "loss": 6.0671,
      "mean_token_accuracy": 0.12919704169034957,
      "num_tokens": 1337770.0,
      "step": 730
    },
    {
      "entropy": 6.254184007644653,
      "epoch": 0.06139325091881056,
      "grad_norm": 1.328125,
      "learning_rate": 0.000367,
      "loss": 6.1822,
      "mean_token_accuracy": 0.1272610329091549,
      "num_tokens": 1346705.0,
      "step": 735
    },
    {
      "entropy": 6.3455441951751705,
      "epoch": 0.06181089208152356,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003695,
      "loss": 6.2394,
      "mean_token_accuracy": 0.12634490579366683,
      "num_tokens": 1354860.0,
      "step": 740
    },
    {
      "entropy": 6.381866407394409,
      "epoch": 0.06222853324423655,
      "grad_norm": 1.234375,
      "learning_rate": 0.000372,
      "loss": 6.2036,
      "mean_token_accuracy": 0.12138592973351478,
      "num_tokens": 1364140.0,
      "step": 745
    },
    {
      "entropy": 6.432730674743652,
      "epoch": 0.06264617440694954,
      "grad_norm": 1.125,
      "learning_rate": 0.0003745,
      "loss": 6.3331,
      "mean_token_accuracy": 0.11733241975307465,
      "num_tokens": 1373645.0,
      "step": 750
    },
    {
      "entropy": 6.377966737747192,
      "epoch": 0.06306381556966255,
      "grad_norm": 1.1875,
      "learning_rate": 0.000377,
      "loss": 6.2159,
      "mean_token_accuracy": 0.13121263906359673,
      "num_tokens": 1383291.0,
      "step": 755
    },
    {
      "entropy": 6.239141368865967,
      "epoch": 0.06348145673237554,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003795,
      "loss": 6.1152,
      "mean_token_accuracy": 0.1310908943414688,
      "num_tokens": 1391172.0,
      "step": 760
    },
    {
      "entropy": 6.377418088912964,
      "epoch": 0.06389909789508853,
      "grad_norm": 1.25,
      "learning_rate": 0.000382,
      "loss": 6.1472,
      "mean_token_accuracy": 0.12764590457081795,
      "num_tokens": 1400864.0,
      "step": 765
    },
    {
      "entropy": 6.295232057571411,
      "epoch": 0.06431673905780154,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003845,
      "loss": 6.2001,
      "mean_token_accuracy": 0.12997447326779366,
      "num_tokens": 1410570.0,
      "step": 770
    },
    {
      "entropy": 6.313883256912232,
      "epoch": 0.06473438022051453,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.1744,
      "mean_token_accuracy": 0.1313069112598896,
      "num_tokens": 1419536.0,
      "step": 775
    },
    {
      "entropy": 6.338523483276367,
      "epoch": 0.06515202138322754,
      "grad_norm": 1.28125,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.091,
      "mean_token_accuracy": 0.13004501089453696,
      "num_tokens": 1428000.0,
      "step": 780
    },
    {
      "entropy": 6.260497093200684,
      "epoch": 0.06556966254594053,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.2992,
      "mean_token_accuracy": 0.12507489323616028,
      "num_tokens": 1437789.0,
      "step": 785
    },
    {
      "entropy": 6.270367765426636,
      "epoch": 0.06598730370865352,
      "grad_norm": 1.359375,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.1214,
      "mean_token_accuracy": 0.1259957693517208,
      "num_tokens": 1446848.0,
      "step": 790
    },
    {
      "entropy": 6.279500675201416,
      "epoch": 0.06640494487136653,
      "grad_norm": 1.25,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.1479,
      "mean_token_accuracy": 0.1257844291627407,
      "num_tokens": 1455749.0,
      "step": 795
    },
    {
      "entropy": 6.253476810455322,
      "epoch": 0.06682258603407952,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003995,
      "loss": 6.0701,
      "mean_token_accuracy": 0.1354301691055298,
      "num_tokens": 1464556.0,
      "step": 800
    },
    {
      "entropy": 6.333716917037964,
      "epoch": 0.06724022719679251,
      "grad_norm": 1.28125,
      "learning_rate": 0.000402,
      "loss": 6.2213,
      "mean_token_accuracy": 0.12829236462712287,
      "num_tokens": 1473617.0,
      "step": 805
    },
    {
      "entropy": 6.306687116622925,
      "epoch": 0.06765786835950552,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004045,
      "loss": 6.2385,
      "mean_token_accuracy": 0.12713917568325997,
      "num_tokens": 1483006.0,
      "step": 810
    },
    {
      "entropy": 6.430218935012817,
      "epoch": 0.06807550952221851,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.2582,
      "mean_token_accuracy": 0.1252996154129505,
      "num_tokens": 1493009.0,
      "step": 815
    },
    {
      "entropy": 6.253592920303345,
      "epoch": 0.0684931506849315,
      "grad_norm": 1.25,
      "learning_rate": 0.0004095,
      "loss": 6.1341,
      "mean_token_accuracy": 0.1317764475941658,
      "num_tokens": 1501261.0,
      "step": 820
    },
    {
      "entropy": 6.3305917263031,
      "epoch": 0.06891079184764451,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000412,
      "loss": 6.2536,
      "mean_token_accuracy": 0.12618372738361358,
      "num_tokens": 1511623.0,
      "step": 825
    },
    {
      "entropy": 6.283257102966308,
      "epoch": 0.0693284330103575,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004145,
      "loss": 6.2422,
      "mean_token_accuracy": 0.12288018986582756,
      "num_tokens": 1521741.0,
      "step": 830
    },
    {
      "entropy": 6.252234697341919,
      "epoch": 0.0697460741730705,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000417,
      "loss": 6.1818,
      "mean_token_accuracy": 0.12901844680309296,
      "num_tokens": 1530988.0,
      "step": 835
    },
    {
      "entropy": 6.257218217849731,
      "epoch": 0.0701637153357835,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004195,
      "loss": 6.0105,
      "mean_token_accuracy": 0.13119714483618736,
      "num_tokens": 1539415.0,
      "step": 840
    },
    {
      "entropy": 6.338153076171875,
      "epoch": 0.07058135649849649,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000422,
      "loss": 6.209,
      "mean_token_accuracy": 0.12688441276550294,
      "num_tokens": 1547969.0,
      "step": 845
    },
    {
      "entropy": 6.164967966079712,
      "epoch": 0.07099899766120948,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004245,
      "loss": 6.08,
      "mean_token_accuracy": 0.1340305209159851,
      "num_tokens": 1557526.0,
      "step": 850
    },
    {
      "entropy": 6.167168045043946,
      "epoch": 0.07141663882392249,
      "grad_norm": 1.296875,
      "learning_rate": 0.000427,
      "loss": 6.0501,
      "mean_token_accuracy": 0.13036102801561356,
      "num_tokens": 1565982.0,
      "step": 855
    },
    {
      "entropy": 6.178088760375976,
      "epoch": 0.07183427998663548,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004295,
      "loss": 6.101,
      "mean_token_accuracy": 0.13804332464933394,
      "num_tokens": 1574492.0,
      "step": 860
    },
    {
      "entropy": 6.236712503433227,
      "epoch": 0.07225192114934847,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000432,
      "loss": 6.0894,
      "mean_token_accuracy": 0.12888288199901582,
      "num_tokens": 1583726.0,
      "step": 865
    },
    {
      "entropy": 6.159904909133911,
      "epoch": 0.07266956231206148,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004345,
      "loss": 6.1032,
      "mean_token_accuracy": 0.12467294186353683,
      "num_tokens": 1593780.0,
      "step": 870
    },
    {
      "entropy": 6.252811336517334,
      "epoch": 0.07308720347477447,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000437,
      "loss": 6.1079,
      "mean_token_accuracy": 0.1311504900455475,
      "num_tokens": 1602746.0,
      "step": 875
    },
    {
      "entropy": 6.235387468338013,
      "epoch": 0.07350484463748747,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004395,
      "loss": 6.1317,
      "mean_token_accuracy": 0.13153513595461847,
      "num_tokens": 1611767.0,
      "step": 880
    },
    {
      "entropy": 6.322087335586548,
      "epoch": 0.07392248580020047,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000442,
      "loss": 6.223,
      "mean_token_accuracy": 0.13359929099678994,
      "num_tokens": 1621568.0,
      "step": 885
    },
    {
      "entropy": 6.234439277648926,
      "epoch": 0.07434012696291346,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004445,
      "loss": 6.1574,
      "mean_token_accuracy": 0.1362864576280117,
      "num_tokens": 1630650.0,
      "step": 890
    },
    {
      "entropy": 6.207277011871338,
      "epoch": 0.07475776812562646,
      "grad_norm": 1.265625,
      "learning_rate": 0.000447,
      "loss": 6.1565,
      "mean_token_accuracy": 0.1267276994884014,
      "num_tokens": 1640482.0,
      "step": 895
    },
    {
      "entropy": 6.311604309082031,
      "epoch": 0.07517540928833946,
      "grad_norm": 1.359375,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.2126,
      "mean_token_accuracy": 0.1277651831507683,
      "num_tokens": 1649344.0,
      "step": 900
    },
    {
      "entropy": 6.199459648132324,
      "epoch": 0.07559305045105245,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.1479,
      "mean_token_accuracy": 0.13086290359497071,
      "num_tokens": 1658962.0,
      "step": 905
    },
    {
      "entropy": 6.197014045715332,
      "epoch": 0.07601069161376545,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.16,
      "mean_token_accuracy": 0.12403983399271964,
      "num_tokens": 1668615.0,
      "step": 910
    },
    {
      "entropy": 6.247787952423096,
      "epoch": 0.07642833277647845,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.0844,
      "mean_token_accuracy": 0.1306306391954422,
      "num_tokens": 1679282.0,
      "step": 915
    },
    {
      "entropy": 6.110937595367432,
      "epoch": 0.07684597393919144,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.1861,
      "mean_token_accuracy": 0.13143347576260567,
      "num_tokens": 1688362.0,
      "step": 920
    },
    {
      "entropy": 6.203617334365845,
      "epoch": 0.07726361510190444,
      "grad_norm": 1.046875,
      "learning_rate": 0.000462,
      "loss": 6.0853,
      "mean_token_accuracy": 0.12813379019498825,
      "num_tokens": 1697593.0,
      "step": 925
    },
    {
      "entropy": 6.284222555160523,
      "epoch": 0.07768125626461744,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004645,
      "loss": 6.135,
      "mean_token_accuracy": 0.1326735071837902,
      "num_tokens": 1706961.0,
      "step": 930
    },
    {
      "entropy": 6.214397430419922,
      "epoch": 0.07809889742733044,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000467,
      "loss": 6.1837,
      "mean_token_accuracy": 0.13653906658291817,
      "num_tokens": 1716222.0,
      "step": 935
    },
    {
      "entropy": 6.132423162460327,
      "epoch": 0.07851653859004344,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004695,
      "loss": 6.0741,
      "mean_token_accuracy": 0.12650192454457282,
      "num_tokens": 1726750.0,
      "step": 940
    },
    {
      "entropy": 6.258601665496826,
      "epoch": 0.07893417975275643,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000472,
      "loss": 6.1501,
      "mean_token_accuracy": 0.12657316625118256,
      "num_tokens": 1736262.0,
      "step": 945
    },
    {
      "entropy": 6.192229986190796,
      "epoch": 0.07935182091546943,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004745,
      "loss": 6.1729,
      "mean_token_accuracy": 0.13067680075764657,
      "num_tokens": 1745555.0,
      "step": 950
    },
    {
      "entropy": 6.050552940368652,
      "epoch": 0.07976946207818243,
      "grad_norm": 1.359375,
      "learning_rate": 0.000477,
      "loss": 6.0643,
      "mean_token_accuracy": 0.12940090522170067,
      "num_tokens": 1754107.0,
      "step": 955
    },
    {
      "entropy": 6.305977058410645,
      "epoch": 0.08018710324089542,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004795,
      "loss": 6.2685,
      "mean_token_accuracy": 0.12384195700287819,
      "num_tokens": 1762982.0,
      "step": 960
    },
    {
      "entropy": 6.171813297271728,
      "epoch": 0.08060474440360842,
      "grad_norm": 1.203125,
      "learning_rate": 0.000482,
      "loss": 6.135,
      "mean_token_accuracy": 0.1338811330497265,
      "num_tokens": 1772545.0,
      "step": 965
    },
    {
      "entropy": 6.244448280334472,
      "epoch": 0.08102238556632142,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004845,
      "loss": 6.2549,
      "mean_token_accuracy": 0.12475204095244408,
      "num_tokens": 1782052.0,
      "step": 970
    },
    {
      "entropy": 6.1756124019622805,
      "epoch": 0.08144002672903441,
      "grad_norm": 1.171875,
      "learning_rate": 0.000487,
      "loss": 6.0588,
      "mean_token_accuracy": 0.1337449759244919,
      "num_tokens": 1791682.0,
      "step": 975
    },
    {
      "entropy": 6.1777324199676515,
      "epoch": 0.08185766789174741,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004895,
      "loss": 6.0759,
      "mean_token_accuracy": 0.13526007980108262,
      "num_tokens": 1800526.0,
      "step": 980
    },
    {
      "entropy": 6.219584560394287,
      "epoch": 0.08227530905446041,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000492,
      "loss": 6.2113,
      "mean_token_accuracy": 0.1325635403394699,
      "num_tokens": 1810062.0,
      "step": 985
    },
    {
      "entropy": 6.163633823394775,
      "epoch": 0.0826929502171734,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004945,
      "loss": 6.1872,
      "mean_token_accuracy": 0.13362277597188948,
      "num_tokens": 1818909.0,
      "step": 990
    },
    {
      "entropy": 6.275831508636474,
      "epoch": 0.0831105913798864,
      "grad_norm": 1.21875,
      "learning_rate": 0.000497,
      "loss": 6.1311,
      "mean_token_accuracy": 0.13871751874685287,
      "num_tokens": 1827978.0,
      "step": 995
    },
    {
      "entropy": 6.068942451477051,
      "epoch": 0.0835282325425994,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004995,
      "loss": 6.1008,
      "mean_token_accuracy": 0.1366957776248455,
      "num_tokens": 1837955.0,
      "step": 1000
    },
    {
      "entropy": 6.199518489837646,
      "epoch": 0.0839458737053124,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004999999987395534,
      "loss": 6.1966,
      "mean_token_accuracy": 0.12947369664907454,
      "num_tokens": 1847699.0,
      "step": 1005
    },
    {
      "entropy": 6.2342846393585205,
      "epoch": 0.08436351486802539,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999999936189888,
      "loss": 6.135,
      "mean_token_accuracy": 0.1308988593518734,
      "num_tokens": 1856490.0,
      "step": 1010
    },
    {
      "entropy": 6.187710952758789,
      "epoch": 0.0847811560307384,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999999845595285,
      "loss": 6.1336,
      "mean_token_accuracy": 0.1302417054772377,
      "num_tokens": 1865644.0,
      "step": 1015
    },
    {
      "entropy": 6.139751815795899,
      "epoch": 0.08519879719345139,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004999999715611727,
      "loss": 6.0628,
      "mean_token_accuracy": 0.13923914581537247,
      "num_tokens": 1874450.0,
      "step": 1020
    },
    {
      "entropy": 6.1254647731781,
      "epoch": 0.08561643835616438,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999999546239215,
      "loss": 6.0891,
      "mean_token_accuracy": 0.13145651072263717,
      "num_tokens": 1883259.0,
      "step": 1025
    },
    {
      "entropy": 6.193487119674683,
      "epoch": 0.08603407951887738,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999999337477753,
      "loss": 6.1689,
      "mean_token_accuracy": 0.12691283002495765,
      "num_tokens": 1891897.0,
      "step": 1030
    },
    {
      "entropy": 6.166404056549072,
      "epoch": 0.08645172068159038,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004999999089327344,
      "loss": 6.1476,
      "mean_token_accuracy": 0.12759166434407235,
      "num_tokens": 1901011.0,
      "step": 1035
    },
    {
      "entropy": 6.058528995513916,
      "epoch": 0.08686936184430337,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004999998801787993,
      "loss": 6.05,
      "mean_token_accuracy": 0.13580385074019433,
      "num_tokens": 1910132.0,
      "step": 1040
    },
    {
      "entropy": 6.21063199043274,
      "epoch": 0.08728700300701638,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004999998474859704,
      "loss": 6.1796,
      "mean_token_accuracy": 0.13001134023070335,
      "num_tokens": 1919938.0,
      "step": 1045
    },
    {
      "entropy": 6.021421146392822,
      "epoch": 0.08770464416972937,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999998108542483,
      "loss": 6.0474,
      "mean_token_accuracy": 0.1303606778383255,
      "num_tokens": 1928199.0,
      "step": 1050
    },
    {
      "entropy": 6.149118709564209,
      "epoch": 0.08812228533244236,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004999997702836339,
      "loss": 6.0384,
      "mean_token_accuracy": 0.1368630550801754,
      "num_tokens": 1937133.0,
      "step": 1055
    },
    {
      "entropy": 6.19862322807312,
      "epoch": 0.08853992649515537,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999997257741274,
      "loss": 6.2007,
      "mean_token_accuracy": 0.13205777183175088,
      "num_tokens": 1946580.0,
      "step": 1060
    },
    {
      "entropy": 6.109546327590943,
      "epoch": 0.08895756765786836,
      "grad_norm": 1.15625,
      "learning_rate": 0.00049999967732573,
      "loss": 6.1303,
      "mean_token_accuracy": 0.1276082694530487,
      "num_tokens": 1955413.0,
      "step": 1065
    },
    {
      "entropy": 6.293388748168946,
      "epoch": 0.08937520882058135,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999996249384425,
      "loss": 6.1897,
      "mean_token_accuracy": 0.1298133671283722,
      "num_tokens": 1964716.0,
      "step": 1070
    },
    {
      "entropy": 6.0309676170349125,
      "epoch": 0.08979284998329436,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999995686122657,
      "loss": 6.0429,
      "mean_token_accuracy": 0.13794475570321083,
      "num_tokens": 1974081.0,
      "step": 1075
    },
    {
      "entropy": 6.057157135009765,
      "epoch": 0.09021049114600735,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004999995083472006,
      "loss": 6.0207,
      "mean_token_accuracy": 0.13763214722275735,
      "num_tokens": 1983035.0,
      "step": 1080
    },
    {
      "entropy": 6.133276653289795,
      "epoch": 0.09062813230872034,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999994441432481,
      "loss": 6.0401,
      "mean_token_accuracy": 0.1399047665297985,
      "num_tokens": 1992475.0,
      "step": 1085
    },
    {
      "entropy": 6.012667560577393,
      "epoch": 0.09104577347143335,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999993760004097,
      "loss": 6.1266,
      "mean_token_accuracy": 0.12739486023783683,
      "num_tokens": 2002729.0,
      "step": 1090
    },
    {
      "entropy": 6.0253236293792725,
      "epoch": 0.09146341463414634,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999993039186864,
      "loss": 5.9623,
      "mean_token_accuracy": 0.13841661661863328,
      "num_tokens": 2011319.0,
      "step": 1095
    },
    {
      "entropy": 6.083655738830567,
      "epoch": 0.09188105579685933,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999992278980794,
      "loss": 5.9924,
      "mean_token_accuracy": 0.1362566351890564,
      "num_tokens": 2020456.0,
      "step": 1100
    },
    {
      "entropy": 6.034357261657715,
      "epoch": 0.09229869695957234,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004999991479385901,
      "loss": 6.0927,
      "mean_token_accuracy": 0.13392627835273743,
      "num_tokens": 2029417.0,
      "step": 1105
    },
    {
      "entropy": 6.154821872711182,
      "epoch": 0.09271633812228533,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999990640402198,
      "loss": 6.1738,
      "mean_token_accuracy": 0.1311173379421234,
      "num_tokens": 2039290.0,
      "step": 1110
    },
    {
      "entropy": 6.101737451553345,
      "epoch": 0.09313397928499834,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004999989762029701,
      "loss": 5.9871,
      "mean_token_accuracy": 0.13381630852818488,
      "num_tokens": 2048485.0,
      "step": 1115
    },
    {
      "entropy": 5.9683490753173825,
      "epoch": 0.09355162044771133,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999988844268425,
      "loss": 5.9595,
      "mean_token_accuracy": 0.14155827090144157,
      "num_tokens": 2057131.0,
      "step": 1120
    },
    {
      "entropy": 6.020996618270874,
      "epoch": 0.09396926161042432,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999987887118387,
      "loss": 5.9606,
      "mean_token_accuracy": 0.1338317573070526,
      "num_tokens": 2066295.0,
      "step": 1125
    },
    {
      "entropy": 6.021601724624634,
      "epoch": 0.09438690277313733,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004999986890579602,
      "loss": 5.9637,
      "mean_token_accuracy": 0.1408805377781391,
      "num_tokens": 2075656.0,
      "step": 1130
    },
    {
      "entropy": 6.0348409652709964,
      "epoch": 0.09480454393585032,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999985854652087,
      "loss": 6.1381,
      "mean_token_accuracy": 0.129783895611763,
      "num_tokens": 2086046.0,
      "step": 1135
    },
    {
      "entropy": 6.1812419414520265,
      "epoch": 0.09522218509856331,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999984779335862,
      "loss": 6.1262,
      "mean_token_accuracy": 0.13610694855451583,
      "num_tokens": 2094808.0,
      "step": 1140
    },
    {
      "entropy": 5.959377193450928,
      "epoch": 0.09563982626127632,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999983664630945,
      "loss": 6.0585,
      "mean_token_accuracy": 0.1390154741704464,
      "num_tokens": 2104155.0,
      "step": 1145
    },
    {
      "entropy": 6.033576393127442,
      "epoch": 0.09605746742398931,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999982510537357,
      "loss": 5.9463,
      "mean_token_accuracy": 0.1419323816895485,
      "num_tokens": 2112517.0,
      "step": 1150
    },
    {
      "entropy": 5.961051750183105,
      "epoch": 0.0964751085867023,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004999981317055116,
      "loss": 5.9185,
      "mean_token_accuracy": 0.1370644360780716,
      "num_tokens": 2121355.0,
      "step": 1155
    },
    {
      "entropy": 6.07532811164856,
      "epoch": 0.09689274974941531,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004999980084184244,
      "loss": 6.0333,
      "mean_token_accuracy": 0.14000216126441956,
      "num_tokens": 2129278.0,
      "step": 1160
    },
    {
      "entropy": 6.112037467956543,
      "epoch": 0.0973103909121283,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999978811924762,
      "loss": 6.209,
      "mean_token_accuracy": 0.13468564450740814,
      "num_tokens": 2139548.0,
      "step": 1165
    },
    {
      "entropy": 6.011156606674194,
      "epoch": 0.09772803207484129,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004999977500276693,
      "loss": 6.1051,
      "mean_token_accuracy": 0.13282948657870292,
      "num_tokens": 2148765.0,
      "step": 1170
    },
    {
      "entropy": 6.156813764572144,
      "epoch": 0.0981456732375543,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999976149240058,
      "loss": 5.9998,
      "mean_token_accuracy": 0.14147626087069512,
      "num_tokens": 2157482.0,
      "step": 1175
    },
    {
      "entropy": 6.06043438911438,
      "epoch": 0.09856331440026729,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999974758814884,
      "loss": 6.118,
      "mean_token_accuracy": 0.12927817851305007,
      "num_tokens": 2167320.0,
      "step": 1180
    },
    {
      "entropy": 6.090041780471802,
      "epoch": 0.09898095556298028,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999973329001193,
      "loss": 6.0481,
      "mean_token_accuracy": 0.13234857842326164,
      "num_tokens": 2175454.0,
      "step": 1185
    },
    {
      "entropy": 6.048929071426391,
      "epoch": 0.09939859672569329,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999971859799009,
      "loss": 5.99,
      "mean_token_accuracy": 0.137349022179842,
      "num_tokens": 2184391.0,
      "step": 1190
    },
    {
      "entropy": 5.9458996772766115,
      "epoch": 0.09981623788840628,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999970351208361,
      "loss": 5.8738,
      "mean_token_accuracy": 0.1423418328166008,
      "num_tokens": 2193309.0,
      "step": 1195
    },
    {
      "entropy": 6.062114095687866,
      "epoch": 0.10023387905111927,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999968803229275,
      "loss": 6.0268,
      "mean_token_accuracy": 0.13685409501194953,
      "num_tokens": 2202057.0,
      "step": 1200
    },
    {
      "entropy": 6.010811853408813,
      "epoch": 0.10065152021383228,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999967215861775,
      "loss": 6.1207,
      "mean_token_accuracy": 0.13319561630487442,
      "num_tokens": 2212384.0,
      "step": 1205
    },
    {
      "entropy": 6.149314880371094,
      "epoch": 0.10106916137654527,
      "grad_norm": 1.125,
      "learning_rate": 0.000499996558910589,
      "loss": 6.0364,
      "mean_token_accuracy": 0.13858548328280448,
      "num_tokens": 2221451.0,
      "step": 1210
    },
    {
      "entropy": 5.897707605361939,
      "epoch": 0.10148680253925826,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000499996392296165,
      "loss": 5.9346,
      "mean_token_accuracy": 0.1363538235425949,
      "num_tokens": 2230440.0,
      "step": 1215
    },
    {
      "entropy": 6.108274269104004,
      "epoch": 0.10190444370197127,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999962217429083,
      "loss": 6.1278,
      "mean_token_accuracy": 0.1263136975467205,
      "num_tokens": 2239985.0,
      "step": 1220
    },
    {
      "entropy": 6.034344911575317,
      "epoch": 0.10232208486468426,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999960472508219,
      "loss": 5.9906,
      "mean_token_accuracy": 0.13609189465641974,
      "num_tokens": 2250391.0,
      "step": 1225
    },
    {
      "entropy": 6.079465055465699,
      "epoch": 0.10273972602739725,
      "grad_norm": 1.140625,
      "learning_rate": 0.000499995868819909,
      "loss": 6.0937,
      "mean_token_accuracy": 0.12762524262070657,
      "num_tokens": 2260620.0,
      "step": 1230
    },
    {
      "entropy": 6.125043344497681,
      "epoch": 0.10315736719011026,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999956864501725,
      "loss": 6.0771,
      "mean_token_accuracy": 0.13054439499974252,
      "num_tokens": 2271176.0,
      "step": 1235
    },
    {
      "entropy": 6.024161291122437,
      "epoch": 0.10357500835282325,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999955001416156,
      "loss": 5.9637,
      "mean_token_accuracy": 0.1394001565873623,
      "num_tokens": 2280371.0,
      "step": 1240
    },
    {
      "entropy": 5.873808765411377,
      "epoch": 0.10399264951553625,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999953098942417,
      "loss": 5.9383,
      "mean_token_accuracy": 0.13987089917063714,
      "num_tokens": 2289015.0,
      "step": 1245
    },
    {
      "entropy": 5.993573093414307,
      "epoch": 0.10441029067824925,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004999951157080541,
      "loss": 6.011,
      "mean_token_accuracy": 0.14008186012506485,
      "num_tokens": 2297409.0,
      "step": 1250
    },
    {
      "entropy": 6.020661783218384,
      "epoch": 0.10482793184096224,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999949175830561,
      "loss": 6.0346,
      "mean_token_accuracy": 0.137710864841938,
      "num_tokens": 2307017.0,
      "step": 1255
    },
    {
      "entropy": 5.966087722778321,
      "epoch": 0.10524557300367524,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999947155192514,
      "loss": 5.8676,
      "mean_token_accuracy": 0.14721441268920898,
      "num_tokens": 2316014.0,
      "step": 1260
    },
    {
      "entropy": 6.035770273208618,
      "epoch": 0.10566321416638824,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999945095166433,
      "loss": 5.9742,
      "mean_token_accuracy": 0.14057080522179605,
      "num_tokens": 2326326.0,
      "step": 1265
    },
    {
      "entropy": 6.115755558013916,
      "epoch": 0.10608085532910123,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004999942995752354,
      "loss": 6.0694,
      "mean_token_accuracy": 0.13691574186086655,
      "num_tokens": 2335108.0,
      "step": 1270
    },
    {
      "entropy": 5.959864377975464,
      "epoch": 0.10649849649181423,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004999940856950315,
      "loss": 5.9885,
      "mean_token_accuracy": 0.14240883886814118,
      "num_tokens": 2343971.0,
      "step": 1275
    },
    {
      "entropy": 6.033252716064453,
      "epoch": 0.10691613765452723,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999938678760354,
      "loss": 6.0546,
      "mean_token_accuracy": 0.13366041257977485,
      "num_tokens": 2353742.0,
      "step": 1280
    },
    {
      "entropy": 5.9920961380004885,
      "epoch": 0.10733377881724022,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999936461182507,
      "loss": 5.9948,
      "mean_token_accuracy": 0.14326683208346366,
      "num_tokens": 2363629.0,
      "step": 1285
    },
    {
      "entropy": 5.8916154384613035,
      "epoch": 0.10775141997995323,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999934204216814,
      "loss": 5.8557,
      "mean_token_accuracy": 0.14521737545728683,
      "num_tokens": 2372705.0,
      "step": 1290
    },
    {
      "entropy": 5.9855352401733395,
      "epoch": 0.10816906114266622,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999931907863315,
      "loss": 5.9681,
      "mean_token_accuracy": 0.14111716747283937,
      "num_tokens": 2382136.0,
      "step": 1295
    },
    {
      "entropy": 5.877673768997193,
      "epoch": 0.10858670230537922,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000499992957212205,
      "loss": 5.9341,
      "mean_token_accuracy": 0.1387516625225544,
      "num_tokens": 2391060.0,
      "step": 1300
    },
    {
      "entropy": 6.04531421661377,
      "epoch": 0.10900434346809222,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999927196993059,
      "loss": 6.0668,
      "mean_token_accuracy": 0.136396811157465,
      "num_tokens": 2399885.0,
      "step": 1305
    },
    {
      "entropy": 6.001427459716797,
      "epoch": 0.10942198463080521,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999924782476384,
      "loss": 5.931,
      "mean_token_accuracy": 0.14695603102445604,
      "num_tokens": 2409654.0,
      "step": 1310
    },
    {
      "entropy": 6.023379850387573,
      "epoch": 0.1098396257935182,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999922328572068,
      "loss": 6.0087,
      "mean_token_accuracy": 0.13872749507427215,
      "num_tokens": 2419342.0,
      "step": 1315
    },
    {
      "entropy": 5.900698280334472,
      "epoch": 0.11025726695623121,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999919835280154,
      "loss": 5.8906,
      "mean_token_accuracy": 0.14329665824770926,
      "num_tokens": 2428845.0,
      "step": 1320
    },
    {
      "entropy": 6.046743535995484,
      "epoch": 0.1106749081189442,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999917302600684,
      "loss": 5.8937,
      "mean_token_accuracy": 0.14875416159629823,
      "num_tokens": 2437134.0,
      "step": 1325
    },
    {
      "entropy": 5.708086442947388,
      "epoch": 0.1110925492816572,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999914730533704,
      "loss": 5.7293,
      "mean_token_accuracy": 0.15531845688819884,
      "num_tokens": 2445438.0,
      "step": 1330
    },
    {
      "entropy": 6.002161979675293,
      "epoch": 0.1115101904443702,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999912119079258,
      "loss": 6.0527,
      "mean_token_accuracy": 0.13914170265197753,
      "num_tokens": 2455039.0,
      "step": 1335
    },
    {
      "entropy": 6.0124329090118405,
      "epoch": 0.1119278316070832,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999909468237393,
      "loss": 5.9927,
      "mean_token_accuracy": 0.13686869218945502,
      "num_tokens": 2465170.0,
      "step": 1340
    },
    {
      "entropy": 5.87831563949585,
      "epoch": 0.11234547276979619,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999906778008153,
      "loss": 5.7413,
      "mean_token_accuracy": 0.1533431127667427,
      "num_tokens": 2473688.0,
      "step": 1345
    },
    {
      "entropy": 5.915750741958618,
      "epoch": 0.11276311393250919,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999904048391588,
      "loss": 6.0191,
      "mean_token_accuracy": 0.13419281020760537,
      "num_tokens": 2482335.0,
      "step": 1350
    },
    {
      "entropy": 5.989591646194458,
      "epoch": 0.11318075509522219,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999901279387743,
      "loss": 5.9312,
      "mean_token_accuracy": 0.14448264613747597,
      "num_tokens": 2491808.0,
      "step": 1355
    },
    {
      "entropy": 6.086487817764282,
      "epoch": 0.11359839625793518,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999898470996669,
      "loss": 6.0697,
      "mean_token_accuracy": 0.13440220579504966,
      "num_tokens": 2501422.0,
      "step": 1360
    },
    {
      "entropy": 5.9091236114501955,
      "epoch": 0.11401603742064818,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999895623218415,
      "loss": 5.8628,
      "mean_token_accuracy": 0.14603495746850967,
      "num_tokens": 2511201.0,
      "step": 1365
    },
    {
      "entropy": 5.936640071868896,
      "epoch": 0.11443367858336118,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999892736053027,
      "loss": 5.8957,
      "mean_token_accuracy": 0.1403353899717331,
      "num_tokens": 2519798.0,
      "step": 1370
    },
    {
      "entropy": 5.891090345382691,
      "epoch": 0.11485131974607417,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999889809500561,
      "loss": 5.8755,
      "mean_token_accuracy": 0.1445053994655609,
      "num_tokens": 2529177.0,
      "step": 1375
    },
    {
      "entropy": 6.015834188461303,
      "epoch": 0.11526896090878717,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999886843561065,
      "loss": 5.8692,
      "mean_token_accuracy": 0.14721576496958733,
      "num_tokens": 2537571.0,
      "step": 1380
    },
    {
      "entropy": 5.768728446960449,
      "epoch": 0.11568660207150017,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999883838234592,
      "loss": 5.7907,
      "mean_token_accuracy": 0.1494941793382168,
      "num_tokens": 2546258.0,
      "step": 1385
    },
    {
      "entropy": 5.879445552825928,
      "epoch": 0.11610424323421316,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999880793521194,
      "loss": 5.8101,
      "mean_token_accuracy": 0.14160490334033965,
      "num_tokens": 2555862.0,
      "step": 1390
    },
    {
      "entropy": 5.703860473632813,
      "epoch": 0.11652188439692616,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999877709420924,
      "loss": 5.8276,
      "mean_token_accuracy": 0.14771698862314225,
      "num_tokens": 2564583.0,
      "step": 1395
    },
    {
      "entropy": 5.946020936965942,
      "epoch": 0.11693952555963916,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999874585933837,
      "loss": 5.9224,
      "mean_token_accuracy": 0.14428764581680298,
      "num_tokens": 2574217.0,
      "step": 1400
    },
    {
      "entropy": 5.874564838409424,
      "epoch": 0.11735716672235215,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999871423059987,
      "loss": 5.8278,
      "mean_token_accuracy": 0.14492169469594957,
      "num_tokens": 2582821.0,
      "step": 1405
    },
    {
      "entropy": 5.90300989151001,
      "epoch": 0.11777480788506516,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499986822079943,
      "loss": 6.0026,
      "mean_token_accuracy": 0.1350848987698555,
      "num_tokens": 2592748.0,
      "step": 1410
    },
    {
      "entropy": 5.979686975479126,
      "epoch": 0.11819244904777815,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999864979152221,
      "loss": 5.8824,
      "mean_token_accuracy": 0.14854978621006013,
      "num_tokens": 2602053.0,
      "step": 1415
    },
    {
      "entropy": 5.808524131774902,
      "epoch": 0.11861009021049114,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999861698118417,
      "loss": 5.9717,
      "mean_token_accuracy": 0.14355539306998252,
      "num_tokens": 2610557.0,
      "step": 1420
    },
    {
      "entropy": 5.955320358276367,
      "epoch": 0.11902773137320415,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999858377698076,
      "loss": 5.8513,
      "mean_token_accuracy": 0.14397183433175087,
      "num_tokens": 2619628.0,
      "step": 1425
    },
    {
      "entropy": 6.067613363265991,
      "epoch": 0.11944537253591714,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999855017891256,
      "loss": 6.0207,
      "mean_token_accuracy": 0.14067387729883193,
      "num_tokens": 2629284.0,
      "step": 1430
    },
    {
      "entropy": 5.882795715332032,
      "epoch": 0.11986301369863013,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999851618698017,
      "loss": 5.9417,
      "mean_token_accuracy": 0.14827706664800644,
      "num_tokens": 2638001.0,
      "step": 1435
    },
    {
      "entropy": 5.891476774215699,
      "epoch": 0.12028065486134314,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999848180118416,
      "loss": 5.9157,
      "mean_token_accuracy": 0.15219354629516602,
      "num_tokens": 2647071.0,
      "step": 1440
    },
    {
      "entropy": 5.973889017105103,
      "epoch": 0.12069829602405613,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999844702152516,
      "loss": 5.8907,
      "mean_token_accuracy": 0.14990584850311278,
      "num_tokens": 2655826.0,
      "step": 1445
    },
    {
      "entropy": 5.833336019515992,
      "epoch": 0.12111593718676913,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999841184800374,
      "loss": 5.7691,
      "mean_token_accuracy": 0.15365617871284484,
      "num_tokens": 2664431.0,
      "step": 1450
    },
    {
      "entropy": 5.761509227752685,
      "epoch": 0.12153357834948213,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999837628062055,
      "loss": 5.6995,
      "mean_token_accuracy": 0.15763423219323158,
      "num_tokens": 2673631.0,
      "step": 1455
    },
    {
      "entropy": 5.807348442077637,
      "epoch": 0.12195121951219512,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999834031937621,
      "loss": 5.9755,
      "mean_token_accuracy": 0.1450788274407387,
      "num_tokens": 2683363.0,
      "step": 1460
    },
    {
      "entropy": 6.004406070709228,
      "epoch": 0.12236886067490813,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999830396427133,
      "loss": 5.9219,
      "mean_token_accuracy": 0.14206716865301133,
      "num_tokens": 2692533.0,
      "step": 1465
    },
    {
      "entropy": 5.8194908618927,
      "epoch": 0.12278650183762112,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999826721530657,
      "loss": 5.9232,
      "mean_token_accuracy": 0.15276242047548294,
      "num_tokens": 2701211.0,
      "step": 1470
    },
    {
      "entropy": 5.959702110290527,
      "epoch": 0.12320414300033411,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999823007248255,
      "loss": 5.9036,
      "mean_token_accuracy": 0.14682099223136902,
      "num_tokens": 2709543.0,
      "step": 1475
    },
    {
      "entropy": 5.9004326343536375,
      "epoch": 0.12362178416304712,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004999819253579992,
      "loss": 5.9041,
      "mean_token_accuracy": 0.14219618514180182,
      "num_tokens": 2718336.0,
      "step": 1480
    },
    {
      "entropy": 5.887391233444214,
      "epoch": 0.12403942532576011,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999815460525938,
      "loss": 5.8491,
      "mean_token_accuracy": 0.15181448459625244,
      "num_tokens": 2726504.0,
      "step": 1485
    },
    {
      "entropy": 5.889515399932861,
      "epoch": 0.1244570664884731,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999811628086155,
      "loss": 5.8877,
      "mean_token_accuracy": 0.14508116692304612,
      "num_tokens": 2736076.0,
      "step": 1490
    },
    {
      "entropy": 5.805719614028931,
      "epoch": 0.1248747076511861,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499980775626071,
      "loss": 5.7985,
      "mean_token_accuracy": 0.1503925770521164,
      "num_tokens": 2745497.0,
      "step": 1495
    },
    {
      "entropy": 5.8484405994415285,
      "epoch": 0.12529234881389908,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999803845049674,
      "loss": 5.9377,
      "mean_token_accuracy": 0.14518412500619887,
      "num_tokens": 2754455.0,
      "step": 1500
    },
    {
      "entropy": 6.031361675262451,
      "epoch": 0.1257099899766121,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999799894453112,
      "loss": 5.9596,
      "mean_token_accuracy": 0.14515393003821372,
      "num_tokens": 2763603.0,
      "step": 1505
    },
    {
      "entropy": 5.842563438415527,
      "epoch": 0.1261276311393251,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999795904471095,
      "loss": 5.8325,
      "mean_token_accuracy": 0.14701997935771943,
      "num_tokens": 2772152.0,
      "step": 1510
    },
    {
      "entropy": 5.936494588851929,
      "epoch": 0.12654527230203808,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999791875103694,
      "loss": 5.9263,
      "mean_token_accuracy": 0.14711305350065232,
      "num_tokens": 2781459.0,
      "step": 1515
    },
    {
      "entropy": 5.912358713150025,
      "epoch": 0.12696291346475108,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999787806350977,
      "loss": 5.8773,
      "mean_token_accuracy": 0.14168751537799834,
      "num_tokens": 2791177.0,
      "step": 1520
    },
    {
      "entropy": 5.809263038635254,
      "epoch": 0.1273805546274641,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999783698213016,
      "loss": 5.8743,
      "mean_token_accuracy": 0.1511882796883583,
      "num_tokens": 2800543.0,
      "step": 1525
    },
    {
      "entropy": 5.718258190155029,
      "epoch": 0.12779819579017707,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999779550689885,
      "loss": 5.7157,
      "mean_token_accuracy": 0.1562495306134224,
      "num_tokens": 2809301.0,
      "step": 1530
    },
    {
      "entropy": 5.946142387390137,
      "epoch": 0.12821583695289007,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999775363781654,
      "loss": 5.9216,
      "mean_token_accuracy": 0.13951645120978357,
      "num_tokens": 2818269.0,
      "step": 1535
    },
    {
      "entropy": 5.855810689926147,
      "epoch": 0.12863347811560308,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999771137488395,
      "loss": 5.7879,
      "mean_token_accuracy": 0.15382847785949708,
      "num_tokens": 2827170.0,
      "step": 1540
    },
    {
      "entropy": 5.764991426467896,
      "epoch": 0.12905111927831608,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999766871810188,
      "loss": 5.7656,
      "mean_token_accuracy": 0.15415056496858598,
      "num_tokens": 2835320.0,
      "step": 1545
    },
    {
      "entropy": 5.850279855728149,
      "epoch": 0.12946876044102906,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999762566747101,
      "loss": 5.8505,
      "mean_token_accuracy": 0.14932266920804976,
      "num_tokens": 2843755.0,
      "step": 1550
    },
    {
      "entropy": 5.914557886123657,
      "epoch": 0.12988640160374207,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999758222299214,
      "loss": 5.8955,
      "mean_token_accuracy": 0.1388828009366989,
      "num_tokens": 2853462.0,
      "step": 1555
    },
    {
      "entropy": 5.80288872718811,
      "epoch": 0.13030404276645507,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999753838466601,
      "loss": 5.8131,
      "mean_token_accuracy": 0.14732857197523117,
      "num_tokens": 2863041.0,
      "step": 1560
    },
    {
      "entropy": 6.004314374923706,
      "epoch": 0.13072168392916805,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999749415249338,
      "loss": 6.0175,
      "mean_token_accuracy": 0.14584584459662436,
      "num_tokens": 2873019.0,
      "step": 1565
    },
    {
      "entropy": 5.770129728317261,
      "epoch": 0.13113932509188106,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999744952647505,
      "loss": 5.7567,
      "mean_token_accuracy": 0.14778472259640693,
      "num_tokens": 2881467.0,
      "step": 1570
    },
    {
      "entropy": 5.8767143249511715,
      "epoch": 0.13155696625459407,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999740450661176,
      "loss": 5.8573,
      "mean_token_accuracy": 0.14663312584161758,
      "num_tokens": 2890914.0,
      "step": 1575
    },
    {
      "entropy": 5.784161996841431,
      "epoch": 0.13197460741730704,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999735909290436,
      "loss": 5.7828,
      "mean_token_accuracy": 0.14994997084140776,
      "num_tokens": 2901164.0,
      "step": 1580
    },
    {
      "entropy": 5.883551549911499,
      "epoch": 0.13239224858002005,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999731328535358,
      "loss": 5.9144,
      "mean_token_accuracy": 0.1454067997634411,
      "num_tokens": 2910560.0,
      "step": 1585
    },
    {
      "entropy": 5.796862983703614,
      "epoch": 0.13280988974273306,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999726708396026,
      "loss": 5.7264,
      "mean_token_accuracy": 0.1551614746451378,
      "num_tokens": 2918980.0,
      "step": 1590
    },
    {
      "entropy": 5.845749855041504,
      "epoch": 0.13322753090544603,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499972204887252,
      "loss": 5.9292,
      "mean_token_accuracy": 0.14284173548221588,
      "num_tokens": 2929516.0,
      "step": 1595
    },
    {
      "entropy": 5.866528034210205,
      "epoch": 0.13364517206815904,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999717349964922,
      "loss": 6.0135,
      "mean_token_accuracy": 0.13401781842112542,
      "num_tokens": 2940603.0,
      "step": 1600
    },
    {
      "entropy": 5.975286054611206,
      "epoch": 0.13406281323087205,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999712611673313,
      "loss": 5.8426,
      "mean_token_accuracy": 0.1439616158604622,
      "num_tokens": 2950082.0,
      "step": 1605
    },
    {
      "entropy": 5.723266506195069,
      "epoch": 0.13448045439358502,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999707833997777,
      "loss": 5.7138,
      "mean_token_accuracy": 0.1633479043841362,
      "num_tokens": 2959098.0,
      "step": 1610
    },
    {
      "entropy": 5.87523307800293,
      "epoch": 0.13489809555629803,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999703016938398,
      "loss": 5.8386,
      "mean_token_accuracy": 0.15109211057424546,
      "num_tokens": 2968585.0,
      "step": 1615
    },
    {
      "entropy": 5.831326818466186,
      "epoch": 0.13531573671901104,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499969816049526,
      "loss": 5.878,
      "mean_token_accuracy": 0.14930004104971886,
      "num_tokens": 2978430.0,
      "step": 1620
    },
    {
      "entropy": 5.996936225891114,
      "epoch": 0.13573337788172402,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999693264668446,
      "loss": 5.9294,
      "mean_token_accuracy": 0.14231408908963203,
      "num_tokens": 2988398.0,
      "step": 1625
    },
    {
      "entropy": 5.878012323379517,
      "epoch": 0.13615101904443702,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999688329458045,
      "loss": 5.9551,
      "mean_token_accuracy": 0.14286330938339234,
      "num_tokens": 2999255.0,
      "step": 1630
    },
    {
      "entropy": 5.927907991409302,
      "epoch": 0.13656866020715003,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999683354864141,
      "loss": 5.8266,
      "mean_token_accuracy": 0.14798136353492736,
      "num_tokens": 3008638.0,
      "step": 1635
    },
    {
      "entropy": 5.776216173171997,
      "epoch": 0.136986301369863,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999678340886823,
      "loss": 5.857,
      "mean_token_accuracy": 0.1482193261384964,
      "num_tokens": 3018070.0,
      "step": 1640
    },
    {
      "entropy": 5.791409111022949,
      "epoch": 0.137403942532576,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999673287526176,
      "loss": 5.8119,
      "mean_token_accuracy": 0.14728376343846322,
      "num_tokens": 3027878.0,
      "step": 1645
    },
    {
      "entropy": 5.978458023071289,
      "epoch": 0.13782158369528902,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999668194782291,
      "loss": 5.9885,
      "mean_token_accuracy": 0.1419169157743454,
      "num_tokens": 3037031.0,
      "step": 1650
    },
    {
      "entropy": 5.949056673049927,
      "epoch": 0.138239224858002,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999663062655257,
      "loss": 5.9234,
      "mean_token_accuracy": 0.140587417781353,
      "num_tokens": 3046311.0,
      "step": 1655
    },
    {
      "entropy": 5.834480047225952,
      "epoch": 0.138656866020715,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999657891145162,
      "loss": 5.737,
      "mean_token_accuracy": 0.1507701277732849,
      "num_tokens": 3055108.0,
      "step": 1660
    },
    {
      "entropy": 5.703247451782227,
      "epoch": 0.139074507183428,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999652680252099,
      "loss": 5.7811,
      "mean_token_accuracy": 0.15281276553869247,
      "num_tokens": 3063492.0,
      "step": 1665
    },
    {
      "entropy": 5.96858491897583,
      "epoch": 0.139492148346141,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999647429976157,
      "loss": 5.839,
      "mean_token_accuracy": 0.15251049250364304,
      "num_tokens": 3072572.0,
      "step": 1670
    },
    {
      "entropy": 5.884686231613159,
      "epoch": 0.139909789508854,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999642140317429,
      "loss": 5.8876,
      "mean_token_accuracy": 0.1460355408489704,
      "num_tokens": 3082942.0,
      "step": 1675
    },
    {
      "entropy": 5.78411111831665,
      "epoch": 0.140327430671567,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999636811276007,
      "loss": 5.8438,
      "mean_token_accuracy": 0.14870428889989853,
      "num_tokens": 3091410.0,
      "step": 1680
    },
    {
      "entropy": 5.933750295639038,
      "epoch": 0.14074507183427998,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999631442851986,
      "loss": 5.9126,
      "mean_token_accuracy": 0.1502613291144371,
      "num_tokens": 3100626.0,
      "step": 1685
    },
    {
      "entropy": 5.747744131088257,
      "epoch": 0.14116271299699298,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999626035045458,
      "loss": 5.7155,
      "mean_token_accuracy": 0.15668077766895294,
      "num_tokens": 3109339.0,
      "step": 1690
    },
    {
      "entropy": 5.869397354125977,
      "epoch": 0.141580354159706,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999620587856519,
      "loss": 5.8462,
      "mean_token_accuracy": 0.1474672093987465,
      "num_tokens": 3118636.0,
      "step": 1695
    },
    {
      "entropy": 5.770104169845581,
      "epoch": 0.14199799532241897,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999615101285262,
      "loss": 5.8111,
      "mean_token_accuracy": 0.14746451154351234,
      "num_tokens": 3128162.0,
      "step": 1700
    },
    {
      "entropy": 5.8655181407928465,
      "epoch": 0.14241563648513197,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999609575331786,
      "loss": 5.8179,
      "mean_token_accuracy": 0.15379944443702698,
      "num_tokens": 3136451.0,
      "step": 1705
    },
    {
      "entropy": 5.857501363754272,
      "epoch": 0.14283327764784498,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999604009996186,
      "loss": 5.8331,
      "mean_token_accuracy": 0.14673274308443068,
      "num_tokens": 3146599.0,
      "step": 1710
    },
    {
      "entropy": 5.884437608718872,
      "epoch": 0.14325091881055796,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499959840527856,
      "loss": 5.8732,
      "mean_token_accuracy": 0.14682791605591775,
      "num_tokens": 3155941.0,
      "step": 1715
    },
    {
      "entropy": 5.824529457092285,
      "epoch": 0.14366855997327097,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999592761179007,
      "loss": 5.8059,
      "mean_token_accuracy": 0.15331910401582718,
      "num_tokens": 3164641.0,
      "step": 1720
    },
    {
      "entropy": 5.833645057678223,
      "epoch": 0.14408620113598397,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999587077697624,
      "loss": 5.8266,
      "mean_token_accuracy": 0.15982660353183747,
      "num_tokens": 3173701.0,
      "step": 1725
    },
    {
      "entropy": 5.684833717346192,
      "epoch": 0.14450384229869695,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999581354834512,
      "loss": 5.6887,
      "mean_token_accuracy": 0.1520286552608013,
      "num_tokens": 3182268.0,
      "step": 1730
    },
    {
      "entropy": 5.74345817565918,
      "epoch": 0.14492148346140996,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000499957559258977,
      "loss": 5.7711,
      "mean_token_accuracy": 0.15220290571451187,
      "num_tokens": 3191513.0,
      "step": 1735
    },
    {
      "entropy": 5.77564492225647,
      "epoch": 0.14533912462412296,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00049995697909635,
      "loss": 5.709,
      "mean_token_accuracy": 0.15256546586751937,
      "num_tokens": 3200580.0,
      "step": 1740
    },
    {
      "entropy": 5.737341022491455,
      "epoch": 0.14575676578683594,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999563949955803,
      "loss": 5.8196,
      "mean_token_accuracy": 0.15687910318374634,
      "num_tokens": 3209053.0,
      "step": 1745
    },
    {
      "entropy": 5.808221817016602,
      "epoch": 0.14617440694954895,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499955806956678,
      "loss": 5.8062,
      "mean_token_accuracy": 0.1543448254466057,
      "num_tokens": 3218241.0,
      "step": 1750
    },
    {
      "entropy": 5.880738401412964,
      "epoch": 0.14659204811226195,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999552149796536,
      "loss": 5.8514,
      "mean_token_accuracy": 0.1490916818380356,
      "num_tokens": 3227887.0,
      "step": 1755
    },
    {
      "entropy": 5.768217468261719,
      "epoch": 0.14700968927497493,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999546190645175,
      "loss": 5.6692,
      "mean_token_accuracy": 0.16196279525756835,
      "num_tokens": 3237281.0,
      "step": 1760
    },
    {
      "entropy": 5.684476280212403,
      "epoch": 0.14742733043768794,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999540192112799,
      "loss": 5.7118,
      "mean_token_accuracy": 0.15405311733484267,
      "num_tokens": 3246907.0,
      "step": 1765
    },
    {
      "entropy": 5.660418939590454,
      "epoch": 0.14784497160040094,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999534154199514,
      "loss": 5.7163,
      "mean_token_accuracy": 0.1486714243888855,
      "num_tokens": 3256312.0,
      "step": 1770
    },
    {
      "entropy": 5.922216749191284,
      "epoch": 0.14826261276311392,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999528076905426,
      "loss": 5.7997,
      "mean_token_accuracy": 0.1525414004921913,
      "num_tokens": 3265554.0,
      "step": 1775
    },
    {
      "entropy": 5.621938705444336,
      "epoch": 0.14868025392582693,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999521960230641,
      "loss": 5.6639,
      "mean_token_accuracy": 0.15837963819503784,
      "num_tokens": 3274731.0,
      "step": 1780
    },
    {
      "entropy": 5.7696099281311035,
      "epoch": 0.14909789508853993,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999515804175268,
      "loss": 5.7894,
      "mean_token_accuracy": 0.1475925050675869,
      "num_tokens": 3283518.0,
      "step": 1785
    },
    {
      "entropy": 5.852998208999634,
      "epoch": 0.1495155362512529,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499950960873941,
      "loss": 5.8082,
      "mean_token_accuracy": 0.1533878356218338,
      "num_tokens": 3292244.0,
      "step": 1790
    },
    {
      "entropy": 5.8209583282470705,
      "epoch": 0.14993317741396592,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999503373923182,
      "loss": 5.7746,
      "mean_token_accuracy": 0.15582873076200485,
      "num_tokens": 3301198.0,
      "step": 1795
    },
    {
      "entropy": 5.733056211471558,
      "epoch": 0.15035081857667892,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999497099726687,
      "loss": 5.7494,
      "mean_token_accuracy": 0.15889813750982285,
      "num_tokens": 3309983.0,
      "step": 1800
    },
    {
      "entropy": 5.81738691329956,
      "epoch": 0.1507684597393919,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999490786150039,
      "loss": 5.8432,
      "mean_token_accuracy": 0.147987063229084,
      "num_tokens": 3319313.0,
      "step": 1805
    },
    {
      "entropy": 5.742344903945923,
      "epoch": 0.1511861009021049,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999484433193348,
      "loss": 5.6971,
      "mean_token_accuracy": 0.15073793679475783,
      "num_tokens": 3327497.0,
      "step": 1810
    },
    {
      "entropy": 5.707638788223266,
      "epoch": 0.15160374206481791,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999478040856722,
      "loss": 5.7082,
      "mean_token_accuracy": 0.1553061291575432,
      "num_tokens": 3337029.0,
      "step": 1815
    },
    {
      "entropy": 5.660393095016479,
      "epoch": 0.1520213832275309,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999471609140276,
      "loss": 5.675,
      "mean_token_accuracy": 0.15133372247219085,
      "num_tokens": 3346243.0,
      "step": 1820
    },
    {
      "entropy": 5.844138050079346,
      "epoch": 0.1524390243902439,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999465138044122,
      "loss": 5.8271,
      "mean_token_accuracy": 0.150594063103199,
      "num_tokens": 3354846.0,
      "step": 1825
    },
    {
      "entropy": 5.748153495788574,
      "epoch": 0.1528566655529569,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999458627568373,
      "loss": 5.7794,
      "mean_token_accuracy": 0.1485428184270859,
      "num_tokens": 3364683.0,
      "step": 1830
    },
    {
      "entropy": 5.829200696945191,
      "epoch": 0.15327430671566988,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999452077713144,
      "loss": 5.7401,
      "mean_token_accuracy": 0.14999438524246217,
      "num_tokens": 3374652.0,
      "step": 1835
    },
    {
      "entropy": 5.692996788024902,
      "epoch": 0.1536919478783829,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999445488478547,
      "loss": 5.7464,
      "mean_token_accuracy": 0.1528235137462616,
      "num_tokens": 3384040.0,
      "step": 1840
    },
    {
      "entropy": 5.801222896575927,
      "epoch": 0.1541095890410959,
      "grad_norm": 1.1875,
      "learning_rate": 0.00049994388598647,
      "loss": 5.8162,
      "mean_token_accuracy": 0.14949264377355576,
      "num_tokens": 3393185.0,
      "step": 1845
    },
    {
      "entropy": 5.708631277084351,
      "epoch": 0.15452723020380887,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999432191871719,
      "loss": 5.7497,
      "mean_token_accuracy": 0.15880197882652283,
      "num_tokens": 3403077.0,
      "step": 1850
    },
    {
      "entropy": 5.772182989120483,
      "epoch": 0.15494487136652188,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999425484499718,
      "loss": 5.7194,
      "mean_token_accuracy": 0.16176845729351044,
      "num_tokens": 3412165.0,
      "step": 1855
    },
    {
      "entropy": 5.772862243652344,
      "epoch": 0.1553625125292349,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999418737748817,
      "loss": 5.7694,
      "mean_token_accuracy": 0.14864142686128617,
      "num_tokens": 3421162.0,
      "step": 1860
    },
    {
      "entropy": 5.872934341430664,
      "epoch": 0.15578015369194786,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999411951619134,
      "loss": 5.8646,
      "mean_token_accuracy": 0.1428262896835804,
      "num_tokens": 3430363.0,
      "step": 1865
    },
    {
      "entropy": 5.807984781265259,
      "epoch": 0.15619779485466087,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999405126110786,
      "loss": 5.7669,
      "mean_token_accuracy": 0.15092551931738854,
      "num_tokens": 3439199.0,
      "step": 1870
    },
    {
      "entropy": 5.957257461547852,
      "epoch": 0.15661543601737388,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999398261223895,
      "loss": 5.8731,
      "mean_token_accuracy": 0.14979226514697075,
      "num_tokens": 3448720.0,
      "step": 1875
    },
    {
      "entropy": 5.749342966079712,
      "epoch": 0.15703307718008688,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999391356958579,
      "loss": 5.8455,
      "mean_token_accuracy": 0.14971398413181305,
      "num_tokens": 3458472.0,
      "step": 1880
    },
    {
      "entropy": 5.749511766433716,
      "epoch": 0.15745071834279986,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499938441331496,
      "loss": 5.7952,
      "mean_token_accuracy": 0.15146801769733428,
      "num_tokens": 3467140.0,
      "step": 1885
    },
    {
      "entropy": 5.867589330673217,
      "epoch": 0.15786835950551287,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999377430293159,
      "loss": 5.7745,
      "mean_token_accuracy": 0.15411789417266847,
      "num_tokens": 3476431.0,
      "step": 1890
    },
    {
      "entropy": 5.766841220855713,
      "epoch": 0.15828600066822587,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999370407893299,
      "loss": 5.6863,
      "mean_token_accuracy": 0.15346913933753967,
      "num_tokens": 3485768.0,
      "step": 1895
    },
    {
      "entropy": 5.807104873657226,
      "epoch": 0.15870364183093885,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999363346115501,
      "loss": 5.8548,
      "mean_token_accuracy": 0.14504164904356004,
      "num_tokens": 3495678.0,
      "step": 1900
    },
    {
      "entropy": 5.785052108764648,
      "epoch": 0.15912128299365186,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999356244959893,
      "loss": 5.8342,
      "mean_token_accuracy": 0.1516023814678192,
      "num_tokens": 3505168.0,
      "step": 1905
    },
    {
      "entropy": 5.740941047668457,
      "epoch": 0.15953892415636486,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999349104426594,
      "loss": 5.7732,
      "mean_token_accuracy": 0.15149827301502228,
      "num_tokens": 3514498.0,
      "step": 1910
    },
    {
      "entropy": 5.848452997207642,
      "epoch": 0.15995656531907784,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999341924515732,
      "loss": 5.6975,
      "mean_token_accuracy": 0.15026968866586685,
      "num_tokens": 3524617.0,
      "step": 1915
    },
    {
      "entropy": 5.668032789230347,
      "epoch": 0.16037420648179085,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999334705227431,
      "loss": 5.673,
      "mean_token_accuracy": 0.15639639049768447,
      "num_tokens": 3533523.0,
      "step": 1920
    },
    {
      "entropy": 5.7723455905914305,
      "epoch": 0.16079184764450385,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499932744656182,
      "loss": 5.7652,
      "mean_token_accuracy": 0.15217868536710738,
      "num_tokens": 3542922.0,
      "step": 1925
    },
    {
      "entropy": 5.797222757339478,
      "epoch": 0.16120948880721683,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999320148519024,
      "loss": 5.7533,
      "mean_token_accuracy": 0.154755599796772,
      "num_tokens": 3551827.0,
      "step": 1930
    },
    {
      "entropy": 5.708801507949829,
      "epoch": 0.16162712996992984,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999312811099171,
      "loss": 5.7244,
      "mean_token_accuracy": 0.15864891856908797,
      "num_tokens": 3560911.0,
      "step": 1935
    },
    {
      "entropy": 5.6975208759307865,
      "epoch": 0.16204477113264285,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999305434302391,
      "loss": 5.7543,
      "mean_token_accuracy": 0.14677345082163812,
      "num_tokens": 3569473.0,
      "step": 1940
    },
    {
      "entropy": 5.7810064315795895,
      "epoch": 0.16246241229535582,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499929801812881,
      "loss": 5.745,
      "mean_token_accuracy": 0.15499909371137618,
      "num_tokens": 3578843.0,
      "step": 1945
    },
    {
      "entropy": 5.788356876373291,
      "epoch": 0.16288005345806883,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999290562578561,
      "loss": 5.7561,
      "mean_token_accuracy": 0.15244330316781998,
      "num_tokens": 3588304.0,
      "step": 1950
    },
    {
      "entropy": 5.7127964973449705,
      "epoch": 0.16329769462078184,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999283067651773,
      "loss": 5.8055,
      "mean_token_accuracy": 0.14917491376399994,
      "num_tokens": 3598354.0,
      "step": 1955
    },
    {
      "entropy": 5.894095182418823,
      "epoch": 0.16371533578349481,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999275533348577,
      "loss": 5.8472,
      "mean_token_accuracy": 0.14917806088924407,
      "num_tokens": 3607503.0,
      "step": 1960
    },
    {
      "entropy": 5.645402431488037,
      "epoch": 0.16413297694620782,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999267959669106,
      "loss": 5.6442,
      "mean_token_accuracy": 0.15993930548429489,
      "num_tokens": 3617467.0,
      "step": 1965
    },
    {
      "entropy": 5.7752259254455565,
      "epoch": 0.16455061810892083,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999260346613491,
      "loss": 5.6732,
      "mean_token_accuracy": 0.15142679512500762,
      "num_tokens": 3627048.0,
      "step": 1970
    },
    {
      "entropy": 5.651623249053955,
      "epoch": 0.1649682592716338,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999252694181867,
      "loss": 5.6978,
      "mean_token_accuracy": 0.1521272361278534,
      "num_tokens": 3636683.0,
      "step": 1975
    },
    {
      "entropy": 5.696958827972412,
      "epoch": 0.1653859004343468,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999245002374367,
      "loss": 5.6235,
      "mean_token_accuracy": 0.16368645280599595,
      "num_tokens": 3645231.0,
      "step": 1980
    },
    {
      "entropy": 5.720323944091797,
      "epoch": 0.16580354159705982,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999237271191125,
      "loss": 5.6822,
      "mean_token_accuracy": 0.15285556465387345,
      "num_tokens": 3653851.0,
      "step": 1985
    },
    {
      "entropy": 5.692817211151123,
      "epoch": 0.1662211827597728,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999229500632279,
      "loss": 5.6642,
      "mean_token_accuracy": 0.1602617844939232,
      "num_tokens": 3662720.0,
      "step": 1990
    },
    {
      "entropy": 5.616102457046509,
      "epoch": 0.1666388239224858,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004999221690697961,
      "loss": 5.6011,
      "mean_token_accuracy": 0.15996311008930206,
      "num_tokens": 3671777.0,
      "step": 1995
    },
    {
      "entropy": 5.6898268699646,
      "epoch": 0.1670564650851988,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999213841388312,
      "loss": 5.7134,
      "mean_token_accuracy": 0.16112901419401168,
      "num_tokens": 3681390.0,
      "step": 2000
    },
    {
      "entropy": 5.791441488265991,
      "epoch": 0.16747410624791179,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999205952703466,
      "loss": 5.7535,
      "mean_token_accuracy": 0.15657739117741584,
      "num_tokens": 3690899.0,
      "step": 2005
    },
    {
      "entropy": 5.80651216506958,
      "epoch": 0.1678917474106248,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999198024643563,
      "loss": 5.8517,
      "mean_token_accuracy": 0.14452483803033828,
      "num_tokens": 3699891.0,
      "step": 2010
    },
    {
      "entropy": 5.6384241580963135,
      "epoch": 0.1683093885733378,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999190057208741,
      "loss": 5.6471,
      "mean_token_accuracy": 0.1593468502163887,
      "num_tokens": 3709860.0,
      "step": 2015
    },
    {
      "entropy": 5.729552555084228,
      "epoch": 0.16872702973605078,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499918205039914,
      "loss": 5.766,
      "mean_token_accuracy": 0.15205028057098388,
      "num_tokens": 3718896.0,
      "step": 2020
    },
    {
      "entropy": 5.710118341445923,
      "epoch": 0.16914467089876378,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999174004214899,
      "loss": 5.5796,
      "mean_token_accuracy": 0.1561296820640564,
      "num_tokens": 3727355.0,
      "step": 2025
    },
    {
      "entropy": 5.6436944007873535,
      "epoch": 0.1695623120614768,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999165918656161,
      "loss": 5.6822,
      "mean_token_accuracy": 0.15167264267802238,
      "num_tokens": 3736191.0,
      "step": 2030
    },
    {
      "entropy": 5.8228062152862545,
      "epoch": 0.16997995322418977,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999157793723065,
      "loss": 5.7164,
      "mean_token_accuracy": 0.1563839003443718,
      "num_tokens": 3744832.0,
      "step": 2035
    },
    {
      "entropy": 5.607315397262573,
      "epoch": 0.17039759438690277,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999149629415755,
      "loss": 5.6646,
      "mean_token_accuracy": 0.15334342569112777,
      "num_tokens": 3753835.0,
      "step": 2040
    },
    {
      "entropy": 5.858393764495849,
      "epoch": 0.17081523554961578,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999141425734374,
      "loss": 5.7431,
      "mean_token_accuracy": 0.15258873701095582,
      "num_tokens": 3762816.0,
      "step": 2045
    },
    {
      "entropy": 5.651172780990601,
      "epoch": 0.17123287671232876,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999133182679064,
      "loss": 5.805,
      "mean_token_accuracy": 0.1495609924197197,
      "num_tokens": 3773244.0,
      "step": 2050
    },
    {
      "entropy": 5.733507108688355,
      "epoch": 0.17165051787504176,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999124900249971,
      "loss": 5.6603,
      "mean_token_accuracy": 0.16301829516887664,
      "num_tokens": 3782122.0,
      "step": 2055
    },
    {
      "entropy": 5.688563394546509,
      "epoch": 0.17206815903775477,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999116578447239,
      "loss": 5.6533,
      "mean_token_accuracy": 0.16714497208595275,
      "num_tokens": 3791303.0,
      "step": 2060
    },
    {
      "entropy": 5.7102601528167725,
      "epoch": 0.17248580020046775,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999108217271015,
      "loss": 5.7161,
      "mean_token_accuracy": 0.15428307950496672,
      "num_tokens": 3800991.0,
      "step": 2065
    },
    {
      "entropy": 5.698825216293335,
      "epoch": 0.17290344136318075,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999099816721444,
      "loss": 5.6302,
      "mean_token_accuracy": 0.15540433079004287,
      "num_tokens": 3811293.0,
      "step": 2070
    },
    {
      "entropy": 5.766291332244873,
      "epoch": 0.17332108252589376,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999091376798673,
      "loss": 5.7416,
      "mean_token_accuracy": 0.15168866515159607,
      "num_tokens": 3820013.0,
      "step": 2075
    },
    {
      "entropy": 5.655336523056031,
      "epoch": 0.17373872368860674,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999082897502849,
      "loss": 5.6204,
      "mean_token_accuracy": 0.1611282467842102,
      "num_tokens": 3829099.0,
      "step": 2080
    },
    {
      "entropy": 5.651612615585327,
      "epoch": 0.17415636485131974,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999074378834123,
      "loss": 5.7326,
      "mean_token_accuracy": 0.15682385265827178,
      "num_tokens": 3838308.0,
      "step": 2085
    },
    {
      "entropy": 5.738585424423218,
      "epoch": 0.17457400601403275,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999065820792644,
      "loss": 5.6786,
      "mean_token_accuracy": 0.15595242530107498,
      "num_tokens": 3847650.0,
      "step": 2090
    },
    {
      "entropy": 5.805827569961548,
      "epoch": 0.17499164717674573,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999057223378559,
      "loss": 5.8611,
      "mean_token_accuracy": 0.14233408570289613,
      "num_tokens": 3857252.0,
      "step": 2095
    },
    {
      "entropy": 5.754422426223755,
      "epoch": 0.17540928833945874,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499904858659202,
      "loss": 5.6955,
      "mean_token_accuracy": 0.15864840149879456,
      "num_tokens": 3865747.0,
      "step": 2100
    },
    {
      "entropy": 5.674372434616089,
      "epoch": 0.17582692950217174,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999039910433179,
      "loss": 5.6711,
      "mean_token_accuracy": 0.16594497859477997,
      "num_tokens": 3874958.0,
      "step": 2105
    },
    {
      "entropy": 5.588525819778442,
      "epoch": 0.17624457066488472,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999031194902187,
      "loss": 5.6174,
      "mean_token_accuracy": 0.16458353102207185,
      "num_tokens": 3885216.0,
      "step": 2110
    },
    {
      "entropy": 5.748322677612305,
      "epoch": 0.17666221182759773,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999022439999198,
      "loss": 5.6728,
      "mean_token_accuracy": 0.16610947996377945,
      "num_tokens": 3893973.0,
      "step": 2115
    },
    {
      "entropy": 5.731654167175293,
      "epoch": 0.17707985299031073,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999013645724362,
      "loss": 5.7414,
      "mean_token_accuracy": 0.15839334726333618,
      "num_tokens": 3903532.0,
      "step": 2120
    },
    {
      "entropy": 5.741130352020264,
      "epoch": 0.1774974941530237,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999004812077836,
      "loss": 5.655,
      "mean_token_accuracy": 0.16057559847831726,
      "num_tokens": 3913427.0,
      "step": 2125
    },
    {
      "entropy": 5.7038579940795895,
      "epoch": 0.17791513531573672,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998995939059773,
      "loss": 5.8097,
      "mean_token_accuracy": 0.14798093736171722,
      "num_tokens": 3923870.0,
      "step": 2130
    },
    {
      "entropy": 5.693835687637329,
      "epoch": 0.17833277647844972,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499898702667033,
      "loss": 5.6247,
      "mean_token_accuracy": 0.16271054297685622,
      "num_tokens": 3933469.0,
      "step": 2135
    },
    {
      "entropy": 5.6254839420318605,
      "epoch": 0.1787504176411627,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004998978074909661,
      "loss": 5.6403,
      "mean_token_accuracy": 0.16687489449977874,
      "num_tokens": 3942427.0,
      "step": 2140
    },
    {
      "entropy": 5.688601303100586,
      "epoch": 0.1791680588038757,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998969083777925,
      "loss": 5.673,
      "mean_token_accuracy": 0.16252454817295076,
      "num_tokens": 3951250.0,
      "step": 2145
    },
    {
      "entropy": 5.682954549789429,
      "epoch": 0.1795856999665887,
      "grad_norm": 1.125,
      "learning_rate": 0.0004998960053275277,
      "loss": 5.6662,
      "mean_token_accuracy": 0.15424496978521346,
      "num_tokens": 3959969.0,
      "step": 2150
    },
    {
      "entropy": 5.671282815933227,
      "epoch": 0.1800033411293017,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004998950983401875,
      "loss": 5.703,
      "mean_token_accuracy": 0.1546568527817726,
      "num_tokens": 3968171.0,
      "step": 2155
    },
    {
      "entropy": 5.702560186386108,
      "epoch": 0.1804209822920147,
      "grad_norm": 1.125,
      "learning_rate": 0.0004998941874157882,
      "loss": 5.6091,
      "mean_token_accuracy": 0.16684609055519103,
      "num_tokens": 3976686.0,
      "step": 2160
    },
    {
      "entropy": 5.5957166194915775,
      "epoch": 0.1808386234547277,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998932725543453,
      "loss": 5.7366,
      "mean_token_accuracy": 0.1542557030916214,
      "num_tokens": 3986552.0,
      "step": 2165
    },
    {
      "entropy": 5.746101808547974,
      "epoch": 0.18125626461744068,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004998923537558749,
      "loss": 5.5971,
      "mean_token_accuracy": 0.15404898673295975,
      "num_tokens": 3995501.0,
      "step": 2170
    },
    {
      "entropy": 5.577417039871216,
      "epoch": 0.1816739057801537,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998914310203932,
      "loss": 5.585,
      "mean_token_accuracy": 0.16384028941392897,
      "num_tokens": 4004423.0,
      "step": 2175
    },
    {
      "entropy": 5.705241107940674,
      "epoch": 0.1820915469428667,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004998905043479163,
      "loss": 5.7633,
      "mean_token_accuracy": 0.15359411537647247,
      "num_tokens": 4013324.0,
      "step": 2180
    },
    {
      "entropy": 5.720277500152588,
      "epoch": 0.18250918810557967,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998895737384605,
      "loss": 5.6616,
      "mean_token_accuracy": 0.15625456869602203,
      "num_tokens": 4021896.0,
      "step": 2185
    },
    {
      "entropy": 5.585773658752442,
      "epoch": 0.18292682926829268,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998886391920418,
      "loss": 5.639,
      "mean_token_accuracy": 0.15853976160287858,
      "num_tokens": 4030590.0,
      "step": 2190
    },
    {
      "entropy": 5.639527130126953,
      "epoch": 0.18334447043100568,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004998877007086769,
      "loss": 5.536,
      "mean_token_accuracy": 0.16841819733381272,
      "num_tokens": 4040041.0,
      "step": 2195
    },
    {
      "entropy": 5.724634075164795,
      "epoch": 0.18376211159371866,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998867582883821,
      "loss": 5.7155,
      "mean_token_accuracy": 0.14944272637367248,
      "num_tokens": 4049119.0,
      "step": 2200
    },
    {
      "entropy": 5.675548553466797,
      "epoch": 0.18417975275643167,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998858119311739,
      "loss": 5.6431,
      "mean_token_accuracy": 0.1540763944387436,
      "num_tokens": 4059265.0,
      "step": 2205
    },
    {
      "entropy": 5.692709350585938,
      "epoch": 0.18459739391914468,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998848616370689,
      "loss": 5.69,
      "mean_token_accuracy": 0.15441271513700486,
      "num_tokens": 4068363.0,
      "step": 2210
    },
    {
      "entropy": 5.612869834899902,
      "epoch": 0.18501503508185768,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998839074060837,
      "loss": 5.575,
      "mean_token_accuracy": 0.16939774602651597,
      "num_tokens": 4077390.0,
      "step": 2215
    },
    {
      "entropy": 5.602217435836792,
      "epoch": 0.18543267624457066,
      "grad_norm": 1.09375,
      "learning_rate": 0.000499882949238235,
      "loss": 5.6605,
      "mean_token_accuracy": 0.15488070547580718,
      "num_tokens": 4087331.0,
      "step": 2220
    },
    {
      "entropy": 5.898349380493164,
      "epoch": 0.18585031740728367,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998819871335396,
      "loss": 5.8585,
      "mean_token_accuracy": 0.1453448213636875,
      "num_tokens": 4096425.0,
      "step": 2225
    },
    {
      "entropy": 5.599613237380981,
      "epoch": 0.18626795856999667,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004998810210920142,
      "loss": 5.6084,
      "mean_token_accuracy": 0.16011494100093843,
      "num_tokens": 4106056.0,
      "step": 2230
    },
    {
      "entropy": 5.683730888366699,
      "epoch": 0.18668559973270965,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998800511136759,
      "loss": 5.6968,
      "mean_token_accuracy": 0.1584601417183876,
      "num_tokens": 4115133.0,
      "step": 2235
    },
    {
      "entropy": 5.667567729949951,
      "epoch": 0.18710324089542266,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998790771985417,
      "loss": 5.5964,
      "mean_token_accuracy": 0.1586738795042038,
      "num_tokens": 4123630.0,
      "step": 2240
    },
    {
      "entropy": 5.640010595321655,
      "epoch": 0.18752088205813566,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998780993466284,
      "loss": 5.7469,
      "mean_token_accuracy": 0.15695928037166595,
      "num_tokens": 4133337.0,
      "step": 2245
    },
    {
      "entropy": 5.728957223892212,
      "epoch": 0.18793852322084864,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998771175579534,
      "loss": 5.6701,
      "mean_token_accuracy": 0.15720749497413636,
      "num_tokens": 4142240.0,
      "step": 2250
    },
    {
      "entropy": 5.650001859664917,
      "epoch": 0.18835616438356165,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998761318325337,
      "loss": 5.6434,
      "mean_token_accuracy": 0.15715699940919875,
      "num_tokens": 4151297.0,
      "step": 2255
    },
    {
      "entropy": 5.6392778873443605,
      "epoch": 0.18877380554627465,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998751421703868,
      "loss": 5.6019,
      "mean_token_accuracy": 0.16495950818061828,
      "num_tokens": 4160193.0,
      "step": 2260
    },
    {
      "entropy": 5.726975584030152,
      "epoch": 0.18919144670898763,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998741485715297,
      "loss": 5.7615,
      "mean_token_accuracy": 0.15852496922016143,
      "num_tokens": 4169948.0,
      "step": 2265
    },
    {
      "entropy": 5.873078060150147,
      "epoch": 0.18960908787170064,
      "grad_norm": 1.015625,
      "learning_rate": 0.00049987315103598,
      "loss": 5.78,
      "mean_token_accuracy": 0.1499393455684185,
      "num_tokens": 4179981.0,
      "step": 2270
    },
    {
      "entropy": 5.655186462402344,
      "epoch": 0.19002672903441364,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004998721495637551,
      "loss": 5.8004,
      "mean_token_accuracy": 0.15085280686616898,
      "num_tokens": 4189172.0,
      "step": 2275
    },
    {
      "entropy": 5.6853272914886475,
      "epoch": 0.19044437019712662,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998711441548726,
      "loss": 5.5864,
      "mean_token_accuracy": 0.15680087357759476,
      "num_tokens": 4197312.0,
      "step": 2280
    },
    {
      "entropy": 5.708002233505249,
      "epoch": 0.19086201135983963,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004998701348093499,
      "loss": 5.6636,
      "mean_token_accuracy": 0.15665152966976165,
      "num_tokens": 4206828.0,
      "step": 2285
    },
    {
      "entropy": 5.730845022201538,
      "epoch": 0.19127965252255263,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000499869121527205,
      "loss": 5.6132,
      "mean_token_accuracy": 0.16018779575824738,
      "num_tokens": 4214835.0,
      "step": 2290
    },
    {
      "entropy": 5.6554999351501465,
      "epoch": 0.1916972936852656,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998681043084553,
      "loss": 5.7972,
      "mean_token_accuracy": 0.147630412876606,
      "num_tokens": 4225119.0,
      "step": 2295
    },
    {
      "entropy": 5.660569190979004,
      "epoch": 0.19211493484797862,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998670831531188,
      "loss": 5.6428,
      "mean_token_accuracy": 0.15618418008089066,
      "num_tokens": 4235257.0,
      "step": 2300
    },
    {
      "entropy": 5.665428733825683,
      "epoch": 0.19253257601069163,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998660580612135,
      "loss": 5.7077,
      "mean_token_accuracy": 0.16061737835407258,
      "num_tokens": 4244211.0,
      "step": 2305
    },
    {
      "entropy": 5.724985074996948,
      "epoch": 0.1929502171734046,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998650290327569,
      "loss": 5.7158,
      "mean_token_accuracy": 0.15692965239286422,
      "num_tokens": 4253149.0,
      "step": 2310
    },
    {
      "entropy": 5.6536462783813475,
      "epoch": 0.1933678583361176,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998639960677675,
      "loss": 5.6279,
      "mean_token_accuracy": 0.15889257192611694,
      "num_tokens": 4262557.0,
      "step": 2315
    },
    {
      "entropy": 5.657335042953491,
      "epoch": 0.19378549949883062,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998629591662631,
      "loss": 5.6482,
      "mean_token_accuracy": 0.16159564554691314,
      "num_tokens": 4272881.0,
      "step": 2320
    },
    {
      "entropy": 5.599450969696045,
      "epoch": 0.1942031406615436,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000499861918328262,
      "loss": 5.6326,
      "mean_token_accuracy": 0.16337416023015977,
      "num_tokens": 4281158.0,
      "step": 2325
    },
    {
      "entropy": 5.6304497718811035,
      "epoch": 0.1946207818242566,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998608735537822,
      "loss": 5.6341,
      "mean_token_accuracy": 0.15982079207897187,
      "num_tokens": 4290319.0,
      "step": 2330
    },
    {
      "entropy": 5.665145397186279,
      "epoch": 0.1950384229869696,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998598248428423,
      "loss": 5.7185,
      "mean_token_accuracy": 0.16070850789546967,
      "num_tokens": 4299276.0,
      "step": 2335
    },
    {
      "entropy": 5.7335333824157715,
      "epoch": 0.19545606414968258,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998587721954604,
      "loss": 5.687,
      "mean_token_accuracy": 0.1566752687096596,
      "num_tokens": 4308291.0,
      "step": 2340
    },
    {
      "entropy": 5.622469186782837,
      "epoch": 0.1958737053123956,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004998577156116551,
      "loss": 5.6283,
      "mean_token_accuracy": 0.155308535695076,
      "num_tokens": 4316654.0,
      "step": 2345
    },
    {
      "entropy": 5.631796360015869,
      "epoch": 0.1962913464751086,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998566550914448,
      "loss": 5.5751,
      "mean_token_accuracy": 0.15566373467445374,
      "num_tokens": 4325911.0,
      "step": 2350
    },
    {
      "entropy": 5.719413614273071,
      "epoch": 0.19670898763782158,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998555906348481,
      "loss": 5.7631,
      "mean_token_accuracy": 0.15533148795366286,
      "num_tokens": 4334846.0,
      "step": 2355
    },
    {
      "entropy": 5.713936042785645,
      "epoch": 0.19712662880053458,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004998545222418836,
      "loss": 5.6411,
      "mean_token_accuracy": 0.16010912209749223,
      "num_tokens": 4343728.0,
      "step": 2360
    },
    {
      "entropy": 5.609404993057251,
      "epoch": 0.1975442699632476,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998534499125701,
      "loss": 5.612,
      "mean_token_accuracy": 0.1628926530480385,
      "num_tokens": 4352774.0,
      "step": 2365
    },
    {
      "entropy": 5.750800704956054,
      "epoch": 0.19796191112596057,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004998523736469263,
      "loss": 5.746,
      "mean_token_accuracy": 0.15442651212215425,
      "num_tokens": 4361663.0,
      "step": 2370
    },
    {
      "entropy": 5.65933666229248,
      "epoch": 0.19837955228867357,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000499851293444971,
      "loss": 5.6002,
      "mean_token_accuracy": 0.1717977210879326,
      "num_tokens": 4370278.0,
      "step": 2375
    },
    {
      "entropy": 5.68487024307251,
      "epoch": 0.19879719345138658,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998502093067231,
      "loss": 5.6569,
      "mean_token_accuracy": 0.16009533107280732,
      "num_tokens": 4379010.0,
      "step": 2380
    },
    {
      "entropy": 5.69346399307251,
      "epoch": 0.19921483461409956,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998491212322017,
      "loss": 5.6422,
      "mean_token_accuracy": 0.1543926328420639,
      "num_tokens": 4388145.0,
      "step": 2385
    },
    {
      "entropy": 5.6455628871917725,
      "epoch": 0.19963247577681256,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004998480292214258,
      "loss": 5.7707,
      "mean_token_accuracy": 0.1530858799815178,
      "num_tokens": 4396321.0,
      "step": 2390
    },
    {
      "entropy": 5.702960968017578,
      "epoch": 0.20005011693952557,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004998469332744144,
      "loss": 5.5412,
      "mean_token_accuracy": 0.16541909873485566,
      "num_tokens": 4404854.0,
      "step": 2395
    },
    {
      "entropy": 5.469099617004394,
      "epoch": 0.20046775810223855,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004998458333911869,
      "loss": 5.5451,
      "mean_token_accuracy": 0.16603398621082305,
      "num_tokens": 4413890.0,
      "step": 2400
    },
    {
      "entropy": 5.7079826354980465,
      "epoch": 0.20088539926495155,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004998447295717624,
      "loss": 5.6872,
      "mean_token_accuracy": 0.15662553459405898,
      "num_tokens": 4423019.0,
      "step": 2405
    },
    {
      "entropy": 5.715851497650147,
      "epoch": 0.20130304042766456,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004998436218161602,
      "loss": 5.6255,
      "mean_token_accuracy": 0.15582710802555083,
      "num_tokens": 4432340.0,
      "step": 2410
    },
    {
      "entropy": 5.611751461029053,
      "epoch": 0.20172068159037754,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998425101243999,
      "loss": 5.6121,
      "mean_token_accuracy": 0.1572697326540947,
      "num_tokens": 4441834.0,
      "step": 2415
    },
    {
      "entropy": 5.6097025871276855,
      "epoch": 0.20213832275309054,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998413944965007,
      "loss": 5.638,
      "mean_token_accuracy": 0.16356834024190903,
      "num_tokens": 4450864.0,
      "step": 2420
    },
    {
      "entropy": 5.6114274024963375,
      "epoch": 0.20255596391580355,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998402749324823,
      "loss": 5.7129,
      "mean_token_accuracy": 0.15567796379327775,
      "num_tokens": 4460697.0,
      "step": 2425
    },
    {
      "entropy": 5.893751335144043,
      "epoch": 0.20297360507851653,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998391514323642,
      "loss": 5.7085,
      "mean_token_accuracy": 0.153797747194767,
      "num_tokens": 4469926.0,
      "step": 2430
    },
    {
      "entropy": 5.583089590072632,
      "epoch": 0.20339124624122953,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004998380239961661,
      "loss": 5.6422,
      "mean_token_accuracy": 0.15996050238609313,
      "num_tokens": 4478818.0,
      "step": 2435
    },
    {
      "entropy": 5.703447008132935,
      "epoch": 0.20380888740394254,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998368926239078,
      "loss": 5.5741,
      "mean_token_accuracy": 0.15723517835140227,
      "num_tokens": 4487451.0,
      "step": 2440
    },
    {
      "entropy": 5.678385782241821,
      "epoch": 0.20422652856665552,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499835757315609,
      "loss": 5.6585,
      "mean_token_accuracy": 0.15110710263252258,
      "num_tokens": 4497674.0,
      "step": 2445
    },
    {
      "entropy": 5.581616687774658,
      "epoch": 0.20464416972936852,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004998346180712898,
      "loss": 5.6004,
      "mean_token_accuracy": 0.16663052439689635,
      "num_tokens": 4507666.0,
      "step": 2450
    },
    {
      "entropy": 5.625546884536743,
      "epoch": 0.20506181089208153,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004998334748909698,
      "loss": 5.7454,
      "mean_token_accuracy": 0.15425530821084976,
      "num_tokens": 4517409.0,
      "step": 2455
    },
    {
      "entropy": 5.657826280593872,
      "epoch": 0.2054794520547945,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998323277746694,
      "loss": 5.5006,
      "mean_token_accuracy": 0.16149946302175522,
      "num_tokens": 4526495.0,
      "step": 2460
    },
    {
      "entropy": 5.665003061294556,
      "epoch": 0.20589709321750752,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998311767224081,
      "loss": 5.6024,
      "mean_token_accuracy": 0.16810937523841857,
      "num_tokens": 4535240.0,
      "step": 2465
    },
    {
      "entropy": 5.631776237487793,
      "epoch": 0.20631473438022052,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998300217342067,
      "loss": 5.7025,
      "mean_token_accuracy": 0.156618569791317,
      "num_tokens": 4544716.0,
      "step": 2470
    },
    {
      "entropy": 5.6214515209198,
      "epoch": 0.2067323755429335,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998288628100851,
      "loss": 5.6206,
      "mean_token_accuracy": 0.1633414626121521,
      "num_tokens": 4553867.0,
      "step": 2475
    },
    {
      "entropy": 5.647698307037354,
      "epoch": 0.2071500167056465,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998276999500636,
      "loss": 5.6668,
      "mean_token_accuracy": 0.1532789349555969,
      "num_tokens": 4562529.0,
      "step": 2480
    },
    {
      "entropy": 5.562749671936035,
      "epoch": 0.2075676578683595,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998265331541627,
      "loss": 5.4629,
      "mean_token_accuracy": 0.1752343535423279,
      "num_tokens": 4571041.0,
      "step": 2485
    },
    {
      "entropy": 5.565519905090332,
      "epoch": 0.2079852990310725,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998253624224026,
      "loss": 5.6044,
      "mean_token_accuracy": 0.161303348839283,
      "num_tokens": 4579914.0,
      "step": 2490
    },
    {
      "entropy": 5.644900465011597,
      "epoch": 0.2084029401937855,
      "grad_norm": 1.25,
      "learning_rate": 0.000499824187754804,
      "loss": 5.6815,
      "mean_token_accuracy": 0.16097000390291213,
      "num_tokens": 4590277.0,
      "step": 2495
    },
    {
      "entropy": 5.66708197593689,
      "epoch": 0.2088205813564985,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004998230091513872,
      "loss": 5.6005,
      "mean_token_accuracy": 0.16419181376695632,
      "num_tokens": 4598719.0,
      "step": 2500
    },
    {
      "entropy": 5.624560022354126,
      "epoch": 0.20923822251921148,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998218266121731,
      "loss": 5.5608,
      "mean_token_accuracy": 0.1620894342660904,
      "num_tokens": 4607804.0,
      "step": 2505
    },
    {
      "entropy": 5.510728168487549,
      "epoch": 0.2096558636819245,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998206401371823,
      "loss": 5.5469,
      "mean_token_accuracy": 0.16698947846889495,
      "num_tokens": 4616529.0,
      "step": 2510
    },
    {
      "entropy": 5.637687015533447,
      "epoch": 0.2100735048446375,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004998194497264357,
      "loss": 5.603,
      "mean_token_accuracy": 0.15697978138923646,
      "num_tokens": 4624545.0,
      "step": 2515
    },
    {
      "entropy": 5.640446758270263,
      "epoch": 0.21049114600735047,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998182553799538,
      "loss": 5.5814,
      "mean_token_accuracy": 0.1600236177444458,
      "num_tokens": 4633316.0,
      "step": 2520
    },
    {
      "entropy": 5.585673189163208,
      "epoch": 0.21090878717006348,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004998170570977579,
      "loss": 5.5394,
      "mean_token_accuracy": 0.16961861550807952,
      "num_tokens": 4643353.0,
      "step": 2525
    },
    {
      "entropy": 5.584037351608276,
      "epoch": 0.21132642833277648,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998158548798687,
      "loss": 5.632,
      "mean_token_accuracy": 0.16008919924497605,
      "num_tokens": 4651998.0,
      "step": 2530
    },
    {
      "entropy": 5.621504116058349,
      "epoch": 0.21174406949548946,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998146487263074,
      "loss": 5.6235,
      "mean_token_accuracy": 0.16083284914493562,
      "num_tokens": 4659937.0,
      "step": 2535
    },
    {
      "entropy": 5.583824634552002,
      "epoch": 0.21216171065820247,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998134386370951,
      "loss": 5.5981,
      "mean_token_accuracy": 0.1597536824643612,
      "num_tokens": 4669396.0,
      "step": 2540
    },
    {
      "entropy": 5.717773294448852,
      "epoch": 0.21257935182091547,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998122246122529,
      "loss": 5.6704,
      "mean_token_accuracy": 0.16459225714206696,
      "num_tokens": 4679320.0,
      "step": 2545
    },
    {
      "entropy": 5.6832643985748295,
      "epoch": 0.21299699298362845,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998110066518021,
      "loss": 5.5986,
      "mean_token_accuracy": 0.16294638812541962,
      "num_tokens": 4688807.0,
      "step": 2550
    },
    {
      "entropy": 5.633354759216308,
      "epoch": 0.21341463414634146,
      "grad_norm": 1.109375,
      "learning_rate": 0.000499809784755764,
      "loss": 5.5983,
      "mean_token_accuracy": 0.17056376934051515,
      "num_tokens": 4697819.0,
      "step": 2555
    },
    {
      "entropy": 5.584991502761841,
      "epoch": 0.21383227530905446,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00049980855892416,
      "loss": 5.5931,
      "mean_token_accuracy": 0.16347702592611313,
      "num_tokens": 4706931.0,
      "step": 2560
    },
    {
      "entropy": 5.608094644546509,
      "epoch": 0.21424991647176747,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998073291570116,
      "loss": 5.5614,
      "mean_token_accuracy": 0.1649039328098297,
      "num_tokens": 4716262.0,
      "step": 2565
    },
    {
      "entropy": 5.627374076843262,
      "epoch": 0.21466755763448045,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998060954543404,
      "loss": 5.5921,
      "mean_token_accuracy": 0.16122105121612548,
      "num_tokens": 4725415.0,
      "step": 2570
    },
    {
      "entropy": 5.546875810623169,
      "epoch": 0.21508519879719346,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004998048578161678,
      "loss": 5.6491,
      "mean_token_accuracy": 0.16040805876255035,
      "num_tokens": 4734580.0,
      "step": 2575
    },
    {
      "entropy": 5.58602876663208,
      "epoch": 0.21550283995990646,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004998036162425155,
      "loss": 5.4665,
      "mean_token_accuracy": 0.17046085894107818,
      "num_tokens": 4742966.0,
      "step": 2580
    },
    {
      "entropy": 5.587724733352661,
      "epoch": 0.21592048112261944,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004998023707334055,
      "loss": 5.6199,
      "mean_token_accuracy": 0.1627469092607498,
      "num_tokens": 4751696.0,
      "step": 2585
    },
    {
      "entropy": 5.655860662460327,
      "epoch": 0.21633812228533245,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998011212888592,
      "loss": 5.6566,
      "mean_token_accuracy": 0.16377044171094896,
      "num_tokens": 4760460.0,
      "step": 2590
    },
    {
      "entropy": 5.668825340270996,
      "epoch": 0.21675576344804545,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997998679088988,
      "loss": 5.583,
      "mean_token_accuracy": 0.16344082206487656,
      "num_tokens": 4769004.0,
      "step": 2595
    },
    {
      "entropy": 5.5735485553741455,
      "epoch": 0.21717340461075843,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997986105935461,
      "loss": 5.5641,
      "mean_token_accuracy": 0.17153629660606384,
      "num_tokens": 4777204.0,
      "step": 2600
    },
    {
      "entropy": 5.688373565673828,
      "epoch": 0.21759104577347144,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997973493428231,
      "loss": 5.6904,
      "mean_token_accuracy": 0.1608117088675499,
      "num_tokens": 4786766.0,
      "step": 2605
    },
    {
      "entropy": 5.573904323577881,
      "epoch": 0.21800868693618444,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004997960841567519,
      "loss": 5.5689,
      "mean_token_accuracy": 0.16307419240474702,
      "num_tokens": 4796676.0,
      "step": 2610
    },
    {
      "entropy": 5.784175872802734,
      "epoch": 0.21842632809889742,
      "grad_norm": 1.125,
      "learning_rate": 0.0004997948150353548,
      "loss": 5.5972,
      "mean_token_accuracy": 0.16198533475399018,
      "num_tokens": 4805514.0,
      "step": 2615
    },
    {
      "entropy": 5.579783344268799,
      "epoch": 0.21884396926161043,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004997935419786537,
      "loss": 5.6752,
      "mean_token_accuracy": 0.15730978101491927,
      "num_tokens": 4815487.0,
      "step": 2620
    },
    {
      "entropy": 5.642004919052124,
      "epoch": 0.21926161042432343,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997922649866712,
      "loss": 5.6203,
      "mean_token_accuracy": 0.162903268635273,
      "num_tokens": 4825044.0,
      "step": 2625
    },
    {
      "entropy": 5.533287715911865,
      "epoch": 0.2196792515870364,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997909840594294,
      "loss": 5.4271,
      "mean_token_accuracy": 0.17537373900413514,
      "num_tokens": 4833875.0,
      "step": 2630
    },
    {
      "entropy": 5.520668888092041,
      "epoch": 0.22009689274974942,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997896991969509,
      "loss": 5.646,
      "mean_token_accuracy": 0.1582174703478813,
      "num_tokens": 4843317.0,
      "step": 2635
    },
    {
      "entropy": 5.6602888107299805,
      "epoch": 0.22051453391246242,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997884103992581,
      "loss": 5.6246,
      "mean_token_accuracy": 0.15556131601333617,
      "num_tokens": 4852514.0,
      "step": 2640
    },
    {
      "entropy": 5.6012287616729735,
      "epoch": 0.2209321750751754,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004997871176663736,
      "loss": 5.5869,
      "mean_token_accuracy": 0.1639481171965599,
      "num_tokens": 4861260.0,
      "step": 2645
    },
    {
      "entropy": 5.607472705841064,
      "epoch": 0.2213498162378884,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004997858209983201,
      "loss": 5.5261,
      "mean_token_accuracy": 0.1611647054553032,
      "num_tokens": 4869654.0,
      "step": 2650
    },
    {
      "entropy": 5.620398616790771,
      "epoch": 0.22176745740060141,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997845203951202,
      "loss": 5.5871,
      "mean_token_accuracy": 0.15774784982204437,
      "num_tokens": 4878311.0,
      "step": 2655
    },
    {
      "entropy": 5.723662614822388,
      "epoch": 0.2221850985633144,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004997832158567967,
      "loss": 5.6628,
      "mean_token_accuracy": 0.15765633732080458,
      "num_tokens": 4887251.0,
      "step": 2660
    },
    {
      "entropy": 5.582074880599976,
      "epoch": 0.2226027397260274,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997819073833724,
      "loss": 5.5383,
      "mean_token_accuracy": 0.16348181664943695,
      "num_tokens": 4896528.0,
      "step": 2665
    },
    {
      "entropy": 5.570661401748657,
      "epoch": 0.2230203808887404,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004997805949748702,
      "loss": 5.575,
      "mean_token_accuracy": 0.16215746700763703,
      "num_tokens": 4904962.0,
      "step": 2670
    },
    {
      "entropy": 5.605434513092041,
      "epoch": 0.22343802205145338,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997792786313132,
      "loss": 5.5917,
      "mean_token_accuracy": 0.15934820771217345,
      "num_tokens": 4914735.0,
      "step": 2675
    },
    {
      "entropy": 5.416878604888916,
      "epoch": 0.2238556632141664,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004997779583527243,
      "loss": 5.4542,
      "mean_token_accuracy": 0.15913522839546204,
      "num_tokens": 4924271.0,
      "step": 2680
    },
    {
      "entropy": 5.731436920166016,
      "epoch": 0.2242733043768794,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004997766341391268,
      "loss": 5.6861,
      "mean_token_accuracy": 0.15840027630329132,
      "num_tokens": 4933928.0,
      "step": 2685
    },
    {
      "entropy": 5.586864900588989,
      "epoch": 0.22469094553959237,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004997753059905437,
      "loss": 5.4023,
      "mean_token_accuracy": 0.1752420797944069,
      "num_tokens": 4942714.0,
      "step": 2690
    },
    {
      "entropy": 5.399338579177856,
      "epoch": 0.22510858670230538,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004997739739069981,
      "loss": 5.5313,
      "mean_token_accuracy": 0.164621265232563,
      "num_tokens": 4952297.0,
      "step": 2695
    },
    {
      "entropy": 5.644370079040527,
      "epoch": 0.22552622786501839,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004997726378885138,
      "loss": 5.4685,
      "mean_token_accuracy": 0.1743430182337761,
      "num_tokens": 4960947.0,
      "step": 2700
    },
    {
      "entropy": 5.593647527694702,
      "epoch": 0.22594386902773136,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004997712979351137,
      "loss": 5.6253,
      "mean_token_accuracy": 0.1603491149842739,
      "num_tokens": 4969705.0,
      "step": 2705
    },
    {
      "entropy": 5.6281603336334225,
      "epoch": 0.22636151019044437,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004997699540468216,
      "loss": 5.5992,
      "mean_token_accuracy": 0.16521542072296141,
      "num_tokens": 4979112.0,
      "step": 2710
    },
    {
      "entropy": 5.5321063041687015,
      "epoch": 0.22677915135315738,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004997686062236609,
      "loss": 5.5237,
      "mean_token_accuracy": 0.16602300256490707,
      "num_tokens": 4987894.0,
      "step": 2715
    },
    {
      "entropy": 5.633674669265747,
      "epoch": 0.22719679251587035,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997672544656553,
      "loss": 5.6604,
      "mean_token_accuracy": 0.1612725928425789,
      "num_tokens": 4997774.0,
      "step": 2720
    },
    {
      "entropy": 5.698086166381836,
      "epoch": 0.22761443367858336,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004997658987728283,
      "loss": 5.5971,
      "mean_token_accuracy": 0.1690062776207924,
      "num_tokens": 5006183.0,
      "step": 2725
    },
    {
      "entropy": 5.591877269744873,
      "epoch": 0.22803207484129637,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997645391452035,
      "loss": 5.5916,
      "mean_token_accuracy": 0.1632835417985916,
      "num_tokens": 5016909.0,
      "step": 2730
    },
    {
      "entropy": 5.610874843597412,
      "epoch": 0.22844971600400935,
      "grad_norm": 1.078125,
      "learning_rate": 0.000499763175582805,
      "loss": 5.6292,
      "mean_token_accuracy": 0.1621117740869522,
      "num_tokens": 5027326.0,
      "step": 2735
    },
    {
      "entropy": 5.632061624526978,
      "epoch": 0.22886735716672235,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997618080856566,
      "loss": 5.594,
      "mean_token_accuracy": 0.15901436507701874,
      "num_tokens": 5037043.0,
      "step": 2740
    },
    {
      "entropy": 5.541202545166016,
      "epoch": 0.22928499832943536,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997604366537823,
      "loss": 5.5703,
      "mean_token_accuracy": 0.16744935959577562,
      "num_tokens": 5045992.0,
      "step": 2745
    },
    {
      "entropy": 5.564315366744995,
      "epoch": 0.22970263949214834,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997590612872058,
      "loss": 5.5393,
      "mean_token_accuracy": 0.16942261159420013,
      "num_tokens": 5055568.0,
      "step": 2750
    },
    {
      "entropy": 5.63058443069458,
      "epoch": 0.23012028065486134,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997576819859515,
      "loss": 5.6298,
      "mean_token_accuracy": 0.16200867295265198,
      "num_tokens": 5065512.0,
      "step": 2755
    },
    {
      "entropy": 5.484599208831787,
      "epoch": 0.23053792181757435,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997562987500434,
      "loss": 5.4663,
      "mean_token_accuracy": 0.17104848176240922,
      "num_tokens": 5074776.0,
      "step": 2760
    },
    {
      "entropy": 5.583745384216309,
      "epoch": 0.23095556298028733,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997549115795058,
      "loss": 5.5334,
      "mean_token_accuracy": 0.16588549315929413,
      "num_tokens": 5084743.0,
      "step": 2765
    },
    {
      "entropy": 5.682448768615723,
      "epoch": 0.23137320414300033,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499753520474363,
      "loss": 5.6935,
      "mean_token_accuracy": 0.16034545600414277,
      "num_tokens": 5095387.0,
      "step": 2770
    },
    {
      "entropy": 5.587224912643433,
      "epoch": 0.23179084530571334,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997521254346391,
      "loss": 5.5973,
      "mean_token_accuracy": 0.16626937463879585,
      "num_tokens": 5104628.0,
      "step": 2775
    },
    {
      "entropy": 5.622297048568726,
      "epoch": 0.23220848646842632,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997507264603589,
      "loss": 5.5347,
      "mean_token_accuracy": 0.16671543717384338,
      "num_tokens": 5113169.0,
      "step": 2780
    },
    {
      "entropy": 5.549988460540772,
      "epoch": 0.23262612763113932,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004997493235515466,
      "loss": 5.6124,
      "mean_token_accuracy": 0.16756106466054915,
      "num_tokens": 5122294.0,
      "step": 2785
    },
    {
      "entropy": 5.531440925598145,
      "epoch": 0.23304376879385233,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997479167082269,
      "loss": 5.4388,
      "mean_token_accuracy": 0.17597562521696092,
      "num_tokens": 5131191.0,
      "step": 2790
    },
    {
      "entropy": 5.549725675582886,
      "epoch": 0.2334614099565653,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997465059304243,
      "loss": 5.6491,
      "mean_token_accuracy": 0.16252151429653167,
      "num_tokens": 5141974.0,
      "step": 2795
    },
    {
      "entropy": 5.6808897972106935,
      "epoch": 0.23387905111927831,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997450912181637,
      "loss": 5.5764,
      "mean_token_accuracy": 0.16537160128355027,
      "num_tokens": 5151412.0,
      "step": 2800
    },
    {
      "entropy": 5.547338819503784,
      "epoch": 0.23429669228199132,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997436725714696,
      "loss": 5.5065,
      "mean_token_accuracy": 0.1739228755235672,
      "num_tokens": 5160989.0,
      "step": 2805
    },
    {
      "entropy": 5.551234436035156,
      "epoch": 0.2347143334447043,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004997422499903671,
      "loss": 5.5758,
      "mean_token_accuracy": 0.16309013664722444,
      "num_tokens": 5170241.0,
      "step": 2810
    },
    {
      "entropy": 5.731640720367432,
      "epoch": 0.2351319746074173,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000499740823474881,
      "loss": 5.7214,
      "mean_token_accuracy": 0.1557489424943924,
      "num_tokens": 5180247.0,
      "step": 2815
    },
    {
      "entropy": 5.54374361038208,
      "epoch": 0.2355496157701303,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004997393930250363,
      "loss": 5.5933,
      "mean_token_accuracy": 0.16376214027404784,
      "num_tokens": 5189144.0,
      "step": 2820
    },
    {
      "entropy": 5.623880338668823,
      "epoch": 0.2359672569328433,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499737958640858,
      "loss": 5.6483,
      "mean_token_accuracy": 0.16290141642093658,
      "num_tokens": 5199410.0,
      "step": 2825
    },
    {
      "entropy": 5.6456780433654785,
      "epoch": 0.2363848980955563,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997365203223711,
      "loss": 5.5117,
      "mean_token_accuracy": 0.17000638246536254,
      "num_tokens": 5209032.0,
      "step": 2830
    },
    {
      "entropy": 5.440157270431518,
      "epoch": 0.2368025392582693,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000499735078069601,
      "loss": 5.4238,
      "mean_token_accuracy": 0.17118311822414398,
      "num_tokens": 5217787.0,
      "step": 2835
    },
    {
      "entropy": 5.555545091629028,
      "epoch": 0.23722018042098228,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004997336318825728,
      "loss": 5.5312,
      "mean_token_accuracy": 0.1703398510813713,
      "num_tokens": 5226962.0,
      "step": 2840
    },
    {
      "entropy": 5.52580771446228,
      "epoch": 0.23763782158369529,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997321817613119,
      "loss": 5.5067,
      "mean_token_accuracy": 0.16689039915800094,
      "num_tokens": 5236111.0,
      "step": 2845
    },
    {
      "entropy": 5.645787096023559,
      "epoch": 0.2380554627464083,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997307277058436,
      "loss": 5.6452,
      "mean_token_accuracy": 0.16265233159065245,
      "num_tokens": 5246628.0,
      "step": 2850
    },
    {
      "entropy": 5.622815704345703,
      "epoch": 0.23847310390912127,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004997292697161935,
      "loss": 5.5583,
      "mean_token_accuracy": 0.15912623330950737,
      "num_tokens": 5255543.0,
      "step": 2855
    },
    {
      "entropy": 5.467281913757324,
      "epoch": 0.23889074507183428,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000499727807792387,
      "loss": 5.4196,
      "mean_token_accuracy": 0.17852753698825835,
      "num_tokens": 5264559.0,
      "step": 2860
    },
    {
      "entropy": 5.447366380691529,
      "epoch": 0.23930838623454728,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997263419344495,
      "loss": 5.5053,
      "mean_token_accuracy": 0.16848173439502717,
      "num_tokens": 5274599.0,
      "step": 2865
    },
    {
      "entropy": 5.553721141815186,
      "epoch": 0.23972602739726026,
      "grad_norm": 1.125,
      "learning_rate": 0.0004997248721424071,
      "loss": 5.473,
      "mean_token_accuracy": 0.17135296761989594,
      "num_tokens": 5283495.0,
      "step": 2870
    },
    {
      "entropy": 5.591934013366699,
      "epoch": 0.24014366855997327,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004997233984162853,
      "loss": 5.5224,
      "mean_token_accuracy": 0.1711369827389717,
      "num_tokens": 5292064.0,
      "step": 2875
    },
    {
      "entropy": 5.527604627609253,
      "epoch": 0.24056130972268627,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997219207561099,
      "loss": 5.535,
      "mean_token_accuracy": 0.16105968952178956,
      "num_tokens": 5301696.0,
      "step": 2880
    },
    {
      "entropy": 5.675354480743408,
      "epoch": 0.24097895088539925,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997204391619068,
      "loss": 5.6537,
      "mean_token_accuracy": 0.15840594172477723,
      "num_tokens": 5311427.0,
      "step": 2885
    },
    {
      "entropy": 5.684924793243408,
      "epoch": 0.24139659204811226,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997189536337018,
      "loss": 5.5821,
      "mean_token_accuracy": 0.16174941658973693,
      "num_tokens": 5321406.0,
      "step": 2890
    },
    {
      "entropy": 5.551355266571045,
      "epoch": 0.24181423321082526,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004997174641715211,
      "loss": 5.6642,
      "mean_token_accuracy": 0.1644423484802246,
      "num_tokens": 5332257.0,
      "step": 2895
    },
    {
      "entropy": 5.504457902908325,
      "epoch": 0.24223187437353827,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997159707753906,
      "loss": 5.4042,
      "mean_token_accuracy": 0.1703709304332733,
      "num_tokens": 5341857.0,
      "step": 2900
    },
    {
      "entropy": 5.511201333999634,
      "epoch": 0.24264951553625125,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004997144734453367,
      "loss": 5.5505,
      "mean_token_accuracy": 0.16949305385351182,
      "num_tokens": 5351367.0,
      "step": 2905
    },
    {
      "entropy": 5.391738796234131,
      "epoch": 0.24306715669896425,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004997129721813853,
      "loss": 5.4013,
      "mean_token_accuracy": 0.17349923253059388,
      "num_tokens": 5360627.0,
      "step": 2910
    },
    {
      "entropy": 5.531809091567993,
      "epoch": 0.24348479786167726,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004997114669835629,
      "loss": 5.4851,
      "mean_token_accuracy": 0.17700317949056626,
      "num_tokens": 5369703.0,
      "step": 2915
    },
    {
      "entropy": 5.5716290950775145,
      "epoch": 0.24390243902439024,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004997099578518959,
      "loss": 5.5563,
      "mean_token_accuracy": 0.16404612362384796,
      "num_tokens": 5378432.0,
      "step": 2920
    },
    {
      "entropy": 5.438390588760376,
      "epoch": 0.24432008018710324,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004997084447864105,
      "loss": 5.3909,
      "mean_token_accuracy": 0.1708747535943985,
      "num_tokens": 5386489.0,
      "step": 2925
    },
    {
      "entropy": 5.610124492645264,
      "epoch": 0.24473772134981625,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997069277871333,
      "loss": 5.6231,
      "mean_token_accuracy": 0.16265049129724501,
      "num_tokens": 5395908.0,
      "step": 2930
    },
    {
      "entropy": 5.710176420211792,
      "epoch": 0.24515536251252923,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004997054068540908,
      "loss": 5.539,
      "mean_token_accuracy": 0.16824057847261428,
      "num_tokens": 5404822.0,
      "step": 2935
    },
    {
      "entropy": 5.442877244949341,
      "epoch": 0.24557300367524224,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004997038819873098,
      "loss": 5.4707,
      "mean_token_accuracy": 0.1711698964238167,
      "num_tokens": 5413980.0,
      "step": 2940
    },
    {
      "entropy": 5.5252721309661865,
      "epoch": 0.24599064483795524,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004997023531868167,
      "loss": 5.5389,
      "mean_token_accuracy": 0.16247909963130952,
      "num_tokens": 5423538.0,
      "step": 2945
    },
    {
      "entropy": 5.515257501602173,
      "epoch": 0.24640828600066822,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004997008204526386,
      "loss": 5.4648,
      "mean_token_accuracy": 0.16875504106283187,
      "num_tokens": 5431930.0,
      "step": 2950
    },
    {
      "entropy": 5.48965220451355,
      "epoch": 0.24682592716338123,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004996992837848021,
      "loss": 5.4508,
      "mean_token_accuracy": 0.1761811926960945,
      "num_tokens": 5440721.0,
      "step": 2955
    },
    {
      "entropy": 5.506892585754395,
      "epoch": 0.24724356832609423,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004996977431833342,
      "loss": 5.5365,
      "mean_token_accuracy": 0.1642148897051811,
      "num_tokens": 5449617.0,
      "step": 2960
    },
    {
      "entropy": 5.5657142162322994,
      "epoch": 0.2476612094888072,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004996961986482618,
      "loss": 5.559,
      "mean_token_accuracy": 0.16744059771299363,
      "num_tokens": 5458827.0,
      "step": 2965
    },
    {
      "entropy": 5.565783166885376,
      "epoch": 0.24807885065152022,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004996946501796119,
      "loss": 5.4315,
      "mean_token_accuracy": 0.17086558043956757,
      "num_tokens": 5467348.0,
      "step": 2970
    },
    {
      "entropy": 5.554220914840698,
      "epoch": 0.24849649181423322,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996930977774118,
      "loss": 5.5211,
      "mean_token_accuracy": 0.15979193449020385,
      "num_tokens": 5476645.0,
      "step": 2975
    },
    {
      "entropy": 5.590994024276734,
      "epoch": 0.2489141329769462,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004996915414416885,
      "loss": 5.5611,
      "mean_token_accuracy": 0.16303199231624604,
      "num_tokens": 5486120.0,
      "step": 2980
    },
    {
      "entropy": 5.6555492877960205,
      "epoch": 0.2493317741396592,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004996899811724693,
      "loss": 5.5378,
      "mean_token_accuracy": 0.1691664695739746,
      "num_tokens": 5495313.0,
      "step": 2985
    },
    {
      "entropy": 5.537061023712158,
      "epoch": 0.2497494153023722,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996884169697816,
      "loss": 5.5371,
      "mean_token_accuracy": 0.16722940653562546,
      "num_tokens": 5504587.0,
      "step": 2990
    },
    {
      "entropy": 5.468687248229981,
      "epoch": 0.2501670564650852,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996868488336526,
      "loss": 5.5464,
      "mean_token_accuracy": 0.16358605027198792,
      "num_tokens": 5514188.0,
      "step": 2995
    },
    {
      "entropy": 5.579990291595459,
      "epoch": 0.25058469762779817,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004996852767641099,
      "loss": 5.5903,
      "mean_token_accuracy": 0.16043460369110107,
      "num_tokens": 5524075.0,
      "step": 3000
    },
    {
      "epoch": 0.25058469762779817,
      "eval_entropy": 5.461522004887654,
      "eval_loss": 5.523029327392578,
      "eval_mean_token_accuracy": 0.17474948998569836,
      "eval_num_tokens": 5524075.0,
      "eval_runtime": 27.2289,
      "eval_samples_per_second": 1386.87,
      "eval_steps_per_second": 173.382,
      "step": 3000
    },
    {
      "entropy": 5.4971150875091555,
      "epoch": 0.2510023387905112,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996837007611811,
      "loss": 5.375,
      "mean_token_accuracy": 0.17106565684080124,
      "num_tokens": 5533323.0,
      "step": 3005
    },
    {
      "entropy": 5.4650226593017575,
      "epoch": 0.2514199799532242,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996821208248935,
      "loss": 5.3986,
      "mean_token_accuracy": 0.17289367467164993,
      "num_tokens": 5542288.0,
      "step": 3010
    },
    {
      "entropy": 5.613493633270264,
      "epoch": 0.25183762111593716,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004996805369552749,
      "loss": 5.5574,
      "mean_token_accuracy": 0.163363116979599,
      "num_tokens": 5550869.0,
      "step": 3015
    },
    {
      "entropy": 5.449234056472778,
      "epoch": 0.2522552622786502,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004996789491523532,
      "loss": 5.4819,
      "mean_token_accuracy": 0.17674408853054047,
      "num_tokens": 5560147.0,
      "step": 3020
    },
    {
      "entropy": 5.469293308258057,
      "epoch": 0.2526729034413632,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004996773574161559,
      "loss": 5.3695,
      "mean_token_accuracy": 0.17348119765520095,
      "num_tokens": 5568358.0,
      "step": 3025
    },
    {
      "entropy": 5.484114408493042,
      "epoch": 0.25309054460407615,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004996757617467112,
      "loss": 5.4993,
      "mean_token_accuracy": 0.1649644508957863,
      "num_tokens": 5577755.0,
      "step": 3030
    },
    {
      "entropy": 5.547186088562012,
      "epoch": 0.2535081857667892,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996741621440466,
      "loss": 5.5348,
      "mean_token_accuracy": 0.1583195373415947,
      "num_tokens": 5587515.0,
      "step": 3035
    },
    {
      "entropy": 5.662641334533691,
      "epoch": 0.25392582692950216,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004996725586081906,
      "loss": 5.6822,
      "mean_token_accuracy": 0.15580229461193085,
      "num_tokens": 5597860.0,
      "step": 3040
    },
    {
      "entropy": 5.5714374542236325,
      "epoch": 0.25434346809221514,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004996709511391709,
      "loss": 5.5374,
      "mean_token_accuracy": 0.1693472683429718,
      "num_tokens": 5607302.0,
      "step": 3045
    },
    {
      "entropy": 5.583249855041504,
      "epoch": 0.2547611092549282,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004996693397370158,
      "loss": 5.619,
      "mean_token_accuracy": 0.16309259384870528,
      "num_tokens": 5616345.0,
      "step": 3050
    },
    {
      "entropy": 5.533199739456177,
      "epoch": 0.25517875041764115,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004996677244017534,
      "loss": 5.4014,
      "mean_token_accuracy": 0.16947476267814637,
      "num_tokens": 5624745.0,
      "step": 3055
    },
    {
      "entropy": 5.5545299530029295,
      "epoch": 0.25559639158035413,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004996661051334121,
      "loss": 5.5197,
      "mean_token_accuracy": 0.15930449813604355,
      "num_tokens": 5634022.0,
      "step": 3060
    },
    {
      "entropy": 5.545503568649292,
      "epoch": 0.25601403274306717,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004996644819320201,
      "loss": 5.4643,
      "mean_token_accuracy": 0.16942096054553984,
      "num_tokens": 5643631.0,
      "step": 3065
    },
    {
      "entropy": 5.583652544021606,
      "epoch": 0.25643167390578014,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004996628547976061,
      "loss": 5.5764,
      "mean_token_accuracy": 0.16339723616838456,
      "num_tokens": 5653155.0,
      "step": 3070
    },
    {
      "entropy": 5.585376787185669,
      "epoch": 0.2568493150684932,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004996612237301983,
      "loss": 5.559,
      "mean_token_accuracy": 0.16842040419578552,
      "num_tokens": 5663091.0,
      "step": 3075
    },
    {
      "entropy": 5.438110208511352,
      "epoch": 0.25726695623120616,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996595887298254,
      "loss": 5.4673,
      "mean_token_accuracy": 0.1753868818283081,
      "num_tokens": 5671885.0,
      "step": 3080
    },
    {
      "entropy": 5.550532436370849,
      "epoch": 0.25768459739391913,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000499657949796516,
      "loss": 5.4727,
      "mean_token_accuracy": 0.1669400379061699,
      "num_tokens": 5680225.0,
      "step": 3085
    },
    {
      "entropy": 5.392565679550171,
      "epoch": 0.25810223855663217,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004996563069302986,
      "loss": 5.3457,
      "mean_token_accuracy": 0.17783385515213013,
      "num_tokens": 5689456.0,
      "step": 3090
    },
    {
      "entropy": 5.4940855503082275,
      "epoch": 0.25851987971934515,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004996546601312023,
      "loss": 5.4581,
      "mean_token_accuracy": 0.17149127870798112,
      "num_tokens": 5698127.0,
      "step": 3095
    },
    {
      "entropy": 5.550233840942383,
      "epoch": 0.2589375208820581,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004996530093992558,
      "loss": 5.5085,
      "mean_token_accuracy": 0.17063496112823487,
      "num_tokens": 5706825.0,
      "step": 3100
    },
    {
      "entropy": 5.553879499435425,
      "epoch": 0.25935516204477116,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004996513547344877,
      "loss": 5.4587,
      "mean_token_accuracy": 0.1669582650065422,
      "num_tokens": 5715677.0,
      "step": 3105
    },
    {
      "entropy": 5.416211986541748,
      "epoch": 0.25977280320748414,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996496961369274,
      "loss": 5.3803,
      "mean_token_accuracy": 0.1699922949075699,
      "num_tokens": 5725371.0,
      "step": 3110
    },
    {
      "entropy": 5.537308883666992,
      "epoch": 0.2601904443701971,
      "grad_norm": 1.125,
      "learning_rate": 0.0004996480336066037,
      "loss": 5.4918,
      "mean_token_accuracy": 0.17023330479860305,
      "num_tokens": 5734664.0,
      "step": 3115
    },
    {
      "entropy": 5.552717208862305,
      "epoch": 0.26060808553291015,
      "grad_norm": 1.125,
      "learning_rate": 0.0004996463671435458,
      "loss": 5.4886,
      "mean_token_accuracy": 0.1722761482000351,
      "num_tokens": 5743870.0,
      "step": 3120
    },
    {
      "entropy": 5.432164907455444,
      "epoch": 0.26102572669562313,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004996446967477829,
      "loss": 5.505,
      "mean_token_accuracy": 0.16911642849445344,
      "num_tokens": 5752855.0,
      "step": 3125
    },
    {
      "entropy": 5.552587652206421,
      "epoch": 0.2614433678583361,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004996430224193441,
      "loss": 5.5603,
      "mean_token_accuracy": 0.16505547910928725,
      "num_tokens": 5762435.0,
      "step": 3130
    },
    {
      "entropy": 5.599203205108642,
      "epoch": 0.26186100902104914,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004996413441582589,
      "loss": 5.5274,
      "mean_token_accuracy": 0.16358712464570999,
      "num_tokens": 5771463.0,
      "step": 3135
    },
    {
      "entropy": 5.583684873580933,
      "epoch": 0.2622786501837621,
      "grad_norm": 1.125,
      "learning_rate": 0.0004996396619645564,
      "loss": 5.5684,
      "mean_token_accuracy": 0.16209518164396286,
      "num_tokens": 5780530.0,
      "step": 3140
    },
    {
      "entropy": 5.507178544998169,
      "epoch": 0.2626962913464751,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004996379758382662,
      "loss": 5.41,
      "mean_token_accuracy": 0.17605946063995362,
      "num_tokens": 5789199.0,
      "step": 3145
    },
    {
      "entropy": 5.605119609832764,
      "epoch": 0.26311393250918813,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499636285779418,
      "loss": 5.6178,
      "mean_token_accuracy": 0.16322802305221557,
      "num_tokens": 5799170.0,
      "step": 3150
    },
    {
      "entropy": 5.559698247909546,
      "epoch": 0.2635315736719011,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004996345917880412,
      "loss": 5.5178,
      "mean_token_accuracy": 0.1654113933444023,
      "num_tokens": 5807662.0,
      "step": 3155
    },
    {
      "entropy": 5.459578895568848,
      "epoch": 0.2639492148346141,
      "grad_norm": 1.125,
      "learning_rate": 0.0004996328938641654,
      "loss": 5.433,
      "mean_token_accuracy": 0.17274057120084763,
      "num_tokens": 5815790.0,
      "step": 3160
    },
    {
      "entropy": 5.572559070587158,
      "epoch": 0.2643668559973271,
      "grad_norm": 1.125,
      "learning_rate": 0.0004996311920078204,
      "loss": 5.4425,
      "mean_token_accuracy": 0.1730218067765236,
      "num_tokens": 5824746.0,
      "step": 3165
    },
    {
      "entropy": 5.449925374984741,
      "epoch": 0.2647844971600401,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004996294862190361,
      "loss": 5.456,
      "mean_token_accuracy": 0.17265160679817199,
      "num_tokens": 5833678.0,
      "step": 3170
    },
    {
      "entropy": 5.483268070220947,
      "epoch": 0.2652021383227531,
      "grad_norm": 1.140625,
      "learning_rate": 0.000499627776497842,
      "loss": 5.4971,
      "mean_token_accuracy": 0.1684225857257843,
      "num_tokens": 5842218.0,
      "step": 3175
    },
    {
      "entropy": 5.695554304122925,
      "epoch": 0.2656197794854661,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004996260628442686,
      "loss": 5.681,
      "mean_token_accuracy": 0.15670068114995955,
      "num_tokens": 5853250.0,
      "step": 3180
    },
    {
      "entropy": 5.482328844070435,
      "epoch": 0.2660374206481791,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004996243452583454,
      "loss": 5.409,
      "mean_token_accuracy": 0.17305591702461243,
      "num_tokens": 5862661.0,
      "step": 3185
    },
    {
      "entropy": 5.569812154769897,
      "epoch": 0.26645506181089207,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996226237401028,
      "loss": 5.5889,
      "mean_token_accuracy": 0.16395057886838912,
      "num_tokens": 5872327.0,
      "step": 3190
    },
    {
      "entropy": 5.441321468353271,
      "epoch": 0.2668727029736051,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004996208982895706,
      "loss": 5.4181,
      "mean_token_accuracy": 0.17318300008773804,
      "num_tokens": 5881539.0,
      "step": 3195
    },
    {
      "entropy": 5.619724702835083,
      "epoch": 0.2672903441363181,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004996191689067794,
      "loss": 5.5345,
      "mean_token_accuracy": 0.16643642708659173,
      "num_tokens": 5891602.0,
      "step": 3200
    },
    {
      "entropy": 5.484193992614746,
      "epoch": 0.26770798529903106,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004996174355917591,
      "loss": 5.5155,
      "mean_token_accuracy": 0.1677042007446289,
      "num_tokens": 5899831.0,
      "step": 3205
    },
    {
      "entropy": 5.6453687191009525,
      "epoch": 0.2681256264617441,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996156983445404,
      "loss": 5.6128,
      "mean_token_accuracy": 0.15973187386989593,
      "num_tokens": 5909193.0,
      "step": 3210
    },
    {
      "entropy": 5.596530437469482,
      "epoch": 0.26854326762445707,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996139571651534,
      "loss": 5.5212,
      "mean_token_accuracy": 0.1631033092737198,
      "num_tokens": 5919353.0,
      "step": 3215
    },
    {
      "entropy": 5.430617761611939,
      "epoch": 0.26896090878717005,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004996122120536288,
      "loss": 5.4745,
      "mean_token_accuracy": 0.17367395162582397,
      "num_tokens": 5927853.0,
      "step": 3220
    },
    {
      "entropy": 5.4728278636932375,
      "epoch": 0.2693785499498831,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996104630099969,
      "loss": 5.5796,
      "mean_token_accuracy": 0.16607874929904937,
      "num_tokens": 5937943.0,
      "step": 3225
    },
    {
      "entropy": 5.565532922744751,
      "epoch": 0.26979619111259606,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004996087100342885,
      "loss": 5.4689,
      "mean_token_accuracy": 0.17290828227996827,
      "num_tokens": 5946213.0,
      "step": 3230
    },
    {
      "entropy": 5.501755237579346,
      "epoch": 0.27021383227530904,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004996069531265345,
      "loss": 5.3754,
      "mean_token_accuracy": 0.17148633301258087,
      "num_tokens": 5954997.0,
      "step": 3235
    },
    {
      "entropy": 5.414116048812867,
      "epoch": 0.2706314734380221,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004996051922867653,
      "loss": 5.4059,
      "mean_token_accuracy": 0.17986717075109482,
      "num_tokens": 5963947.0,
      "step": 3240
    },
    {
      "entropy": 5.482024765014648,
      "epoch": 0.27104911460073505,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004996034275150118,
      "loss": 5.4712,
      "mean_token_accuracy": 0.16966199427843093,
      "num_tokens": 5972077.0,
      "step": 3245
    },
    {
      "entropy": 5.4911558628082275,
      "epoch": 0.27146675576344803,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000499601658811305,
      "loss": 5.4668,
      "mean_token_accuracy": 0.16965609639883042,
      "num_tokens": 5982649.0,
      "step": 3250
    },
    {
      "entropy": 5.577063179016113,
      "epoch": 0.27188439692616106,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004995998861756758,
      "loss": 5.4771,
      "mean_token_accuracy": 0.16891886293888092,
      "num_tokens": 5991349.0,
      "step": 3255
    },
    {
      "entropy": 5.471480846405029,
      "epoch": 0.27230203808887404,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004995981096081552,
      "loss": 5.4678,
      "mean_token_accuracy": 0.1755162000656128,
      "num_tokens": 6000555.0,
      "step": 3260
    },
    {
      "entropy": 5.650067710876465,
      "epoch": 0.272719679251587,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004995963291087744,
      "loss": 5.5778,
      "mean_token_accuracy": 0.15589477270841598,
      "num_tokens": 6009554.0,
      "step": 3265
    },
    {
      "entropy": 5.569346714019775,
      "epoch": 0.27313732041430006,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004995945446775645,
      "loss": 5.6646,
      "mean_token_accuracy": 0.16038435250520705,
      "num_tokens": 6019330.0,
      "step": 3270
    },
    {
      "entropy": 5.587334251403808,
      "epoch": 0.27355496157701303,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004995927563145566,
      "loss": 5.6125,
      "mean_token_accuracy": 0.16468510627746583,
      "num_tokens": 6028527.0,
      "step": 3275
    },
    {
      "entropy": 5.563228559494019,
      "epoch": 0.273972602739726,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004995909640197825,
      "loss": 5.4647,
      "mean_token_accuracy": 0.17105194479227065,
      "num_tokens": 6038223.0,
      "step": 3280
    },
    {
      "entropy": 5.444104480743408,
      "epoch": 0.27439024390243905,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004995891677932729,
      "loss": 5.4045,
      "mean_token_accuracy": 0.17711416333913804,
      "num_tokens": 6047459.0,
      "step": 3285
    },
    {
      "entropy": 5.602990055084229,
      "epoch": 0.274807885065152,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004995873676350597,
      "loss": 5.5954,
      "mean_token_accuracy": 0.16400577127933502,
      "num_tokens": 6057322.0,
      "step": 3290
    },
    {
      "entropy": 5.64007306098938,
      "epoch": 0.275225526227865,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004995855635451743,
      "loss": 5.7239,
      "mean_token_accuracy": 0.15748985335230828,
      "num_tokens": 6067520.0,
      "step": 3295
    },
    {
      "entropy": 5.563010215759277,
      "epoch": 0.27564316739057804,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004995837555236484,
      "loss": 5.5978,
      "mean_token_accuracy": 0.16345635652542115,
      "num_tokens": 6076729.0,
      "step": 3300
    },
    {
      "entropy": 5.6180760860443115,
      "epoch": 0.276060808553291,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004995819435705133,
      "loss": 5.534,
      "mean_token_accuracy": 0.16132914572954177,
      "num_tokens": 6086529.0,
      "step": 3305
    },
    {
      "entropy": 5.601102066040039,
      "epoch": 0.276478449716004,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004995801276858011,
      "loss": 5.5461,
      "mean_token_accuracy": 0.17339718639850615,
      "num_tokens": 6095784.0,
      "step": 3310
    },
    {
      "entropy": 5.643276929855347,
      "epoch": 0.276896090878717,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004995783078695434,
      "loss": 5.5606,
      "mean_token_accuracy": 0.1667674422264099,
      "num_tokens": 6104626.0,
      "step": 3315
    },
    {
      "entropy": 5.467655038833618,
      "epoch": 0.27731373204143,
      "grad_norm": 1.125,
      "learning_rate": 0.000499576484121772,
      "loss": 5.4733,
      "mean_token_accuracy": 0.1719300702214241,
      "num_tokens": 6113758.0,
      "step": 3320
    },
    {
      "entropy": 5.589902544021607,
      "epoch": 0.277731373204143,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004995746564425191,
      "loss": 5.5087,
      "mean_token_accuracy": 0.17106839418411254,
      "num_tokens": 6123265.0,
      "step": 3325
    },
    {
      "entropy": 5.580819034576416,
      "epoch": 0.278149014366856,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004995728248318163,
      "loss": 5.5676,
      "mean_token_accuracy": 0.16482732594013214,
      "num_tokens": 6132855.0,
      "step": 3330
    },
    {
      "entropy": 5.468076133728028,
      "epoch": 0.278566655529569,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004995709892896961,
      "loss": 5.4537,
      "mean_token_accuracy": 0.16931613385677338,
      "num_tokens": 6142739.0,
      "step": 3335
    },
    {
      "entropy": 5.483571529388428,
      "epoch": 0.278984296692282,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004995691498161903,
      "loss": 5.3865,
      "mean_token_accuracy": 0.1695520266890526,
      "num_tokens": 6152169.0,
      "step": 3340
    },
    {
      "entropy": 5.3815367221832275,
      "epoch": 0.279401937854995,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004995673064113313,
      "loss": 5.4103,
      "mean_token_accuracy": 0.1787784531712532,
      "num_tokens": 6160235.0,
      "step": 3345
    },
    {
      "entropy": 5.49701042175293,
      "epoch": 0.279819579017708,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004995654590751513,
      "loss": 5.4065,
      "mean_token_accuracy": 0.16954068541526796,
      "num_tokens": 6169440.0,
      "step": 3350
    },
    {
      "entropy": 5.473315906524658,
      "epoch": 0.28023722018042097,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004995636078076825,
      "loss": 5.4342,
      "mean_token_accuracy": 0.1779964342713356,
      "num_tokens": 6178456.0,
      "step": 3355
    },
    {
      "entropy": 5.461573886871338,
      "epoch": 0.280654861343134,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004995617526089575,
      "loss": 5.5125,
      "mean_token_accuracy": 0.16690640151500702,
      "num_tokens": 6186844.0,
      "step": 3360
    },
    {
      "entropy": 5.531739711761475,
      "epoch": 0.281072502505847,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004995598934790087,
      "loss": 5.4251,
      "mean_token_accuracy": 0.17161790579557418,
      "num_tokens": 6196117.0,
      "step": 3365
    },
    {
      "entropy": 5.4806375980377195,
      "epoch": 0.28149014366855996,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004995580304178687,
      "loss": 5.4232,
      "mean_token_accuracy": 0.17214852422475815,
      "num_tokens": 6204636.0,
      "step": 3370
    },
    {
      "entropy": 5.398985528945923,
      "epoch": 0.281907784831273,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004995561634255701,
      "loss": 5.4457,
      "mean_token_accuracy": 0.1673709973692894,
      "num_tokens": 6214537.0,
      "step": 3375
    },
    {
      "entropy": 5.4943078517913815,
      "epoch": 0.28232542599398597,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004995542925021455,
      "loss": 5.5147,
      "mean_token_accuracy": 0.16867199391126633,
      "num_tokens": 6224753.0,
      "step": 3380
    },
    {
      "entropy": 5.747494173049927,
      "epoch": 0.28274306715669895,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004995524176476277,
      "loss": 5.692,
      "mean_token_accuracy": 0.16499322801828384,
      "num_tokens": 6233842.0,
      "step": 3385
    },
    {
      "entropy": 5.621364450454712,
      "epoch": 0.283160708319412,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004995505388620497,
      "loss": 5.5243,
      "mean_token_accuracy": 0.17118486315011977,
      "num_tokens": 6243467.0,
      "step": 3390
    },
    {
      "entropy": 5.429578542709351,
      "epoch": 0.28357834948212496,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000499548656145444,
      "loss": 5.5671,
      "mean_token_accuracy": 0.1568418487906456,
      "num_tokens": 6253263.0,
      "step": 3395
    },
    {
      "entropy": 5.573283243179321,
      "epoch": 0.28399599064483794,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004995467694978439,
      "loss": 5.402,
      "mean_token_accuracy": 0.17850250601768494,
      "num_tokens": 6262249.0,
      "step": 3400
    },
    {
      "entropy": 5.463663387298584,
      "epoch": 0.28441363180755097,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004995448789192823,
      "loss": 5.4295,
      "mean_token_accuracy": 0.1717173531651497,
      "num_tokens": 6271154.0,
      "step": 3405
    },
    {
      "entropy": 5.4220223903656,
      "epoch": 0.28483127297026395,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004995429844097922,
      "loss": 5.5015,
      "mean_token_accuracy": 0.1784306287765503,
      "num_tokens": 6280377.0,
      "step": 3410
    },
    {
      "entropy": 5.57632474899292,
      "epoch": 0.2852489141329769,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004995410859694071,
      "loss": 5.5089,
      "mean_token_accuracy": 0.1689310923218727,
      "num_tokens": 6289941.0,
      "step": 3415
    },
    {
      "entropy": 5.488957262039184,
      "epoch": 0.28566655529568996,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995391835981599,
      "loss": 5.5573,
      "mean_token_accuracy": 0.16624608933925628,
      "num_tokens": 6299281.0,
      "step": 3420
    },
    {
      "entropy": 5.595778846740723,
      "epoch": 0.28608419645840294,
      "grad_norm": 1.140625,
      "learning_rate": 0.000499537277296084,
      "loss": 5.5249,
      "mean_token_accuracy": 0.16632925122976303,
      "num_tokens": 6309408.0,
      "step": 3425
    },
    {
      "entropy": 5.50165581703186,
      "epoch": 0.2865018376211159,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004995353670632127,
      "loss": 5.5242,
      "mean_token_accuracy": 0.17109988778829574,
      "num_tokens": 6318979.0,
      "step": 3430
    },
    {
      "entropy": 5.483353614807129,
      "epoch": 0.28691947878382895,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004995334528995797,
      "loss": 5.3567,
      "mean_token_accuracy": 0.17871294766664506,
      "num_tokens": 6327616.0,
      "step": 3435
    },
    {
      "entropy": 5.409444570541382,
      "epoch": 0.28733711994654193,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004995315348052182,
      "loss": 5.3874,
      "mean_token_accuracy": 0.1814719870686531,
      "num_tokens": 6336486.0,
      "step": 3440
    },
    {
      "entropy": 5.418563032150269,
      "epoch": 0.2877547611092549,
      "grad_norm": 1.265625,
      "learning_rate": 0.000499529612780162,
      "loss": 5.4481,
      "mean_token_accuracy": 0.16858992278575896,
      "num_tokens": 6345671.0,
      "step": 3445
    },
    {
      "entropy": 5.611011266708374,
      "epoch": 0.28817240227196794,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004995276868244446,
      "loss": 5.5455,
      "mean_token_accuracy": 0.16791318058967591,
      "num_tokens": 6354534.0,
      "step": 3450
    },
    {
      "entropy": 5.499365186691284,
      "epoch": 0.2885900434346809,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004995257569380998,
      "loss": 5.3989,
      "mean_token_accuracy": 0.1744227558374405,
      "num_tokens": 6364052.0,
      "step": 3455
    },
    {
      "entropy": 5.545021820068359,
      "epoch": 0.2890076845973939,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004995238231211614,
      "loss": 5.5207,
      "mean_token_accuracy": 0.1676422566175461,
      "num_tokens": 6373305.0,
      "step": 3460
    },
    {
      "entropy": 5.512579870223999,
      "epoch": 0.28942532576010693,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004995218853736631,
      "loss": 5.4487,
      "mean_token_accuracy": 0.1660549208521843,
      "num_tokens": 6382673.0,
      "step": 3465
    },
    {
      "entropy": 5.534687423706055,
      "epoch": 0.2898429669228199,
      "grad_norm": 1.125,
      "learning_rate": 0.0004995199436956391,
      "loss": 5.555,
      "mean_token_accuracy": 0.16030752062797546,
      "num_tokens": 6392122.0,
      "step": 3470
    },
    {
      "entropy": 5.603588008880616,
      "epoch": 0.2902606080855329,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004995179980871231,
      "loss": 5.5507,
      "mean_token_accuracy": 0.1647276908159256,
      "num_tokens": 6401130.0,
      "step": 3475
    },
    {
      "entropy": 5.497493600845337,
      "epoch": 0.2906782492482459,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004995160485481493,
      "loss": 5.4269,
      "mean_token_accuracy": 0.17015280574560165,
      "num_tokens": 6409471.0,
      "step": 3480
    },
    {
      "entropy": 5.416902780532837,
      "epoch": 0.2910958904109589,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499514095078752,
      "loss": 5.4452,
      "mean_token_accuracy": 0.16760370135307312,
      "num_tokens": 6418902.0,
      "step": 3485
    },
    {
      "entropy": 5.552921962738037,
      "epoch": 0.2915135315736719,
      "grad_norm": 1.09375,
      "learning_rate": 0.000499512137678965,
      "loss": 5.5363,
      "mean_token_accuracy": 0.1662679061293602,
      "num_tokens": 6428529.0,
      "step": 3490
    },
    {
      "entropy": 5.490020132064819,
      "epoch": 0.2919311727363849,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004995101763488229,
      "loss": 5.3835,
      "mean_token_accuracy": 0.17481506168842315,
      "num_tokens": 6437460.0,
      "step": 3495
    },
    {
      "entropy": 5.474491262435913,
      "epoch": 0.2923488138990979,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004995082110883599,
      "loss": 5.4209,
      "mean_token_accuracy": 0.1738422319293022,
      "num_tokens": 6446005.0,
      "step": 3500
    },
    {
      "entropy": 5.478244733810425,
      "epoch": 0.29276645506181087,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995062418976104,
      "loss": 5.4825,
      "mean_token_accuracy": 0.17011081129312516,
      "num_tokens": 6455241.0,
      "step": 3505
    },
    {
      "entropy": 5.546651840209961,
      "epoch": 0.2931840962245239,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004995042687766089,
      "loss": 5.4695,
      "mean_token_accuracy": 0.1717524617910385,
      "num_tokens": 6464748.0,
      "step": 3510
    },
    {
      "entropy": 5.584757280349732,
      "epoch": 0.2936017373872369,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004995022917253899,
      "loss": 5.524,
      "mean_token_accuracy": 0.16998693495988845,
      "num_tokens": 6473834.0,
      "step": 3515
    },
    {
      "entropy": 5.469439220428467,
      "epoch": 0.29401937854994986,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004995003107439882,
      "loss": 5.4556,
      "mean_token_accuracy": 0.1739688917994499,
      "num_tokens": 6482169.0,
      "step": 3520
    },
    {
      "entropy": 5.410130739212036,
      "epoch": 0.2944370197126629,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004994983258324383,
      "loss": 5.4014,
      "mean_token_accuracy": 0.17411208897829056,
      "num_tokens": 6491798.0,
      "step": 3525
    },
    {
      "entropy": 5.402098608016968,
      "epoch": 0.2948546608753759,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004994963369907748,
      "loss": 5.3449,
      "mean_token_accuracy": 0.17219721227884294,
      "num_tokens": 6500143.0,
      "step": 3530
    },
    {
      "entropy": 5.369478321075439,
      "epoch": 0.29527230203808885,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004994943442190328,
      "loss": 5.4815,
      "mean_token_accuracy": 0.16979831457138062,
      "num_tokens": 6509745.0,
      "step": 3535
    },
    {
      "entropy": 5.628178644180298,
      "epoch": 0.2956899432008019,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994923475172471,
      "loss": 5.5531,
      "mean_token_accuracy": 0.163117353618145,
      "num_tokens": 6520931.0,
      "step": 3540
    },
    {
      "entropy": 5.5175331115722654,
      "epoch": 0.29610758436351486,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004994903468854527,
      "loss": 5.4896,
      "mean_token_accuracy": 0.17120033651590347,
      "num_tokens": 6529774.0,
      "step": 3545
    },
    {
      "entropy": 5.425450468063355,
      "epoch": 0.29652522552622784,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004994883423236843,
      "loss": 5.3989,
      "mean_token_accuracy": 0.16883310824632644,
      "num_tokens": 6538922.0,
      "step": 3550
    },
    {
      "entropy": 5.456408739089966,
      "epoch": 0.2969428666889409,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004994863338319775,
      "loss": 5.452,
      "mean_token_accuracy": 0.1734740749001503,
      "num_tokens": 6548891.0,
      "step": 3555
    },
    {
      "entropy": 5.396758222579956,
      "epoch": 0.29736050785165385,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004994843214103672,
      "loss": 5.2893,
      "mean_token_accuracy": 0.16934687346220018,
      "num_tokens": 6557025.0,
      "step": 3560
    },
    {
      "entropy": 5.520787620544434,
      "epoch": 0.29777814901436683,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004994823050588886,
      "loss": 5.4657,
      "mean_token_accuracy": 0.17021430730819703,
      "num_tokens": 6567037.0,
      "step": 3565
    },
    {
      "entropy": 5.481213045120239,
      "epoch": 0.29819579017707987,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000499480284777577,
      "loss": 5.5072,
      "mean_token_accuracy": 0.17070177495479583,
      "num_tokens": 6576766.0,
      "step": 3570
    },
    {
      "entropy": 5.488879442214966,
      "epoch": 0.29861343133979285,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004994782605664677,
      "loss": 5.4616,
      "mean_token_accuracy": 0.17389861792325972,
      "num_tokens": 6585576.0,
      "step": 3575
    },
    {
      "entropy": 5.533296680450439,
      "epoch": 0.2990310725025058,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004994762324255965,
      "loss": 5.507,
      "mean_token_accuracy": 0.16581012904644013,
      "num_tokens": 6594463.0,
      "step": 3580
    },
    {
      "entropy": 5.436222791671753,
      "epoch": 0.29944871366521886,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004994742003549986,
      "loss": 5.406,
      "mean_token_accuracy": 0.16841213405132294,
      "num_tokens": 6603651.0,
      "step": 3585
    },
    {
      "entropy": 5.515470027923584,
      "epoch": 0.29986635482793184,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004994721643547095,
      "loss": 5.4564,
      "mean_token_accuracy": 0.17240571826696396,
      "num_tokens": 6612950.0,
      "step": 3590
    },
    {
      "entropy": 5.525665330886841,
      "epoch": 0.3002839959906448,
      "grad_norm": 1.109375,
      "learning_rate": 0.000499470124424765,
      "loss": 5.4161,
      "mean_token_accuracy": 0.1754685029387474,
      "num_tokens": 6622013.0,
      "step": 3595
    },
    {
      "entropy": 5.503145170211792,
      "epoch": 0.30070163715335785,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004994680805652008,
      "loss": 5.4908,
      "mean_token_accuracy": 0.17114405781030656,
      "num_tokens": 6632046.0,
      "step": 3600
    },
    {
      "entropy": 5.509345245361328,
      "epoch": 0.3011192783160708,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004994660327760527,
      "loss": 5.5412,
      "mean_token_accuracy": 0.17287831604480744,
      "num_tokens": 6641169.0,
      "step": 3605
    },
    {
      "entropy": 5.460749101638794,
      "epoch": 0.3015369194787838,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004994639810573565,
      "loss": 5.3378,
      "mean_token_accuracy": 0.18106640875339508,
      "num_tokens": 6649663.0,
      "step": 3610
    },
    {
      "entropy": 5.547510719299316,
      "epoch": 0.30195456064149684,
      "grad_norm": 1.140625,
      "learning_rate": 0.000499461925409148,
      "loss": 5.5401,
      "mean_token_accuracy": 0.16402754783630372,
      "num_tokens": 6659448.0,
      "step": 3615
    },
    {
      "entropy": 5.565422201156617,
      "epoch": 0.3023722018042098,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004994598658314634,
      "loss": 5.5074,
      "mean_token_accuracy": 0.17040248960256577,
      "num_tokens": 6670314.0,
      "step": 3620
    },
    {
      "entropy": 5.487205886840821,
      "epoch": 0.3027898429669228,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004994578023243386,
      "loss": 5.4488,
      "mean_token_accuracy": 0.17701876014471055,
      "num_tokens": 6679413.0,
      "step": 3625
    },
    {
      "entropy": 5.462418937683106,
      "epoch": 0.30320748412963583,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004994557348878098,
      "loss": 5.425,
      "mean_token_accuracy": 0.17467287182807922,
      "num_tokens": 6688439.0,
      "step": 3630
    },
    {
      "entropy": 5.500660133361817,
      "epoch": 0.3036251252923488,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004994536635219133,
      "loss": 5.4111,
      "mean_token_accuracy": 0.17729663252830505,
      "num_tokens": 6697471.0,
      "step": 3635
    },
    {
      "entropy": 5.53426079750061,
      "epoch": 0.3040427664550618,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004994515882266852,
      "loss": 5.4637,
      "mean_token_accuracy": 0.1736002594232559,
      "num_tokens": 6706698.0,
      "step": 3640
    },
    {
      "entropy": 5.389495468139648,
      "epoch": 0.3044604076177748,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004994495090021618,
      "loss": 5.3802,
      "mean_token_accuracy": 0.1806873559951782,
      "num_tokens": 6715676.0,
      "step": 3645
    },
    {
      "entropy": 5.371573352813721,
      "epoch": 0.3048780487804878,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004994474258483796,
      "loss": 5.4234,
      "mean_token_accuracy": 0.17871695756912231,
      "num_tokens": 6723818.0,
      "step": 3650
    },
    {
      "entropy": 5.5329201221466064,
      "epoch": 0.3052956899432008,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004994453387653751,
      "loss": 5.4393,
      "mean_token_accuracy": 0.17155028581619264,
      "num_tokens": 6733450.0,
      "step": 3655
    },
    {
      "entropy": 5.5077282905578615,
      "epoch": 0.3057133311059138,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004994432477531847,
      "loss": 5.3792,
      "mean_token_accuracy": 0.1691980928182602,
      "num_tokens": 6743142.0,
      "step": 3660
    },
    {
      "entropy": 5.3698314189910885,
      "epoch": 0.3061309722686268,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004994411528118452,
      "loss": 5.3371,
      "mean_token_accuracy": 0.17547308504581452,
      "num_tokens": 6752185.0,
      "step": 3665
    },
    {
      "entropy": 5.382838296890259,
      "epoch": 0.30654861343133977,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004994390539413931,
      "loss": 5.4441,
      "mean_token_accuracy": 0.16950728744268417,
      "num_tokens": 6761549.0,
      "step": 3670
    },
    {
      "entropy": 5.464862394332886,
      "epoch": 0.3069662545940528,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004994369511418652,
      "loss": 5.4435,
      "mean_token_accuracy": 0.17317559719085693,
      "num_tokens": 6770395.0,
      "step": 3675
    },
    {
      "entropy": 5.665520048141479,
      "epoch": 0.3073838957567658,
      "grad_norm": 1.25,
      "learning_rate": 0.0004994348444132983,
      "loss": 5.5202,
      "mean_token_accuracy": 0.17058894634246827,
      "num_tokens": 6779154.0,
      "step": 3680
    },
    {
      "entropy": 5.412521505355835,
      "epoch": 0.30780153691947876,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004994327337557293,
      "loss": 5.418,
      "mean_token_accuracy": 0.18299008756875992,
      "num_tokens": 6789263.0,
      "step": 3685
    },
    {
      "entropy": 5.45570969581604,
      "epoch": 0.3082191780821918,
      "grad_norm": 1.125,
      "learning_rate": 0.0004994306191691952,
      "loss": 5.4908,
      "mean_token_accuracy": 0.166125063598156,
      "num_tokens": 6797985.0,
      "step": 3690
    },
    {
      "entropy": 5.499072170257568,
      "epoch": 0.30863681924490477,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004994285006537329,
      "loss": 5.3968,
      "mean_token_accuracy": 0.17336190789937972,
      "num_tokens": 6807566.0,
      "step": 3695
    },
    {
      "entropy": 5.451264333724976,
      "epoch": 0.30905446040761775,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004994263782093796,
      "loss": 5.4346,
      "mean_token_accuracy": 0.17076193541288376,
      "num_tokens": 6815637.0,
      "step": 3700
    },
    {
      "entropy": 5.440542793273925,
      "epoch": 0.3094721015703308,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004994242518361723,
      "loss": 5.3301,
      "mean_token_accuracy": 0.17853307276964187,
      "num_tokens": 6823721.0,
      "step": 3705
    },
    {
      "entropy": 5.552204942703247,
      "epoch": 0.30988974273304376,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004994221215341484,
      "loss": 5.4947,
      "mean_token_accuracy": 0.16985803246498107,
      "num_tokens": 6833437.0,
      "step": 3710
    },
    {
      "entropy": 5.461875104904175,
      "epoch": 0.31030738389575674,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004994199873033452,
      "loss": 5.54,
      "mean_token_accuracy": 0.16892331689596177,
      "num_tokens": 6842287.0,
      "step": 3715
    },
    {
      "entropy": 5.549509859085083,
      "epoch": 0.3107250250584698,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004994178491437999,
      "loss": 5.3781,
      "mean_token_accuracy": 0.16952035278081895,
      "num_tokens": 6851520.0,
      "step": 3720
    },
    {
      "entropy": 5.418500137329102,
      "epoch": 0.31114266622118275,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994157070555501,
      "loss": 5.4484,
      "mean_token_accuracy": 0.17235054969787597,
      "num_tokens": 6861219.0,
      "step": 3725
    },
    {
      "entropy": 5.520118331909179,
      "epoch": 0.31156030738389573,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004994135610386331,
      "loss": 5.5773,
      "mean_token_accuracy": 0.16754324436187745,
      "num_tokens": 6870525.0,
      "step": 3730
    },
    {
      "entropy": 5.609558916091919,
      "epoch": 0.31197794854660876,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004994114110930867,
      "loss": 5.5009,
      "mean_token_accuracy": 0.16462523490190506,
      "num_tokens": 6880183.0,
      "step": 3735
    },
    {
      "entropy": 5.485359001159668,
      "epoch": 0.31239558970932174,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004994092572189483,
      "loss": 5.511,
      "mean_token_accuracy": 0.16772003024816512,
      "num_tokens": 6890442.0,
      "step": 3740
    },
    {
      "entropy": 5.494738817214966,
      "epoch": 0.3128132308720348,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004994070994162558,
      "loss": 5.4216,
      "mean_token_accuracy": 0.17486200034618377,
      "num_tokens": 6900091.0,
      "step": 3745
    },
    {
      "entropy": 5.4695563316345215,
      "epoch": 0.31323087203474775,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004994049376850471,
      "loss": 5.4689,
      "mean_token_accuracy": 0.17180595993995668,
      "num_tokens": 6910489.0,
      "step": 3750
    },
    {
      "entropy": 5.496180200576783,
      "epoch": 0.31364851319746073,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004994027720253596,
      "loss": 5.4565,
      "mean_token_accuracy": 0.1682304173707962,
      "num_tokens": 6919895.0,
      "step": 3755
    },
    {
      "entropy": 5.693031644821167,
      "epoch": 0.31406615436017377,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004994006024372315,
      "loss": 5.5842,
      "mean_token_accuracy": 0.16491315960884095,
      "num_tokens": 6929511.0,
      "step": 3760
    },
    {
      "entropy": 5.454358005523682,
      "epoch": 0.31448379552288674,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004993984289207007,
      "loss": 5.3617,
      "mean_token_accuracy": 0.18178811967372893,
      "num_tokens": 6938059.0,
      "step": 3765
    },
    {
      "entropy": 5.404311084747315,
      "epoch": 0.3149014366855997,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004993962514758054,
      "loss": 5.3914,
      "mean_token_accuracy": 0.17555590122938156,
      "num_tokens": 6946894.0,
      "step": 3770
    },
    {
      "entropy": 5.5242852687835695,
      "epoch": 0.31531907784831276,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004993940701025836,
      "loss": 5.3853,
      "mean_token_accuracy": 0.1679450422525406,
      "num_tokens": 6955494.0,
      "step": 3775
    },
    {
      "entropy": 5.482671546936035,
      "epoch": 0.31573671901102573,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004993918848010734,
      "loss": 5.4947,
      "mean_token_accuracy": 0.1700427144765854,
      "num_tokens": 6964347.0,
      "step": 3780
    },
    {
      "entropy": 5.541811227798462,
      "epoch": 0.3161543601737387,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004993896955713133,
      "loss": 5.4704,
      "mean_token_accuracy": 0.16363491863012314,
      "num_tokens": 6974750.0,
      "step": 3785
    },
    {
      "entropy": 5.4976883888244625,
      "epoch": 0.31657200133645175,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004993875024133414,
      "loss": 5.4639,
      "mean_token_accuracy": 0.1697466790676117,
      "num_tokens": 6984228.0,
      "step": 3790
    },
    {
      "entropy": 5.449398231506348,
      "epoch": 0.3169896424991647,
      "grad_norm": 1.25,
      "learning_rate": 0.0004993853053271962,
      "loss": 5.4505,
      "mean_token_accuracy": 0.17396792620420456,
      "num_tokens": 6993173.0,
      "step": 3795
    },
    {
      "entropy": 5.663131284713745,
      "epoch": 0.3174072836618777,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004993831043129161,
      "loss": 5.6454,
      "mean_token_accuracy": 0.1585911437869072,
      "num_tokens": 7003486.0,
      "step": 3800
    },
    {
      "entropy": 5.448543548583984,
      "epoch": 0.31782492482459074,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004993808993705397,
      "loss": 5.3884,
      "mean_token_accuracy": 0.17939800918102264,
      "num_tokens": 7011985.0,
      "step": 3805
    },
    {
      "entropy": 5.462221574783325,
      "epoch": 0.3182425659873037,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004993786905001055,
      "loss": 5.3647,
      "mean_token_accuracy": 0.176728555560112,
      "num_tokens": 7021148.0,
      "step": 3810
    },
    {
      "entropy": 5.46138858795166,
      "epoch": 0.3186602071500167,
      "grad_norm": 1.125,
      "learning_rate": 0.0004993764777016524,
      "loss": 5.4079,
      "mean_token_accuracy": 0.1701902762055397,
      "num_tokens": 7030110.0,
      "step": 3815
    },
    {
      "entropy": 5.465259981155396,
      "epoch": 0.31907784831272973,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004993742609752188,
      "loss": 5.4838,
      "mean_token_accuracy": 0.1755315348505974,
      "num_tokens": 7039428.0,
      "step": 3820
    },
    {
      "entropy": 5.609820795059204,
      "epoch": 0.3194954894754427,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004993720403208438,
      "loss": 5.5255,
      "mean_token_accuracy": 0.1672436237335205,
      "num_tokens": 7048703.0,
      "step": 3825
    },
    {
      "entropy": 5.4761299133300785,
      "epoch": 0.3199131306381557,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000499369815738566,
      "loss": 5.4339,
      "mean_token_accuracy": 0.16614928096532822,
      "num_tokens": 7057612.0,
      "step": 3830
    },
    {
      "entropy": 5.477329874038697,
      "epoch": 0.3203307718008687,
      "grad_norm": 1.125,
      "learning_rate": 0.0004993675872284248,
      "loss": 5.3872,
      "mean_token_accuracy": 0.1730633869767189,
      "num_tokens": 7066660.0,
      "step": 3835
    },
    {
      "entropy": 5.435452938079834,
      "epoch": 0.3207484129635817,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004993653547904586,
      "loss": 5.3691,
      "mean_token_accuracy": 0.16408761739730834,
      "num_tokens": 7075368.0,
      "step": 3840
    },
    {
      "entropy": 5.454200029373169,
      "epoch": 0.3211660541262947,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000499363118424707,
      "loss": 5.4654,
      "mean_token_accuracy": 0.17837097644805908,
      "num_tokens": 7085276.0,
      "step": 3845
    },
    {
      "entropy": 5.4827862739562985,
      "epoch": 0.3215836952890077,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004993608781312088,
      "loss": 5.4342,
      "mean_token_accuracy": 0.16890606433153152,
      "num_tokens": 7094292.0,
      "step": 3850
    },
    {
      "entropy": 5.619381523132324,
      "epoch": 0.3220013364517207,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004993586339100035,
      "loss": 5.6588,
      "mean_token_accuracy": 0.16210833191871643,
      "num_tokens": 7104712.0,
      "step": 3855
    },
    {
      "entropy": 5.581846046447754,
      "epoch": 0.32241897761443367,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004993563857611301,
      "loss": 5.4461,
      "mean_token_accuracy": 0.18202234506607057,
      "num_tokens": 7113297.0,
      "step": 3860
    },
    {
      "entropy": 5.394202089309692,
      "epoch": 0.3228366187771467,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004993541336846283,
      "loss": 5.4533,
      "mean_token_accuracy": 0.1653970956802368,
      "num_tokens": 7122706.0,
      "step": 3865
    },
    {
      "entropy": 5.512542057037353,
      "epoch": 0.3232542599398597,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004993518776805372,
      "loss": 5.4188,
      "mean_token_accuracy": 0.17372276186943053,
      "num_tokens": 7131939.0,
      "step": 3870
    },
    {
      "entropy": 5.3669188022613525,
      "epoch": 0.32367190110257266,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004993496177488965,
      "loss": 5.3126,
      "mean_token_accuracy": 0.18179777413606643,
      "num_tokens": 7141016.0,
      "step": 3875
    },
    {
      "entropy": 5.382031631469727,
      "epoch": 0.3240895422652857,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004993473538897457,
      "loss": 5.2542,
      "mean_token_accuracy": 0.18906042873859405,
      "num_tokens": 7149177.0,
      "step": 3880
    },
    {
      "entropy": 5.469439220428467,
      "epoch": 0.32450718342799867,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004993450861031243,
      "loss": 5.3876,
      "mean_token_accuracy": 0.17814804166555404,
      "num_tokens": 7157839.0,
      "step": 3885
    },
    {
      "entropy": 5.582052564620971,
      "epoch": 0.32492482459071165,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004993428143890722,
      "loss": 5.5406,
      "mean_token_accuracy": 0.16474749743938447,
      "num_tokens": 7167759.0,
      "step": 3890
    },
    {
      "entropy": 5.6012077808380125,
      "epoch": 0.3253424657534247,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004993405387476291,
      "loss": 5.4721,
      "mean_token_accuracy": 0.16357818096876145,
      "num_tokens": 7177141.0,
      "step": 3895
    },
    {
      "entropy": 5.377801370620728,
      "epoch": 0.32576010691613766,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004993382591788349,
      "loss": 5.3816,
      "mean_token_accuracy": 0.17757417857646943,
      "num_tokens": 7185350.0,
      "step": 3900
    },
    {
      "entropy": 5.352601861953735,
      "epoch": 0.32617774807885064,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004993359756827293,
      "loss": 5.3258,
      "mean_token_accuracy": 0.17330418080091475,
      "num_tokens": 7194073.0,
      "step": 3905
    },
    {
      "entropy": 5.51496434211731,
      "epoch": 0.32659538924156367,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004993336882593524,
      "loss": 5.5013,
      "mean_token_accuracy": 0.16401236206293107,
      "num_tokens": 7203421.0,
      "step": 3910
    },
    {
      "entropy": 5.389591217041016,
      "epoch": 0.32701303040427665,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004993313969087444,
      "loss": 5.2743,
      "mean_token_accuracy": 0.17551216781139373,
      "num_tokens": 7212548.0,
      "step": 3915
    },
    {
      "entropy": 5.519296264648437,
      "epoch": 0.32743067156698963,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004993291016309452,
      "loss": 5.512,
      "mean_token_accuracy": 0.1692492887377739,
      "num_tokens": 7221668.0,
      "step": 3920
    },
    {
      "entropy": 5.46710615158081,
      "epoch": 0.32784831272970266,
      "grad_norm": 1.15625,
      "learning_rate": 0.000499326802425995,
      "loss": 5.4019,
      "mean_token_accuracy": 0.1661408171057701,
      "num_tokens": 7231775.0,
      "step": 3925
    },
    {
      "entropy": 5.508456468582153,
      "epoch": 0.32826595389241564,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004993244992939342,
      "loss": 5.4339,
      "mean_token_accuracy": 0.17021936625242234,
      "num_tokens": 7241264.0,
      "step": 3930
    },
    {
      "entropy": 5.513637828826904,
      "epoch": 0.3286835950551286,
      "grad_norm": 1.171875,
      "learning_rate": 0.000499322192234803,
      "loss": 5.4714,
      "mean_token_accuracy": 0.16775393933057786,
      "num_tokens": 7250519.0,
      "step": 3935
    },
    {
      "entropy": 5.558846855163575,
      "epoch": 0.32910123621784165,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004993198812486418,
      "loss": 5.5018,
      "mean_token_accuracy": 0.16846247613430024,
      "num_tokens": 7259533.0,
      "step": 3940
    },
    {
      "entropy": 5.421068096160889,
      "epoch": 0.32951887738055463,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000499317566335491,
      "loss": 5.3448,
      "mean_token_accuracy": 0.17600925415754318,
      "num_tokens": 7268980.0,
      "step": 3945
    },
    {
      "entropy": 5.460406970977783,
      "epoch": 0.3299365185432676,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004993152474953913,
      "loss": 5.4394,
      "mean_token_accuracy": 0.16838183552026748,
      "num_tokens": 7277126.0,
      "step": 3950
    },
    {
      "entropy": 5.517564821243286,
      "epoch": 0.33035415970598064,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004993129247283831,
      "loss": 5.3928,
      "mean_token_accuracy": 0.18117072880268098,
      "num_tokens": 7286223.0,
      "step": 3955
    },
    {
      "entropy": 5.446171569824219,
      "epoch": 0.3307718008686936,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004993105980345071,
      "loss": 5.4018,
      "mean_token_accuracy": 0.17625371664762496,
      "num_tokens": 7295054.0,
      "step": 3960
    },
    {
      "entropy": 5.449082612991333,
      "epoch": 0.3311894420314066,
      "grad_norm": 1.125,
      "learning_rate": 0.0004993082674138042,
      "loss": 5.4285,
      "mean_token_accuracy": 0.17082341611385346,
      "num_tokens": 7304502.0,
      "step": 3965
    },
    {
      "entropy": 5.513706970214844,
      "epoch": 0.33160708319411963,
      "grad_norm": 1.078125,
      "learning_rate": 0.000499305932866315,
      "loss": 5.4101,
      "mean_token_accuracy": 0.17128008902072905,
      "num_tokens": 7314093.0,
      "step": 3970
    },
    {
      "entropy": 5.487787532806396,
      "epoch": 0.3320247243568326,
      "grad_norm": 1.25,
      "learning_rate": 0.0004993035943920805,
      "loss": 5.4315,
      "mean_token_accuracy": 0.17096492648124695,
      "num_tokens": 7323059.0,
      "step": 3975
    },
    {
      "entropy": 5.577842712402344,
      "epoch": 0.3324423655195456,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004993012519911416,
      "loss": 5.5601,
      "mean_token_accuracy": 0.16351205110549927,
      "num_tokens": 7332596.0,
      "step": 3980
    },
    {
      "entropy": 5.440038442611694,
      "epoch": 0.3328600066822586,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004992989056635392,
      "loss": 5.3579,
      "mean_token_accuracy": 0.17744496315717698,
      "num_tokens": 7341671.0,
      "step": 3985
    },
    {
      "entropy": 5.377516365051269,
      "epoch": 0.3332776478449716,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004992965554093146,
      "loss": 5.4915,
      "mean_token_accuracy": 0.17166960388422012,
      "num_tokens": 7350344.0,
      "step": 3990
    },
    {
      "entropy": 5.5424505233764645,
      "epoch": 0.3336952890076846,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004992942012285086,
      "loss": 5.4988,
      "mean_token_accuracy": 0.17335852086544037,
      "num_tokens": 7359844.0,
      "step": 3995
    },
    {
      "entropy": 5.500360536575317,
      "epoch": 0.3341129301703976,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004992918431211629,
      "loss": 5.4235,
      "mean_token_accuracy": 0.17371536046266556,
      "num_tokens": 7369377.0,
      "step": 4000
    },
    {
      "entropy": 5.457582235336304,
      "epoch": 0.3345305713331106,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004992894810873183,
      "loss": 5.3979,
      "mean_token_accuracy": 0.1774983748793602,
      "num_tokens": 7378814.0,
      "step": 4005
    },
    {
      "entropy": 5.443440914154053,
      "epoch": 0.33494821249582357,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004992871151270165,
      "loss": 5.3996,
      "mean_token_accuracy": 0.1764570638537407,
      "num_tokens": 7387333.0,
      "step": 4010
    },
    {
      "entropy": 5.421675539016723,
      "epoch": 0.3353658536585366,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004992847452402986,
      "loss": 5.4407,
      "mean_token_accuracy": 0.17375606745481492,
      "num_tokens": 7396895.0,
      "step": 4015
    },
    {
      "entropy": 5.496244049072265,
      "epoch": 0.3357834948212496,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004992823714272064,
      "loss": 5.4299,
      "mean_token_accuracy": 0.17418541759252548,
      "num_tokens": 7406430.0,
      "step": 4020
    },
    {
      "entropy": 5.433465576171875,
      "epoch": 0.33620113598396256,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004992799936877813,
      "loss": 5.2732,
      "mean_token_accuracy": 0.18776014894247056,
      "num_tokens": 7415224.0,
      "step": 4025
    },
    {
      "entropy": 5.561156797409057,
      "epoch": 0.3366187771466756,
      "grad_norm": 1.375,
      "learning_rate": 0.0004992776120220649,
      "loss": 5.607,
      "mean_token_accuracy": 0.161159510910511,
      "num_tokens": 7424250.0,
      "step": 4030
    },
    {
      "entropy": 5.531838226318359,
      "epoch": 0.3370364183093886,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004992752264300989,
      "loss": 5.4354,
      "mean_token_accuracy": 0.1759601816534996,
      "num_tokens": 7433579.0,
      "step": 4035
    },
    {
      "entropy": 5.38196702003479,
      "epoch": 0.33745405947210155,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004992728369119252,
      "loss": 5.402,
      "mean_token_accuracy": 0.17414322048425673,
      "num_tokens": 7442621.0,
      "step": 4040
    },
    {
      "entropy": 5.533233785629273,
      "epoch": 0.3378717006348146,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004992704434675855,
      "loss": 5.4952,
      "mean_token_accuracy": 0.16609173864126206,
      "num_tokens": 7452077.0,
      "step": 4045
    },
    {
      "entropy": 5.438749504089356,
      "epoch": 0.33828934179752757,
      "grad_norm": 1.125,
      "learning_rate": 0.0004992680460971218,
      "loss": 5.4049,
      "mean_token_accuracy": 0.1703796476125717,
      "num_tokens": 7461146.0,
      "step": 4050
    },
    {
      "entropy": 5.381596326828003,
      "epoch": 0.33870698296024054,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004992656448005758,
      "loss": 5.3923,
      "mean_token_accuracy": 0.175481915473938,
      "num_tokens": 7470335.0,
      "step": 4055
    },
    {
      "entropy": 5.518864488601684,
      "epoch": 0.3391246241229536,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004992632395779899,
      "loss": 5.3962,
      "mean_token_accuracy": 0.1677387461066246,
      "num_tokens": 7479458.0,
      "step": 4060
    },
    {
      "entropy": 5.304780340194702,
      "epoch": 0.33954226528566656,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004992608304294061,
      "loss": 5.3504,
      "mean_token_accuracy": 0.1870820254087448,
      "num_tokens": 7488005.0,
      "step": 4065
    },
    {
      "entropy": 5.345000314712524,
      "epoch": 0.33995990644837953,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004992584173548664,
      "loss": 5.3477,
      "mean_token_accuracy": 0.17666814774274825,
      "num_tokens": 7496807.0,
      "step": 4070
    },
    {
      "entropy": 5.497731494903564,
      "epoch": 0.34037754761109257,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004992560003544131,
      "loss": 5.3833,
      "mean_token_accuracy": 0.18096455782651902,
      "num_tokens": 7505383.0,
      "step": 4075
    },
    {
      "entropy": 5.504910802841186,
      "epoch": 0.34079518877380555,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004992535794280887,
      "loss": 5.4622,
      "mean_token_accuracy": 0.16868881583213807,
      "num_tokens": 7515093.0,
      "step": 4080
    },
    {
      "entropy": 5.515105199813843,
      "epoch": 0.3412128299365185,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004992511545759354,
      "loss": 5.4388,
      "mean_token_accuracy": 0.16774399131536483,
      "num_tokens": 7524067.0,
      "step": 4085
    },
    {
      "entropy": 5.307038259506226,
      "epoch": 0.34163047109923156,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004992487257979958,
      "loss": 5.3634,
      "mean_token_accuracy": 0.1704171895980835,
      "num_tokens": 7533197.0,
      "step": 4090
    },
    {
      "entropy": 5.53603777885437,
      "epoch": 0.34204811226194454,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004992462930943123,
      "loss": 5.378,
      "mean_token_accuracy": 0.16514680087566375,
      "num_tokens": 7543030.0,
      "step": 4095
    },
    {
      "entropy": 5.470230436325073,
      "epoch": 0.3424657534246575,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004992438564649275,
      "loss": 5.4426,
      "mean_token_accuracy": 0.16946968883275987,
      "num_tokens": 7551556.0,
      "step": 4100
    },
    {
      "entropy": 5.487739992141724,
      "epoch": 0.34288339458737055,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004992414159098841,
      "loss": 5.3658,
      "mean_token_accuracy": 0.17220384776592254,
      "num_tokens": 7559284.0,
      "step": 4105
    },
    {
      "entropy": 5.544583940505982,
      "epoch": 0.3433010357500835,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004992389714292247,
      "loss": 5.4741,
      "mean_token_accuracy": 0.1721594214439392,
      "num_tokens": 7568750.0,
      "step": 4110
    },
    {
      "entropy": 5.374774599075318,
      "epoch": 0.3437186769127965,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004992365230229922,
      "loss": 5.3389,
      "mean_token_accuracy": 0.17926724553108214,
      "num_tokens": 7578309.0,
      "step": 4115
    },
    {
      "entropy": 5.357023048400879,
      "epoch": 0.34413631807550954,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004992340706912295,
      "loss": 5.3611,
      "mean_token_accuracy": 0.17205804735422134,
      "num_tokens": 7586760.0,
      "step": 4120
    },
    {
      "entropy": 5.409601068496704,
      "epoch": 0.3445539592382225,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004992316144339794,
      "loss": 5.3262,
      "mean_token_accuracy": 0.1809419348835945,
      "num_tokens": 7595472.0,
      "step": 4125
    },
    {
      "entropy": 5.4892051219940186,
      "epoch": 0.3449716004009355,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004992291542512852,
      "loss": 5.519,
      "mean_token_accuracy": 0.15911627560853958,
      "num_tokens": 7605663.0,
      "step": 4130
    },
    {
      "entropy": 5.534612464904785,
      "epoch": 0.34538924156364853,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004992266901431895,
      "loss": 5.4135,
      "mean_token_accuracy": 0.1721968099474907,
      "num_tokens": 7616110.0,
      "step": 4135
    },
    {
      "entropy": 5.4140078067779545,
      "epoch": 0.3458068827263615,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004992242221097358,
      "loss": 5.3598,
      "mean_token_accuracy": 0.17137768119573593,
      "num_tokens": 7625311.0,
      "step": 4140
    },
    {
      "entropy": 5.474600219726563,
      "epoch": 0.3462245238890745,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004992217501509672,
      "loss": 5.3866,
      "mean_token_accuracy": 0.1727500230073929,
      "num_tokens": 7634562.0,
      "step": 4145
    },
    {
      "entropy": 5.388373994827271,
      "epoch": 0.3466421650517875,
      "grad_norm": 1.125,
      "learning_rate": 0.000499219274266927,
      "loss": 5.3311,
      "mean_token_accuracy": 0.17528713196516038,
      "num_tokens": 7643231.0,
      "step": 4150
    },
    {
      "entropy": 5.455113363265991,
      "epoch": 0.3470598062145005,
      "grad_norm": 1.125,
      "learning_rate": 0.0004992167944576584,
      "loss": 5.4623,
      "mean_token_accuracy": 0.17543056905269622,
      "num_tokens": 7653556.0,
      "step": 4155
    },
    {
      "entropy": 5.442866277694702,
      "epoch": 0.3474774473772135,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000499214310723205,
      "loss": 5.3582,
      "mean_token_accuracy": 0.179081429541111,
      "num_tokens": 7662295.0,
      "step": 4160
    },
    {
      "entropy": 5.40164589881897,
      "epoch": 0.3478950885399265,
      "grad_norm": 1.125,
      "learning_rate": 0.0004992118230636102,
      "loss": 5.3895,
      "mean_token_accuracy": 0.17207947075366975,
      "num_tokens": 7671429.0,
      "step": 4165
    },
    {
      "entropy": 5.3517780780792235,
      "epoch": 0.3483127297026395,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004992093314789176,
      "loss": 5.3151,
      "mean_token_accuracy": 0.17298997491598128,
      "num_tokens": 7680589.0,
      "step": 4170
    },
    {
      "entropy": 5.411142587661743,
      "epoch": 0.34873037086535247,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004992068359691707,
      "loss": 5.3612,
      "mean_token_accuracy": 0.178671196103096,
      "num_tokens": 7689005.0,
      "step": 4175
    },
    {
      "entropy": 5.334112358093262,
      "epoch": 0.3491480120280655,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004992043365344132,
      "loss": 5.2888,
      "mean_token_accuracy": 0.17627826929092408,
      "num_tokens": 7698231.0,
      "step": 4180
    },
    {
      "entropy": 5.532045602798462,
      "epoch": 0.3495656531907785,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004992018331746891,
      "loss": 5.3888,
      "mean_token_accuracy": 0.17696067690849304,
      "num_tokens": 7706678.0,
      "step": 4185
    },
    {
      "entropy": 5.509274339675903,
      "epoch": 0.34998329435349146,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004991993258900418,
      "loss": 5.4457,
      "mean_token_accuracy": 0.17433779835700988,
      "num_tokens": 7715506.0,
      "step": 4190
    },
    {
      "entropy": 5.482197046279907,
      "epoch": 0.3504009355162045,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004991968146805156,
      "loss": 5.4174,
      "mean_token_accuracy": 0.1729229286313057,
      "num_tokens": 7726549.0,
      "step": 4195
    },
    {
      "entropy": 5.44385256767273,
      "epoch": 0.35081857667891747,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004991942995461542,
      "loss": 5.5096,
      "mean_token_accuracy": 0.16654945462942122,
      "num_tokens": 7735563.0,
      "step": 4200
    },
    {
      "entropy": 5.555334234237671,
      "epoch": 0.35123621784163045,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004991917804870018,
      "loss": 5.4325,
      "mean_token_accuracy": 0.17427832037210464,
      "num_tokens": 7746144.0,
      "step": 4205
    },
    {
      "entropy": 5.482233190536499,
      "epoch": 0.3516538590043435,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004991892575031024,
      "loss": 5.4359,
      "mean_token_accuracy": 0.17850037515163422,
      "num_tokens": 7755394.0,
      "step": 4210
    },
    {
      "entropy": 5.3845151424407955,
      "epoch": 0.35207150016705646,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004991867305945001,
      "loss": 5.4009,
      "mean_token_accuracy": 0.17939516454935073,
      "num_tokens": 7764855.0,
      "step": 4215
    },
    {
      "entropy": 5.381243753433227,
      "epoch": 0.35248914132976944,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004991841997612394,
      "loss": 5.3151,
      "mean_token_accuracy": 0.16915800273418427,
      "num_tokens": 7773394.0,
      "step": 4220
    },
    {
      "entropy": 5.479357719421387,
      "epoch": 0.3529067824924825,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004991816650033643,
      "loss": 5.3801,
      "mean_token_accuracy": 0.16887461990118027,
      "num_tokens": 7783111.0,
      "step": 4225
    },
    {
      "entropy": 5.550615119934082,
      "epoch": 0.35332442365519545,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004991791263209194,
      "loss": 5.5084,
      "mean_token_accuracy": 0.17079004645347595,
      "num_tokens": 7792569.0,
      "step": 4230
    },
    {
      "entropy": 5.416995286941528,
      "epoch": 0.35374206481790843,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004991765837139491,
      "loss": 5.4008,
      "mean_token_accuracy": 0.17731060981750488,
      "num_tokens": 7801483.0,
      "step": 4235
    },
    {
      "entropy": 5.509868335723877,
      "epoch": 0.35415970598062146,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004991740371824977,
      "loss": 5.5186,
      "mean_token_accuracy": 0.1633282110095024,
      "num_tokens": 7812274.0,
      "step": 4240
    },
    {
      "entropy": 5.4453041553497314,
      "epoch": 0.35457734714333444,
      "grad_norm": 1.125,
      "learning_rate": 0.00049917148672661,
      "loss": 5.3439,
      "mean_token_accuracy": 0.17422096878290178,
      "num_tokens": 7821101.0,
      "step": 4245
    },
    {
      "entropy": 5.492163610458374,
      "epoch": 0.3549949883060474,
      "grad_norm": 1.125,
      "learning_rate": 0.0004991689323463306,
      "loss": 5.3311,
      "mean_token_accuracy": 0.1771793395280838,
      "num_tokens": 7829805.0,
      "step": 4250
    },
    {
      "entropy": 5.395871925354004,
      "epoch": 0.35541262946876045,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004991663740417042,
      "loss": 5.4292,
      "mean_token_accuracy": 0.16598358154296874,
      "num_tokens": 7839772.0,
      "step": 4255
    },
    {
      "entropy": 5.419190835952759,
      "epoch": 0.35583027063147343,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004991638118127757,
      "loss": 5.504,
      "mean_token_accuracy": 0.16540692001581192,
      "num_tokens": 7850290.0,
      "step": 4260
    },
    {
      "entropy": 5.612323188781739,
      "epoch": 0.3562479117941864,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004991612456595896,
      "loss": 5.4877,
      "mean_token_accuracy": 0.165177820622921,
      "num_tokens": 7860180.0,
      "step": 4265
    },
    {
      "entropy": 5.363939237594605,
      "epoch": 0.35666555295689945,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004991586755821913,
      "loss": 5.3392,
      "mean_token_accuracy": 0.1823292627930641,
      "num_tokens": 7869625.0,
      "step": 4270
    },
    {
      "entropy": 5.370387601852417,
      "epoch": 0.3570831941196124,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004991561015806254,
      "loss": 5.2875,
      "mean_token_accuracy": 0.1758848696947098,
      "num_tokens": 7878671.0,
      "step": 4275
    },
    {
      "entropy": 5.464903545379639,
      "epoch": 0.3575008352823254,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004991535236549372,
      "loss": 5.4066,
      "mean_token_accuracy": 0.18005224019289018,
      "num_tokens": 7887225.0,
      "step": 4280
    },
    {
      "entropy": 5.377832698822021,
      "epoch": 0.35791847644503844,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004991509418051718,
      "loss": 5.2979,
      "mean_token_accuracy": 0.18242234736680984,
      "num_tokens": 7895504.0,
      "step": 4285
    },
    {
      "entropy": 5.526500511169433,
      "epoch": 0.3583361176077514,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004991483560313743,
      "loss": 5.3679,
      "mean_token_accuracy": 0.175752092897892,
      "num_tokens": 7904875.0,
      "step": 4290
    },
    {
      "entropy": 5.467843961715698,
      "epoch": 0.3587537587704644,
      "grad_norm": 1.03125,
      "learning_rate": 0.00049914576633359,
      "loss": 5.4368,
      "mean_token_accuracy": 0.17353366315364838,
      "num_tokens": 7914747.0,
      "step": 4295
    },
    {
      "entropy": 5.422105979919434,
      "epoch": 0.3591713999331774,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004991431727118643,
      "loss": 5.3989,
      "mean_token_accuracy": 0.1721815288066864,
      "num_tokens": 7923858.0,
      "step": 4300
    },
    {
      "entropy": 5.432664155960083,
      "epoch": 0.3595890410958904,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004991405751662426,
      "loss": 5.3636,
      "mean_token_accuracy": 0.17276132702827454,
      "num_tokens": 7932396.0,
      "step": 4305
    },
    {
      "entropy": 5.454059314727783,
      "epoch": 0.3600066822586034,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004991379736967703,
      "loss": 5.3517,
      "mean_token_accuracy": 0.1802230089902878,
      "num_tokens": 7940763.0,
      "step": 4310
    },
    {
      "entropy": 5.351620006561279,
      "epoch": 0.3604243234213164,
      "grad_norm": 1.25,
      "learning_rate": 0.0004991353683034929,
      "loss": 5.3739,
      "mean_token_accuracy": 0.16648114025592803,
      "num_tokens": 7949424.0,
      "step": 4315
    },
    {
      "entropy": 5.438655281066895,
      "epoch": 0.3608419645840294,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004991327589864563,
      "loss": 5.3538,
      "mean_token_accuracy": 0.18122622966766358,
      "num_tokens": 7958805.0,
      "step": 4320
    },
    {
      "entropy": 5.5220118999481205,
      "epoch": 0.3612596057467424,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004991301457457057,
      "loss": 5.459,
      "mean_token_accuracy": 0.17593965083360671,
      "num_tokens": 7967859.0,
      "step": 4325
    },
    {
      "entropy": 5.565468502044678,
      "epoch": 0.3616772469094554,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004991275285812873,
      "loss": 5.5861,
      "mean_token_accuracy": 0.16238001137971877,
      "num_tokens": 7978357.0,
      "step": 4330
    },
    {
      "entropy": 5.5675276756286625,
      "epoch": 0.3620948880721684,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004991249074932466,
      "loss": 5.4316,
      "mean_token_accuracy": 0.17491962164640426,
      "num_tokens": 7986637.0,
      "step": 4335
    },
    {
      "entropy": 5.391841077804566,
      "epoch": 0.36251252923488136,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004991222824816296,
      "loss": 5.35,
      "mean_token_accuracy": 0.17362402230501175,
      "num_tokens": 7995954.0,
      "step": 4340
    },
    {
      "entropy": 5.500148963928223,
      "epoch": 0.3629301703975944,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004991196535464824,
      "loss": 5.4214,
      "mean_token_accuracy": 0.16786658018827438,
      "num_tokens": 8005193.0,
      "step": 4345
    },
    {
      "entropy": 5.545471715927124,
      "epoch": 0.3633478115603074,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004991170206878508,
      "loss": 5.4557,
      "mean_token_accuracy": 0.17339856028556824,
      "num_tokens": 8014419.0,
      "step": 4350
    },
    {
      "entropy": 5.448093557357788,
      "epoch": 0.36376545272302035,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000499114383905781,
      "loss": 5.3852,
      "mean_token_accuracy": 0.16827281564474106,
      "num_tokens": 8024133.0,
      "step": 4355
    },
    {
      "entropy": 5.407823801040649,
      "epoch": 0.3641830938857334,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004991117432003192,
      "loss": 5.3613,
      "mean_token_accuracy": 0.1798083484172821,
      "num_tokens": 8033059.0,
      "step": 4360
    },
    {
      "entropy": 5.4817417621612545,
      "epoch": 0.36460073504844637,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004991090985715116,
      "loss": 5.413,
      "mean_token_accuracy": 0.17707774192094802,
      "num_tokens": 8041866.0,
      "step": 4365
    },
    {
      "entropy": 5.485598945617676,
      "epoch": 0.36501837621115935,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004991064500194043,
      "loss": 5.4607,
      "mean_token_accuracy": 0.169550259411335,
      "num_tokens": 8051541.0,
      "step": 4370
    },
    {
      "entropy": 5.580405807495117,
      "epoch": 0.3654360173738724,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004991037975440439,
      "loss": 5.5277,
      "mean_token_accuracy": 0.16821610182523727,
      "num_tokens": 8061964.0,
      "step": 4375
    },
    {
      "entropy": 5.369140577316284,
      "epoch": 0.36585365853658536,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004991011411454769,
      "loss": 5.334,
      "mean_token_accuracy": 0.1683669298887253,
      "num_tokens": 8071006.0,
      "step": 4380
    },
    {
      "entropy": 5.381167030334472,
      "epoch": 0.36627129969929834,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004990984808237495,
      "loss": 5.3319,
      "mean_token_accuracy": 0.17772201001644133,
      "num_tokens": 8080104.0,
      "step": 4385
    },
    {
      "entropy": 5.47397289276123,
      "epoch": 0.36668894086201137,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004990958165789086,
      "loss": 5.3862,
      "mean_token_accuracy": 0.17768747508525848,
      "num_tokens": 8089397.0,
      "step": 4390
    },
    {
      "entropy": 5.458541488647461,
      "epoch": 0.36710658202472435,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004990931484110005,
      "loss": 5.4254,
      "mean_token_accuracy": 0.17549906820058822,
      "num_tokens": 8099055.0,
      "step": 4395
    },
    {
      "entropy": 5.469098377227783,
      "epoch": 0.3675242231874373,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004990904763200722,
      "loss": 5.3577,
      "mean_token_accuracy": 0.17544038593769073,
      "num_tokens": 8107609.0,
      "step": 4400
    },
    {
      "entropy": 5.4731487274169925,
      "epoch": 0.36794186435015036,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004990878003061705,
      "loss": 5.3848,
      "mean_token_accuracy": 0.18331958800554277,
      "num_tokens": 8116905.0,
      "step": 4405
    },
    {
      "entropy": 5.367913866043091,
      "epoch": 0.36835950551286334,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004990851203693419,
      "loss": 5.366,
      "mean_token_accuracy": 0.1775816261768341,
      "num_tokens": 8125580.0,
      "step": 4410
    },
    {
      "entropy": 5.379986810684204,
      "epoch": 0.3687771466755763,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004990824365096336,
      "loss": 5.3032,
      "mean_token_accuracy": 0.17180398106575012,
      "num_tokens": 8134568.0,
      "step": 4415
    },
    {
      "entropy": 5.334994173049926,
      "epoch": 0.36919478783828935,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004990797487270926,
      "loss": 5.2734,
      "mean_token_accuracy": 0.1837972655892372,
      "num_tokens": 8143044.0,
      "step": 4420
    },
    {
      "entropy": 5.4388011455535885,
      "epoch": 0.36961242900100233,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004990770570217659,
      "loss": 5.4079,
      "mean_token_accuracy": 0.17175349444150925,
      "num_tokens": 8153701.0,
      "step": 4425
    },
    {
      "entropy": 5.403829574584961,
      "epoch": 0.37003007016371536,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990743613937005,
      "loss": 5.3074,
      "mean_token_accuracy": 0.1803583025932312,
      "num_tokens": 8163264.0,
      "step": 4430
    },
    {
      "entropy": 5.365253925323486,
      "epoch": 0.37044771132642834,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004990716618429436,
      "loss": 5.397,
      "mean_token_accuracy": 0.1691998392343521,
      "num_tokens": 8173332.0,
      "step": 4435
    },
    {
      "entropy": 5.477272033691406,
      "epoch": 0.3708653524891413,
      "grad_norm": 1.25,
      "learning_rate": 0.0004990689583695428,
      "loss": 5.3903,
      "mean_token_accuracy": 0.17341262996196746,
      "num_tokens": 8181881.0,
      "step": 4440
    },
    {
      "entropy": 5.485300540924072,
      "epoch": 0.37128299365185435,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000499066250973545,
      "loss": 5.4363,
      "mean_token_accuracy": 0.1675601840019226,
      "num_tokens": 8190771.0,
      "step": 4445
    },
    {
      "entropy": 5.41284670829773,
      "epoch": 0.37170063481456733,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004990635396549978,
      "loss": 5.3752,
      "mean_token_accuracy": 0.169596266746521,
      "num_tokens": 8200740.0,
      "step": 4450
    },
    {
      "entropy": 5.361733341217041,
      "epoch": 0.3721182759772803,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004990608244139486,
      "loss": 5.3695,
      "mean_token_accuracy": 0.17281728684902192,
      "num_tokens": 8209935.0,
      "step": 4455
    },
    {
      "entropy": 5.480435037612915,
      "epoch": 0.37253591713999334,
      "grad_norm": 1.15625,
      "learning_rate": 0.000499058105250445,
      "loss": 5.4323,
      "mean_token_accuracy": 0.17257851213216782,
      "num_tokens": 8219137.0,
      "step": 4460
    },
    {
      "entropy": 5.5401466369628904,
      "epoch": 0.3729535583027063,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004990553821645346,
      "loss": 5.398,
      "mean_token_accuracy": 0.17586126327514648,
      "num_tokens": 8228076.0,
      "step": 4465
    },
    {
      "entropy": 5.352044868469238,
      "epoch": 0.3733711994654193,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000499052655156265,
      "loss": 5.3622,
      "mean_token_accuracy": 0.17877929359674455,
      "num_tokens": 8237581.0,
      "step": 4470
    },
    {
      "entropy": 5.456056451797485,
      "epoch": 0.37378884062813234,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004990499242256839,
      "loss": 5.4018,
      "mean_token_accuracy": 0.16744532585144042,
      "num_tokens": 8247384.0,
      "step": 4475
    },
    {
      "entropy": 5.5242781162261965,
      "epoch": 0.3742064817908453,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004990471893728394,
      "loss": 5.4786,
      "mean_token_accuracy": 0.17573148906230926,
      "num_tokens": 8256383.0,
      "step": 4480
    },
    {
      "entropy": 5.433877658843994,
      "epoch": 0.3746241229535583,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000499044450597779,
      "loss": 5.3368,
      "mean_token_accuracy": 0.1774474561214447,
      "num_tokens": 8264848.0,
      "step": 4485
    },
    {
      "entropy": 5.404270029067993,
      "epoch": 0.3750417641162713,
      "grad_norm": 1.125,
      "learning_rate": 0.0004990417079005508,
      "loss": 5.3714,
      "mean_token_accuracy": 0.17153847068548203,
      "num_tokens": 8274403.0,
      "step": 4490
    },
    {
      "entropy": 5.402362918853759,
      "epoch": 0.3754594052789843,
      "grad_norm": 1.21875,
      "learning_rate": 0.000499038961281203,
      "loss": 5.2962,
      "mean_token_accuracy": 0.18346159905195236,
      "num_tokens": 8282700.0,
      "step": 4495
    },
    {
      "entropy": 5.384265756607055,
      "epoch": 0.3758770464416973,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004990362107397834,
      "loss": 5.3261,
      "mean_token_accuracy": 0.18079284429550171,
      "num_tokens": 8291718.0,
      "step": 4500
    },
    {
      "entropy": 5.3855781078338625,
      "epoch": 0.3762946876044103,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004990334562763401,
      "loss": 5.3805,
      "mean_token_accuracy": 0.1737850561738014,
      "num_tokens": 8301113.0,
      "step": 4505
    },
    {
      "entropy": 5.481485652923584,
      "epoch": 0.3767123287671233,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004990306978909216,
      "loss": 5.4836,
      "mean_token_accuracy": 0.17322600334882737,
      "num_tokens": 8310357.0,
      "step": 4510
    },
    {
      "entropy": 5.519826745986938,
      "epoch": 0.3771299699298363,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004990279355835761,
      "loss": 5.4215,
      "mean_token_accuracy": 0.17298668026924133,
      "num_tokens": 8319680.0,
      "step": 4515
    },
    {
      "entropy": 5.442970657348633,
      "epoch": 0.3775476110925493,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004990251693543519,
      "loss": 5.4149,
      "mean_token_accuracy": 0.17046769857406616,
      "num_tokens": 8329127.0,
      "step": 4520
    },
    {
      "entropy": 5.341222906112671,
      "epoch": 0.3779652522552623,
      "grad_norm": 1.25,
      "learning_rate": 0.0004990223992032975,
      "loss": 5.2696,
      "mean_token_accuracy": 0.17672463953495027,
      "num_tokens": 8338308.0,
      "step": 4525
    },
    {
      "entropy": 5.412484121322632,
      "epoch": 0.37838289341797526,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004990196251304612,
      "loss": 5.3117,
      "mean_token_accuracy": 0.17839513421058656,
      "num_tokens": 8347491.0,
      "step": 4530
    },
    {
      "entropy": 5.337620973587036,
      "epoch": 0.3788005345806883,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000499016847135892,
      "loss": 5.3287,
      "mean_token_accuracy": 0.17211673855781556,
      "num_tokens": 8357112.0,
      "step": 4535
    },
    {
      "entropy": 5.385684013366699,
      "epoch": 0.3792181757434013,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000499014065219638,
      "loss": 5.3142,
      "mean_token_accuracy": 0.1788911998271942,
      "num_tokens": 8366285.0,
      "step": 4540
    },
    {
      "entropy": 5.390375089645386,
      "epoch": 0.37963581690611425,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004990112793817481,
      "loss": 5.2255,
      "mean_token_accuracy": 0.1836492419242859,
      "num_tokens": 8374362.0,
      "step": 4545
    },
    {
      "entropy": 5.383315944671631,
      "epoch": 0.3800534580688273,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004990084896222713,
      "loss": 5.4595,
      "mean_token_accuracy": 0.17287468016147614,
      "num_tokens": 8383297.0,
      "step": 4550
    },
    {
      "entropy": 5.403417682647705,
      "epoch": 0.38047109923154027,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004990056959412562,
      "loss": 5.3412,
      "mean_token_accuracy": 0.18178841322660447,
      "num_tokens": 8391926.0,
      "step": 4555
    },
    {
      "entropy": 5.517784595489502,
      "epoch": 0.38088874039425324,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004990028983387517,
      "loss": 5.3183,
      "mean_token_accuracy": 0.175078184902668,
      "num_tokens": 8401226.0,
      "step": 4560
    },
    {
      "entropy": 5.393605709075928,
      "epoch": 0.3813063815569663,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004990000968148068,
      "loss": 5.3501,
      "mean_token_accuracy": 0.17882849872112275,
      "num_tokens": 8409966.0,
      "step": 4565
    },
    {
      "entropy": 5.295736312866211,
      "epoch": 0.38172402271967926,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004989972913694707,
      "loss": 5.2016,
      "mean_token_accuracy": 0.1859448716044426,
      "num_tokens": 8418281.0,
      "step": 4570
    },
    {
      "entropy": 5.2785608768463135,
      "epoch": 0.38214166388239224,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004989944820027923,
      "loss": 5.32,
      "mean_token_accuracy": 0.18151139318943024,
      "num_tokens": 8427436.0,
      "step": 4575
    },
    {
      "entropy": 5.460799217224121,
      "epoch": 0.38255930504510527,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004989916687148209,
      "loss": 5.3463,
      "mean_token_accuracy": 0.17874803096055986,
      "num_tokens": 8436066.0,
      "step": 4580
    },
    {
      "entropy": 5.419190311431885,
      "epoch": 0.38297694620781825,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004989888515056057,
      "loss": 5.3677,
      "mean_token_accuracy": 0.17505231499671936,
      "num_tokens": 8445572.0,
      "step": 4585
    },
    {
      "entropy": 5.444575595855713,
      "epoch": 0.3833945873705312,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000498986030375196,
      "loss": 5.3963,
      "mean_token_accuracy": 0.17158419489860535,
      "num_tokens": 8454781.0,
      "step": 4590
    },
    {
      "entropy": 5.506966257095337,
      "epoch": 0.38381222853324426,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004989832053236413,
      "loss": 5.4777,
      "mean_token_accuracy": 0.16998900324106217,
      "num_tokens": 8463239.0,
      "step": 4595
    },
    {
      "entropy": 5.521318006515503,
      "epoch": 0.38422986969595724,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000498980376350991,
      "loss": 5.3731,
      "mean_token_accuracy": 0.17278695106506348,
      "num_tokens": 8472664.0,
      "step": 4600
    },
    {
      "entropy": 5.39429235458374,
      "epoch": 0.3846475108586702,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004989775434572945,
      "loss": 5.4378,
      "mean_token_accuracy": 0.17073088139295578,
      "num_tokens": 8483149.0,
      "step": 4605
    },
    {
      "entropy": 5.548193073272705,
      "epoch": 0.38506515202138325,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004989747066426015,
      "loss": 5.3843,
      "mean_token_accuracy": 0.17684950828552246,
      "num_tokens": 8492297.0,
      "step": 4610
    },
    {
      "entropy": 5.465753412246704,
      "epoch": 0.38548279318409623,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004989718659069618,
      "loss": 5.4503,
      "mean_token_accuracy": 0.171021269261837,
      "num_tokens": 8502092.0,
      "step": 4615
    },
    {
      "entropy": 5.4493889808654785,
      "epoch": 0.3859004343468092,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000498969021250425,
      "loss": 5.361,
      "mean_token_accuracy": 0.17595359086990356,
      "num_tokens": 8510816.0,
      "step": 4620
    },
    {
      "entropy": 5.475935745239258,
      "epoch": 0.38631807550952224,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004989661726730407,
      "loss": 5.3667,
      "mean_token_accuracy": 0.17920992225408555,
      "num_tokens": 8520055.0,
      "step": 4625
    },
    {
      "entropy": 5.4450061321258545,
      "epoch": 0.3867357166722352,
      "grad_norm": 1.125,
      "learning_rate": 0.0004989633201748591,
      "loss": 5.3865,
      "mean_token_accuracy": 0.1788432076573372,
      "num_tokens": 8528271.0,
      "step": 4630
    },
    {
      "entropy": 5.379427671432495,
      "epoch": 0.3871533578349482,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004989604637559299,
      "loss": 5.2925,
      "mean_token_accuracy": 0.18760859370231628,
      "num_tokens": 8537360.0,
      "step": 4635
    },
    {
      "entropy": 5.320774173736572,
      "epoch": 0.38757099899766123,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004989576034163033,
      "loss": 5.2445,
      "mean_token_accuracy": 0.1910588413476944,
      "num_tokens": 8546574.0,
      "step": 4640
    },
    {
      "entropy": 5.38830828666687,
      "epoch": 0.3879886401603742,
      "grad_norm": 1.125,
      "learning_rate": 0.0004989547391560293,
      "loss": 5.3569,
      "mean_token_accuracy": 0.17621755599975586,
      "num_tokens": 8555581.0,
      "step": 4645
    },
    {
      "entropy": 5.429186201095581,
      "epoch": 0.3884062813230872,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000498951870975158,
      "loss": 5.3921,
      "mean_token_accuracy": 0.18038074523210526,
      "num_tokens": 8564539.0,
      "step": 4650
    },
    {
      "entropy": 5.448959732055664,
      "epoch": 0.3888239224858002,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004989489988737395,
      "loss": 5.4328,
      "mean_token_accuracy": 0.1747396320104599,
      "num_tokens": 8573804.0,
      "step": 4655
    },
    {
      "entropy": 5.450884819030762,
      "epoch": 0.3892415636485132,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004989461228518243,
      "loss": 5.3603,
      "mean_token_accuracy": 0.1754891902208328,
      "num_tokens": 8583262.0,
      "step": 4660
    },
    {
      "entropy": 5.43102765083313,
      "epoch": 0.3896592048112262,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004989432429094627,
      "loss": 5.4228,
      "mean_token_accuracy": 0.1750425562262535,
      "num_tokens": 8592763.0,
      "step": 4665
    },
    {
      "entropy": 5.493407201766968,
      "epoch": 0.3900768459739392,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004989403590467049,
      "loss": 5.2917,
      "mean_token_accuracy": 0.178767366707325,
      "num_tokens": 8601884.0,
      "step": 4670
    },
    {
      "entropy": 5.363888597488403,
      "epoch": 0.3904944871366522,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004989374712636017,
      "loss": 5.4072,
      "mean_token_accuracy": 0.1706961750984192,
      "num_tokens": 8611207.0,
      "step": 4675
    },
    {
      "entropy": 5.304356241226197,
      "epoch": 0.39091212829936517,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004989345795602036,
      "loss": 5.1684,
      "mean_token_accuracy": 0.19175629168748856,
      "num_tokens": 8620104.0,
      "step": 4680
    },
    {
      "entropy": 5.444139051437378,
      "epoch": 0.3913297694620782,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004989316839365609,
      "loss": 5.3749,
      "mean_token_accuracy": 0.17283373326063156,
      "num_tokens": 8629552.0,
      "step": 4685
    },
    {
      "entropy": 5.383452367782593,
      "epoch": 0.3917474106247912,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004989287843927246,
      "loss": 5.3047,
      "mean_token_accuracy": 0.17732003927230836,
      "num_tokens": 8638961.0,
      "step": 4690
    },
    {
      "entropy": 5.368448734283447,
      "epoch": 0.39216505178750416,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004989258809287455,
      "loss": 5.3312,
      "mean_token_accuracy": 0.17787770181894302,
      "num_tokens": 8648051.0,
      "step": 4695
    },
    {
      "entropy": 5.459696626663208,
      "epoch": 0.3925826929502172,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004989229735446743,
      "loss": 5.3316,
      "mean_token_accuracy": 0.17427610903978347,
      "num_tokens": 8656863.0,
      "step": 4700
    },
    {
      "entropy": 5.235874509811401,
      "epoch": 0.39300033411293017,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004989200622405619,
      "loss": 5.162,
      "mean_token_accuracy": 0.190497462451458,
      "num_tokens": 8665455.0,
      "step": 4705
    },
    {
      "entropy": 5.341135787963867,
      "epoch": 0.39341797527564315,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004989171470164592,
      "loss": 5.2642,
      "mean_token_accuracy": 0.18486505299806594,
      "num_tokens": 8674119.0,
      "step": 4710
    },
    {
      "entropy": 5.447092962265015,
      "epoch": 0.3938356164383562,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004989142278724173,
      "loss": 5.3712,
      "mean_token_accuracy": 0.17342129051685334,
      "num_tokens": 8682979.0,
      "step": 4715
    },
    {
      "entropy": 5.436543989181518,
      "epoch": 0.39425325760106916,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004989113048084874,
      "loss": 5.2756,
      "mean_token_accuracy": 0.1758004441857338,
      "num_tokens": 8692392.0,
      "step": 4720
    },
    {
      "entropy": 5.3052526950836185,
      "epoch": 0.39467089876378214,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004989083778247206,
      "loss": 5.3185,
      "mean_token_accuracy": 0.18246010392904283,
      "num_tokens": 8700932.0,
      "step": 4725
    },
    {
      "entropy": 5.563783454895019,
      "epoch": 0.3950885399264952,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000498905446921168,
      "loss": 5.4978,
      "mean_token_accuracy": 0.16393214464187622,
      "num_tokens": 8710683.0,
      "step": 4730
    },
    {
      "entropy": 5.347655344009399,
      "epoch": 0.39550618108920815,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004989025120978812,
      "loss": 5.327,
      "mean_token_accuracy": 0.1803911566734314,
      "num_tokens": 8719939.0,
      "step": 4735
    },
    {
      "entropy": 5.438155937194824,
      "epoch": 0.39592382225192113,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004988995733549113,
      "loss": 5.3816,
      "mean_token_accuracy": 0.1774156004190445,
      "num_tokens": 8729275.0,
      "step": 4740
    },
    {
      "entropy": 5.342011165618897,
      "epoch": 0.39634146341463417,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004988966306923099,
      "loss": 5.2584,
      "mean_token_accuracy": 0.18361410945653917,
      "num_tokens": 8738048.0,
      "step": 4745
    },
    {
      "entropy": 5.422035169601441,
      "epoch": 0.39675910457734714,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004988936841101284,
      "loss": 5.3187,
      "mean_token_accuracy": 0.1750010848045349,
      "num_tokens": 8746967.0,
      "step": 4750
    },
    {
      "entropy": 5.353234767913818,
      "epoch": 0.3971767457400601,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004988907336084186,
      "loss": 5.2678,
      "mean_token_accuracy": 0.18189094662666322,
      "num_tokens": 8754908.0,
      "step": 4755
    },
    {
      "entropy": 5.369625997543335,
      "epoch": 0.39759438690277316,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004988877791872319,
      "loss": 5.3166,
      "mean_token_accuracy": 0.18776927441358565,
      "num_tokens": 8764489.0,
      "step": 4760
    },
    {
      "entropy": 5.504715442657471,
      "epoch": 0.39801202806548613,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004988848208466202,
      "loss": 5.4869,
      "mean_token_accuracy": 0.15817637890577316,
      "num_tokens": 8774365.0,
      "step": 4765
    },
    {
      "entropy": 5.444577646255493,
      "epoch": 0.3984296692281991,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004988818585866351,
      "loss": 5.2834,
      "mean_token_accuracy": 0.17653939872980118,
      "num_tokens": 8783538.0,
      "step": 4770
    },
    {
      "entropy": 5.358550214767456,
      "epoch": 0.39884731039091215,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004988788924073288,
      "loss": 5.3763,
      "mean_token_accuracy": 0.17767149358987808,
      "num_tokens": 8792171.0,
      "step": 4775
    },
    {
      "entropy": 5.404060745239258,
      "epoch": 0.3992649515536251,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004988759223087528,
      "loss": 5.3828,
      "mean_token_accuracy": 0.17320301830768586,
      "num_tokens": 8801325.0,
      "step": 4780
    },
    {
      "entropy": 5.445061063766479,
      "epoch": 0.3996825927163381,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004988729482909595,
      "loss": 5.3207,
      "mean_token_accuracy": 0.18149494826793672,
      "num_tokens": 8810566.0,
      "step": 4785
    },
    {
      "entropy": 5.3350725173950195,
      "epoch": 0.40010023387905114,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004988699703540006,
      "loss": 5.2704,
      "mean_token_accuracy": 0.18389502614736558,
      "num_tokens": 8819735.0,
      "step": 4790
    },
    {
      "entropy": 5.356999015808105,
      "epoch": 0.4005178750417641,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004988669884979284,
      "loss": 5.3955,
      "mean_token_accuracy": 0.17396778911352156,
      "num_tokens": 8828820.0,
      "step": 4795
    },
    {
      "entropy": 5.423072004318238,
      "epoch": 0.4009355162044771,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004988640027227951,
      "loss": 5.2685,
      "mean_token_accuracy": 0.17254929840564728,
      "num_tokens": 8837858.0,
      "step": 4800
    },
    {
      "entropy": 5.387976026535034,
      "epoch": 0.4013531573671901,
      "grad_norm": 1.296875,
      "learning_rate": 0.000498861013028653,
      "loss": 5.255,
      "mean_token_accuracy": 0.18501665890216829,
      "num_tokens": 8847172.0,
      "step": 4805
    },
    {
      "entropy": 5.371137475967407,
      "epoch": 0.4017707985299031,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004988580194155544,
      "loss": 5.3041,
      "mean_token_accuracy": 0.17536961734294892,
      "num_tokens": 8856700.0,
      "step": 4810
    },
    {
      "entropy": 5.294289445877075,
      "epoch": 0.4021884396926161,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004988550218835517,
      "loss": 5.1902,
      "mean_token_accuracy": 0.18726896494627,
      "num_tokens": 8866108.0,
      "step": 4815
    },
    {
      "entropy": 5.357527589797973,
      "epoch": 0.4026060808553291,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004988520204326975,
      "loss": 5.2898,
      "mean_token_accuracy": 0.18764663934707643,
      "num_tokens": 8874833.0,
      "step": 4820
    },
    {
      "entropy": 5.440208768844604,
      "epoch": 0.4030237220180421,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004988490150630441,
      "loss": 5.3687,
      "mean_token_accuracy": 0.17777276188135147,
      "num_tokens": 8884211.0,
      "step": 4825
    },
    {
      "entropy": 5.470197916030884,
      "epoch": 0.4034413631807551,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004988460057746443,
      "loss": 5.4681,
      "mean_token_accuracy": 0.17849488854408263,
      "num_tokens": 8893443.0,
      "step": 4830
    },
    {
      "entropy": 5.328664779663086,
      "epoch": 0.4038590043434681,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004988429925675507,
      "loss": 5.3296,
      "mean_token_accuracy": 0.17989784330129624,
      "num_tokens": 8903544.0,
      "step": 4835
    },
    {
      "entropy": 5.440176105499267,
      "epoch": 0.4042766455061811,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004988399754418161,
      "loss": 5.3234,
      "mean_token_accuracy": 0.17989243417978287,
      "num_tokens": 8912270.0,
      "step": 4840
    },
    {
      "entropy": 5.539068412780762,
      "epoch": 0.40469428666889407,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004988369543974932,
      "loss": 5.4391,
      "mean_token_accuracy": 0.17070268467068672,
      "num_tokens": 8922155.0,
      "step": 4845
    },
    {
      "entropy": 5.381599950790405,
      "epoch": 0.4051119278316071,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004988339294346351,
      "loss": 5.4313,
      "mean_token_accuracy": 0.17393961548805237,
      "num_tokens": 8931844.0,
      "step": 4850
    },
    {
      "entropy": 5.4250129699707035,
      "epoch": 0.4055295689943201,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004988309005532946,
      "loss": 5.2831,
      "mean_token_accuracy": 0.18272562175989152,
      "num_tokens": 8940874.0,
      "step": 4855
    },
    {
      "entropy": 5.370961380004883,
      "epoch": 0.40594721015703306,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004988278677535247,
      "loss": 5.3523,
      "mean_token_accuracy": 0.18261157870292663,
      "num_tokens": 8950108.0,
      "step": 4860
    },
    {
      "entropy": 5.42987847328186,
      "epoch": 0.4063648513197461,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004988248310353788,
      "loss": 5.3788,
      "mean_token_accuracy": 0.17099629342556,
      "num_tokens": 8960610.0,
      "step": 4865
    },
    {
      "entropy": 5.376508474349976,
      "epoch": 0.40678249248245907,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004988217903989095,
      "loss": 5.3048,
      "mean_token_accuracy": 0.17501236051321029,
      "num_tokens": 8969853.0,
      "step": 4870
    },
    {
      "entropy": 5.462110900878907,
      "epoch": 0.40720013364517205,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004988187458441704,
      "loss": 5.2908,
      "mean_token_accuracy": 0.17498612254858018,
      "num_tokens": 8979800.0,
      "step": 4875
    },
    {
      "entropy": 5.426138687133789,
      "epoch": 0.4076177748078851,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004988156973712149,
      "loss": 5.3455,
      "mean_token_accuracy": 0.17958766967058182,
      "num_tokens": 8988667.0,
      "step": 4880
    },
    {
      "entropy": 5.289855289459228,
      "epoch": 0.40803541597059806,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004988126449800961,
      "loss": 5.2302,
      "mean_token_accuracy": 0.1811926245689392,
      "num_tokens": 8997385.0,
      "step": 4885
    },
    {
      "entropy": 5.332584190368652,
      "epoch": 0.40845305713331104,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004988095886708676,
      "loss": 5.2574,
      "mean_token_accuracy": 0.18289996385574342,
      "num_tokens": 9005999.0,
      "step": 4890
    },
    {
      "entropy": 5.475919389724732,
      "epoch": 0.40887069829602407,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004988065284435828,
      "loss": 5.376,
      "mean_token_accuracy": 0.18217089772224426,
      "num_tokens": 9014837.0,
      "step": 4895
    },
    {
      "entropy": 5.474930286407471,
      "epoch": 0.40928833945873705,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004988034642982954,
      "loss": 5.4324,
      "mean_token_accuracy": 0.1781845659017563,
      "num_tokens": 9023956.0,
      "step": 4900
    },
    {
      "entropy": 5.425081396102906,
      "epoch": 0.40970598062145,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004988003962350589,
      "loss": 5.38,
      "mean_token_accuracy": 0.17241764217615127,
      "num_tokens": 9032945.0,
      "step": 4905
    },
    {
      "entropy": 5.421437549591064,
      "epoch": 0.41012362178416306,
      "grad_norm": 1.1875,
      "learning_rate": 0.000498797324253927,
      "loss": 5.3073,
      "mean_token_accuracy": 0.18210716545581818,
      "num_tokens": 9042109.0,
      "step": 4910
    },
    {
      "entropy": 5.321787977218628,
      "epoch": 0.41054126294687604,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004987942483549537,
      "loss": 5.2942,
      "mean_token_accuracy": 0.17567227482795716,
      "num_tokens": 9052210.0,
      "step": 4915
    },
    {
      "entropy": 5.35203971862793,
      "epoch": 0.410958904109589,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004987911685381926,
      "loss": 5.2493,
      "mean_token_accuracy": 0.18420227468013764,
      "num_tokens": 9060089.0,
      "step": 4920
    },
    {
      "entropy": 5.4903594017028805,
      "epoch": 0.41137654527230205,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004987880848036977,
      "loss": 5.3719,
      "mean_token_accuracy": 0.1727112799882889,
      "num_tokens": 9069235.0,
      "step": 4925
    },
    {
      "entropy": 5.364435195922852,
      "epoch": 0.41179418643501503,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004987849971515232,
      "loss": 5.316,
      "mean_token_accuracy": 0.18092474788427354,
      "num_tokens": 9077509.0,
      "step": 4930
    },
    {
      "entropy": 5.3612274646759035,
      "epoch": 0.412211827597728,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004987819055817227,
      "loss": 5.2291,
      "mean_token_accuracy": 0.19235456585884095,
      "num_tokens": 9085493.0,
      "step": 4935
    },
    {
      "entropy": 5.198392677307129,
      "epoch": 0.41262946876044104,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004987788100943507,
      "loss": 5.1979,
      "mean_token_accuracy": 0.18430102318525315,
      "num_tokens": 9094748.0,
      "step": 4940
    },
    {
      "entropy": 5.369553852081299,
      "epoch": 0.413047109923154,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004987757106894613,
      "loss": 5.2769,
      "mean_token_accuracy": 0.18725120276212692,
      "num_tokens": 9103678.0,
      "step": 4945
    },
    {
      "entropy": 5.419416522979736,
      "epoch": 0.413464751085867,
      "grad_norm": 1.25,
      "learning_rate": 0.0004987726073671086,
      "loss": 5.3523,
      "mean_token_accuracy": 0.1809053748846054,
      "num_tokens": 9112851.0,
      "step": 4950
    },
    {
      "entropy": 5.473608207702637,
      "epoch": 0.41388239224858003,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004987695001273471,
      "loss": 5.5562,
      "mean_token_accuracy": 0.16060083359479904,
      "num_tokens": 9122106.0,
      "step": 4955
    },
    {
      "entropy": 5.39730806350708,
      "epoch": 0.414300033411293,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004987663889702313,
      "loss": 5.2691,
      "mean_token_accuracy": 0.1870732367038727,
      "num_tokens": 9131580.0,
      "step": 4960
    },
    {
      "entropy": 5.452555704116821,
      "epoch": 0.414717674574006,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004987632738958155,
      "loss": 5.4085,
      "mean_token_accuracy": 0.17870815396308898,
      "num_tokens": 9141642.0,
      "step": 4965
    },
    {
      "entropy": 5.485944128036499,
      "epoch": 0.415135315736719,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004987601549041541,
      "loss": 5.4501,
      "mean_token_accuracy": 0.16735972315073014,
      "num_tokens": 9151192.0,
      "step": 4970
    },
    {
      "entropy": 5.405943584442139,
      "epoch": 0.415552956899432,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004987570319953018,
      "loss": 5.3474,
      "mean_token_accuracy": 0.17551329880952835,
      "num_tokens": 9161316.0,
      "step": 4975
    },
    {
      "entropy": 5.416285467147827,
      "epoch": 0.415970598062145,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004987539051693134,
      "loss": 5.3477,
      "mean_token_accuracy": 0.1752282813191414,
      "num_tokens": 9169535.0,
      "step": 4980
    },
    {
      "entropy": 5.4347559928894045,
      "epoch": 0.416388239224858,
      "grad_norm": 1.25,
      "learning_rate": 0.0004987507744262436,
      "loss": 5.3748,
      "mean_token_accuracy": 0.17112898379564284,
      "num_tokens": 9178187.0,
      "step": 4985
    },
    {
      "entropy": 5.336753177642822,
      "epoch": 0.416805880387571,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004987476397661472,
      "loss": 5.2621,
      "mean_token_accuracy": 0.17952335774898528,
      "num_tokens": 9186342.0,
      "step": 4990
    },
    {
      "entropy": 5.433024072647095,
      "epoch": 0.41722352155028397,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004987445011890789,
      "loss": 5.3902,
      "mean_token_accuracy": 0.17493401765823363,
      "num_tokens": 9196470.0,
      "step": 4995
    },
    {
      "entropy": 5.40567421913147,
      "epoch": 0.417641162712997,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000498741358695094,
      "loss": 5.3013,
      "mean_token_accuracy": 0.1819268971681595,
      "num_tokens": 9206587.0,
      "step": 5000
    },
    {
      "entropy": 5.351897239685059,
      "epoch": 0.41805880387571,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000498738212284247,
      "loss": 5.3097,
      "mean_token_accuracy": 0.17583694010972978,
      "num_tokens": 9216459.0,
      "step": 5005
    },
    {
      "entropy": 5.543639469146728,
      "epoch": 0.41847644503842296,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004987350619565934,
      "loss": 5.4109,
      "mean_token_accuracy": 0.173588390648365,
      "num_tokens": 9227301.0,
      "step": 5010
    },
    {
      "entropy": 5.440727043151855,
      "epoch": 0.418894086201136,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004987319077121884,
      "loss": 5.3436,
      "mean_token_accuracy": 0.1722140923142433,
      "num_tokens": 9236577.0,
      "step": 5015
    },
    {
      "entropy": 5.341638374328613,
      "epoch": 0.419311727363849,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004987287495510869,
      "loss": 5.3458,
      "mean_token_accuracy": 0.18006064742803574,
      "num_tokens": 9246451.0,
      "step": 5020
    },
    {
      "entropy": 5.386490249633789,
      "epoch": 0.41972936852656195,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004987255874733444,
      "loss": 5.2222,
      "mean_token_accuracy": 0.17657777816057205,
      "num_tokens": 9256108.0,
      "step": 5025
    },
    {
      "entropy": 5.293013095855713,
      "epoch": 0.420147009689275,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004987224214790161,
      "loss": 5.2966,
      "mean_token_accuracy": 0.1810378149151802,
      "num_tokens": 9265173.0,
      "step": 5030
    },
    {
      "entropy": 5.3749912738800045,
      "epoch": 0.42056465085198796,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004987192515681575,
      "loss": 5.3944,
      "mean_token_accuracy": 0.1769643172621727,
      "num_tokens": 9273979.0,
      "step": 5035
    },
    {
      "entropy": 5.359067821502686,
      "epoch": 0.42098229201470094,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004987160777408243,
      "loss": 5.3176,
      "mean_token_accuracy": 0.1761663630604744,
      "num_tokens": 9281893.0,
      "step": 5040
    },
    {
      "entropy": 5.4864922046661375,
      "epoch": 0.421399933177414,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004987128999970717,
      "loss": 5.4408,
      "mean_token_accuracy": 0.17744186595082284,
      "num_tokens": 9290799.0,
      "step": 5045
    },
    {
      "entropy": 5.495283555984497,
      "epoch": 0.42181757434012696,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004987097183369555,
      "loss": 5.4117,
      "mean_token_accuracy": 0.17196936160326004,
      "num_tokens": 9300562.0,
      "step": 5050
    },
    {
      "entropy": 5.305688667297363,
      "epoch": 0.42223521550283993,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004987065327605315,
      "loss": 5.1819,
      "mean_token_accuracy": 0.18683690279722215,
      "num_tokens": 9308889.0,
      "step": 5055
    },
    {
      "entropy": 5.289650535583496,
      "epoch": 0.42265285666555297,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004987033432678553,
      "loss": 5.2763,
      "mean_token_accuracy": 0.17782160639762878,
      "num_tokens": 9317957.0,
      "step": 5060
    },
    {
      "entropy": 5.473822784423828,
      "epoch": 0.42307049782826595,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004987001498589828,
      "loss": 5.4711,
      "mean_token_accuracy": 0.16878805607557296,
      "num_tokens": 9327849.0,
      "step": 5065
    },
    {
      "entropy": 5.426246404647827,
      "epoch": 0.4234881389909789,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00049869695253397,
      "loss": 5.3342,
      "mean_token_accuracy": 0.17408983856439592,
      "num_tokens": 9337165.0,
      "step": 5070
    },
    {
      "entropy": 5.373740291595459,
      "epoch": 0.42390578015369196,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004986937512928727,
      "loss": 5.2397,
      "mean_token_accuracy": 0.18019959777593614,
      "num_tokens": 9346817.0,
      "step": 5075
    },
    {
      "entropy": 5.473091650009155,
      "epoch": 0.42432342131640494,
      "grad_norm": 1.171875,
      "learning_rate": 0.000498690546135747,
      "loss": 5.365,
      "mean_token_accuracy": 0.1764334335923195,
      "num_tokens": 9356420.0,
      "step": 5080
    },
    {
      "entropy": 5.322289037704468,
      "epoch": 0.4247410624791179,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004986873370626491,
      "loss": 5.287,
      "mean_token_accuracy": 0.18042603135108948,
      "num_tokens": 9365580.0,
      "step": 5085
    },
    {
      "entropy": 5.360691070556641,
      "epoch": 0.42515870364183095,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000498684124073635,
      "loss": 5.3531,
      "mean_token_accuracy": 0.17079622596502303,
      "num_tokens": 9375116.0,
      "step": 5090
    },
    {
      "entropy": 5.448138332366943,
      "epoch": 0.4255763448045439,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004986809071687612,
      "loss": 5.3255,
      "mean_token_accuracy": 0.17626895606517792,
      "num_tokens": 9384129.0,
      "step": 5095
    },
    {
      "entropy": 5.353268814086914,
      "epoch": 0.4259939859672569,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004986776863480838,
      "loss": 5.2626,
      "mean_token_accuracy": 0.17728972136974336,
      "num_tokens": 9392757.0,
      "step": 5100
    },
    {
      "entropy": 5.281383848190307,
      "epoch": 0.42641162712996994,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004986744616116593,
      "loss": 5.2399,
      "mean_token_accuracy": 0.18091434091329575,
      "num_tokens": 9401753.0,
      "step": 5105
    },
    {
      "entropy": 5.399411725997925,
      "epoch": 0.4268292682926829,
      "grad_norm": 1.21875,
      "learning_rate": 0.000498671232959544,
      "loss": 5.4345,
      "mean_token_accuracy": 0.1719147503376007,
      "num_tokens": 9411400.0,
      "step": 5110
    },
    {
      "entropy": 5.466083431243897,
      "epoch": 0.42724690945539595,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004986680003917947,
      "loss": 5.2904,
      "mean_token_accuracy": 0.18127259761095046,
      "num_tokens": 9420206.0,
      "step": 5115
    },
    {
      "entropy": 5.359201669692993,
      "epoch": 0.42766455061810893,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004986647639084677,
      "loss": 5.2895,
      "mean_token_accuracy": 0.1818522483110428,
      "num_tokens": 9428418.0,
      "step": 5120
    },
    {
      "entropy": 5.450143098831177,
      "epoch": 0.4280821917808219,
      "grad_norm": 1.234375,
      "learning_rate": 0.00049866152350962,
      "loss": 5.3298,
      "mean_token_accuracy": 0.17638176530599595,
      "num_tokens": 9437624.0,
      "step": 5125
    },
    {
      "entropy": 5.328964281082153,
      "epoch": 0.42849983294353494,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004986582791953078,
      "loss": 5.2455,
      "mean_token_accuracy": 0.18318878710269929,
      "num_tokens": 9446857.0,
      "step": 5130
    },
    {
      "entropy": 5.368569469451904,
      "epoch": 0.4289174741062479,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004986550309655884,
      "loss": 5.3666,
      "mean_token_accuracy": 0.1741127997636795,
      "num_tokens": 9456024.0,
      "step": 5135
    },
    {
      "entropy": 5.385288286209106,
      "epoch": 0.4293351152689609,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004986517788205184,
      "loss": 5.2471,
      "mean_token_accuracy": 0.18674221485853196,
      "num_tokens": 9464018.0,
      "step": 5140
    },
    {
      "entropy": 5.322151327133179,
      "epoch": 0.42975275643167393,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004986485227601548,
      "loss": 5.3711,
      "mean_token_accuracy": 0.18010206669569015,
      "num_tokens": 9472430.0,
      "step": 5145
    },
    {
      "entropy": 5.3633181095123295,
      "epoch": 0.4301703975943869,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004986452627845547,
      "loss": 5.2925,
      "mean_token_accuracy": 0.1861819475889206,
      "num_tokens": 9480592.0,
      "step": 5150
    },
    {
      "entropy": 5.374464511871338,
      "epoch": 0.4305880387570999,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000498641998893775,
      "loss": 5.2301,
      "mean_token_accuracy": 0.1861698091030121,
      "num_tokens": 9489162.0,
      "step": 5155
    },
    {
      "entropy": 5.34882435798645,
      "epoch": 0.4310056799198129,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004986387310878728,
      "loss": 5.3057,
      "mean_token_accuracy": 0.18348312079906465,
      "num_tokens": 9499210.0,
      "step": 5160
    },
    {
      "entropy": 5.419818878173828,
      "epoch": 0.4314233210825259,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004986354593669056,
      "loss": 5.342,
      "mean_token_accuracy": 0.18277039229869843,
      "num_tokens": 9508062.0,
      "step": 5165
    },
    {
      "entropy": 5.442840671539306,
      "epoch": 0.4318409622452389,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004986321837309304,
      "loss": 5.2918,
      "mean_token_accuracy": 0.18474480360746384,
      "num_tokens": 9516756.0,
      "step": 5170
    },
    {
      "entropy": 5.345137739181519,
      "epoch": 0.4322586034079519,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004986289041800046,
      "loss": 5.2826,
      "mean_token_accuracy": 0.18207624405622483,
      "num_tokens": 9525305.0,
      "step": 5175
    },
    {
      "entropy": 5.369629335403443,
      "epoch": 0.4326762445706649,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004986256207141857,
      "loss": 5.2597,
      "mean_token_accuracy": 0.18715112209320067,
      "num_tokens": 9534310.0,
      "step": 5180
    },
    {
      "entropy": 5.473608112335205,
      "epoch": 0.43309388573337787,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004986223333335311,
      "loss": 5.3746,
      "mean_token_accuracy": 0.17821064591407776,
      "num_tokens": 9543648.0,
      "step": 5185
    },
    {
      "entropy": 5.32420244216919,
      "epoch": 0.4335115268960909,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004986190420380985,
      "loss": 5.2916,
      "mean_token_accuracy": 0.17708334922790528,
      "num_tokens": 9553081.0,
      "step": 5190
    },
    {
      "entropy": 5.454599094390869,
      "epoch": 0.4339291680588039,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004986157468279453,
      "loss": 5.3266,
      "mean_token_accuracy": 0.17734083980321885,
      "num_tokens": 9563184.0,
      "step": 5195
    },
    {
      "entropy": 5.3927582740783695,
      "epoch": 0.43434680922151686,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004986124477031293,
      "loss": 5.2708,
      "mean_token_accuracy": 0.1813582643866539,
      "num_tokens": 9572550.0,
      "step": 5200
    },
    {
      "entropy": 5.31060209274292,
      "epoch": 0.4347644503842299,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004986091446637083,
      "loss": 5.3474,
      "mean_token_accuracy": 0.17946574985980987,
      "num_tokens": 9580900.0,
      "step": 5205
    },
    {
      "entropy": 5.402362823486328,
      "epoch": 0.4351820915469429,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00049860583770974,
      "loss": 5.3189,
      "mean_token_accuracy": 0.18296932131052018,
      "num_tokens": 9589340.0,
      "step": 5210
    },
    {
      "entropy": 5.423223400115967,
      "epoch": 0.43559973270965585,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004986025268412823,
      "loss": 5.4521,
      "mean_token_accuracy": 0.1766061156988144,
      "num_tokens": 9599316.0,
      "step": 5215
    },
    {
      "entropy": 5.457963228225708,
      "epoch": 0.4360173738723689,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004985992120583933,
      "loss": 5.3374,
      "mean_token_accuracy": 0.17567280381917955,
      "num_tokens": 9609350.0,
      "step": 5220
    },
    {
      "entropy": 5.489837169647217,
      "epoch": 0.43643501503508186,
      "grad_norm": 1.265625,
      "learning_rate": 0.000498595893361131,
      "loss": 5.3962,
      "mean_token_accuracy": 0.18076670318841934,
      "num_tokens": 9618491.0,
      "step": 5225
    },
    {
      "entropy": 5.389662647247315,
      "epoch": 0.43685265619779484,
      "grad_norm": 1.125,
      "learning_rate": 0.0004985925707495533,
      "loss": 5.4583,
      "mean_token_accuracy": 0.16912828981876374,
      "num_tokens": 9629613.0,
      "step": 5230
    },
    {
      "entropy": 5.350446987152099,
      "epoch": 0.4372702973605079,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004985892442237186,
      "loss": 5.3291,
      "mean_token_accuracy": 0.17980844527482986,
      "num_tokens": 9638035.0,
      "step": 5235
    },
    {
      "entropy": 5.479083681106568,
      "epoch": 0.43768793852322085,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004985859137836851,
      "loss": 5.3899,
      "mean_token_accuracy": 0.17037070393562317,
      "num_tokens": 9647378.0,
      "step": 5240
    },
    {
      "entropy": 5.439112854003906,
      "epoch": 0.43810557968593383,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004985825794295109,
      "loss": 5.3092,
      "mean_token_accuracy": 0.18021205067634583,
      "num_tokens": 9655728.0,
      "step": 5245
    },
    {
      "entropy": 5.371652317047119,
      "epoch": 0.43852322084864687,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004985792411612546,
      "loss": 5.3426,
      "mean_token_accuracy": 0.18552555292844772,
      "num_tokens": 9665063.0,
      "step": 5250
    },
    {
      "entropy": 5.4210280418396,
      "epoch": 0.43894086201135984,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004985758989789746,
      "loss": 5.3247,
      "mean_token_accuracy": 0.1793728917837143,
      "num_tokens": 9674406.0,
      "step": 5255
    },
    {
      "entropy": 5.350600242614746,
      "epoch": 0.4393585031740728,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004985725528827293,
      "loss": 5.2842,
      "mean_token_accuracy": 0.17866162061691285,
      "num_tokens": 9683303.0,
      "step": 5260
    },
    {
      "entropy": 5.358454942703247,
      "epoch": 0.43977614433678586,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004985692028725774,
      "loss": 5.2941,
      "mean_token_accuracy": 0.17754778861999512,
      "num_tokens": 9692938.0,
      "step": 5265
    },
    {
      "entropy": 5.429182291030884,
      "epoch": 0.44019378549949884,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004985658489485774,
      "loss": 5.3549,
      "mean_token_accuracy": 0.18046221733093262,
      "num_tokens": 9702192.0,
      "step": 5270
    },
    {
      "entropy": 5.387715005874634,
      "epoch": 0.4406114266622118,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004985624911107882,
      "loss": 5.3418,
      "mean_token_accuracy": 0.1722553476691246,
      "num_tokens": 9711301.0,
      "step": 5275
    },
    {
      "entropy": 5.342666530609131,
      "epoch": 0.44102906782492485,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004985591293592685,
      "loss": 5.3347,
      "mean_token_accuracy": 0.17638202905654907,
      "num_tokens": 9720768.0,
      "step": 5280
    },
    {
      "entropy": 5.433481502532959,
      "epoch": 0.4414467089876378,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004985557636940771,
      "loss": 5.3717,
      "mean_token_accuracy": 0.17637849897146224,
      "num_tokens": 9730007.0,
      "step": 5285
    },
    {
      "entropy": 5.497163009643555,
      "epoch": 0.4418643501503508,
      "grad_norm": 1.328125,
      "learning_rate": 0.000498552394115273,
      "loss": 5.3995,
      "mean_token_accuracy": 0.17280904501676558,
      "num_tokens": 9739546.0,
      "step": 5290
    },
    {
      "entropy": 5.344863986968994,
      "epoch": 0.44228199131306384,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004985490206229151,
      "loss": 5.2902,
      "mean_token_accuracy": 0.18390890061855317,
      "num_tokens": 9749615.0,
      "step": 5295
    },
    {
      "entropy": 5.466024589538574,
      "epoch": 0.4426996324757768,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004985456432170626,
      "loss": 5.3927,
      "mean_token_accuracy": 0.17096854895353317,
      "num_tokens": 9759285.0,
      "step": 5300
    },
    {
      "entropy": 5.338745927810669,
      "epoch": 0.4431172736384898,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004985422618977744,
      "loss": 5.2538,
      "mean_token_accuracy": 0.1765191838145256,
      "num_tokens": 9769396.0,
      "step": 5305
    },
    {
      "entropy": 5.379651117324829,
      "epoch": 0.44353491480120283,
      "grad_norm": 1.234375,
      "learning_rate": 0.00049853887666511,
      "loss": 5.3788,
      "mean_token_accuracy": 0.17065683305263518,
      "num_tokens": 9779632.0,
      "step": 5310
    },
    {
      "entropy": 5.511491537094116,
      "epoch": 0.4439525559639158,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004985354875191283,
      "loss": 5.3043,
      "mean_token_accuracy": 0.18027861267328263,
      "num_tokens": 9788090.0,
      "step": 5315
    },
    {
      "entropy": 5.255291843414307,
      "epoch": 0.4443701971266288,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004985320944598889,
      "loss": 5.2171,
      "mean_token_accuracy": 0.18982870131731033,
      "num_tokens": 9795844.0,
      "step": 5320
    },
    {
      "entropy": 5.249967241287232,
      "epoch": 0.4447878382893418,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004985286974874512,
      "loss": 5.2153,
      "mean_token_accuracy": 0.1883803591132164,
      "num_tokens": 9805119.0,
      "step": 5325
    },
    {
      "entropy": 5.439846801757812,
      "epoch": 0.4452054794520548,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004985252966018745,
      "loss": 5.3253,
      "mean_token_accuracy": 0.18204206973314285,
      "num_tokens": 9813842.0,
      "step": 5330
    },
    {
      "entropy": 5.441546678543091,
      "epoch": 0.4456231206147678,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004985218918032184,
      "loss": 5.3271,
      "mean_token_accuracy": 0.17488853931427,
      "num_tokens": 9822872.0,
      "step": 5335
    },
    {
      "entropy": 5.438081836700439,
      "epoch": 0.4460407617774808,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004985184830915426,
      "loss": 5.4444,
      "mean_token_accuracy": 0.16981665343046187,
      "num_tokens": 9832710.0,
      "step": 5340
    },
    {
      "entropy": 5.504905652999878,
      "epoch": 0.4464584029401938,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004985150704669067,
      "loss": 5.3417,
      "mean_token_accuracy": 0.17520561516284944,
      "num_tokens": 9841047.0,
      "step": 5345
    },
    {
      "entropy": 5.4110023975372314,
      "epoch": 0.44687604410290677,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004985116539293704,
      "loss": 5.3015,
      "mean_token_accuracy": 0.17695399969816208,
      "num_tokens": 9850168.0,
      "step": 5350
    },
    {
      "entropy": 5.354893350601197,
      "epoch": 0.4472936852656198,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004985082334789936,
      "loss": 5.3746,
      "mean_token_accuracy": 0.1664320170879364,
      "num_tokens": 9859550.0,
      "step": 5355
    },
    {
      "entropy": 5.461143445968628,
      "epoch": 0.4477113264283328,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004985048091158361,
      "loss": 5.3081,
      "mean_token_accuracy": 0.17618091255426407,
      "num_tokens": 9868317.0,
      "step": 5360
    },
    {
      "entropy": 5.520910835266113,
      "epoch": 0.44812896759104576,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004985013808399578,
      "loss": 5.532,
      "mean_token_accuracy": 0.16146971583366393,
      "num_tokens": 9878421.0,
      "step": 5365
    },
    {
      "entropy": 5.452930593490601,
      "epoch": 0.4485466087537588,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004984979486514189,
      "loss": 5.2693,
      "mean_token_accuracy": 0.18397300988435744,
      "num_tokens": 9887213.0,
      "step": 5370
    },
    {
      "entropy": 5.377728796005249,
      "epoch": 0.44896424991647177,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004984945125502795,
      "loss": 5.3968,
      "mean_token_accuracy": 0.17348188757896424,
      "num_tokens": 9897274.0,
      "step": 5375
    },
    {
      "entropy": 5.336976003646851,
      "epoch": 0.44938189107918475,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004984910725365994,
      "loss": 5.3118,
      "mean_token_accuracy": 0.17537881433963776,
      "num_tokens": 9906765.0,
      "step": 5380
    },
    {
      "entropy": 5.482185411453247,
      "epoch": 0.4497995322418978,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004984876286104392,
      "loss": 5.3263,
      "mean_token_accuracy": 0.17005700916051864,
      "num_tokens": 9916149.0,
      "step": 5385
    },
    {
      "entropy": 5.419423675537109,
      "epoch": 0.45021717340461076,
      "grad_norm": 1.25,
      "learning_rate": 0.0004984841807718591,
      "loss": 5.3256,
      "mean_token_accuracy": 0.17702989429235458,
      "num_tokens": 9925657.0,
      "step": 5390
    },
    {
      "entropy": 5.366682529449463,
      "epoch": 0.45063481456732374,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004984807290209193,
      "loss": 5.3034,
      "mean_token_accuracy": 0.17391487210988998,
      "num_tokens": 9934740.0,
      "step": 5395
    },
    {
      "entropy": 5.554361343383789,
      "epoch": 0.45105245573003677,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004984772733576803,
      "loss": 5.461,
      "mean_token_accuracy": 0.1729282855987549,
      "num_tokens": 9943516.0,
      "step": 5400
    },
    {
      "entropy": 5.4826559066772464,
      "epoch": 0.45147009689274975,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004984738137822028,
      "loss": 5.3969,
      "mean_token_accuracy": 0.1681110993027687,
      "num_tokens": 9952998.0,
      "step": 5405
    },
    {
      "entropy": 5.384304571151733,
      "epoch": 0.45188773805546273,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000498470350294547,
      "loss": 5.2152,
      "mean_token_accuracy": 0.18386478573083878,
      "num_tokens": 9961698.0,
      "step": 5410
    },
    {
      "entropy": 5.374800205230713,
      "epoch": 0.45230537921817576,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004984668828947739,
      "loss": 5.3546,
      "mean_token_accuracy": 0.17178037017583847,
      "num_tokens": 9972015.0,
      "step": 5415
    },
    {
      "entropy": 5.40453519821167,
      "epoch": 0.45272302038088874,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004984634115829439,
      "loss": 5.4344,
      "mean_token_accuracy": 0.17172151505947114,
      "num_tokens": 9981834.0,
      "step": 5420
    },
    {
      "entropy": 5.331679344177246,
      "epoch": 0.4531406615436017,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000498459936359118,
      "loss": 5.2877,
      "mean_token_accuracy": 0.18475869297981262,
      "num_tokens": 9991113.0,
      "step": 5425
    },
    {
      "entropy": 5.386208820343017,
      "epoch": 0.45355830270631475,
      "grad_norm": 1.25,
      "learning_rate": 0.0004984564572233568,
      "loss": 5.2612,
      "mean_token_accuracy": 0.185154689848423,
      "num_tokens": 10000048.0,
      "step": 5430
    },
    {
      "entropy": 5.427082777023315,
      "epoch": 0.45397594386902773,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004984529741757216,
      "loss": 5.3136,
      "mean_token_accuracy": 0.18205272108316423,
      "num_tokens": 10009389.0,
      "step": 5435
    },
    {
      "entropy": 5.335471868515015,
      "epoch": 0.4543935850317407,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004984494872162728,
      "loss": 5.1676,
      "mean_token_accuracy": 0.18400882184505463,
      "num_tokens": 10018342.0,
      "step": 5440
    },
    {
      "entropy": 5.422104072570801,
      "epoch": 0.45481122619445374,
      "grad_norm": 1.453125,
      "learning_rate": 0.000498445996345072,
      "loss": 5.352,
      "mean_token_accuracy": 0.1807814508676529,
      "num_tokens": 10027219.0,
      "step": 5445
    },
    {
      "entropy": 5.291424751281738,
      "epoch": 0.4552288673571667,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004984425015621799,
      "loss": 5.2568,
      "mean_token_accuracy": 0.17668753862380981,
      "num_tokens": 10036635.0,
      "step": 5450
    },
    {
      "entropy": 5.309024143218994,
      "epoch": 0.4556465085198797,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000498439002867658,
      "loss": 5.3069,
      "mean_token_accuracy": 0.1802414745092392,
      "num_tokens": 10045655.0,
      "step": 5455
    },
    {
      "entropy": 5.528062772750855,
      "epoch": 0.45606414968259273,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004984355002615673,
      "loss": 5.3658,
      "mean_token_accuracy": 0.16974907964468003,
      "num_tokens": 10054366.0,
      "step": 5460
    },
    {
      "entropy": 5.380119657516479,
      "epoch": 0.4564817908453057,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004984319937439693,
      "loss": 5.3169,
      "mean_token_accuracy": 0.1861657440662384,
      "num_tokens": 10062548.0,
      "step": 5465
    },
    {
      "entropy": 5.329405069351196,
      "epoch": 0.4568994320080187,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004984284833149252,
      "loss": 5.2996,
      "mean_token_accuracy": 0.17789616137742997,
      "num_tokens": 10071333.0,
      "step": 5470
    },
    {
      "entropy": 5.279886627197266,
      "epoch": 0.4573170731707317,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004984249689744966,
      "loss": 5.1906,
      "mean_token_accuracy": 0.18658134043216706,
      "num_tokens": 10079502.0,
      "step": 5475
    },
    {
      "entropy": 5.323281526565552,
      "epoch": 0.4577347143334447,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000498421450722745,
      "loss": 5.2308,
      "mean_token_accuracy": 0.18677007406949997,
      "num_tokens": 10088648.0,
      "step": 5480
    },
    {
      "entropy": 5.322705841064453,
      "epoch": 0.4581523554961577,
      "grad_norm": 1.265625,
      "learning_rate": 0.000498417928559732,
      "loss": 5.2418,
      "mean_token_accuracy": 0.18585171401500702,
      "num_tokens": 10097742.0,
      "step": 5485
    },
    {
      "entropy": 5.403957986831665,
      "epoch": 0.4585699966588707,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004984144024855191,
      "loss": 5.3563,
      "mean_token_accuracy": 0.1740902915596962,
      "num_tokens": 10107228.0,
      "step": 5490
    },
    {
      "entropy": 5.359741067886352,
      "epoch": 0.4589876378215837,
      "grad_norm": 1.5,
      "learning_rate": 0.0004984108725001683,
      "loss": 5.2611,
      "mean_token_accuracy": 0.18054186105728148,
      "num_tokens": 10115488.0,
      "step": 5495
    },
    {
      "entropy": 5.276429319381714,
      "epoch": 0.45940527898429667,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004984073386037413,
      "loss": 5.1774,
      "mean_token_accuracy": 0.18593759089708328,
      "num_tokens": 10124235.0,
      "step": 5500
    },
    {
      "entropy": 5.282084178924561,
      "epoch": 0.4598229201470097,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004984038007962997,
      "loss": 5.2832,
      "mean_token_accuracy": 0.18730263113975526,
      "num_tokens": 10133208.0,
      "step": 5505
    },
    {
      "entropy": 5.338069248199463,
      "epoch": 0.4602405613097227,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004984002590779059,
      "loss": 5.2134,
      "mean_token_accuracy": 0.18483959138393402,
      "num_tokens": 10142754.0,
      "step": 5510
    },
    {
      "entropy": 5.421583557128907,
      "epoch": 0.46065820247243566,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004983967134486215,
      "loss": 5.3281,
      "mean_token_accuracy": 0.18001955896615982,
      "num_tokens": 10152114.0,
      "step": 5515
    },
    {
      "entropy": 5.357906293869019,
      "epoch": 0.4610758436351487,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004983931639085088,
      "loss": 5.244,
      "mean_token_accuracy": 0.18192300349473953,
      "num_tokens": 10161562.0,
      "step": 5520
    },
    {
      "entropy": 5.395240306854248,
      "epoch": 0.4614934847978617,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004983896104576298,
      "loss": 5.3466,
      "mean_token_accuracy": 0.17601957768201829,
      "num_tokens": 10170461.0,
      "step": 5525
    },
    {
      "entropy": 5.41256160736084,
      "epoch": 0.46191112596057465,
      "grad_norm": 1.328125,
      "learning_rate": 0.000498386053096047,
      "loss": 5.3461,
      "mean_token_accuracy": 0.17777144759893418,
      "num_tokens": 10179243.0,
      "step": 5530
    },
    {
      "entropy": 5.392285537719727,
      "epoch": 0.4623287671232877,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004983824918238221,
      "loss": 5.2858,
      "mean_token_accuracy": 0.18495858013629912,
      "num_tokens": 10187935.0,
      "step": 5535
    },
    {
      "entropy": 5.435108327865601,
      "epoch": 0.46274640828600067,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000498378926641018,
      "loss": 5.3696,
      "mean_token_accuracy": 0.17514135539531708,
      "num_tokens": 10196185.0,
      "step": 5540
    },
    {
      "entropy": 5.4282838821411135,
      "epoch": 0.46316404944871364,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004983753575476969,
      "loss": 5.3395,
      "mean_token_accuracy": 0.1728462278842926,
      "num_tokens": 10205373.0,
      "step": 5545
    },
    {
      "entropy": 5.3686912059783936,
      "epoch": 0.4635816906114267,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004983717845439212,
      "loss": 5.3222,
      "mean_token_accuracy": 0.18750559240579606,
      "num_tokens": 10214464.0,
      "step": 5550
    },
    {
      "entropy": 5.397244310379028,
      "epoch": 0.46399933177413966,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004983682076297536,
      "loss": 5.3975,
      "mean_token_accuracy": 0.1748291254043579,
      "num_tokens": 10224462.0,
      "step": 5555
    },
    {
      "entropy": 5.490514850616455,
      "epoch": 0.46441697293685263,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004983646268052567,
      "loss": 5.3247,
      "mean_token_accuracy": 0.18098523914813996,
      "num_tokens": 10233123.0,
      "step": 5560
    },
    {
      "entropy": 5.338079309463501,
      "epoch": 0.46483461409956567,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004983610420704931,
      "loss": 5.1601,
      "mean_token_accuracy": 0.1845409706234932,
      "num_tokens": 10241613.0,
      "step": 5565
    },
    {
      "entropy": 5.43339262008667,
      "epoch": 0.46525225526227865,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004983574534255256,
      "loss": 5.5158,
      "mean_token_accuracy": 0.16590092927217484,
      "num_tokens": 10251809.0,
      "step": 5570
    },
    {
      "entropy": 5.359250164031982,
      "epoch": 0.4656698964249916,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000498353860870417,
      "loss": 5.2021,
      "mean_token_accuracy": 0.18127120435237884,
      "num_tokens": 10260198.0,
      "step": 5575
    },
    {
      "entropy": 5.4148660659790036,
      "epoch": 0.46608753758770466,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004983502644052303,
      "loss": 5.2674,
      "mean_token_accuracy": 0.1783775269985199,
      "num_tokens": 10268906.0,
      "step": 5580
    },
    {
      "entropy": 5.307451057434082,
      "epoch": 0.46650517875041764,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004983466640300282,
      "loss": 5.1726,
      "mean_token_accuracy": 0.18028734028339385,
      "num_tokens": 10278526.0,
      "step": 5585
    },
    {
      "entropy": 5.332444190979004,
      "epoch": 0.4669228199131306,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004983430597448741,
      "loss": 5.3036,
      "mean_token_accuracy": 0.18263361006975173,
      "num_tokens": 10287428.0,
      "step": 5590
    },
    {
      "entropy": 5.506376791000366,
      "epoch": 0.46734046107584365,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004983394515498308,
      "loss": 5.2966,
      "mean_token_accuracy": 0.17286923974752427,
      "num_tokens": 10297136.0,
      "step": 5595
    },
    {
      "entropy": 5.388471555709839,
      "epoch": 0.46775810223855663,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004983358394449616,
      "loss": 5.3562,
      "mean_token_accuracy": 0.177946974337101,
      "num_tokens": 10305678.0,
      "step": 5600
    },
    {
      "entropy": 5.251855993270874,
      "epoch": 0.4681757434012696,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004983322234303297,
      "loss": 5.276,
      "mean_token_accuracy": 0.1755787819623947,
      "num_tokens": 10314690.0,
      "step": 5605
    },
    {
      "entropy": 5.3856062412261965,
      "epoch": 0.46859338456398264,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004983286035059984,
      "loss": 5.2237,
      "mean_token_accuracy": 0.17978821098804473,
      "num_tokens": 10323521.0,
      "step": 5610
    },
    {
      "entropy": 5.373954200744629,
      "epoch": 0.4690110257266956,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004983249796720311,
      "loss": 5.2539,
      "mean_token_accuracy": 0.18739869743585585,
      "num_tokens": 10333077.0,
      "step": 5615
    },
    {
      "entropy": 5.315990686416626,
      "epoch": 0.4694286668894086,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004983213519284912,
      "loss": 5.2979,
      "mean_token_accuracy": 0.1791898638010025,
      "num_tokens": 10342194.0,
      "step": 5620
    },
    {
      "entropy": 5.298461961746216,
      "epoch": 0.46984630805212163,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004983177202754421,
      "loss": 5.1452,
      "mean_token_accuracy": 0.19010407328605652,
      "num_tokens": 10351009.0,
      "step": 5625
    },
    {
      "entropy": 5.410635614395142,
      "epoch": 0.4702639492148346,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004983140847129476,
      "loss": 5.3391,
      "mean_token_accuracy": 0.17841810435056688,
      "num_tokens": 10360872.0,
      "step": 5630
    },
    {
      "entropy": 5.368702793121338,
      "epoch": 0.4706815903775476,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004983104452410713,
      "loss": 5.2621,
      "mean_token_accuracy": 0.18015502095222474,
      "num_tokens": 10370159.0,
      "step": 5635
    },
    {
      "entropy": 5.339200353622436,
      "epoch": 0.4710992315402606,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004983068018598768,
      "loss": 5.251,
      "mean_token_accuracy": 0.1864570513367653,
      "num_tokens": 10379219.0,
      "step": 5640
    },
    {
      "entropy": 5.379155874252319,
      "epoch": 0.4715168727029736,
      "grad_norm": 1.296875,
      "learning_rate": 0.000498303154569428,
      "loss": 5.2945,
      "mean_token_accuracy": 0.17841103971004485,
      "num_tokens": 10389425.0,
      "step": 5645
    },
    {
      "entropy": 5.391329765319824,
      "epoch": 0.4719345138656866,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004982995033697887,
      "loss": 5.3186,
      "mean_token_accuracy": 0.18286772966384887,
      "num_tokens": 10398023.0,
      "step": 5650
    },
    {
      "entropy": 5.385508298873901,
      "epoch": 0.4723521550283996,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004982958482610226,
      "loss": 5.2503,
      "mean_token_accuracy": 0.1814128801226616,
      "num_tokens": 10406578.0,
      "step": 5655
    },
    {
      "entropy": 5.273907423019409,
      "epoch": 0.4727697961911126,
      "grad_norm": 1.28125,
      "learning_rate": 0.000498292189243194,
      "loss": 5.2583,
      "mean_token_accuracy": 0.18512603789567947,
      "num_tokens": 10416448.0,
      "step": 5660
    },
    {
      "entropy": 5.385401105880737,
      "epoch": 0.47318743735382557,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000498288526316367,
      "loss": 5.2807,
      "mean_token_accuracy": 0.17727738469839097,
      "num_tokens": 10425414.0,
      "step": 5665
    },
    {
      "entropy": 5.361488723754883,
      "epoch": 0.4736050785165386,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004982848594806054,
      "loss": 5.2749,
      "mean_token_accuracy": 0.18584574311971663,
      "num_tokens": 10433900.0,
      "step": 5670
    },
    {
      "entropy": 5.416553497314453,
      "epoch": 0.4740227196792516,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004982811887359736,
      "loss": 5.375,
      "mean_token_accuracy": 0.17688911110162736,
      "num_tokens": 10443077.0,
      "step": 5675
    },
    {
      "entropy": 5.431783676147461,
      "epoch": 0.47444036084196456,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004982775140825359,
      "loss": 5.3065,
      "mean_token_accuracy": 0.1801500827074051,
      "num_tokens": 10452791.0,
      "step": 5680
    },
    {
      "entropy": 5.464850807189942,
      "epoch": 0.4748580020046776,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004982738355203565,
      "loss": 5.478,
      "mean_token_accuracy": 0.16853803396224976,
      "num_tokens": 10462087.0,
      "step": 5685
    },
    {
      "entropy": 5.489383697509766,
      "epoch": 0.47527564316739057,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004982701530494998,
      "loss": 5.4765,
      "mean_token_accuracy": 0.1620257392525673,
      "num_tokens": 10472576.0,
      "step": 5690
    },
    {
      "entropy": 5.422889518737793,
      "epoch": 0.47569328433010355,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004982664666700304,
      "loss": 5.2016,
      "mean_token_accuracy": 0.18758722692728041,
      "num_tokens": 10481176.0,
      "step": 5695
    },
    {
      "entropy": 5.33063063621521,
      "epoch": 0.4761109254928166,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004982627763820128,
      "loss": 5.3086,
      "mean_token_accuracy": 0.17557826787233352,
      "num_tokens": 10490558.0,
      "step": 5700
    },
    {
      "entropy": 5.420174217224121,
      "epoch": 0.47652856665552956,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004982590821855115,
      "loss": 5.3597,
      "mean_token_accuracy": 0.17991894334554673,
      "num_tokens": 10499543.0,
      "step": 5705
    },
    {
      "entropy": 5.348049879074097,
      "epoch": 0.47694620781824254,
      "grad_norm": 1.625,
      "learning_rate": 0.0004982553840805912,
      "loss": 5.2476,
      "mean_token_accuracy": 0.18451592773199083,
      "num_tokens": 10508162.0,
      "step": 5710
    },
    {
      "entropy": 5.332060480117798,
      "epoch": 0.4773638489809556,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004982516820673167,
      "loss": 5.1524,
      "mean_token_accuracy": 0.1859285429120064,
      "num_tokens": 10516837.0,
      "step": 5715
    },
    {
      "entropy": 5.348451137542725,
      "epoch": 0.47778149014366855,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004982479761457528,
      "loss": 5.2843,
      "mean_token_accuracy": 0.18323100209236146,
      "num_tokens": 10526208.0,
      "step": 5720
    },
    {
      "entropy": 5.4435309886932375,
      "epoch": 0.47819913130638153,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004982442663159643,
      "loss": 5.3526,
      "mean_token_accuracy": 0.1724609911441803,
      "num_tokens": 10536002.0,
      "step": 5725
    },
    {
      "entropy": 5.425581312179565,
      "epoch": 0.47861677246909456,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004982405525780162,
      "loss": 5.3181,
      "mean_token_accuracy": 0.17911448329687119,
      "num_tokens": 10545861.0,
      "step": 5730
    },
    {
      "entropy": 5.382137107849121,
      "epoch": 0.47903441363180754,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004982368349319736,
      "loss": 5.2315,
      "mean_token_accuracy": 0.1853106439113617,
      "num_tokens": 10555357.0,
      "step": 5735
    },
    {
      "entropy": 5.369589853286743,
      "epoch": 0.4794520547945205,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004982331133779014,
      "loss": 5.3045,
      "mean_token_accuracy": 0.17767890989780427,
      "num_tokens": 10564744.0,
      "step": 5740
    },
    {
      "entropy": 5.314902353286743,
      "epoch": 0.47986969595723356,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004982293879158649,
      "loss": 5.2694,
      "mean_token_accuracy": 0.18325549662113189,
      "num_tokens": 10573736.0,
      "step": 5745
    },
    {
      "entropy": 5.375504016876221,
      "epoch": 0.48028733711994653,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004982256585459293,
      "loss": 5.3302,
      "mean_token_accuracy": 0.1777355268597603,
      "num_tokens": 10583238.0,
      "step": 5750
    },
    {
      "entropy": 5.455562448501587,
      "epoch": 0.4807049782826595,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004982219252681597,
      "loss": 5.3942,
      "mean_token_accuracy": 0.1724510073661804,
      "num_tokens": 10592316.0,
      "step": 5755
    },
    {
      "entropy": 5.407722091674804,
      "epoch": 0.48112261944537255,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004982181880826218,
      "loss": 5.3938,
      "mean_token_accuracy": 0.18294421583414078,
      "num_tokens": 10601418.0,
      "step": 5760
    },
    {
      "entropy": 5.333720922470093,
      "epoch": 0.4815402606080855,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004982144469893807,
      "loss": 5.2696,
      "mean_token_accuracy": 0.18337585180997848,
      "num_tokens": 10610513.0,
      "step": 5765
    },
    {
      "entropy": 5.368336915969849,
      "epoch": 0.4819579017707985,
      "grad_norm": 1.296875,
      "learning_rate": 0.000498210701988502,
      "loss": 5.2205,
      "mean_token_accuracy": 0.18447786718606948,
      "num_tokens": 10619529.0,
      "step": 5770
    },
    {
      "entropy": 5.485368061065674,
      "epoch": 0.48237554293351154,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004982069530800514,
      "loss": 5.3216,
      "mean_token_accuracy": 0.1866314858198166,
      "num_tokens": 10628263.0,
      "step": 5775
    },
    {
      "entropy": 5.438887739181519,
      "epoch": 0.4827931840962245,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004982032002640943,
      "loss": 5.3464,
      "mean_token_accuracy": 0.18147321790456772,
      "num_tokens": 10637605.0,
      "step": 5780
    },
    {
      "entropy": 5.40214958190918,
      "epoch": 0.48321082525893755,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004981994435406965,
      "loss": 5.4084,
      "mean_token_accuracy": 0.1765257492661476,
      "num_tokens": 10647113.0,
      "step": 5785
    },
    {
      "entropy": 5.340604686737061,
      "epoch": 0.4836284664216505,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004981956829099239,
      "loss": 5.1619,
      "mean_token_accuracy": 0.18975128680467607,
      "num_tokens": 10655464.0,
      "step": 5790
    },
    {
      "entropy": 5.403401851654053,
      "epoch": 0.4840461075843635,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004981919183718422,
      "loss": 5.2674,
      "mean_token_accuracy": 0.18043090552091598,
      "num_tokens": 10664689.0,
      "step": 5795
    },
    {
      "entropy": 5.264755392074585,
      "epoch": 0.48446374874707654,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004981881499265173,
      "loss": 5.1747,
      "mean_token_accuracy": 0.18198387920856476,
      "num_tokens": 10673440.0,
      "step": 5800
    },
    {
      "entropy": 5.299979114532471,
      "epoch": 0.4848813899097895,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004981843775740151,
      "loss": 5.3026,
      "mean_token_accuracy": 0.18126818984746934,
      "num_tokens": 10683075.0,
      "step": 5805
    },
    {
      "entropy": 5.384060907363891,
      "epoch": 0.4852990310725025,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004981806013144017,
      "loss": 5.2763,
      "mean_token_accuracy": 0.18361429870128632,
      "num_tokens": 10692663.0,
      "step": 5810
    },
    {
      "entropy": 5.366247510910034,
      "epoch": 0.48571667223521553,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004981768211477434,
      "loss": 5.177,
      "mean_token_accuracy": 0.1851995274424553,
      "num_tokens": 10702068.0,
      "step": 5815
    },
    {
      "entropy": 5.3900166034698485,
      "epoch": 0.4861343133979285,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000498173037074106,
      "loss": 5.3275,
      "mean_token_accuracy": 0.1716886818408966,
      "num_tokens": 10711230.0,
      "step": 5820
    },
    {
      "entropy": 5.321808528900147,
      "epoch": 0.4865519545606415,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004981692490935561,
      "loss": 5.2848,
      "mean_token_accuracy": 0.18573111295700073,
      "num_tokens": 10719651.0,
      "step": 5825
    },
    {
      "entropy": 5.370608282089234,
      "epoch": 0.4869695957233545,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004981654572061597,
      "loss": 5.3009,
      "mean_token_accuracy": 0.18043226301670073,
      "num_tokens": 10729342.0,
      "step": 5830
    },
    {
      "entropy": 5.434235668182373,
      "epoch": 0.4873872368860675,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004981616614119836,
      "loss": 5.2954,
      "mean_token_accuracy": 0.185221067070961,
      "num_tokens": 10738617.0,
      "step": 5835
    },
    {
      "entropy": 5.412237596511841,
      "epoch": 0.4878048780487805,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004981578617110938,
      "loss": 5.3353,
      "mean_token_accuracy": 0.1808386668562889,
      "num_tokens": 10746881.0,
      "step": 5840
    },
    {
      "entropy": 5.320700454711914,
      "epoch": 0.4882225192114935,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004981540581035571,
      "loss": 5.1917,
      "mean_token_accuracy": 0.19013732820749282,
      "num_tokens": 10755848.0,
      "step": 5845
    },
    {
      "entropy": 5.322195291519165,
      "epoch": 0.4886401603742065,
      "grad_norm": 1.359375,
      "learning_rate": 0.00049815025058944,
      "loss": 5.1859,
      "mean_token_accuracy": 0.1823713093996048,
      "num_tokens": 10765519.0,
      "step": 5850
    },
    {
      "entropy": 5.298445796966552,
      "epoch": 0.48905780153691947,
      "grad_norm": 1.25,
      "learning_rate": 0.0004981464391688092,
      "loss": 5.3266,
      "mean_token_accuracy": 0.17206095159053802,
      "num_tokens": 10776408.0,
      "step": 5855
    },
    {
      "entropy": 5.365063190460205,
      "epoch": 0.4894754426996325,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004981426238417312,
      "loss": 5.3158,
      "mean_token_accuracy": 0.18536051958799363,
      "num_tokens": 10785499.0,
      "step": 5860
    },
    {
      "entropy": 5.382756567001342,
      "epoch": 0.4898930838623455,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004981388046082731,
      "loss": 5.2718,
      "mean_token_accuracy": 0.1825641930103302,
      "num_tokens": 10794736.0,
      "step": 5865
    },
    {
      "entropy": 5.368501281738281,
      "epoch": 0.49031072502505846,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004981349814685016,
      "loss": 5.3274,
      "mean_token_accuracy": 0.17694495171308516,
      "num_tokens": 10804125.0,
      "step": 5870
    },
    {
      "entropy": 5.485472869873047,
      "epoch": 0.4907283661877715,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004981311544224838,
      "loss": 5.3985,
      "mean_token_accuracy": 0.18018668740987778,
      "num_tokens": 10813465.0,
      "step": 5875
    },
    {
      "entropy": 5.334915399551392,
      "epoch": 0.49114600735048447,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004981273234702864,
      "loss": 5.2073,
      "mean_token_accuracy": 0.1875336542725563,
      "num_tokens": 10821550.0,
      "step": 5880
    },
    {
      "entropy": 5.338493299484253,
      "epoch": 0.49156364851319745,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004981234886119767,
      "loss": 5.245,
      "mean_token_accuracy": 0.18219776451587677,
      "num_tokens": 10830683.0,
      "step": 5885
    },
    {
      "entropy": 5.369595909118653,
      "epoch": 0.4919812896759105,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004981196498476217,
      "loss": 5.2527,
      "mean_token_accuracy": 0.18770304918289185,
      "num_tokens": 10839492.0,
      "step": 5890
    },
    {
      "entropy": 5.3159058570861815,
      "epoch": 0.49239893083862346,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004981158071772887,
      "loss": 5.1701,
      "mean_token_accuracy": 0.19166890382766724,
      "num_tokens": 10847998.0,
      "step": 5895
    },
    {
      "entropy": 5.240080451965332,
      "epoch": 0.49281657200133644,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004981119606010449,
      "loss": 5.1771,
      "mean_token_accuracy": 0.189774052798748,
      "num_tokens": 10856438.0,
      "step": 5900
    },
    {
      "entropy": 5.267599725723267,
      "epoch": 0.4932342131640495,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004981081101189577,
      "loss": 5.1876,
      "mean_token_accuracy": 0.181645205616951,
      "num_tokens": 10865936.0,
      "step": 5905
    },
    {
      "entropy": 5.358993434906006,
      "epoch": 0.49365185432676245,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004981042557310945,
      "loss": 5.1959,
      "mean_token_accuracy": 0.18553776443004608,
      "num_tokens": 10874621.0,
      "step": 5910
    },
    {
      "entropy": 5.33198766708374,
      "epoch": 0.49406949548947543,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004981003974375227,
      "loss": 5.1729,
      "mean_token_accuracy": 0.1845684453845024,
      "num_tokens": 10883516.0,
      "step": 5915
    },
    {
      "entropy": 5.3517989158630375,
      "epoch": 0.49448713665218846,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004980965352383099,
      "loss": 5.307,
      "mean_token_accuracy": 0.1766266942024231,
      "num_tokens": 10892580.0,
      "step": 5920
    },
    {
      "entropy": 5.357056951522827,
      "epoch": 0.49490477781490144,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004980926691335238,
      "loss": 5.2308,
      "mean_token_accuracy": 0.18829175233840942,
      "num_tokens": 10901227.0,
      "step": 5925
    },
    {
      "entropy": 5.3624979019165036,
      "epoch": 0.4953224189776144,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004980887991232318,
      "loss": 5.2728,
      "mean_token_accuracy": 0.17593953758478165,
      "num_tokens": 10909454.0,
      "step": 5930
    },
    {
      "entropy": 5.474798393249512,
      "epoch": 0.49574006014032745,
      "grad_norm": 1.4375,
      "learning_rate": 0.000498084925207502,
      "loss": 5.3562,
      "mean_token_accuracy": 0.17890977412462233,
      "num_tokens": 10919792.0,
      "step": 5935
    },
    {
      "entropy": 5.422196006774902,
      "epoch": 0.49615770130304043,
      "grad_norm": 1.3125,
      "learning_rate": 0.000498081047386402,
      "loss": 5.3897,
      "mean_token_accuracy": 0.17535832524299622,
      "num_tokens": 10930523.0,
      "step": 5940
    },
    {
      "entropy": 5.304169750213623,
      "epoch": 0.4965753424657534,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004980771656599998,
      "loss": 5.2476,
      "mean_token_accuracy": 0.18457962125539779,
      "num_tokens": 10940075.0,
      "step": 5945
    },
    {
      "entropy": 5.379092025756836,
      "epoch": 0.49699298362846644,
      "grad_norm": 1.25,
      "learning_rate": 0.0004980732800283633,
      "loss": 5.2687,
      "mean_token_accuracy": 0.17703412622213363,
      "num_tokens": 10949059.0,
      "step": 5950
    },
    {
      "entropy": 5.374813175201416,
      "epoch": 0.4974106247911794,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004980693904915604,
      "loss": 5.2363,
      "mean_token_accuracy": 0.1839677095413208,
      "num_tokens": 10958313.0,
      "step": 5955
    },
    {
      "entropy": 5.287323093414306,
      "epoch": 0.4978282659538924,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004980654970496595,
      "loss": 5.2889,
      "mean_token_accuracy": 0.1837708458304405,
      "num_tokens": 10967666.0,
      "step": 5960
    },
    {
      "entropy": 5.385119342803955,
      "epoch": 0.49824590711660544,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004980615997027284,
      "loss": 5.3045,
      "mean_token_accuracy": 0.1810262307524681,
      "num_tokens": 10976519.0,
      "step": 5965
    },
    {
      "entropy": 5.468370676040649,
      "epoch": 0.4986635482793184,
      "grad_norm": 1.375,
      "learning_rate": 0.0004980576984508356,
      "loss": 5.4155,
      "mean_token_accuracy": 0.1726203069090843,
      "num_tokens": 10987138.0,
      "step": 5970
    },
    {
      "entropy": 5.394054555892945,
      "epoch": 0.4990811894420314,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004980537932940492,
      "loss": 5.2881,
      "mean_token_accuracy": 0.18088141530752183,
      "num_tokens": 10996517.0,
      "step": 5975
    },
    {
      "entropy": 5.35601921081543,
      "epoch": 0.4994988306047444,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004980498842324377,
      "loss": 5.1878,
      "mean_token_accuracy": 0.1789805054664612,
      "num_tokens": 11006312.0,
      "step": 5980
    },
    {
      "entropy": 5.3022911071777346,
      "epoch": 0.4999164717674574,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004980459712660694,
      "loss": 5.2933,
      "mean_token_accuracy": 0.18762718588113786,
      "num_tokens": 11015535.0,
      "step": 5985
    },
    {
      "entropy": 5.340489625930786,
      "epoch": 0.5003341129301704,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004980420543950131,
      "loss": 5.2993,
      "mean_token_accuracy": 0.18379236608743668,
      "num_tokens": 11024887.0,
      "step": 5990
    },
    {
      "entropy": 5.253097724914551,
      "epoch": 0.5007517540928834,
      "grad_norm": 1.25,
      "learning_rate": 0.0004980381336193371,
      "loss": 5.1465,
      "mean_token_accuracy": 0.1922702133655548,
      "num_tokens": 11033147.0,
      "step": 5995
    },
    {
      "entropy": 5.330597829818726,
      "epoch": 0.5011693952555963,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004980342089391099,
      "loss": 5.3002,
      "mean_token_accuracy": 0.18446871489286423,
      "num_tokens": 11042726.0,
      "step": 6000
    },
    {
      "epoch": 0.5011693952555963,
      "eval_entropy": 5.122171156892734,
      "eval_loss": 5.276597023010254,
      "eval_mean_token_accuracy": 0.19089796238106782,
      "eval_num_tokens": 11042726.0,
      "eval_runtime": 27.3546,
      "eval_samples_per_second": 1380.499,
      "eval_steps_per_second": 172.585,
      "step": 6000
    },
    {
      "entropy": 5.3341082572937015,
      "epoch": 0.5015870364183094,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004980302803544006,
      "loss": 5.2947,
      "mean_token_accuracy": 0.1802077740430832,
      "num_tokens": 11052181.0,
      "step": 6005
    },
    {
      "entropy": 5.4133617877960205,
      "epoch": 0.5020046775810224,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004980263478652778,
      "loss": 5.3263,
      "mean_token_accuracy": 0.1810848146677017,
      "num_tokens": 11061586.0,
      "step": 6010
    },
    {
      "entropy": 5.479144477844239,
      "epoch": 0.5024223187437353,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004980224114718102,
      "loss": 5.348,
      "mean_token_accuracy": 0.1872189611196518,
      "num_tokens": 11070914.0,
      "step": 6015
    },
    {
      "entropy": 5.252730751037598,
      "epoch": 0.5028399599064484,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004980184711740669,
      "loss": 5.167,
      "mean_token_accuracy": 0.1886384144425392,
      "num_tokens": 11079836.0,
      "step": 6020
    },
    {
      "entropy": 5.248090744018555,
      "epoch": 0.5032576010691614,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004980145269721168,
      "loss": 5.1897,
      "mean_token_accuracy": 0.1809512495994568,
      "num_tokens": 11088320.0,
      "step": 6025
    },
    {
      "entropy": 5.397958278656006,
      "epoch": 0.5036752422318743,
      "grad_norm": 1.296875,
      "learning_rate": 0.000498010578866029,
      "loss": 5.3541,
      "mean_token_accuracy": 0.18020374923944474,
      "num_tokens": 11097803.0,
      "step": 6030
    },
    {
      "entropy": 5.469384574890137,
      "epoch": 0.5040928833945874,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004980066268558726,
      "loss": 5.3733,
      "mean_token_accuracy": 0.17793723195791245,
      "num_tokens": 11107218.0,
      "step": 6035
    },
    {
      "entropy": 5.43264799118042,
      "epoch": 0.5045105245573004,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004980026709417167,
      "loss": 5.2653,
      "mean_token_accuracy": 0.18189581483602524,
      "num_tokens": 11115599.0,
      "step": 6040
    },
    {
      "entropy": 5.2718737602233885,
      "epoch": 0.5049281657200133,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004979987111236306,
      "loss": 5.151,
      "mean_token_accuracy": 0.18838154375553132,
      "num_tokens": 11124312.0,
      "step": 6045
    },
    {
      "entropy": 5.3155444145202635,
      "epoch": 0.5053458068827263,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004979947474016837,
      "loss": 5.2877,
      "mean_token_accuracy": 0.18409484922885894,
      "num_tokens": 11133117.0,
      "step": 6050
    },
    {
      "entropy": 5.507103681564331,
      "epoch": 0.5057634480454394,
      "grad_norm": 1.25,
      "learning_rate": 0.0004979907797759453,
      "loss": 5.3329,
      "mean_token_accuracy": 0.17988639920949936,
      "num_tokens": 11141972.0,
      "step": 6055
    },
    {
      "entropy": 5.437141990661621,
      "epoch": 0.5061810892081523,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004979868082464849,
      "loss": 5.3501,
      "mean_token_accuracy": 0.17766742259263993,
      "num_tokens": 11151416.0,
      "step": 6060
    },
    {
      "entropy": 5.344732427597046,
      "epoch": 0.5065987303708653,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004979828328133719,
      "loss": 5.299,
      "mean_token_accuracy": 0.17817157208919526,
      "num_tokens": 11160727.0,
      "step": 6065
    },
    {
      "entropy": 5.356343793869018,
      "epoch": 0.5070163715335784,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004979788534766762,
      "loss": 5.2537,
      "mean_token_accuracy": 0.17333008646965026,
      "num_tokens": 11171350.0,
      "step": 6070
    },
    {
      "entropy": 5.357258081436157,
      "epoch": 0.5074340126962913,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004979748702364672,
      "loss": 5.2315,
      "mean_token_accuracy": 0.17688867598772048,
      "num_tokens": 11180233.0,
      "step": 6075
    },
    {
      "entropy": 5.3647864818572994,
      "epoch": 0.5078516538590043,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004979708830928147,
      "loss": 5.2232,
      "mean_token_accuracy": 0.18030519783496857,
      "num_tokens": 11189269.0,
      "step": 6080
    },
    {
      "entropy": 5.394839859008789,
      "epoch": 0.5082692950217174,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004979668920457884,
      "loss": 5.3028,
      "mean_token_accuracy": 0.17938395589590073,
      "num_tokens": 11199072.0,
      "step": 6085
    },
    {
      "entropy": 5.323867988586426,
      "epoch": 0.5086869361844303,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004979628970954584,
      "loss": 5.2023,
      "mean_token_accuracy": 0.18903121054172517,
      "num_tokens": 11207164.0,
      "step": 6090
    },
    {
      "entropy": 5.287879514694214,
      "epoch": 0.5091045773471433,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004979588982418944,
      "loss": 5.263,
      "mean_token_accuracy": 0.1728105887770653,
      "num_tokens": 11216557.0,
      "step": 6095
    },
    {
      "entropy": 5.535755586624146,
      "epoch": 0.5095222185098564,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004979548954851666,
      "loss": 5.4028,
      "mean_token_accuracy": 0.17514478415250778,
      "num_tokens": 11227277.0,
      "step": 6100
    },
    {
      "entropy": 5.391517972946167,
      "epoch": 0.5099398596725693,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000497950888825345,
      "loss": 5.1965,
      "mean_token_accuracy": 0.18890394270420074,
      "num_tokens": 11236226.0,
      "step": 6105
    },
    {
      "entropy": 5.3458089351654055,
      "epoch": 0.5103575008352823,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004979468782624996,
      "loss": 5.3087,
      "mean_token_accuracy": 0.18111661821603775,
      "num_tokens": 11245246.0,
      "step": 6110
    },
    {
      "entropy": 5.333164882659912,
      "epoch": 0.5107751419979953,
      "grad_norm": 1.375,
      "learning_rate": 0.0004979428637967009,
      "loss": 5.3793,
      "mean_token_accuracy": 0.1781592220067978,
      "num_tokens": 11254437.0,
      "step": 6115
    },
    {
      "entropy": 5.41632308959961,
      "epoch": 0.5111927831607083,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004979388454280189,
      "loss": 5.2039,
      "mean_token_accuracy": 0.19035724103450774,
      "num_tokens": 11262413.0,
      "step": 6120
    },
    {
      "entropy": 5.405248832702637,
      "epoch": 0.5116104243234213,
      "grad_norm": 1.5546875,
      "learning_rate": 0.000497934823156524,
      "loss": 5.396,
      "mean_token_accuracy": 0.17728246599435807,
      "num_tokens": 11272061.0,
      "step": 6125
    },
    {
      "entropy": 5.3317732334136965,
      "epoch": 0.5120280654861343,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004979307969822869,
      "loss": 5.2133,
      "mean_token_accuracy": 0.18845614194869995,
      "num_tokens": 11281623.0,
      "step": 6130
    },
    {
      "entropy": 5.251134586334229,
      "epoch": 0.5124457066488474,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004979267669053776,
      "loss": 5.0858,
      "mean_token_accuracy": 0.19723618477582933,
      "num_tokens": 11290202.0,
      "step": 6135
    },
    {
      "entropy": 5.342445135116577,
      "epoch": 0.5128633478115603,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004979227329258671,
      "loss": 5.372,
      "mean_token_accuracy": 0.18314483612775803,
      "num_tokens": 11299040.0,
      "step": 6140
    },
    {
      "entropy": 5.418828630447388,
      "epoch": 0.5132809889742733,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004979186950438256,
      "loss": 5.2915,
      "mean_token_accuracy": 0.18070457875728607,
      "num_tokens": 11307938.0,
      "step": 6145
    },
    {
      "entropy": 5.378529262542725,
      "epoch": 0.5136986301369864,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004979146532593243,
      "loss": 5.2766,
      "mean_token_accuracy": 0.18071874380111694,
      "num_tokens": 11316915.0,
      "step": 6150
    },
    {
      "entropy": 5.346374654769898,
      "epoch": 0.5141162712996993,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004979106075724334,
      "loss": 5.2551,
      "mean_token_accuracy": 0.176848366856575,
      "num_tokens": 11326634.0,
      "step": 6155
    },
    {
      "entropy": 5.443122911453247,
      "epoch": 0.5145339124624123,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004979065579832242,
      "loss": 5.4262,
      "mean_token_accuracy": 0.1679534450173378,
      "num_tokens": 11336502.0,
      "step": 6160
    },
    {
      "entropy": 5.377861785888672,
      "epoch": 0.5149515536251253,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004979025044917673,
      "loss": 5.3264,
      "mean_token_accuracy": 0.1782236009836197,
      "num_tokens": 11346125.0,
      "step": 6165
    },
    {
      "entropy": 5.454143762588501,
      "epoch": 0.5153691947878383,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004978984470981339,
      "loss": 5.4072,
      "mean_token_accuracy": 0.16891748160123826,
      "num_tokens": 11355713.0,
      "step": 6170
    },
    {
      "entropy": 5.386297178268433,
      "epoch": 0.5157868359505513,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004978943858023946,
      "loss": 5.2348,
      "mean_token_accuracy": 0.1821313038468361,
      "num_tokens": 11364726.0,
      "step": 6175
    },
    {
      "entropy": 5.37279748916626,
      "epoch": 0.5162044771132643,
      "grad_norm": 1.34375,
      "learning_rate": 0.000497890320604621,
      "loss": 5.2176,
      "mean_token_accuracy": 0.1816943421959877,
      "num_tokens": 11373779.0,
      "step": 6180
    },
    {
      "entropy": 5.277878665924073,
      "epoch": 0.5166221182759773,
      "grad_norm": 1.359375,
      "learning_rate": 0.000497886251504884,
      "loss": 5.2351,
      "mean_token_accuracy": 0.18144305646419526,
      "num_tokens": 11382641.0,
      "step": 6185
    },
    {
      "entropy": 5.302442121505737,
      "epoch": 0.5170397594386903,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004978821785032549,
      "loss": 5.1861,
      "mean_token_accuracy": 0.18470952212810515,
      "num_tokens": 11392200.0,
      "step": 6190
    },
    {
      "entropy": 5.387860345840454,
      "epoch": 0.5174574006014033,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004978781015998048,
      "loss": 5.3157,
      "mean_token_accuracy": 0.17309316545724868,
      "num_tokens": 11401884.0,
      "step": 6195
    },
    {
      "entropy": 5.348043489456177,
      "epoch": 0.5178750417641163,
      "grad_norm": 1.375,
      "learning_rate": 0.0004978740207946056,
      "loss": 5.2126,
      "mean_token_accuracy": 0.1886736273765564,
      "num_tokens": 11410710.0,
      "step": 6200
    },
    {
      "entropy": 5.267181253433227,
      "epoch": 0.5182926829268293,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004978699360877281,
      "loss": 5.1645,
      "mean_token_accuracy": 0.19061697721481324,
      "num_tokens": 11419622.0,
      "step": 6205
    },
    {
      "entropy": 5.267860460281372,
      "epoch": 0.5187103240895423,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004978658474792442,
      "loss": 5.1691,
      "mean_token_accuracy": 0.19042179584503174,
      "num_tokens": 11427631.0,
      "step": 6210
    },
    {
      "entropy": 5.332252073287964,
      "epoch": 0.5191279652522552,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004978617549692253,
      "loss": 5.2481,
      "mean_token_accuracy": 0.1809620276093483,
      "num_tokens": 11437630.0,
      "step": 6215
    },
    {
      "entropy": 5.458483409881592,
      "epoch": 0.5195456064149683,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004978576585577432,
      "loss": 5.4633,
      "mean_token_accuracy": 0.1654739946126938,
      "num_tokens": 11447962.0,
      "step": 6220
    },
    {
      "entropy": 5.490985774993897,
      "epoch": 0.5199632475776813,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004978535582448695,
      "loss": 5.3176,
      "mean_token_accuracy": 0.17904573380947114,
      "num_tokens": 11458154.0,
      "step": 6225
    },
    {
      "entropy": 5.315806674957275,
      "epoch": 0.5203808887403942,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004978494540306761,
      "loss": 5.1594,
      "mean_token_accuracy": 0.18527811467647554,
      "num_tokens": 11467449.0,
      "step": 6230
    },
    {
      "entropy": 5.3177930355072025,
      "epoch": 0.5207985299031073,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004978453459152346,
      "loss": 5.4445,
      "mean_token_accuracy": 0.17112648785114287,
      "num_tokens": 11477775.0,
      "step": 6235
    },
    {
      "entropy": 5.40331506729126,
      "epoch": 0.5212161710658203,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004978412338986172,
      "loss": 5.2034,
      "mean_token_accuracy": 0.18052331507205963,
      "num_tokens": 11486921.0,
      "step": 6240
    },
    {
      "entropy": 5.360177278518677,
      "epoch": 0.5216338122285332,
      "grad_norm": 1.90625,
      "learning_rate": 0.0004978371179808958,
      "loss": 5.2395,
      "mean_token_accuracy": 0.18757482320070268,
      "num_tokens": 11496562.0,
      "step": 6245
    },
    {
      "entropy": 5.310189580917358,
      "epoch": 0.5220514533912463,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004978329981621424,
      "loss": 5.1893,
      "mean_token_accuracy": 0.18194289207458497,
      "num_tokens": 11504906.0,
      "step": 6250
    },
    {
      "entropy": 5.298901605606079,
      "epoch": 0.5224690945539593,
      "grad_norm": 1.3125,
      "learning_rate": 0.000497828874442429,
      "loss": 5.2343,
      "mean_token_accuracy": 0.18324219882488252,
      "num_tokens": 11513757.0,
      "step": 6255
    },
    {
      "entropy": 5.2840211391448975,
      "epoch": 0.5228867357166722,
      "grad_norm": 1.34375,
      "learning_rate": 0.000497824746821828,
      "loss": 5.2653,
      "mean_token_accuracy": 0.17947673201560974,
      "num_tokens": 11523195.0,
      "step": 6260
    },
    {
      "entropy": 5.385505962371826,
      "epoch": 0.5233043768793852,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004978206153004117,
      "loss": 5.2997,
      "mean_token_accuracy": 0.180406354367733,
      "num_tokens": 11532348.0,
      "step": 6265
    },
    {
      "entropy": 5.402559614181518,
      "epoch": 0.5237220180420983,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004978164798782523,
      "loss": 5.2936,
      "mean_token_accuracy": 0.1870359420776367,
      "num_tokens": 11542356.0,
      "step": 6270
    },
    {
      "entropy": 5.347551870346069,
      "epoch": 0.5241396592048112,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004978123405554221,
      "loss": 5.2981,
      "mean_token_accuracy": 0.18448646366596222,
      "num_tokens": 11550914.0,
      "step": 6275
    },
    {
      "entropy": 5.362148094177246,
      "epoch": 0.5245573003675242,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004978081973319938,
      "loss": 5.2482,
      "mean_token_accuracy": 0.18454992920160293,
      "num_tokens": 11559314.0,
      "step": 6280
    },
    {
      "entropy": 5.28530478477478,
      "epoch": 0.5249749415302373,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004978040502080398,
      "loss": 5.2172,
      "mean_token_accuracy": 0.1808277413249016,
      "num_tokens": 11567990.0,
      "step": 6285
    },
    {
      "entropy": 5.3501801013946535,
      "epoch": 0.5253925826929502,
      "grad_norm": 2.671875,
      "learning_rate": 0.0004977998991836327,
      "loss": 5.2573,
      "mean_token_accuracy": 0.17986629903316498,
      "num_tokens": 11577133.0,
      "step": 6290
    },
    {
      "entropy": 5.369371223449707,
      "epoch": 0.5258102238556632,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004977957442588451,
      "loss": 5.3297,
      "mean_token_accuracy": 0.17925186157226564,
      "num_tokens": 11585330.0,
      "step": 6295
    },
    {
      "entropy": 5.421820878982544,
      "epoch": 0.5262278650183763,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004977915854337498,
      "loss": 5.2774,
      "mean_token_accuracy": 0.17785076349973677,
      "num_tokens": 11595005.0,
      "step": 6300
    },
    {
      "entropy": 5.323221492767334,
      "epoch": 0.5266455061810892,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004977874227084197,
      "loss": 5.1651,
      "mean_token_accuracy": 0.1909732237458229,
      "num_tokens": 11603969.0,
      "step": 6305
    },
    {
      "entropy": 5.273255443572998,
      "epoch": 0.5270631473438022,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004977832560829277,
      "loss": 5.2189,
      "mean_token_accuracy": 0.18179108798503876,
      "num_tokens": 11613118.0,
      "step": 6310
    },
    {
      "entropy": 5.305721616744995,
      "epoch": 0.5274807885065153,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004977790855573464,
      "loss": 5.147,
      "mean_token_accuracy": 0.18721737563610077,
      "num_tokens": 11621799.0,
      "step": 6315
    },
    {
      "entropy": 5.288804006576538,
      "epoch": 0.5278984296692282,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004977749111317492,
      "loss": 5.2437,
      "mean_token_accuracy": 0.18319330960512162,
      "num_tokens": 11630557.0,
      "step": 6320
    },
    {
      "entropy": 5.36851134300232,
      "epoch": 0.5283160708319412,
      "grad_norm": 1.421875,
      "learning_rate": 0.000497770732806209,
      "loss": 5.2609,
      "mean_token_accuracy": 0.18259372413158417,
      "num_tokens": 11639179.0,
      "step": 6325
    },
    {
      "entropy": 5.415849304199218,
      "epoch": 0.5287337119946542,
      "grad_norm": 1.5234375,
      "learning_rate": 0.000497766550580799,
      "loss": 5.3701,
      "mean_token_accuracy": 0.17368093580007554,
      "num_tokens": 11648273.0,
      "step": 6330
    },
    {
      "entropy": 5.332431697845459,
      "epoch": 0.5291513531573672,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004977623644555923,
      "loss": 5.2629,
      "mean_token_accuracy": 0.18116851449012755,
      "num_tokens": 11656372.0,
      "step": 6335
    },
    {
      "entropy": 5.327931547164917,
      "epoch": 0.5295689943200802,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004977581744306624,
      "loss": 5.218,
      "mean_token_accuracy": 0.18387259691953659,
      "num_tokens": 11665007.0,
      "step": 6340
    },
    {
      "entropy": 5.37454047203064,
      "epoch": 0.5299866354827932,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004977539805060824,
      "loss": 5.3286,
      "mean_token_accuracy": 0.18091593384742738,
      "num_tokens": 11674519.0,
      "step": 6345
    },
    {
      "entropy": 5.404172754287719,
      "epoch": 0.5304042766455062,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004977497826819259,
      "loss": 5.2521,
      "mean_token_accuracy": 0.18294285833835602,
      "num_tokens": 11683705.0,
      "step": 6350
    },
    {
      "entropy": 5.350062894821167,
      "epoch": 0.5308219178082192,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004977455809582664,
      "loss": 5.2317,
      "mean_token_accuracy": 0.18395105600357056,
      "num_tokens": 11693152.0,
      "step": 6355
    },
    {
      "entropy": 5.365998888015747,
      "epoch": 0.5312395589709322,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004977413753351772,
      "loss": 5.2489,
      "mean_token_accuracy": 0.17858757078647614,
      "num_tokens": 11702584.0,
      "step": 6360
    },
    {
      "entropy": 5.336689805984497,
      "epoch": 0.5316572001336451,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004977371658127323,
      "loss": 5.2792,
      "mean_token_accuracy": 0.18464104682207108,
      "num_tokens": 11711633.0,
      "step": 6365
    },
    {
      "entropy": 5.288880968093872,
      "epoch": 0.5320748412963582,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004977329523910051,
      "loss": 5.2277,
      "mean_token_accuracy": 0.18570223748683928,
      "num_tokens": 11720116.0,
      "step": 6370
    },
    {
      "entropy": 5.385643768310547,
      "epoch": 0.5324924824590712,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004977287350700695,
      "loss": 5.3146,
      "mean_token_accuracy": 0.17931191474199296,
      "num_tokens": 11729564.0,
      "step": 6375
    },
    {
      "entropy": 5.341831684112549,
      "epoch": 0.5329101236217841,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004977245138499993,
      "loss": 5.179,
      "mean_token_accuracy": 0.19463744014501572,
      "num_tokens": 11737935.0,
      "step": 6380
    },
    {
      "entropy": 5.315127897262573,
      "epoch": 0.5333277647844972,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004977202887308684,
      "loss": 5.1643,
      "mean_token_accuracy": 0.186013887822628,
      "num_tokens": 11745989.0,
      "step": 6385
    },
    {
      "entropy": 5.346630334854126,
      "epoch": 0.5337454059472102,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004977160597127507,
      "loss": 5.36,
      "mean_token_accuracy": 0.18391820639371873,
      "num_tokens": 11755530.0,
      "step": 6390
    },
    {
      "entropy": 5.392025470733643,
      "epoch": 0.5341630471099231,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004977118267957204,
      "loss": 5.3045,
      "mean_token_accuracy": 0.17761826366186143,
      "num_tokens": 11764546.0,
      "step": 6395
    },
    {
      "entropy": 5.4786036014556885,
      "epoch": 0.5345806882726362,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004977075899798513,
      "loss": 5.4267,
      "mean_token_accuracy": 0.17362408041954042,
      "num_tokens": 11774117.0,
      "step": 6400
    },
    {
      "entropy": 5.394445466995239,
      "epoch": 0.5349983294353492,
      "grad_norm": 1.296875,
      "learning_rate": 0.000497703349265218,
      "loss": 5.1905,
      "mean_token_accuracy": 0.19093178063631058,
      "num_tokens": 11782952.0,
      "step": 6405
    },
    {
      "entropy": 5.368710231781006,
      "epoch": 0.5354159705980621,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004976991046518943,
      "loss": 5.3612,
      "mean_token_accuracy": 0.17221387028694152,
      "num_tokens": 11792369.0,
      "step": 6410
    },
    {
      "entropy": 5.316056966781616,
      "epoch": 0.5358336117607752,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004976948561399549,
      "loss": 5.2136,
      "mean_token_accuracy": 0.18843212723731995,
      "num_tokens": 11801365.0,
      "step": 6415
    },
    {
      "entropy": 5.381639766693115,
      "epoch": 0.5362512529234882,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004976906037294738,
      "loss": 5.2888,
      "mean_token_accuracy": 0.18221500366926194,
      "num_tokens": 11809785.0,
      "step": 6420
    },
    {
      "entropy": 5.366283702850342,
      "epoch": 0.5366688940862011,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004976863474205257,
      "loss": 5.1736,
      "mean_token_accuracy": 0.18574026674032212,
      "num_tokens": 11817978.0,
      "step": 6425
    },
    {
      "entropy": 5.356812143325806,
      "epoch": 0.5370865352489141,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004976820872131851,
      "loss": 5.2892,
      "mean_token_accuracy": 0.18172308206558227,
      "num_tokens": 11827796.0,
      "step": 6430
    },
    {
      "entropy": 5.255517959594727,
      "epoch": 0.5375041764116272,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004976778231075266,
      "loss": 5.1449,
      "mean_token_accuracy": 0.18679563701152802,
      "num_tokens": 11836439.0,
      "step": 6435
    },
    {
      "entropy": 5.324428701400757,
      "epoch": 0.5379218175743401,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004976735551036245,
      "loss": 5.3341,
      "mean_token_accuracy": 0.17596723586320878,
      "num_tokens": 11845268.0,
      "step": 6440
    },
    {
      "entropy": 5.403821802139282,
      "epoch": 0.5383394587370531,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000497669283201554,
      "loss": 5.2689,
      "mean_token_accuracy": 0.17970362454652786,
      "num_tokens": 11855185.0,
      "step": 6445
    },
    {
      "entropy": 5.315844249725342,
      "epoch": 0.5387570998997662,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004976650074013895,
      "loss": 5.2925,
      "mean_token_accuracy": 0.18441420644521714,
      "num_tokens": 11865073.0,
      "step": 6450
    },
    {
      "entropy": 5.36966233253479,
      "epoch": 0.5391747410624791,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004976607277032061,
      "loss": 5.2809,
      "mean_token_accuracy": 0.18345875293016434,
      "num_tokens": 11875004.0,
      "step": 6455
    },
    {
      "entropy": 5.3333975315094,
      "epoch": 0.5395923822251921,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004976564441070788,
      "loss": 5.2828,
      "mean_token_accuracy": 0.1773657128214836,
      "num_tokens": 11884905.0,
      "step": 6460
    },
    {
      "entropy": 5.417646837234497,
      "epoch": 0.5400100233879052,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004976521566130823,
      "loss": 5.2855,
      "mean_token_accuracy": 0.1761803388595581,
      "num_tokens": 11894669.0,
      "step": 6465
    },
    {
      "entropy": 5.512452220916748,
      "epoch": 0.5404276645506181,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004976478652212918,
      "loss": 5.4672,
      "mean_token_accuracy": 0.1686572551727295,
      "num_tokens": 11904127.0,
      "step": 6470
    },
    {
      "entropy": 5.4485691547393795,
      "epoch": 0.5408453057133311,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004976435699317824,
      "loss": 5.2688,
      "mean_token_accuracy": 0.18013800978660582,
      "num_tokens": 11912031.0,
      "step": 6475
    },
    {
      "entropy": 5.291551780700684,
      "epoch": 0.5412629468760441,
      "grad_norm": 14.625,
      "learning_rate": 0.0004976392707446293,
      "loss": 5.1936,
      "mean_token_accuracy": 0.19379432648420333,
      "num_tokens": 11920961.0,
      "step": 6480
    },
    {
      "entropy": 5.406322240829468,
      "epoch": 0.5416805880387571,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004976349676599078,
      "loss": 5.3511,
      "mean_token_accuracy": 0.17953124791383743,
      "num_tokens": 11930454.0,
      "step": 6485
    },
    {
      "entropy": 5.31783766746521,
      "epoch": 0.5420982292014701,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004976306606776932,
      "loss": 5.1366,
      "mean_token_accuracy": 0.18430352956056595,
      "num_tokens": 11938610.0,
      "step": 6490
    },
    {
      "entropy": 5.341507530212402,
      "epoch": 0.5425158703641831,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004976263497980609,
      "loss": 5.3085,
      "mean_token_accuracy": 0.18581376671791078,
      "num_tokens": 11947882.0,
      "step": 6495
    },
    {
      "entropy": 5.37786717414856,
      "epoch": 0.5429335115268961,
      "grad_norm": 1.5,
      "learning_rate": 0.0004976220350210864,
      "loss": 5.3043,
      "mean_token_accuracy": 0.17941107898950576,
      "num_tokens": 11957552.0,
      "step": 6500
    },
    {
      "entropy": 5.35598783493042,
      "epoch": 0.5433511526896091,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004976177163468451,
      "loss": 5.2416,
      "mean_token_accuracy": 0.1838558703660965,
      "num_tokens": 11967350.0,
      "step": 6505
    },
    {
      "entropy": 5.350469017028809,
      "epoch": 0.5437687938523221,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004976133937754128,
      "loss": 5.2353,
      "mean_token_accuracy": 0.1827439084649086,
      "num_tokens": 11976107.0,
      "step": 6510
    },
    {
      "entropy": 5.373798465728759,
      "epoch": 0.544186435015035,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004976090673068651,
      "loss": 5.2208,
      "mean_token_accuracy": 0.18852132707834243,
      "num_tokens": 11985059.0,
      "step": 6515
    },
    {
      "entropy": 5.3296410083770756,
      "epoch": 0.5446040761777481,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004976047369412778,
      "loss": 5.2103,
      "mean_token_accuracy": 0.1862734004855156,
      "num_tokens": 11993453.0,
      "step": 6520
    },
    {
      "entropy": 5.236110258102417,
      "epoch": 0.5450217173404611,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004976004026787266,
      "loss": 5.1652,
      "mean_token_accuracy": 0.1915011450648308,
      "num_tokens": 12002524.0,
      "step": 6525
    },
    {
      "entropy": 5.261706590652466,
      "epoch": 0.545439358503174,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004975960645192873,
      "loss": 5.2274,
      "mean_token_accuracy": 0.18687387853860854,
      "num_tokens": 12011516.0,
      "step": 6530
    },
    {
      "entropy": 5.389655351638794,
      "epoch": 0.5458569996658871,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004975917224630359,
      "loss": 5.2799,
      "mean_token_accuracy": 0.17607234567403793,
      "num_tokens": 12020058.0,
      "step": 6535
    },
    {
      "entropy": 5.430317354202271,
      "epoch": 0.5462746408286001,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004975873765100487,
      "loss": 5.3742,
      "mean_token_accuracy": 0.17935390025377274,
      "num_tokens": 12029167.0,
      "step": 6540
    },
    {
      "entropy": 5.413573980331421,
      "epoch": 0.546692281991313,
      "grad_norm": 1.25,
      "learning_rate": 0.0004975830266604013,
      "loss": 5.3049,
      "mean_token_accuracy": 0.18287244141101838,
      "num_tokens": 12038298.0,
      "step": 6545
    },
    {
      "entropy": 5.399274921417236,
      "epoch": 0.5471099231540261,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004975786729141702,
      "loss": 5.2424,
      "mean_token_accuracy": 0.18066378235816954,
      "num_tokens": 12047437.0,
      "step": 6550
    },
    {
      "entropy": 5.372140121459961,
      "epoch": 0.5475275643167391,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004975743152714314,
      "loss": 5.2819,
      "mean_token_accuracy": 0.1853661999106407,
      "num_tokens": 12055821.0,
      "step": 6555
    },
    {
      "entropy": 5.335363817214966,
      "epoch": 0.547945205479452,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004975699537322616,
      "loss": 5.1656,
      "mean_token_accuracy": 0.19055379927158356,
      "num_tokens": 12064378.0,
      "step": 6560
    },
    {
      "entropy": 5.275246906280517,
      "epoch": 0.5483628466421651,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004975655882967366,
      "loss": 5.2227,
      "mean_token_accuracy": 0.18328310996294023,
      "num_tokens": 12073110.0,
      "step": 6565
    },
    {
      "entropy": 5.386295366287231,
      "epoch": 0.5487804878048781,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004975612189649331,
      "loss": 5.253,
      "mean_token_accuracy": 0.18681494295597076,
      "num_tokens": 12081224.0,
      "step": 6570
    },
    {
      "entropy": 5.444203853607178,
      "epoch": 0.549198128967591,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004975568457369276,
      "loss": 5.3016,
      "mean_token_accuracy": 0.1844761922955513,
      "num_tokens": 12090279.0,
      "step": 6575
    },
    {
      "entropy": 5.360468578338623,
      "epoch": 0.549615770130304,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004975524686127966,
      "loss": 5.3047,
      "mean_token_accuracy": 0.1804049789905548,
      "num_tokens": 12098818.0,
      "step": 6580
    },
    {
      "entropy": 5.335679626464843,
      "epoch": 0.5500334112930171,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004975480875926168,
      "loss": 5.2456,
      "mean_token_accuracy": 0.18969240933656692,
      "num_tokens": 12107358.0,
      "step": 6585
    },
    {
      "entropy": 5.354188394546509,
      "epoch": 0.55045105245573,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004975437026764648,
      "loss": 5.3011,
      "mean_token_accuracy": 0.17559485733509064,
      "num_tokens": 12116802.0,
      "step": 6590
    },
    {
      "entropy": 5.372743368148804,
      "epoch": 0.550868693618443,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004975393138644175,
      "loss": 5.2239,
      "mean_token_accuracy": 0.18506044149398804,
      "num_tokens": 12126347.0,
      "step": 6595
    },
    {
      "entropy": 5.2828162670135494,
      "epoch": 0.5512863347811561,
      "grad_norm": 1.5,
      "learning_rate": 0.0004975349211565515,
      "loss": 5.2015,
      "mean_token_accuracy": 0.1851177468895912,
      "num_tokens": 12136138.0,
      "step": 6600
    },
    {
      "entropy": 5.269084787368774,
      "epoch": 0.551703975943869,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000497530524552944,
      "loss": 5.2677,
      "mean_token_accuracy": 0.18324491530656814,
      "num_tokens": 12146436.0,
      "step": 6605
    },
    {
      "entropy": 5.394599580764771,
      "epoch": 0.552121617106582,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004975261240536716,
      "loss": 5.2683,
      "mean_token_accuracy": 0.1781994119286537,
      "num_tokens": 12157003.0,
      "step": 6610
    },
    {
      "entropy": 5.396362686157227,
      "epoch": 0.5525392582692951,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004975217196588117,
      "loss": 5.2153,
      "mean_token_accuracy": 0.19297451823949813,
      "num_tokens": 12166671.0,
      "step": 6615
    },
    {
      "entropy": 5.309810495376587,
      "epoch": 0.552956899432008,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004975173113684413,
      "loss": 5.2322,
      "mean_token_accuracy": 0.18909222781658172,
      "num_tokens": 12176474.0,
      "step": 6620
    },
    {
      "entropy": 5.369884920120239,
      "epoch": 0.553374540594721,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004975128991826374,
      "loss": 5.2496,
      "mean_token_accuracy": 0.18444299697875977,
      "num_tokens": 12185307.0,
      "step": 6625
    },
    {
      "entropy": 5.3402153015136715,
      "epoch": 0.553792181757434,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004975084831014774,
      "loss": 5.2215,
      "mean_token_accuracy": 0.18116243183612823,
      "num_tokens": 12195179.0,
      "step": 6630
    },
    {
      "entropy": 5.241352272033692,
      "epoch": 0.554209822920147,
      "grad_norm": 1.9375,
      "learning_rate": 0.0004975040631250386,
      "loss": 5.2909,
      "mean_token_accuracy": 0.1890150398015976,
      "num_tokens": 12204083.0,
      "step": 6635
    },
    {
      "entropy": 5.283638286590576,
      "epoch": 0.55462746408286,
      "grad_norm": 1.5,
      "learning_rate": 0.0004974996392533983,
      "loss": 5.2094,
      "mean_token_accuracy": 0.19065252989530562,
      "num_tokens": 12213322.0,
      "step": 6640
    },
    {
      "entropy": 5.343265342712402,
      "epoch": 0.555045105245573,
      "grad_norm": 1.4375,
      "learning_rate": 0.000497495211486634,
      "loss": 5.2042,
      "mean_token_accuracy": 0.18546016663312911,
      "num_tokens": 12222305.0,
      "step": 6645
    },
    {
      "entropy": 5.381438827514648,
      "epoch": 0.555462746408286,
      "grad_norm": 1.625,
      "learning_rate": 0.0004974907798248232,
      "loss": 5.3211,
      "mean_token_accuracy": 0.17728150337934495,
      "num_tokens": 12232196.0,
      "step": 6650
    },
    {
      "entropy": 5.40497350692749,
      "epoch": 0.555880387570999,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004974863442680434,
      "loss": 5.2544,
      "mean_token_accuracy": 0.1843208283185959,
      "num_tokens": 12241914.0,
      "step": 6655
    },
    {
      "entropy": 5.254424905776977,
      "epoch": 0.556298028733712,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004974819048163725,
      "loss": 5.1286,
      "mean_token_accuracy": 0.1960676372051239,
      "num_tokens": 12250185.0,
      "step": 6660
    },
    {
      "entropy": 5.338935422897339,
      "epoch": 0.556715669896425,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000497477461469888,
      "loss": 5.311,
      "mean_token_accuracy": 0.18130231350660325,
      "num_tokens": 12258954.0,
      "step": 6665
    },
    {
      "entropy": 5.422602319717408,
      "epoch": 0.557133311059138,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004974730142286677,
      "loss": 5.3484,
      "mean_token_accuracy": 0.17598465085029602,
      "num_tokens": 12268320.0,
      "step": 6670
    },
    {
      "entropy": 5.267890453338623,
      "epoch": 0.557550952221851,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004974685630927895,
      "loss": 5.2085,
      "mean_token_accuracy": 0.1799112319946289,
      "num_tokens": 12276804.0,
      "step": 6675
    },
    {
      "entropy": 5.440449333190918,
      "epoch": 0.557968593384564,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004974641080623314,
      "loss": 5.2744,
      "mean_token_accuracy": 0.17954966872930528,
      "num_tokens": 12286024.0,
      "step": 6680
    },
    {
      "entropy": 5.4408424377441404,
      "epoch": 0.558386234547277,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004974596491373711,
      "loss": 5.3013,
      "mean_token_accuracy": 0.18255170434713364,
      "num_tokens": 12295154.0,
      "step": 6685
    },
    {
      "entropy": 5.397327518463134,
      "epoch": 0.55880387570999,
      "grad_norm": 1.5,
      "learning_rate": 0.0004974551863179871,
      "loss": 5.3663,
      "mean_token_accuracy": 0.1708329826593399,
      "num_tokens": 12306349.0,
      "step": 6690
    },
    {
      "entropy": 5.368606328964233,
      "epoch": 0.5592215168727029,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004974507196042572,
      "loss": 5.3065,
      "mean_token_accuracy": 0.18280256688594818,
      "num_tokens": 12315496.0,
      "step": 6695
    },
    {
      "entropy": 5.2260963916778564,
      "epoch": 0.559639158035416,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004974462489962597,
      "loss": 5.1566,
      "mean_token_accuracy": 0.18630271255970002,
      "num_tokens": 12325070.0,
      "step": 6700
    },
    {
      "entropy": 5.328977012634278,
      "epoch": 0.560056799198129,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004974417744940729,
      "loss": 5.2401,
      "mean_token_accuracy": 0.17674152851104735,
      "num_tokens": 12333509.0,
      "step": 6705
    },
    {
      "entropy": 5.396507406234742,
      "epoch": 0.5604744403608419,
      "grad_norm": 1.40625,
      "learning_rate": 0.000497437296097775,
      "loss": 5.3019,
      "mean_token_accuracy": 0.18092448711395265,
      "num_tokens": 12342495.0,
      "step": 6710
    },
    {
      "entropy": 5.354132032394409,
      "epoch": 0.560892081523555,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004974328138074445,
      "loss": 5.1923,
      "mean_token_accuracy": 0.18410516530275345,
      "num_tokens": 12351560.0,
      "step": 6715
    },
    {
      "entropy": 5.321539878845215,
      "epoch": 0.561309722686268,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004974283276231599,
      "loss": 5.3202,
      "mean_token_accuracy": 0.18035029768943786,
      "num_tokens": 12360773.0,
      "step": 6720
    },
    {
      "entropy": 5.375478458404541,
      "epoch": 0.5617273638489809,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004974238375449996,
      "loss": 5.3132,
      "mean_token_accuracy": 0.17522581219673156,
      "num_tokens": 12370519.0,
      "step": 6725
    },
    {
      "entropy": 5.3802210807800295,
      "epoch": 0.562145005011694,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004974193435730424,
      "loss": 5.3093,
      "mean_token_accuracy": 0.1791001543402672,
      "num_tokens": 12380563.0,
      "step": 6730
    },
    {
      "entropy": 5.437316846847534,
      "epoch": 0.562562646174407,
      "grad_norm": 1.625,
      "learning_rate": 0.0004974148457073668,
      "loss": 5.2999,
      "mean_token_accuracy": 0.18318154364824296,
      "num_tokens": 12389291.0,
      "step": 6735
    },
    {
      "entropy": 5.401554870605469,
      "epoch": 0.5629802873371199,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004974103439480517,
      "loss": 5.2931,
      "mean_token_accuracy": 0.18205004781484604,
      "num_tokens": 12399037.0,
      "step": 6740
    },
    {
      "entropy": 5.295340442657471,
      "epoch": 0.563397928499833,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004974058382951757,
      "loss": 5.0877,
      "mean_token_accuracy": 0.1885618582367897,
      "num_tokens": 12407553.0,
      "step": 6745
    },
    {
      "entropy": 5.246711587905883,
      "epoch": 0.563815569662546,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004974013287488177,
      "loss": 5.1679,
      "mean_token_accuracy": 0.1863757476210594,
      "num_tokens": 12416298.0,
      "step": 6750
    },
    {
      "entropy": 5.3530566692352295,
      "epoch": 0.5642332108252589,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004973968153090569,
      "loss": 5.3277,
      "mean_token_accuracy": 0.17042685747146608,
      "num_tokens": 12425918.0,
      "step": 6755
    },
    {
      "entropy": 5.436422061920166,
      "epoch": 0.5646508519879719,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004973922979759722,
      "loss": 5.2571,
      "mean_token_accuracy": 0.18476276099681854,
      "num_tokens": 12434693.0,
      "step": 6760
    },
    {
      "entropy": 5.2864480972290036,
      "epoch": 0.565068493150685,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004973877767496424,
      "loss": 5.1198,
      "mean_token_accuracy": 0.18800848424434663,
      "num_tokens": 12444055.0,
      "step": 6765
    },
    {
      "entropy": 5.280451107025146,
      "epoch": 0.5654861343133979,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004973832516301472,
      "loss": 5.1922,
      "mean_token_accuracy": 0.1876409873366356,
      "num_tokens": 12453012.0,
      "step": 6770
    },
    {
      "entropy": 5.295413780212402,
      "epoch": 0.5659037754761109,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004973787226175652,
      "loss": 5.2304,
      "mean_token_accuracy": 0.18886158466339112,
      "num_tokens": 12462292.0,
      "step": 6775
    },
    {
      "entropy": 5.292002534866333,
      "epoch": 0.566321416638824,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004973741897119761,
      "loss": 5.1956,
      "mean_token_accuracy": 0.19201604574918746,
      "num_tokens": 12471178.0,
      "step": 6780
    },
    {
      "entropy": 5.315654134750366,
      "epoch": 0.5667390578015369,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004973696529134592,
      "loss": 5.2287,
      "mean_token_accuracy": 0.18952189832925798,
      "num_tokens": 12479977.0,
      "step": 6785
    },
    {
      "entropy": 5.33144588470459,
      "epoch": 0.5671566989642499,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004973651122220938,
      "loss": 5.2977,
      "mean_token_accuracy": 0.18171906769275664,
      "num_tokens": 12489004.0,
      "step": 6790
    },
    {
      "entropy": 5.384362840652466,
      "epoch": 0.567574340126963,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004973605676379595,
      "loss": 5.2938,
      "mean_token_accuracy": 0.1782091274857521,
      "num_tokens": 12498979.0,
      "step": 6795
    },
    {
      "entropy": 5.387359476089477,
      "epoch": 0.5679919812896759,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004973560191611357,
      "loss": 5.3494,
      "mean_token_accuracy": 0.17661321610212327,
      "num_tokens": 12508293.0,
      "step": 6800
    },
    {
      "entropy": 5.517708206176758,
      "epoch": 0.5684096224523889,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004973514667917021,
      "loss": 5.3591,
      "mean_token_accuracy": 0.17277874499559404,
      "num_tokens": 12517828.0,
      "step": 6805
    },
    {
      "entropy": 5.382287263870239,
      "epoch": 0.5688272636151019,
      "grad_norm": 2.578125,
      "learning_rate": 0.0004973469105297386,
      "loss": 5.2169,
      "mean_token_accuracy": 0.1828594297170639,
      "num_tokens": 12526569.0,
      "step": 6810
    },
    {
      "entropy": 5.34793701171875,
      "epoch": 0.5692449047778149,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004973423503753247,
      "loss": 5.3519,
      "mean_token_accuracy": 0.17847722619771958,
      "num_tokens": 12535706.0,
      "step": 6815
    },
    {
      "entropy": 5.418070983886719,
      "epoch": 0.5696625459405279,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004973377863285403,
      "loss": 5.2238,
      "mean_token_accuracy": 0.18181915283203126,
      "num_tokens": 12544689.0,
      "step": 6820
    },
    {
      "entropy": 5.332527208328247,
      "epoch": 0.5700801871032409,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004973332183894653,
      "loss": 5.2291,
      "mean_token_accuracy": 0.18357685655355455,
      "num_tokens": 12553318.0,
      "step": 6825
    },
    {
      "entropy": 5.286745500564575,
      "epoch": 0.5704978282659539,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004973286465581798,
      "loss": 5.1869,
      "mean_token_accuracy": 0.19722501188516617,
      "num_tokens": 12562009.0,
      "step": 6830
    },
    {
      "entropy": 5.440903806686402,
      "epoch": 0.5709154694286669,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004973240708347637,
      "loss": 5.3354,
      "mean_token_accuracy": 0.18447528183460235,
      "num_tokens": 12571043.0,
      "step": 6835
    },
    {
      "entropy": 5.487530755996704,
      "epoch": 0.5713331105913799,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004973194912192971,
      "loss": 5.264,
      "mean_token_accuracy": 0.18016863763332366,
      "num_tokens": 12581151.0,
      "step": 6840
    },
    {
      "entropy": 5.394162034988403,
      "epoch": 0.5717507517540928,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004973149077118602,
      "loss": 5.2894,
      "mean_token_accuracy": 0.18303738087415694,
      "num_tokens": 12590671.0,
      "step": 6845
    },
    {
      "entropy": 5.329015398025513,
      "epoch": 0.5721683929168059,
      "grad_norm": 1.25,
      "learning_rate": 0.0004973103203125333,
      "loss": 5.3212,
      "mean_token_accuracy": 0.18117813766002655,
      "num_tokens": 12599693.0,
      "step": 6850
    },
    {
      "entropy": 5.351763391494751,
      "epoch": 0.5725860340795189,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004973057290213965,
      "loss": 5.2291,
      "mean_token_accuracy": 0.18840961754322053,
      "num_tokens": 12609238.0,
      "step": 6855
    },
    {
      "entropy": 5.418849325180053,
      "epoch": 0.5730036752422318,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004973011338385304,
      "loss": 5.2352,
      "mean_token_accuracy": 0.18453244715929032,
      "num_tokens": 12618273.0,
      "step": 6860
    },
    {
      "entropy": 5.37387866973877,
      "epoch": 0.5734213164049449,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004972965347640154,
      "loss": 5.2834,
      "mean_token_accuracy": 0.17970575392246246,
      "num_tokens": 12628229.0,
      "step": 6865
    },
    {
      "entropy": 5.366715145111084,
      "epoch": 0.5738389575676579,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000497291931797932,
      "loss": 5.1776,
      "mean_token_accuracy": 0.19457005858421325,
      "num_tokens": 12636533.0,
      "step": 6870
    },
    {
      "entropy": 5.288800144195557,
      "epoch": 0.5742565987303708,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004972873249403608,
      "loss": 5.245,
      "mean_token_accuracy": 0.18472814112901687,
      "num_tokens": 12645503.0,
      "step": 6875
    },
    {
      "entropy": 5.354816198348999,
      "epoch": 0.5746742398930839,
      "grad_norm": 1.5,
      "learning_rate": 0.0004972827141913822,
      "loss": 5.2285,
      "mean_token_accuracy": 0.18137926906347274,
      "num_tokens": 12654355.0,
      "step": 6880
    },
    {
      "entropy": 5.415995168685913,
      "epoch": 0.5750918810557969,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004972780995510773,
      "loss": 5.2232,
      "mean_token_accuracy": 0.18634467124938964,
      "num_tokens": 12663774.0,
      "step": 6885
    },
    {
      "entropy": 5.2255223274230955,
      "epoch": 0.5755095222185098,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004972734810195266,
      "loss": 5.252,
      "mean_token_accuracy": 0.18272623121738435,
      "num_tokens": 12674436.0,
      "step": 6890
    },
    {
      "entropy": 5.394057369232177,
      "epoch": 0.5759271633812229,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004972688585968112,
      "loss": 5.1898,
      "mean_token_accuracy": 0.1799162283539772,
      "num_tokens": 12682891.0,
      "step": 6895
    },
    {
      "entropy": 5.271082019805908,
      "epoch": 0.5763448045439359,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004972642322830118,
      "loss": 5.0987,
      "mean_token_accuracy": 0.19736090451478958,
      "num_tokens": 12691170.0,
      "step": 6900
    },
    {
      "entropy": 5.26110053062439,
      "epoch": 0.5767624457066488,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004972596020782096,
      "loss": 5.1984,
      "mean_token_accuracy": 0.1865182638168335,
      "num_tokens": 12700314.0,
      "step": 6905
    },
    {
      "entropy": 5.320359992980957,
      "epoch": 0.5771800868693618,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004972549679824853,
      "loss": 5.2574,
      "mean_token_accuracy": 0.18710872828960418,
      "num_tokens": 12709418.0,
      "step": 6910
    },
    {
      "entropy": 5.372145700454712,
      "epoch": 0.5775977280320749,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004972503299959204,
      "loss": 5.2205,
      "mean_token_accuracy": 0.17934612184762955,
      "num_tokens": 12719287.0,
      "step": 6915
    },
    {
      "entropy": 5.327325820922852,
      "epoch": 0.5780153691947878,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000497245688118596,
      "loss": 5.1637,
      "mean_token_accuracy": 0.19210369884967804,
      "num_tokens": 12728836.0,
      "step": 6920
    },
    {
      "entropy": 5.356794595718384,
      "epoch": 0.5784330103575008,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004972410423505932,
      "loss": 5.276,
      "mean_token_accuracy": 0.17989090979099273,
      "num_tokens": 12738173.0,
      "step": 6925
    },
    {
      "entropy": 5.308476448059082,
      "epoch": 0.5788506515202139,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004972363926919935,
      "loss": 5.2692,
      "mean_token_accuracy": 0.18660370409488677,
      "num_tokens": 12747428.0,
      "step": 6930
    },
    {
      "entropy": 5.349858331680298,
      "epoch": 0.5792682926829268,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004972317391428782,
      "loss": 5.3145,
      "mean_token_accuracy": 0.17450422048568726,
      "num_tokens": 12757148.0,
      "step": 6935
    },
    {
      "entropy": 5.4827470779418945,
      "epoch": 0.5796859338456398,
      "grad_norm": 1.375,
      "learning_rate": 0.0004972270817033288,
      "loss": 5.3398,
      "mean_token_accuracy": 0.17227669209241867,
      "num_tokens": 12766820.0,
      "step": 6940
    },
    {
      "entropy": 5.4383704662323,
      "epoch": 0.5801035750083529,
      "grad_norm": 2.546875,
      "learning_rate": 0.0004972224203734269,
      "loss": 5.3232,
      "mean_token_accuracy": 0.17507237195968628,
      "num_tokens": 12776082.0,
      "step": 6945
    },
    {
      "entropy": 5.291722059249878,
      "epoch": 0.5805212161710658,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000497217755153254,
      "loss": 5.3014,
      "mean_token_accuracy": 0.18722181618213654,
      "num_tokens": 12785057.0,
      "step": 6950
    },
    {
      "entropy": 5.311150979995728,
      "epoch": 0.5809388573337788,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004972130860428918,
      "loss": 5.2884,
      "mean_token_accuracy": 0.19013078063726424,
      "num_tokens": 12793599.0,
      "step": 6955
    },
    {
      "entropy": 5.400493526458741,
      "epoch": 0.5813564984964918,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004972084130424222,
      "loss": 5.2757,
      "mean_token_accuracy": 0.17608216851949693,
      "num_tokens": 12803775.0,
      "step": 6960
    },
    {
      "entropy": 5.351313972473145,
      "epoch": 0.5817741396592048,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004972037361519267,
      "loss": 5.2041,
      "mean_token_accuracy": 0.1831551507115364,
      "num_tokens": 12812327.0,
      "step": 6965
    },
    {
      "entropy": 5.352539682388306,
      "epoch": 0.5821917808219178,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004971990553714874,
      "loss": 5.1806,
      "mean_token_accuracy": 0.1919349417090416,
      "num_tokens": 12821218.0,
      "step": 6970
    },
    {
      "entropy": 5.350169467926025,
      "epoch": 0.5826094219846308,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004971943707011861,
      "loss": 5.2666,
      "mean_token_accuracy": 0.18081563264131545,
      "num_tokens": 12831579.0,
      "step": 6975
    },
    {
      "entropy": 5.395135354995728,
      "epoch": 0.5830270631473438,
      "grad_norm": 1.4375,
      "learning_rate": 0.000497189682141105,
      "loss": 5.336,
      "mean_token_accuracy": 0.18033974319696428,
      "num_tokens": 12841026.0,
      "step": 6980
    },
    {
      "entropy": 5.288710975646973,
      "epoch": 0.5834447043100568,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004971849896913259,
      "loss": 5.2635,
      "mean_token_accuracy": 0.17693693190813065,
      "num_tokens": 12850136.0,
      "step": 6985
    },
    {
      "entropy": 5.420898962020874,
      "epoch": 0.5838623454727698,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004971802933519312,
      "loss": 5.3363,
      "mean_token_accuracy": 0.1791166216135025,
      "num_tokens": 12860345.0,
      "step": 6990
    },
    {
      "entropy": 5.33952693939209,
      "epoch": 0.5842799866354828,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004971755931230031,
      "loss": 5.2564,
      "mean_token_accuracy": 0.1843261569738388,
      "num_tokens": 12868995.0,
      "step": 6995
    },
    {
      "entropy": 5.439844036102295,
      "epoch": 0.5846976277981958,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004971708890046236,
      "loss": 5.4114,
      "mean_token_accuracy": 0.17483607530593873,
      "num_tokens": 12878043.0,
      "step": 7000
    },
    {
      "entropy": 5.341176462173462,
      "epoch": 0.5851152689609088,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004971661809968755,
      "loss": 5.2454,
      "mean_token_accuracy": 0.18325130194425582,
      "num_tokens": 12887090.0,
      "step": 7005
    },
    {
      "entropy": 5.326449537277222,
      "epoch": 0.5855329101236217,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004971614690998409,
      "loss": 5.1738,
      "mean_token_accuracy": 0.19785908460617066,
      "num_tokens": 12896349.0,
      "step": 7010
    },
    {
      "entropy": 5.385544300079346,
      "epoch": 0.5859505512863348,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004971567533136024,
      "loss": 5.2267,
      "mean_token_accuracy": 0.1823991373181343,
      "num_tokens": 12905698.0,
      "step": 7015
    },
    {
      "entropy": 5.27918815612793,
      "epoch": 0.5863681924490478,
      "grad_norm": 1.25,
      "learning_rate": 0.0004971520336382424,
      "loss": 5.2251,
      "mean_token_accuracy": 0.1906791776418686,
      "num_tokens": 12915146.0,
      "step": 7020
    },
    {
      "entropy": 5.205868577957153,
      "epoch": 0.5867858336117607,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004971473100738437,
      "loss": 5.0812,
      "mean_token_accuracy": 0.1962924838066101,
      "num_tokens": 12924933.0,
      "step": 7025
    },
    {
      "entropy": 5.396005582809448,
      "epoch": 0.5872034747744738,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000497142582620489,
      "loss": 5.3284,
      "mean_token_accuracy": 0.17793653458356856,
      "num_tokens": 12934959.0,
      "step": 7030
    },
    {
      "entropy": 5.342319345474243,
      "epoch": 0.5876211159371868,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004971378512782609,
      "loss": 5.1724,
      "mean_token_accuracy": 0.19512684494256974,
      "num_tokens": 12943470.0,
      "step": 7035
    },
    {
      "entropy": 5.198559999465942,
      "epoch": 0.5880387570998997,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004971331160472424,
      "loss": 5.1912,
      "mean_token_accuracy": 0.19097486287355422,
      "num_tokens": 12953592.0,
      "step": 7040
    },
    {
      "entropy": 5.412131643295288,
      "epoch": 0.5884563982626128,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004971283769275163,
      "loss": 5.2948,
      "mean_token_accuracy": 0.17394380867481232,
      "num_tokens": 12962893.0,
      "step": 7045
    },
    {
      "entropy": 5.350337505340576,
      "epoch": 0.5888740394253258,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004971236339191656,
      "loss": 5.2017,
      "mean_token_accuracy": 0.18977117091417312,
      "num_tokens": 12971635.0,
      "step": 7050
    },
    {
      "entropy": 5.201319980621338,
      "epoch": 0.5892916805880387,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004971188870222734,
      "loss": 5.1694,
      "mean_token_accuracy": 0.19135574549436568,
      "num_tokens": 12981040.0,
      "step": 7055
    },
    {
      "entropy": 5.355551242828369,
      "epoch": 0.5897093217507517,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004971141362369226,
      "loss": 5.1816,
      "mean_token_accuracy": 0.18889567255973816,
      "num_tokens": 12989898.0,
      "step": 7060
    },
    {
      "entropy": 5.38842282295227,
      "epoch": 0.5901269629134648,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004971093815631965,
      "loss": 5.1727,
      "mean_token_accuracy": 0.18946857899427413,
      "num_tokens": 12998584.0,
      "step": 7065
    },
    {
      "entropy": 5.267561054229736,
      "epoch": 0.5905446040761777,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004971046230011783,
      "loss": 5.2168,
      "mean_token_accuracy": 0.18272117227315904,
      "num_tokens": 13008007.0,
      "step": 7070
    },
    {
      "entropy": 5.3609912395477295,
      "epoch": 0.5909622452388907,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004970998605509514,
      "loss": 5.3365,
      "mean_token_accuracy": 0.1790837287902832,
      "num_tokens": 13017034.0,
      "step": 7075
    },
    {
      "entropy": 5.466132068634034,
      "epoch": 0.5913798864016038,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000497095094212599,
      "loss": 5.398,
      "mean_token_accuracy": 0.1699241131544113,
      "num_tokens": 13025916.0,
      "step": 7080
    },
    {
      "entropy": 5.443122339248657,
      "epoch": 0.5917975275643167,
      "grad_norm": 1.25,
      "learning_rate": 0.0004970903239862047,
      "loss": 5.2852,
      "mean_token_accuracy": 0.1773509755730629,
      "num_tokens": 13035242.0,
      "step": 7085
    },
    {
      "entropy": 5.343547201156616,
      "epoch": 0.5922151687270297,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004970855498718519,
      "loss": 5.2112,
      "mean_token_accuracy": 0.18576087355613707,
      "num_tokens": 13044123.0,
      "step": 7090
    },
    {
      "entropy": 5.339623832702637,
      "epoch": 0.5926328098897428,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004970807718696243,
      "loss": 5.2569,
      "mean_token_accuracy": 0.17808963507413864,
      "num_tokens": 13053503.0,
      "step": 7095
    },
    {
      "entropy": 5.338616991043091,
      "epoch": 0.5930504510524557,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004970759899796055,
      "loss": 5.2362,
      "mean_token_accuracy": 0.1856381982564926,
      "num_tokens": 13062106.0,
      "step": 7100
    },
    {
      "entropy": 5.375444459915161,
      "epoch": 0.5934680922151687,
      "grad_norm": 1.5,
      "learning_rate": 0.0004970712042018791,
      "loss": 5.2402,
      "mean_token_accuracy": 0.18249546736478806,
      "num_tokens": 13071509.0,
      "step": 7105
    },
    {
      "entropy": 5.29193081855774,
      "epoch": 0.5938857333778818,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004970664145365289,
      "loss": 5.1719,
      "mean_token_accuracy": 0.18653131276369095,
      "num_tokens": 13080469.0,
      "step": 7110
    },
    {
      "entropy": 5.317806673049927,
      "epoch": 0.5943033745405947,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000497061620983639,
      "loss": 5.1336,
      "mean_token_accuracy": 0.1925976499915123,
      "num_tokens": 13089826.0,
      "step": 7115
    },
    {
      "entropy": 5.362063884735107,
      "epoch": 0.5947210157033077,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004970568235432931,
      "loss": 5.3293,
      "mean_token_accuracy": 0.17771779298782348,
      "num_tokens": 13098630.0,
      "step": 7120
    },
    {
      "entropy": 5.333346843719482,
      "epoch": 0.5951386568660207,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004970520222155751,
      "loss": 5.2213,
      "mean_token_accuracy": 0.18532659113407135,
      "num_tokens": 13106819.0,
      "step": 7125
    },
    {
      "entropy": 5.388774776458741,
      "epoch": 0.5955562980287337,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004970472170005692,
      "loss": 5.2399,
      "mean_token_accuracy": 0.18373559415340424,
      "num_tokens": 13115471.0,
      "step": 7130
    },
    {
      "entropy": 5.285980701446533,
      "epoch": 0.5959739391914467,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004970424078983596,
      "loss": 5.1147,
      "mean_token_accuracy": 0.19347502142190934,
      "num_tokens": 13124537.0,
      "step": 7135
    },
    {
      "entropy": 5.302227973937988,
      "epoch": 0.5963915803541597,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004970375949090302,
      "loss": 5.2592,
      "mean_token_accuracy": 0.18323099315166474,
      "num_tokens": 13135326.0,
      "step": 7140
    },
    {
      "entropy": 5.431037044525146,
      "epoch": 0.5968092215168727,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004970327780326655,
      "loss": 5.3579,
      "mean_token_accuracy": 0.17709604054689407,
      "num_tokens": 13145502.0,
      "step": 7145
    },
    {
      "entropy": 5.390539216995239,
      "epoch": 0.5972268626795857,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004970279572693498,
      "loss": 5.2538,
      "mean_token_accuracy": 0.1896309569478035,
      "num_tokens": 13154808.0,
      "step": 7150
    },
    {
      "entropy": 5.285022211074829,
      "epoch": 0.5976445038422987,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004970231326191674,
      "loss": 5.3793,
      "mean_token_accuracy": 0.17192819863557815,
      "num_tokens": 13163982.0,
      "step": 7155
    },
    {
      "entropy": 5.379360580444336,
      "epoch": 0.5980621450050116,
      "grad_norm": 1.375,
      "learning_rate": 0.0004970183040822029,
      "loss": 5.2595,
      "mean_token_accuracy": 0.18212141543626786,
      "num_tokens": 13172446.0,
      "step": 7160
    },
    {
      "entropy": 5.3797863006591795,
      "epoch": 0.5984797861677247,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004970134716585407,
      "loss": 5.2232,
      "mean_token_accuracy": 0.17862303853034972,
      "num_tokens": 13181671.0,
      "step": 7165
    },
    {
      "entropy": 5.264983320236206,
      "epoch": 0.5988974273304377,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004970086353482654,
      "loss": 5.1644,
      "mean_token_accuracy": 0.18219546973705292,
      "num_tokens": 13190207.0,
      "step": 7170
    },
    {
      "entropy": 5.283562469482422,
      "epoch": 0.5993150684931506,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004970037951514617,
      "loss": 5.1438,
      "mean_token_accuracy": 0.1897499144077301,
      "num_tokens": 13198866.0,
      "step": 7175
    },
    {
      "entropy": 5.266381168365479,
      "epoch": 0.5997327096558637,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004969989510682143,
      "loss": 5.1849,
      "mean_token_accuracy": 0.19003028124570848,
      "num_tokens": 13207921.0,
      "step": 7180
    },
    {
      "entropy": 5.441741847991944,
      "epoch": 0.6001503508185767,
      "grad_norm": 2.46875,
      "learning_rate": 0.0004969941030986082,
      "loss": 5.3539,
      "mean_token_accuracy": 0.17731461375951768,
      "num_tokens": 13217759.0,
      "step": 7185
    },
    {
      "entropy": 5.310184001922607,
      "epoch": 0.6005679919812896,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000496989251242728,
      "loss": 5.147,
      "mean_token_accuracy": 0.1870572417974472,
      "num_tokens": 13226592.0,
      "step": 7190
    },
    {
      "entropy": 5.2749436378479,
      "epoch": 0.6009856331440027,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004969843955006588,
      "loss": 5.1968,
      "mean_token_accuracy": 0.18484863340854646,
      "num_tokens": 13236123.0,
      "step": 7195
    },
    {
      "entropy": 5.3280675411224365,
      "epoch": 0.6014032743067157,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004969795358724855,
      "loss": 5.3064,
      "mean_token_accuracy": 0.18016528859734535,
      "num_tokens": 13246169.0,
      "step": 7200
    },
    {
      "entropy": 5.3149556636810305,
      "epoch": 0.6018209154694286,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004969746723582932,
      "loss": 5.167,
      "mean_token_accuracy": 0.19074517488479614,
      "num_tokens": 13256141.0,
      "step": 7205
    },
    {
      "entropy": 5.384066438674926,
      "epoch": 0.6022385566321417,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004969698049581672,
      "loss": 5.3445,
      "mean_token_accuracy": 0.18261153548955916,
      "num_tokens": 13265491.0,
      "step": 7210
    },
    {
      "entropy": 5.488926124572754,
      "epoch": 0.6026561977948547,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004969649336721924,
      "loss": 5.3456,
      "mean_token_accuracy": 0.18465238213539123,
      "num_tokens": 13274805.0,
      "step": 7215
    },
    {
      "entropy": 5.40759916305542,
      "epoch": 0.6030738389575676,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004969600585004544,
      "loss": 5.2881,
      "mean_token_accuracy": 0.18210784196853638,
      "num_tokens": 13284589.0,
      "step": 7220
    },
    {
      "entropy": 5.267890977859497,
      "epoch": 0.6034914801202806,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004969551794430383,
      "loss": 5.1208,
      "mean_token_accuracy": 0.19712562263011932,
      "num_tokens": 13293370.0,
      "step": 7225
    },
    {
      "entropy": 5.411610698699951,
      "epoch": 0.6039091212829937,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004969502965000297,
      "loss": 5.3005,
      "mean_token_accuracy": 0.18510427922010422,
      "num_tokens": 13302555.0,
      "step": 7230
    },
    {
      "entropy": 5.279111528396607,
      "epoch": 0.6043267624457066,
      "grad_norm": 1.390625,
      "learning_rate": 0.000496945409671514,
      "loss": 5.2364,
      "mean_token_accuracy": 0.18550454825162888,
      "num_tokens": 13311632.0,
      "step": 7235
    },
    {
      "entropy": 5.320348691940308,
      "epoch": 0.6047444036084196,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004969405189575768,
      "loss": 5.1686,
      "mean_token_accuracy": 0.19313837885856627,
      "num_tokens": 13320593.0,
      "step": 7240
    },
    {
      "entropy": 5.328238487243652,
      "epoch": 0.6051620447711327,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004969356243583036,
      "loss": 5.1539,
      "mean_token_accuracy": 0.19189489483833314,
      "num_tokens": 13329095.0,
      "step": 7245
    },
    {
      "entropy": 5.264534997940063,
      "epoch": 0.6055796859338456,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004969307258737802,
      "loss": 5.2195,
      "mean_token_accuracy": 0.19089837074279786,
      "num_tokens": 13338404.0,
      "step": 7250
    },
    {
      "entropy": 5.287572193145752,
      "epoch": 0.6059973270965586,
      "grad_norm": 1.375,
      "learning_rate": 0.0004969258235040921,
      "loss": 5.1897,
      "mean_token_accuracy": 0.190699964761734,
      "num_tokens": 13347095.0,
      "step": 7255
    },
    {
      "entropy": 5.331546020507813,
      "epoch": 0.6064149682592717,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004969209172493255,
      "loss": 5.248,
      "mean_token_accuracy": 0.18628860563039779,
      "num_tokens": 13356606.0,
      "step": 7260
    },
    {
      "entropy": 5.3542115688323975,
      "epoch": 0.6068326094219846,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004969160071095661,
      "loss": 5.2565,
      "mean_token_accuracy": 0.17968995571136476,
      "num_tokens": 13366889.0,
      "step": 7265
    },
    {
      "entropy": 5.3427972316741945,
      "epoch": 0.6072502505846976,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004969110930848999,
      "loss": 5.2496,
      "mean_token_accuracy": 0.1801399767398834,
      "num_tokens": 13376761.0,
      "step": 7270
    },
    {
      "entropy": 5.345357751846313,
      "epoch": 0.6076678917474106,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004969061751754129,
      "loss": 5.1761,
      "mean_token_accuracy": 0.18946549892425538,
      "num_tokens": 13386226.0,
      "step": 7275
    },
    {
      "entropy": 5.24541597366333,
      "epoch": 0.6080855329101236,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004969012533811911,
      "loss": 5.1109,
      "mean_token_accuracy": 0.19180119186639785,
      "num_tokens": 13394243.0,
      "step": 7280
    },
    {
      "entropy": 5.297079992294312,
      "epoch": 0.6085031740728366,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004968963277023208,
      "loss": 5.3852,
      "mean_token_accuracy": 0.1712947428226471,
      "num_tokens": 13403708.0,
      "step": 7285
    },
    {
      "entropy": 5.289767599105835,
      "epoch": 0.6089208152355496,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004968913981388882,
      "loss": 5.2149,
      "mean_token_accuracy": 0.1913897216320038,
      "num_tokens": 13412911.0,
      "step": 7290
    },
    {
      "entropy": 5.420920896530151,
      "epoch": 0.6093384563982626,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004968864646909797,
      "loss": 5.3699,
      "mean_token_accuracy": 0.18042888194322587,
      "num_tokens": 13422423.0,
      "step": 7295
    },
    {
      "entropy": 5.339856433868408,
      "epoch": 0.6097560975609756,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004968815273586814,
      "loss": 5.2024,
      "mean_token_accuracy": 0.18344418108463287,
      "num_tokens": 13430984.0,
      "step": 7300
    },
    {
      "entropy": 5.313500881195068,
      "epoch": 0.6101737387236886,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004968765861420799,
      "loss": 5.2711,
      "mean_token_accuracy": 0.18218379020690917,
      "num_tokens": 13440492.0,
      "step": 7305
    },
    {
      "entropy": 5.378695392608643,
      "epoch": 0.6105913798864016,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004968716410412617,
      "loss": 5.2371,
      "mean_token_accuracy": 0.1780386224389076,
      "num_tokens": 13449222.0,
      "step": 7310
    },
    {
      "entropy": 5.316129922866821,
      "epoch": 0.6110090210491146,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004968666920563134,
      "loss": 5.2233,
      "mean_token_accuracy": 0.1849824532866478,
      "num_tokens": 13459126.0,
      "step": 7315
    },
    {
      "entropy": 5.447695922851563,
      "epoch": 0.6114266622118276,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004968617391873216,
      "loss": 5.3668,
      "mean_token_accuracy": 0.17276448756456375,
      "num_tokens": 13468811.0,
      "step": 7320
    },
    {
      "entropy": 5.394797897338867,
      "epoch": 0.6118443033745405,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004968567824343729,
      "loss": 5.1539,
      "mean_token_accuracy": 0.18289622515439988,
      "num_tokens": 13477765.0,
      "step": 7325
    },
    {
      "entropy": 5.354283380508423,
      "epoch": 0.6122619445372536,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004968518217975542,
      "loss": 5.2958,
      "mean_token_accuracy": 0.17394291758537292,
      "num_tokens": 13487184.0,
      "step": 7330
    },
    {
      "entropy": 5.325632810592651,
      "epoch": 0.6126795856999666,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004968468572769524,
      "loss": 5.2445,
      "mean_token_accuracy": 0.18050160109996796,
      "num_tokens": 13496784.0,
      "step": 7335
    },
    {
      "entropy": 5.360996294021606,
      "epoch": 0.6130972268626795,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004968418888726543,
      "loss": 5.2756,
      "mean_token_accuracy": 0.1839147076010704,
      "num_tokens": 13505824.0,
      "step": 7340
    },
    {
      "entropy": 5.321667623519898,
      "epoch": 0.6135148680253926,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004968369165847469,
      "loss": 5.1401,
      "mean_token_accuracy": 0.19074717164039612,
      "num_tokens": 13515111.0,
      "step": 7345
    },
    {
      "entropy": 5.265428972244263,
      "epoch": 0.6139325091881056,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004968319404133173,
      "loss": 5.1529,
      "mean_token_accuracy": 0.19337067902088165,
      "num_tokens": 13524163.0,
      "step": 7350
    },
    {
      "entropy": 5.2426989555358885,
      "epoch": 0.6143501503508185,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004968269603584525,
      "loss": 5.1243,
      "mean_token_accuracy": 0.19729869216680526,
      "num_tokens": 13533368.0,
      "step": 7355
    },
    {
      "entropy": 5.348956775665283,
      "epoch": 0.6147677915135316,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004968219764202397,
      "loss": 5.3445,
      "mean_token_accuracy": 0.1793224424123764,
      "num_tokens": 13543708.0,
      "step": 7360
    },
    {
      "entropy": 5.328090858459473,
      "epoch": 0.6151854326762446,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004968169885987663,
      "loss": 5.2711,
      "mean_token_accuracy": 0.178127159178257,
      "num_tokens": 13552561.0,
      "step": 7365
    },
    {
      "entropy": 5.493785858154297,
      "epoch": 0.6156030738389575,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004968119968941193,
      "loss": 5.3535,
      "mean_token_accuracy": 0.17047833055257797,
      "num_tokens": 13562050.0,
      "step": 7370
    },
    {
      "entropy": 5.292812633514404,
      "epoch": 0.6160207150016705,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004968070013063866,
      "loss": 5.1878,
      "mean_token_accuracy": 0.18223292082548143,
      "num_tokens": 13571356.0,
      "step": 7375
    },
    {
      "entropy": 5.280111265182495,
      "epoch": 0.6164383561643836,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004968020018356552,
      "loss": 5.1778,
      "mean_token_accuracy": 0.17824804931879043,
      "num_tokens": 13581357.0,
      "step": 7380
    },
    {
      "entropy": 5.334889364242554,
      "epoch": 0.6168559973270965,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004967969984820128,
      "loss": 5.1663,
      "mean_token_accuracy": 0.18418300598859788,
      "num_tokens": 13589392.0,
      "step": 7385
    },
    {
      "entropy": 5.3280601978302,
      "epoch": 0.6172736384898095,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004967919912455469,
      "loss": 5.1409,
      "mean_token_accuracy": 0.19772139340639114,
      "num_tokens": 13598376.0,
      "step": 7390
    },
    {
      "entropy": 5.216180658340454,
      "epoch": 0.6176912796525226,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004967869801263454,
      "loss": 5.0551,
      "mean_token_accuracy": 0.201987686753273,
      "num_tokens": 13607014.0,
      "step": 7395
    },
    {
      "entropy": 5.273418521881103,
      "epoch": 0.6181089208152355,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004967819651244957,
      "loss": 5.2351,
      "mean_token_accuracy": 0.18187806010246277,
      "num_tokens": 13615700.0,
      "step": 7400
    },
    {
      "entropy": 5.40106315612793,
      "epoch": 0.6185265619779485,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004967769462400857,
      "loss": 5.2765,
      "mean_token_accuracy": 0.1862255647778511,
      "num_tokens": 13625230.0,
      "step": 7405
    },
    {
      "entropy": 5.480621004104615,
      "epoch": 0.6189442031406616,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004967719234732033,
      "loss": 5.4279,
      "mean_token_accuracy": 0.18136966973543167,
      "num_tokens": 13635258.0,
      "step": 7410
    },
    {
      "entropy": 5.23916802406311,
      "epoch": 0.6193618443033745,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004967668968239365,
      "loss": 5.1311,
      "mean_token_accuracy": 0.18889791518449783,
      "num_tokens": 13643688.0,
      "step": 7415
    },
    {
      "entropy": 5.297773933410644,
      "epoch": 0.6197794854660875,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004967618662923732,
      "loss": 5.2874,
      "mean_token_accuracy": 0.18282823413610458,
      "num_tokens": 13652894.0,
      "step": 7420
    },
    {
      "entropy": 5.33356122970581,
      "epoch": 0.6201971266288006,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004967568318786016,
      "loss": 5.2184,
      "mean_token_accuracy": 0.18684618920087814,
      "num_tokens": 13662852.0,
      "step": 7425
    },
    {
      "entropy": 5.36697096824646,
      "epoch": 0.6206147677915135,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004967517935827097,
      "loss": 5.1812,
      "mean_token_accuracy": 0.1854433074593544,
      "num_tokens": 13671457.0,
      "step": 7430
    },
    {
      "entropy": 5.349347496032715,
      "epoch": 0.6210324089542265,
      "grad_norm": 1.375,
      "learning_rate": 0.0004967467514047856,
      "loss": 5.2687,
      "mean_token_accuracy": 0.18863174319267273,
      "num_tokens": 13681002.0,
      "step": 7435
    },
    {
      "entropy": 5.352931785583496,
      "epoch": 0.6214500501169395,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004967417053449178,
      "loss": 5.2267,
      "mean_token_accuracy": 0.18377165794372557,
      "num_tokens": 13690677.0,
      "step": 7440
    },
    {
      "entropy": 5.292918729782104,
      "epoch": 0.6218676912796525,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004967366554031945,
      "loss": 5.0591,
      "mean_token_accuracy": 0.19274694174528123,
      "num_tokens": 13699151.0,
      "step": 7445
    },
    {
      "entropy": 5.2490105628967285,
      "epoch": 0.6222853324423655,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004967316015797042,
      "loss": 5.2427,
      "mean_token_accuracy": 0.18493798971176148,
      "num_tokens": 13708105.0,
      "step": 7450
    },
    {
      "entropy": 5.3417497158050535,
      "epoch": 0.6227029736050785,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004967265438745352,
      "loss": 5.3898,
      "mean_token_accuracy": 0.17622340694069863,
      "num_tokens": 13717094.0,
      "step": 7455
    },
    {
      "entropy": 5.4337269306182865,
      "epoch": 0.6231206147677915,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004967214822877763,
      "loss": 5.2287,
      "mean_token_accuracy": 0.1915486201643944,
      "num_tokens": 13726639.0,
      "step": 7460
    },
    {
      "entropy": 5.335785913467407,
      "epoch": 0.6235382559305045,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004967164168195159,
      "loss": 5.1911,
      "mean_token_accuracy": 0.18517189621925353,
      "num_tokens": 13735220.0,
      "step": 7465
    },
    {
      "entropy": 5.207172584533692,
      "epoch": 0.6239558970932175,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004967113474698428,
      "loss": 5.1302,
      "mean_token_accuracy": 0.19569428712129594,
      "num_tokens": 13743277.0,
      "step": 7470
    },
    {
      "entropy": 5.3403395652771,
      "epoch": 0.6243735382559304,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004967062742388456,
      "loss": 5.2229,
      "mean_token_accuracy": 0.1846537321805954,
      "num_tokens": 13752946.0,
      "step": 7475
    },
    {
      "entropy": 5.3574121475219725,
      "epoch": 0.6247911794186435,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004967011971266132,
      "loss": 5.2699,
      "mean_token_accuracy": 0.18475013822317124,
      "num_tokens": 13761297.0,
      "step": 7480
    },
    {
      "entropy": 5.327942609786987,
      "epoch": 0.6252088205813565,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004966961161332346,
      "loss": 5.2022,
      "mean_token_accuracy": 0.18386684060096742,
      "num_tokens": 13771238.0,
      "step": 7485
    },
    {
      "entropy": 5.364589881896973,
      "epoch": 0.6256264617440696,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004966910312587985,
      "loss": 5.2195,
      "mean_token_accuracy": 0.18339993059635162,
      "num_tokens": 13781420.0,
      "step": 7490
    },
    {
      "entropy": 5.308527183532715,
      "epoch": 0.6260441029067825,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004966859425033941,
      "loss": 5.1926,
      "mean_token_accuracy": 0.1881234809756279,
      "num_tokens": 13789833.0,
      "step": 7495
    },
    {
      "entropy": 5.390172338485717,
      "epoch": 0.6264617440694955,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004966808498671104,
      "loss": 5.3149,
      "mean_token_accuracy": 0.17690399289131165,
      "num_tokens": 13799202.0,
      "step": 7500
    },
    {
      "entropy": 5.399218368530273,
      "epoch": 0.6268793852322085,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004966757533500367,
      "loss": 5.2979,
      "mean_token_accuracy": 0.18487776964902877,
      "num_tokens": 13809097.0,
      "step": 7505
    },
    {
      "entropy": 5.355049562454224,
      "epoch": 0.6272970263949215,
      "grad_norm": 1.5546875,
      "learning_rate": 0.000496670652952262,
      "loss": 5.2038,
      "mean_token_accuracy": 0.18850983381271363,
      "num_tokens": 13818437.0,
      "step": 7510
    },
    {
      "entropy": 5.331247568130493,
      "epoch": 0.6277146675576345,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004966655486738757,
      "loss": 5.2152,
      "mean_token_accuracy": 0.18493678718805312,
      "num_tokens": 13827083.0,
      "step": 7515
    },
    {
      "entropy": 5.306726598739624,
      "epoch": 0.6281323087203475,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004966604405149672,
      "loss": 5.1731,
      "mean_token_accuracy": 0.1843095228075981,
      "num_tokens": 13835905.0,
      "step": 7520
    },
    {
      "entropy": 5.290187835693359,
      "epoch": 0.6285499498830605,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004966553284756257,
      "loss": 5.1087,
      "mean_token_accuracy": 0.19366546273231505,
      "num_tokens": 13845136.0,
      "step": 7525
    },
    {
      "entropy": 5.283191108703614,
      "epoch": 0.6289675910457735,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004966502125559411,
      "loss": 5.1777,
      "mean_token_accuracy": 0.1859496995806694,
      "num_tokens": 13855087.0,
      "step": 7530
    },
    {
      "entropy": 5.367642831802368,
      "epoch": 0.6293852322084865,
      "grad_norm": 1.625,
      "learning_rate": 0.0004966450927560026,
      "loss": 5.2489,
      "mean_token_accuracy": 0.18497766703367233,
      "num_tokens": 13864077.0,
      "step": 7535
    },
    {
      "entropy": 5.348160266876221,
      "epoch": 0.6298028733711994,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004966399690759,
      "loss": 5.2252,
      "mean_token_accuracy": 0.18485676795244216,
      "num_tokens": 13873404.0,
      "step": 7540
    },
    {
      "entropy": 5.4055322170257565,
      "epoch": 0.6302205145339125,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000496634841515723,
      "loss": 5.3175,
      "mean_token_accuracy": 0.18193374574184418,
      "num_tokens": 13883231.0,
      "step": 7545
    },
    {
      "entropy": 5.205918598175049,
      "epoch": 0.6306381556966255,
      "grad_norm": 1.5,
      "learning_rate": 0.0004966297100755613,
      "loss": 5.0591,
      "mean_token_accuracy": 0.1982281342148781,
      "num_tokens": 13892314.0,
      "step": 7550
    },
    {
      "entropy": 5.220934677124023,
      "epoch": 0.6310557968593384,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004966245747555047,
      "loss": 5.1708,
      "mean_token_accuracy": 0.1904767259955406,
      "num_tokens": 13900771.0,
      "step": 7555
    },
    {
      "entropy": 5.313705539703369,
      "epoch": 0.6314734380220515,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004966194355556432,
      "loss": 5.1452,
      "mean_token_accuracy": 0.19047312587499618,
      "num_tokens": 13909615.0,
      "step": 7560
    },
    {
      "entropy": 5.355904197692871,
      "epoch": 0.6318910791847645,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004966142924760666,
      "loss": 5.2563,
      "mean_token_accuracy": 0.1856594130396843,
      "num_tokens": 13918491.0,
      "step": 7565
    },
    {
      "entropy": 5.301749563217163,
      "epoch": 0.6323087203474774,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004966091455168652,
      "loss": 5.2628,
      "mean_token_accuracy": 0.17914954125881194,
      "num_tokens": 13927935.0,
      "step": 7570
    },
    {
      "entropy": 5.298324489593506,
      "epoch": 0.6327263615101905,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004966039946781289,
      "loss": 5.0816,
      "mean_token_accuracy": 0.19273121207952498,
      "num_tokens": 13936524.0,
      "step": 7575
    },
    {
      "entropy": 5.23538384437561,
      "epoch": 0.6331440026729035,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000496598839959948,
      "loss": 5.1649,
      "mean_token_accuracy": 0.18347153812646866,
      "num_tokens": 13945544.0,
      "step": 7580
    },
    {
      "entropy": 5.293307542800903,
      "epoch": 0.6335616438356164,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004965936813624126,
      "loss": 5.1998,
      "mean_token_accuracy": 0.193753482401371,
      "num_tokens": 13955016.0,
      "step": 7585
    },
    {
      "entropy": 5.325612735748291,
      "epoch": 0.6339792849983295,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000496588518885613,
      "loss": 5.2004,
      "mean_token_accuracy": 0.18419731855392457,
      "num_tokens": 13964045.0,
      "step": 7590
    },
    {
      "entropy": 5.332293176651001,
      "epoch": 0.6343969261610425,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004965833525296397,
      "loss": 5.2046,
      "mean_token_accuracy": 0.18768487274646758,
      "num_tokens": 13973554.0,
      "step": 7595
    },
    {
      "entropy": 5.290938854217529,
      "epoch": 0.6348145673237554,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004965781822945831,
      "loss": 5.1893,
      "mean_token_accuracy": 0.18466798812150956,
      "num_tokens": 13982669.0,
      "step": 7600
    },
    {
      "entropy": 5.369865846633911,
      "epoch": 0.6352322084864684,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004965730081805338,
      "loss": 5.3103,
      "mean_token_accuracy": 0.17964046746492385,
      "num_tokens": 13993392.0,
      "step": 7605
    },
    {
      "entropy": 5.4114116668701175,
      "epoch": 0.6356498496491815,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004965678301875821,
      "loss": 5.3049,
      "mean_token_accuracy": 0.17811712324619294,
      "num_tokens": 14002659.0,
      "step": 7610
    },
    {
      "entropy": 5.315918636322022,
      "epoch": 0.6360674908118944,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000496562648315819,
      "loss": 5.238,
      "mean_token_accuracy": 0.18533693999052048,
      "num_tokens": 14011561.0,
      "step": 7615
    },
    {
      "entropy": 5.342669439315796,
      "epoch": 0.6364851319746074,
      "grad_norm": 1.40625,
      "learning_rate": 0.000496557462565335,
      "loss": 5.228,
      "mean_token_accuracy": 0.18481769561767578,
      "num_tokens": 14021286.0,
      "step": 7620
    },
    {
      "entropy": 5.307096338272094,
      "epoch": 0.6369027731373205,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004965522729362209,
      "loss": 5.2196,
      "mean_token_accuracy": 0.19055305868387223,
      "num_tokens": 14031901.0,
      "step": 7625
    },
    {
      "entropy": 5.328595161437988,
      "epoch": 0.6373204143000334,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004965470794285676,
      "loss": 5.2903,
      "mean_token_accuracy": 0.18768050372600556,
      "num_tokens": 14041000.0,
      "step": 7630
    },
    {
      "entropy": 5.415085124969482,
      "epoch": 0.6377380554627464,
      "grad_norm": 1.421875,
      "learning_rate": 0.000496541882042466,
      "loss": 5.3154,
      "mean_token_accuracy": 0.17772594690322877,
      "num_tokens": 14050567.0,
      "step": 7635
    },
    {
      "entropy": 5.298159551620484,
      "epoch": 0.6381556966254595,
      "grad_norm": 1.375,
      "learning_rate": 0.0004965366807780071,
      "loss": 5.1515,
      "mean_token_accuracy": 0.18856868594884874,
      "num_tokens": 14059990.0,
      "step": 7640
    },
    {
      "entropy": 5.282280111312867,
      "epoch": 0.6385733377881724,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004965314756352818,
      "loss": 5.2896,
      "mean_token_accuracy": 0.18284620642662047,
      "num_tokens": 14069317.0,
      "step": 7645
    },
    {
      "entropy": 5.34216046333313,
      "epoch": 0.6389909789508854,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004965262666143816,
      "loss": 5.2257,
      "mean_token_accuracy": 0.18454798310995102,
      "num_tokens": 14078809.0,
      "step": 7650
    },
    {
      "entropy": 5.365179538726807,
      "epoch": 0.6394086201135984,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004965210537153973,
      "loss": 5.267,
      "mean_token_accuracy": 0.17834798991680145,
      "num_tokens": 14087660.0,
      "step": 7655
    },
    {
      "entropy": 5.392992258071899,
      "epoch": 0.6398262612763114,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004965158369384204,
      "loss": 5.246,
      "mean_token_accuracy": 0.17948950380086898,
      "num_tokens": 14096585.0,
      "step": 7660
    },
    {
      "entropy": 5.28318452835083,
      "epoch": 0.6402439024390244,
      "grad_norm": 1.421875,
      "learning_rate": 0.000496510616283542,
      "loss": 5.172,
      "mean_token_accuracy": 0.19040281027555467,
      "num_tokens": 14106079.0,
      "step": 7665
    },
    {
      "entropy": 5.296985054016114,
      "epoch": 0.6406615436017374,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004965053917508537,
      "loss": 5.299,
      "mean_token_accuracy": 0.1832717940211296,
      "num_tokens": 14115988.0,
      "step": 7670
    },
    {
      "entropy": 5.363583993911743,
      "epoch": 0.6410791847644504,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004965001633404469,
      "loss": 5.3465,
      "mean_token_accuracy": 0.17442837357521057,
      "num_tokens": 14126980.0,
      "step": 7675
    },
    {
      "entropy": 5.401338386535644,
      "epoch": 0.6414968259271634,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004964949310524131,
      "loss": 5.2618,
      "mean_token_accuracy": 0.18376351296901702,
      "num_tokens": 14135780.0,
      "step": 7680
    },
    {
      "entropy": 5.393174266815185,
      "epoch": 0.6419144670898764,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004964896948868439,
      "loss": 5.2884,
      "mean_token_accuracy": 0.17311108410358428,
      "num_tokens": 14145655.0,
      "step": 7685
    },
    {
      "entropy": 5.334845304489136,
      "epoch": 0.6423321082525894,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004964844548438311,
      "loss": 5.1807,
      "mean_token_accuracy": 0.19475809633731841,
      "num_tokens": 14154157.0,
      "step": 7690
    },
    {
      "entropy": 5.3640824317932125,
      "epoch": 0.6427497494153024,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004964792109234662,
      "loss": 5.3063,
      "mean_token_accuracy": 0.18169961273670196,
      "num_tokens": 14163749.0,
      "step": 7695
    },
    {
      "entropy": 5.284685802459717,
      "epoch": 0.6431673905780154,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004964739631258411,
      "loss": 5.2376,
      "mean_token_accuracy": 0.18628758788108826,
      "num_tokens": 14174170.0,
      "step": 7700
    },
    {
      "entropy": 5.33451657295227,
      "epoch": 0.6435850317407283,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0004964687114510478,
      "loss": 5.2179,
      "mean_token_accuracy": 0.18481380492448807,
      "num_tokens": 14182625.0,
      "step": 7705
    },
    {
      "entropy": 5.381241226196289,
      "epoch": 0.6440026729034414,
      "grad_norm": 1.484375,
      "learning_rate": 0.000496463455899178,
      "loss": 5.2226,
      "mean_token_accuracy": 0.18117601424455643,
      "num_tokens": 14192351.0,
      "step": 7710
    },
    {
      "entropy": 5.215759181976319,
      "epoch": 0.6444203140661544,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004964581964703239,
      "loss": 5.1375,
      "mean_token_accuracy": 0.18823303878307343,
      "num_tokens": 14201107.0,
      "step": 7715
    },
    {
      "entropy": 5.334169054031372,
      "epoch": 0.6448379552288673,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004964529331645776,
      "loss": 5.1763,
      "mean_token_accuracy": 0.19227225780487062,
      "num_tokens": 14209540.0,
      "step": 7720
    },
    {
      "entropy": 5.29780330657959,
      "epoch": 0.6452555963915804,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000496447665982031,
      "loss": 5.0993,
      "mean_token_accuracy": 0.19178706109523774,
      "num_tokens": 14217944.0,
      "step": 7725
    },
    {
      "entropy": 5.256038570404053,
      "epoch": 0.6456732375542934,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004964423949227765,
      "loss": 5.1434,
      "mean_token_accuracy": 0.1919618457555771,
      "num_tokens": 14226044.0,
      "step": 7730
    },
    {
      "entropy": 5.3067868709564205,
      "epoch": 0.6460908787170063,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004964371199869062,
      "loss": 5.2251,
      "mean_token_accuracy": 0.17858000248670577,
      "num_tokens": 14235566.0,
      "step": 7735
    },
    {
      "entropy": 5.320413303375244,
      "epoch": 0.6465085198797194,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004964318411745127,
      "loss": 5.1664,
      "mean_token_accuracy": 0.19364046007394792,
      "num_tokens": 14244511.0,
      "step": 7740
    },
    {
      "entropy": 5.3206816673278805,
      "epoch": 0.6469261610424324,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004964265584856883,
      "loss": 5.2597,
      "mean_token_accuracy": 0.17933887243270874,
      "num_tokens": 14254299.0,
      "step": 7745
    },
    {
      "entropy": 5.242252397537231,
      "epoch": 0.6473438022051453,
      "grad_norm": 1.375,
      "learning_rate": 0.0004964212719205254,
      "loss": 5.1651,
      "mean_token_accuracy": 0.1874264359474182,
      "num_tokens": 14262915.0,
      "step": 7750
    },
    {
      "entropy": 5.305627775192261,
      "epoch": 0.6477614433678583,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004964159814791167,
      "loss": 5.2086,
      "mean_token_accuracy": 0.18672461062669754,
      "num_tokens": 14272606.0,
      "step": 7755
    },
    {
      "entropy": 5.325691652297974,
      "epoch": 0.6481790845305714,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004964106871615547,
      "loss": 5.3095,
      "mean_token_accuracy": 0.1769751340150833,
      "num_tokens": 14282117.0,
      "step": 7760
    },
    {
      "entropy": 5.266233634948731,
      "epoch": 0.6485967256932843,
      "grad_norm": 1.546875,
      "learning_rate": 0.000496405388967932,
      "loss": 5.2039,
      "mean_token_accuracy": 0.19108301848173143,
      "num_tokens": 14290421.0,
      "step": 7765
    },
    {
      "entropy": 5.417497158050537,
      "epoch": 0.6490143668559973,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004964000868983416,
      "loss": 5.3029,
      "mean_token_accuracy": 0.18283947557210922,
      "num_tokens": 14299870.0,
      "step": 7770
    },
    {
      "entropy": 5.4342268943786625,
      "epoch": 0.6494320080187104,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004963947809528761,
      "loss": 5.2441,
      "mean_token_accuracy": 0.18015625774860383,
      "num_tokens": 14309916.0,
      "step": 7775
    },
    {
      "entropy": 5.295298337936401,
      "epoch": 0.6498496491814233,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004963894711316285,
      "loss": 5.2238,
      "mean_token_accuracy": 0.18146929442882537,
      "num_tokens": 14319061.0,
      "step": 7780
    },
    {
      "entropy": 5.293241739273071,
      "epoch": 0.6502672903441363,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004963841574346917,
      "loss": 5.1766,
      "mean_token_accuracy": 0.19493246227502822,
      "num_tokens": 14327241.0,
      "step": 7785
    },
    {
      "entropy": 5.354121112823487,
      "epoch": 0.6506849315068494,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004963788398621587,
      "loss": 5.2843,
      "mean_token_accuracy": 0.18239463865756989,
      "num_tokens": 14336586.0,
      "step": 7790
    },
    {
      "entropy": 5.3281457901000975,
      "epoch": 0.6511025726695623,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004963735184141227,
      "loss": 5.2584,
      "mean_token_accuracy": 0.1824356123805046,
      "num_tokens": 14346484.0,
      "step": 7795
    },
    {
      "entropy": 5.422971963882446,
      "epoch": 0.6515202138322753,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004963681930906768,
      "loss": 5.3272,
      "mean_token_accuracy": 0.17493996769189835,
      "num_tokens": 14356255.0,
      "step": 7800
    },
    {
      "entropy": 5.365948581695557,
      "epoch": 0.6519378549949884,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004963628638919141,
      "loss": 5.2612,
      "mean_token_accuracy": 0.18347094655036927,
      "num_tokens": 14365680.0,
      "step": 7805
    },
    {
      "entropy": 5.353703498840332,
      "epoch": 0.6523554961577013,
      "grad_norm": 2.125,
      "learning_rate": 0.0004963575308179281,
      "loss": 5.2335,
      "mean_token_accuracy": 0.1850593239068985,
      "num_tokens": 14374546.0,
      "step": 7810
    },
    {
      "entropy": 5.357265901565552,
      "epoch": 0.6527731373204143,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004963521938688122,
      "loss": 5.2453,
      "mean_token_accuracy": 0.1900917574763298,
      "num_tokens": 14383624.0,
      "step": 7815
    },
    {
      "entropy": 5.378614139556885,
      "epoch": 0.6531907784831273,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004963468530446596,
      "loss": 5.3436,
      "mean_token_accuracy": 0.17858231961727142,
      "num_tokens": 14392482.0,
      "step": 7820
    },
    {
      "entropy": 5.3555823802948,
      "epoch": 0.6536084196458403,
      "grad_norm": 1.984375,
      "learning_rate": 0.0004963415083455639,
      "loss": 5.2496,
      "mean_token_accuracy": 0.1824125111103058,
      "num_tokens": 14401940.0,
      "step": 7825
    },
    {
      "entropy": 5.3768706798553465,
      "epoch": 0.6540260608085533,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004963361597716186,
      "loss": 5.2924,
      "mean_token_accuracy": 0.18103145509958268,
      "num_tokens": 14411127.0,
      "step": 7830
    },
    {
      "entropy": 5.329895257949829,
      "epoch": 0.6544437019712663,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004963308073229175,
      "loss": 5.2134,
      "mean_token_accuracy": 0.1840163514018059,
      "num_tokens": 14421358.0,
      "step": 7835
    },
    {
      "entropy": 5.319847917556762,
      "epoch": 0.6548613431339793,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004963254509995542,
      "loss": 5.2414,
      "mean_token_accuracy": 0.17936811447143555,
      "num_tokens": 14430507.0,
      "step": 7840
    },
    {
      "entropy": 5.300984907150268,
      "epoch": 0.6552789842966923,
      "grad_norm": 1.75,
      "learning_rate": 0.0004963200908016224,
      "loss": 5.2252,
      "mean_token_accuracy": 0.18744567632675171,
      "num_tokens": 14439723.0,
      "step": 7845
    },
    {
      "entropy": 5.364639043807983,
      "epoch": 0.6556966254594053,
      "grad_norm": 1.4609375,
      "learning_rate": 0.000496314726729216,
      "loss": 5.1911,
      "mean_token_accuracy": 0.18130225986242293,
      "num_tokens": 14449011.0,
      "step": 7850
    },
    {
      "entropy": 5.322962141036987,
      "epoch": 0.6561142666221182,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004963093587824289,
      "loss": 5.187,
      "mean_token_accuracy": 0.1883500710129738,
      "num_tokens": 14457676.0,
      "step": 7855
    },
    {
      "entropy": 5.317316293716431,
      "epoch": 0.6565319077848313,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004963039869613551,
      "loss": 5.2558,
      "mean_token_accuracy": 0.18281662315130234,
      "num_tokens": 14466956.0,
      "step": 7860
    },
    {
      "entropy": 5.330993938446045,
      "epoch": 0.6569495489475443,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004962986112660887,
      "loss": 5.1527,
      "mean_token_accuracy": 0.18978717774152756,
      "num_tokens": 14476083.0,
      "step": 7865
    },
    {
      "entropy": 5.380226421356201,
      "epoch": 0.6573671901102572,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004962932316967237,
      "loss": 5.2595,
      "mean_token_accuracy": 0.1838265299797058,
      "num_tokens": 14484980.0,
      "step": 7870
    },
    {
      "entropy": 5.340812683105469,
      "epoch": 0.6577848312729703,
      "grad_norm": 1.5,
      "learning_rate": 0.0004962878482533544,
      "loss": 5.2757,
      "mean_token_accuracy": 0.18076233118772506,
      "num_tokens": 14493965.0,
      "step": 7875
    },
    {
      "entropy": 5.385230588912964,
      "epoch": 0.6582024724356833,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004962824609360748,
      "loss": 5.232,
      "mean_token_accuracy": 0.1849704399704933,
      "num_tokens": 14501914.0,
      "step": 7880
    },
    {
      "entropy": 5.274803495407104,
      "epoch": 0.6586201135983962,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004962770697449794,
      "loss": 5.1846,
      "mean_token_accuracy": 0.18560997247695923,
      "num_tokens": 14511447.0,
      "step": 7885
    },
    {
      "entropy": 5.26126184463501,
      "epoch": 0.6590377547611093,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004962716746801625,
      "loss": 5.1827,
      "mean_token_accuracy": 0.18252527862787246,
      "num_tokens": 14520558.0,
      "step": 7890
    },
    {
      "entropy": 5.332695484161377,
      "epoch": 0.6594553959238223,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004962662757417187,
      "loss": 5.1825,
      "mean_token_accuracy": 0.18705461770296097,
      "num_tokens": 14529995.0,
      "step": 7895
    },
    {
      "entropy": 5.33423376083374,
      "epoch": 0.6598730370865352,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004962608729297424,
      "loss": 5.2965,
      "mean_token_accuracy": 0.1774620771408081,
      "num_tokens": 14538972.0,
      "step": 7900
    },
    {
      "entropy": 5.345441484451294,
      "epoch": 0.6602906782492483,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004962554662443281,
      "loss": 5.234,
      "mean_token_accuracy": 0.18200966119766235,
      "num_tokens": 14548255.0,
      "step": 7905
    },
    {
      "entropy": 5.284280252456665,
      "epoch": 0.6607083194119613,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004962500556855706,
      "loss": 5.1967,
      "mean_token_accuracy": 0.18816303312778473,
      "num_tokens": 14557265.0,
      "step": 7910
    },
    {
      "entropy": 5.380446290969848,
      "epoch": 0.6611259605746742,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004962446412535645,
      "loss": 5.2708,
      "mean_token_accuracy": 0.1826505497097969,
      "num_tokens": 14566650.0,
      "step": 7915
    },
    {
      "entropy": 5.3239155292510985,
      "epoch": 0.6615436017373872,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004962392229484046,
      "loss": 5.1519,
      "mean_token_accuracy": 0.19616436660289766,
      "num_tokens": 14575550.0,
      "step": 7920
    },
    {
      "entropy": 5.337485313415527,
      "epoch": 0.6619612429001003,
      "grad_norm": 1.4375,
      "learning_rate": 0.000496233800770186,
      "loss": 5.1699,
      "mean_token_accuracy": 0.18520848751068114,
      "num_tokens": 14584459.0,
      "step": 7925
    },
    {
      "entropy": 5.392370510101318,
      "epoch": 0.6623788840628132,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004962283747190033,
      "loss": 5.3014,
      "mean_token_accuracy": 0.1802950844168663,
      "num_tokens": 14593709.0,
      "step": 7930
    },
    {
      "entropy": 5.258808040618897,
      "epoch": 0.6627965252255262,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004962229447949515,
      "loss": 5.1583,
      "mean_token_accuracy": 0.19028683751821518,
      "num_tokens": 14602071.0,
      "step": 7935
    },
    {
      "entropy": 5.41706166267395,
      "epoch": 0.6632141663882393,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004962175109981259,
      "loss": 5.353,
      "mean_token_accuracy": 0.17619150280952453,
      "num_tokens": 14612842.0,
      "step": 7940
    },
    {
      "entropy": 5.463220453262329,
      "epoch": 0.6636318075509522,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004962120733286215,
      "loss": 5.3466,
      "mean_token_accuracy": 0.18361104130744935,
      "num_tokens": 14621645.0,
      "step": 7945
    },
    {
      "entropy": 5.245592975616455,
      "epoch": 0.6640494487136652,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004962066317865334,
      "loss": 5.1579,
      "mean_token_accuracy": 0.1886135458946228,
      "num_tokens": 14630447.0,
      "step": 7950
    },
    {
      "entropy": 5.263873529434204,
      "epoch": 0.6644670898763783,
      "grad_norm": 1.40625,
      "learning_rate": 0.000496201186371957,
      "loss": 5.2422,
      "mean_token_accuracy": 0.191739284992218,
      "num_tokens": 14639371.0,
      "step": 7955
    },
    {
      "entropy": 5.375819110870362,
      "epoch": 0.6648847310390912,
      "grad_norm": 1.5,
      "learning_rate": 0.0004961957370849875,
      "loss": 5.1581,
      "mean_token_accuracy": 0.18274612724781036,
      "num_tokens": 14648540.0,
      "step": 7960
    },
    {
      "entropy": 5.384054374694824,
      "epoch": 0.6653023722018042,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004961902839257205,
      "loss": 5.3403,
      "mean_token_accuracy": 0.183157841861248,
      "num_tokens": 14658130.0,
      "step": 7965
    },
    {
      "entropy": 5.386115550994873,
      "epoch": 0.6657200133645172,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004961848268942513,
      "loss": 5.2793,
      "mean_token_accuracy": 0.1771568238735199,
      "num_tokens": 14667538.0,
      "step": 7970
    },
    {
      "entropy": 5.374794006347656,
      "epoch": 0.6661376545272302,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004961793659906754,
      "loss": 5.2495,
      "mean_token_accuracy": 0.177203768491745,
      "num_tokens": 14677115.0,
      "step": 7975
    },
    {
      "entropy": 5.437082386016845,
      "epoch": 0.6665552956899432,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004961739012150885,
      "loss": 5.2215,
      "mean_token_accuracy": 0.18960783183574675,
      "num_tokens": 14686444.0,
      "step": 7980
    },
    {
      "entropy": 5.325315284729004,
      "epoch": 0.6669729368526562,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004961684325675862,
      "loss": 5.1871,
      "mean_token_accuracy": 0.19141027480363845,
      "num_tokens": 14695951.0,
      "step": 7985
    },
    {
      "entropy": 5.31355094909668,
      "epoch": 0.6673905780153692,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004961629600482643,
      "loss": 5.3027,
      "mean_token_accuracy": 0.18038068562746049,
      "num_tokens": 14705415.0,
      "step": 7990
    },
    {
      "entropy": 5.417062187194825,
      "epoch": 0.6678082191780822,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004961574836572186,
      "loss": 5.2723,
      "mean_token_accuracy": 0.18219208866357803,
      "num_tokens": 14714793.0,
      "step": 7995
    },
    {
      "entropy": 5.276196432113648,
      "epoch": 0.6682258603407952,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004961520033945448,
      "loss": 5.0514,
      "mean_token_accuracy": 0.19502789080142974,
      "num_tokens": 14723820.0,
      "step": 8000
    },
    {
      "entropy": 5.288319492340088,
      "epoch": 0.6686435015035082,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004961465192603391,
      "loss": 5.1675,
      "mean_token_accuracy": 0.19178766310214995,
      "num_tokens": 14733055.0,
      "step": 8005
    },
    {
      "entropy": 5.338773727416992,
      "epoch": 0.6690611426662212,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004961410312546974,
      "loss": 5.2023,
      "mean_token_accuracy": 0.1862553045153618,
      "num_tokens": 14742416.0,
      "step": 8010
    },
    {
      "entropy": 5.295219945907593,
      "epoch": 0.6694787838289342,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004961355393777158,
      "loss": 5.1967,
      "mean_token_accuracy": 0.18526092320680618,
      "num_tokens": 14751384.0,
      "step": 8015
    },
    {
      "entropy": 5.250601530075073,
      "epoch": 0.6698964249916471,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004961300436294902,
      "loss": 5.0915,
      "mean_token_accuracy": 0.18689117580652237,
      "num_tokens": 14760525.0,
      "step": 8020
    },
    {
      "entropy": 5.3468163967132565,
      "epoch": 0.6703140661543602,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004961245440101171,
      "loss": 5.2616,
      "mean_token_accuracy": 0.1819304719567299,
      "num_tokens": 14769405.0,
      "step": 8025
    },
    {
      "entropy": 5.238767528533936,
      "epoch": 0.6707317073170732,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004961190405196928,
      "loss": 5.1253,
      "mean_token_accuracy": 0.18605050891637803,
      "num_tokens": 14778945.0,
      "step": 8030
    },
    {
      "entropy": 5.373712873458862,
      "epoch": 0.6711493484797861,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004961135331583135,
      "loss": 5.2964,
      "mean_token_accuracy": 0.1771259590983391,
      "num_tokens": 14788289.0,
      "step": 8035
    },
    {
      "entropy": 5.292560672760009,
      "epoch": 0.6715669896424992,
      "grad_norm": 1.5,
      "learning_rate": 0.0004961080219260755,
      "loss": 5.2144,
      "mean_token_accuracy": 0.18261197507381438,
      "num_tokens": 14797761.0,
      "step": 8040
    },
    {
      "entropy": 5.344115161895752,
      "epoch": 0.6719846308052122,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004961025068230757,
      "loss": 5.2678,
      "mean_token_accuracy": 0.1875170111656189,
      "num_tokens": 14806424.0,
      "step": 8045
    },
    {
      "entropy": 5.29541711807251,
      "epoch": 0.6724022719679251,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004960969878494103,
      "loss": 5.1995,
      "mean_token_accuracy": 0.1792940229177475,
      "num_tokens": 14815816.0,
      "step": 8050
    },
    {
      "entropy": 5.263209342956543,
      "epoch": 0.6728199131306382,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004960914650051758,
      "loss": 5.1901,
      "mean_token_accuracy": 0.1867182433605194,
      "num_tokens": 14824877.0,
      "step": 8055
    },
    {
      "entropy": 5.24476809501648,
      "epoch": 0.6732375542933512,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004960859382904691,
      "loss": 5.103,
      "mean_token_accuracy": 0.19761217534542083,
      "num_tokens": 14833095.0,
      "step": 8060
    },
    {
      "entropy": 5.392882919311523,
      "epoch": 0.6736551954560641,
      "grad_norm": 1.59375,
      "learning_rate": 0.000496080407705387,
      "loss": 5.2434,
      "mean_token_accuracy": 0.180871120095253,
      "num_tokens": 14842963.0,
      "step": 8065
    },
    {
      "entropy": 5.411832475662232,
      "epoch": 0.6740728366187771,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004960748732500263,
      "loss": 5.3141,
      "mean_token_accuracy": 0.18388109803199768,
      "num_tokens": 14851552.0,
      "step": 8070
    },
    {
      "entropy": 5.34446907043457,
      "epoch": 0.6744904777814902,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004960693349244838,
      "loss": 5.2227,
      "mean_token_accuracy": 0.18630302846431732,
      "num_tokens": 14859997.0,
      "step": 8075
    },
    {
      "entropy": 5.321320629119873,
      "epoch": 0.6749081189442031,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004960637927288566,
      "loss": 5.1375,
      "mean_token_accuracy": 0.18813525140285492,
      "num_tokens": 14868050.0,
      "step": 8080
    },
    {
      "entropy": 5.301049709320068,
      "epoch": 0.6753257601069161,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004960582466632414,
      "loss": 5.239,
      "mean_token_accuracy": 0.1822244793176651,
      "num_tokens": 14877774.0,
      "step": 8085
    },
    {
      "entropy": 5.240277290344238,
      "epoch": 0.6757434012696292,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004960526967277356,
      "loss": 5.1728,
      "mean_token_accuracy": 0.1926938235759735,
      "num_tokens": 14886971.0,
      "step": 8090
    },
    {
      "entropy": 5.238359546661377,
      "epoch": 0.6761610424323421,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004960471429224363,
      "loss": 5.139,
      "mean_token_accuracy": 0.1871116504073143,
      "num_tokens": 14895815.0,
      "step": 8095
    },
    {
      "entropy": 5.296923542022705,
      "epoch": 0.6765786835950551,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004960415852474406,
      "loss": 5.1515,
      "mean_token_accuracy": 0.19421091228723525,
      "num_tokens": 14904082.0,
      "step": 8100
    },
    {
      "entropy": 5.280825757980347,
      "epoch": 0.6769963247577682,
      "grad_norm": 1.5625,
      "learning_rate": 0.000496036023702846,
      "loss": 5.1832,
      "mean_token_accuracy": 0.1913072794675827,
      "num_tokens": 14913533.0,
      "step": 8105
    },
    {
      "entropy": 5.277411365509034,
      "epoch": 0.6774139659204811,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004960304582887497,
      "loss": 5.1652,
      "mean_token_accuracy": 0.18801499307155609,
      "num_tokens": 14922063.0,
      "step": 8110
    },
    {
      "entropy": 5.314997434616089,
      "epoch": 0.6778316070831941,
      "grad_norm": 1.375,
      "learning_rate": 0.0004960248890052493,
      "loss": 5.1968,
      "mean_token_accuracy": 0.1813928097486496,
      "num_tokens": 14930219.0,
      "step": 8115
    },
    {
      "entropy": 5.381176519393921,
      "epoch": 0.6782492482459072,
      "grad_norm": 1.578125,
      "learning_rate": 0.000496019315852442,
      "loss": 5.2923,
      "mean_token_accuracy": 0.18256254643201827,
      "num_tokens": 14939463.0,
      "step": 8120
    },
    {
      "entropy": 5.402235221862793,
      "epoch": 0.6786668894086201,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004960137388304255,
      "loss": 5.1932,
      "mean_token_accuracy": 0.1897577852010727,
      "num_tokens": 14948289.0,
      "step": 8125
    },
    {
      "entropy": 5.347008609771729,
      "epoch": 0.6790845305713331,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004960081579392975,
      "loss": 5.2933,
      "mean_token_accuracy": 0.18622678965330125,
      "num_tokens": 14958172.0,
      "step": 8130
    },
    {
      "entropy": 5.241091251373291,
      "epoch": 0.6795021717340461,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004960025731791558,
      "loss": 5.0896,
      "mean_token_accuracy": 0.19144905507564544,
      "num_tokens": 14967362.0,
      "step": 8135
    },
    {
      "entropy": 5.273113155364991,
      "epoch": 0.6799198128967591,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004959969845500979,
      "loss": 5.3005,
      "mean_token_accuracy": 0.1805402860045433,
      "num_tokens": 14975942.0,
      "step": 8140
    },
    {
      "entropy": 5.323067092895508,
      "epoch": 0.6803374540594721,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004959913920522218,
      "loss": 5.2146,
      "mean_token_accuracy": 0.18757303953170776,
      "num_tokens": 14985428.0,
      "step": 8145
    },
    {
      "entropy": 5.32785005569458,
      "epoch": 0.6807550952221851,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004959857956856254,
      "loss": 5.1571,
      "mean_token_accuracy": 0.1838672325015068,
      "num_tokens": 14994558.0,
      "step": 8150
    },
    {
      "entropy": 5.326738452911377,
      "epoch": 0.6811727363848981,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004959801954504067,
      "loss": 5.1632,
      "mean_token_accuracy": 0.1867782175540924,
      "num_tokens": 15003674.0,
      "step": 8155
    },
    {
      "entropy": 5.315521955490112,
      "epoch": 0.6815903775476111,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004959745913466635,
      "loss": 5.2272,
      "mean_token_accuracy": 0.18782551288604737,
      "num_tokens": 15012569.0,
      "step": 8160
    },
    {
      "entropy": 5.327926349639893,
      "epoch": 0.6820080187103241,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004959689833744942,
      "loss": 5.2356,
      "mean_token_accuracy": 0.18866464048624038,
      "num_tokens": 15021863.0,
      "step": 8165
    },
    {
      "entropy": 5.3411026954650875,
      "epoch": 0.682425659873037,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004959633715339969,
      "loss": 5.2798,
      "mean_token_accuracy": 0.17385224550962447,
      "num_tokens": 15032272.0,
      "step": 8170
    },
    {
      "entropy": 5.29858512878418,
      "epoch": 0.6828433010357501,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004959577558252698,
      "loss": 5.2051,
      "mean_token_accuracy": 0.18027757704257966,
      "num_tokens": 15041066.0,
      "step": 8175
    },
    {
      "entropy": 5.4229888916015625,
      "epoch": 0.6832609421984631,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004959521362484112,
      "loss": 5.294,
      "mean_token_accuracy": 0.18273967504501343,
      "num_tokens": 15050855.0,
      "step": 8180
    },
    {
      "entropy": 5.31981406211853,
      "epoch": 0.683678583361176,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004959465128035195,
      "loss": 5.1874,
      "mean_token_accuracy": 0.192647847533226,
      "num_tokens": 15060810.0,
      "step": 8185
    },
    {
      "entropy": 5.296506929397583,
      "epoch": 0.6840962245238891,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004959408854906932,
      "loss": 5.1767,
      "mean_token_accuracy": 0.1906803071498871,
      "num_tokens": 15069761.0,
      "step": 8190
    },
    {
      "entropy": 5.253224277496338,
      "epoch": 0.6845138656866021,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004959352543100306,
      "loss": 5.1457,
      "mean_token_accuracy": 0.18826841562986374,
      "num_tokens": 15078691.0,
      "step": 8195
    },
    {
      "entropy": 5.253979873657227,
      "epoch": 0.684931506849315,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004959296192616305,
      "loss": 5.1472,
      "mean_token_accuracy": 0.19228082299232482,
      "num_tokens": 15087068.0,
      "step": 8200
    },
    {
      "entropy": 5.297430467605591,
      "epoch": 0.6853491480120281,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004959239803455916,
      "loss": 5.0477,
      "mean_token_accuracy": 0.1919408455491066,
      "num_tokens": 15095962.0,
      "step": 8205
    },
    {
      "entropy": 5.271142101287841,
      "epoch": 0.6857667891747411,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004959183375620123,
      "loss": 5.2207,
      "mean_token_accuracy": 0.18081678003072738,
      "num_tokens": 15105835.0,
      "step": 8210
    },
    {
      "entropy": 5.2731605052948,
      "epoch": 0.686184430337454,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004959126909109918,
      "loss": 5.138,
      "mean_token_accuracy": 0.19526263922452927,
      "num_tokens": 15114140.0,
      "step": 8215
    },
    {
      "entropy": 5.344247579574585,
      "epoch": 0.686602071500167,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004959070403926286,
      "loss": 5.2346,
      "mean_token_accuracy": 0.18727920651435853,
      "num_tokens": 15123253.0,
      "step": 8220
    },
    {
      "entropy": 5.445370769500732,
      "epoch": 0.6870197126628801,
      "grad_norm": 2.4375,
      "learning_rate": 0.0004959013860070217,
      "loss": 5.3406,
      "mean_token_accuracy": 0.17764439433813095,
      "num_tokens": 15133699.0,
      "step": 8225
    },
    {
      "entropy": 5.3401844024658205,
      "epoch": 0.687437353825593,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004958957277542701,
      "loss": 5.0982,
      "mean_token_accuracy": 0.19175320714712143,
      "num_tokens": 15142133.0,
      "step": 8230
    },
    {
      "entropy": 5.279022932052612,
      "epoch": 0.687854994988306,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004958900656344731,
      "loss": 5.1824,
      "mean_token_accuracy": 0.18504478335380553,
      "num_tokens": 15151367.0,
      "step": 8235
    },
    {
      "entropy": 5.283271741867066,
      "epoch": 0.6882726361510191,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004958843996477296,
      "loss": 5.2127,
      "mean_token_accuracy": 0.18199769407510757,
      "num_tokens": 15160938.0,
      "step": 8240
    },
    {
      "entropy": 5.4045368194580075,
      "epoch": 0.688690277313732,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004958787297941386,
      "loss": 5.2598,
      "mean_token_accuracy": 0.1807729035615921,
      "num_tokens": 15170507.0,
      "step": 8245
    },
    {
      "entropy": 5.393487977981567,
      "epoch": 0.689107918476445,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004958730560737998,
      "loss": 5.2622,
      "mean_token_accuracy": 0.18554342314600944,
      "num_tokens": 15179822.0,
      "step": 8250
    },
    {
      "entropy": 5.3177588939666744,
      "epoch": 0.6895255596391581,
      "grad_norm": 1.5,
      "learning_rate": 0.0004958673784868121,
      "loss": 5.1533,
      "mean_token_accuracy": 0.18469705432653427,
      "num_tokens": 15189353.0,
      "step": 8255
    },
    {
      "entropy": 5.343029642105103,
      "epoch": 0.689943200801871,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004958616970332752,
      "loss": 5.2452,
      "mean_token_accuracy": 0.17982738912105561,
      "num_tokens": 15197772.0,
      "step": 8260
    },
    {
      "entropy": 5.288300085067749,
      "epoch": 0.690360841964584,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004958560117132884,
      "loss": 5.1982,
      "mean_token_accuracy": 0.18857772499322892,
      "num_tokens": 15207453.0,
      "step": 8265
    },
    {
      "entropy": 5.289036798477173,
      "epoch": 0.6907784831272971,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004958503225269513,
      "loss": 5.1189,
      "mean_token_accuracy": 0.19024553447961806,
      "num_tokens": 15215787.0,
      "step": 8270
    },
    {
      "entropy": 5.41863865852356,
      "epoch": 0.69119612429001,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004958446294743633,
      "loss": 5.2361,
      "mean_token_accuracy": 0.18338089287281037,
      "num_tokens": 15225148.0,
      "step": 8275
    },
    {
      "entropy": 5.372235584259033,
      "epoch": 0.691613765452723,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004958389325556245,
      "loss": 5.2674,
      "mean_token_accuracy": 0.18188745975494386,
      "num_tokens": 15234533.0,
      "step": 8280
    },
    {
      "entropy": 5.300735425949097,
      "epoch": 0.692031406615436,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004958332317708341,
      "loss": 5.2568,
      "mean_token_accuracy": 0.17238677889108658,
      "num_tokens": 15243805.0,
      "step": 8285
    },
    {
      "entropy": 5.406217336654663,
      "epoch": 0.692449047778149,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004958275271200924,
      "loss": 5.2783,
      "mean_token_accuracy": 0.18325183987617494,
      "num_tokens": 15253298.0,
      "step": 8290
    },
    {
      "entropy": 5.314294815063477,
      "epoch": 0.692866688940862,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004958218186034988,
      "loss": 5.2217,
      "mean_token_accuracy": 0.1871773287653923,
      "num_tokens": 15261860.0,
      "step": 8295
    },
    {
      "entropy": 5.314128017425537,
      "epoch": 0.693284330103575,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004958161062211537,
      "loss": 5.152,
      "mean_token_accuracy": 0.18709763884544373,
      "num_tokens": 15271021.0,
      "step": 8300
    },
    {
      "entropy": 5.282918977737427,
      "epoch": 0.693701971266288,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004958103899731566,
      "loss": 5.1934,
      "mean_token_accuracy": 0.18143055588006973,
      "num_tokens": 15280489.0,
      "step": 8305
    },
    {
      "entropy": 5.330470991134644,
      "epoch": 0.694119612429001,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004958046698596081,
      "loss": 5.2072,
      "mean_token_accuracy": 0.19208525270223617,
      "num_tokens": 15289387.0,
      "step": 8310
    },
    {
      "entropy": 5.350530004501342,
      "epoch": 0.694537253591714,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004957989458806079,
      "loss": 5.2873,
      "mean_token_accuracy": 0.1837986633181572,
      "num_tokens": 15299334.0,
      "step": 8315
    },
    {
      "entropy": 5.330388498306275,
      "epoch": 0.694954894754427,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004957932180362564,
      "loss": 5.2336,
      "mean_token_accuracy": 0.18512164205312728,
      "num_tokens": 15309174.0,
      "step": 8320
    },
    {
      "entropy": 5.378217935562134,
      "epoch": 0.69537253591714,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004957874863266538,
      "loss": 5.3008,
      "mean_token_accuracy": 0.18431434631347657,
      "num_tokens": 15318722.0,
      "step": 8325
    },
    {
      "entropy": 5.436281538009643,
      "epoch": 0.695790177079853,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004957817507519006,
      "loss": 5.2331,
      "mean_token_accuracy": 0.18671645671129228,
      "num_tokens": 15327606.0,
      "step": 8330
    },
    {
      "entropy": 5.309534358978271,
      "epoch": 0.696207818242566,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004957760113120971,
      "loss": 5.2058,
      "mean_token_accuracy": 0.19387727379798889,
      "num_tokens": 15336520.0,
      "step": 8335
    },
    {
      "entropy": 5.378321361541748,
      "epoch": 0.696625459405279,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004957702680073437,
      "loss": 5.2587,
      "mean_token_accuracy": 0.18018345087766646,
      "num_tokens": 15345437.0,
      "step": 8340
    },
    {
      "entropy": 5.409728097915649,
      "epoch": 0.697043100567992,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004957645208377411,
      "loss": 5.2033,
      "mean_token_accuracy": 0.186541286110878,
      "num_tokens": 15354660.0,
      "step": 8345
    },
    {
      "entropy": 5.334236431121826,
      "epoch": 0.6974607417307049,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004957587698033898,
      "loss": 5.253,
      "mean_token_accuracy": 0.1862293854355812,
      "num_tokens": 15363958.0,
      "step": 8350
    },
    {
      "entropy": 5.34509129524231,
      "epoch": 0.697878382893418,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004957530149043905,
      "loss": 5.1946,
      "mean_token_accuracy": 0.19260832369327546,
      "num_tokens": 15373351.0,
      "step": 8355
    },
    {
      "entropy": 5.418985795974732,
      "epoch": 0.698296024056131,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004957472561408438,
      "loss": 5.2004,
      "mean_token_accuracy": 0.1887256145477295,
      "num_tokens": 15383069.0,
      "step": 8360
    },
    {
      "entropy": 5.3014484405517575,
      "epoch": 0.6987136652188439,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004957414935128509,
      "loss": 5.1864,
      "mean_token_accuracy": 0.18785572946071624,
      "num_tokens": 15391394.0,
      "step": 8365
    },
    {
      "entropy": 5.318035697937011,
      "epoch": 0.699131306381557,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004957357270205123,
      "loss": 5.1866,
      "mean_token_accuracy": 0.1832756981253624,
      "num_tokens": 15399993.0,
      "step": 8370
    },
    {
      "entropy": 5.291863489151001,
      "epoch": 0.69954894754427,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004957299566639292,
      "loss": 5.1826,
      "mean_token_accuracy": 0.18668273985385894,
      "num_tokens": 15409886.0,
      "step": 8375
    },
    {
      "entropy": 5.346201705932617,
      "epoch": 0.6999665887069829,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004957241824432025,
      "loss": 5.1871,
      "mean_token_accuracy": 0.18907375484704972,
      "num_tokens": 15419772.0,
      "step": 8380
    },
    {
      "entropy": 5.353672170639038,
      "epoch": 0.700384229869696,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004957184043584333,
      "loss": 5.1934,
      "mean_token_accuracy": 0.1868671089410782,
      "num_tokens": 15429631.0,
      "step": 8385
    },
    {
      "entropy": 5.304797792434693,
      "epoch": 0.700801871032409,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004957126224097227,
      "loss": 5.1247,
      "mean_token_accuracy": 0.1953100398182869,
      "num_tokens": 15438411.0,
      "step": 8390
    },
    {
      "entropy": 5.218867683410645,
      "epoch": 0.7012195121951219,
      "grad_norm": 1.8359375,
      "learning_rate": 0.000495706836597172,
      "loss": 5.1851,
      "mean_token_accuracy": 0.1908182442188263,
      "num_tokens": 15448943.0,
      "step": 8395
    },
    {
      "entropy": 5.322815656661987,
      "epoch": 0.7016371533578349,
      "grad_norm": 1.984375,
      "learning_rate": 0.0004957010469208825,
      "loss": 5.134,
      "mean_token_accuracy": 0.1907452329993248,
      "num_tokens": 15457908.0,
      "step": 8400
    },
    {
      "entropy": 5.398818445205689,
      "epoch": 0.702054794520548,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004956952533809554,
      "loss": 5.2215,
      "mean_token_accuracy": 0.18379063457250594,
      "num_tokens": 15467390.0,
      "step": 8405
    },
    {
      "entropy": 5.355314254760742,
      "epoch": 0.7024724356832609,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004956894559774923,
      "loss": 5.2242,
      "mean_token_accuracy": 0.185761496424675,
      "num_tokens": 15476419.0,
      "step": 8410
    },
    {
      "entropy": 5.22340760231018,
      "epoch": 0.7028900768459739,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004956836547105947,
      "loss": 5.0728,
      "mean_token_accuracy": 0.1945248782634735,
      "num_tokens": 15485147.0,
      "step": 8415
    },
    {
      "entropy": 5.319608354568482,
      "epoch": 0.703307718008687,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004956778495803641,
      "loss": 5.1862,
      "mean_token_accuracy": 0.18875861316919326,
      "num_tokens": 15493816.0,
      "step": 8420
    },
    {
      "entropy": 5.342964315414429,
      "epoch": 0.7037253591713999,
      "grad_norm": 1.7734375,
      "learning_rate": 0.000495672040586902,
      "loss": 5.2987,
      "mean_token_accuracy": 0.1840755820274353,
      "num_tokens": 15503051.0,
      "step": 8425
    },
    {
      "entropy": 5.316436672210694,
      "epoch": 0.7041430003341129,
      "grad_norm": 2.0,
      "learning_rate": 0.0004956662277303103,
      "loss": 5.1175,
      "mean_token_accuracy": 0.19207129925489425,
      "num_tokens": 15511854.0,
      "step": 8430
    },
    {
      "entropy": 5.30217661857605,
      "epoch": 0.704560641496826,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004956604110106905,
      "loss": 5.2394,
      "mean_token_accuracy": 0.18537282645702363,
      "num_tokens": 15521492.0,
      "step": 8435
    },
    {
      "entropy": 5.464894580841064,
      "epoch": 0.7049782826595389,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004956545904281447,
      "loss": 5.3147,
      "mean_token_accuracy": 0.1746738463640213,
      "num_tokens": 15530983.0,
      "step": 8440
    },
    {
      "entropy": 5.297878646850586,
      "epoch": 0.7053959238222519,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004956487659827748,
      "loss": 5.1515,
      "mean_token_accuracy": 0.19397141933441162,
      "num_tokens": 15539722.0,
      "step": 8445
    },
    {
      "entropy": 5.2263692855834964,
      "epoch": 0.705813564984965,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004956429376746825,
      "loss": 5.1295,
      "mean_token_accuracy": 0.1891203999519348,
      "num_tokens": 15548738.0,
      "step": 8450
    },
    {
      "entropy": 5.235093307495117,
      "epoch": 0.7062312061476779,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004956371055039701,
      "loss": 5.1649,
      "mean_token_accuracy": 0.19618828296661378,
      "num_tokens": 15557520.0,
      "step": 8455
    },
    {
      "entropy": 5.387892389297486,
      "epoch": 0.7066488473103909,
      "grad_norm": 2.234375,
      "learning_rate": 0.0004956312694707395,
      "loss": 5.2469,
      "mean_token_accuracy": 0.17590499818325042,
      "num_tokens": 15567860.0,
      "step": 8460
    },
    {
      "entropy": 5.376634550094605,
      "epoch": 0.7070664884731039,
      "grad_norm": 1.6953125,
      "learning_rate": 0.000495625429575093,
      "loss": 5.2539,
      "mean_token_accuracy": 0.17584866136312485,
      "num_tokens": 15577174.0,
      "step": 8465
    },
    {
      "entropy": 5.230661010742187,
      "epoch": 0.7074841296358169,
      "grad_norm": 1.375,
      "learning_rate": 0.0004956195858171327,
      "loss": 5.044,
      "mean_token_accuracy": 0.19701918065547944,
      "num_tokens": 15585476.0,
      "step": 8470
    },
    {
      "entropy": 5.195006275177002,
      "epoch": 0.7079017707985299,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004956137381969611,
      "loss": 5.1821,
      "mean_token_accuracy": 0.18857980370521546,
      "num_tokens": 15595422.0,
      "step": 8475
    },
    {
      "entropy": 5.3309242725372314,
      "epoch": 0.7083194119612429,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004956078867146805,
      "loss": 5.2941,
      "mean_token_accuracy": 0.18028816431760789,
      "num_tokens": 15604223.0,
      "step": 8480
    },
    {
      "entropy": 5.39313907623291,
      "epoch": 0.7087370531239559,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004956020313703933,
      "loss": 5.2744,
      "mean_token_accuracy": 0.18230718821287156,
      "num_tokens": 15613715.0,
      "step": 8485
    },
    {
      "entropy": 5.282086181640625,
      "epoch": 0.7091546942866689,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004955961721642018,
      "loss": 5.0979,
      "mean_token_accuracy": 0.19790386408567429,
      "num_tokens": 15623826.0,
      "step": 8490
    },
    {
      "entropy": 5.3074768543243405,
      "epoch": 0.7095723354493819,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000495590309096209,
      "loss": 5.2168,
      "mean_token_accuracy": 0.1797385260462761,
      "num_tokens": 15633311.0,
      "step": 8495
    },
    {
      "entropy": 5.309714984893799,
      "epoch": 0.7099899766120948,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004955844421665171,
      "loss": 5.1824,
      "mean_token_accuracy": 0.18832795321941376,
      "num_tokens": 15642785.0,
      "step": 8500
    },
    {
      "entropy": 5.471587228775024,
      "epoch": 0.7104076177748079,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004955785713752292,
      "loss": 5.3778,
      "mean_token_accuracy": 0.18039360791444778,
      "num_tokens": 15653009.0,
      "step": 8505
    },
    {
      "entropy": 5.359288024902344,
      "epoch": 0.7108252589375209,
      "grad_norm": 1.5,
      "learning_rate": 0.0004955726967224479,
      "loss": 5.1956,
      "mean_token_accuracy": 0.1898520588874817,
      "num_tokens": 15661672.0,
      "step": 8510
    },
    {
      "entropy": 5.261266231536865,
      "epoch": 0.7112429001002338,
      "grad_norm": 1.703125,
      "learning_rate": 0.000495566818208276,
      "loss": 5.173,
      "mean_token_accuracy": 0.17999346405267716,
      "num_tokens": 15672186.0,
      "step": 8515
    },
    {
      "entropy": 5.263257932662964,
      "epoch": 0.7116605412629469,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004955609358328163,
      "loss": 5.2018,
      "mean_token_accuracy": 0.19061159789562226,
      "num_tokens": 15680351.0,
      "step": 8520
    },
    {
      "entropy": 5.33342432975769,
      "epoch": 0.7120781824256599,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004955550495961722,
      "loss": 5.1845,
      "mean_token_accuracy": 0.18149695098400115,
      "num_tokens": 15690140.0,
      "step": 8525
    },
    {
      "entropy": 5.353526830673218,
      "epoch": 0.7124958235883728,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004955491594984462,
      "loss": 5.2746,
      "mean_token_accuracy": 0.17878417074680328,
      "num_tokens": 15699968.0,
      "step": 8530
    },
    {
      "entropy": 5.304271030426025,
      "epoch": 0.7129134647510859,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004955432655397419,
      "loss": 5.1269,
      "mean_token_accuracy": 0.19037148207426072,
      "num_tokens": 15709061.0,
      "step": 8535
    },
    {
      "entropy": 5.357570028305053,
      "epoch": 0.7133311059137989,
      "grad_norm": 2.0,
      "learning_rate": 0.0004955373677201621,
      "loss": 5.2665,
      "mean_token_accuracy": 0.17969775050878525,
      "num_tokens": 15718350.0,
      "step": 8540
    },
    {
      "entropy": 5.321164226531982,
      "epoch": 0.7137487470765118,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004955314660398105,
      "loss": 5.1793,
      "mean_token_accuracy": 0.19132979661226274,
      "num_tokens": 15727120.0,
      "step": 8545
    },
    {
      "entropy": 5.318919515609741,
      "epoch": 0.7141663882392248,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004955255604987899,
      "loss": 5.2633,
      "mean_token_accuracy": 0.18388740569353104,
      "num_tokens": 15735865.0,
      "step": 8550
    },
    {
      "entropy": 5.334926176071167,
      "epoch": 0.7145840294019379,
      "grad_norm": 1.421875,
      "learning_rate": 0.000495519651097204,
      "loss": 5.2147,
      "mean_token_accuracy": 0.18114018738269805,
      "num_tokens": 15745239.0,
      "step": 8555
    },
    {
      "entropy": 5.339488744735718,
      "epoch": 0.7150016705646508,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004955137378351561,
      "loss": 5.1265,
      "mean_token_accuracy": 0.1935678794980049,
      "num_tokens": 15754418.0,
      "step": 8560
    },
    {
      "entropy": 5.23320837020874,
      "epoch": 0.7154193117273638,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004955078207127498,
      "loss": 5.1937,
      "mean_token_accuracy": 0.1880308657884598,
      "num_tokens": 15764175.0,
      "step": 8565
    },
    {
      "entropy": 5.28746280670166,
      "epoch": 0.7158369528900769,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004955018997300886,
      "loss": 5.1879,
      "mean_token_accuracy": 0.1876701757311821,
      "num_tokens": 15772766.0,
      "step": 8570
    },
    {
      "entropy": 5.26781816482544,
      "epoch": 0.7162545940527898,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004954959748872764,
      "loss": 5.2039,
      "mean_token_accuracy": 0.188557930290699,
      "num_tokens": 15781768.0,
      "step": 8575
    },
    {
      "entropy": 5.360330533981323,
      "epoch": 0.7166722352155028,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004954900461844167,
      "loss": 5.2512,
      "mean_token_accuracy": 0.18298709094524385,
      "num_tokens": 15791706.0,
      "step": 8580
    },
    {
      "entropy": 5.3711076259613035,
      "epoch": 0.7170898763782159,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004954841136216132,
      "loss": 5.2638,
      "mean_token_accuracy": 0.182969132065773,
      "num_tokens": 15802015.0,
      "step": 8585
    },
    {
      "entropy": 5.390833330154419,
      "epoch": 0.7175075175409288,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004954781771989701,
      "loss": 5.2102,
      "mean_token_accuracy": 0.18598561882972717,
      "num_tokens": 15811116.0,
      "step": 8590
    },
    {
      "entropy": 5.370823907852173,
      "epoch": 0.7179251587036418,
      "grad_norm": 1.5234375,
      "learning_rate": 0.000495472236916591,
      "loss": 5.1818,
      "mean_token_accuracy": 0.18811384439468384,
      "num_tokens": 15819240.0,
      "step": 8595
    },
    {
      "entropy": 5.246731090545654,
      "epoch": 0.7183427998663549,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004954662927745799,
      "loss": 5.1822,
      "mean_token_accuracy": 0.1792804092168808,
      "num_tokens": 15827740.0,
      "step": 8600
    },
    {
      "entropy": 5.252020025253296,
      "epoch": 0.7187604410290678,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004954603447730411,
      "loss": 5.2498,
      "mean_token_accuracy": 0.1856865718960762,
      "num_tokens": 15836763.0,
      "step": 8605
    },
    {
      "entropy": 5.484684085845947,
      "epoch": 0.7191780821917808,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004954543929120786,
      "loss": 5.3126,
      "mean_token_accuracy": 0.18425522446632386,
      "num_tokens": 15846624.0,
      "step": 8610
    },
    {
      "entropy": 5.3179261684417725,
      "epoch": 0.7195957233544938,
      "grad_norm": 1.5,
      "learning_rate": 0.0004954484371917967,
      "loss": 5.2187,
      "mean_token_accuracy": 0.18258151710033416,
      "num_tokens": 15855713.0,
      "step": 8615
    },
    {
      "entropy": 5.25519208908081,
      "epoch": 0.7200133645172068,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004954424776122993,
      "loss": 5.0484,
      "mean_token_accuracy": 0.1937964841723442,
      "num_tokens": 15863577.0,
      "step": 8620
    },
    {
      "entropy": 5.310612058639526,
      "epoch": 0.7204310056799198,
      "grad_norm": 1.5,
      "learning_rate": 0.0004954365141736913,
      "loss": 5.2178,
      "mean_token_accuracy": 0.18976670354604722,
      "num_tokens": 15872479.0,
      "step": 8625
    },
    {
      "entropy": 5.366736268997192,
      "epoch": 0.7208486468426328,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004954305468760765,
      "loss": 5.3233,
      "mean_token_accuracy": 0.18595676273107528,
      "num_tokens": 15882170.0,
      "step": 8630
    },
    {
      "entropy": 5.297217178344726,
      "epoch": 0.7212662880053458,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004954245757195598,
      "loss": 5.1957,
      "mean_token_accuracy": 0.19112035632133484,
      "num_tokens": 15891694.0,
      "step": 8635
    },
    {
      "entropy": 5.342066907882691,
      "epoch": 0.7216839291680588,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004954186007042454,
      "loss": 5.3065,
      "mean_token_accuracy": 0.1751521810889244,
      "num_tokens": 15902232.0,
      "step": 8640
    },
    {
      "entropy": 5.355078458786011,
      "epoch": 0.7221015703307718,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004954126218302382,
      "loss": 5.1585,
      "mean_token_accuracy": 0.1850830689072609,
      "num_tokens": 15912218.0,
      "step": 8645
    },
    {
      "entropy": 5.295679426193237,
      "epoch": 0.7225192114934847,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004954066390976427,
      "loss": 5.1407,
      "mean_token_accuracy": 0.1954933747649193,
      "num_tokens": 15922011.0,
      "step": 8650
    },
    {
      "entropy": 5.214598512649536,
      "epoch": 0.7229368526561978,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004954006525065639,
      "loss": 5.1124,
      "mean_token_accuracy": 0.18943319767713546,
      "num_tokens": 15930394.0,
      "step": 8655
    },
    {
      "entropy": 5.2756795406341555,
      "epoch": 0.7233544938189108,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004953946620571061,
      "loss": 5.2237,
      "mean_token_accuracy": 0.1842021256685257,
      "num_tokens": 15939880.0,
      "step": 8660
    },
    {
      "entropy": 5.294782829284668,
      "epoch": 0.7237721349816237,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004953886677493746,
      "loss": 5.0967,
      "mean_token_accuracy": 0.19848918318748474,
      "num_tokens": 15948521.0,
      "step": 8665
    },
    {
      "entropy": 5.2965775489807125,
      "epoch": 0.7241897761443368,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004953826695834741,
      "loss": 5.2168,
      "mean_token_accuracy": 0.18476293385028839,
      "num_tokens": 15956811.0,
      "step": 8670
    },
    {
      "entropy": 5.289892244338989,
      "epoch": 0.7246074173070498,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004953766675595098,
      "loss": 5.257,
      "mean_token_accuracy": 0.18657170236110687,
      "num_tokens": 15967019.0,
      "step": 8675
    },
    {
      "entropy": 5.426598834991455,
      "epoch": 0.7250250584697627,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004953706616775866,
      "loss": 5.2212,
      "mean_token_accuracy": 0.18381640315055847,
      "num_tokens": 15976149.0,
      "step": 8680
    },
    {
      "entropy": 5.339119338989258,
      "epoch": 0.7254426996324758,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004953646519378097,
      "loss": 5.0742,
      "mean_token_accuracy": 0.1894015073776245,
      "num_tokens": 15986711.0,
      "step": 8685
    },
    {
      "entropy": 5.259220027923584,
      "epoch": 0.7258603407951888,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004953586383402844,
      "loss": 5.2079,
      "mean_token_accuracy": 0.18817801922559738,
      "num_tokens": 15996482.0,
      "step": 8690
    },
    {
      "entropy": 5.132692193984985,
      "epoch": 0.7262779819579017,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004953526208851157,
      "loss": 5.1042,
      "mean_token_accuracy": 0.18728736639022828,
      "num_tokens": 16005801.0,
      "step": 8695
    },
    {
      "entropy": 5.223638677597046,
      "epoch": 0.7266956231206148,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004953465995724092,
      "loss": 5.1295,
      "mean_token_accuracy": 0.18736834675073624,
      "num_tokens": 16014807.0,
      "step": 8700
    },
    {
      "entropy": 5.381020355224609,
      "epoch": 0.7271132642833278,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0004953405744022705,
      "loss": 5.2837,
      "mean_token_accuracy": 0.18292548656463622,
      "num_tokens": 16023326.0,
      "step": 8705
    },
    {
      "entropy": 5.430472993850708,
      "epoch": 0.7275309054460407,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004953345453748045,
      "loss": 5.23,
      "mean_token_accuracy": 0.18610840290784836,
      "num_tokens": 16031648.0,
      "step": 8710
    },
    {
      "entropy": 5.307134485244751,
      "epoch": 0.7279485466087537,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004953285124901171,
      "loss": 5.1438,
      "mean_token_accuracy": 0.19232644140720367,
      "num_tokens": 16040179.0,
      "step": 8715
    },
    {
      "entropy": 5.301759386062622,
      "epoch": 0.7283661877714668,
      "grad_norm": 1.5234375,
      "learning_rate": 0.000495322475748314,
      "loss": 5.3203,
      "mean_token_accuracy": 0.17622517049312592,
      "num_tokens": 16050770.0,
      "step": 8720
    },
    {
      "entropy": 5.340561103820801,
      "epoch": 0.7287838289341797,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004953164351495007,
      "loss": 5.181,
      "mean_token_accuracy": 0.18876151144504547,
      "num_tokens": 16060504.0,
      "step": 8725
    },
    {
      "entropy": 5.281233406066894,
      "epoch": 0.7292014700968927,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004953103906937831,
      "loss": 5.0292,
      "mean_token_accuracy": 0.20289447158575058,
      "num_tokens": 16069240.0,
      "step": 8730
    },
    {
      "entropy": 5.201758241653442,
      "epoch": 0.7296191112596058,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004953043423812669,
      "loss": 5.2745,
      "mean_token_accuracy": 0.18410247266292573,
      "num_tokens": 16079098.0,
      "step": 8735
    },
    {
      "entropy": 5.213766145706177,
      "epoch": 0.7300367524223187,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004952982902120579,
      "loss": 5.0874,
      "mean_token_accuracy": 0.19202603101730348,
      "num_tokens": 16088454.0,
      "step": 8740
    },
    {
      "entropy": 5.362329149246216,
      "epoch": 0.7304543935850317,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004952922341862622,
      "loss": 5.2434,
      "mean_token_accuracy": 0.18529306799173356,
      "num_tokens": 16096985.0,
      "step": 8745
    },
    {
      "entropy": 5.428350830078125,
      "epoch": 0.7308720347477448,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004952861743039858,
      "loss": 5.3341,
      "mean_token_accuracy": 0.1782863602042198,
      "num_tokens": 16105434.0,
      "step": 8750
    },
    {
      "entropy": 5.224010896682739,
      "epoch": 0.7312896759104577,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004952801105653347,
      "loss": 5.0705,
      "mean_token_accuracy": 0.1973409339785576,
      "num_tokens": 16114933.0,
      "step": 8755
    },
    {
      "entropy": 5.372373008728028,
      "epoch": 0.7317073170731707,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004952740429704152,
      "loss": 5.2499,
      "mean_token_accuracy": 0.18186335861682892,
      "num_tokens": 16124954.0,
      "step": 8760
    },
    {
      "entropy": 5.326339387893677,
      "epoch": 0.7321249582358837,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004952679715193334,
      "loss": 5.0645,
      "mean_token_accuracy": 0.19299455881118774,
      "num_tokens": 16134442.0,
      "step": 8765
    },
    {
      "entropy": 5.337293004989624,
      "epoch": 0.7325425993985967,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004952618962121956,
      "loss": 5.2484,
      "mean_token_accuracy": 0.18174122124910355,
      "num_tokens": 16143994.0,
      "step": 8770
    },
    {
      "entropy": 5.2761390209198,
      "epoch": 0.7329602405613097,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004952558170491082,
      "loss": 5.143,
      "mean_token_accuracy": 0.18572575896978377,
      "num_tokens": 16152931.0,
      "step": 8775
    },
    {
      "entropy": 5.270718765258789,
      "epoch": 0.7333778817240227,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004952497340301776,
      "loss": 5.1735,
      "mean_token_accuracy": 0.18109920769929885,
      "num_tokens": 16162716.0,
      "step": 8780
    },
    {
      "entropy": 5.346599483489991,
      "epoch": 0.7337955228867357,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004952436471555103,
      "loss": 5.23,
      "mean_token_accuracy": 0.18129942566156387,
      "num_tokens": 16172371.0,
      "step": 8785
    },
    {
      "entropy": 5.311984920501709,
      "epoch": 0.7342131640494487,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004952375564252128,
      "loss": 5.3044,
      "mean_token_accuracy": 0.18387712836265563,
      "num_tokens": 16181823.0,
      "step": 8790
    },
    {
      "entropy": 5.302162504196167,
      "epoch": 0.7346308052121617,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004952314618393919,
      "loss": 5.202,
      "mean_token_accuracy": 0.18727488070726395,
      "num_tokens": 16190720.0,
      "step": 8795
    },
    {
      "entropy": 5.323683071136474,
      "epoch": 0.7350484463748747,
      "grad_norm": 1.90625,
      "learning_rate": 0.000495225363398154,
      "loss": 5.1427,
      "mean_token_accuracy": 0.18795779943466187,
      "num_tokens": 16199589.0,
      "step": 8800
    },
    {
      "entropy": 5.39074182510376,
      "epoch": 0.7354660875375877,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004952192611016061,
      "loss": 5.3305,
      "mean_token_accuracy": 0.18400349467992783,
      "num_tokens": 16208800.0,
      "step": 8805
    },
    {
      "entropy": 5.388135766983032,
      "epoch": 0.7358837287003007,
      "grad_norm": 1.65625,
      "learning_rate": 0.000495213154949855,
      "loss": 5.2849,
      "mean_token_accuracy": 0.18048720359802245,
      "num_tokens": 16218835.0,
      "step": 8810
    },
    {
      "entropy": 5.376900386810303,
      "epoch": 0.7363013698630136,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004952070449430075,
      "loss": 5.2467,
      "mean_token_accuracy": 0.1812291622161865,
      "num_tokens": 16228441.0,
      "step": 8815
    },
    {
      "entropy": 5.403039264678955,
      "epoch": 0.7367190110257267,
      "grad_norm": 2.6875,
      "learning_rate": 0.0004952009310811706,
      "loss": 5.2191,
      "mean_token_accuracy": 0.19410694092512132,
      "num_tokens": 16237877.0,
      "step": 8820
    },
    {
      "entropy": 5.265282154083252,
      "epoch": 0.7371366521884397,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004951948133644514,
      "loss": 5.1385,
      "mean_token_accuracy": 0.19013114124536515,
      "num_tokens": 16246809.0,
      "step": 8825
    },
    {
      "entropy": 5.3557374477386475,
      "epoch": 0.7375542933511526,
      "grad_norm": 1.7109375,
      "learning_rate": 0.000495188691792957,
      "loss": 5.2226,
      "mean_token_accuracy": 0.19548689126968383,
      "num_tokens": 16255754.0,
      "step": 8830
    },
    {
      "entropy": 5.33818097114563,
      "epoch": 0.7379719345138657,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004951825663667943,
      "loss": 5.2541,
      "mean_token_accuracy": 0.1831888809800148,
      "num_tokens": 16264861.0,
      "step": 8835
    },
    {
      "entropy": 5.298843860626221,
      "epoch": 0.7383895756765787,
      "grad_norm": 1.6328125,
      "learning_rate": 0.000495176437086071,
      "loss": 5.184,
      "mean_token_accuracy": 0.1878280982375145,
      "num_tokens": 16273628.0,
      "step": 8840
    },
    {
      "entropy": 5.345774793624878,
      "epoch": 0.7388072168392917,
      "grad_norm": 2.0625,
      "learning_rate": 0.000495170303950894,
      "loss": 5.1998,
      "mean_token_accuracy": 0.1875653460621834,
      "num_tokens": 16282958.0,
      "step": 8845
    },
    {
      "entropy": 5.316007518768311,
      "epoch": 0.7392248580020047,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004951641669613708,
      "loss": 5.2169,
      "mean_token_accuracy": 0.18460226953029632,
      "num_tokens": 16292763.0,
      "step": 8850
    },
    {
      "entropy": 5.329687404632568,
      "epoch": 0.7396424991647177,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004951580261176088,
      "loss": 5.1999,
      "mean_token_accuracy": 0.19027262479066848,
      "num_tokens": 16302216.0,
      "step": 8855
    },
    {
      "entropy": 5.23100438117981,
      "epoch": 0.7400601403274307,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004951518814197156,
      "loss": 5.1137,
      "mean_token_accuracy": 0.19242411851882935,
      "num_tokens": 16311672.0,
      "step": 8860
    },
    {
      "entropy": 5.275502109527588,
      "epoch": 0.7404777814901436,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004951457328677989,
      "loss": 5.1435,
      "mean_token_accuracy": 0.18861838132143022,
      "num_tokens": 16321027.0,
      "step": 8865
    },
    {
      "entropy": 5.375608444213867,
      "epoch": 0.7408954226528567,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004951395804619659,
      "loss": 5.2432,
      "mean_token_accuracy": 0.18391497284173966,
      "num_tokens": 16329471.0,
      "step": 8870
    },
    {
      "entropy": 5.309133386611938,
      "epoch": 0.7413130638155697,
      "grad_norm": 3.28125,
      "learning_rate": 0.0004951334242023247,
      "loss": 5.205,
      "mean_token_accuracy": 0.18458560705184937,
      "num_tokens": 16339085.0,
      "step": 8875
    },
    {
      "entropy": 5.363067388534546,
      "epoch": 0.7417307049782826,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004951272640889828,
      "loss": 5.2659,
      "mean_token_accuracy": 0.19278847724199294,
      "num_tokens": 16347950.0,
      "step": 8880
    },
    {
      "entropy": 5.304359006881714,
      "epoch": 0.7421483461409957,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004951211001220484,
      "loss": 5.0592,
      "mean_token_accuracy": 0.1945688992738724,
      "num_tokens": 16357129.0,
      "step": 8885
    },
    {
      "entropy": 5.282190561294556,
      "epoch": 0.7425659873037087,
      "grad_norm": 1.578125,
      "learning_rate": 0.000495114932301629,
      "loss": 5.182,
      "mean_token_accuracy": 0.18669179528951646,
      "num_tokens": 16365768.0,
      "step": 8890
    },
    {
      "entropy": 5.2501928329467775,
      "epoch": 0.7429836284664216,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004951087606278328,
      "loss": 5.1428,
      "mean_token_accuracy": 0.1911521926522255,
      "num_tokens": 16375548.0,
      "step": 8895
    },
    {
      "entropy": 5.325199222564697,
      "epoch": 0.7434012696291347,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004951025851007678,
      "loss": 5.2023,
      "mean_token_accuracy": 0.19053663462400436,
      "num_tokens": 16385472.0,
      "step": 8900
    },
    {
      "entropy": 5.321870899200439,
      "epoch": 0.7438189107918477,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004950964057205421,
      "loss": 5.0785,
      "mean_token_accuracy": 0.18921963274478912,
      "num_tokens": 16395048.0,
      "step": 8905
    },
    {
      "entropy": 5.388301944732666,
      "epoch": 0.7442365519545606,
      "grad_norm": 1.59375,
      "learning_rate": 0.000495090222487264,
      "loss": 5.3106,
      "mean_token_accuracy": 0.1784229427576065,
      "num_tokens": 16405378.0,
      "step": 8910
    },
    {
      "entropy": 5.316306924819946,
      "epoch": 0.7446541931172737,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004950840354010415,
      "loss": 5.1607,
      "mean_token_accuracy": 0.19073249250650406,
      "num_tokens": 16415206.0,
      "step": 8915
    },
    {
      "entropy": 5.375656604766846,
      "epoch": 0.7450718342799867,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004950778444619831,
      "loss": 5.1945,
      "mean_token_accuracy": 0.19712818264961243,
      "num_tokens": 16424185.0,
      "step": 8920
    },
    {
      "entropy": 5.329453420639038,
      "epoch": 0.7454894754426996,
      "grad_norm": 1.8046875,
      "learning_rate": 0.000495071649670197,
      "loss": 5.1654,
      "mean_token_accuracy": 0.1838183030486107,
      "num_tokens": 16433330.0,
      "step": 8925
    },
    {
      "entropy": 5.271111345291137,
      "epoch": 0.7459071166054126,
      "grad_norm": 1.8125,
      "learning_rate": 0.000495065451025792,
      "loss": 5.1994,
      "mean_token_accuracy": 0.1852631837129593,
      "num_tokens": 16442777.0,
      "step": 8930
    },
    {
      "entropy": 5.310748481750489,
      "epoch": 0.7463247577681257,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004950592485288761,
      "loss": 5.1774,
      "mean_token_accuracy": 0.19424661993980408,
      "num_tokens": 16451838.0,
      "step": 8935
    },
    {
      "entropy": 5.314354085922242,
      "epoch": 0.7467423989308386,
      "grad_norm": 2.265625,
      "learning_rate": 0.0004950530421795584,
      "loss": 5.131,
      "mean_token_accuracy": 0.19498806148767472,
      "num_tokens": 16460022.0,
      "step": 8940
    },
    {
      "entropy": 5.21478157043457,
      "epoch": 0.7471600400935516,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0004950468319779472,
      "loss": 5.0241,
      "mean_token_accuracy": 0.19858380109071733,
      "num_tokens": 16468891.0,
      "step": 8945
    },
    {
      "entropy": 5.1760947704315186,
      "epoch": 0.7475776812562647,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004950406179241515,
      "loss": 5.1171,
      "mean_token_accuracy": 0.20099257677793503,
      "num_tokens": 16478020.0,
      "step": 8950
    },
    {
      "entropy": 5.3247462749481205,
      "epoch": 0.7479953224189776,
      "grad_norm": 2.34375,
      "learning_rate": 0.0004950344000182798,
      "loss": 5.2944,
      "mean_token_accuracy": 0.185774627327919,
      "num_tokens": 16487536.0,
      "step": 8955
    },
    {
      "entropy": 5.343104982376099,
      "epoch": 0.7484129635816906,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004950281782604411,
      "loss": 5.2046,
      "mean_token_accuracy": 0.17981323897838591,
      "num_tokens": 16495786.0,
      "step": 8960
    },
    {
      "entropy": 5.380094528198242,
      "epoch": 0.7488306047444037,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004950219526507442,
      "loss": 5.276,
      "mean_token_accuracy": 0.1816750258207321,
      "num_tokens": 16505535.0,
      "step": 8965
    },
    {
      "entropy": 5.417230463027954,
      "epoch": 0.7492482459071166,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004950157231892983,
      "loss": 5.2806,
      "mean_token_accuracy": 0.18207372725009918,
      "num_tokens": 16515387.0,
      "step": 8970
    },
    {
      "entropy": 5.290445327758789,
      "epoch": 0.7496658870698296,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004950094898762123,
      "loss": 5.2302,
      "mean_token_accuracy": 0.18433275073766708,
      "num_tokens": 16524421.0,
      "step": 8975
    },
    {
      "entropy": 5.390057373046875,
      "epoch": 0.7500835282325427,
      "grad_norm": 3.875,
      "learning_rate": 0.0004950032527115954,
      "loss": 5.2178,
      "mean_token_accuracy": 0.18798471987247467,
      "num_tokens": 16534533.0,
      "step": 8980
    },
    {
      "entropy": 5.400622415542602,
      "epoch": 0.7505011693952556,
      "grad_norm": 2.390625,
      "learning_rate": 0.0004949970116955566,
      "loss": 5.2418,
      "mean_token_accuracy": 0.17809109389781952,
      "num_tokens": 16543712.0,
      "step": 8985
    },
    {
      "entropy": 5.364700078964233,
      "epoch": 0.7509188105579686,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004949907668282054,
      "loss": 5.206,
      "mean_token_accuracy": 0.1881457805633545,
      "num_tokens": 16552475.0,
      "step": 8990
    },
    {
      "entropy": 5.34286322593689,
      "epoch": 0.7513364517206816,
      "grad_norm": 1.734375,
      "learning_rate": 0.000494984518109651,
      "loss": 5.2192,
      "mean_token_accuracy": 0.18579161614179612,
      "num_tokens": 16562237.0,
      "step": 8995
    },
    {
      "entropy": 5.346833086013794,
      "epoch": 0.7517540928833946,
      "grad_norm": 1.90625,
      "learning_rate": 0.0004949782655400028,
      "loss": 5.1874,
      "mean_token_accuracy": 0.19135233908891677,
      "num_tokens": 16570508.0,
      "step": 9000
    },
    {
      "epoch": 0.7517540928833946,
      "eval_entropy": 5.133677991005708,
      "eval_loss": 5.168975830078125,
      "eval_mean_token_accuracy": 0.19775329205923428,
      "eval_num_tokens": 16570508.0,
      "eval_runtime": 27.3441,
      "eval_samples_per_second": 1381.027,
      "eval_steps_per_second": 172.651,
      "step": 9000
    },
    {
      "entropy": 5.272634363174438,
      "epoch": 0.7521717340461076,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004949720091193704,
      "loss": 5.0851,
      "mean_token_accuracy": 0.19876011908054353,
      "num_tokens": 16578561.0,
      "step": 9005
    },
    {
      "entropy": 5.300364828109741,
      "epoch": 0.7525893752088206,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000494965748847863,
      "loss": 5.0734,
      "mean_token_accuracy": 0.19513107985258102,
      "num_tokens": 16587168.0,
      "step": 9010
    },
    {
      "entropy": 5.328412818908691,
      "epoch": 0.7530070163715336,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004949594847255905,
      "loss": 5.2357,
      "mean_token_accuracy": 0.18871351927518845,
      "num_tokens": 16597252.0,
      "step": 9015
    },
    {
      "entropy": 5.2942945003509525,
      "epoch": 0.7534246575342466,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004949532167526625,
      "loss": 5.0876,
      "mean_token_accuracy": 0.19655826091766357,
      "num_tokens": 16607558.0,
      "step": 9020
    },
    {
      "entropy": 5.419779872894287,
      "epoch": 0.7538422986969596,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004949469449291886,
      "loss": 5.2756,
      "mean_token_accuracy": 0.1777003213763237,
      "num_tokens": 16617019.0,
      "step": 9025
    },
    {
      "entropy": 5.312467765808106,
      "epoch": 0.7542599398596725,
      "grad_norm": 2.453125,
      "learning_rate": 0.0004949406692552787,
      "loss": 5.1217,
      "mean_token_accuracy": 0.1900804117321968,
      "num_tokens": 16625815.0,
      "step": 9030
    },
    {
      "entropy": 5.358976984024048,
      "epoch": 0.7546775810223856,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004949343897310426,
      "loss": 5.329,
      "mean_token_accuracy": 0.17723113745450975,
      "num_tokens": 16634876.0,
      "step": 9035
    },
    {
      "entropy": 5.319064283370972,
      "epoch": 0.7550952221850986,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004949281063565904,
      "loss": 5.2199,
      "mean_token_accuracy": 0.18531542420387268,
      "num_tokens": 16643770.0,
      "step": 9040
    },
    {
      "entropy": 5.3930223941802975,
      "epoch": 0.7555128633478115,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004949218191320319,
      "loss": 5.166,
      "mean_token_accuracy": 0.18985310196876526,
      "num_tokens": 16652845.0,
      "step": 9045
    },
    {
      "entropy": 5.336880540847778,
      "epoch": 0.7559305045105246,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004949155280574772,
      "loss": 5.1343,
      "mean_token_accuracy": 0.1917647212743759,
      "num_tokens": 16662224.0,
      "step": 9050
    },
    {
      "entropy": 5.3161591529846195,
      "epoch": 0.7563481456732376,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0004949092331330365,
      "loss": 5.1616,
      "mean_token_accuracy": 0.18797622621059418,
      "num_tokens": 16671114.0,
      "step": 9055
    },
    {
      "entropy": 5.347621059417724,
      "epoch": 0.7567657868359505,
      "grad_norm": 2.0625,
      "learning_rate": 0.00049490293435882,
      "loss": 5.2289,
      "mean_token_accuracy": 0.1742393746972084,
      "num_tokens": 16680395.0,
      "step": 9060
    },
    {
      "entropy": 5.377555990219117,
      "epoch": 0.7571834279986636,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004948966317349379,
      "loss": 5.1723,
      "mean_token_accuracy": 0.19361136853694916,
      "num_tokens": 16690251.0,
      "step": 9065
    },
    {
      "entropy": 5.27882399559021,
      "epoch": 0.7576010691613766,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004948903252615006,
      "loss": 5.0676,
      "mean_token_accuracy": 0.2000073954463005,
      "num_tokens": 16698945.0,
      "step": 9070
    },
    {
      "entropy": 5.349974203109741,
      "epoch": 0.7580187103240895,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004948840149386184,
      "loss": 5.2575,
      "mean_token_accuracy": 0.18208573907613754,
      "num_tokens": 16708325.0,
      "step": 9075
    },
    {
      "entropy": 5.315873098373413,
      "epoch": 0.7584363514868026,
      "grad_norm": 1.5234375,
      "learning_rate": 0.000494877700766402,
      "loss": 5.1602,
      "mean_token_accuracy": 0.1909972459077835,
      "num_tokens": 16716392.0,
      "step": 9080
    },
    {
      "entropy": 5.320262050628662,
      "epoch": 0.7588539926495156,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004948713827449616,
      "loss": 5.1721,
      "mean_token_accuracy": 0.18527687340974808,
      "num_tokens": 16725284.0,
      "step": 9085
    },
    {
      "entropy": 5.2729230403900145,
      "epoch": 0.7592716338122285,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004948650608744081,
      "loss": 5.196,
      "mean_token_accuracy": 0.1946970194578171,
      "num_tokens": 16734658.0,
      "step": 9090
    },
    {
      "entropy": 5.419935989379883,
      "epoch": 0.7596892749749415,
      "grad_norm": 2.515625,
      "learning_rate": 0.000494858735154852,
      "loss": 5.3093,
      "mean_token_accuracy": 0.1810906082391739,
      "num_tokens": 16744403.0,
      "step": 9095
    },
    {
      "entropy": 5.41253628730774,
      "epoch": 0.7601069161376546,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004948524055864041,
      "loss": 5.2691,
      "mean_token_accuracy": 0.18740759789943695,
      "num_tokens": 16754254.0,
      "step": 9100
    },
    {
      "entropy": 5.3168446063995365,
      "epoch": 0.7605245573003675,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004948460721691752,
      "loss": 5.2766,
      "mean_token_accuracy": 0.18386004269123077,
      "num_tokens": 16763941.0,
      "step": 9105
    },
    {
      "entropy": 5.364061737060547,
      "epoch": 0.7609421984630805,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0004948397349032762,
      "loss": 5.1873,
      "mean_token_accuracy": 0.1827145352959633,
      "num_tokens": 16773181.0,
      "step": 9110
    },
    {
      "entropy": 5.2379027843475345,
      "epoch": 0.7613598396257936,
      "grad_norm": 1.6484375,
      "learning_rate": 0.000494833393788818,
      "loss": 5.0208,
      "mean_token_accuracy": 0.1977047637104988,
      "num_tokens": 16782622.0,
      "step": 9115
    },
    {
      "entropy": 5.358839702606201,
      "epoch": 0.7617774807885065,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004948270488259117,
      "loss": 5.2721,
      "mean_token_accuracy": 0.1830137401819229,
      "num_tokens": 16792805.0,
      "step": 9120
    },
    {
      "entropy": 5.32723560333252,
      "epoch": 0.7621951219512195,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0004948207000146682,
      "loss": 5.227,
      "mean_token_accuracy": 0.18766830265522003,
      "num_tokens": 16802641.0,
      "step": 9125
    },
    {
      "entropy": 5.391316843032837,
      "epoch": 0.7626127631139326,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004948143473551989,
      "loss": 5.2224,
      "mean_token_accuracy": 0.1805815115571022,
      "num_tokens": 16812037.0,
      "step": 9130
    },
    {
      "entropy": 5.2934071063995365,
      "epoch": 0.7630304042766455,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004948079908476147,
      "loss": 5.1753,
      "mean_token_accuracy": 0.19688214212656022,
      "num_tokens": 16820279.0,
      "step": 9135
    },
    {
      "entropy": 5.285008192062378,
      "epoch": 0.7634480454393585,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0004948016304920271,
      "loss": 5.1601,
      "mean_token_accuracy": 0.18657118529081346,
      "num_tokens": 16829396.0,
      "step": 9140
    },
    {
      "entropy": 5.311561822891235,
      "epoch": 0.7638656866020715,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004947952662885473,
      "loss": 5.139,
      "mean_token_accuracy": 0.194387049973011,
      "num_tokens": 16838389.0,
      "step": 9145
    },
    {
      "entropy": 5.374364137649536,
      "epoch": 0.7642833277647845,
      "grad_norm": 1.5625,
      "learning_rate": 0.000494788898237287,
      "loss": 5.2669,
      "mean_token_accuracy": 0.18704009801149368,
      "num_tokens": 16848012.0,
      "step": 9150
    },
    {
      "entropy": 5.387968873977661,
      "epoch": 0.7647009689274975,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004947825263383573,
      "loss": 5.2326,
      "mean_token_accuracy": 0.18611233234405516,
      "num_tokens": 16857591.0,
      "step": 9155
    },
    {
      "entropy": 5.392985248565674,
      "epoch": 0.7651186100902105,
      "grad_norm": 1.546875,
      "learning_rate": 0.00049477615059187,
      "loss": 5.2136,
      "mean_token_accuracy": 0.17910126298666001,
      "num_tokens": 16866792.0,
      "step": 9160
    },
    {
      "entropy": 5.325544166564941,
      "epoch": 0.7655362512529235,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004947697709979367,
      "loss": 5.2457,
      "mean_token_accuracy": 0.18787939697504044,
      "num_tokens": 16876602.0,
      "step": 9165
    },
    {
      "entropy": 5.3107171058654785,
      "epoch": 0.7659538924156365,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004947633875566688,
      "loss": 5.2456,
      "mean_token_accuracy": 0.18358664959669113,
      "num_tokens": 16886992.0,
      "step": 9170
    },
    {
      "entropy": 5.3244751453399655,
      "epoch": 0.7663715335783495,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004947570002681785,
      "loss": 5.1464,
      "mean_token_accuracy": 0.1884155884385109,
      "num_tokens": 16895716.0,
      "step": 9175
    },
    {
      "entropy": 5.3303642749786375,
      "epoch": 0.7667891747410625,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004947506091325772,
      "loss": 5.1467,
      "mean_token_accuracy": 0.19776746332645417,
      "num_tokens": 16904568.0,
      "step": 9180
    },
    {
      "entropy": 5.410282278060913,
      "epoch": 0.7672068159037755,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004947442141499771,
      "loss": 5.2502,
      "mean_token_accuracy": 0.18748794198036195,
      "num_tokens": 16913667.0,
      "step": 9185
    },
    {
      "entropy": 5.33154182434082,
      "epoch": 0.7676244570664885,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004947378153204899,
      "loss": 5.1772,
      "mean_token_accuracy": 0.18723595887422562,
      "num_tokens": 16922897.0,
      "step": 9190
    },
    {
      "entropy": 5.321815347671508,
      "epoch": 0.7680420982292014,
      "grad_norm": 2.234375,
      "learning_rate": 0.0004947314126442279,
      "loss": 5.1931,
      "mean_token_accuracy": 0.18120656311511993,
      "num_tokens": 16931559.0,
      "step": 9195
    },
    {
      "entropy": 5.326811933517456,
      "epoch": 0.7684597393919145,
      "grad_norm": 1.7578125,
      "learning_rate": 0.000494725006121303,
      "loss": 5.1614,
      "mean_token_accuracy": 0.1848924219608307,
      "num_tokens": 16940191.0,
      "step": 9200
    },
    {
      "entropy": 5.371112871170044,
      "epoch": 0.7688773805546275,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004947185957518273,
      "loss": 5.24,
      "mean_token_accuracy": 0.18787964135408403,
      "num_tokens": 16948971.0,
      "step": 9205
    },
    {
      "entropy": 5.376923370361328,
      "epoch": 0.7692950217173404,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004947121815359133,
      "loss": 5.1824,
      "mean_token_accuracy": 0.19100566506385802,
      "num_tokens": 16957944.0,
      "step": 9210
    },
    {
      "entropy": 5.32720103263855,
      "epoch": 0.7697126628800535,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004947057634736729,
      "loss": 5.1513,
      "mean_token_accuracy": 0.18390686362981795,
      "num_tokens": 16967418.0,
      "step": 9215
    },
    {
      "entropy": 5.265461206436157,
      "epoch": 0.7701303040427665,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004946993415652186,
      "loss": 5.1423,
      "mean_token_accuracy": 0.1907045751810074,
      "num_tokens": 16977233.0,
      "step": 9220
    },
    {
      "entropy": 5.309878253936768,
      "epoch": 0.7705479452054794,
      "grad_norm": 1.703125,
      "learning_rate": 0.000494692915810663,
      "loss": 5.1846,
      "mean_token_accuracy": 0.1941950038075447,
      "num_tokens": 16986023.0,
      "step": 9225
    },
    {
      "entropy": 5.371839141845703,
      "epoch": 0.7709655863681925,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004946864862101185,
      "loss": 5.1379,
      "mean_token_accuracy": 0.18917000889778138,
      "num_tokens": 16994987.0,
      "step": 9230
    },
    {
      "entropy": 5.297115755081177,
      "epoch": 0.7713832275309055,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004946800527636975,
      "loss": 5.1623,
      "mean_token_accuracy": 0.18845341652631759,
      "num_tokens": 17003902.0,
      "step": 9235
    },
    {
      "entropy": 5.363648509979248,
      "epoch": 0.7718008686936184,
      "grad_norm": 2.125,
      "learning_rate": 0.0004946736154715129,
      "loss": 5.2557,
      "mean_token_accuracy": 0.18313714861869812,
      "num_tokens": 17013258.0,
      "step": 9240
    },
    {
      "entropy": 5.406899166107178,
      "epoch": 0.7722185098563314,
      "grad_norm": 2.125,
      "learning_rate": 0.0004946671743336771,
      "loss": 5.2462,
      "mean_token_accuracy": 0.1832134172320366,
      "num_tokens": 17022258.0,
      "step": 9245
    },
    {
      "entropy": 5.3464460372924805,
      "epoch": 0.7726361510190445,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004946607293503031,
      "loss": 5.1778,
      "mean_token_accuracy": 0.18468582928180693,
      "num_tokens": 17031485.0,
      "step": 9250
    },
    {
      "entropy": 5.22146167755127,
      "epoch": 0.7730537921817574,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004946542805215036,
      "loss": 5.1238,
      "mean_token_accuracy": 0.19387171864509584,
      "num_tokens": 17039916.0,
      "step": 9255
    },
    {
      "entropy": 5.280299949645996,
      "epoch": 0.7734714333444704,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004946478278473914,
      "loss": 5.1706,
      "mean_token_accuracy": 0.1912824898958206,
      "num_tokens": 17048153.0,
      "step": 9260
    },
    {
      "entropy": 5.310075855255127,
      "epoch": 0.7738890745071835,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004946413713280797,
      "loss": 5.2034,
      "mean_token_accuracy": 0.18531114161014556,
      "num_tokens": 17057473.0,
      "step": 9265
    },
    {
      "entropy": 5.363498067855835,
      "epoch": 0.7743067156698964,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004946349109636814,
      "loss": 5.1999,
      "mean_token_accuracy": 0.1903122827410698,
      "num_tokens": 17066431.0,
      "step": 9270
    },
    {
      "entropy": 5.39309287071228,
      "epoch": 0.7747243568326094,
      "grad_norm": 2.5,
      "learning_rate": 0.0004946284467543096,
      "loss": 5.3107,
      "mean_token_accuracy": 0.1767013356089592,
      "num_tokens": 17076519.0,
      "step": 9275
    },
    {
      "entropy": 5.35628457069397,
      "epoch": 0.7751419979953225,
      "grad_norm": 1.9375,
      "learning_rate": 0.0004946219787000776,
      "loss": 5.2118,
      "mean_token_accuracy": 0.18584041744470597,
      "num_tokens": 17084651.0,
      "step": 9280
    },
    {
      "entropy": 5.425425672531128,
      "epoch": 0.7755596391580354,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004946155068010984,
      "loss": 5.2682,
      "mean_token_accuracy": 0.1853787124156952,
      "num_tokens": 17094943.0,
      "step": 9285
    },
    {
      "entropy": 5.2977934837341305,
      "epoch": 0.7759772803207484,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004946090310574854,
      "loss": 5.1227,
      "mean_token_accuracy": 0.19857468008995055,
      "num_tokens": 17103548.0,
      "step": 9290
    },
    {
      "entropy": 5.27626314163208,
      "epoch": 0.7763949214834615,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004946025514693519,
      "loss": 5.1027,
      "mean_token_accuracy": 0.18831444829702376,
      "num_tokens": 17112472.0,
      "step": 9295
    },
    {
      "entropy": 5.424504995346069,
      "epoch": 0.7768125626461744,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004945960680368115,
      "loss": 5.2452,
      "mean_token_accuracy": 0.1821184054017067,
      "num_tokens": 17121536.0,
      "step": 9300
    },
    {
      "entropy": 5.452235221862793,
      "epoch": 0.7772302038088874,
      "grad_norm": 2.515625,
      "learning_rate": 0.0004945895807599775,
      "loss": 5.286,
      "mean_token_accuracy": 0.1766143783926964,
      "num_tokens": 17130708.0,
      "step": 9305
    },
    {
      "entropy": 5.259407567977905,
      "epoch": 0.7776478449716004,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004945830896389636,
      "loss": 5.0701,
      "mean_token_accuracy": 0.19628924280405044,
      "num_tokens": 17139437.0,
      "step": 9310
    },
    {
      "entropy": 5.295243883132935,
      "epoch": 0.7780654861343134,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004945765946738835,
      "loss": 5.2454,
      "mean_token_accuracy": 0.18049858659505844,
      "num_tokens": 17149107.0,
      "step": 9315
    },
    {
      "entropy": 5.3829522132873535,
      "epoch": 0.7784831272970264,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004945700958648507,
      "loss": 5.2267,
      "mean_token_accuracy": 0.18175995647907256,
      "num_tokens": 17158866.0,
      "step": 9320
    },
    {
      "entropy": 5.346366882324219,
      "epoch": 0.7789007684597394,
      "grad_norm": 1.765625,
      "learning_rate": 0.000494563593211979,
      "loss": 5.108,
      "mean_token_accuracy": 0.19358669072389603,
      "num_tokens": 17167909.0,
      "step": 9325
    },
    {
      "entropy": 5.34268593788147,
      "epoch": 0.7793184096224524,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004945570867153826,
      "loss": 5.2358,
      "mean_token_accuracy": 0.1824558898806572,
      "num_tokens": 17176746.0,
      "step": 9330
    },
    {
      "entropy": 5.31865668296814,
      "epoch": 0.7797360507851654,
      "grad_norm": 2.1875,
      "learning_rate": 0.0004945505763751749,
      "loss": 5.2378,
      "mean_token_accuracy": 0.1797788992524147,
      "num_tokens": 17186219.0,
      "step": 9335
    },
    {
      "entropy": 5.359292221069336,
      "epoch": 0.7801536919478784,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004945440621914702,
      "loss": 5.0901,
      "mean_token_accuracy": 0.19835327565670013,
      "num_tokens": 17195811.0,
      "step": 9340
    },
    {
      "entropy": 5.29733943939209,
      "epoch": 0.7805713331105913,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004945375441643823,
      "loss": 5.237,
      "mean_token_accuracy": 0.18875968903303147,
      "num_tokens": 17205333.0,
      "step": 9345
    },
    {
      "entropy": 5.36309585571289,
      "epoch": 0.7809889742733044,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004945310222940254,
      "loss": 5.1722,
      "mean_token_accuracy": 0.18497316241264344,
      "num_tokens": 17214493.0,
      "step": 9350
    },
    {
      "entropy": 5.447697019577026,
      "epoch": 0.7814066154360174,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004945244965805138,
      "loss": 5.2195,
      "mean_token_accuracy": 0.18682510107755662,
      "num_tokens": 17223602.0,
      "step": 9355
    },
    {
      "entropy": 5.441306400299072,
      "epoch": 0.7818242565987303,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004945179670239616,
      "loss": 5.2394,
      "mean_token_accuracy": 0.17852559983730315,
      "num_tokens": 17232915.0,
      "step": 9360
    },
    {
      "entropy": 5.338666677474976,
      "epoch": 0.7822418977614434,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004945114336244832,
      "loss": 5.1955,
      "mean_token_accuracy": 0.18359384685754776,
      "num_tokens": 17242489.0,
      "step": 9365
    },
    {
      "entropy": 5.22138524055481,
      "epoch": 0.7826595389241564,
      "grad_norm": 1.5546875,
      "learning_rate": 0.000494504896382193,
      "loss": 4.9625,
      "mean_token_accuracy": 0.1995940700173378,
      "num_tokens": 17251496.0,
      "step": 9370
    },
    {
      "entropy": 5.272387266159058,
      "epoch": 0.7830771800868693,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004944983552972053,
      "loss": 5.1484,
      "mean_token_accuracy": 0.19191693365573884,
      "num_tokens": 17260550.0,
      "step": 9375
    },
    {
      "entropy": 5.31194658279419,
      "epoch": 0.7834948212495824,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0004944918103696347,
      "loss": 5.1109,
      "mean_token_accuracy": 0.19449665546417236,
      "num_tokens": 17269085.0,
      "step": 9380
    },
    {
      "entropy": 5.404148912429809,
      "epoch": 0.7839124624122954,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004944852615995956,
      "loss": 5.3652,
      "mean_token_accuracy": 0.18360071033239364,
      "num_tokens": 17278766.0,
      "step": 9385
    },
    {
      "entropy": 5.463492679595947,
      "epoch": 0.7843301035750083,
      "grad_norm": 1.6640625,
      "learning_rate": 0.000494478708987203,
      "loss": 5.2165,
      "mean_token_accuracy": 0.18161397576332092,
      "num_tokens": 17288316.0,
      "step": 9390
    },
    {
      "entropy": 5.2378357410430905,
      "epoch": 0.7847477447377214,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004944721525325713,
      "loss": 5.0727,
      "mean_token_accuracy": 0.18655068576335906,
      "num_tokens": 17297829.0,
      "step": 9395
    },
    {
      "entropy": 5.304569292068481,
      "epoch": 0.7851653859004344,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004944655922358154,
      "loss": 5.1548,
      "mean_token_accuracy": 0.18702153265476226,
      "num_tokens": 17306631.0,
      "step": 9400
    },
    {
      "entropy": 5.32991132736206,
      "epoch": 0.7855830270631473,
      "grad_norm": 1.5,
      "learning_rate": 0.0004944590280970503,
      "loss": 5.145,
      "mean_token_accuracy": 0.19350167214870453,
      "num_tokens": 17316829.0,
      "step": 9405
    },
    {
      "entropy": 5.319064426422119,
      "epoch": 0.7860006682258603,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004944524601163905,
      "loss": 5.1833,
      "mean_token_accuracy": 0.18373969346284866,
      "num_tokens": 17326568.0,
      "step": 9410
    },
    {
      "entropy": 5.375158071517944,
      "epoch": 0.7864183093885734,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004944458882939513,
      "loss": 5.2468,
      "mean_token_accuracy": 0.1819099396467209,
      "num_tokens": 17335524.0,
      "step": 9415
    },
    {
      "entropy": 5.443592643737793,
      "epoch": 0.7868359505512863,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004944393126298478,
      "loss": 5.2737,
      "mean_token_accuracy": 0.18035196363925934,
      "num_tokens": 17346198.0,
      "step": 9420
    },
    {
      "entropy": 5.311717510223389,
      "epoch": 0.7872535917139993,
      "grad_norm": 1.375,
      "learning_rate": 0.000494432733124195,
      "loss": 5.1182,
      "mean_token_accuracy": 0.18798684179782868,
      "num_tokens": 17354696.0,
      "step": 9425
    },
    {
      "entropy": 5.210242652893067,
      "epoch": 0.7876712328767124,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000494426149777108,
      "loss": 5.0869,
      "mean_token_accuracy": 0.1907953992486,
      "num_tokens": 17363314.0,
      "step": 9430
    },
    {
      "entropy": 5.270390796661377,
      "epoch": 0.7880888740394253,
      "grad_norm": 1.8359375,
      "learning_rate": 0.000494419562588702,
      "loss": 5.1519,
      "mean_token_accuracy": 0.19028382301330565,
      "num_tokens": 17372662.0,
      "step": 9435
    },
    {
      "entropy": 5.316920042037964,
      "epoch": 0.7885065152021383,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004944129715590927,
      "loss": 5.1268,
      "mean_token_accuracy": 0.1886994644999504,
      "num_tokens": 17381931.0,
      "step": 9440
    },
    {
      "entropy": 5.3894792079925535,
      "epoch": 0.7889241563648514,
      "grad_norm": 1.96875,
      "learning_rate": 0.000494406376688395,
      "loss": 5.2182,
      "mean_token_accuracy": 0.18322570770978927,
      "num_tokens": 17391210.0,
      "step": 9445
    },
    {
      "entropy": 5.243765878677368,
      "epoch": 0.7893417975275643,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004943997779767248,
      "loss": 5.0887,
      "mean_token_accuracy": 0.19628691226243972,
      "num_tokens": 17401102.0,
      "step": 9450
    },
    {
      "entropy": 5.261760663986206,
      "epoch": 0.7897594386902773,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004943931754241972,
      "loss": 5.1302,
      "mean_token_accuracy": 0.1869055986404419,
      "num_tokens": 17410036.0,
      "step": 9455
    },
    {
      "entropy": 5.318477916717529,
      "epoch": 0.7901770798529903,
      "grad_norm": 1.828125,
      "learning_rate": 0.000494386569030928,
      "loss": 5.1692,
      "mean_token_accuracy": 0.18524761795997619,
      "num_tokens": 17419348.0,
      "step": 9460
    },
    {
      "entropy": 5.360423946380616,
      "epoch": 0.7905947210157033,
      "grad_norm": 2.40625,
      "learning_rate": 0.0004943799587970328,
      "loss": 5.1765,
      "mean_token_accuracy": 0.1848449930548668,
      "num_tokens": 17427937.0,
      "step": 9465
    },
    {
      "entropy": 5.277559375762939,
      "epoch": 0.7910123621784163,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004943733447226275,
      "loss": 5.0565,
      "mean_token_accuracy": 0.19531989693641663,
      "num_tokens": 17436259.0,
      "step": 9470
    },
    {
      "entropy": 5.35192346572876,
      "epoch": 0.7914300033411293,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004943667268078277,
      "loss": 5.2486,
      "mean_token_accuracy": 0.18139349669218063,
      "num_tokens": 17445202.0,
      "step": 9475
    },
    {
      "entropy": 5.317629814147949,
      "epoch": 0.7918476445038423,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004943601050527493,
      "loss": 5.11,
      "mean_token_accuracy": 0.19007206559181214,
      "num_tokens": 17453928.0,
      "step": 9480
    },
    {
      "entropy": 5.3239970207214355,
      "epoch": 0.7922652856665553,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004943534794575081,
      "loss": 5.1595,
      "mean_token_accuracy": 0.1924855336546898,
      "num_tokens": 17462609.0,
      "step": 9485
    },
    {
      "entropy": 5.336129665374756,
      "epoch": 0.7926829268292683,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004943468500222203,
      "loss": 5.176,
      "mean_token_accuracy": 0.18830626904964448,
      "num_tokens": 17471326.0,
      "step": 9490
    },
    {
      "entropy": 5.284777736663818,
      "epoch": 0.7931005679919813,
      "grad_norm": 1.625,
      "learning_rate": 0.0004943402167470019,
      "loss": 5.1383,
      "mean_token_accuracy": 0.19210643768310548,
      "num_tokens": 17480296.0,
      "step": 9495
    },
    {
      "entropy": 5.250183916091919,
      "epoch": 0.7935182091546943,
      "grad_norm": 1.703125,
      "learning_rate": 0.000494333579631969,
      "loss": 5.0905,
      "mean_token_accuracy": 0.1895009309053421,
      "num_tokens": 17489779.0,
      "step": 9500
    },
    {
      "entropy": 5.308423089981079,
      "epoch": 0.7939358503174073,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004943269386772376,
      "loss": 5.2094,
      "mean_token_accuracy": 0.18017989993095399,
      "num_tokens": 17499796.0,
      "step": 9505
    },
    {
      "entropy": 5.392049121856689,
      "epoch": 0.7943534914801202,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004943202938829243,
      "loss": 5.287,
      "mean_token_accuracy": 0.18167537897825242,
      "num_tokens": 17509223.0,
      "step": 9510
    },
    {
      "entropy": 5.340440464019776,
      "epoch": 0.7947711326428333,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004943136452491452,
      "loss": 5.1864,
      "mean_token_accuracy": 0.18735254257917405,
      "num_tokens": 17518712.0,
      "step": 9515
    },
    {
      "entropy": 5.278647708892822,
      "epoch": 0.7951887738055463,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004943069927760168,
      "loss": 5.117,
      "mean_token_accuracy": 0.19221056997776031,
      "num_tokens": 17528648.0,
      "step": 9520
    },
    {
      "entropy": 5.215114593505859,
      "epoch": 0.7956064149682592,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004943003364636554,
      "loss": 5.1399,
      "mean_token_accuracy": 0.18678269982337953,
      "num_tokens": 17537859.0,
      "step": 9525
    },
    {
      "entropy": 5.372722578048706,
      "epoch": 0.7960240561309723,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004942936763121777,
      "loss": 5.2746,
      "mean_token_accuracy": 0.18087439835071564,
      "num_tokens": 17546438.0,
      "step": 9530
    },
    {
      "entropy": 5.311570692062378,
      "epoch": 0.7964416972936853,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004942870123217002,
      "loss": 5.0906,
      "mean_token_accuracy": 0.19542259871959686,
      "num_tokens": 17554537.0,
      "step": 9535
    },
    {
      "entropy": 5.428062295913696,
      "epoch": 0.7968593384563982,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004942803444923398,
      "loss": 5.329,
      "mean_token_accuracy": 0.17997895628213884,
      "num_tokens": 17564239.0,
      "step": 9540
    },
    {
      "entropy": 5.310004758834839,
      "epoch": 0.7972769796191113,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004942736728242127,
      "loss": 5.2042,
      "mean_token_accuracy": 0.18477356284856797,
      "num_tokens": 17574437.0,
      "step": 9545
    },
    {
      "entropy": 5.249059724807739,
      "epoch": 0.7976946207818243,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004942669973174363,
      "loss": 5.138,
      "mean_token_accuracy": 0.18902096301317214,
      "num_tokens": 17583173.0,
      "step": 9550
    },
    {
      "entropy": 5.3453527927398685,
      "epoch": 0.7981122619445372,
      "grad_norm": 1.7109375,
      "learning_rate": 0.000494260317972127,
      "loss": 5.1849,
      "mean_token_accuracy": 0.1859835907816887,
      "num_tokens": 17591731.0,
      "step": 9555
    },
    {
      "entropy": 5.38207745552063,
      "epoch": 0.7985299031072502,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004942536347884019,
      "loss": 5.2059,
      "mean_token_accuracy": 0.18607254177331925,
      "num_tokens": 17600778.0,
      "step": 9560
    },
    {
      "entropy": 5.273932456970215,
      "epoch": 0.7989475442699633,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004942469477663782,
      "loss": 5.1716,
      "mean_token_accuracy": 0.2019364818930626,
      "num_tokens": 17609470.0,
      "step": 9565
    },
    {
      "entropy": 5.2391273975372314,
      "epoch": 0.7993651854326762,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004942402569061726,
      "loss": 5.0865,
      "mean_token_accuracy": 0.19261219650506972,
      "num_tokens": 17618434.0,
      "step": 9570
    },
    {
      "entropy": 5.2402228832244875,
      "epoch": 0.7997828265953892,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004942335622079025,
      "loss": 5.2054,
      "mean_token_accuracy": 0.1846891924738884,
      "num_tokens": 17628170.0,
      "step": 9575
    },
    {
      "entropy": 5.277979469299316,
      "epoch": 0.8002004677581023,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004942268636716848,
      "loss": 5.1303,
      "mean_token_accuracy": 0.18975522816181184,
      "num_tokens": 17637391.0,
      "step": 9580
    },
    {
      "entropy": 5.34949049949646,
      "epoch": 0.8006181089208152,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004942201612976372,
      "loss": 5.21,
      "mean_token_accuracy": 0.19083760529756547,
      "num_tokens": 17647466.0,
      "step": 9585
    },
    {
      "entropy": 5.313105869293213,
      "epoch": 0.8010357500835282,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004942134550858767,
      "loss": 5.1392,
      "mean_token_accuracy": 0.19590833485126496,
      "num_tokens": 17656155.0,
      "step": 9590
    },
    {
      "entropy": 5.417249298095703,
      "epoch": 0.8014533912462413,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004942067450365208,
      "loss": 5.2324,
      "mean_token_accuracy": 0.18625636547803878,
      "num_tokens": 17665242.0,
      "step": 9595
    },
    {
      "entropy": 5.3511634349823,
      "epoch": 0.8018710324089542,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004942000311496871,
      "loss": 5.2572,
      "mean_token_accuracy": 0.18960424214601518,
      "num_tokens": 17674964.0,
      "step": 9600
    },
    {
      "entropy": 5.402841711044312,
      "epoch": 0.8022886735716672,
      "grad_norm": 1.75,
      "learning_rate": 0.0004941933134254928,
      "loss": 5.2394,
      "mean_token_accuracy": 0.1859276622533798,
      "num_tokens": 17683275.0,
      "step": 9605
    },
    {
      "entropy": 5.266304683685303,
      "epoch": 0.8027063147343803,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004941865918640559,
      "loss": 5.1913,
      "mean_token_accuracy": 0.18445036709308624,
      "num_tokens": 17691841.0,
      "step": 9610
    },
    {
      "entropy": 5.318730354309082,
      "epoch": 0.8031239558970932,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004941798664654937,
      "loss": 5.1629,
      "mean_token_accuracy": 0.19045049995183944,
      "num_tokens": 17701258.0,
      "step": 9615
    },
    {
      "entropy": 5.415330648422241,
      "epoch": 0.8035415970598062,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004941731372299241,
      "loss": 5.2355,
      "mean_token_accuracy": 0.1848721519112587,
      "num_tokens": 17711028.0,
      "step": 9620
    },
    {
      "entropy": 5.370194816589356,
      "epoch": 0.8039592382225192,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004941664041574651,
      "loss": 5.1717,
      "mean_token_accuracy": 0.2002640411257744,
      "num_tokens": 17720175.0,
      "step": 9625
    },
    {
      "entropy": 5.322130966186523,
      "epoch": 0.8043768793852322,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004941596672482343,
      "loss": 5.158,
      "mean_token_accuracy": 0.18256354182958603,
      "num_tokens": 17730420.0,
      "step": 9630
    },
    {
      "entropy": 5.32376012802124,
      "epoch": 0.8047945205479452,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004941529265023498,
      "loss": 5.2532,
      "mean_token_accuracy": 0.1815265953540802,
      "num_tokens": 17739802.0,
      "step": 9635
    },
    {
      "entropy": 5.34448390007019,
      "epoch": 0.8052121617106582,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004941461819199293,
      "loss": 5.1352,
      "mean_token_accuracy": 0.192313115298748,
      "num_tokens": 17749450.0,
      "step": 9640
    },
    {
      "entropy": 5.457966899871826,
      "epoch": 0.8056298028733712,
      "grad_norm": 1.625,
      "learning_rate": 0.0004941394335010912,
      "loss": 5.3361,
      "mean_token_accuracy": 0.1806011214852333,
      "num_tokens": 17758733.0,
      "step": 9645
    },
    {
      "entropy": 5.298712062835693,
      "epoch": 0.8060474440360842,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004941326812459535,
      "loss": 5.2275,
      "mean_token_accuracy": 0.18545981049537658,
      "num_tokens": 17769037.0,
      "step": 9650
    },
    {
      "entropy": 5.266812705993653,
      "epoch": 0.8064650851987972,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004941259251546346,
      "loss": 5.1853,
      "mean_token_accuracy": 0.17927865833044052,
      "num_tokens": 17777603.0,
      "step": 9655
    },
    {
      "entropy": 5.304557132720947,
      "epoch": 0.8068827263615101,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004941191652272526,
      "loss": 5.1931,
      "mean_token_accuracy": 0.1796041786670685,
      "num_tokens": 17787513.0,
      "step": 9660
    },
    {
      "entropy": 5.31368932723999,
      "epoch": 0.8073003675242232,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004941124014639258,
      "loss": 5.1645,
      "mean_token_accuracy": 0.19038289189338684,
      "num_tokens": 17796720.0,
      "step": 9665
    },
    {
      "entropy": 5.333065700531006,
      "epoch": 0.8077180086869362,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004941056338647726,
      "loss": 5.1458,
      "mean_token_accuracy": 0.19311497509479522,
      "num_tokens": 17806024.0,
      "step": 9670
    },
    {
      "entropy": 5.235565567016602,
      "epoch": 0.8081356498496491,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004940988624299116,
      "loss": 5.1258,
      "mean_token_accuracy": 0.18659957349300385,
      "num_tokens": 17815247.0,
      "step": 9675
    },
    {
      "entropy": 5.419852590560913,
      "epoch": 0.8085532910123622,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004940920871594613,
      "loss": 5.1757,
      "mean_token_accuracy": 0.18847472071647645,
      "num_tokens": 17824024.0,
      "step": 9680
    },
    {
      "entropy": 5.292865324020386,
      "epoch": 0.8089709321750752,
      "grad_norm": 2.265625,
      "learning_rate": 0.0004940853080535403,
      "loss": 5.1548,
      "mean_token_accuracy": 0.18051009625196457,
      "num_tokens": 17832782.0,
      "step": 9685
    },
    {
      "entropy": 5.228757667541504,
      "epoch": 0.8093885733377881,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004940785251122673,
      "loss": 5.1138,
      "mean_token_accuracy": 0.1954245463013649,
      "num_tokens": 17841961.0,
      "step": 9690
    },
    {
      "entropy": 5.27298641204834,
      "epoch": 0.8098062145005012,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004940717383357608,
      "loss": 5.221,
      "mean_token_accuracy": 0.19371180683374406,
      "num_tokens": 17851364.0,
      "step": 9695
    },
    {
      "entropy": 5.415110397338867,
      "epoch": 0.8102238556632142,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004940649477241401,
      "loss": 5.2833,
      "mean_token_accuracy": 0.18041903972625734,
      "num_tokens": 17860871.0,
      "step": 9700
    },
    {
      "entropy": 5.341606140136719,
      "epoch": 0.8106414968259271,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004940581532775237,
      "loss": 5.1961,
      "mean_token_accuracy": 0.18526162207126617,
      "num_tokens": 17869687.0,
      "step": 9705
    },
    {
      "entropy": 5.299842596054077,
      "epoch": 0.8110591379886402,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004940513549960306,
      "loss": 5.1409,
      "mean_token_accuracy": 0.1847122862935066,
      "num_tokens": 17878815.0,
      "step": 9710
    },
    {
      "entropy": 5.254646492004395,
      "epoch": 0.8114767791513532,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004940445528797798,
      "loss": 5.0882,
      "mean_token_accuracy": 0.190871062874794,
      "num_tokens": 17888302.0,
      "step": 9715
    },
    {
      "entropy": 5.218020725250244,
      "epoch": 0.8118944203140661,
      "grad_norm": 1.625,
      "learning_rate": 0.0004940377469288906,
      "loss": 5.0379,
      "mean_token_accuracy": 0.19915541261434555,
      "num_tokens": 17896807.0,
      "step": 9720
    },
    {
      "entropy": 5.281565999984741,
      "epoch": 0.8123120614767791,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004940309371434819,
      "loss": 5.1181,
      "mean_token_accuracy": 0.19083558171987533,
      "num_tokens": 17905988.0,
      "step": 9725
    },
    {
      "entropy": 5.387138223648071,
      "epoch": 0.8127297026394922,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004940241235236731,
      "loss": 5.2532,
      "mean_token_accuracy": 0.18090667426586152,
      "num_tokens": 17915917.0,
      "step": 9730
    },
    {
      "entropy": 5.409555864334107,
      "epoch": 0.8131473438022051,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0004940173060695833,
      "loss": 5.24,
      "mean_token_accuracy": 0.1853630855679512,
      "num_tokens": 17926158.0,
      "step": 9735
    },
    {
      "entropy": 5.284359407424927,
      "epoch": 0.8135649849649181,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004940104847813319,
      "loss": 5.1996,
      "mean_token_accuracy": 0.18690650314092636,
      "num_tokens": 17935488.0,
      "step": 9740
    },
    {
      "entropy": 5.35828652381897,
      "epoch": 0.8139826261276312,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004940036596590384,
      "loss": 5.1131,
      "mean_token_accuracy": 0.18699391931295395,
      "num_tokens": 17944850.0,
      "step": 9745
    },
    {
      "entropy": 5.310150384902954,
      "epoch": 0.8144002672903441,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004939968307028221,
      "loss": 5.1529,
      "mean_token_accuracy": 0.19170355647802353,
      "num_tokens": 17954796.0,
      "step": 9750
    },
    {
      "entropy": 5.236877679824829,
      "epoch": 0.8148179084530571,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004939899979128028,
      "loss": 5.1258,
      "mean_token_accuracy": 0.19769725501537322,
      "num_tokens": 17963875.0,
      "step": 9755
    },
    {
      "entropy": 5.264697742462158,
      "epoch": 0.8152355496157702,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004939831612890999,
      "loss": 5.1758,
      "mean_token_accuracy": 0.18781137466430664,
      "num_tokens": 17972789.0,
      "step": 9760
    },
    {
      "entropy": 5.26484956741333,
      "epoch": 0.8156531907784831,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004939763208318332,
      "loss": 5.183,
      "mean_token_accuracy": 0.1855485737323761,
      "num_tokens": 17982355.0,
      "step": 9765
    },
    {
      "entropy": 5.362791919708252,
      "epoch": 0.8160708319411961,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004939694765411225,
      "loss": 5.2074,
      "mean_token_accuracy": 0.1848495841026306,
      "num_tokens": 17992046.0,
      "step": 9770
    },
    {
      "entropy": 5.287937593460083,
      "epoch": 0.8164884731039092,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004939626284170874,
      "loss": 5.134,
      "mean_token_accuracy": 0.19881495088338852,
      "num_tokens": 18001440.0,
      "step": 9775
    },
    {
      "entropy": 5.4196573257446286,
      "epoch": 0.8169061142666221,
      "grad_norm": 1.7421875,
      "learning_rate": 0.000493955776459848,
      "loss": 5.2775,
      "mean_token_accuracy": 0.17994987219572067,
      "num_tokens": 18010817.0,
      "step": 9780
    },
    {
      "entropy": 5.3791358947753904,
      "epoch": 0.8173237554293351,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004939489206695241,
      "loss": 5.1783,
      "mean_token_accuracy": 0.18664075583219528,
      "num_tokens": 18019547.0,
      "step": 9785
    },
    {
      "entropy": 5.309901094436645,
      "epoch": 0.8177413965920481,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004939420610462359,
      "loss": 5.1801,
      "mean_token_accuracy": 0.1869453087449074,
      "num_tokens": 18029191.0,
      "step": 9790
    },
    {
      "entropy": 5.341063547134399,
      "epoch": 0.8181590377547611,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004939351975901034,
      "loss": 5.2286,
      "mean_token_accuracy": 0.1871086061000824,
      "num_tokens": 18038831.0,
      "step": 9795
    },
    {
      "entropy": 5.357484006881714,
      "epoch": 0.8185766789174741,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004939283303012467,
      "loss": 5.1504,
      "mean_token_accuracy": 0.19173508286476135,
      "num_tokens": 18049243.0,
      "step": 9800
    },
    {
      "entropy": 5.23955454826355,
      "epoch": 0.8189943200801871,
      "grad_norm": 1.703125,
      "learning_rate": 0.000493921459179786,
      "loss": 5.1032,
      "mean_token_accuracy": 0.1906172439455986,
      "num_tokens": 18057960.0,
      "step": 9805
    },
    {
      "entropy": 5.30379090309143,
      "epoch": 0.8194119612429,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004939145842258418,
      "loss": 5.2163,
      "mean_token_accuracy": 0.1840106263756752,
      "num_tokens": 18066677.0,
      "step": 9810
    },
    {
      "entropy": 5.389127445220947,
      "epoch": 0.8198296024056131,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004939077054395341,
      "loss": 5.1906,
      "mean_token_accuracy": 0.1844009429216385,
      "num_tokens": 18075830.0,
      "step": 9815
    },
    {
      "entropy": 5.340755653381348,
      "epoch": 0.8202472435683261,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004939008228209836,
      "loss": 5.0974,
      "mean_token_accuracy": 0.18784936964511872,
      "num_tokens": 18085335.0,
      "step": 9820
    },
    {
      "entropy": 5.30837197303772,
      "epoch": 0.820664884731039,
      "grad_norm": 2.453125,
      "learning_rate": 0.0004938939363703106,
      "loss": 5.1508,
      "mean_token_accuracy": 0.18218090534210205,
      "num_tokens": 18094531.0,
      "step": 9825
    },
    {
      "entropy": 5.2695659637451175,
      "epoch": 0.8210825258937521,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004938870460876359,
      "loss": 5.1278,
      "mean_token_accuracy": 0.19451568573713302,
      "num_tokens": 18103593.0,
      "step": 9830
    },
    {
      "entropy": 5.316597127914429,
      "epoch": 0.8215001670564651,
      "grad_norm": 2.3125,
      "learning_rate": 0.00049388015197308,
      "loss": 5.1864,
      "mean_token_accuracy": 0.19050071984529496,
      "num_tokens": 18112826.0,
      "step": 9835
    },
    {
      "entropy": 5.365003681182861,
      "epoch": 0.821917808219178,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004938732540267635,
      "loss": 5.2466,
      "mean_token_accuracy": 0.18888215273618697,
      "num_tokens": 18121687.0,
      "step": 9840
    },
    {
      "entropy": 5.345579147338867,
      "epoch": 0.8223354493818911,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004938663522488072,
      "loss": 5.0953,
      "mean_token_accuracy": 0.19906752705574035,
      "num_tokens": 18130071.0,
      "step": 9845
    },
    {
      "entropy": 5.380211639404297,
      "epoch": 0.8227530905446041,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004938594466393321,
      "loss": 5.3378,
      "mean_token_accuracy": 0.1846168205142021,
      "num_tokens": 18139077.0,
      "step": 9850
    },
    {
      "entropy": 5.34048023223877,
      "epoch": 0.823170731707317,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004938525371984588,
      "loss": 5.1978,
      "mean_token_accuracy": 0.19228179156780242,
      "num_tokens": 18147173.0,
      "step": 9855
    },
    {
      "entropy": 5.235449695587159,
      "epoch": 0.8235883728700301,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004938456239263085,
      "loss": 5.1805,
      "mean_token_accuracy": 0.19033245891332626,
      "num_tokens": 18157061.0,
      "step": 9860
    },
    {
      "entropy": 5.2925516128540036,
      "epoch": 0.8240060140327431,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004938387068230021,
      "loss": 5.1136,
      "mean_token_accuracy": 0.1912828728556633,
      "num_tokens": 18166115.0,
      "step": 9865
    },
    {
      "entropy": 5.3132106304168705,
      "epoch": 0.824423655195456,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004938317858886606,
      "loss": 5.1381,
      "mean_token_accuracy": 0.19255006164312363,
      "num_tokens": 18174978.0,
      "step": 9870
    },
    {
      "entropy": 5.307817506790161,
      "epoch": 0.824841296358169,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004938248611234055,
      "loss": 5.2144,
      "mean_token_accuracy": 0.1895279109477997,
      "num_tokens": 18184948.0,
      "step": 9875
    },
    {
      "entropy": 5.3290262699127195,
      "epoch": 0.8252589375208821,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004938179325273577,
      "loss": 5.1494,
      "mean_token_accuracy": 0.19005334973335267,
      "num_tokens": 18193443.0,
      "step": 9880
    },
    {
      "entropy": 5.354461336135865,
      "epoch": 0.825676578683595,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004938110001006385,
      "loss": 5.1634,
      "mean_token_accuracy": 0.1891029730439186,
      "num_tokens": 18202512.0,
      "step": 9885
    },
    {
      "entropy": 5.361838436126709,
      "epoch": 0.826094219846308,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004938040638433694,
      "loss": 5.1741,
      "mean_token_accuracy": 0.19064506739377976,
      "num_tokens": 18212328.0,
      "step": 9890
    },
    {
      "entropy": 5.320434808731079,
      "epoch": 0.8265118610090211,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004937971237556719,
      "loss": 5.1231,
      "mean_token_accuracy": 0.1904003605246544,
      "num_tokens": 18221155.0,
      "step": 9895
    },
    {
      "entropy": 5.325121593475342,
      "epoch": 0.826929502171734,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004937901798376672,
      "loss": 5.1895,
      "mean_token_accuracy": 0.1933050587773323,
      "num_tokens": 18230604.0,
      "step": 9900
    },
    {
      "entropy": 5.316837215423584,
      "epoch": 0.827347143334447,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004937832320894772,
      "loss": 5.0963,
      "mean_token_accuracy": 0.191922527551651,
      "num_tokens": 18239126.0,
      "step": 9905
    },
    {
      "entropy": 5.282761669158935,
      "epoch": 0.8277647844971601,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004937762805112233,
      "loss": 5.1499,
      "mean_token_accuracy": 0.18285557478666306,
      "num_tokens": 18248634.0,
      "step": 9910
    },
    {
      "entropy": 5.326156663894653,
      "epoch": 0.828182425659873,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004937693251030273,
      "loss": 5.2975,
      "mean_token_accuracy": 0.17551200985908508,
      "num_tokens": 18258500.0,
      "step": 9915
    },
    {
      "entropy": 5.399630689620972,
      "epoch": 0.828600066822586,
      "grad_norm": 2.21875,
      "learning_rate": 0.000493762365865011,
      "loss": 5.1146,
      "mean_token_accuracy": 0.18955700248479843,
      "num_tokens": 18268374.0,
      "step": 9920
    },
    {
      "entropy": 5.373224115371704,
      "epoch": 0.8290177079852991,
      "grad_norm": 3.203125,
      "learning_rate": 0.0004937554027972961,
      "loss": 5.2311,
      "mean_token_accuracy": 0.18239429891109465,
      "num_tokens": 18277667.0,
      "step": 9925
    },
    {
      "entropy": 5.306133794784546,
      "epoch": 0.829435349148012,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004937484359000046,
      "loss": 5.247,
      "mean_token_accuracy": 0.18764786720275878,
      "num_tokens": 18287516.0,
      "step": 9930
    },
    {
      "entropy": 5.292248153686524,
      "epoch": 0.829852990310725,
      "grad_norm": 3.015625,
      "learning_rate": 0.0004937414651732584,
      "loss": 5.0986,
      "mean_token_accuracy": 0.19411901384592056,
      "num_tokens": 18296534.0,
      "step": 9935
    },
    {
      "entropy": 5.35420503616333,
      "epoch": 0.830270631473438,
      "grad_norm": 2.609375,
      "learning_rate": 0.0004937344906171795,
      "loss": 5.23,
      "mean_token_accuracy": 0.18965759575366975,
      "num_tokens": 18305743.0,
      "step": 9940
    },
    {
      "entropy": 5.36882586479187,
      "epoch": 0.830688272636151,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004937275122318902,
      "loss": 5.2444,
      "mean_token_accuracy": 0.18101778775453567,
      "num_tokens": 18315524.0,
      "step": 9945
    },
    {
      "entropy": 5.297546863555908,
      "epoch": 0.831105913798864,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004937205300175123,
      "loss": 5.1095,
      "mean_token_accuracy": 0.19682240039110183,
      "num_tokens": 18324280.0,
      "step": 9950
    },
    {
      "entropy": 5.363136672973633,
      "epoch": 0.831523554961577,
      "grad_norm": 2.125,
      "learning_rate": 0.0004937135439741682,
      "loss": 5.2043,
      "mean_token_accuracy": 0.18807094991207124,
      "num_tokens": 18332661.0,
      "step": 9955
    },
    {
      "entropy": 5.384078884124756,
      "epoch": 0.83194119612429,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004937065541019806,
      "loss": 5.3094,
      "mean_token_accuracy": 0.1700368896126747,
      "num_tokens": 18342986.0,
      "step": 9960
    },
    {
      "entropy": 5.396446323394775,
      "epoch": 0.832358837287003,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004936995604010711,
      "loss": 5.2473,
      "mean_token_accuracy": 0.18414016664028168,
      "num_tokens": 18351873.0,
      "step": 9965
    },
    {
      "entropy": 5.399761438369751,
      "epoch": 0.832776478449716,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004936925628715628,
      "loss": 5.3191,
      "mean_token_accuracy": 0.18092354089021684,
      "num_tokens": 18362174.0,
      "step": 9970
    },
    {
      "entropy": 5.335303449630738,
      "epoch": 0.833194119612429,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004936855615135779,
      "loss": 5.1173,
      "mean_token_accuracy": 0.19510533511638642,
      "num_tokens": 18371272.0,
      "step": 9975
    },
    {
      "entropy": 5.361533784866333,
      "epoch": 0.833611760775142,
      "grad_norm": 2.640625,
      "learning_rate": 0.000493678556327239,
      "loss": 5.236,
      "mean_token_accuracy": 0.18942816257476808,
      "num_tokens": 18380526.0,
      "step": 9980
    },
    {
      "entropy": 5.2963732242584225,
      "epoch": 0.834029401937855,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004936715473126688,
      "loss": 5.1356,
      "mean_token_accuracy": 0.1921591728925705,
      "num_tokens": 18390027.0,
      "step": 9985
    },
    {
      "entropy": 5.242372179031372,
      "epoch": 0.8344470431005679,
      "grad_norm": 2.03125,
      "learning_rate": 0.00049366453446999,
      "loss": 5.1533,
      "mean_token_accuracy": 0.19045700430870055,
      "num_tokens": 18399012.0,
      "step": 9990
    },
    {
      "entropy": 5.216147613525391,
      "epoch": 0.834864684263281,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004936575177993251,
      "loss": 5.0701,
      "mean_token_accuracy": 0.19057831168174744,
      "num_tokens": 18408891.0,
      "step": 9995
    },
    {
      "entropy": 5.264990663528442,
      "epoch": 0.835282325425994,
      "grad_norm": 1.625,
      "learning_rate": 0.0004936504973007973,
      "loss": 5.1106,
      "mean_token_accuracy": 0.19354889839887618,
      "num_tokens": 18417909.0,
      "step": 10000
    },
    {
      "entropy": 5.27159948348999,
      "epoch": 0.8356999665887069,
      "grad_norm": 1.375,
      "learning_rate": 0.0004936434729745294,
      "loss": 5.1755,
      "mean_token_accuracy": 0.19167029708623887,
      "num_tokens": 18427250.0,
      "step": 10005
    },
    {
      "entropy": 5.294535732269287,
      "epoch": 0.83611760775142,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004936364448206442,
      "loss": 5.1591,
      "mean_token_accuracy": 0.19149200171232222,
      "num_tokens": 18435851.0,
      "step": 10010
    },
    {
      "entropy": 5.231304550170899,
      "epoch": 0.836535248914133,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004936294128392649,
      "loss": 5.1461,
      "mean_token_accuracy": 0.1829630732536316,
      "num_tokens": 18444535.0,
      "step": 10015
    },
    {
      "entropy": 5.337705469131469,
      "epoch": 0.8369528900768459,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004936223770305146,
      "loss": 5.2628,
      "mean_token_accuracy": 0.1871430516242981,
      "num_tokens": 18454453.0,
      "step": 10020
    },
    {
      "entropy": 5.411758995056152,
      "epoch": 0.837370531239559,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004936153373945165,
      "loss": 5.2999,
      "mean_token_accuracy": 0.18302317559719086,
      "num_tokens": 18465033.0,
      "step": 10025
    },
    {
      "entropy": 5.334819173812866,
      "epoch": 0.837788172402272,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004936082939313936,
      "loss": 5.1752,
      "mean_token_accuracy": 0.1861498937010765,
      "num_tokens": 18473786.0,
      "step": 10030
    },
    {
      "entropy": 5.284872436523438,
      "epoch": 0.8382058135649849,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004936012466412695,
      "loss": 5.1391,
      "mean_token_accuracy": 0.1824897602200508,
      "num_tokens": 18483282.0,
      "step": 10035
    },
    {
      "entropy": 5.4622749328613285,
      "epoch": 0.838623454727698,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004935941955242674,
      "loss": 5.3098,
      "mean_token_accuracy": 0.1818353682756424,
      "num_tokens": 18492612.0,
      "step": 10040
    },
    {
      "entropy": 5.348860359191894,
      "epoch": 0.839041095890411,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004935871405805109,
      "loss": 5.1289,
      "mean_token_accuracy": 0.19514228105545045,
      "num_tokens": 18501120.0,
      "step": 10045
    },
    {
      "entropy": 5.301193523406982,
      "epoch": 0.8394587370531239,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004935800818101232,
      "loss": 5.2369,
      "mean_token_accuracy": 0.18511967957019806,
      "num_tokens": 18510613.0,
      "step": 10050
    },
    {
      "entropy": 5.27965784072876,
      "epoch": 0.8398763782158369,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004935730192132282,
      "loss": 5.1105,
      "mean_token_accuracy": 0.19652514904737473,
      "num_tokens": 18518854.0,
      "step": 10055
    },
    {
      "entropy": 5.348846578598023,
      "epoch": 0.84029401937855,
      "grad_norm": 4.34375,
      "learning_rate": 0.0004935659527899493,
      "loss": 5.2051,
      "mean_token_accuracy": 0.1909124955534935,
      "num_tokens": 18527592.0,
      "step": 10060
    },
    {
      "entropy": 5.223753356933594,
      "epoch": 0.8407116605412629,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004935588825404104,
      "loss": 5.0669,
      "mean_token_accuracy": 0.2033865198493004,
      "num_tokens": 18536396.0,
      "step": 10065
    },
    {
      "entropy": 5.2362908840179445,
      "epoch": 0.8411293017039759,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004935518084647351,
      "loss": 5.0412,
      "mean_token_accuracy": 0.20289033353328706,
      "num_tokens": 18545201.0,
      "step": 10070
    },
    {
      "entropy": 5.311696910858155,
      "epoch": 0.841546942866689,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004935447305630474,
      "loss": 5.2413,
      "mean_token_accuracy": 0.18510085046291352,
      "num_tokens": 18554944.0,
      "step": 10075
    },
    {
      "entropy": 5.298465013504028,
      "epoch": 0.8419645840294019,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004935376488354711,
      "loss": 5.1372,
      "mean_token_accuracy": 0.18965242207050323,
      "num_tokens": 18564642.0,
      "step": 10080
    },
    {
      "entropy": 5.321731233596802,
      "epoch": 0.8423822251921149,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004935305632821303,
      "loss": 5.1482,
      "mean_token_accuracy": 0.19197022020816804,
      "num_tokens": 18573897.0,
      "step": 10085
    },
    {
      "entropy": 5.34338345527649,
      "epoch": 0.842799866354828,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004935234739031488,
      "loss": 5.188,
      "mean_token_accuracy": 0.18661972284317016,
      "num_tokens": 18583174.0,
      "step": 10090
    },
    {
      "entropy": 5.326038074493408,
      "epoch": 0.8432175075175409,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004935163806986508,
      "loss": 5.086,
      "mean_token_accuracy": 0.19799598753452302,
      "num_tokens": 18592540.0,
      "step": 10095
    },
    {
      "entropy": 5.268729209899902,
      "epoch": 0.8436351486802539,
      "grad_norm": 2.1875,
      "learning_rate": 0.0004935092836687607,
      "loss": 5.0957,
      "mean_token_accuracy": 0.19314343929290773,
      "num_tokens": 18602743.0,
      "step": 10100
    },
    {
      "entropy": 5.22485122680664,
      "epoch": 0.8440527898429669,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004935021828136025,
      "loss": 5.0851,
      "mean_token_accuracy": 0.19164522588253022,
      "num_tokens": 18611544.0,
      "step": 10105
    },
    {
      "entropy": 5.3677003383636475,
      "epoch": 0.8444704310056799,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004934950781333005,
      "loss": 5.1831,
      "mean_token_accuracy": 0.18606801629066466,
      "num_tokens": 18620841.0,
      "step": 10110
    },
    {
      "entropy": 5.367251348495484,
      "epoch": 0.8448880721683929,
      "grad_norm": 1.5,
      "learning_rate": 0.0004934879696279793,
      "loss": 5.2326,
      "mean_token_accuracy": 0.18908178210258483,
      "num_tokens": 18630890.0,
      "step": 10115
    },
    {
      "entropy": 5.355392265319824,
      "epoch": 0.8453057133311059,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004934808572977632,
      "loss": 5.2247,
      "mean_token_accuracy": 0.18310176432132722,
      "num_tokens": 18639768.0,
      "step": 10120
    },
    {
      "entropy": 5.249336194992066,
      "epoch": 0.8457233544938189,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004934737411427765,
      "loss": 5.0478,
      "mean_token_accuracy": 0.19338246583938598,
      "num_tokens": 18649578.0,
      "step": 10125
    },
    {
      "entropy": 5.265984725952149,
      "epoch": 0.8461409956565319,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004934666211631442,
      "loss": 5.1359,
      "mean_token_accuracy": 0.1938577711582184,
      "num_tokens": 18658470.0,
      "step": 10130
    },
    {
      "entropy": 5.264990043640137,
      "epoch": 0.8465586368192449,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004934594973589906,
      "loss": 5.1135,
      "mean_token_accuracy": 0.1872977375984192,
      "num_tokens": 18668037.0,
      "step": 10135
    },
    {
      "entropy": 5.295865106582641,
      "epoch": 0.8469762779819578,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004934523697304406,
      "loss": 5.2325,
      "mean_token_accuracy": 0.18623916655778885,
      "num_tokens": 18677841.0,
      "step": 10140
    },
    {
      "entropy": 5.404227352142334,
      "epoch": 0.8473939191446709,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004934452382776189,
      "loss": 5.1883,
      "mean_token_accuracy": 0.18454785794019699,
      "num_tokens": 18686244.0,
      "step": 10145
    },
    {
      "entropy": 5.313640069961548,
      "epoch": 0.8478115603073839,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004934381030006504,
      "loss": 5.2222,
      "mean_token_accuracy": 0.18657472729682922,
      "num_tokens": 18695129.0,
      "step": 10150
    },
    {
      "entropy": 5.181084680557251,
      "epoch": 0.8482292014700968,
      "grad_norm": 1.9375,
      "learning_rate": 0.00049343096389966,
      "loss": 5.028,
      "mean_token_accuracy": 0.20085298717021943,
      "num_tokens": 18703566.0,
      "step": 10155
    },
    {
      "entropy": 5.384372615814209,
      "epoch": 0.8486468426328099,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004934238209747724,
      "loss": 5.2419,
      "mean_token_accuracy": 0.17773670852184295,
      "num_tokens": 18713616.0,
      "step": 10160
    },
    {
      "entropy": 5.421914100646973,
      "epoch": 0.8490644837955229,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004934166742261132,
      "loss": 5.1721,
      "mean_token_accuracy": 0.1845224305987358,
      "num_tokens": 18722851.0,
      "step": 10165
    },
    {
      "entropy": 5.356164598464966,
      "epoch": 0.8494821249582358,
      "grad_norm": 1.6953125,
      "learning_rate": 0.000493409523653807,
      "loss": 5.1633,
      "mean_token_accuracy": 0.1861880376935005,
      "num_tokens": 18731961.0,
      "step": 10170
    },
    {
      "entropy": 5.365734815597534,
      "epoch": 0.8498997661209489,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004934023692579793,
      "loss": 5.3294,
      "mean_token_accuracy": 0.18228207230567933,
      "num_tokens": 18741267.0,
      "step": 10175
    },
    {
      "entropy": 5.303253555297852,
      "epoch": 0.8503174072836619,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004933952110387553,
      "loss": 5.1679,
      "mean_token_accuracy": 0.18911945819854736,
      "num_tokens": 18750266.0,
      "step": 10180
    },
    {
      "entropy": 5.2705018520355225,
      "epoch": 0.8507350484463748,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004933880489962602,
      "loss": 5.1493,
      "mean_token_accuracy": 0.18801131546497346,
      "num_tokens": 18759910.0,
      "step": 10185
    },
    {
      "entropy": 5.439163112640381,
      "epoch": 0.8511526896090879,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004933808831306194,
      "loss": 5.302,
      "mean_token_accuracy": 0.17844336181879045,
      "num_tokens": 18769510.0,
      "step": 10190
    },
    {
      "entropy": 5.300149202346802,
      "epoch": 0.8515703307718009,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004933737134419583,
      "loss": 5.1083,
      "mean_token_accuracy": 0.1858849734067917,
      "num_tokens": 18779893.0,
      "step": 10195
    },
    {
      "entropy": 5.327555704116821,
      "epoch": 0.8519879719345138,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004933665399304025,
      "loss": 5.1711,
      "mean_token_accuracy": 0.19460691660642623,
      "num_tokens": 18788802.0,
      "step": 10200
    },
    {
      "entropy": 5.280136299133301,
      "epoch": 0.8524056130972268,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004933593625960777,
      "loss": 5.0825,
      "mean_token_accuracy": 0.18462813794612884,
      "num_tokens": 18797154.0,
      "step": 10205
    },
    {
      "entropy": 5.303889799118042,
      "epoch": 0.8528232542599399,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004933521814391094,
      "loss": 5.1826,
      "mean_token_accuracy": 0.18324453979730607,
      "num_tokens": 18806151.0,
      "step": 10210
    },
    {
      "entropy": 5.385970687866211,
      "epoch": 0.8532408954226529,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004933449964596232,
      "loss": 5.0983,
      "mean_token_accuracy": 0.19556251168251038,
      "num_tokens": 18815162.0,
      "step": 10215
    },
    {
      "entropy": 5.244508266448975,
      "epoch": 0.8536585365853658,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004933378076577451,
      "loss": 5.077,
      "mean_token_accuracy": 0.18822017908096314,
      "num_tokens": 18823676.0,
      "step": 10220
    },
    {
      "entropy": 5.357374906539917,
      "epoch": 0.8540761777480789,
      "grad_norm": 3.1875,
      "learning_rate": 0.0004933306150336009,
      "loss": 5.3943,
      "mean_token_accuracy": 0.17783361673355103,
      "num_tokens": 18832991.0,
      "step": 10225
    },
    {
      "entropy": 5.480168008804322,
      "epoch": 0.8544938189107919,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004933234185873164,
      "loss": 5.2504,
      "mean_token_accuracy": 0.18567426204681398,
      "num_tokens": 18842614.0,
      "step": 10230
    },
    {
      "entropy": 5.451560068130493,
      "epoch": 0.8549114600735048,
      "grad_norm": 2.171875,
      "learning_rate": 0.0004933162183190177,
      "loss": 5.3161,
      "mean_token_accuracy": 0.17856184989213944,
      "num_tokens": 18852747.0,
      "step": 10235
    },
    {
      "entropy": 5.244213247299195,
      "epoch": 0.8553291012362179,
      "grad_norm": 1.9375,
      "learning_rate": 0.0004933090142288308,
      "loss": 5.1642,
      "mean_token_accuracy": 0.19237858057022095,
      "num_tokens": 18862005.0,
      "step": 10240
    },
    {
      "entropy": 5.309420824050903,
      "epoch": 0.8557467423989309,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004933018063168819,
      "loss": 5.1093,
      "mean_token_accuracy": 0.1995083510875702,
      "num_tokens": 18870172.0,
      "step": 10245
    },
    {
      "entropy": 5.3396923542022705,
      "epoch": 0.8561643835616438,
      "grad_norm": 1.421875,
      "learning_rate": 0.000493294594583297,
      "loss": 5.2014,
      "mean_token_accuracy": 0.18443857431411742,
      "num_tokens": 18879428.0,
      "step": 10250
    },
    {
      "entropy": 5.296653985977173,
      "epoch": 0.8565820247243568,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004932873790282026,
      "loss": 5.1454,
      "mean_token_accuracy": 0.2006555438041687,
      "num_tokens": 18888694.0,
      "step": 10255
    },
    {
      "entropy": 5.408766222000122,
      "epoch": 0.8569996658870699,
      "grad_norm": 1.625,
      "learning_rate": 0.0004932801596517248,
      "loss": 5.203,
      "mean_token_accuracy": 0.18389252275228501,
      "num_tokens": 18898317.0,
      "step": 10260
    },
    {
      "entropy": 5.274123239517212,
      "epoch": 0.8574173070497828,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004932729364539901,
      "loss": 5.1946,
      "mean_token_accuracy": 0.18400730639696122,
      "num_tokens": 18906749.0,
      "step": 10265
    },
    {
      "entropy": 5.280352401733398,
      "epoch": 0.8578349482124958,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004932657094351248,
      "loss": 5.0676,
      "mean_token_accuracy": 0.19361914098262786,
      "num_tokens": 18915404.0,
      "step": 10270
    },
    {
      "entropy": 5.2740716457366945,
      "epoch": 0.8582525893752089,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004932584785952555,
      "loss": 5.1794,
      "mean_token_accuracy": 0.18665087223052979,
      "num_tokens": 18925512.0,
      "step": 10275
    },
    {
      "entropy": 5.284380483627319,
      "epoch": 0.8586702305379218,
      "grad_norm": 1.515625,
      "learning_rate": 0.000493251243934509,
      "loss": 5.1032,
      "mean_token_accuracy": 0.19272940307855607,
      "num_tokens": 18934278.0,
      "step": 10280
    },
    {
      "entropy": 5.322548866271973,
      "epoch": 0.8590878717006348,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004932440054530116,
      "loss": 5.1802,
      "mean_token_accuracy": 0.18642934262752534,
      "num_tokens": 18943126.0,
      "step": 10285
    },
    {
      "entropy": 5.337692451477051,
      "epoch": 0.8595055128633479,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004932367631508903,
      "loss": 5.1761,
      "mean_token_accuracy": 0.19143159240484237,
      "num_tokens": 18951716.0,
      "step": 10290
    },
    {
      "entropy": 5.29817156791687,
      "epoch": 0.8599231540260608,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004932295170282717,
      "loss": 5.2517,
      "mean_token_accuracy": 0.18123940676450728,
      "num_tokens": 18960692.0,
      "step": 10295
    },
    {
      "entropy": 5.280509328842163,
      "epoch": 0.8603407951887738,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004932222670852828,
      "loss": 5.1992,
      "mean_token_accuracy": 0.1924230858683586,
      "num_tokens": 18969813.0,
      "step": 10300
    },
    {
      "entropy": 5.352678060531616,
      "epoch": 0.8607584363514869,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004932150133220504,
      "loss": 5.0526,
      "mean_token_accuracy": 0.193808476626873,
      "num_tokens": 18978536.0,
      "step": 10305
    },
    {
      "entropy": 5.290709447860718,
      "epoch": 0.8611760775141998,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004932077557387015,
      "loss": 5.1542,
      "mean_token_accuracy": 0.18996871411800384,
      "num_tokens": 18987637.0,
      "step": 10310
    },
    {
      "entropy": 5.175679302215576,
      "epoch": 0.8615937186769128,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004932004943353633,
      "loss": 4.9682,
      "mean_token_accuracy": 0.20098690539598466,
      "num_tokens": 18997010.0,
      "step": 10315
    },
    {
      "entropy": 5.327095413208008,
      "epoch": 0.8620113598396258,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004931932291121627,
      "loss": 5.1458,
      "mean_token_accuracy": 0.1873937502503395,
      "num_tokens": 19005685.0,
      "step": 10320
    },
    {
      "entropy": 5.307133674621582,
      "epoch": 0.8624290010023388,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004931859600692269,
      "loss": 5.1659,
      "mean_token_accuracy": 0.1900626540184021,
      "num_tokens": 19014836.0,
      "step": 10325
    },
    {
      "entropy": 5.344813346862793,
      "epoch": 0.8628466421650518,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004931786872066833,
      "loss": 5.1915,
      "mean_token_accuracy": 0.1814238429069519,
      "num_tokens": 19024442.0,
      "step": 10330
    },
    {
      "entropy": 5.33070740699768,
      "epoch": 0.8632642833277648,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004931714105246592,
      "loss": 5.1468,
      "mean_token_accuracy": 0.19128657281398773,
      "num_tokens": 19033730.0,
      "step": 10335
    },
    {
      "entropy": 5.310273504257202,
      "epoch": 0.8636819244904778,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004931641300232819,
      "loss": 5.1005,
      "mean_token_accuracy": 0.2026464730501175,
      "num_tokens": 19042765.0,
      "step": 10340
    },
    {
      "entropy": 5.308903694152832,
      "epoch": 0.8640995656531908,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004931568457026789,
      "loss": 5.1066,
      "mean_token_accuracy": 0.18763562738895417,
      "num_tokens": 19051408.0,
      "step": 10345
    },
    {
      "entropy": 5.32097806930542,
      "epoch": 0.8645172068159038,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004931495575629778,
      "loss": 5.2228,
      "mean_token_accuracy": 0.17876803129911423,
      "num_tokens": 19061145.0,
      "step": 10350
    },
    {
      "entropy": 5.31494722366333,
      "epoch": 0.8649348479786167,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000493142265604306,
      "loss": 5.2006,
      "mean_token_accuracy": 0.18291258364915847,
      "num_tokens": 19070157.0,
      "step": 10355
    },
    {
      "entropy": 5.256790208816528,
      "epoch": 0.8653524891413298,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004931349698267914,
      "loss": 5.0146,
      "mean_token_accuracy": 0.2009235605597496,
      "num_tokens": 19078532.0,
      "step": 10360
    },
    {
      "entropy": 5.263900661468506,
      "epoch": 0.8657701303040428,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004931276702305615,
      "loss": 5.0908,
      "mean_token_accuracy": 0.1976345658302307,
      "num_tokens": 19087283.0,
      "step": 10365
    },
    {
      "entropy": 5.322926759719849,
      "epoch": 0.8661877714667557,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004931203668157441,
      "loss": 5.2604,
      "mean_token_accuracy": 0.19001127183437347,
      "num_tokens": 19097568.0,
      "step": 10370
    },
    {
      "entropy": 5.374041128158569,
      "epoch": 0.8666054126294688,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004931130595824672,
      "loss": 5.2286,
      "mean_token_accuracy": 0.18117907494306565,
      "num_tokens": 19107218.0,
      "step": 10375
    },
    {
      "entropy": 5.334418821334839,
      "epoch": 0.8670230537921818,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004931057485308587,
      "loss": 5.1696,
      "mean_token_accuracy": 0.1926760733127594,
      "num_tokens": 19116128.0,
      "step": 10380
    },
    {
      "entropy": 5.229773235321045,
      "epoch": 0.8674406949548947,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004930984336610465,
      "loss": 5.1577,
      "mean_token_accuracy": 0.19678619652986526,
      "num_tokens": 19124885.0,
      "step": 10385
    },
    {
      "entropy": 5.4001775741577145,
      "epoch": 0.8678583361176078,
      "grad_norm": 2.171875,
      "learning_rate": 0.0004930911149731587,
      "loss": 5.2216,
      "mean_token_accuracy": 0.19020774364471435,
      "num_tokens": 19134125.0,
      "step": 10390
    },
    {
      "entropy": 5.273541212081909,
      "epoch": 0.8682759772803208,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004930837924673233,
      "loss": 5.0966,
      "mean_token_accuracy": 0.1925998941063881,
      "num_tokens": 19141925.0,
      "step": 10395
    },
    {
      "entropy": 5.30088586807251,
      "epoch": 0.8686936184430337,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004930764661436688,
      "loss": 5.1868,
      "mean_token_accuracy": 0.19022163450717927,
      "num_tokens": 19151440.0,
      "step": 10400
    },
    {
      "entropy": 5.368042755126953,
      "epoch": 0.8691112596057468,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004930691360023231,
      "loss": 5.2532,
      "mean_token_accuracy": 0.19069868624210357,
      "num_tokens": 19161447.0,
      "step": 10405
    },
    {
      "entropy": 5.383994102478027,
      "epoch": 0.8695289007684598,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004930618020434147,
      "loss": 5.1782,
      "mean_token_accuracy": 0.18452608585357666,
      "num_tokens": 19170804.0,
      "step": 10410
    },
    {
      "entropy": 5.431319618225098,
      "epoch": 0.8699465419311727,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004930544642670721,
      "loss": 5.2801,
      "mean_token_accuracy": 0.18168249875307083,
      "num_tokens": 19179898.0,
      "step": 10415
    },
    {
      "entropy": 5.241486978530884,
      "epoch": 0.8703641830938857,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004930471226734235,
      "loss": 5.1091,
      "mean_token_accuracy": 0.19783590137958526,
      "num_tokens": 19188919.0,
      "step": 10420
    },
    {
      "entropy": 5.33616042137146,
      "epoch": 0.8707818242565988,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004930397772625976,
      "loss": 5.1933,
      "mean_token_accuracy": 0.18918959349393843,
      "num_tokens": 19197745.0,
      "step": 10425
    },
    {
      "entropy": 5.2575071334838865,
      "epoch": 0.8711994654193117,
      "grad_norm": 1.515625,
      "learning_rate": 0.000493032428034723,
      "loss": 5.1872,
      "mean_token_accuracy": 0.18823064863681793,
      "num_tokens": 19206681.0,
      "step": 10430
    },
    {
      "entropy": 5.380467176437378,
      "epoch": 0.8716171065820247,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004930250749899282,
      "loss": 5.1801,
      "mean_token_accuracy": 0.1915568307042122,
      "num_tokens": 19215806.0,
      "step": 10435
    },
    {
      "entropy": 5.374777555465698,
      "epoch": 0.8720347477447378,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004930177181283422,
      "loss": 5.2069,
      "mean_token_accuracy": 0.18793948143720626,
      "num_tokens": 19224681.0,
      "step": 10440
    },
    {
      "entropy": 5.287902545928955,
      "epoch": 0.8724523889074507,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004930103574500934,
      "loss": 5.205,
      "mean_token_accuracy": 0.1879182666540146,
      "num_tokens": 19232684.0,
      "step": 10445
    },
    {
      "entropy": 5.311669063568115,
      "epoch": 0.8728700300701637,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004930029929553109,
      "loss": 5.1095,
      "mean_token_accuracy": 0.18654640167951583,
      "num_tokens": 19242538.0,
      "step": 10450
    },
    {
      "entropy": 5.341206121444702,
      "epoch": 0.8732876712328768,
      "grad_norm": 1.75,
      "learning_rate": 0.0004929956246441237,
      "loss": 5.1558,
      "mean_token_accuracy": 0.1860884830355644,
      "num_tokens": 19251146.0,
      "step": 10455
    },
    {
      "entropy": 5.400697851181031,
      "epoch": 0.8737053123955897,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004929882525166608,
      "loss": 5.1929,
      "mean_token_accuracy": 0.18426415473222732,
      "num_tokens": 19259931.0,
      "step": 10460
    },
    {
      "entropy": 5.306517648696899,
      "epoch": 0.8741229535583027,
      "grad_norm": 1.5,
      "learning_rate": 0.0004929808765730509,
      "loss": 5.2037,
      "mean_token_accuracy": 0.1874943718314171,
      "num_tokens": 19268796.0,
      "step": 10465
    },
    {
      "entropy": 5.295291614532471,
      "epoch": 0.8745405947210158,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004929734968134235,
      "loss": 5.2156,
      "mean_token_accuracy": 0.18509959131479264,
      "num_tokens": 19278396.0,
      "step": 10470
    },
    {
      "entropy": 5.361218738555908,
      "epoch": 0.8749582358837287,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004929661132379076,
      "loss": 5.2311,
      "mean_token_accuracy": 0.1819717988371849,
      "num_tokens": 19287914.0,
      "step": 10475
    },
    {
      "entropy": 5.388542318344117,
      "epoch": 0.8753758770464417,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004929587258466326,
      "loss": 5.201,
      "mean_token_accuracy": 0.1837903916835785,
      "num_tokens": 19297304.0,
      "step": 10480
    },
    {
      "entropy": 5.357698440551758,
      "epoch": 0.8757935182091547,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004929513346397278,
      "loss": 5.1662,
      "mean_token_accuracy": 0.19044652879238128,
      "num_tokens": 19306245.0,
      "step": 10485
    },
    {
      "entropy": 5.345926094055176,
      "epoch": 0.8762111593718677,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004929439396173223,
      "loss": 5.1817,
      "mean_token_accuracy": 0.17997460663318635,
      "num_tokens": 19316160.0,
      "step": 10490
    },
    {
      "entropy": 5.294673013687134,
      "epoch": 0.8766288005345807,
      "grad_norm": 1.734375,
      "learning_rate": 0.000492936540779546,
      "loss": 5.2467,
      "mean_token_accuracy": 0.18372495323419571,
      "num_tokens": 19325411.0,
      "step": 10495
    },
    {
      "entropy": 5.377767562866211,
      "epoch": 0.8770464416972937,
      "grad_norm": 4.875,
      "learning_rate": 0.0004929291381265281,
      "loss": 5.1617,
      "mean_token_accuracy": 0.18904377818107604,
      "num_tokens": 19334297.0,
      "step": 10500
    },
    {
      "entropy": 5.402296447753907,
      "epoch": 0.8774640828600067,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004929217316583985,
      "loss": 5.2646,
      "mean_token_accuracy": 0.18171188086271287,
      "num_tokens": 19344121.0,
      "step": 10505
    },
    {
      "entropy": 5.380285930633545,
      "epoch": 0.8778817240227197,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004929143213752865,
      "loss": 5.2007,
      "mean_token_accuracy": 0.18814597725868226,
      "num_tokens": 19353428.0,
      "step": 10510
    },
    {
      "entropy": 5.40724811553955,
      "epoch": 0.8782993651854327,
      "grad_norm": 1.8203125,
      "learning_rate": 0.000492906907277322,
      "loss": 5.2546,
      "mean_token_accuracy": 0.1868235781788826,
      "num_tokens": 19362661.0,
      "step": 10515
    },
    {
      "entropy": 5.2036622047424315,
      "epoch": 0.8787170063481456,
      "grad_norm": 1.8671875,
      "learning_rate": 0.000492899489364635,
      "loss": 5.0407,
      "mean_token_accuracy": 0.20076248794794083,
      "num_tokens": 19371246.0,
      "step": 10520
    },
    {
      "entropy": 5.288510274887085,
      "epoch": 0.8791346475108587,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004928920676373549,
      "loss": 5.1087,
      "mean_token_accuracy": 0.19001537412405015,
      "num_tokens": 19379777.0,
      "step": 10525
    },
    {
      "entropy": 5.292198038101196,
      "epoch": 0.8795522886735717,
      "grad_norm": 2.734375,
      "learning_rate": 0.000492884642095612,
      "loss": 5.1426,
      "mean_token_accuracy": 0.18770332336425782,
      "num_tokens": 19388717.0,
      "step": 10530
    },
    {
      "entropy": 5.430406284332276,
      "epoch": 0.8799699298362846,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004928772127395361,
      "loss": 5.2594,
      "mean_token_accuracy": 0.17967021018266677,
      "num_tokens": 19397936.0,
      "step": 10535
    },
    {
      "entropy": 5.449111366271973,
      "epoch": 0.8803875709989977,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004928697795692575,
      "loss": 5.2377,
      "mean_token_accuracy": 0.1796219006180763,
      "num_tokens": 19406830.0,
      "step": 10540
    },
    {
      "entropy": 5.391684436798096,
      "epoch": 0.8808052121617107,
      "grad_norm": 1.8515625,
      "learning_rate": 0.000492862342584906,
      "loss": 5.2393,
      "mean_token_accuracy": 0.18780584782361984,
      "num_tokens": 19415131.0,
      "step": 10545
    },
    {
      "entropy": 5.293669986724853,
      "epoch": 0.8812228533244236,
      "grad_norm": 1.5,
      "learning_rate": 0.000492854901786612,
      "loss": 5.0449,
      "mean_token_accuracy": 0.20062102228403092,
      "num_tokens": 19423585.0,
      "step": 10550
    },
    {
      "entropy": 5.236867094039917,
      "epoch": 0.8816404944871367,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004928474571745058,
      "loss": 5.0575,
      "mean_token_accuracy": 0.2019955188035965,
      "num_tokens": 19431890.0,
      "step": 10555
    },
    {
      "entropy": 5.3149364471435545,
      "epoch": 0.8820581356498497,
      "grad_norm": 2.1875,
      "learning_rate": 0.0004928400087487175,
      "loss": 5.2024,
      "mean_token_accuracy": 0.18816973268985748,
      "num_tokens": 19441377.0,
      "step": 10560
    },
    {
      "entropy": 5.341273832321167,
      "epoch": 0.8824757768125626,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004928325565093777,
      "loss": 5.1982,
      "mean_token_accuracy": 0.18512748777866364,
      "num_tokens": 19451003.0,
      "step": 10565
    },
    {
      "entropy": 5.263963937759399,
      "epoch": 0.8828934179752757,
      "grad_norm": 1.625,
      "learning_rate": 0.0004928251004566167,
      "loss": 5.0857,
      "mean_token_accuracy": 0.1944951444864273,
      "num_tokens": 19460166.0,
      "step": 10570
    },
    {
      "entropy": 5.249623775482178,
      "epoch": 0.8833110591379887,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004928176405905653,
      "loss": 5.1925,
      "mean_token_accuracy": 0.19052540510892868,
      "num_tokens": 19469417.0,
      "step": 10575
    },
    {
      "entropy": 5.357008075714111,
      "epoch": 0.8837287003007016,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004928101769113537,
      "loss": 5.2282,
      "mean_token_accuracy": 0.19121133685112,
      "num_tokens": 19479275.0,
      "step": 10580
    },
    {
      "entropy": 5.377585363388062,
      "epoch": 0.8841463414634146,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000492802709419113,
      "loss": 5.16,
      "mean_token_accuracy": 0.1852693811058998,
      "num_tokens": 19488804.0,
      "step": 10585
    },
    {
      "entropy": 5.372596406936646,
      "epoch": 0.8845639826261277,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004927952381139735,
      "loss": 5.1224,
      "mean_token_accuracy": 0.19373662322759627,
      "num_tokens": 19497701.0,
      "step": 10590
    },
    {
      "entropy": 5.350929021835327,
      "epoch": 0.8849816237888406,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004927877629960663,
      "loss": 5.1562,
      "mean_token_accuracy": 0.1899461030960083,
      "num_tokens": 19508177.0,
      "step": 10595
    },
    {
      "entropy": 5.321989917755127,
      "epoch": 0.8853992649515536,
      "grad_norm": 2.53125,
      "learning_rate": 0.0004927802840655221,
      "loss": 5.2128,
      "mean_token_accuracy": 0.18899890333414077,
      "num_tokens": 19516866.0,
      "step": 10600
    },
    {
      "entropy": 5.357509136199951,
      "epoch": 0.8858169061142667,
      "grad_norm": 1.6171875,
      "learning_rate": 0.000492772801322472,
      "loss": 5.2131,
      "mean_token_accuracy": 0.1834711253643036,
      "num_tokens": 19525788.0,
      "step": 10605
    },
    {
      "entropy": 5.330137538909912,
      "epoch": 0.8862345472769796,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004927653147670468,
      "loss": 5.0386,
      "mean_token_accuracy": 0.19043076783418655,
      "num_tokens": 19534748.0,
      "step": 10610
    },
    {
      "entropy": 5.2439979076385494,
      "epoch": 0.8866521884396926,
      "grad_norm": 2.171875,
      "learning_rate": 0.0004927578243993777,
      "loss": 5.128,
      "mean_token_accuracy": 0.19365649223327636,
      "num_tokens": 19544083.0,
      "step": 10615
    },
    {
      "entropy": 5.21814603805542,
      "epoch": 0.8870698296024057,
      "grad_norm": 1.5,
      "learning_rate": 0.0004927503302195957,
      "loss": 5.1503,
      "mean_token_accuracy": 0.18991438597440718,
      "num_tokens": 19553978.0,
      "step": 10620
    },
    {
      "entropy": 5.35388216972351,
      "epoch": 0.8874874707651186,
      "grad_norm": 1.53125,
      "learning_rate": 0.000492742832227832,
      "loss": 5.212,
      "mean_token_accuracy": 0.18995462059974672,
      "num_tokens": 19563265.0,
      "step": 10625
    },
    {
      "entropy": 5.413037967681885,
      "epoch": 0.8879051119278316,
      "grad_norm": 1.8125,
      "learning_rate": 0.000492735330424218,
      "loss": 5.2376,
      "mean_token_accuracy": 0.17343059033155442,
      "num_tokens": 19572904.0,
      "step": 10630
    },
    {
      "entropy": 5.40406641960144,
      "epoch": 0.8883227530905446,
      "grad_norm": 1.640625,
      "learning_rate": 0.000492727824808885,
      "loss": 5.2239,
      "mean_token_accuracy": 0.1931409627199173,
      "num_tokens": 19581680.0,
      "step": 10635
    },
    {
      "entropy": 5.270676708221435,
      "epoch": 0.8887403942532576,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004927203153819643,
      "loss": 5.1051,
      "mean_token_accuracy": 0.1901499405503273,
      "num_tokens": 19590885.0,
      "step": 10640
    },
    {
      "entropy": 5.241280460357666,
      "epoch": 0.8891580354159706,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004927128021435875,
      "loss": 5.0942,
      "mean_token_accuracy": 0.19374241679906845,
      "num_tokens": 19599901.0,
      "step": 10645
    },
    {
      "entropy": 5.370092678070068,
      "epoch": 0.8895756765786836,
      "grad_norm": 1.59375,
      "learning_rate": 0.000492705285093886,
      "loss": 5.2227,
      "mean_token_accuracy": 0.18349129408597947,
      "num_tokens": 19608745.0,
      "step": 10650
    },
    {
      "entropy": 5.277809715270996,
      "epoch": 0.8899933177413966,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004926977642329913,
      "loss": 5.0146,
      "mean_token_accuracy": 0.19165204763412474,
      "num_tokens": 19618418.0,
      "step": 10655
    },
    {
      "entropy": 5.279439640045166,
      "epoch": 0.8904109589041096,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004926902395610353,
      "loss": 5.0896,
      "mean_token_accuracy": 0.18721594661474228,
      "num_tokens": 19627118.0,
      "step": 10660
    },
    {
      "entropy": 5.3316254138946535,
      "epoch": 0.8908286000668226,
      "grad_norm": 1.5,
      "learning_rate": 0.0004926827110781496,
      "loss": 5.2442,
      "mean_token_accuracy": 0.18570893853902817,
      "num_tokens": 19636565.0,
      "step": 10665
    },
    {
      "entropy": 5.270682525634766,
      "epoch": 0.8912462412295356,
      "grad_norm": 1.9765625,
      "learning_rate": 0.000492675178784466,
      "loss": 5.1062,
      "mean_token_accuracy": 0.1896353021264076,
      "num_tokens": 19645038.0,
      "step": 10670
    },
    {
      "entropy": 5.279138278961182,
      "epoch": 0.8916638823922486,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004926676426801164,
      "loss": 5.1373,
      "mean_token_accuracy": 0.19063107967376708,
      "num_tokens": 19654088.0,
      "step": 10675
    },
    {
      "entropy": 5.321343564987183,
      "epoch": 0.8920815235549616,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004926601027652328,
      "loss": 5.2231,
      "mean_token_accuracy": 0.18899362981319429,
      "num_tokens": 19662551.0,
      "step": 10680
    },
    {
      "entropy": 5.398276138305664,
      "epoch": 0.8924991647176745,
      "grad_norm": 1.90625,
      "learning_rate": 0.000492652559039947,
      "loss": 5.1596,
      "mean_token_accuracy": 0.19150290936231612,
      "num_tokens": 19672405.0,
      "step": 10685
    },
    {
      "entropy": 5.265410947799682,
      "epoch": 0.8929168058803876,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004926450115043912,
      "loss": 5.1589,
      "mean_token_accuracy": 0.18920089155435563,
      "num_tokens": 19681288.0,
      "step": 10690
    },
    {
      "entropy": 5.2466329574584964,
      "epoch": 0.8933344470431006,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004926374601586975,
      "loss": 5.1056,
      "mean_token_accuracy": 0.1912362590432167,
      "num_tokens": 19690723.0,
      "step": 10695
    },
    {
      "entropy": 5.358538722991943,
      "epoch": 0.8937520882058135,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004926299050029981,
      "loss": 5.1343,
      "mean_token_accuracy": 0.18955470472574235,
      "num_tokens": 19699642.0,
      "step": 10700
    },
    {
      "entropy": 5.320053577423096,
      "epoch": 0.8941697293685266,
      "grad_norm": 3.234375,
      "learning_rate": 0.0004926223460374251,
      "loss": 5.0762,
      "mean_token_accuracy": 0.1930806428194046,
      "num_tokens": 19707563.0,
      "step": 10705
    },
    {
      "entropy": 5.3053521633148195,
      "epoch": 0.8945873705312396,
      "grad_norm": 1.625,
      "learning_rate": 0.0004926147832621111,
      "loss": 5.2436,
      "mean_token_accuracy": 0.18802300244569778,
      "num_tokens": 19717109.0,
      "step": 10710
    },
    {
      "entropy": 5.22956428527832,
      "epoch": 0.8950050116939525,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004926072166771884,
      "loss": 5.0662,
      "mean_token_accuracy": 0.19896597117185594,
      "num_tokens": 19725585.0,
      "step": 10715
    },
    {
      "entropy": 5.175854825973511,
      "epoch": 0.8954226528566656,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004925996462827895,
      "loss": 5.0034,
      "mean_token_accuracy": 0.20378232896327972,
      "num_tokens": 19733502.0,
      "step": 10720
    },
    {
      "entropy": 5.299879360198974,
      "epoch": 0.8958402940193786,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004925920720790466,
      "loss": 5.1465,
      "mean_token_accuracy": 0.1861718401312828,
      "num_tokens": 19742926.0,
      "step": 10725
    },
    {
      "entropy": 5.264319705963135,
      "epoch": 0.8962579351820915,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004925844940660928,
      "loss": 5.1276,
      "mean_token_accuracy": 0.19325936138629912,
      "num_tokens": 19752488.0,
      "step": 10730
    },
    {
      "entropy": 5.363840627670288,
      "epoch": 0.8966755763448045,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004925769122440606,
      "loss": 5.26,
      "mean_token_accuracy": 0.17976808696985244,
      "num_tokens": 19762705.0,
      "step": 10735
    },
    {
      "entropy": 5.22624888420105,
      "epoch": 0.8970932175075176,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004925693266130824,
      "loss": 5.0485,
      "mean_token_accuracy": 0.2002018615603447,
      "num_tokens": 19771156.0,
      "step": 10740
    },
    {
      "entropy": 5.264958715438842,
      "epoch": 0.8975108586702305,
      "grad_norm": 2.546875,
      "learning_rate": 0.0004925617371732914,
      "loss": 5.0965,
      "mean_token_accuracy": 0.1936448633670807,
      "num_tokens": 19780788.0,
      "step": 10745
    },
    {
      "entropy": 5.199323940277099,
      "epoch": 0.8979284998329435,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004925541439248203,
      "loss": 5.0699,
      "mean_token_accuracy": 0.20016645640134811,
      "num_tokens": 19790169.0,
      "step": 10750
    },
    {
      "entropy": 5.388787126541137,
      "epoch": 0.8983461409956566,
      "grad_norm": 1.5625,
      "learning_rate": 0.000492546546867802,
      "loss": 5.2905,
      "mean_token_accuracy": 0.17736679315567017,
      "num_tokens": 19800249.0,
      "step": 10755
    },
    {
      "entropy": 5.318563222885132,
      "epoch": 0.8987637821583695,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004925389460023696,
      "loss": 5.1454,
      "mean_token_accuracy": 0.19099222868680954,
      "num_tokens": 19809797.0,
      "step": 10760
    },
    {
      "entropy": 5.230153274536133,
      "epoch": 0.8991814233210825,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004925313413286559,
      "loss": 5.0364,
      "mean_token_accuracy": 0.1953825056552887,
      "num_tokens": 19818826.0,
      "step": 10765
    },
    {
      "entropy": 5.286236763000488,
      "epoch": 0.8995990644837956,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004925237328467945,
      "loss": 5.2179,
      "mean_token_accuracy": 0.19076265394687653,
      "num_tokens": 19827947.0,
      "step": 10770
    },
    {
      "entropy": 5.361205387115478,
      "epoch": 0.9000167056465085,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004925161205569182,
      "loss": 5.2106,
      "mean_token_accuracy": 0.18659184724092484,
      "num_tokens": 19835734.0,
      "step": 10775
    },
    {
      "entropy": 5.411896657943726,
      "epoch": 0.9004343468092215,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004925085044591604,
      "loss": 5.3222,
      "mean_token_accuracy": 0.17887253314256668,
      "num_tokens": 19845175.0,
      "step": 10780
    },
    {
      "entropy": 5.305286407470703,
      "epoch": 0.9008519879719346,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004925008845536544,
      "loss": 5.153,
      "mean_token_accuracy": 0.18973954319953917,
      "num_tokens": 19853684.0,
      "step": 10785
    },
    {
      "entropy": 5.348168087005615,
      "epoch": 0.9012696291346475,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004924932608405336,
      "loss": 5.1293,
      "mean_token_accuracy": 0.1882624715566635,
      "num_tokens": 19862813.0,
      "step": 10790
    },
    {
      "entropy": 5.347144985198975,
      "epoch": 0.9016872702973605,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004924856333199315,
      "loss": 5.1737,
      "mean_token_accuracy": 0.1923440143465996,
      "num_tokens": 19871545.0,
      "step": 10795
    },
    {
      "entropy": 5.308362054824829,
      "epoch": 0.9021049114600735,
      "grad_norm": 2.203125,
      "learning_rate": 0.0004924780019919816,
      "loss": 5.1371,
      "mean_token_accuracy": 0.186128169298172,
      "num_tokens": 19880060.0,
      "step": 10800
    },
    {
      "entropy": 5.289190149307251,
      "epoch": 0.9025225526227865,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004924703668568174,
      "loss": 5.1752,
      "mean_token_accuracy": 0.18785108178853988,
      "num_tokens": 19889151.0,
      "step": 10805
    },
    {
      "entropy": 5.363161087036133,
      "epoch": 0.9029401937854995,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004924627279145728,
      "loss": 5.1159,
      "mean_token_accuracy": 0.1965825855731964,
      "num_tokens": 19898087.0,
      "step": 10810
    },
    {
      "entropy": 5.298778772354126,
      "epoch": 0.9033578349482125,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004924550851653814,
      "loss": 5.1203,
      "mean_token_accuracy": 0.19995987564325332,
      "num_tokens": 19906196.0,
      "step": 10815
    },
    {
      "entropy": 5.31768798828125,
      "epoch": 0.9037754761109255,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004924474386093769,
      "loss": 5.1626,
      "mean_token_accuracy": 0.18674472272396087,
      "num_tokens": 19915837.0,
      "step": 10820
    },
    {
      "entropy": 5.273045253753662,
      "epoch": 0.9041931172736385,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004924397882466933,
      "loss": 5.1587,
      "mean_token_accuracy": 0.18707826882600784,
      "num_tokens": 19926012.0,
      "step": 10825
    },
    {
      "entropy": 5.275276613235474,
      "epoch": 0.9046107584363515,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004924321340774644,
      "loss": 5.1238,
      "mean_token_accuracy": 0.1945568010210991,
      "num_tokens": 19935987.0,
      "step": 10830
    },
    {
      "entropy": 5.277724266052246,
      "epoch": 0.9050283995990644,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004924244761018243,
      "loss": 5.1609,
      "mean_token_accuracy": 0.18308883309364318,
      "num_tokens": 19945865.0,
      "step": 10835
    },
    {
      "entropy": 5.3237203598022464,
      "epoch": 0.9054460407617775,
      "grad_norm": 1.84375,
      "learning_rate": 0.000492416814319907,
      "loss": 5.224,
      "mean_token_accuracy": 0.19275479316711425,
      "num_tokens": 19954824.0,
      "step": 10840
    },
    {
      "entropy": 5.264316654205322,
      "epoch": 0.9058636819244905,
      "grad_norm": 3.125,
      "learning_rate": 0.0004924091487318468,
      "loss": 5.1235,
      "mean_token_accuracy": 0.18456410467624665,
      "num_tokens": 19964740.0,
      "step": 10845
    },
    {
      "entropy": 5.248150300979614,
      "epoch": 0.9062813230872034,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004924014793377776,
      "loss": 5.0879,
      "mean_token_accuracy": 0.19337785840034485,
      "num_tokens": 19974310.0,
      "step": 10850
    },
    {
      "entropy": 5.235246229171753,
      "epoch": 0.9066989642499165,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004923938061378339,
      "loss": 5.0035,
      "mean_token_accuracy": 0.19750896394252776,
      "num_tokens": 19982943.0,
      "step": 10855
    },
    {
      "entropy": 5.2883970737457275,
      "epoch": 0.9071166054126295,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004923861291321499,
      "loss": 5.1282,
      "mean_token_accuracy": 0.1913818299770355,
      "num_tokens": 19991971.0,
      "step": 10860
    },
    {
      "entropy": 5.354921817779541,
      "epoch": 0.9075342465753424,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004923784483208601,
      "loss": 5.1884,
      "mean_token_accuracy": 0.18568297028541564,
      "num_tokens": 20001491.0,
      "step": 10865
    },
    {
      "entropy": 5.290411901473999,
      "epoch": 0.9079518877380555,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004923707637040989,
      "loss": 5.0595,
      "mean_token_accuracy": 0.1971973344683647,
      "num_tokens": 20010605.0,
      "step": 10870
    },
    {
      "entropy": 5.130425500869751,
      "epoch": 0.9083695289007685,
      "grad_norm": 2.671875,
      "learning_rate": 0.0004923630752820008,
      "loss": 4.9342,
      "mean_token_accuracy": 0.2001164361834526,
      "num_tokens": 20019291.0,
      "step": 10875
    },
    {
      "entropy": 5.232851934432984,
      "epoch": 0.9087871700634814,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0004923553830547004,
      "loss": 5.2368,
      "mean_token_accuracy": 0.18514347672462464,
      "num_tokens": 20029048.0,
      "step": 10880
    },
    {
      "entropy": 5.386471128463745,
      "epoch": 0.9092048112261945,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004923476870223324,
      "loss": 5.1938,
      "mean_token_accuracy": 0.18818991929292678,
      "num_tokens": 20037878.0,
      "step": 10885
    },
    {
      "entropy": 5.286022520065307,
      "epoch": 0.9096224523889075,
      "grad_norm": 2.515625,
      "learning_rate": 0.0004923399871850315,
      "loss": 5.1188,
      "mean_token_accuracy": 0.1949710115790367,
      "num_tokens": 20046662.0,
      "step": 10890
    },
    {
      "entropy": 5.223567628860474,
      "epoch": 0.9100400935516204,
      "grad_norm": 1.75,
      "learning_rate": 0.0004923322835429326,
      "loss": 5.1004,
      "mean_token_accuracy": 0.19120468497276305,
      "num_tokens": 20055329.0,
      "step": 10895
    },
    {
      "entropy": 5.195465993881226,
      "epoch": 0.9104577347143334,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004923245760961704,
      "loss": 5.011,
      "mean_token_accuracy": 0.2026669815182686,
      "num_tokens": 20064024.0,
      "step": 10900
    },
    {
      "entropy": 5.33611044883728,
      "epoch": 0.9108753758770465,
      "grad_norm": 2.125,
      "learning_rate": 0.0004923168648448799,
      "loss": 5.2224,
      "mean_token_accuracy": 0.18506976813077927,
      "num_tokens": 20073022.0,
      "step": 10905
    },
    {
      "entropy": 5.30430588722229,
      "epoch": 0.9112930170397594,
      "grad_norm": 2.328125,
      "learning_rate": 0.0004923091497891961,
      "loss": 5.1726,
      "mean_token_accuracy": 0.18535736501216887,
      "num_tokens": 20081495.0,
      "step": 10910
    },
    {
      "entropy": 5.344881772994995,
      "epoch": 0.9117106582024724,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004923014309292541,
      "loss": 5.1097,
      "mean_token_accuracy": 0.19913057386875152,
      "num_tokens": 20090545.0,
      "step": 10915
    },
    {
      "entropy": 5.327037811279297,
      "epoch": 0.9121282993651855,
      "grad_norm": 1.9296875,
      "learning_rate": 0.000492293708265189,
      "loss": 5.1835,
      "mean_token_accuracy": 0.19455758184194566,
      "num_tokens": 20100236.0,
      "step": 10920
    },
    {
      "entropy": 5.265479278564453,
      "epoch": 0.9125459405278984,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004922859817971359,
      "loss": 5.1328,
      "mean_token_accuracy": 0.18306612074375153,
      "num_tokens": 20109522.0,
      "step": 10925
    },
    {
      "entropy": 5.263588905334473,
      "epoch": 0.9129635816906114,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004922782515252302,
      "loss": 5.0669,
      "mean_token_accuracy": 0.19643136113882065,
      "num_tokens": 20118248.0,
      "step": 10930
    },
    {
      "entropy": 5.158023881912231,
      "epoch": 0.9133812228533245,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004922705174496072,
      "loss": 5.1207,
      "mean_token_accuracy": 0.188995860517025,
      "num_tokens": 20127096.0,
      "step": 10935
    },
    {
      "entropy": 5.232941007614135,
      "epoch": 0.9137988640160374,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004922627795704022,
      "loss": 5.1304,
      "mean_token_accuracy": 0.18677591234445573,
      "num_tokens": 20136815.0,
      "step": 10940
    },
    {
      "entropy": 5.323242568969727,
      "epoch": 0.9142165051787504,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004922550378877507,
      "loss": 5.1271,
      "mean_token_accuracy": 0.19914378374814987,
      "num_tokens": 20145455.0,
      "step": 10945
    },
    {
      "entropy": 5.339954042434693,
      "epoch": 0.9146341463414634,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004922472924017882,
      "loss": 5.2196,
      "mean_token_accuracy": 0.18386811316013335,
      "num_tokens": 20156139.0,
      "step": 10950
    },
    {
      "entropy": 5.345303297042847,
      "epoch": 0.9150517875041764,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004922395431126506,
      "loss": 5.1843,
      "mean_token_accuracy": 0.18810480535030366,
      "num_tokens": 20166266.0,
      "step": 10955
    },
    {
      "entropy": 5.416538572311401,
      "epoch": 0.9154694286668894,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004922317900204732,
      "loss": 5.2465,
      "mean_token_accuracy": 0.18291556239128112,
      "num_tokens": 20175005.0,
      "step": 10960
    },
    {
      "entropy": 5.225204706192017,
      "epoch": 0.9158870698296024,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004922240331253916,
      "loss": 5.0377,
      "mean_token_accuracy": 0.1997062399983406,
      "num_tokens": 20183488.0,
      "step": 10965
    },
    {
      "entropy": 5.266636800765991,
      "epoch": 0.9163047109923154,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004922162724275421,
      "loss": 5.1463,
      "mean_token_accuracy": 0.1962377279996872,
      "num_tokens": 20193724.0,
      "step": 10970
    },
    {
      "entropy": 5.342124700546265,
      "epoch": 0.9167223521550284,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004922085079270602,
      "loss": 5.238,
      "mean_token_accuracy": 0.1881588250398636,
      "num_tokens": 20203733.0,
      "step": 10975
    },
    {
      "entropy": 5.452734661102295,
      "epoch": 0.9171399933177414,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004922007396240819,
      "loss": 5.2073,
      "mean_token_accuracy": 0.19434769153594972,
      "num_tokens": 20213448.0,
      "step": 10980
    },
    {
      "entropy": 5.356554937362671,
      "epoch": 0.9175576344804544,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004921929675187433,
      "loss": 5.1263,
      "mean_token_accuracy": 0.18734360635280609,
      "num_tokens": 20222837.0,
      "step": 10985
    },
    {
      "entropy": 5.323911523818969,
      "epoch": 0.9179752756431674,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004921851916111802,
      "loss": 5.2899,
      "mean_token_accuracy": 0.1847895488142967,
      "num_tokens": 20232056.0,
      "step": 10990
    },
    {
      "entropy": 5.361779308319091,
      "epoch": 0.9183929168058804,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004921774119015289,
      "loss": 5.2158,
      "mean_token_accuracy": 0.18765031099319457,
      "num_tokens": 20241052.0,
      "step": 10995
    },
    {
      "entropy": 5.379041576385498,
      "epoch": 0.9188105579685933,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004921696283899257,
      "loss": 5.2776,
      "mean_token_accuracy": 0.18354066982865333,
      "num_tokens": 20251015.0,
      "step": 11000
    },
    {
      "entropy": 5.271229314804077,
      "epoch": 0.9192281991313064,
      "grad_norm": 2.4375,
      "learning_rate": 0.0004921618410765067,
      "loss": 5.1153,
      "mean_token_accuracy": 0.19168516099452973,
      "num_tokens": 20259388.0,
      "step": 11005
    },
    {
      "entropy": 5.378978967666626,
      "epoch": 0.9196458402940194,
      "grad_norm": 2.59375,
      "learning_rate": 0.0004921540499614083,
      "loss": 5.1008,
      "mean_token_accuracy": 0.19158013612031938,
      "num_tokens": 20268146.0,
      "step": 11010
    },
    {
      "entropy": 5.273536682128906,
      "epoch": 0.9200634814567323,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004921462550447669,
      "loss": 5.0677,
      "mean_token_accuracy": 0.1963006317615509,
      "num_tokens": 20277559.0,
      "step": 11015
    },
    {
      "entropy": 5.276912784576416,
      "epoch": 0.9204811226194454,
      "grad_norm": 2.84375,
      "learning_rate": 0.0004921384563267189,
      "loss": 5.1328,
      "mean_token_accuracy": 0.18395720720291137,
      "num_tokens": 20285661.0,
      "step": 11020
    },
    {
      "entropy": 5.326535844802857,
      "epoch": 0.9208987637821584,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004921306538074008,
      "loss": 5.1978,
      "mean_token_accuracy": 0.1907956123352051,
      "num_tokens": 20294636.0,
      "step": 11025
    },
    {
      "entropy": 5.3194504737854,
      "epoch": 0.9213164049448713,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004921228474869493,
      "loss": 5.1017,
      "mean_token_accuracy": 0.1956602394580841,
      "num_tokens": 20302372.0,
      "step": 11030
    },
    {
      "entropy": 5.30044379234314,
      "epoch": 0.9217340461075844,
      "grad_norm": 1.7578125,
      "learning_rate": 0.000492115037365501,
      "loss": 5.1339,
      "mean_token_accuracy": 0.19145287573337555,
      "num_tokens": 20311974.0,
      "step": 11035
    },
    {
      "entropy": 5.297578382492065,
      "epoch": 0.9221516872702974,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004921072234431925,
      "loss": 5.1489,
      "mean_token_accuracy": 0.1912020057439804,
      "num_tokens": 20321720.0,
      "step": 11040
    },
    {
      "entropy": 5.326252603530884,
      "epoch": 0.9225693284330103,
      "grad_norm": 1.5859375,
      "learning_rate": 0.000492099405720161,
      "loss": 5.2084,
      "mean_token_accuracy": 0.18220922201871873,
      "num_tokens": 20331324.0,
      "step": 11045
    },
    {
      "entropy": 5.349126148223877,
      "epoch": 0.9229869695957233,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004920915841965429,
      "loss": 5.2731,
      "mean_token_accuracy": 0.185405170917511,
      "num_tokens": 20340921.0,
      "step": 11050
    },
    {
      "entropy": 5.402704906463623,
      "epoch": 0.9234046107584364,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004920837588724753,
      "loss": 5.1698,
      "mean_token_accuracy": 0.19052685499191285,
      "num_tokens": 20349871.0,
      "step": 11055
    },
    {
      "entropy": 5.206499862670898,
      "epoch": 0.9238222519211493,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004920759297480953,
      "loss": 4.9386,
      "mean_token_accuracy": 0.19570755064487458,
      "num_tokens": 20359171.0,
      "step": 11060
    },
    {
      "entropy": 5.247551012039184,
      "epoch": 0.9242398930838623,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004920680968235397,
      "loss": 5.1494,
      "mean_token_accuracy": 0.19701757580041884,
      "num_tokens": 20367779.0,
      "step": 11065
    },
    {
      "entropy": 5.256858015060425,
      "epoch": 0.9246575342465754,
      "grad_norm": 2.609375,
      "learning_rate": 0.0004920602600989459,
      "loss": 5.0731,
      "mean_token_accuracy": 0.19889040440320968,
      "num_tokens": 20377568.0,
      "step": 11070
    },
    {
      "entropy": 5.340265560150146,
      "epoch": 0.9250751754092883,
      "grad_norm": 2.359375,
      "learning_rate": 0.0004920524195744509,
      "loss": 5.2268,
      "mean_token_accuracy": 0.18503626585006713,
      "num_tokens": 20385915.0,
      "step": 11075
    },
    {
      "entropy": 5.243323707580567,
      "epoch": 0.9254928165720013,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004920445752501921,
      "loss": 5.0694,
      "mean_token_accuracy": 0.1943541631102562,
      "num_tokens": 20394877.0,
      "step": 11080
    },
    {
      "entropy": 5.265145397186279,
      "epoch": 0.9259104577347144,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004920367271263066,
      "loss": 5.0823,
      "mean_token_accuracy": 0.1932800829410553,
      "num_tokens": 20403741.0,
      "step": 11085
    },
    {
      "entropy": 5.271869754791259,
      "epoch": 0.9263280988974273,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004920288752029321,
      "loss": 5.1256,
      "mean_token_accuracy": 0.19450776129961014,
      "num_tokens": 20412702.0,
      "step": 11090
    },
    {
      "entropy": 5.250715827941894,
      "epoch": 0.9267457400601403,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004920210194802059,
      "loss": 5.1626,
      "mean_token_accuracy": 0.19392989724874496,
      "num_tokens": 20421776.0,
      "step": 11095
    },
    {
      "entropy": 5.363069295883179,
      "epoch": 0.9271633812228534,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004920131599582654,
      "loss": 5.2699,
      "mean_token_accuracy": 0.1774836540222168,
      "num_tokens": 20432204.0,
      "step": 11100
    },
    {
      "entropy": 5.401186466217041,
      "epoch": 0.9275810223855663,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004920052966372485,
      "loss": 5.1652,
      "mean_token_accuracy": 0.18065235167741775,
      "num_tokens": 20441344.0,
      "step": 11105
    },
    {
      "entropy": 5.299026727676392,
      "epoch": 0.9279986635482793,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004919974295172926,
      "loss": 5.0697,
      "mean_token_accuracy": 0.193643781542778,
      "num_tokens": 20449800.0,
      "step": 11110
    },
    {
      "entropy": 5.238025188446045,
      "epoch": 0.9284163047109923,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004919895585985353,
      "loss": 5.017,
      "mean_token_accuracy": 0.20537518113851547,
      "num_tokens": 20458326.0,
      "step": 11115
    },
    {
      "entropy": 5.28767580986023,
      "epoch": 0.9288339458737053,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004919816838811149,
      "loss": 5.2846,
      "mean_token_accuracy": 0.18285572230815889,
      "num_tokens": 20468050.0,
      "step": 11120
    },
    {
      "entropy": 5.346377992630005,
      "epoch": 0.9292515870364183,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004919738053651687,
      "loss": 5.0514,
      "mean_token_accuracy": 0.19551948606967925,
      "num_tokens": 20477046.0,
      "step": 11125
    },
    {
      "entropy": 5.274564123153686,
      "epoch": 0.9296692281991313,
      "grad_norm": 1.6640625,
      "learning_rate": 0.000491965923050835,
      "loss": 5.151,
      "mean_token_accuracy": 0.18730415105819703,
      "num_tokens": 20486228.0,
      "step": 11130
    },
    {
      "entropy": 5.310534954071045,
      "epoch": 0.9300868693618443,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004919580369382516,
      "loss": 5.1831,
      "mean_token_accuracy": 0.18330602049827577,
      "num_tokens": 20494764.0,
      "step": 11135
    },
    {
      "entropy": 5.310151481628418,
      "epoch": 0.9305045105245573,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004919501470275567,
      "loss": 5.0927,
      "mean_token_accuracy": 0.19177802354097367,
      "num_tokens": 20502805.0,
      "step": 11140
    },
    {
      "entropy": 5.246659469604492,
      "epoch": 0.9309221516872703,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004919422533188882,
      "loss": 5.1541,
      "mean_token_accuracy": 0.1869480162858963,
      "num_tokens": 20512322.0,
      "step": 11145
    },
    {
      "entropy": 5.272397565841675,
      "epoch": 0.9313397928499833,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004919343558123845,
      "loss": 5.1346,
      "mean_token_accuracy": 0.18849266916513444,
      "num_tokens": 20520680.0,
      "step": 11150
    },
    {
      "entropy": 5.285646057128906,
      "epoch": 0.9317574340126963,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004919264545081838,
      "loss": 4.9893,
      "mean_token_accuracy": 0.2047054126858711,
      "num_tokens": 20528864.0,
      "step": 11155
    },
    {
      "entropy": 5.260857439041137,
      "epoch": 0.9321750751754093,
      "grad_norm": 2.125,
      "learning_rate": 0.0004919185494064244,
      "loss": 5.1262,
      "mean_token_accuracy": 0.1858099415898323,
      "num_tokens": 20538195.0,
      "step": 11160
    },
    {
      "entropy": 5.279745483398438,
      "epoch": 0.9325927163381222,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0004919106405072446,
      "loss": 5.1563,
      "mean_token_accuracy": 0.19002094566822053,
      "num_tokens": 20547312.0,
      "step": 11165
    },
    {
      "entropy": 5.331996822357178,
      "epoch": 0.9330103575008353,
      "grad_norm": 1.53125,
      "learning_rate": 0.000491902727810783,
      "loss": 5.1893,
      "mean_token_accuracy": 0.18345918357372284,
      "num_tokens": 20557184.0,
      "step": 11170
    },
    {
      "entropy": 5.2742408275604244,
      "epoch": 0.9334279986635483,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004918948113171781,
      "loss": 5.1348,
      "mean_token_accuracy": 0.19250834733247757,
      "num_tokens": 20565819.0,
      "step": 11175
    },
    {
      "entropy": 5.260569286346436,
      "epoch": 0.9338456398262612,
      "grad_norm": 1.875,
      "learning_rate": 0.0004918868910265684,
      "loss": 5.1838,
      "mean_token_accuracy": 0.19066485911607742,
      "num_tokens": 20574761.0,
      "step": 11180
    },
    {
      "entropy": 5.406101369857788,
      "epoch": 0.9342632809889743,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004918789669390927,
      "loss": 5.2245,
      "mean_token_accuracy": 0.18671819418668748,
      "num_tokens": 20584003.0,
      "step": 11185
    },
    {
      "entropy": 5.283585691452027,
      "epoch": 0.9346809221516873,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004918710390548895,
      "loss": 5.0894,
      "mean_token_accuracy": 0.18976902812719346,
      "num_tokens": 20593123.0,
      "step": 11190
    },
    {
      "entropy": 5.250103616714478,
      "epoch": 0.9350985633144002,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004918631073740977,
      "loss": 5.1595,
      "mean_token_accuracy": 0.18998567759990692,
      "num_tokens": 20602582.0,
      "step": 11195
    },
    {
      "entropy": 5.296892309188843,
      "epoch": 0.9355162044771133,
      "grad_norm": 1.90625,
      "learning_rate": 0.0004918551718968562,
      "loss": 5.0745,
      "mean_token_accuracy": 0.19461484998464584,
      "num_tokens": 20611681.0,
      "step": 11200
    },
    {
      "entropy": 5.290851354598999,
      "epoch": 0.9359338456398263,
      "grad_norm": 2.171875,
      "learning_rate": 0.000491847232623304,
      "loss": 5.1398,
      "mean_token_accuracy": 0.1950131133198738,
      "num_tokens": 20620671.0,
      "step": 11205
    },
    {
      "entropy": 5.283709478378296,
      "epoch": 0.9363514868025392,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004918392895535798,
      "loss": 5.1054,
      "mean_token_accuracy": 0.18826131671667098,
      "num_tokens": 20629986.0,
      "step": 11210
    },
    {
      "entropy": 5.3349779605865475,
      "epoch": 0.9367691279652522,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000491831342687823,
      "loss": 5.1054,
      "mean_token_accuracy": 0.19026129245758056,
      "num_tokens": 20639274.0,
      "step": 11215
    },
    {
      "entropy": 5.308081912994385,
      "epoch": 0.9371867691279653,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0004918233920261723,
      "loss": 5.123,
      "mean_token_accuracy": 0.1903127670288086,
      "num_tokens": 20647753.0,
      "step": 11220
    },
    {
      "entropy": 5.17316951751709,
      "epoch": 0.9376044102906782,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004918154375687673,
      "loss": 5.0179,
      "mean_token_accuracy": 0.20007274597883223,
      "num_tokens": 20656226.0,
      "step": 11225
    },
    {
      "entropy": 5.194850063323974,
      "epoch": 0.9380220514533912,
      "grad_norm": 2.125,
      "learning_rate": 0.0004918074793157472,
      "loss": 5.0213,
      "mean_token_accuracy": 0.19764335155487062,
      "num_tokens": 20664682.0,
      "step": 11230
    },
    {
      "entropy": 5.331245851516724,
      "epoch": 0.9384396926161043,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004917995172672511,
      "loss": 5.1377,
      "mean_token_accuracy": 0.18766049295663834,
      "num_tokens": 20672909.0,
      "step": 11235
    },
    {
      "entropy": 5.212302780151367,
      "epoch": 0.9388573337788172,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004917915514234184,
      "loss": 4.9808,
      "mean_token_accuracy": 0.2021414890885353,
      "num_tokens": 20681316.0,
      "step": 11240
    },
    {
      "entropy": 5.279926443099976,
      "epoch": 0.9392749749415302,
      "grad_norm": 2.515625,
      "learning_rate": 0.0004917835817843887,
      "loss": 5.0968,
      "mean_token_accuracy": 0.18689426630735398,
      "num_tokens": 20690360.0,
      "step": 11245
    },
    {
      "entropy": 5.330228090286255,
      "epoch": 0.9396926161042433,
      "grad_norm": 1.625,
      "learning_rate": 0.0004917756083503014,
      "loss": 5.146,
      "mean_token_accuracy": 0.18522235304117202,
      "num_tokens": 20700398.0,
      "step": 11250
    },
    {
      "entropy": 5.363566207885742,
      "epoch": 0.9401102572669562,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004917676311212962,
      "loss": 5.2367,
      "mean_token_accuracy": 0.18600541204214097,
      "num_tokens": 20709705.0,
      "step": 11255
    },
    {
      "entropy": 5.238001728057862,
      "epoch": 0.9405278984296692,
      "grad_norm": 2.265625,
      "learning_rate": 0.0004917596500975127,
      "loss": 5.0333,
      "mean_token_accuracy": 0.1972285643219948,
      "num_tokens": 20717774.0,
      "step": 11260
    },
    {
      "entropy": 5.296810054779053,
      "epoch": 0.9409455395923823,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004917516652790907,
      "loss": 5.2007,
      "mean_token_accuracy": 0.1933752939105034,
      "num_tokens": 20725966.0,
      "step": 11265
    },
    {
      "entropy": 5.339372205734253,
      "epoch": 0.9413631807550952,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004917436766661697,
      "loss": 5.0924,
      "mean_token_accuracy": 0.19547368437051774,
      "num_tokens": 20735254.0,
      "step": 11270
    },
    {
      "entropy": 5.292354536056519,
      "epoch": 0.9417808219178082,
      "grad_norm": 1.8828125,
      "learning_rate": 0.00049173568425889,
      "loss": 5.0094,
      "mean_token_accuracy": 0.20186227560043335,
      "num_tokens": 20743331.0,
      "step": 11275
    },
    {
      "entropy": 5.275941276550293,
      "epoch": 0.9421984630805212,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004917276880573911,
      "loss": 5.134,
      "mean_token_accuracy": 0.1867146834731102,
      "num_tokens": 20753966.0,
      "step": 11280
    },
    {
      "entropy": 5.270225381851196,
      "epoch": 0.9426161042432342,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004917196880618132,
      "loss": 5.0735,
      "mean_token_accuracy": 0.19647087901830673,
      "num_tokens": 20762608.0,
      "step": 11285
    },
    {
      "entropy": 5.209722185134888,
      "epoch": 0.9430337454059472,
      "grad_norm": 1.875,
      "learning_rate": 0.0004917116842722962,
      "loss": 5.0087,
      "mean_token_accuracy": 0.19446980357170104,
      "num_tokens": 20771194.0,
      "step": 11290
    },
    {
      "entropy": 5.278521490097046,
      "epoch": 0.9434513865686602,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004917036766889804,
      "loss": 5.1193,
      "mean_token_accuracy": 0.19536812752485275,
      "num_tokens": 20779700.0,
      "step": 11295
    },
    {
      "entropy": 5.364008235931396,
      "epoch": 0.9438690277313732,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004916956653120058,
      "loss": 5.2433,
      "mean_token_accuracy": 0.18282273262739182,
      "num_tokens": 20788117.0,
      "step": 11300
    },
    {
      "entropy": 5.3220068454742435,
      "epoch": 0.9442866688940862,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004916876501415129,
      "loss": 5.2287,
      "mean_token_accuracy": 0.18652146607637404,
      "num_tokens": 20798041.0,
      "step": 11305
    },
    {
      "entropy": 5.310404253005982,
      "epoch": 0.9447043100567992,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004916796311776417,
      "loss": 5.1465,
      "mean_token_accuracy": 0.1839068502187729,
      "num_tokens": 20807423.0,
      "step": 11310
    },
    {
      "entropy": 5.2947508811950685,
      "epoch": 0.9451219512195121,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004916716084205328,
      "loss": 5.0623,
      "mean_token_accuracy": 0.19841802716255189,
      "num_tokens": 20816727.0,
      "step": 11315
    },
    {
      "entropy": 5.262984037399292,
      "epoch": 0.9455395923822252,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004916635818703266,
      "loss": 5.0625,
      "mean_token_accuracy": 0.18846407979726792,
      "num_tokens": 20824855.0,
      "step": 11320
    },
    {
      "entropy": 5.201770353317261,
      "epoch": 0.9459572335449382,
      "grad_norm": 1.9375,
      "learning_rate": 0.0004916555515271635,
      "loss": 5.0727,
      "mean_token_accuracy": 0.196591155230999,
      "num_tokens": 20833096.0,
      "step": 11325
    },
    {
      "entropy": 5.2126647472381595,
      "epoch": 0.9463748747076511,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004916475173911844,
      "loss": 5.0424,
      "mean_token_accuracy": 0.19711389988660813,
      "num_tokens": 20842398.0,
      "step": 11330
    },
    {
      "entropy": 5.326529884338379,
      "epoch": 0.9467925158703642,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004916394794625296,
      "loss": 5.2461,
      "mean_token_accuracy": 0.18619543313980103,
      "num_tokens": 20851745.0,
      "step": 11335
    },
    {
      "entropy": 5.459332418441773,
      "epoch": 0.9472101570330772,
      "grad_norm": 2.328125,
      "learning_rate": 0.0004916314377413399,
      "loss": 5.1886,
      "mean_token_accuracy": 0.18649569600820542,
      "num_tokens": 20860935.0,
      "step": 11340
    },
    {
      "entropy": 5.332683086395264,
      "epoch": 0.9476277981957901,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004916233922277562,
      "loss": 5.1536,
      "mean_token_accuracy": 0.18652458041906356,
      "num_tokens": 20871088.0,
      "step": 11345
    },
    {
      "entropy": 5.267989206314087,
      "epoch": 0.9480454393585032,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004916153429219192,
      "loss": 5.1899,
      "mean_token_accuracy": 0.19276488721370696,
      "num_tokens": 20879313.0,
      "step": 11350
    },
    {
      "entropy": 5.288082408905029,
      "epoch": 0.9484630805212162,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004916072898239699,
      "loss": 5.133,
      "mean_token_accuracy": 0.188817335665226,
      "num_tokens": 20888392.0,
      "step": 11355
    },
    {
      "entropy": 5.3171515464782715,
      "epoch": 0.9488807216839291,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004915992329340493,
      "loss": 5.2257,
      "mean_token_accuracy": 0.18004927039146423,
      "num_tokens": 20897456.0,
      "step": 11360
    },
    {
      "entropy": 5.319647026062012,
      "epoch": 0.9492983628466422,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004915911722522984,
      "loss": 5.144,
      "mean_token_accuracy": 0.1874665066599846,
      "num_tokens": 20906621.0,
      "step": 11365
    },
    {
      "entropy": 5.4354753494262695,
      "epoch": 0.9497160040093552,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004915831077788583,
      "loss": 5.3414,
      "mean_token_accuracy": 0.18478410840034484,
      "num_tokens": 20916929.0,
      "step": 11370
    },
    {
      "entropy": 5.323004484176636,
      "epoch": 0.9501336451720681,
      "grad_norm": 2.4375,
      "learning_rate": 0.0004915750395138702,
      "loss": 5.1464,
      "mean_token_accuracy": 0.19146731793880462,
      "num_tokens": 20926028.0,
      "step": 11375
    },
    {
      "entropy": 5.309450626373291,
      "epoch": 0.9505512863347811,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004915669674574754,
      "loss": 5.25,
      "mean_token_accuracy": 0.18093163669109344,
      "num_tokens": 20935516.0,
      "step": 11380
    },
    {
      "entropy": 5.2704212188720705,
      "epoch": 0.9509689274974942,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004915588916098151,
      "loss": 5.0336,
      "mean_token_accuracy": 0.19577372819185257,
      "num_tokens": 20944100.0,
      "step": 11385
    },
    {
      "entropy": 5.292933511734009,
      "epoch": 0.9513865686602071,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004915508119710307,
      "loss": 5.1717,
      "mean_token_accuracy": 0.18894966691732407,
      "num_tokens": 20953733.0,
      "step": 11390
    },
    {
      "entropy": 5.375700569152832,
      "epoch": 0.9518042098229201,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004915427285412636,
      "loss": 5.1813,
      "mean_token_accuracy": 0.1801777586340904,
      "num_tokens": 20962413.0,
      "step": 11395
    },
    {
      "entropy": 5.241631650924683,
      "epoch": 0.9522218509856332,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004915346413206555,
      "loss": 5.0426,
      "mean_token_accuracy": 0.2037292703986168,
      "num_tokens": 20971172.0,
      "step": 11400
    },
    {
      "entropy": 5.231201124191284,
      "epoch": 0.9526394921483461,
      "grad_norm": 2.46875,
      "learning_rate": 0.0004915265503093478,
      "loss": 5.0709,
      "mean_token_accuracy": 0.1976637974381447,
      "num_tokens": 20979450.0,
      "step": 11405
    },
    {
      "entropy": 5.260703754425049,
      "epoch": 0.9530571333110591,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004915184555074824,
      "loss": 5.1409,
      "mean_token_accuracy": 0.18479856103658676,
      "num_tokens": 20988518.0,
      "step": 11410
    },
    {
      "entropy": 5.44167914390564,
      "epoch": 0.9534747744737722,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004915103569152005,
      "loss": 5.3107,
      "mean_token_accuracy": 0.18017488569021226,
      "num_tokens": 20998721.0,
      "step": 11415
    },
    {
      "entropy": 5.3801109313964846,
      "epoch": 0.9538924156364851,
      "grad_norm": 2.5625,
      "learning_rate": 0.0004915022545326444,
      "loss": 5.1781,
      "mean_token_accuracy": 0.1904640778899193,
      "num_tokens": 21008007.0,
      "step": 11420
    },
    {
      "entropy": 5.293911361694336,
      "epoch": 0.9543100567991981,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004914941483599558,
      "loss": 5.2452,
      "mean_token_accuracy": 0.18404674232006074,
      "num_tokens": 21017199.0,
      "step": 11425
    },
    {
      "entropy": 5.188850021362304,
      "epoch": 0.9547276979619111,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004914860383972763,
      "loss": 5.0462,
      "mean_token_accuracy": 0.19653122425079345,
      "num_tokens": 21025571.0,
      "step": 11430
    },
    {
      "entropy": 5.330324983596801,
      "epoch": 0.9551453391246241,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004914779246447483,
      "loss": 5.1478,
      "mean_token_accuracy": 0.19278790205717086,
      "num_tokens": 21035040.0,
      "step": 11435
    },
    {
      "entropy": 5.365421342849731,
      "epoch": 0.9555629802873371,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004914698071025136,
      "loss": 5.1696,
      "mean_token_accuracy": 0.19380325675010682,
      "num_tokens": 21044018.0,
      "step": 11440
    },
    {
      "entropy": 5.254361629486084,
      "epoch": 0.9559806214500501,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004914616857707143,
      "loss": 5.0717,
      "mean_token_accuracy": 0.19683144390583038,
      "num_tokens": 21052489.0,
      "step": 11445
    },
    {
      "entropy": 5.305207538604736,
      "epoch": 0.9563982626127631,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004914535606494927,
      "loss": 5.1449,
      "mean_token_accuracy": 0.19891838133335113,
      "num_tokens": 21061153.0,
      "step": 11450
    },
    {
      "entropy": 5.240456485748291,
      "epoch": 0.9568159037754761,
      "grad_norm": 1.5390625,
      "learning_rate": 0.000491445431738991,
      "loss": 5.0367,
      "mean_token_accuracy": 0.1911842942237854,
      "num_tokens": 21069671.0,
      "step": 11455
    },
    {
      "entropy": 5.408563566207886,
      "epoch": 0.9572335449381891,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004914372990393515,
      "loss": 5.3486,
      "mean_token_accuracy": 0.17980122417211533,
      "num_tokens": 21079339.0,
      "step": 11460
    },
    {
      "entropy": 5.296997117996216,
      "epoch": 0.957651186100902,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004914291625507164,
      "loss": 5.0455,
      "mean_token_accuracy": 0.19713663309812546,
      "num_tokens": 21087868.0,
      "step": 11465
    },
    {
      "entropy": 5.303809928894043,
      "epoch": 0.9580688272636151,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004914210222732284,
      "loss": 5.1488,
      "mean_token_accuracy": 0.19415508657693864,
      "num_tokens": 21096258.0,
      "step": 11470
    },
    {
      "entropy": 5.257193231582642,
      "epoch": 0.9584864684263281,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004914128782070299,
      "loss": 5.1573,
      "mean_token_accuracy": 0.18823350816965104,
      "num_tokens": 21104504.0,
      "step": 11475
    },
    {
      "entropy": 5.261191463470459,
      "epoch": 0.958904109589041,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004914047303522636,
      "loss": 5.0956,
      "mean_token_accuracy": 0.19104788154363633,
      "num_tokens": 21114402.0,
      "step": 11480
    },
    {
      "entropy": 5.26935601234436,
      "epoch": 0.9593217507517541,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004913965787090718,
      "loss": 5.1131,
      "mean_token_accuracy": 0.19183927476406099,
      "num_tokens": 21123572.0,
      "step": 11485
    },
    {
      "entropy": 5.382525205612183,
      "epoch": 0.9597393919144671,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004913884232775974,
      "loss": 5.3069,
      "mean_token_accuracy": 0.1748875617980957,
      "num_tokens": 21132385.0,
      "step": 11490
    },
    {
      "entropy": 5.203315162658692,
      "epoch": 0.96015703307718,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004913802640579833,
      "loss": 5.0582,
      "mean_token_accuracy": 0.19596102833747864,
      "num_tokens": 21140383.0,
      "step": 11495
    },
    {
      "entropy": 5.394109916687012,
      "epoch": 0.9605746742398931,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004913721010503722,
      "loss": 5.1642,
      "mean_token_accuracy": 0.18963964134454728,
      "num_tokens": 21149793.0,
      "step": 11500
    },
    {
      "entropy": 5.343473768234253,
      "epoch": 0.9609923154026061,
      "grad_norm": 2.34375,
      "learning_rate": 0.000491363934254907,
      "loss": 5.1417,
      "mean_token_accuracy": 0.18909443020820618,
      "num_tokens": 21158955.0,
      "step": 11505
    },
    {
      "entropy": 5.326027345657349,
      "epoch": 0.961409956565319,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004913557636717307,
      "loss": 5.1401,
      "mean_token_accuracy": 0.18643326908349991,
      "num_tokens": 21167900.0,
      "step": 11510
    },
    {
      "entropy": 5.24142804145813,
      "epoch": 0.9618275977280321,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004913475893009864,
      "loss": 5.0908,
      "mean_token_accuracy": 0.19366782605648042,
      "num_tokens": 21177243.0,
      "step": 11515
    },
    {
      "entropy": 5.299542379379273,
      "epoch": 0.9622452388907451,
      "grad_norm": 1.640625,
      "learning_rate": 0.000491339411142817,
      "loss": 5.1733,
      "mean_token_accuracy": 0.19068634510040283,
      "num_tokens": 21186118.0,
      "step": 11520
    },
    {
      "entropy": 5.347597646713257,
      "epoch": 0.962662880053458,
      "grad_norm": 2.90625,
      "learning_rate": 0.0004913312291973659,
      "loss": 5.1909,
      "mean_token_accuracy": 0.18620602041482925,
      "num_tokens": 21195032.0,
      "step": 11525
    },
    {
      "entropy": 5.380717658996582,
      "epoch": 0.963080521216171,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004913230434647762,
      "loss": 5.1805,
      "mean_token_accuracy": 0.1849704220890999,
      "num_tokens": 21205388.0,
      "step": 11530
    },
    {
      "entropy": 5.302324724197388,
      "epoch": 0.9634981623788841,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004913148539451912,
      "loss": 5.0599,
      "mean_token_accuracy": 0.19199014604091644,
      "num_tokens": 21215181.0,
      "step": 11535
    },
    {
      "entropy": 5.189516878128051,
      "epoch": 0.963915803541597,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004913066606387544,
      "loss": 4.9804,
      "mean_token_accuracy": 0.19661485999822617,
      "num_tokens": 21223668.0,
      "step": 11540
    },
    {
      "entropy": 5.189825487136841,
      "epoch": 0.96433344470431,
      "grad_norm": 1.625,
      "learning_rate": 0.000491298463545609,
      "loss": 5.1138,
      "mean_token_accuracy": 0.18589077442884444,
      "num_tokens": 21232448.0,
      "step": 11545
    },
    {
      "entropy": 5.2891865253448485,
      "epoch": 0.9647510858670231,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004912902626658988,
      "loss": 5.0887,
      "mean_token_accuracy": 0.19474891871213912,
      "num_tokens": 21240777.0,
      "step": 11550
    },
    {
      "entropy": 5.354679107666016,
      "epoch": 0.965168727029736,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004912820579997672,
      "loss": 5.2119,
      "mean_token_accuracy": 0.18303453028202057,
      "num_tokens": 21248803.0,
      "step": 11555
    },
    {
      "entropy": 5.3409021377563475,
      "epoch": 0.965586368192449,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004912738495473577,
      "loss": 5.1719,
      "mean_token_accuracy": 0.1852123960852623,
      "num_tokens": 21257861.0,
      "step": 11560
    },
    {
      "entropy": 5.314940071105957,
      "epoch": 0.9660040093551621,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004912656373088142,
      "loss": 5.0666,
      "mean_token_accuracy": 0.20077644139528275,
      "num_tokens": 21267140.0,
      "step": 11565
    },
    {
      "entropy": 5.28928689956665,
      "epoch": 0.9664216505178751,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004912574212842804,
      "loss": 5.1815,
      "mean_token_accuracy": 0.1886407658457756,
      "num_tokens": 21276313.0,
      "step": 11570
    },
    {
      "entropy": 5.23785252571106,
      "epoch": 0.966839291680588,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004912492014739001,
      "loss": 5.0942,
      "mean_token_accuracy": 0.19174057245254517,
      "num_tokens": 21286156.0,
      "step": 11575
    },
    {
      "entropy": 5.362173986434937,
      "epoch": 0.967256932843301,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004912409778778173,
      "loss": 5.1886,
      "mean_token_accuracy": 0.19431511759757997,
      "num_tokens": 21295420.0,
      "step": 11580
    },
    {
      "entropy": 5.403459548950195,
      "epoch": 0.9676745740060141,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004912327504961758,
      "loss": 5.3534,
      "mean_token_accuracy": 0.18076042234897613,
      "num_tokens": 21306292.0,
      "step": 11585
    },
    {
      "entropy": 5.322366619110108,
      "epoch": 0.968092215168727,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004912245193291198,
      "loss": 5.1864,
      "mean_token_accuracy": 0.18871198445558549,
      "num_tokens": 21315165.0,
      "step": 11590
    },
    {
      "entropy": 5.335630273818969,
      "epoch": 0.96850985633144,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004912162843767932,
      "loss": 5.1686,
      "mean_token_accuracy": 0.18643238097429277,
      "num_tokens": 21324608.0,
      "step": 11595
    },
    {
      "entropy": 5.315397024154663,
      "epoch": 0.9689274974941531,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004912080456393404,
      "loss": 5.1183,
      "mean_token_accuracy": 0.18949589729309083,
      "num_tokens": 21333609.0,
      "step": 11600
    },
    {
      "entropy": 5.347077798843384,
      "epoch": 0.969345138656866,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0004911998031169056,
      "loss": 5.2402,
      "mean_token_accuracy": 0.18186257034540176,
      "num_tokens": 21344397.0,
      "step": 11605
    },
    {
      "entropy": 5.331781530380249,
      "epoch": 0.969762779819579,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004911915568096328,
      "loss": 5.2206,
      "mean_token_accuracy": 0.1849284753203392,
      "num_tokens": 21354111.0,
      "step": 11610
    },
    {
      "entropy": 5.214372158050537,
      "epoch": 0.9701804209822921,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004911833067176666,
      "loss": 5.0248,
      "mean_token_accuracy": 0.19430746287107467,
      "num_tokens": 21362807.0,
      "step": 11615
    },
    {
      "entropy": 5.325587177276612,
      "epoch": 0.970598062145005,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004911750528411515,
      "loss": 5.1995,
      "mean_token_accuracy": 0.1845437690615654,
      "num_tokens": 21372609.0,
      "step": 11620
    },
    {
      "entropy": 5.27841272354126,
      "epoch": 0.971015703307718,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004911667951802318,
      "loss": 5.1263,
      "mean_token_accuracy": 0.18883326947689055,
      "num_tokens": 21381182.0,
      "step": 11625
    },
    {
      "entropy": 5.225793695449829,
      "epoch": 0.9714333444704311,
      "grad_norm": 1.375,
      "learning_rate": 0.0004911585337350524,
      "loss": 5.1286,
      "mean_token_accuracy": 0.196099953353405,
      "num_tokens": 21389930.0,
      "step": 11630
    },
    {
      "entropy": 5.333341789245606,
      "epoch": 0.971850985633144,
      "grad_norm": 1.5,
      "learning_rate": 0.0004911502685057573,
      "loss": 5.1792,
      "mean_token_accuracy": 0.18482903987169266,
      "num_tokens": 21399315.0,
      "step": 11635
    },
    {
      "entropy": 5.329329252243042,
      "epoch": 0.972268626795857,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004911419994924918,
      "loss": 5.1831,
      "mean_token_accuracy": 0.18919351249933242,
      "num_tokens": 21408857.0,
      "step": 11640
    },
    {
      "entropy": 5.29211835861206,
      "epoch": 0.97268626795857,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004911337266954004,
      "loss": 5.1092,
      "mean_token_accuracy": 0.19436143785715104,
      "num_tokens": 21417452.0,
      "step": 11645
    },
    {
      "entropy": 5.260520315170288,
      "epoch": 0.973103909121283,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004911254501146279,
      "loss": 5.1564,
      "mean_token_accuracy": 0.19158775210380555,
      "num_tokens": 21427315.0,
      "step": 11650
    },
    {
      "entropy": 5.3140082359313965,
      "epoch": 0.973521550283996,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004911171697503193,
      "loss": 5.1919,
      "mean_token_accuracy": 0.1894884452223778,
      "num_tokens": 21437080.0,
      "step": 11655
    },
    {
      "entropy": 5.281603193283081,
      "epoch": 0.973939191446709,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004911088856026194,
      "loss": 5.0699,
      "mean_token_accuracy": 0.2041848510503769,
      "num_tokens": 21445421.0,
      "step": 11660
    },
    {
      "entropy": 5.306150197982788,
      "epoch": 0.974356832609422,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004911005976716734,
      "loss": 5.0709,
      "mean_token_accuracy": 0.19587866216897964,
      "num_tokens": 21454684.0,
      "step": 11665
    },
    {
      "entropy": 5.159938526153565,
      "epoch": 0.974774473772135,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004910923059576264,
      "loss": 5.0385,
      "mean_token_accuracy": 0.19672162532806398,
      "num_tokens": 21464287.0,
      "step": 11670
    },
    {
      "entropy": 5.216953611373901,
      "epoch": 0.975192114934848,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004910840104606234,
      "loss": 5.1675,
      "mean_token_accuracy": 0.1888881504535675,
      "num_tokens": 21472835.0,
      "step": 11675
    },
    {
      "entropy": 5.312102174758911,
      "epoch": 0.975609756097561,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004910757111808097,
      "loss": 5.063,
      "mean_token_accuracy": 0.19668349921703338,
      "num_tokens": 21482327.0,
      "step": 11680
    },
    {
      "entropy": 5.332135915756226,
      "epoch": 0.976027397260274,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004910674081183305,
      "loss": 5.1794,
      "mean_token_accuracy": 0.1831653743982315,
      "num_tokens": 21492257.0,
      "step": 11685
    },
    {
      "entropy": 5.208573675155639,
      "epoch": 0.976445038422987,
      "grad_norm": 2.421875,
      "learning_rate": 0.0004910591012733313,
      "loss": 5.1601,
      "mean_token_accuracy": 0.18686803430318832,
      "num_tokens": 21501930.0,
      "step": 11690
    },
    {
      "entropy": 5.351539659500122,
      "epoch": 0.9768626795857,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004910507906459575,
      "loss": 5.1643,
      "mean_token_accuracy": 0.18271103650331497,
      "num_tokens": 21511924.0,
      "step": 11695
    },
    {
      "entropy": 5.38795576095581,
      "epoch": 0.977280320748413,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004910424762363545,
      "loss": 5.1788,
      "mean_token_accuracy": 0.18678227066993713,
      "num_tokens": 21522279.0,
      "step": 11700
    },
    {
      "entropy": 5.247772216796875,
      "epoch": 0.977697961911126,
      "grad_norm": 1.9609375,
      "learning_rate": 0.000491034158044668,
      "loss": 5.1605,
      "mean_token_accuracy": 0.18947492092847823,
      "num_tokens": 21532016.0,
      "step": 11705
    },
    {
      "entropy": 5.293446969985962,
      "epoch": 0.9781156030738389,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004910258360710434,
      "loss": 5.2093,
      "mean_token_accuracy": 0.18152908682823182,
      "num_tokens": 21541952.0,
      "step": 11710
    },
    {
      "entropy": 5.299660634994507,
      "epoch": 0.978533244236552,
      "grad_norm": 2.203125,
      "learning_rate": 0.0004910175103156266,
      "loss": 5.0346,
      "mean_token_accuracy": 0.20440198630094528,
      "num_tokens": 21550185.0,
      "step": 11715
    },
    {
      "entropy": 5.325966167449951,
      "epoch": 0.978950885399265,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004910091807785632,
      "loss": 5.1981,
      "mean_token_accuracy": 0.19157318621873856,
      "num_tokens": 21558891.0,
      "step": 11720
    },
    {
      "entropy": 5.286233568191529,
      "epoch": 0.9793685265619779,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004910008474599991,
      "loss": 5.1618,
      "mean_token_accuracy": 0.1881268322467804,
      "num_tokens": 21568808.0,
      "step": 11725
    },
    {
      "entropy": 5.3441784381866455,
      "epoch": 0.979786167724691,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004909925103600801,
      "loss": 5.2066,
      "mean_token_accuracy": 0.182724791765213,
      "num_tokens": 21578838.0,
      "step": 11730
    },
    {
      "entropy": 5.234168863296508,
      "epoch": 0.980203808887404,
      "grad_norm": 1.375,
      "learning_rate": 0.0004909841694789523,
      "loss": 5.0788,
      "mean_token_accuracy": 0.19667238146066665,
      "num_tokens": 21588008.0,
      "step": 11735
    },
    {
      "entropy": 5.287701797485352,
      "epoch": 0.9806214500501169,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004909758248167616,
      "loss": 5.1934,
      "mean_token_accuracy": 0.18925960808992387,
      "num_tokens": 21597036.0,
      "step": 11740
    },
    {
      "entropy": 5.349890089035034,
      "epoch": 0.98103909121283,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004909674763736542,
      "loss": 5.1596,
      "mean_token_accuracy": 0.18791257888078688,
      "num_tokens": 21606028.0,
      "step": 11745
    },
    {
      "entropy": 5.362050580978393,
      "epoch": 0.981456732375543,
      "grad_norm": 2.5,
      "learning_rate": 0.0004909591241497761,
      "loss": 5.2329,
      "mean_token_accuracy": 0.19296300411224365,
      "num_tokens": 21615377.0,
      "step": 11750
    },
    {
      "entropy": 5.435769844055176,
      "epoch": 0.9818743735382559,
      "grad_norm": 2.234375,
      "learning_rate": 0.0004909507681452736,
      "loss": 5.3028,
      "mean_token_accuracy": 0.17882558107376098,
      "num_tokens": 21625751.0,
      "step": 11755
    },
    {
      "entropy": 5.416766548156739,
      "epoch": 0.9822920147009689,
      "grad_norm": 2.1875,
      "learning_rate": 0.0004909424083602929,
      "loss": 5.2494,
      "mean_token_accuracy": 0.18734355568885802,
      "num_tokens": 21634825.0,
      "step": 11760
    },
    {
      "entropy": 5.399048328399658,
      "epoch": 0.982709655863682,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004909340447949805,
      "loss": 5.0992,
      "mean_token_accuracy": 0.19860923737287522,
      "num_tokens": 21644286.0,
      "step": 11765
    },
    {
      "entropy": 5.369190263748169,
      "epoch": 0.9831272970263949,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004909256774494827,
      "loss": 5.1945,
      "mean_token_accuracy": 0.19318927228450775,
      "num_tokens": 21653384.0,
      "step": 11770
    },
    {
      "entropy": 5.287894582748413,
      "epoch": 0.9835449381891079,
      "grad_norm": 1.8203125,
      "learning_rate": 0.000490917306323946,
      "loss": 5.1139,
      "mean_token_accuracy": 0.19105462729930878,
      "num_tokens": 21662126.0,
      "step": 11775
    },
    {
      "entropy": 5.267673778533935,
      "epoch": 0.983962579351821,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004909089314185169,
      "loss": 5.1336,
      "mean_token_accuracy": 0.18622952252626418,
      "num_tokens": 21671940.0,
      "step": 11780
    },
    {
      "entropy": 5.209247636795044,
      "epoch": 0.9843802205145339,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004909005527333421,
      "loss": 5.0404,
      "mean_token_accuracy": 0.1980053663253784,
      "num_tokens": 21680928.0,
      "step": 11785
    },
    {
      "entropy": 5.254646968841553,
      "epoch": 0.9847978616772469,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004908921702685683,
      "loss": 5.1197,
      "mean_token_accuracy": 0.18791358321905136,
      "num_tokens": 21689712.0,
      "step": 11790
    },
    {
      "entropy": 5.310435771942139,
      "epoch": 0.98521550283996,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004908837840243421,
      "loss": 5.0987,
      "mean_token_accuracy": 0.19554044753313066,
      "num_tokens": 21698325.0,
      "step": 11795
    },
    {
      "entropy": 5.2651325225830075,
      "epoch": 0.9856331440026729,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004908753940008105,
      "loss": 5.0731,
      "mean_token_accuracy": 0.19559344202280043,
      "num_tokens": 21707607.0,
      "step": 11800
    },
    {
      "entropy": 5.1734973907470705,
      "epoch": 0.9860507851653859,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004908670001981201,
      "loss": 5.0449,
      "mean_token_accuracy": 0.19553141593933104,
      "num_tokens": 21715887.0,
      "step": 11805
    },
    {
      "entropy": 5.235254240036011,
      "epoch": 0.986468426328099,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004908586026164182,
      "loss": 5.1716,
      "mean_token_accuracy": 0.19380661100149155,
      "num_tokens": 21724992.0,
      "step": 11810
    },
    {
      "entropy": 5.3405735969543455,
      "epoch": 0.9868860674908119,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004908502012558516,
      "loss": 5.1665,
      "mean_token_accuracy": 0.18751619160175323,
      "num_tokens": 21734317.0,
      "step": 11815
    },
    {
      "entropy": 5.3303934097290036,
      "epoch": 0.9873037086535249,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004908417961165674,
      "loss": 5.2115,
      "mean_token_accuracy": 0.18783560991287232,
      "num_tokens": 21743014.0,
      "step": 11820
    },
    {
      "entropy": 5.33513298034668,
      "epoch": 0.9877213498162379,
      "grad_norm": 1.5,
      "learning_rate": 0.0004908333871987126,
      "loss": 5.1203,
      "mean_token_accuracy": 0.19073102325201036,
      "num_tokens": 21751058.0,
      "step": 11825
    },
    {
      "entropy": 5.371034145355225,
      "epoch": 0.9881389909789509,
      "grad_norm": 1.5,
      "learning_rate": 0.0004908249745024347,
      "loss": 5.219,
      "mean_token_accuracy": 0.18869234770536422,
      "num_tokens": 21760549.0,
      "step": 11830
    },
    {
      "entropy": 5.293810033798218,
      "epoch": 0.9885566321416639,
      "grad_norm": 1.7109375,
      "learning_rate": 0.000490816558027881,
      "loss": 5.14,
      "mean_token_accuracy": 0.1943634942173958,
      "num_tokens": 21769873.0,
      "step": 11835
    },
    {
      "entropy": 5.346495151519775,
      "epoch": 0.9889742733043769,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004908081377751985,
      "loss": 5.2217,
      "mean_token_accuracy": 0.19356146156787873,
      "num_tokens": 21779036.0,
      "step": 11840
    },
    {
      "entropy": 5.422462463378906,
      "epoch": 0.9893919144670899,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004907997137445347,
      "loss": 5.2363,
      "mean_token_accuracy": 0.18284029066562651,
      "num_tokens": 21787671.0,
      "step": 11845
    },
    {
      "entropy": 5.393016958236695,
      "epoch": 0.9898095556298029,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004907912859360373,
      "loss": 5.1843,
      "mean_token_accuracy": 0.1853944331407547,
      "num_tokens": 21797269.0,
      "step": 11850
    },
    {
      "entropy": 5.389364719390869,
      "epoch": 0.9902271967925159,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004907828543498536,
      "loss": 5.247,
      "mean_token_accuracy": 0.18878703266382219,
      "num_tokens": 21806738.0,
      "step": 11855
    },
    {
      "entropy": 5.315618324279785,
      "epoch": 0.9906448379552288,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004907744189861313,
      "loss": 5.152,
      "mean_token_accuracy": 0.1888543501496315,
      "num_tokens": 21815612.0,
      "step": 11860
    },
    {
      "entropy": 5.297834730148315,
      "epoch": 0.9910624791179419,
      "grad_norm": 1.6796875,
      "learning_rate": 0.000490765979845018,
      "loss": 5.1603,
      "mean_token_accuracy": 0.19539257884025574,
      "num_tokens": 21824222.0,
      "step": 11865
    },
    {
      "entropy": 5.372282075881958,
      "epoch": 0.9914801202806549,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004907575369266617,
      "loss": 5.1616,
      "mean_token_accuracy": 0.19222169518470764,
      "num_tokens": 21833929.0,
      "step": 11870
    },
    {
      "entropy": 5.306639719009399,
      "epoch": 0.9918977614433678,
      "grad_norm": 2.3125,
      "learning_rate": 0.0004907490902312099,
      "loss": 5.1967,
      "mean_token_accuracy": 0.18780184388160706,
      "num_tokens": 21843261.0,
      "step": 11875
    },
    {
      "entropy": 5.365342092514038,
      "epoch": 0.9923154026060809,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004907406397588106,
      "loss": 5.2507,
      "mean_token_accuracy": 0.18705663979053497,
      "num_tokens": 21853004.0,
      "step": 11880
    },
    {
      "entropy": 5.315323257446289,
      "epoch": 0.9927330437687939,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004907321855096117,
      "loss": 5.0994,
      "mean_token_accuracy": 0.18834714442491532,
      "num_tokens": 21862628.0,
      "step": 11885
    },
    {
      "entropy": 5.276562643051148,
      "epoch": 0.9931506849315068,
      "grad_norm": 2.59375,
      "learning_rate": 0.0004907237274837613,
      "loss": 5.0889,
      "mean_token_accuracy": 0.19038716852664947,
      "num_tokens": 21872041.0,
      "step": 11890
    },
    {
      "entropy": 5.309222984313965,
      "epoch": 0.9935683260942199,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004907152656814072,
      "loss": 5.1846,
      "mean_token_accuracy": 0.18281338661909102,
      "num_tokens": 21881834.0,
      "step": 11895
    },
    {
      "entropy": 5.4300117015838625,
      "epoch": 0.9939859672569329,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004907068001026978,
      "loss": 5.2902,
      "mean_token_accuracy": 0.18365519046783446,
      "num_tokens": 21892349.0,
      "step": 11900
    },
    {
      "entropy": 5.383771467208862,
      "epoch": 0.9944036084196458,
      "grad_norm": 1.75,
      "learning_rate": 0.0004906983307477813,
      "loss": 5.1386,
      "mean_token_accuracy": 0.19515896290540696,
      "num_tokens": 21901427.0,
      "step": 11905
    },
    {
      "entropy": 5.267907905578613,
      "epoch": 0.9948212495823588,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004906898576168057,
      "loss": 5.0808,
      "mean_token_accuracy": 0.19798390865325927,
      "num_tokens": 21910654.0,
      "step": 11910
    },
    {
      "entropy": 5.219382905960083,
      "epoch": 0.9952388907450719,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004906813807099197,
      "loss": 5.1081,
      "mean_token_accuracy": 0.20003349632024764,
      "num_tokens": 21919583.0,
      "step": 11915
    },
    {
      "entropy": 5.2093384742736815,
      "epoch": 0.9956565319077848,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004906729000272714,
      "loss": 5.0318,
      "mean_token_accuracy": 0.1949403002858162,
      "num_tokens": 21927843.0,
      "step": 11920
    },
    {
      "entropy": 5.280549144744873,
      "epoch": 0.9960741730704978,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004906644155690095,
      "loss": 5.1152,
      "mean_token_accuracy": 0.19122098833322526,
      "num_tokens": 21937269.0,
      "step": 11925
    },
    {
      "entropy": 5.496305274963379,
      "epoch": 0.9964918142332109,
      "grad_norm": 2.625,
      "learning_rate": 0.0004906559273352823,
      "loss": 5.2893,
      "mean_token_accuracy": 0.18277825117111207,
      "num_tokens": 21946965.0,
      "step": 11930
    },
    {
      "entropy": 5.353962135314942,
      "epoch": 0.9969094553959238,
      "grad_norm": 1.5,
      "learning_rate": 0.0004906474353262386,
      "loss": 5.1074,
      "mean_token_accuracy": 0.18749611526727678,
      "num_tokens": 21955957.0,
      "step": 11935
    },
    {
      "entropy": 5.250110530853272,
      "epoch": 0.9973270965586368,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004906389395420269,
      "loss": 5.0643,
      "mean_token_accuracy": 0.19533012956380844,
      "num_tokens": 21965643.0,
      "step": 11940
    },
    {
      "entropy": 5.307873106002807,
      "epoch": 0.9977447377213499,
      "grad_norm": 1.7734375,
      "learning_rate": 0.000490630439982796,
      "loss": 5.1395,
      "mean_token_accuracy": 0.19904470890760423,
      "num_tokens": 21974789.0,
      "step": 11945
    },
    {
      "entropy": 5.25768461227417,
      "epoch": 0.9981623788840628,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004906219366486947,
      "loss": 5.0335,
      "mean_token_accuracy": 0.20398582220077516,
      "num_tokens": 21983411.0,
      "step": 11950
    },
    {
      "entropy": 5.322871732711792,
      "epoch": 0.9985800200467758,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004906134295398719,
      "loss": 5.1427,
      "mean_token_accuracy": 0.1877253919839859,
      "num_tokens": 21992570.0,
      "step": 11955
    },
    {
      "entropy": 5.268395280838012,
      "epoch": 0.9989976612094889,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004906049186564764,
      "loss": 5.0122,
      "mean_token_accuracy": 0.20317024737596512,
      "num_tokens": 22001411.0,
      "step": 11960
    },
    {
      "entropy": 5.4146442890167235,
      "epoch": 0.9994153023722018,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004905964039986574,
      "loss": 5.2699,
      "mean_token_accuracy": 0.1825779139995575,
      "num_tokens": 22010882.0,
      "step": 11965
    },
    {
      "entropy": 5.368969488143921,
      "epoch": 0.9998329435349148,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004905878855665637,
      "loss": 5.2377,
      "mean_token_accuracy": 0.18160374611616134,
      "num_tokens": 22020885.0,
      "step": 11970
    },
    {
      "entropy": 5.37403244972229,
      "epoch": 1.0002505846976277,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004905793633603446,
      "loss": 4.9856,
      "mean_token_accuracy": 0.19897338449954988,
      "num_tokens": 22030278.0,
      "step": 11975
    },
    {
      "entropy": 5.339780664443969,
      "epoch": 1.0006682258603408,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004905708373801492,
      "loss": 5.1908,
      "mean_token_accuracy": 0.18481556475162506,
      "num_tokens": 22039601.0,
      "step": 11980
    },
    {
      "entropy": 5.226790142059326,
      "epoch": 1.0010858670230538,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004905623076261269,
      "loss": 5.0162,
      "mean_token_accuracy": 0.19655622839927672,
      "num_tokens": 22049158.0,
      "step": 11985
    },
    {
      "entropy": 5.328135681152344,
      "epoch": 1.0015035081857668,
      "grad_norm": 1.53125,
      "learning_rate": 0.000490553774098427,
      "loss": 5.1471,
      "mean_token_accuracy": 0.18671782016754152,
      "num_tokens": 22059278.0,
      "step": 11990
    },
    {
      "entropy": 5.360319375991821,
      "epoch": 1.0019211493484799,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004905452367971986,
      "loss": 5.1016,
      "mean_token_accuracy": 0.1948260560631752,
      "num_tokens": 22068415.0,
      "step": 11995
    },
    {
      "entropy": 5.3181774616241455,
      "epoch": 1.0023387905111927,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004905366957225916,
      "loss": 5.0707,
      "mean_token_accuracy": 0.19402292519807815,
      "num_tokens": 22077815.0,
      "step": 12000
    },
    {
      "epoch": 1.0023387905111927,
      "eval_entropy": 5.0716266818431395,
      "eval_loss": 5.101118087768555,
      "eval_mean_token_accuracy": 0.203080904427612,
      "eval_num_tokens": 22077815.0,
      "eval_runtime": 27.2587,
      "eval_samples_per_second": 1385.357,
      "eval_steps_per_second": 173.193,
      "step": 12000
    },
    {
      "entropy": 5.3564841747283936,
      "epoch": 1.0027564316739057,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004905281508747552,
      "loss": 5.1376,
      "mean_token_accuracy": 0.18673287332057953,
      "num_tokens": 22087606.0,
      "step": 12005
    },
    {
      "entropy": 5.2614648818969725,
      "epoch": 1.0031740728366187,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004905196022538391,
      "loss": 4.979,
      "mean_token_accuracy": 0.19654346108436585,
      "num_tokens": 22096300.0,
      "step": 12010
    },
    {
      "entropy": 5.364877796173095,
      "epoch": 1.0035917139993318,
      "grad_norm": 1.5,
      "learning_rate": 0.000490511049859993,
      "loss": 5.1873,
      "mean_token_accuracy": 0.1917300522327423,
      "num_tokens": 22105750.0,
      "step": 12015
    },
    {
      "entropy": 5.348353290557862,
      "epoch": 1.0040093551620448,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004905024936933664,
      "loss": 5.1653,
      "mean_token_accuracy": 0.18933407813310624,
      "num_tokens": 22115163.0,
      "step": 12020
    },
    {
      "entropy": 5.365058279037475,
      "epoch": 1.0044269963247578,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004904939337541093,
      "loss": 5.1537,
      "mean_token_accuracy": 0.18768314272165298,
      "num_tokens": 22125580.0,
      "step": 12025
    },
    {
      "entropy": 5.2469371318817135,
      "epoch": 1.0048446374874707,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004904853700423716,
      "loss": 5.0365,
      "mean_token_accuracy": 0.19901788532733916,
      "num_tokens": 22134985.0,
      "step": 12030
    },
    {
      "entropy": 5.319075870513916,
      "epoch": 1.0052622786501837,
      "grad_norm": 2.171875,
      "learning_rate": 0.0004904768025583031,
      "loss": 5.0153,
      "mean_token_accuracy": 0.19664503782987594,
      "num_tokens": 22142530.0,
      "step": 12035
    },
    {
      "entropy": 5.3517999172210695,
      "epoch": 1.0056799198128967,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004904682313020537,
      "loss": 5.1159,
      "mean_token_accuracy": 0.1914706364274025,
      "num_tokens": 22151696.0,
      "step": 12040
    },
    {
      "entropy": 5.279608488082886,
      "epoch": 1.0060975609756098,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004904596562737737,
      "loss": 5.0536,
      "mean_token_accuracy": 0.19465957581996918,
      "num_tokens": 22161286.0,
      "step": 12045
    },
    {
      "entropy": 5.2205061435699465,
      "epoch": 1.0065152021383228,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000490451077473613,
      "loss": 5.0118,
      "mean_token_accuracy": 0.20078013390302657,
      "num_tokens": 22169678.0,
      "step": 12050
    },
    {
      "entropy": 5.32427077293396,
      "epoch": 1.0069328433010358,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004904424949017218,
      "loss": 5.0789,
      "mean_token_accuracy": 0.20604410767555237,
      "num_tokens": 22178618.0,
      "step": 12055
    },
    {
      "entropy": 5.24556303024292,
      "epoch": 1.0073504844637486,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004904339085582504,
      "loss": 5.0445,
      "mean_token_accuracy": 0.19719454199075698,
      "num_tokens": 22187574.0,
      "step": 12060
    },
    {
      "entropy": 5.292127656936645,
      "epoch": 1.0077681256264617,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004904253184433491,
      "loss": 5.2186,
      "mean_token_accuracy": 0.1820813775062561,
      "num_tokens": 22198840.0,
      "step": 12065
    },
    {
      "entropy": 5.310047197341919,
      "epoch": 1.0081857667891747,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004904167245571684,
      "loss": 5.0125,
      "mean_token_accuracy": 0.2010982885956764,
      "num_tokens": 22207166.0,
      "step": 12070
    },
    {
      "entropy": 5.256866455078125,
      "epoch": 1.0086034079518877,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004904081268998586,
      "loss": 5.0248,
      "mean_token_accuracy": 0.1909558966755867,
      "num_tokens": 22216232.0,
      "step": 12075
    },
    {
      "entropy": 5.287031745910644,
      "epoch": 1.0090210491146008,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004903995254715702,
      "loss": 5.0256,
      "mean_token_accuracy": 0.1938977286219597,
      "num_tokens": 22224853.0,
      "step": 12080
    },
    {
      "entropy": 5.251332473754883,
      "epoch": 1.0094386902773138,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004903909202724539,
      "loss": 5.0097,
      "mean_token_accuracy": 0.1959120512008667,
      "num_tokens": 22233971.0,
      "step": 12085
    },
    {
      "entropy": 5.226678371429443,
      "epoch": 1.0098563314400266,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004903823113026602,
      "loss": 4.9231,
      "mean_token_accuracy": 0.2032647103071213,
      "num_tokens": 22242548.0,
      "step": 12090
    },
    {
      "entropy": 5.144503164291382,
      "epoch": 1.0102739726027397,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00049037369856234,
      "loss": 4.8977,
      "mean_token_accuracy": 0.20532550811767578,
      "num_tokens": 22251086.0,
      "step": 12095
    },
    {
      "entropy": 5.2448217391967775,
      "epoch": 1.0106916137654527,
      "grad_norm": 1.5,
      "learning_rate": 0.0004903650820516438,
      "loss": 5.0991,
      "mean_token_accuracy": 0.1895515277981758,
      "num_tokens": 22260035.0,
      "step": 12100
    },
    {
      "entropy": 5.316429376602173,
      "epoch": 1.0111092549281657,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004903564617707227,
      "loss": 5.0185,
      "mean_token_accuracy": 0.1963577851653099,
      "num_tokens": 22269134.0,
      "step": 12105
    },
    {
      "entropy": 5.299637699127198,
      "epoch": 1.0115268960908788,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004903478377197275,
      "loss": 5.1079,
      "mean_token_accuracy": 0.1995877131819725,
      "num_tokens": 22278123.0,
      "step": 12110
    },
    {
      "entropy": 5.2202794551849365,
      "epoch": 1.0119445372535918,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004903392098988092,
      "loss": 5.0995,
      "mean_token_accuracy": 0.18974104672670364,
      "num_tokens": 22287557.0,
      "step": 12115
    },
    {
      "entropy": 5.25773434638977,
      "epoch": 1.0123621784163046,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004903305783081188,
      "loss": 5.0376,
      "mean_token_accuracy": 0.19454010874032973,
      "num_tokens": 22296705.0,
      "step": 12120
    },
    {
      "entropy": 5.252296733856201,
      "epoch": 1.0127798195790176,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004903219429478075,
      "loss": 5.0071,
      "mean_token_accuracy": 0.19256707578897475,
      "num_tokens": 22305904.0,
      "step": 12125
    },
    {
      "entropy": 5.2542510509490965,
      "epoch": 1.0131974607417307,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004903133038180263,
      "loss": 5.0416,
      "mean_token_accuracy": 0.19820158332586288,
      "num_tokens": 22315051.0,
      "step": 12130
    },
    {
      "entropy": 5.245576667785644,
      "epoch": 1.0136151019044437,
      "grad_norm": 2.1875,
      "learning_rate": 0.0004903046609189266,
      "loss": 5.1062,
      "mean_token_accuracy": 0.19255658537149428,
      "num_tokens": 22324469.0,
      "step": 12135
    },
    {
      "entropy": 5.38571572303772,
      "epoch": 1.0140327430671567,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004902960142506597,
      "loss": 5.1577,
      "mean_token_accuracy": 0.18887430280447007,
      "num_tokens": 22334407.0,
      "step": 12140
    },
    {
      "entropy": 5.36635046005249,
      "epoch": 1.0144503842298698,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004902873638133768,
      "loss": 5.1754,
      "mean_token_accuracy": 0.19001132398843765,
      "num_tokens": 22342975.0,
      "step": 12145
    },
    {
      "entropy": 5.248002672195435,
      "epoch": 1.0148680253925826,
      "grad_norm": 2.0,
      "learning_rate": 0.0004902787096072296,
      "loss": 4.9716,
      "mean_token_accuracy": 0.19113945811986924,
      "num_tokens": 22351649.0,
      "step": 12150
    },
    {
      "entropy": 5.2755434036254885,
      "epoch": 1.0152856665552956,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004902700516323693,
      "loss": 5.0716,
      "mean_token_accuracy": 0.19035932272672654,
      "num_tokens": 22361956.0,
      "step": 12155
    },
    {
      "entropy": 5.305901145935058,
      "epoch": 1.0157033077180087,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004902613898889477,
      "loss": 5.0773,
      "mean_token_accuracy": 0.18565924912691117,
      "num_tokens": 22372354.0,
      "step": 12160
    },
    {
      "entropy": 5.237570762634277,
      "epoch": 1.0161209488807217,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004902527243771164,
      "loss": 4.9574,
      "mean_token_accuracy": 0.20020891427993776,
      "num_tokens": 22381225.0,
      "step": 12165
    },
    {
      "entropy": 5.190505027770996,
      "epoch": 1.0165385900434347,
      "grad_norm": 2.296875,
      "learning_rate": 0.000490244055097027,
      "loss": 4.9804,
      "mean_token_accuracy": 0.2005234807729721,
      "num_tokens": 22390666.0,
      "step": 12170
    },
    {
      "entropy": 5.236529922485351,
      "epoch": 1.0169562312061478,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004902353820488314,
      "loss": 4.9703,
      "mean_token_accuracy": 0.20961204320192336,
      "num_tokens": 22399528.0,
      "step": 12175
    },
    {
      "entropy": 5.230377674102783,
      "epoch": 1.0173738723688606,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004902267052326812,
      "loss": 5.0057,
      "mean_token_accuracy": 0.20272872000932693,
      "num_tokens": 22408827.0,
      "step": 12180
    },
    {
      "entropy": 5.196474838256836,
      "epoch": 1.0177915135315736,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004902180246487285,
      "loss": 5.0181,
      "mean_token_accuracy": 0.19436388164758683,
      "num_tokens": 22418098.0,
      "step": 12185
    },
    {
      "entropy": 5.118693399429321,
      "epoch": 1.0182091546942866,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004902093402971254,
      "loss": 4.9027,
      "mean_token_accuracy": 0.20587823539972305,
      "num_tokens": 22427417.0,
      "step": 12190
    },
    {
      "entropy": 5.266181230545044,
      "epoch": 1.0186267958569997,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004902006521780236,
      "loss": 5.0871,
      "mean_token_accuracy": 0.1938799202442169,
      "num_tokens": 22436938.0,
      "step": 12195
    },
    {
      "entropy": 5.302451229095459,
      "epoch": 1.0190444370197127,
      "grad_norm": 1.75,
      "learning_rate": 0.0004901919602915754,
      "loss": 5.1309,
      "mean_token_accuracy": 0.18617192208766936,
      "num_tokens": 22447361.0,
      "step": 12200
    },
    {
      "entropy": 5.363397836685181,
      "epoch": 1.0194620781824257,
      "grad_norm": 2.484375,
      "learning_rate": 0.0004901832646379328,
      "loss": 5.1512,
      "mean_token_accuracy": 0.19016380310058595,
      "num_tokens": 22457431.0,
      "step": 12205
    },
    {
      "entropy": 5.348782062530518,
      "epoch": 1.0198797193451385,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004901745652172483,
      "loss": 5.1273,
      "mean_token_accuracy": 0.19151293337345124,
      "num_tokens": 22466219.0,
      "step": 12210
    },
    {
      "entropy": 5.325495958328247,
      "epoch": 1.0202973605078516,
      "grad_norm": 1.5625,
      "learning_rate": 0.000490165862029674,
      "loss": 5.1626,
      "mean_token_accuracy": 0.19051584154367446,
      "num_tokens": 22476231.0,
      "step": 12215
    },
    {
      "entropy": 5.2457421779632565,
      "epoch": 1.0207150016705646,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004901571550753622,
      "loss": 4.9422,
      "mean_token_accuracy": 0.204918797314167,
      "num_tokens": 22484288.0,
      "step": 12220
    },
    {
      "entropy": 5.2824334621429445,
      "epoch": 1.0211326428332776,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004901484443544655,
      "loss": 5.0931,
      "mean_token_accuracy": 0.19519853442907334,
      "num_tokens": 22493619.0,
      "step": 12225
    },
    {
      "entropy": 5.17180609703064,
      "epoch": 1.0215502839959907,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004901397298671362,
      "loss": 4.9781,
      "mean_token_accuracy": 0.2009566158056259,
      "num_tokens": 22501606.0,
      "step": 12230
    },
    {
      "entropy": 5.245876407623291,
      "epoch": 1.0219679251587037,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004901310116135271,
      "loss": 4.989,
      "mean_token_accuracy": 0.20166365504264833,
      "num_tokens": 22510410.0,
      "step": 12235
    },
    {
      "entropy": 5.295335149765014,
      "epoch": 1.0223855663214167,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004901222895937905,
      "loss": 5.1092,
      "mean_token_accuracy": 0.19832412898540497,
      "num_tokens": 22518796.0,
      "step": 12240
    },
    {
      "entropy": 5.256699323654175,
      "epoch": 1.0228032074841296,
      "grad_norm": 2.234375,
      "learning_rate": 0.0004901135638080796,
      "loss": 4.9306,
      "mean_token_accuracy": 0.19821864366531372,
      "num_tokens": 22527532.0,
      "step": 12245
    },
    {
      "entropy": 5.230990982055664,
      "epoch": 1.0232208486468426,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004901048342565465,
      "loss": 5.0167,
      "mean_token_accuracy": 0.19369079172611237,
      "num_tokens": 22536882.0,
      "step": 12250
    },
    {
      "entropy": 5.19123477935791,
      "epoch": 1.0236384898095556,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004900961009393445,
      "loss": 4.9332,
      "mean_token_accuracy": 0.19752179682254792,
      "num_tokens": 22545549.0,
      "step": 12255
    },
    {
      "entropy": 5.291220664978027,
      "epoch": 1.0240561309722687,
      "grad_norm": 2.28125,
      "learning_rate": 0.0004900873638566263,
      "loss": 5.1583,
      "mean_token_accuracy": 0.18746659457683562,
      "num_tokens": 22554941.0,
      "step": 12260
    },
    {
      "entropy": 5.19574875831604,
      "epoch": 1.0244737721349817,
      "grad_norm": 1.6328125,
      "learning_rate": 0.000490078623008545,
      "loss": 5.0484,
      "mean_token_accuracy": 0.1940171927213669,
      "num_tokens": 22565159.0,
      "step": 12265
    },
    {
      "entropy": 5.196165275573731,
      "epoch": 1.0248914132976945,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004900698783952533,
      "loss": 4.981,
      "mean_token_accuracy": 0.19427304565906525,
      "num_tokens": 22574184.0,
      "step": 12270
    },
    {
      "entropy": 5.241775274276733,
      "epoch": 1.0253090544604075,
      "grad_norm": 2.578125,
      "learning_rate": 0.0004900611300169048,
      "loss": 5.1111,
      "mean_token_accuracy": 0.19012198746204376,
      "num_tokens": 22584062.0,
      "step": 12275
    },
    {
      "entropy": 5.249707841873169,
      "epoch": 1.0257266956231206,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004900523778736521,
      "loss": 5.0373,
      "mean_token_accuracy": 0.19855550974607467,
      "num_tokens": 22592845.0,
      "step": 12280
    },
    {
      "entropy": 5.314531898498535,
      "epoch": 1.0261443367858336,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004900436219656488,
      "loss": 5.0716,
      "mean_token_accuracy": 0.19319814294576645,
      "num_tokens": 22602750.0,
      "step": 12285
    },
    {
      "entropy": 5.274512577056885,
      "epoch": 1.0265619779485466,
      "grad_norm": 1.6015625,
      "learning_rate": 0.000490034862293048,
      "loss": 5.0781,
      "mean_token_accuracy": 0.18839733451604843,
      "num_tokens": 22611788.0,
      "step": 12290
    },
    {
      "entropy": 5.211328792572021,
      "epoch": 1.0269796191112597,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004900260988560031,
      "loss": 4.9816,
      "mean_token_accuracy": 0.19597660899162292,
      "num_tokens": 22620234.0,
      "step": 12295
    },
    {
      "entropy": 5.137051153182983,
      "epoch": 1.0273972602739727,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004900173316546675,
      "loss": 4.9353,
      "mean_token_accuracy": 0.20167066901922226,
      "num_tokens": 22629155.0,
      "step": 12300
    },
    {
      "entropy": 5.311109161376953,
      "epoch": 1.0278149014366855,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004900085606891947,
      "loss": 5.1488,
      "mean_token_accuracy": 0.1848938912153244,
      "num_tokens": 22638647.0,
      "step": 12305
    },
    {
      "entropy": 5.323642635345459,
      "epoch": 1.0282325425993986,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004899997859597382,
      "loss": 5.0611,
      "mean_token_accuracy": 0.18786784559488295,
      "num_tokens": 22648295.0,
      "step": 12310
    },
    {
      "entropy": 5.26903977394104,
      "epoch": 1.0286501837621116,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004899910074664518,
      "loss": 5.0512,
      "mean_token_accuracy": 0.19077493250370026,
      "num_tokens": 22657342.0,
      "step": 12315
    },
    {
      "entropy": 5.202337646484375,
      "epoch": 1.0290678249248246,
      "grad_norm": 5.0625,
      "learning_rate": 0.0004899822252094889,
      "loss": 4.9947,
      "mean_token_accuracy": 0.19804390519857407,
      "num_tokens": 22665230.0,
      "step": 12320
    },
    {
      "entropy": 5.202148008346557,
      "epoch": 1.0294854660875377,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004899734391890034,
      "loss": 5.0215,
      "mean_token_accuracy": 0.1964283838868141,
      "num_tokens": 22674385.0,
      "step": 12325
    },
    {
      "entropy": 5.173833751678467,
      "epoch": 1.0299031072502507,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004899646494051491,
      "loss": 4.8979,
      "mean_token_accuracy": 0.2144990473985672,
      "num_tokens": 22683493.0,
      "step": 12330
    },
    {
      "entropy": 5.252021265029907,
      "epoch": 1.0303207484129635,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004899558558580798,
      "loss": 4.9783,
      "mean_token_accuracy": 0.1992649331688881,
      "num_tokens": 22691878.0,
      "step": 12335
    },
    {
      "entropy": 5.139744234085083,
      "epoch": 1.0307383895756765,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004899470585479495,
      "loss": 4.9779,
      "mean_token_accuracy": 0.19737347066402436,
      "num_tokens": 22700112.0,
      "step": 12340
    },
    {
      "entropy": 5.223740816116333,
      "epoch": 1.0311560307383896,
      "grad_norm": 2.375,
      "learning_rate": 0.0004899382574749124,
      "loss": 5.0161,
      "mean_token_accuracy": 0.19964801669120788,
      "num_tokens": 22709218.0,
      "step": 12345
    },
    {
      "entropy": 5.300681447982788,
      "epoch": 1.0315736719011026,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004899294526391221,
      "loss": 5.0398,
      "mean_token_accuracy": 0.19755626767873763,
      "num_tokens": 22718423.0,
      "step": 12350
    },
    {
      "entropy": 5.242780160903931,
      "epoch": 1.0319913130638156,
      "grad_norm": 2.4375,
      "learning_rate": 0.0004899206440407332,
      "loss": 5.1107,
      "mean_token_accuracy": 0.18800341635942458,
      "num_tokens": 22727333.0,
      "step": 12355
    },
    {
      "entropy": 5.2333526611328125,
      "epoch": 1.0324089542265287,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004899118316798996,
      "loss": 5.0374,
      "mean_token_accuracy": 0.19742570519447328,
      "num_tokens": 22737248.0,
      "step": 12360
    },
    {
      "entropy": 5.213828611373901,
      "epoch": 1.0328265953892415,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004899030155567759,
      "loss": 5.0634,
      "mean_token_accuracy": 0.1908201813697815,
      "num_tokens": 22747515.0,
      "step": 12365
    },
    {
      "entropy": 5.314577865600586,
      "epoch": 1.0332442365519545,
      "grad_norm": 2.046875,
      "learning_rate": 0.000489894195671516,
      "loss": 5.0704,
      "mean_token_accuracy": 0.19596044570207596,
      "num_tokens": 22756670.0,
      "step": 12370
    },
    {
      "entropy": 5.3372204303741455,
      "epoch": 1.0336618777146676,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004898853720242747,
      "loss": 5.1719,
      "mean_token_accuracy": 0.1850533291697502,
      "num_tokens": 22766332.0,
      "step": 12375
    },
    {
      "entropy": 5.318588733673096,
      "epoch": 1.0340795188773806,
      "grad_norm": 2.203125,
      "learning_rate": 0.0004898765446152062,
      "loss": 5.0141,
      "mean_token_accuracy": 0.19768150001764298,
      "num_tokens": 22774999.0,
      "step": 12380
    },
    {
      "entropy": 5.2346514701843265,
      "epoch": 1.0344971600400936,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004898677134444652,
      "loss": 5.0377,
      "mean_token_accuracy": 0.1933281883597374,
      "num_tokens": 22784578.0,
      "step": 12385
    },
    {
      "entropy": 5.237772464752197,
      "epoch": 1.0349148012028067,
      "grad_norm": 2.96875,
      "learning_rate": 0.0004898588785122063,
      "loss": 5.06,
      "mean_token_accuracy": 0.19854413121938705,
      "num_tokens": 22794508.0,
      "step": 12390
    },
    {
      "entropy": 5.2888405323028564,
      "epoch": 1.0353324423655195,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004898500398185841,
      "loss": 5.1788,
      "mean_token_accuracy": 0.18963636159896852,
      "num_tokens": 22804399.0,
      "step": 12395
    },
    {
      "entropy": 5.316745328903198,
      "epoch": 1.0357500835282325,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004898411973637532,
      "loss": 5.0627,
      "mean_token_accuracy": 0.19071345627307892,
      "num_tokens": 22814361.0,
      "step": 12400
    },
    {
      "entropy": 5.233664703369141,
      "epoch": 1.0361677246909455,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004898323511478686,
      "loss": 4.9911,
      "mean_token_accuracy": 0.20224276781082154,
      "num_tokens": 22823780.0,
      "step": 12405
    },
    {
      "entropy": 5.184096479415894,
      "epoch": 1.0365853658536586,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004898235011710851,
      "loss": 4.9833,
      "mean_token_accuracy": 0.19100701212882995,
      "num_tokens": 22832575.0,
      "step": 12410
    },
    {
      "entropy": 5.158435678482055,
      "epoch": 1.0370030070163716,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004898146474335577,
      "loss": 4.8523,
      "mean_token_accuracy": 0.2075952410697937,
      "num_tokens": 22841822.0,
      "step": 12415
    },
    {
      "entropy": 5.3491050720214846,
      "epoch": 1.0374206481790846,
      "grad_norm": 2.515625,
      "learning_rate": 0.0004898057899354412,
      "loss": 5.1388,
      "mean_token_accuracy": 0.1887414962053299,
      "num_tokens": 22851835.0,
      "step": 12420
    },
    {
      "entropy": 5.2412700176239015,
      "epoch": 1.0378382893417974,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000489796928676891,
      "loss": 5.0567,
      "mean_token_accuracy": 0.1934041440486908,
      "num_tokens": 22861025.0,
      "step": 12425
    },
    {
      "entropy": 5.247186899185181,
      "epoch": 1.0382559305045105,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004897880636580618,
      "loss": 5.0563,
      "mean_token_accuracy": 0.19145672768354416,
      "num_tokens": 22869721.0,
      "step": 12430
    },
    {
      "entropy": 5.194805335998535,
      "epoch": 1.0386735716672235,
      "grad_norm": 1.515625,
      "learning_rate": 0.000489779194879109,
      "loss": 4.9564,
      "mean_token_accuracy": 0.20724873393774032,
      "num_tokens": 22878225.0,
      "step": 12435
    },
    {
      "entropy": 5.2599772930145265,
      "epoch": 1.0390912128299366,
      "grad_norm": 1.5,
      "learning_rate": 0.000489770322340188,
      "loss": 5.1821,
      "mean_token_accuracy": 0.18695750534534455,
      "num_tokens": 22888598.0,
      "step": 12440
    },
    {
      "entropy": 5.3204656600952145,
      "epoch": 1.0395088539926496,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004897614460414539,
      "loss": 5.0819,
      "mean_token_accuracy": 0.18755998760461806,
      "num_tokens": 22897326.0,
      "step": 12445
    },
    {
      "entropy": 5.31657600402832,
      "epoch": 1.0399264951553626,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004897525659830623,
      "loss": 5.0655,
      "mean_token_accuracy": 0.19301057010889053,
      "num_tokens": 22906372.0,
      "step": 12450
    },
    {
      "entropy": 5.165547275543213,
      "epoch": 1.0403441363180754,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004897436821651683,
      "loss": 4.9648,
      "mean_token_accuracy": 0.19965047836303712,
      "num_tokens": 22915987.0,
      "step": 12455
    },
    {
      "entropy": 5.2289777278900145,
      "epoch": 1.0407617774807885,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004897347945879278,
      "loss": 5.0802,
      "mean_token_accuracy": 0.19071826934814454,
      "num_tokens": 22926118.0,
      "step": 12460
    },
    {
      "entropy": 5.2511763095855715,
      "epoch": 1.0411794186435015,
      "grad_norm": 1.5,
      "learning_rate": 0.0004897259032514963,
      "loss": 5.0502,
      "mean_token_accuracy": 0.1948823869228363,
      "num_tokens": 22934813.0,
      "step": 12465
    },
    {
      "entropy": 5.305549621582031,
      "epoch": 1.0415970598062145,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004897170081560292,
      "loss": 5.1429,
      "mean_token_accuracy": 0.1903460443019867,
      "num_tokens": 22944579.0,
      "step": 12470
    },
    {
      "entropy": 5.19605393409729,
      "epoch": 1.0420147009689276,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004897081093016827,
      "loss": 4.957,
      "mean_token_accuracy": 0.20196807980537415,
      "num_tokens": 22953515.0,
      "step": 12475
    },
    {
      "entropy": 5.261943769454956,
      "epoch": 1.0424323421316406,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004896992066886121,
      "loss": 5.1221,
      "mean_token_accuracy": 0.183735953271389,
      "num_tokens": 22962599.0,
      "step": 12480
    },
    {
      "entropy": 5.274242687225342,
      "epoch": 1.0428499832943534,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004896903003169735,
      "loss": 5.1029,
      "mean_token_accuracy": 0.1980666399002075,
      "num_tokens": 22971717.0,
      "step": 12485
    },
    {
      "entropy": 5.153327941894531,
      "epoch": 1.0432676244570664,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004896813901869229,
      "loss": 4.8546,
      "mean_token_accuracy": 0.20338601171970366,
      "num_tokens": 22980587.0,
      "step": 12490
    },
    {
      "entropy": 5.21031551361084,
      "epoch": 1.0436852656197795,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004896724762986159,
      "loss": 5.0535,
      "mean_token_accuracy": 0.19984702467918397,
      "num_tokens": 22990197.0,
      "step": 12495
    },
    {
      "entropy": 5.220157718658447,
      "epoch": 1.0441029067824925,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004896635586522089,
      "loss": 5.1202,
      "mean_token_accuracy": 0.1914472311735153,
      "num_tokens": 22999670.0,
      "step": 12500
    },
    {
      "entropy": 5.311928558349609,
      "epoch": 1.0445205479452055,
      "grad_norm": 1.5,
      "learning_rate": 0.0004896546372478581,
      "loss": 5.0662,
      "mean_token_accuracy": 0.1965074732899666,
      "num_tokens": 23008125.0,
      "step": 12505
    },
    {
      "entropy": 5.1955492973327635,
      "epoch": 1.0449381891079186,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004896457120857192,
      "loss": 4.9079,
      "mean_token_accuracy": 0.20469543635845183,
      "num_tokens": 23017469.0,
      "step": 12510
    },
    {
      "entropy": 5.191381311416626,
      "epoch": 1.0453558302706314,
      "grad_norm": 1.328125,
      "learning_rate": 0.000489636783165949,
      "loss": 5.0598,
      "mean_token_accuracy": 0.19465763568878175,
      "num_tokens": 23026468.0,
      "step": 12515
    },
    {
      "entropy": 5.376666641235351,
      "epoch": 1.0457734714333444,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004896278504887033,
      "loss": 5.0929,
      "mean_token_accuracy": 0.19848301112651826,
      "num_tokens": 23034944.0,
      "step": 12520
    },
    {
      "entropy": 5.1970946311950685,
      "epoch": 1.0461911125960575,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000489618914054139,
      "loss": 4.9333,
      "mean_token_accuracy": 0.20223904252052308,
      "num_tokens": 23043834.0,
      "step": 12525
    },
    {
      "entropy": 5.194126605987549,
      "epoch": 1.0466087537587705,
      "grad_norm": 1.390625,
      "learning_rate": 0.000489609973862412,
      "loss": 5.0127,
      "mean_token_accuracy": 0.19905549436807632,
      "num_tokens": 23052542.0,
      "step": 12530
    },
    {
      "entropy": 5.34961895942688,
      "epoch": 1.0470263949214835,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004896010299136793,
      "loss": 5.0868,
      "mean_token_accuracy": 0.19590330123901367,
      "num_tokens": 23061438.0,
      "step": 12535
    },
    {
      "entropy": 5.339598798751831,
      "epoch": 1.0474440360841966,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004895920822080972,
      "loss": 5.1856,
      "mean_token_accuracy": 0.18428489714860916,
      "num_tokens": 23071777.0,
      "step": 12540
    },
    {
      "entropy": 5.312030458450318,
      "epoch": 1.0478616772469094,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004895831307458224,
      "loss": 5.0559,
      "mean_token_accuracy": 0.1971406489610672,
      "num_tokens": 23080889.0,
      "step": 12545
    },
    {
      "entropy": 5.216479253768921,
      "epoch": 1.0482793184096224,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004895741755270115,
      "loss": 4.9408,
      "mean_token_accuracy": 0.20071810483932495,
      "num_tokens": 23089737.0,
      "step": 12550
    },
    {
      "entropy": 5.174747180938721,
      "epoch": 1.0486969595723354,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004895652165518215,
      "loss": 5.0116,
      "mean_token_accuracy": 0.20586864799261093,
      "num_tokens": 23098861.0,
      "step": 12555
    },
    {
      "entropy": 5.244504785537719,
      "epoch": 1.0491146007350485,
      "grad_norm": 1.5,
      "learning_rate": 0.0004895562538204091,
      "loss": 5.1178,
      "mean_token_accuracy": 0.19042373895645143,
      "num_tokens": 23107798.0,
      "step": 12560
    },
    {
      "entropy": 5.271414041519165,
      "epoch": 1.0495322418977615,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004895472873329311,
      "loss": 4.9578,
      "mean_token_accuracy": 0.2030344769358635,
      "num_tokens": 23116698.0,
      "step": 12565
    },
    {
      "entropy": 5.286085891723633,
      "epoch": 1.0499498830604745,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004895383170895447,
      "loss": 5.1369,
      "mean_token_accuracy": 0.19007036685943604,
      "num_tokens": 23125880.0,
      "step": 12570
    },
    {
      "entropy": 5.308351278305054,
      "epoch": 1.0503675242231874,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004895293430904067,
      "loss": 5.16,
      "mean_token_accuracy": 0.19246817827224733,
      "num_tokens": 23135633.0,
      "step": 12575
    },
    {
      "entropy": 5.247073650360107,
      "epoch": 1.0507851653859004,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004895203653356745,
      "loss": 4.9922,
      "mean_token_accuracy": 0.19710817486047744,
      "num_tokens": 23144952.0,
      "step": 12580
    },
    {
      "entropy": 5.180167293548584,
      "epoch": 1.0512028065486134,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004895113838255049,
      "loss": 5.0058,
      "mean_token_accuracy": 0.2022145539522171,
      "num_tokens": 23154128.0,
      "step": 12585
    },
    {
      "entropy": 5.145247459411621,
      "epoch": 1.0516204477113265,
      "grad_norm": 1.625,
      "learning_rate": 0.0004895023985600555,
      "loss": 4.9286,
      "mean_token_accuracy": 0.20536952763795852,
      "num_tokens": 23162579.0,
      "step": 12590
    },
    {
      "entropy": 5.225745010375976,
      "epoch": 1.0520380888740395,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004894934095394832,
      "loss": 5.066,
      "mean_token_accuracy": 0.19630144387483597,
      "num_tokens": 23171889.0,
      "step": 12595
    },
    {
      "entropy": 5.158015441894531,
      "epoch": 1.0524557300367525,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004894844167639458,
      "loss": 4.9205,
      "mean_token_accuracy": 0.20611272007226944,
      "num_tokens": 23180876.0,
      "step": 12600
    },
    {
      "entropy": 5.182101011276245,
      "epoch": 1.0528733711994653,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004894754202336004,
      "loss": 4.9656,
      "mean_token_accuracy": 0.20411943048238754,
      "num_tokens": 23189567.0,
      "step": 12605
    },
    {
      "entropy": 5.324113845825195,
      "epoch": 1.0532910123621784,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004894664199486047,
      "loss": 4.9814,
      "mean_token_accuracy": 0.19692541807889938,
      "num_tokens": 23198142.0,
      "step": 12610
    },
    {
      "entropy": 5.254392004013061,
      "epoch": 1.0537086535248914,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004894574159091161,
      "loss": 5.0877,
      "mean_token_accuracy": 0.1905139923095703,
      "num_tokens": 23207143.0,
      "step": 12615
    },
    {
      "entropy": 5.278478670120239,
      "epoch": 1.0541262946876044,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004894484081152922,
      "loss": 4.999,
      "mean_token_accuracy": 0.19988663345575333,
      "num_tokens": 23215637.0,
      "step": 12620
    },
    {
      "entropy": 5.292605352401734,
      "epoch": 1.0545439358503175,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004894393965672909,
      "loss": 5.0329,
      "mean_token_accuracy": 0.19582016617059708,
      "num_tokens": 23224402.0,
      "step": 12625
    },
    {
      "entropy": 5.430939149856568,
      "epoch": 1.0549615770130305,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004894303812652698,
      "loss": 5.2741,
      "mean_token_accuracy": 0.18121656626462937,
      "num_tokens": 23232669.0,
      "step": 12630
    },
    {
      "entropy": 5.195004987716675,
      "epoch": 1.0553792181757433,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004894213622093867,
      "loss": 4.9943,
      "mean_token_accuracy": 0.19472752660512924,
      "num_tokens": 23242204.0,
      "step": 12635
    },
    {
      "entropy": 5.265455102920532,
      "epoch": 1.0557968593384564,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004894123393997997,
      "loss": 5.0486,
      "mean_token_accuracy": 0.193990857899189,
      "num_tokens": 23250765.0,
      "step": 12640
    },
    {
      "entropy": 5.2288980960845945,
      "epoch": 1.0562145005011694,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004894033128366665,
      "loss": 4.9769,
      "mean_token_accuracy": 0.19706922620534897,
      "num_tokens": 23259986.0,
      "step": 12645
    },
    {
      "entropy": 5.153661155700684,
      "epoch": 1.0566321416638824,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004893942825201453,
      "loss": 4.8858,
      "mean_token_accuracy": 0.20731846690177919,
      "num_tokens": 23268599.0,
      "step": 12650
    },
    {
      "entropy": 5.166029930114746,
      "epoch": 1.0570497828265955,
      "grad_norm": 1.578125,
      "learning_rate": 0.000489385248450394,
      "loss": 4.9655,
      "mean_token_accuracy": 0.20276941806077958,
      "num_tokens": 23277509.0,
      "step": 12655
    },
    {
      "entropy": 5.274418210983276,
      "epoch": 1.0574674239893085,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000489376210627571,
      "loss": 5.0052,
      "mean_token_accuracy": 0.19347480833530425,
      "num_tokens": 23286673.0,
      "step": 12660
    },
    {
      "entropy": 5.287510967254638,
      "epoch": 1.0578850651520213,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004893671690518343,
      "loss": 4.9833,
      "mean_token_accuracy": 0.20537156164646148,
      "num_tokens": 23295775.0,
      "step": 12665
    },
    {
      "entropy": 5.159037923812866,
      "epoch": 1.0583027063147343,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004893581237233423,
      "loss": 4.9961,
      "mean_token_accuracy": 0.20119742900133133,
      "num_tokens": 23305321.0,
      "step": 12670
    },
    {
      "entropy": 5.2546546936035154,
      "epoch": 1.0587203474774474,
      "grad_norm": 2.125,
      "learning_rate": 0.0004893490746422533,
      "loss": 4.9859,
      "mean_token_accuracy": 0.1984270617365837,
      "num_tokens": 23313798.0,
      "step": 12675
    },
    {
      "entropy": 5.139239645004272,
      "epoch": 1.0591379886401604,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004893400218087258,
      "loss": 4.871,
      "mean_token_accuracy": 0.2094012215733528,
      "num_tokens": 23322308.0,
      "step": 12680
    },
    {
      "entropy": 5.253513193130493,
      "epoch": 1.0595556298028734,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004893309652229183,
      "loss": 5.0935,
      "mean_token_accuracy": 0.19093754887580872,
      "num_tokens": 23331311.0,
      "step": 12685
    },
    {
      "entropy": 5.324931669235229,
      "epoch": 1.0599732709655865,
      "grad_norm": 1.75,
      "learning_rate": 0.0004893219048849891,
      "loss": 5.171,
      "mean_token_accuracy": 0.19184111803770065,
      "num_tokens": 23340655.0,
      "step": 12690
    },
    {
      "entropy": 5.334091281890869,
      "epoch": 1.0603909121282993,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004893128407950971,
      "loss": 5.2092,
      "mean_token_accuracy": 0.1892612248659134,
      "num_tokens": 23350999.0,
      "step": 12695
    },
    {
      "entropy": 5.246554136276245,
      "epoch": 1.0608085532910123,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004893037729534007,
      "loss": 5.0047,
      "mean_token_accuracy": 0.20068952441215515,
      "num_tokens": 23359311.0,
      "step": 12700
    },
    {
      "entropy": 5.286236572265625,
      "epoch": 1.0612261944537253,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004892947013600589,
      "loss": 5.042,
      "mean_token_accuracy": 0.19421575218439102,
      "num_tokens": 23367925.0,
      "step": 12705
    },
    {
      "entropy": 5.263022947311401,
      "epoch": 1.0616438356164384,
      "grad_norm": 2.34375,
      "learning_rate": 0.0004892856260152304,
      "loss": 5.1568,
      "mean_token_accuracy": 0.1914500206708908,
      "num_tokens": 23377159.0,
      "step": 12710
    },
    {
      "entropy": 5.399932241439819,
      "epoch": 1.0620614767791514,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000489276546919074,
      "loss": 5.1938,
      "mean_token_accuracy": 0.18248753994703293,
      "num_tokens": 23387699.0,
      "step": 12715
    },
    {
      "entropy": 5.314846515655518,
      "epoch": 1.0624791179418644,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004892674640717488,
      "loss": 5.1216,
      "mean_token_accuracy": 0.19561748802661896,
      "num_tokens": 23397520.0,
      "step": 12720
    },
    {
      "entropy": 5.185560512542724,
      "epoch": 1.0628967591045773,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004892583774734138,
      "loss": 5.0264,
      "mean_token_accuracy": 0.19200746864080429,
      "num_tokens": 23406853.0,
      "step": 12725
    },
    {
      "entropy": 5.175264120101929,
      "epoch": 1.0633144002672903,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004892492871242278,
      "loss": 4.981,
      "mean_token_accuracy": 0.20046818852424622,
      "num_tokens": 23416235.0,
      "step": 12730
    },
    {
      "entropy": 5.2853508472442625,
      "epoch": 1.0637320414300033,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004892401930243504,
      "loss": 5.0525,
      "mean_token_accuracy": 0.19408349096775054,
      "num_tokens": 23425347.0,
      "step": 12735
    },
    {
      "entropy": 5.312745475769043,
      "epoch": 1.0641496825927164,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004892310951739404,
      "loss": 5.2047,
      "mean_token_accuracy": 0.1898958146572113,
      "num_tokens": 23434905.0,
      "step": 12740
    },
    {
      "entropy": 5.296979522705078,
      "epoch": 1.0645673237554294,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004892219935731571,
      "loss": 5.0194,
      "mean_token_accuracy": 0.20163357853889466,
      "num_tokens": 23444495.0,
      "step": 12745
    },
    {
      "entropy": 5.174061965942383,
      "epoch": 1.0649849649181424,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00048921288822216,
      "loss": 4.9681,
      "mean_token_accuracy": 0.19993997663259505,
      "num_tokens": 23452224.0,
      "step": 12750
    },
    {
      "entropy": 5.216191959381104,
      "epoch": 1.0654026060808552,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004892037791211086,
      "loss": 5.1355,
      "mean_token_accuracy": 0.18996180444955826,
      "num_tokens": 23461987.0,
      "step": 12755
    },
    {
      "entropy": 5.163602304458618,
      "epoch": 1.0658202472435683,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004891946662701621,
      "loss": 4.9508,
      "mean_token_accuracy": 0.2081431820988655,
      "num_tokens": 23470645.0,
      "step": 12760
    },
    {
      "entropy": 5.2784381866455075,
      "epoch": 1.0662378884062813,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004891855496694801,
      "loss": 5.1335,
      "mean_token_accuracy": 0.19354406446218492,
      "num_tokens": 23480169.0,
      "step": 12765
    },
    {
      "entropy": 5.2956990718841555,
      "epoch": 1.0666555295689943,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004891764293192222,
      "loss": 5.0659,
      "mean_token_accuracy": 0.1961946338415146,
      "num_tokens": 23488630.0,
      "step": 12770
    },
    {
      "entropy": 5.190297937393188,
      "epoch": 1.0670731707317074,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004891673052195482,
      "loss": 4.9489,
      "mean_token_accuracy": 0.19633503407239913,
      "num_tokens": 23497800.0,
      "step": 12775
    },
    {
      "entropy": 5.21078143119812,
      "epoch": 1.0674908118944204,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004891581773706177,
      "loss": 5.1297,
      "mean_token_accuracy": 0.19167614430189134,
      "num_tokens": 23507400.0,
      "step": 12780
    },
    {
      "entropy": 5.338127708435058,
      "epoch": 1.0679084530571332,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004891490457725906,
      "loss": 5.1504,
      "mean_token_accuracy": 0.17745302021503448,
      "num_tokens": 23516441.0,
      "step": 12785
    },
    {
      "entropy": 5.253556871414185,
      "epoch": 1.0683260942198463,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004891399104256266,
      "loss": 4.9725,
      "mean_token_accuracy": 0.1929553836584091,
      "num_tokens": 23525066.0,
      "step": 12790
    },
    {
      "entropy": 5.135930967330933,
      "epoch": 1.0687437353825593,
      "grad_norm": 1.375,
      "learning_rate": 0.0004891307713298857,
      "loss": 4.9428,
      "mean_token_accuracy": 0.20225382298231126,
      "num_tokens": 23533892.0,
      "step": 12795
    },
    {
      "entropy": 5.183720254898072,
      "epoch": 1.0691613765452723,
      "grad_norm": 1.421875,
      "learning_rate": 0.000489121628485528,
      "loss": 5.0843,
      "mean_token_accuracy": 0.19113122373819352,
      "num_tokens": 23542680.0,
      "step": 12800
    },
    {
      "entropy": 5.226210832595825,
      "epoch": 1.0695790177079854,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004891124818927133,
      "loss": 4.8788,
      "mean_token_accuracy": 0.21258097887039185,
      "num_tokens": 23550573.0,
      "step": 12805
    },
    {
      "entropy": 5.1033806800842285,
      "epoch": 1.0699966588706984,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004891033315516019,
      "loss": 5.0339,
      "mean_token_accuracy": 0.19297184497117997,
      "num_tokens": 23561281.0,
      "step": 12810
    },
    {
      "entropy": 5.184213066101075,
      "epoch": 1.0704143000334112,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004890941774623541,
      "loss": 4.9274,
      "mean_token_accuracy": 0.20642451494932174,
      "num_tokens": 23569846.0,
      "step": 12815
    },
    {
      "entropy": 5.22930498123169,
      "epoch": 1.0708319411961242,
      "grad_norm": 2.125,
      "learning_rate": 0.0004890850196251299,
      "loss": 4.9339,
      "mean_token_accuracy": 0.2084512308239937,
      "num_tokens": 23578145.0,
      "step": 12820
    },
    {
      "entropy": 5.200729608535767,
      "epoch": 1.0712495823588373,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004890758580400898,
      "loss": 5.0374,
      "mean_token_accuracy": 0.19567439705133438,
      "num_tokens": 23587494.0,
      "step": 12825
    },
    {
      "entropy": 5.315274620056153,
      "epoch": 1.0716672235215503,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004890666927073941,
      "loss": 5.1299,
      "mean_token_accuracy": 0.18970830738544464,
      "num_tokens": 23596599.0,
      "step": 12830
    },
    {
      "entropy": 5.219842195510864,
      "epoch": 1.0720848646842633,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004890575236272032,
      "loss": 5.0839,
      "mean_token_accuracy": 0.19962160736322404,
      "num_tokens": 23606170.0,
      "step": 12835
    },
    {
      "entropy": 5.20392632484436,
      "epoch": 1.0725025058469764,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004890483507996778,
      "loss": 4.8782,
      "mean_token_accuracy": 0.205711966753006,
      "num_tokens": 23614558.0,
      "step": 12840
    },
    {
      "entropy": 5.284447813034058,
      "epoch": 1.0729201470096892,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004890391742249783,
      "loss": 5.0784,
      "mean_token_accuracy": 0.1932889774441719,
      "num_tokens": 23624125.0,
      "step": 12845
    },
    {
      "entropy": 5.317376899719238,
      "epoch": 1.0733377881724022,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004890299939032654,
      "loss": 5.0433,
      "mean_token_accuracy": 0.20024346709251403,
      "num_tokens": 23633025.0,
      "step": 12850
    },
    {
      "entropy": 5.1511646747589115,
      "epoch": 1.0737554293351153,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004890208098346999,
      "loss": 5.0001,
      "mean_token_accuracy": 0.19418468177318574,
      "num_tokens": 23643387.0,
      "step": 12855
    },
    {
      "entropy": 5.262984895706177,
      "epoch": 1.0741730704978283,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004890116220194425,
      "loss": 5.0698,
      "mean_token_accuracy": 0.19583483040332794,
      "num_tokens": 23653486.0,
      "step": 12860
    },
    {
      "entropy": 5.201009273529053,
      "epoch": 1.0745907116605413,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000489002430457654,
      "loss": 4.9848,
      "mean_token_accuracy": 0.19988911151885985,
      "num_tokens": 23663483.0,
      "step": 12865
    },
    {
      "entropy": 5.150922203063965,
      "epoch": 1.0750083528232544,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004889932351494954,
      "loss": 5.0184,
      "mean_token_accuracy": 0.20066923201084136,
      "num_tokens": 23672453.0,
      "step": 12870
    },
    {
      "entropy": 5.23501672744751,
      "epoch": 1.0754259939859672,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004889840360951277,
      "loss": 5.0271,
      "mean_token_accuracy": 0.19712787717580796,
      "num_tokens": 23681351.0,
      "step": 12875
    },
    {
      "entropy": 5.247542905807495,
      "epoch": 1.0758436351486802,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004889748332947118,
      "loss": 5.0066,
      "mean_token_accuracy": 0.1982413947582245,
      "num_tokens": 23689581.0,
      "step": 12880
    },
    {
      "entropy": 5.258001470565796,
      "epoch": 1.0762612763113932,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004889656267484089,
      "loss": 5.1071,
      "mean_token_accuracy": 0.18993108719587326,
      "num_tokens": 23699415.0,
      "step": 12885
    },
    {
      "entropy": 5.26158447265625,
      "epoch": 1.0766789174741063,
      "grad_norm": 2.0,
      "learning_rate": 0.00048895641645638,
      "loss": 5.0355,
      "mean_token_accuracy": 0.20100877732038497,
      "num_tokens": 23708933.0,
      "step": 12890
    },
    {
      "entropy": 5.214012813568115,
      "epoch": 1.0770965586368193,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004889472024187867,
      "loss": 5.0783,
      "mean_token_accuracy": 0.186471064388752,
      "num_tokens": 23718423.0,
      "step": 12895
    },
    {
      "entropy": 5.287875747680664,
      "epoch": 1.0775141997995323,
      "grad_norm": 1.453125,
      "learning_rate": 0.00048893798463579,
      "loss": 5.0605,
      "mean_token_accuracy": 0.19091844260692598,
      "num_tokens": 23728131.0,
      "step": 12900
    },
    {
      "entropy": 5.211427307128906,
      "epoch": 1.0779318409622451,
      "grad_norm": 2.0,
      "learning_rate": 0.0004889287631075513,
      "loss": 5.0213,
      "mean_token_accuracy": 0.20042961984872817,
      "num_tokens": 23737378.0,
      "step": 12905
    },
    {
      "entropy": 5.395720338821411,
      "epoch": 1.0783494821249582,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004889195378342322,
      "loss": 5.1276,
      "mean_token_accuracy": 0.18550482839345933,
      "num_tokens": 23747156.0,
      "step": 12910
    },
    {
      "entropy": 5.290826368331909,
      "epoch": 1.0787671232876712,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004889103088159941,
      "loss": 5.0734,
      "mean_token_accuracy": 0.1926543101668358,
      "num_tokens": 23757090.0,
      "step": 12915
    },
    {
      "entropy": 5.236329746246338,
      "epoch": 1.0791847644503842,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004889010760529986,
      "loss": 5.0686,
      "mean_token_accuracy": 0.19342664629220963,
      "num_tokens": 23766679.0,
      "step": 12920
    },
    {
      "entropy": 5.22120885848999,
      "epoch": 1.0796024056130973,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004888918395454071,
      "loss": 4.9529,
      "mean_token_accuracy": 0.2114442393183708,
      "num_tokens": 23775109.0,
      "step": 12925
    },
    {
      "entropy": 5.229614353179931,
      "epoch": 1.0800200467758103,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004888825992933817,
      "loss": 5.0466,
      "mean_token_accuracy": 0.2030608519911766,
      "num_tokens": 23784698.0,
      "step": 12930
    },
    {
      "entropy": 5.099056196212769,
      "epoch": 1.0804376879385231,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004888733552970839,
      "loss": 4.9582,
      "mean_token_accuracy": 0.20638607293367386,
      "num_tokens": 23793590.0,
      "step": 12935
    },
    {
      "entropy": 5.2329333305358885,
      "epoch": 1.0808553291012362,
      "grad_norm": 1.625,
      "learning_rate": 0.0004888641075566755,
      "loss": 4.9701,
      "mean_token_accuracy": 0.19810496419668197,
      "num_tokens": 23802421.0,
      "step": 12940
    },
    {
      "entropy": 5.284356641769409,
      "epoch": 1.0812729702639492,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004888548560723184,
      "loss": 5.0931,
      "mean_token_accuracy": 0.19796355664730073,
      "num_tokens": 23811610.0,
      "step": 12945
    },
    {
      "entropy": 5.322061777114868,
      "epoch": 1.0816906114266622,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004888456008441748,
      "loss": 5.1229,
      "mean_token_accuracy": 0.19698571413755417,
      "num_tokens": 23821197.0,
      "step": 12950
    },
    {
      "entropy": 5.286272573471069,
      "epoch": 1.0821082525893753,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004888363418724064,
      "loss": 5.0621,
      "mean_token_accuracy": 0.19100695997476577,
      "num_tokens": 23829757.0,
      "step": 12955
    },
    {
      "entropy": 5.256100702285766,
      "epoch": 1.0825258937520883,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004888270791571756,
      "loss": 5.0456,
      "mean_token_accuracy": 0.1903567612171173,
      "num_tokens": 23839242.0,
      "step": 12960
    },
    {
      "entropy": 5.229103946685791,
      "epoch": 1.082943534914801,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004888178126986444,
      "loss": 4.9889,
      "mean_token_accuracy": 0.20518402606248856,
      "num_tokens": 23848193.0,
      "step": 12965
    },
    {
      "entropy": 5.206324100494385,
      "epoch": 1.0833611760775141,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004888085424969748,
      "loss": 4.9702,
      "mean_token_accuracy": 0.1973526030778885,
      "num_tokens": 23857623.0,
      "step": 12970
    },
    {
      "entropy": 5.28991436958313,
      "epoch": 1.0837788172402272,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004887992685523295,
      "loss": 5.0857,
      "mean_token_accuracy": 0.19137540608644485,
      "num_tokens": 23866424.0,
      "step": 12975
    },
    {
      "entropy": 5.283302259445191,
      "epoch": 1.0841964584029402,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004887899908648706,
      "loss": 5.1234,
      "mean_token_accuracy": 0.19656671732664108,
      "num_tokens": 23874855.0,
      "step": 12980
    },
    {
      "entropy": 5.307825946807862,
      "epoch": 1.0846140995656532,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004887807094347606,
      "loss": 5.0773,
      "mean_token_accuracy": 0.1940826505422592,
      "num_tokens": 23884634.0,
      "step": 12985
    },
    {
      "entropy": 5.356455898284912,
      "epoch": 1.0850317407283663,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004887714242621619,
      "loss": 5.1184,
      "mean_token_accuracy": 0.19090297371149062,
      "num_tokens": 23894307.0,
      "step": 12990
    },
    {
      "entropy": 5.260967969894409,
      "epoch": 1.085449381891079,
      "grad_norm": 3.28125,
      "learning_rate": 0.0004887621353472371,
      "loss": 5.0135,
      "mean_token_accuracy": 0.20085800141096116,
      "num_tokens": 23903274.0,
      "step": 12995
    },
    {
      "entropy": 5.228359842300415,
      "epoch": 1.0858670230537921,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004887528426901488,
      "loss": 5.0505,
      "mean_token_accuracy": 0.1971916824579239,
      "num_tokens": 23912787.0,
      "step": 13000
    },
    {
      "entropy": 5.305121803283692,
      "epoch": 1.0862846642165052,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004887435462910597,
      "loss": 5.087,
      "mean_token_accuracy": 0.1834319144487381,
      "num_tokens": 23921558.0,
      "step": 13005
    },
    {
      "entropy": 5.2337846755981445,
      "epoch": 1.0867023053792182,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004887342461501326,
      "loss": 4.9902,
      "mean_token_accuracy": 0.19493157267570496,
      "num_tokens": 23929868.0,
      "step": 13010
    },
    {
      "entropy": 5.270093250274658,
      "epoch": 1.0871199465419312,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004887249422675301,
      "loss": 5.0917,
      "mean_token_accuracy": 0.18756240159273146,
      "num_tokens": 23939206.0,
      "step": 13015
    },
    {
      "entropy": 5.269919538497925,
      "epoch": 1.0875375877046443,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004887156346434154,
      "loss": 5.0032,
      "mean_token_accuracy": 0.1940973773598671,
      "num_tokens": 23948213.0,
      "step": 13020
    },
    {
      "entropy": 5.233686733245849,
      "epoch": 1.087955228867357,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004887063232779513,
      "loss": 5.0288,
      "mean_token_accuracy": 0.19157015681266784,
      "num_tokens": 23957269.0,
      "step": 13025
    },
    {
      "entropy": 5.326388120651245,
      "epoch": 1.08837287003007,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004886970081713006,
      "loss": 5.1614,
      "mean_token_accuracy": 0.19049212485551834,
      "num_tokens": 23967300.0,
      "step": 13030
    },
    {
      "entropy": 5.2333118438720705,
      "epoch": 1.0887905111927831,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004886876893236266,
      "loss": 5.0297,
      "mean_token_accuracy": 0.19713301956653595,
      "num_tokens": 23976934.0,
      "step": 13035
    },
    {
      "entropy": 5.192884397506714,
      "epoch": 1.0892081523554962,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004886783667350925,
      "loss": 5.037,
      "mean_token_accuracy": 0.19572983235120772,
      "num_tokens": 23986119.0,
      "step": 13040
    },
    {
      "entropy": 5.240741920471192,
      "epoch": 1.0896257935182092,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004886690404058613,
      "loss": 5.0642,
      "mean_token_accuracy": 0.1981247693300247,
      "num_tokens": 23996093.0,
      "step": 13045
    },
    {
      "entropy": 5.392534875869751,
      "epoch": 1.0900434346809222,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004886597103360963,
      "loss": 5.189,
      "mean_token_accuracy": 0.18296193778514863,
      "num_tokens": 24006585.0,
      "step": 13050
    },
    {
      "entropy": 5.325340700149536,
      "epoch": 1.090461075843635,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004886503765259609,
      "loss": 5.1316,
      "mean_token_accuracy": 0.1919292241334915,
      "num_tokens": 24016398.0,
      "step": 13055
    },
    {
      "entropy": 5.242816305160522,
      "epoch": 1.090878717006348,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004886410389756186,
      "loss": 4.9616,
      "mean_token_accuracy": 0.19498070925474167,
      "num_tokens": 24025988.0,
      "step": 13060
    },
    {
      "entropy": 5.242601346969605,
      "epoch": 1.0912963581690611,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004886316976852326,
      "loss": 5.0588,
      "mean_token_accuracy": 0.19371406883001327,
      "num_tokens": 24036448.0,
      "step": 13065
    },
    {
      "entropy": 5.269687032699585,
      "epoch": 1.0917139993317742,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004886223526549666,
      "loss": 5.1802,
      "mean_token_accuracy": 0.18471720665693284,
      "num_tokens": 24047011.0,
      "step": 13070
    },
    {
      "entropy": 5.243171215057373,
      "epoch": 1.0921316404944872,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004886130038849843,
      "loss": 4.9401,
      "mean_token_accuracy": 0.20318590253591537,
      "num_tokens": 24056332.0,
      "step": 13075
    },
    {
      "entropy": 5.257252979278564,
      "epoch": 1.0925492816572002,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004886036513754491,
      "loss": 5.117,
      "mean_token_accuracy": 0.19600074738264084,
      "num_tokens": 24065538.0,
      "step": 13080
    },
    {
      "entropy": 5.230002689361572,
      "epoch": 1.092966922819913,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004885942951265251,
      "loss": 5.051,
      "mean_token_accuracy": 0.19860922545194626,
      "num_tokens": 24074950.0,
      "step": 13085
    },
    {
      "entropy": 5.302611637115478,
      "epoch": 1.093384563982626,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004885849351383756,
      "loss": 5.1202,
      "mean_token_accuracy": 0.19223973155021667,
      "num_tokens": 24084571.0,
      "step": 13090
    },
    {
      "entropy": 5.286756324768066,
      "epoch": 1.093802205145339,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004885755714111649,
      "loss": 5.116,
      "mean_token_accuracy": 0.1949808567762375,
      "num_tokens": 24093693.0,
      "step": 13095
    },
    {
      "entropy": 5.255267190933227,
      "epoch": 1.0942198463080521,
      "grad_norm": 1.625,
      "learning_rate": 0.0004885662039450565,
      "loss": 5.0214,
      "mean_token_accuracy": 0.1948310077190399,
      "num_tokens": 24104140.0,
      "step": 13100
    },
    {
      "entropy": 5.269103288650513,
      "epoch": 1.0946374874707652,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0004885568327402149,
      "loss": 5.007,
      "mean_token_accuracy": 0.20086413025856018,
      "num_tokens": 24113107.0,
      "step": 13105
    },
    {
      "entropy": 5.215589952468872,
      "epoch": 1.0950551286334782,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004885474577968038,
      "loss": 5.0513,
      "mean_token_accuracy": 0.19714694023132323,
      "num_tokens": 24122271.0,
      "step": 13110
    },
    {
      "entropy": 5.265857124328614,
      "epoch": 1.095472769796191,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004885380791149874,
      "loss": 5.1154,
      "mean_token_accuracy": 0.18980410993099212,
      "num_tokens": 24132068.0,
      "step": 13115
    },
    {
      "entropy": 5.295174503326416,
      "epoch": 1.095890410958904,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004885286966949298,
      "loss": 5.0081,
      "mean_token_accuracy": 0.19659265726804734,
      "num_tokens": 24140429.0,
      "step": 13120
    },
    {
      "entropy": 5.263141345977783,
      "epoch": 1.096308052121617,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004885193105367955,
      "loss": 5.1294,
      "mean_token_accuracy": 0.19368053376674652,
      "num_tokens": 24150563.0,
      "step": 13125
    },
    {
      "entropy": 5.253929615020752,
      "epoch": 1.0967256932843301,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004885099206407486,
      "loss": 5.0363,
      "mean_token_accuracy": 0.19127064496278762,
      "num_tokens": 24159059.0,
      "step": 13130
    },
    {
      "entropy": 5.381769418716431,
      "epoch": 1.0971433344470432,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004885005270069535,
      "loss": 5.1722,
      "mean_token_accuracy": 0.18928706496953965,
      "num_tokens": 24168858.0,
      "step": 13135
    },
    {
      "entropy": 5.260519313812256,
      "epoch": 1.0975609756097562,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004884911296355746,
      "loss": 5.0646,
      "mean_token_accuracy": 0.19336674660444259,
      "num_tokens": 24177573.0,
      "step": 13140
    },
    {
      "entropy": 5.247461318969727,
      "epoch": 1.097978616772469,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004884817285267766,
      "loss": 5.013,
      "mean_token_accuracy": 0.19790593087673186,
      "num_tokens": 24186832.0,
      "step": 13145
    },
    {
      "entropy": 5.196414232254028,
      "epoch": 1.098396257935182,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004884723236807239,
      "loss": 5.026,
      "mean_token_accuracy": 0.19572710543870925,
      "num_tokens": 24195281.0,
      "step": 13150
    },
    {
      "entropy": 5.229942560195923,
      "epoch": 1.098813899097895,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004884629150975814,
      "loss": 5.0398,
      "mean_token_accuracy": 0.2007513165473938,
      "num_tokens": 24204499.0,
      "step": 13155
    },
    {
      "entropy": 5.162657928466797,
      "epoch": 1.099231540260608,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004884535027775135,
      "loss": 4.9237,
      "mean_token_accuracy": 0.2136691316962242,
      "num_tokens": 24213418.0,
      "step": 13160
    },
    {
      "entropy": 5.1497570991516115,
      "epoch": 1.0996491814233211,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004884440867206851,
      "loss": 4.9347,
      "mean_token_accuracy": 0.20316660404205322,
      "num_tokens": 24222834.0,
      "step": 13165
    },
    {
      "entropy": 5.217789840698242,
      "epoch": 1.1000668225860342,
      "grad_norm": 1.78125,
      "learning_rate": 0.000488434666927261,
      "loss": 5.0279,
      "mean_token_accuracy": 0.19720485508441926,
      "num_tokens": 24231821.0,
      "step": 13170
    },
    {
      "entropy": 5.316047859191895,
      "epoch": 1.100484463748747,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004884252433974062,
      "loss": 5.0344,
      "mean_token_accuracy": 0.19637175649404526,
      "num_tokens": 24240694.0,
      "step": 13175
    },
    {
      "entropy": 5.2582377910614015,
      "epoch": 1.10090210491146,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004884158161312856,
      "loss": 5.0542,
      "mean_token_accuracy": 0.18963941931724548,
      "num_tokens": 24249927.0,
      "step": 13180
    },
    {
      "entropy": 5.249260377883911,
      "epoch": 1.101319746074173,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004884063851290642,
      "loss": 5.0174,
      "mean_token_accuracy": 0.19640277028083802,
      "num_tokens": 24259174.0,
      "step": 13185
    },
    {
      "entropy": 5.260727643966675,
      "epoch": 1.101737387236886,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004883969503909073,
      "loss": 5.0581,
      "mean_token_accuracy": 0.19531009942293168,
      "num_tokens": 24268381.0,
      "step": 13190
    },
    {
      "entropy": 5.305428409576416,
      "epoch": 1.1021550283995991,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004883875119169798,
      "loss": 5.1373,
      "mean_token_accuracy": 0.19397216737270356,
      "num_tokens": 24277916.0,
      "step": 13195
    },
    {
      "entropy": 5.267978048324585,
      "epoch": 1.1025726695623121,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000488378069707447,
      "loss": 5.118,
      "mean_token_accuracy": 0.19240879118442536,
      "num_tokens": 24287530.0,
      "step": 13200
    },
    {
      "entropy": 5.2984882354736325,
      "epoch": 1.102990310725025,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004883686237624744,
      "loss": 5.2397,
      "mean_token_accuracy": 0.18264338374137878,
      "num_tokens": 24298239.0,
      "step": 13205
    },
    {
      "entropy": 5.356864500045776,
      "epoch": 1.103407951887738,
      "grad_norm": 1.375,
      "learning_rate": 0.0004883591740822271,
      "loss": 5.0528,
      "mean_token_accuracy": 0.19344450384378434,
      "num_tokens": 24307001.0,
      "step": 13210
    },
    {
      "entropy": 5.266033506393432,
      "epoch": 1.103825593050451,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004883497206668707,
      "loss": 5.0212,
      "mean_token_accuracy": 0.19660103768110276,
      "num_tokens": 24315995.0,
      "step": 13215
    },
    {
      "entropy": 5.207292461395264,
      "epoch": 1.104243234213164,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004883402635165705,
      "loss": 4.9579,
      "mean_token_accuracy": 0.20370802879333497,
      "num_tokens": 24324421.0,
      "step": 13220
    },
    {
      "entropy": 5.278978109359741,
      "epoch": 1.104660875375877,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004883308026314923,
      "loss": 5.0997,
      "mean_token_accuracy": 0.19509113281965257,
      "num_tokens": 24333715.0,
      "step": 13225
    },
    {
      "entropy": 5.204811239242554,
      "epoch": 1.1050785165385901,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004883213380118016,
      "loss": 4.9924,
      "mean_token_accuracy": 0.20283719152212143,
      "num_tokens": 24342085.0,
      "step": 13230
    },
    {
      "entropy": 5.259184503555298,
      "epoch": 1.105496157701303,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004883118696576642,
      "loss": 5.1064,
      "mean_token_accuracy": 0.1954349160194397,
      "num_tokens": 24350906.0,
      "step": 13235
    },
    {
      "entropy": 5.188865375518799,
      "epoch": 1.105913798864016,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004883023975692457,
      "loss": 4.953,
      "mean_token_accuracy": 0.19601337164640426,
      "num_tokens": 24359948.0,
      "step": 13240
    },
    {
      "entropy": 5.142308902740479,
      "epoch": 1.106331440026729,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004882929217467118,
      "loss": 4.9598,
      "mean_token_accuracy": 0.20338685661554337,
      "num_tokens": 24369440.0,
      "step": 13245
    },
    {
      "entropy": 5.2372642993927006,
      "epoch": 1.106749081189442,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004882834421902287,
      "loss": 5.0747,
      "mean_token_accuracy": 0.19608121365308762,
      "num_tokens": 24378442.0,
      "step": 13250
    },
    {
      "entropy": 5.264742136001587,
      "epoch": 1.107166722352155,
      "grad_norm": 1.59375,
      "learning_rate": 0.00048827395889996226,
      "loss": 5.0223,
      "mean_token_accuracy": 0.1981010064482689,
      "num_tokens": 24387750.0,
      "step": 13255
    },
    {
      "entropy": 5.204302549362183,
      "epoch": 1.107584363514868,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004882644718760784,
      "loss": 4.8587,
      "mean_token_accuracy": 0.2038038894534111,
      "num_tokens": 24396380.0,
      "step": 13260
    },
    {
      "entropy": 5.124483442306518,
      "epoch": 1.108002004677581,
      "grad_norm": 1.578125,
      "learning_rate": 0.00048825498111874325,
      "loss": 4.932,
      "mean_token_accuracy": 0.204781374335289,
      "num_tokens": 24405135.0,
      "step": 13265
    },
    {
      "entropy": 5.193187189102173,
      "epoch": 1.108419645840294,
      "grad_norm": 1.515625,
      "learning_rate": 0.00048824548662812294,
      "loss": 5.0426,
      "mean_token_accuracy": 0.1967284247279167,
      "num_tokens": 24413517.0,
      "step": 13270
    },
    {
      "entropy": 5.298235940933227,
      "epoch": 1.108837287003007,
      "grad_norm": 1.640625,
      "learning_rate": 0.00048823598840438384,
      "loss": 5.0465,
      "mean_token_accuracy": 0.19728272259235383,
      "num_tokens": 24422170.0,
      "step": 13275
    },
    {
      "entropy": 5.334612703323364,
      "epoch": 1.10925492816572,
      "grad_norm": 1.8515625,
      "learning_rate": 0.00048822648644769194,
      "loss": 5.1996,
      "mean_token_accuracy": 0.18619959950447082,
      "num_tokens": 24431083.0,
      "step": 13280
    },
    {
      "entropy": 5.237791728973389,
      "epoch": 1.109672569328433,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004882169807582138,
      "loss": 5.0778,
      "mean_token_accuracy": 0.1943157583475113,
      "num_tokens": 24440582.0,
      "step": 13285
    },
    {
      "entropy": 5.264667558670044,
      "epoch": 1.110090210491146,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004882074713361157,
      "loss": 5.0356,
      "mean_token_accuracy": 0.20301843881607057,
      "num_tokens": 24450047.0,
      "step": 13290
    },
    {
      "entropy": 5.187220430374145,
      "epoch": 1.110507851653859,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004881979581815642,
      "loss": 4.8988,
      "mean_token_accuracy": 0.20455289036035537,
      "num_tokens": 24457955.0,
      "step": 13295
    },
    {
      "entropy": 5.115291452407837,
      "epoch": 1.110925492816572,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0004881884412947259,
      "loss": 4.8895,
      "mean_token_accuracy": 0.2024856835603714,
      "num_tokens": 24467394.0,
      "step": 13300
    },
    {
      "entropy": 5.141928958892822,
      "epoch": 1.111343133979285,
      "grad_norm": 1.734375,
      "learning_rate": 0.00048817892067576726,
      "loss": 4.9635,
      "mean_token_accuracy": 0.1982668787240982,
      "num_tokens": 24476386.0,
      "step": 13305
    },
    {
      "entropy": 5.174110078811646,
      "epoch": 1.111760775141998,
      "grad_norm": 1.6171875,
      "learning_rate": 0.000488169396324855,
      "loss": 5.0192,
      "mean_token_accuracy": 0.2011033833026886,
      "num_tokens": 24484800.0,
      "step": 13310
    },
    {
      "entropy": 5.249092149734497,
      "epoch": 1.112178416304711,
      "grad_norm": 1.640625,
      "learning_rate": 0.00048815986824215585,
      "loss": 4.9802,
      "mean_token_accuracy": 0.20633817762136458,
      "num_tokens": 24494252.0,
      "step": 13315
    },
    {
      "entropy": 5.257562589645386,
      "epoch": 1.112596057467424,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004881503364278366,
      "loss": 5.0336,
      "mean_token_accuracy": 0.20128801614046096,
      "num_tokens": 24503315.0,
      "step": 13320
    },
    {
      "entropy": 5.2461193084716795,
      "epoch": 1.1130136986301369,
      "grad_norm": 2.875,
      "learning_rate": 0.0004881408008820642,
      "loss": 5.0026,
      "mean_token_accuracy": 0.2026966318488121,
      "num_tokens": 24512660.0,
      "step": 13325
    },
    {
      "entropy": 5.220983171463013,
      "epoch": 1.11343133979285,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00048813126160500555,
      "loss": 5.0421,
      "mean_token_accuracy": 0.195218263566494,
      "num_tokens": 24521712.0,
      "step": 13330
    },
    {
      "entropy": 5.142714118957519,
      "epoch": 1.113848980955563,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004881217185968276,
      "loss": 4.999,
      "mean_token_accuracy": 0.2046705111861229,
      "num_tokens": 24529851.0,
      "step": 13335
    },
    {
      "entropy": 5.2647895336151125,
      "epoch": 1.114266622118276,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004881121718576974,
      "loss": 5.0577,
      "mean_token_accuracy": 0.19243974536657332,
      "num_tokens": 24538425.0,
      "step": 13340
    },
    {
      "entropy": 5.215370082855225,
      "epoch": 1.114684263280989,
      "grad_norm": 1.515625,
      "learning_rate": 0.00048810262138778213,
      "loss": 5.1254,
      "mean_token_accuracy": 0.19348261058330535,
      "num_tokens": 24547477.0,
      "step": 13345
    },
    {
      "entropy": 5.202165031433106,
      "epoch": 1.115101904443702,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00048809306718724886,
      "loss": 5.077,
      "mean_token_accuracy": 0.18883096128702165,
      "num_tokens": 24556230.0,
      "step": 13350
    },
    {
      "entropy": 5.207986164093017,
      "epoch": 1.115519545606415,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00048808350925626514,
      "loss": 5.043,
      "mean_token_accuracy": 0.20241668522357942,
      "num_tokens": 24565021.0,
      "step": 13355
    },
    {
      "entropy": 5.302781009674073,
      "epoch": 1.115937186769128,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00048807394759499797,
      "loss": 5.028,
      "mean_token_accuracy": 0.19695805311203002,
      "num_tokens": 24575012.0,
      "step": 13360
    },
    {
      "entropy": 5.323993110656739,
      "epoch": 1.116354827931841,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004880643822036149,
      "loss": 5.1712,
      "mean_token_accuracy": 0.18490291386842728,
      "num_tokens": 24584267.0,
      "step": 13365
    },
    {
      "entropy": 5.262844467163086,
      "epoch": 1.116772469094554,
      "grad_norm": 1.421875,
      "learning_rate": 0.00048805481308228333,
      "loss": 5.0648,
      "mean_token_accuracy": 0.1954543873667717,
      "num_tokens": 24594299.0,
      "step": 13370
    },
    {
      "entropy": 5.28668909072876,
      "epoch": 1.117190110257267,
      "grad_norm": 1.5,
      "learning_rate": 0.0004880452402311709,
      "loss": 5.0562,
      "mean_token_accuracy": 0.19331935346126555,
      "num_tokens": 24603717.0,
      "step": 13375
    },
    {
      "entropy": 5.2826121807098385,
      "epoch": 1.11760775141998,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048803566365044493,
      "loss": 5.11,
      "mean_token_accuracy": 0.19348817467689514,
      "num_tokens": 24613141.0,
      "step": 13380
    },
    {
      "entropy": 5.237560844421386,
      "epoch": 1.1180253925826928,
      "grad_norm": 1.625,
      "learning_rate": 0.0004880260833402734,
      "loss": 5.0575,
      "mean_token_accuracy": 0.19855398088693618,
      "num_tokens": 24621837.0,
      "step": 13385
    },
    {
      "entropy": 5.338270282745361,
      "epoch": 1.1184430337454059,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048801649930082375,
      "loss": 5.0852,
      "mean_token_accuracy": 0.19058642983436586,
      "num_tokens": 24630657.0,
      "step": 13390
    },
    {
      "entropy": 5.408084058761597,
      "epoch": 1.118860674908119,
      "grad_norm": 1.359375,
      "learning_rate": 0.000488006911532264,
      "loss": 5.1726,
      "mean_token_accuracy": 0.182516947388649,
      "num_tokens": 24639929.0,
      "step": 13395
    },
    {
      "entropy": 5.312208986282348,
      "epoch": 1.119278316070832,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004879973200347617,
      "loss": 5.1126,
      "mean_token_accuracy": 0.1973265901207924,
      "num_tokens": 24648537.0,
      "step": 13400
    },
    {
      "entropy": 5.177673101425171,
      "epoch": 1.119695957233545,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00048798772480848504,
      "loss": 5.0585,
      "mean_token_accuracy": 0.19730731695890427,
      "num_tokens": 24656809.0,
      "step": 13405
    },
    {
      "entropy": 5.217769384384155,
      "epoch": 1.120113598396258,
      "grad_norm": 1.546875,
      "learning_rate": 0.00048797812585360187,
      "loss": 5.0422,
      "mean_token_accuracy": 0.19020529240369796,
      "num_tokens": 24665523.0,
      "step": 13410
    },
    {
      "entropy": 5.320910215377808,
      "epoch": 1.120531239558971,
      "grad_norm": 2.34375,
      "learning_rate": 0.00048796852317028024,
      "loss": 5.0044,
      "mean_token_accuracy": 0.19939696788787842,
      "num_tokens": 24675618.0,
      "step": 13415
    },
    {
      "entropy": 5.263426780700684,
      "epoch": 1.1209488807216839,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004879589167586882,
      "loss": 5.0608,
      "mean_token_accuracy": 0.19728926122188567,
      "num_tokens": 24684922.0,
      "step": 13420
    },
    {
      "entropy": 5.206707382202149,
      "epoch": 1.121366521884397,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048794930661899414,
      "loss": 5.0792,
      "mean_token_accuracy": 0.19284846782684326,
      "num_tokens": 24693791.0,
      "step": 13425
    },
    {
      "entropy": 5.213493537902832,
      "epoch": 1.12178416304711,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048793969275136593,
      "loss": 4.9884,
      "mean_token_accuracy": 0.19988018721342088,
      "num_tokens": 24702661.0,
      "step": 13430
    },
    {
      "entropy": 5.229031181335449,
      "epoch": 1.122201804209823,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004879300751559722,
      "loss": 5.0581,
      "mean_token_accuracy": 0.1931751325726509,
      "num_tokens": 24711019.0,
      "step": 13435
    },
    {
      "entropy": 5.296931648254395,
      "epoch": 1.122619445372536,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004879204538329812,
      "loss": 5.093,
      "mean_token_accuracy": 0.19077321738004685,
      "num_tokens": 24720456.0,
      "step": 13440
    },
    {
      "entropy": 5.226991891860962,
      "epoch": 1.1230370865352488,
      "grad_norm": 2.0,
      "learning_rate": 0.0004879108287825614,
      "loss": 5.0419,
      "mean_token_accuracy": 0.2036843180656433,
      "num_tokens": 24729395.0,
      "step": 13445
    },
    {
      "entropy": 5.215022993087769,
      "epoch": 1.1234547276979618,
      "grad_norm": 1.421875,
      "learning_rate": 0.00048790120000488115,
      "loss": 5.0159,
      "mean_token_accuracy": 0.20135164856910706,
      "num_tokens": 24738543.0,
      "step": 13450
    },
    {
      "entropy": 5.282989263534546,
      "epoch": 1.1238723688606749,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004878915675001092,
      "loss": 5.0202,
      "mean_token_accuracy": 0.1959486797451973,
      "num_tokens": 24747860.0,
      "step": 13455
    },
    {
      "entropy": 5.124939489364624,
      "epoch": 1.124290010023388,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004878819312684141,
      "loss": 4.9877,
      "mean_token_accuracy": 0.19528626203536986,
      "num_tokens": 24757457.0,
      "step": 13460
    },
    {
      "entropy": 5.298675298690796,
      "epoch": 1.124707651186101,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004878722913099645,
      "loss": 5.0563,
      "mean_token_accuracy": 0.19277603477239608,
      "num_tokens": 24766844.0,
      "step": 13465
    },
    {
      "entropy": 5.287698030471802,
      "epoch": 1.125125292348814,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004878626476249293,
      "loss": 5.1077,
      "mean_token_accuracy": 0.18728177696466447,
      "num_tokens": 24776041.0,
      "step": 13470
    },
    {
      "entropy": 5.255358076095581,
      "epoch": 1.125542933511527,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004878530002134772,
      "loss": 5.0662,
      "mean_token_accuracy": 0.2004278928041458,
      "num_tokens": 24785515.0,
      "step": 13475
    },
    {
      "entropy": 5.350152587890625,
      "epoch": 1.1259605746742398,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004878433490757771,
      "loss": 5.1022,
      "mean_token_accuracy": 0.20173366069793702,
      "num_tokens": 24795251.0,
      "step": 13480
    },
    {
      "entropy": 5.188641834259033,
      "epoch": 1.1263782158369529,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004878336942119979,
      "loss": 5.0051,
      "mean_token_accuracy": 0.20361573994159698,
      "num_tokens": 24804672.0,
      "step": 13485
    },
    {
      "entropy": 5.27924690246582,
      "epoch": 1.126795856999666,
      "grad_norm": 1.53125,
      "learning_rate": 0.00048782403562230874,
      "loss": 5.0497,
      "mean_token_accuracy": 0.19386534839868547,
      "num_tokens": 24813520.0,
      "step": 13490
    },
    {
      "entropy": 5.225813150405884,
      "epoch": 1.127213498162379,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004878143733068787,
      "loss": 4.9132,
      "mean_token_accuracy": 0.20406006425619125,
      "num_tokens": 24821368.0,
      "step": 13495
    },
    {
      "entropy": 5.217352247238159,
      "epoch": 1.127631139325092,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004878047072658769,
      "loss": 5.0087,
      "mean_token_accuracy": 0.19408257454633712,
      "num_tokens": 24830071.0,
      "step": 13500
    },
    {
      "entropy": 5.133703994750976,
      "epoch": 1.1280487804878048,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004877950374994725,
      "loss": 4.9162,
      "mean_token_accuracy": 0.2027101680636406,
      "num_tokens": 24838982.0,
      "step": 13505
    },
    {
      "entropy": 5.268516159057617,
      "epoch": 1.1284664216505178,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004877853640078349,
      "loss": 5.1063,
      "mean_token_accuracy": 0.18830694556236266,
      "num_tokens": 24848313.0,
      "step": 13510
    },
    {
      "entropy": 5.22919716835022,
      "epoch": 1.1288840628132308,
      "grad_norm": 3.015625,
      "learning_rate": 0.0004877756867911333,
      "loss": 5.0705,
      "mean_token_accuracy": 0.1979207679629326,
      "num_tokens": 24857477.0,
      "step": 13515
    },
    {
      "entropy": 5.236756181716919,
      "epoch": 1.1293017039759439,
      "grad_norm": 1.625,
      "learning_rate": 0.0004877660058495372,
      "loss": 5.028,
      "mean_token_accuracy": 0.19710905700922013,
      "num_tokens": 24866561.0,
      "step": 13520
    },
    {
      "entropy": 5.27832932472229,
      "epoch": 1.129719345138657,
      "grad_norm": 2.125,
      "learning_rate": 0.0004877563211832161,
      "loss": 5.1203,
      "mean_token_accuracy": 0.18596124798059463,
      "num_tokens": 24876113.0,
      "step": 13525
    },
    {
      "entropy": 5.3604912757873535,
      "epoch": 1.13013698630137,
      "grad_norm": 2.359375,
      "learning_rate": 0.00048774663279233947,
      "loss": 5.1763,
      "mean_token_accuracy": 0.19035030901432037,
      "num_tokens": 24885974.0,
      "step": 13530
    },
    {
      "entropy": 5.262937355041504,
      "epoch": 1.130554627464083,
      "grad_norm": 1.671875,
      "learning_rate": 0.00048773694067707693,
      "loss": 5.0132,
      "mean_token_accuracy": 0.2005309209227562,
      "num_tokens": 24895046.0,
      "step": 13535
    },
    {
      "entropy": 5.284610652923584,
      "epoch": 1.1309722686267958,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048772724483759825,
      "loss": 5.1498,
      "mean_token_accuracy": 0.19009493738412858,
      "num_tokens": 24903315.0,
      "step": 13540
    },
    {
      "entropy": 5.291896152496338,
      "epoch": 1.1313899097895088,
      "grad_norm": 1.5234375,
      "learning_rate": 0.000487717545274073,
      "loss": 5.1529,
      "mean_token_accuracy": 0.19181631952524186,
      "num_tokens": 24912556.0,
      "step": 13545
    },
    {
      "entropy": 5.339671230316162,
      "epoch": 1.1318075509522219,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004877078419866711,
      "loss": 5.0539,
      "mean_token_accuracy": 0.19764002710580825,
      "num_tokens": 24921854.0,
      "step": 13550
    },
    {
      "entropy": 5.263269662857056,
      "epoch": 1.1322251921149349,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00048769813497556236,
      "loss": 5.0025,
      "mean_token_accuracy": 0.20629232823848725,
      "num_tokens": 24930691.0,
      "step": 13555
    },
    {
      "entropy": 5.2161369800567625,
      "epoch": 1.132642833277648,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004876884242409168,
      "loss": 5.0934,
      "mean_token_accuracy": 0.18924072831869126,
      "num_tokens": 24940573.0,
      "step": 13560
    },
    {
      "entropy": 5.234211015701294,
      "epoch": 1.1330604744403607,
      "grad_norm": 1.9765625,
      "learning_rate": 0.00048767870978290434,
      "loss": 4.9839,
      "mean_token_accuracy": 0.19487117677927018,
      "num_tokens": 24950902.0,
      "step": 13565
    },
    {
      "entropy": 5.221404790878296,
      "epoch": 1.1334781156030738,
      "grad_norm": 2.0,
      "learning_rate": 0.0004876689916016951,
      "loss": 5.0121,
      "mean_token_accuracy": 0.19780682474374772,
      "num_tokens": 24960055.0,
      "step": 13570
    },
    {
      "entropy": 5.260713815689087,
      "epoch": 1.1338957567657868,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004876592696974591,
      "loss": 5.0285,
      "mean_token_accuracy": 0.1954428642988205,
      "num_tokens": 24969896.0,
      "step": 13575
    },
    {
      "entropy": 5.26274528503418,
      "epoch": 1.1343133979284998,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004876495440703667,
      "loss": 5.003,
      "mean_token_accuracy": 0.19974747598171233,
      "num_tokens": 24978467.0,
      "step": 13580
    },
    {
      "entropy": 5.2191413879394535,
      "epoch": 1.1347310390912129,
      "grad_norm": 2.234375,
      "learning_rate": 0.0004876398147205879,
      "loss": 5.0495,
      "mean_token_accuracy": 0.20350105911493302,
      "num_tokens": 24987038.0,
      "step": 13585
    },
    {
      "entropy": 5.359581089019775,
      "epoch": 1.135148680253926,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004876300816482933,
      "loss": 5.1063,
      "mean_token_accuracy": 0.1911684662103653,
      "num_tokens": 24997228.0,
      "step": 13590
    },
    {
      "entropy": 5.244908857345581,
      "epoch": 1.135566321416639,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004876203448536532,
      "loss": 4.9827,
      "mean_token_accuracy": 0.19602168798446656,
      "num_tokens": 25005784.0,
      "step": 13595
    },
    {
      "entropy": 5.2006336688995365,
      "epoch": 1.1359839625793517,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004876106043368381,
      "loss": 5.1017,
      "mean_token_accuracy": 0.1919988989830017,
      "num_tokens": 25015863.0,
      "step": 13600
    },
    {
      "entropy": 5.102196550369262,
      "epoch": 1.1364016037420648,
      "grad_norm": 2.171875,
      "learning_rate": 0.0004876008600980183,
      "loss": 4.9283,
      "mean_token_accuracy": 0.20608919560909272,
      "num_tokens": 25025391.0,
      "step": 13605
    },
    {
      "entropy": 5.3048131465911865,
      "epoch": 1.1368192449047778,
      "grad_norm": 1.734375,
      "learning_rate": 0.00048759111213736455,
      "loss": 5.1365,
      "mean_token_accuracy": 0.19356234818696977,
      "num_tokens": 25035387.0,
      "step": 13610
    },
    {
      "entropy": 5.209079885482788,
      "epoch": 1.1372368860674908,
      "grad_norm": 1.75,
      "learning_rate": 0.0004875813604550476,
      "loss": 4.8821,
      "mean_token_accuracy": 0.20501834154129028,
      "num_tokens": 25044485.0,
      "step": 13615
    },
    {
      "entropy": 5.1651664733886715,
      "epoch": 1.1376545272302039,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004875716050512379,
      "loss": 5.0246,
      "mean_token_accuracy": 0.19918796718120574,
      "num_tokens": 25053079.0,
      "step": 13620
    },
    {
      "entropy": 5.314040851593018,
      "epoch": 1.1380721683929167,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004875618459261065,
      "loss": 5.1145,
      "mean_token_accuracy": 0.19197810292243958,
      "num_tokens": 25063262.0,
      "step": 13625
    },
    {
      "entropy": 5.275153017044067,
      "epoch": 1.1384898095556297,
      "grad_norm": 2.140625,
      "learning_rate": 0.00048755208307982414,
      "loss": 5.0005,
      "mean_token_accuracy": 0.2050935998558998,
      "num_tokens": 25071720.0,
      "step": 13630
    },
    {
      "entropy": 5.151277208328247,
      "epoch": 1.1389074507183428,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004875423165125616,
      "loss": 5.0376,
      "mean_token_accuracy": 0.20022284835577012,
      "num_tokens": 25080460.0,
      "step": 13635
    },
    {
      "entropy": 5.202970933914185,
      "epoch": 1.1393250918810558,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00048753254622449006,
      "loss": 5.0435,
      "mean_token_accuracy": 0.19837188422679902,
      "num_tokens": 25089162.0,
      "step": 13640
    },
    {
      "entropy": 5.194352722167968,
      "epoch": 1.1397427330437688,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004875227722157804,
      "loss": 4.9361,
      "mean_token_accuracy": 0.19701631516218185,
      "num_tokens": 25098234.0,
      "step": 13645
    },
    {
      "entropy": 5.157417678833008,
      "epoch": 1.1401603742064819,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004875129944866039,
      "loss": 4.9937,
      "mean_token_accuracy": 0.1979946404695511,
      "num_tokens": 25107832.0,
      "step": 13650
    },
    {
      "entropy": 5.273838233947754,
      "epoch": 1.140578015369195,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004875032130371316,
      "loss": 5.0019,
      "mean_token_accuracy": 0.19866594672203064,
      "num_tokens": 25116118.0,
      "step": 13655
    },
    {
      "entropy": 5.178330183029175,
      "epoch": 1.1409956565319077,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004874934278675347,
      "loss": 4.8475,
      "mean_token_accuracy": 0.20748500227928163,
      "num_tokens": 25125025.0,
      "step": 13660
    },
    {
      "entropy": 5.181360149383545,
      "epoch": 1.1414132976946207,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004874836389779846,
      "loss": 5.1469,
      "mean_token_accuracy": 0.19253448247909546,
      "num_tokens": 25135763.0,
      "step": 13665
    },
    {
      "entropy": 5.395813989639282,
      "epoch": 1.1418309388573338,
      "grad_norm": 2.125,
      "learning_rate": 0.00048747384636865267,
      "loss": 5.167,
      "mean_token_accuracy": 0.1867660254240036,
      "num_tokens": 25145510.0,
      "step": 13670
    },
    {
      "entropy": 5.282507753372192,
      "epoch": 1.1422485800200468,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004874640500397103,
      "loss": 4.9929,
      "mean_token_accuracy": 0.19667802453041078,
      "num_tokens": 25153910.0,
      "step": 13675
    },
    {
      "entropy": 5.222048282623291,
      "epoch": 1.1426662211827598,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004874542499913289,
      "loss": 5.0309,
      "mean_token_accuracy": 0.1937414214015007,
      "num_tokens": 25163357.0,
      "step": 13680
    },
    {
      "entropy": 5.228209114074707,
      "epoch": 1.1430838623454727,
      "grad_norm": 1.71875,
      "learning_rate": 0.00048744444622368015,
      "loss": 5.0562,
      "mean_token_accuracy": 0.19904932826757432,
      "num_tokens": 25172423.0,
      "step": 13685
    },
    {
      "entropy": 5.252134704589844,
      "epoch": 1.1435015035081857,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00048743463873693576,
      "loss": 4.9744,
      "mean_token_accuracy": 0.19328828305006027,
      "num_tokens": 25181508.0,
      "step": 13690
    },
    {
      "entropy": 5.252773809432983,
      "epoch": 1.1439191446708987,
      "grad_norm": 2.28125,
      "learning_rate": 0.0004874248275312673,
      "loss": 5.0956,
      "mean_token_accuracy": 0.19804529398679732,
      "num_tokens": 25190741.0,
      "step": 13695
    },
    {
      "entropy": 5.315851306915283,
      "epoch": 1.1443367858336118,
      "grad_norm": 1.609375,
      "learning_rate": 0.00048741501260684644,
      "loss": 5.1753,
      "mean_token_accuracy": 0.18677868396043779,
      "num_tokens": 25201288.0,
      "step": 13700
    },
    {
      "entropy": 5.279872751235962,
      "epoch": 1.1447544269963248,
      "grad_norm": 1.8984375,
      "learning_rate": 0.00048740519396384515,
      "loss": 5.004,
      "mean_token_accuracy": 0.20154963433742523,
      "num_tokens": 25210111.0,
      "step": 13705
    },
    {
      "entropy": 5.247532987594605,
      "epoch": 1.1451720681590378,
      "grad_norm": 1.84375,
      "learning_rate": 0.00048739537160243524,
      "loss": 5.055,
      "mean_token_accuracy": 0.19249035269021988,
      "num_tokens": 25220061.0,
      "step": 13710
    },
    {
      "entropy": 5.269809627532959,
      "epoch": 1.1455897093217509,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00048738554552278874,
      "loss": 5.0364,
      "mean_token_accuracy": 0.19996514767408372,
      "num_tokens": 25229336.0,
      "step": 13715
    },
    {
      "entropy": 5.251982831954956,
      "epoch": 1.1460073504844637,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00048737571572507753,
      "loss": 4.9972,
      "mean_token_accuracy": 0.2053694874048233,
      "num_tokens": 25238539.0,
      "step": 13720
    },
    {
      "entropy": 5.234113550186157,
      "epoch": 1.1464249916471767,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00048736588220947384,
      "loss": 5.0408,
      "mean_token_accuracy": 0.19419148415327073,
      "num_tokens": 25247880.0,
      "step": 13725
    },
    {
      "entropy": 5.240157604217529,
      "epoch": 1.1468426328098897,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048735604497614974,
      "loss": 5.0717,
      "mean_token_accuracy": 0.19779163748025894,
      "num_tokens": 25257295.0,
      "step": 13730
    },
    {
      "entropy": 5.237091398239135,
      "epoch": 1.1472602739726028,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004873462040252775,
      "loss": 5.0633,
      "mean_token_accuracy": 0.1952240377664566,
      "num_tokens": 25266784.0,
      "step": 13735
    },
    {
      "entropy": 5.332303190231324,
      "epoch": 1.1476779151353158,
      "grad_norm": 1.375,
      "learning_rate": 0.0004873363593570293,
      "loss": 5.1335,
      "mean_token_accuracy": 0.18732615411281586,
      "num_tokens": 25276910.0,
      "step": 13740
    },
    {
      "entropy": 5.310872554779053,
      "epoch": 1.1480955562980286,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004873265109715776,
      "loss": 5.0207,
      "mean_token_accuracy": 0.2046460136771202,
      "num_tokens": 25285776.0,
      "step": 13745
    },
    {
      "entropy": 5.268605041503906,
      "epoch": 1.1485131974607417,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004873166588690947,
      "loss": 4.972,
      "mean_token_accuracy": 0.20372209548950196,
      "num_tokens": 25294319.0,
      "step": 13750
    },
    {
      "entropy": 5.186012887954712,
      "epoch": 1.1489308386234547,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004873068030497532,
      "loss": 4.9977,
      "mean_token_accuracy": 0.1955571174621582,
      "num_tokens": 25304396.0,
      "step": 13755
    },
    {
      "entropy": 5.140398216247559,
      "epoch": 1.1493484797861677,
      "grad_norm": 1.53125,
      "learning_rate": 0.00048729694351372545,
      "loss": 4.9548,
      "mean_token_accuracy": 0.20507626086473466,
      "num_tokens": 25313095.0,
      "step": 13760
    },
    {
      "entropy": 5.218362331390381,
      "epoch": 1.1497661209488808,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004872870802611842,
      "loss": 5.1552,
      "mean_token_accuracy": 0.19242347478866578,
      "num_tokens": 25322948.0,
      "step": 13765
    },
    {
      "entropy": 5.1958261013031,
      "epoch": 1.1501837621115938,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004872772132923021,
      "loss": 4.9977,
      "mean_token_accuracy": 0.19651063084602355,
      "num_tokens": 25332707.0,
      "step": 13770
    },
    {
      "entropy": 5.234188222885132,
      "epoch": 1.1506014032743068,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048726734260725186,
      "loss": 5.0138,
      "mean_token_accuracy": 0.20153516083955764,
      "num_tokens": 25342029.0,
      "step": 13775
    },
    {
      "entropy": 5.202038383483886,
      "epoch": 1.1510190444370196,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004872574682062063,
      "loss": 4.9271,
      "mean_token_accuracy": 0.21056569069623948,
      "num_tokens": 25350374.0,
      "step": 13780
    },
    {
      "entropy": 5.266944360733032,
      "epoch": 1.1514366855997327,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004872475900893382,
      "loss": 5.0873,
      "mean_token_accuracy": 0.1954547196626663,
      "num_tokens": 25359370.0,
      "step": 13785
    },
    {
      "entropy": 5.187385368347168,
      "epoch": 1.1518543267624457,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004872377082568206,
      "loss": 5.0047,
      "mean_token_accuracy": 0.2040065795183182,
      "num_tokens": 25368329.0,
      "step": 13790
    },
    {
      "entropy": 5.227781057357788,
      "epoch": 1.1522719679251587,
      "grad_norm": 1.9921875,
      "learning_rate": 0.00048722782270882643,
      "loss": 5.1348,
      "mean_token_accuracy": 0.18927734345197678,
      "num_tokens": 25377795.0,
      "step": 13795
    },
    {
      "entropy": 5.31145486831665,
      "epoch": 1.1526896090878718,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004872179334455288,
      "loss": 5.0818,
      "mean_token_accuracy": 0.19981986433267593,
      "num_tokens": 25387134.0,
      "step": 13800
    },
    {
      "entropy": 5.344278240203858,
      "epoch": 1.1531072502505846,
      "grad_norm": 1.5625,
      "learning_rate": 0.00048720804046710074,
      "loss": 5.1182,
      "mean_token_accuracy": 0.19466662555932998,
      "num_tokens": 25396320.0,
      "step": 13805
    },
    {
      "entropy": 5.265701627731323,
      "epoch": 1.1535248914132976,
      "grad_norm": 1.71875,
      "learning_rate": 0.00048719814377371553,
      "loss": 5.0656,
      "mean_token_accuracy": 0.19596205800771713,
      "num_tokens": 25404727.0,
      "step": 13810
    },
    {
      "entropy": 5.153612661361694,
      "epoch": 1.1539425325760106,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004871882433655464,
      "loss": 4.9724,
      "mean_token_accuracy": 0.1954722821712494,
      "num_tokens": 25414137.0,
      "step": 13815
    },
    {
      "entropy": 5.274288940429687,
      "epoch": 1.1543601737387237,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004871783392427666,
      "loss": 5.0482,
      "mean_token_accuracy": 0.20331157743930817,
      "num_tokens": 25423297.0,
      "step": 13820
    },
    {
      "entropy": 5.3333171844482425,
      "epoch": 1.1547778149014367,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004871684314055497,
      "loss": 5.134,
      "mean_token_accuracy": 0.1921887442469597,
      "num_tokens": 25433185.0,
      "step": 13825
    },
    {
      "entropy": 5.246662425994873,
      "epoch": 1.1551954560641498,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004871585198540689,
      "loss": 5.0566,
      "mean_token_accuracy": 0.20300233513116836,
      "num_tokens": 25441769.0,
      "step": 13830
    },
    {
      "entropy": 5.232418394088745,
      "epoch": 1.1556130972268628,
      "grad_norm": 1.984375,
      "learning_rate": 0.00048714860458849794,
      "loss": 4.9832,
      "mean_token_accuracy": 0.19768329858779907,
      "num_tokens": 25450650.0,
      "step": 13835
    },
    {
      "entropy": 5.200397729873657,
      "epoch": 1.1560307383895756,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00048713868560901013,
      "loss": 4.9986,
      "mean_token_accuracy": 0.21136417239904404,
      "num_tokens": 25458721.0,
      "step": 13840
    },
    {
      "entropy": 5.250268650054932,
      "epoch": 1.1564483795522886,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004871287629157794,
      "loss": 5.0564,
      "mean_token_accuracy": 0.19784608036279677,
      "num_tokens": 25468004.0,
      "step": 13845
    },
    {
      "entropy": 5.344364356994629,
      "epoch": 1.1568660207150017,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004871188365089793,
      "loss": 5.1821,
      "mean_token_accuracy": 0.19265294224023818,
      "num_tokens": 25478042.0,
      "step": 13850
    },
    {
      "entropy": 5.319950342178345,
      "epoch": 1.1572836618777147,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004871089063887837,
      "loss": 5.0752,
      "mean_token_accuracy": 0.19470391869544984,
      "num_tokens": 25488439.0,
      "step": 13855
    },
    {
      "entropy": 5.220446586608887,
      "epoch": 1.1577013030404277,
      "grad_norm": 4.1875,
      "learning_rate": 0.00048709897255536626,
      "loss": 5.0579,
      "mean_token_accuracy": 0.19195529222488403,
      "num_tokens": 25497928.0,
      "step": 13860
    },
    {
      "entropy": 5.250965881347656,
      "epoch": 1.1581189442031405,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004870890350089011,
      "loss": 5.0944,
      "mean_token_accuracy": 0.19550800919532776,
      "num_tokens": 25506522.0,
      "step": 13865
    },
    {
      "entropy": 5.3145708560943605,
      "epoch": 1.1585365853658536,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004870790937495621,
      "loss": 5.0795,
      "mean_token_accuracy": 0.18546274602413176,
      "num_tokens": 25516017.0,
      "step": 13870
    },
    {
      "entropy": 5.284396028518676,
      "epoch": 1.1589542265285666,
      "grad_norm": 2.171875,
      "learning_rate": 0.00048706914877752325,
      "loss": 5.0402,
      "mean_token_accuracy": 0.19727193266153337,
      "num_tokens": 25525262.0,
      "step": 13875
    },
    {
      "entropy": 5.141807603836059,
      "epoch": 1.1593718676912796,
      "grad_norm": 2.0,
      "learning_rate": 0.00048705920009295867,
      "loss": 5.0254,
      "mean_token_accuracy": 0.2004509076476097,
      "num_tokens": 25534479.0,
      "step": 13880
    },
    {
      "entropy": 5.244466781616211,
      "epoch": 1.1597895088539927,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004870492476960426,
      "loss": 5.0484,
      "mean_token_accuracy": 0.20021468102931977,
      "num_tokens": 25543338.0,
      "step": 13885
    },
    {
      "entropy": 5.381829881668091,
      "epoch": 1.1602071500167057,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004870392915869491,
      "loss": 5.1784,
      "mean_token_accuracy": 0.18872259110212325,
      "num_tokens": 25552286.0,
      "step": 13890
    },
    {
      "entropy": 5.2721415042877195,
      "epoch": 1.1606247911794187,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004870293317658526,
      "loss": 5.1039,
      "mean_token_accuracy": 0.1914939358830452,
      "num_tokens": 25561978.0,
      "step": 13895
    },
    {
      "entropy": 5.172778081893921,
      "epoch": 1.1610424323421316,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004870193682329275,
      "loss": 4.9799,
      "mean_token_accuracy": 0.20485321134328843,
      "num_tokens": 25570139.0,
      "step": 13900
    },
    {
      "entropy": 5.22913498878479,
      "epoch": 1.1614600735048446,
      "grad_norm": 2.265625,
      "learning_rate": 0.00048700940098834805,
      "loss": 5.0206,
      "mean_token_accuracy": 0.19571435898542405,
      "num_tokens": 25579761.0,
      "step": 13905
    },
    {
      "entropy": 5.255637693405151,
      "epoch": 1.1618777146675576,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004869994300322889,
      "loss": 5.0182,
      "mean_token_accuracy": 0.19819843471050264,
      "num_tokens": 25588740.0,
      "step": 13910
    },
    {
      "entropy": 5.261387205123901,
      "epoch": 1.1622953558302707,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00048698945536492454,
      "loss": 5.0487,
      "mean_token_accuracy": 0.19226504266262054,
      "num_tokens": 25597963.0,
      "step": 13915
    },
    {
      "entropy": 5.206467866897583,
      "epoch": 1.1627129969929837,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004869794769864296,
      "loss": 5.0216,
      "mean_token_accuracy": 0.19539858400821686,
      "num_tokens": 25606782.0,
      "step": 13920
    },
    {
      "entropy": 5.3103350639343265,
      "epoch": 1.1631306381556965,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0004869694948969787,
      "loss": 5.1479,
      "mean_token_accuracy": 0.1867156818509102,
      "num_tokens": 25616458.0,
      "step": 13925
    },
    {
      "entropy": 5.29889817237854,
      "epoch": 1.1635482793184095,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00048695950909674666,
      "loss": 5.0864,
      "mean_token_accuracy": 0.1868084728717804,
      "num_tokens": 25625081.0,
      "step": 13930
    },
    {
      "entropy": 5.266204833984375,
      "epoch": 1.1639659204811226,
      "grad_norm": 1.8125,
      "learning_rate": 0.00048694951958590835,
      "loss": 5.1248,
      "mean_token_accuracy": 0.19338326156139374,
      "num_tokens": 25634000.0,
      "step": 13935
    },
    {
      "entropy": 5.281264400482177,
      "epoch": 1.1643835616438356,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00048693952636463854,
      "loss": 4.9998,
      "mean_token_accuracy": 0.1972396343946457,
      "num_tokens": 25642715.0,
      "step": 13940
    },
    {
      "entropy": 5.137683963775634,
      "epoch": 1.1648012028065486,
      "grad_norm": 1.625,
      "learning_rate": 0.00048692952943311214,
      "loss": 4.8311,
      "mean_token_accuracy": 0.21715322732925416,
      "num_tokens": 25650979.0,
      "step": 13945
    },
    {
      "entropy": 5.222711229324341,
      "epoch": 1.1652188439692617,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004869195287915043,
      "loss": 5.0941,
      "mean_token_accuracy": 0.1946940928697586,
      "num_tokens": 25660860.0,
      "step": 13950
    },
    {
      "entropy": 5.165192604064941,
      "epoch": 1.1656364851319747,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00048690952443999,
      "loss": 4.8932,
      "mean_token_accuracy": 0.2030971094965935,
      "num_tokens": 25670331.0,
      "step": 13955
    },
    {
      "entropy": 5.346300268173218,
      "epoch": 1.1660541262946875,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048689951637874445,
      "loss": 5.1541,
      "mean_token_accuracy": 0.1872032731771469,
      "num_tokens": 25679276.0,
      "step": 13960
    },
    {
      "entropy": 5.305884933471679,
      "epoch": 1.1664717674574006,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004868895046079428,
      "loss": 5.0692,
      "mean_token_accuracy": 0.20190114080905913,
      "num_tokens": 25688509.0,
      "step": 13965
    },
    {
      "entropy": 5.1893651485443115,
      "epoch": 1.1668894086201136,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004868794891277602,
      "loss": 4.9101,
      "mean_token_accuracy": 0.20364498049020768,
      "num_tokens": 25697776.0,
      "step": 13970
    },
    {
      "entropy": 5.190750026702881,
      "epoch": 1.1673070497828266,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004868694699383722,
      "loss": 5.079,
      "mean_token_accuracy": 0.19715775400400162,
      "num_tokens": 25707528.0,
      "step": 13975
    },
    {
      "entropy": 5.247435808181763,
      "epoch": 1.1677246909455397,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004868594470399541,
      "loss": 5.0206,
      "mean_token_accuracy": 0.19994188696146012,
      "num_tokens": 25716568.0,
      "step": 13980
    },
    {
      "entropy": 5.2728743076324465,
      "epoch": 1.1681423321082525,
      "grad_norm": 1.640625,
      "learning_rate": 0.00048684942043268137,
      "loss": 5.0711,
      "mean_token_accuracy": 0.19502151906490325,
      "num_tokens": 25726441.0,
      "step": 13985
    },
    {
      "entropy": 5.213754510879516,
      "epoch": 1.1685599732709655,
      "grad_norm": 1.8828125,
      "learning_rate": 0.00048683939011672953,
      "loss": 5.0452,
      "mean_token_accuracy": 0.20185276120901108,
      "num_tokens": 25735386.0,
      "step": 13990
    },
    {
      "entropy": 5.337268829345703,
      "epoch": 1.1689776144336785,
      "grad_norm": 1.859375,
      "learning_rate": 0.00048682935609227414,
      "loss": 5.0175,
      "mean_token_accuracy": 0.19358722418546676,
      "num_tokens": 25744758.0,
      "step": 13995
    },
    {
      "entropy": 5.191312408447265,
      "epoch": 1.1693952555963916,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00048681931835949083,
      "loss": 5.0614,
      "mean_token_accuracy": 0.19836339354515076,
      "num_tokens": 25753836.0,
      "step": 14000
    },
    {
      "entropy": 5.219338989257812,
      "epoch": 1.1698128967591046,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004868092769185555,
      "loss": 5.0901,
      "mean_token_accuracy": 0.18236420899629593,
      "num_tokens": 25763302.0,
      "step": 14005
    },
    {
      "entropy": 5.28340220451355,
      "epoch": 1.1702305379218176,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00048679923176964384,
      "loss": 4.9571,
      "mean_token_accuracy": 0.1990026131272316,
      "num_tokens": 25772204.0,
      "step": 14010
    },
    {
      "entropy": 5.259957218170166,
      "epoch": 1.1706481790845307,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004867891829129316,
      "loss": 5.0282,
      "mean_token_accuracy": 0.19278707653284072,
      "num_tokens": 25781762.0,
      "step": 14015
    },
    {
      "entropy": 5.213589239120483,
      "epoch": 1.1710658202472435,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004867791303485948,
      "loss": 5.0931,
      "mean_token_accuracy": 0.1855207771062851,
      "num_tokens": 25791573.0,
      "step": 14020
    },
    {
      "entropy": 5.21425199508667,
      "epoch": 1.1714834614099565,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004867690740768094,
      "loss": 5.0422,
      "mean_token_accuracy": 0.2028242751955986,
      "num_tokens": 25800777.0,
      "step": 14025
    },
    {
      "entropy": 5.219263553619385,
      "epoch": 1.1719011025726696,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048675901409775147,
      "loss": 5.0003,
      "mean_token_accuracy": 0.19964272379875184,
      "num_tokens": 25809145.0,
      "step": 14030
    },
    {
      "entropy": 5.128988742828369,
      "epoch": 1.1723187437353826,
      "grad_norm": 1.5625,
      "learning_rate": 0.00048674895041159714,
      "loss": 4.9279,
      "mean_token_accuracy": 0.20562102645635605,
      "num_tokens": 25817884.0,
      "step": 14035
    },
    {
      "entropy": 5.263066196441651,
      "epoch": 1.1727363848980956,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00048673888301852247,
      "loss": 5.0606,
      "mean_token_accuracy": 0.20210368782281876,
      "num_tokens": 25826953.0,
      "step": 14040
    },
    {
      "entropy": 5.305962133407593,
      "epoch": 1.1731540260608084,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004867288119187039,
      "loss": 5.0771,
      "mean_token_accuracy": 0.1973983973264694,
      "num_tokens": 25835597.0,
      "step": 14045
    },
    {
      "entropy": 5.233417606353759,
      "epoch": 1.1735716672235215,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004867187371123175,
      "loss": 5.0872,
      "mean_token_accuracy": 0.19305889010429383,
      "num_tokens": 25845072.0,
      "step": 14050
    },
    {
      "entropy": 5.214178276062012,
      "epoch": 1.1739893083862345,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00048670865859953984,
      "loss": 4.9595,
      "mean_token_accuracy": 0.20093905180692673,
      "num_tokens": 25854902.0,
      "step": 14055
    },
    {
      "entropy": 5.2834234714508055,
      "epoch": 1.1744069495489475,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004866985763805472,
      "loss": 5.1152,
      "mean_token_accuracy": 0.19185637533664704,
      "num_tokens": 25863807.0,
      "step": 14060
    },
    {
      "entropy": 5.261442613601685,
      "epoch": 1.1748245907116606,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004866884904555162,
      "loss": 5.0999,
      "mean_token_accuracy": 0.19316508173942565,
      "num_tokens": 25873709.0,
      "step": 14065
    },
    {
      "entropy": 5.2690423965454105,
      "epoch": 1.1752422318743736,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004866784008246233,
      "loss": 5.0655,
      "mean_token_accuracy": 0.19769184589385985,
      "num_tokens": 25883425.0,
      "step": 14070
    },
    {
      "entropy": 5.217103290557861,
      "epoch": 1.1756598730370866,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004866683074880453,
      "loss": 4.9527,
      "mean_token_accuracy": 0.20166141539812088,
      "num_tokens": 25892317.0,
      "step": 14075
    },
    {
      "entropy": 5.199564599990845,
      "epoch": 1.1760775141997994,
      "grad_norm": 1.546875,
      "learning_rate": 0.00048665821044595884,
      "loss": 5.022,
      "mean_token_accuracy": 0.19604338109493255,
      "num_tokens": 25901733.0,
      "step": 14080
    },
    {
      "entropy": 5.1453604221344,
      "epoch": 1.1764951553625125,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004866481096985405,
      "loss": 4.9916,
      "mean_token_accuracy": 0.20075306296348572,
      "num_tokens": 25911574.0,
      "step": 14085
    },
    {
      "entropy": 5.244553947448731,
      "epoch": 1.1769127965252255,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004866380052459673,
      "loss": 5.0878,
      "mean_token_accuracy": 0.19271158576011657,
      "num_tokens": 25921152.0,
      "step": 14090
    },
    {
      "entropy": 5.344291496276855,
      "epoch": 1.1773304376879385,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004866278970884161,
      "loss": 5.0464,
      "mean_token_accuracy": 0.19440077394247054,
      "num_tokens": 25930455.0,
      "step": 14095
    },
    {
      "entropy": 5.208934497833252,
      "epoch": 1.1777480788506516,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00048661778522606376,
      "loss": 4.9856,
      "mean_token_accuracy": 0.193764266371727,
      "num_tokens": 25939298.0,
      "step": 14100
    },
    {
      "entropy": 5.199110603332519,
      "epoch": 1.1781657200133644,
      "grad_norm": 1.8828125,
      "learning_rate": 0.00048660766965908736,
      "loss": 5.0486,
      "mean_token_accuracy": 0.19498568922281265,
      "num_tokens": 25948426.0,
      "step": 14105
    },
    {
      "entropy": 5.274981164932251,
      "epoch": 1.1785833611760774,
      "grad_norm": 1.765625,
      "learning_rate": 0.00048659755038766406,
      "loss": 5.0364,
      "mean_token_accuracy": 0.19440480917692185,
      "num_tokens": 25958469.0,
      "step": 14110
    },
    {
      "entropy": 5.223921489715576,
      "epoch": 1.1790010023387905,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00048658742741197086,
      "loss": 4.9311,
      "mean_token_accuracy": 0.20939817428588867,
      "num_tokens": 25967173.0,
      "step": 14115
    },
    {
      "entropy": 5.292565822601318,
      "epoch": 1.1794186435015035,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00048657730073218517,
      "loss": 5.074,
      "mean_token_accuracy": 0.1931035563349724,
      "num_tokens": 25976474.0,
      "step": 14120
    },
    {
      "entropy": 5.273281669616699,
      "epoch": 1.1798362846642165,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004865671703484841,
      "loss": 5.0371,
      "mean_token_accuracy": 0.19877732396125794,
      "num_tokens": 25984745.0,
      "step": 14125
    },
    {
      "entropy": 5.226378726959228,
      "epoch": 1.1802539258269296,
      "grad_norm": 1.421875,
      "learning_rate": 0.00048655703626104493,
      "loss": 4.9661,
      "mean_token_accuracy": 0.20514098256826402,
      "num_tokens": 25993265.0,
      "step": 14130
    },
    {
      "entropy": 5.281987428665161,
      "epoch": 1.1806715669896426,
      "grad_norm": 1.8984375,
      "learning_rate": 0.00048654689847004535,
      "loss": 5.0154,
      "mean_token_accuracy": 0.19895192235708237,
      "num_tokens": 26002390.0,
      "step": 14135
    },
    {
      "entropy": 5.331594181060791,
      "epoch": 1.1810892081523554,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004865367569756626,
      "loss": 5.1704,
      "mean_token_accuracy": 0.20136239528656005,
      "num_tokens": 26013086.0,
      "step": 14140
    },
    {
      "entropy": 5.308658695220947,
      "epoch": 1.1815068493150684,
      "grad_norm": 2.125,
      "learning_rate": 0.00048652661177807426,
      "loss": 5.051,
      "mean_token_accuracy": 0.19561164379119872,
      "num_tokens": 26022119.0,
      "step": 14145
    },
    {
      "entropy": 5.273245573043823,
      "epoch": 1.1819244904777815,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00048651646287745803,
      "loss": 5.0894,
      "mean_token_accuracy": 0.19708743840456008,
      "num_tokens": 26031805.0,
      "step": 14150
    },
    {
      "entropy": 5.253956317901611,
      "epoch": 1.1823421316404945,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00048650631027399155,
      "loss": 5.0497,
      "mean_token_accuracy": 0.19077426344156265,
      "num_tokens": 26040965.0,
      "step": 14155
    },
    {
      "entropy": 5.239748907089234,
      "epoch": 1.1827597728032075,
      "grad_norm": 1.5625,
      "learning_rate": 0.00048649615396785246,
      "loss": 5.0453,
      "mean_token_accuracy": 0.1952982500195503,
      "num_tokens": 26049618.0,
      "step": 14160
    },
    {
      "entropy": 5.246959638595581,
      "epoch": 1.1831774139659206,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00048648599395921863,
      "loss": 4.9952,
      "mean_token_accuracy": 0.1976350486278534,
      "num_tokens": 26058335.0,
      "step": 14165
    },
    {
      "entropy": 5.134028005599975,
      "epoch": 1.1835950551286334,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00048647583024826786,
      "loss": 4.8813,
      "mean_token_accuracy": 0.20854868590831757,
      "num_tokens": 26067361.0,
      "step": 14170
    },
    {
      "entropy": 5.290346717834472,
      "epoch": 1.1840126962913464,
      "grad_norm": 1.78125,
      "learning_rate": 0.00048646566283517824,
      "loss": 5.1163,
      "mean_token_accuracy": 0.18856153190135955,
      "num_tokens": 26077401.0,
      "step": 14175
    },
    {
      "entropy": 5.2107587337493895,
      "epoch": 1.1844303374540595,
      "grad_norm": 1.40625,
      "learning_rate": 0.00048645549172012756,
      "loss": 4.9323,
      "mean_token_accuracy": 0.20594017207622528,
      "num_tokens": 26086422.0,
      "step": 14180
    },
    {
      "entropy": 5.226652812957764,
      "epoch": 1.1848479786167725,
      "grad_norm": 1.453125,
      "learning_rate": 0.000486445316903294,
      "loss": 5.093,
      "mean_token_accuracy": 0.19316518008708955,
      "num_tokens": 26095974.0,
      "step": 14185
    },
    {
      "entropy": 5.216159725189209,
      "epoch": 1.1852656197794855,
      "grad_norm": 2.265625,
      "learning_rate": 0.0004864351383848556,
      "loss": 4.9799,
      "mean_token_accuracy": 0.20305544286966323,
      "num_tokens": 26104782.0,
      "step": 14190
    },
    {
      "entropy": 5.321679925918579,
      "epoch": 1.1856832609421986,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004864249561649906,
      "loss": 5.0609,
      "mean_token_accuracy": 0.1967729389667511,
      "num_tokens": 26113769.0,
      "step": 14195
    },
    {
      "entropy": 5.275844764709473,
      "epoch": 1.1861009021049114,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004864147702438773,
      "loss": 5.0869,
      "mean_token_accuracy": 0.19221366494894027,
      "num_tokens": 26123759.0,
      "step": 14200
    },
    {
      "entropy": 5.262006759643555,
      "epoch": 1.1865185432676244,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000486404580621694,
      "loss": 4.9716,
      "mean_token_accuracy": 0.2061242163181305,
      "num_tokens": 26133392.0,
      "step": 14205
    },
    {
      "entropy": 5.310753917694091,
      "epoch": 1.1869361844303374,
      "grad_norm": 1.4375,
      "learning_rate": 0.000486394387298619,
      "loss": 5.1246,
      "mean_token_accuracy": 0.19420162588357925,
      "num_tokens": 26142102.0,
      "step": 14210
    },
    {
      "entropy": 5.205290746688843,
      "epoch": 1.1873538255930505,
      "grad_norm": 1.671875,
      "learning_rate": 0.00048638419027483075,
      "loss": 5.0075,
      "mean_token_accuracy": 0.20034173429012297,
      "num_tokens": 26151504.0,
      "step": 14215
    },
    {
      "entropy": 5.228722381591797,
      "epoch": 1.1877714667557635,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004863739895505079,
      "loss": 5.1417,
      "mean_token_accuracy": 0.18748248368501663,
      "num_tokens": 26160717.0,
      "step": 14220
    },
    {
      "entropy": 5.280159616470337,
      "epoch": 1.1881891079184765,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048636378512582884,
      "loss": 5.0075,
      "mean_token_accuracy": 0.20164833217859268,
      "num_tokens": 26169788.0,
      "step": 14225
    },
    {
      "entropy": 5.251110553741455,
      "epoch": 1.1886067490811894,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00048635357700097236,
      "loss": 4.9861,
      "mean_token_accuracy": 0.19518522024154664,
      "num_tokens": 26178904.0,
      "step": 14230
    },
    {
      "entropy": 5.2481200218200685,
      "epoch": 1.1890243902439024,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004863433651761171,
      "loss": 5.0839,
      "mean_token_accuracy": 0.19130893647670746,
      "num_tokens": 26188832.0,
      "step": 14235
    },
    {
      "entropy": 5.303866910934448,
      "epoch": 1.1894420314066154,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004863331496514418,
      "loss": 5.125,
      "mean_token_accuracy": 0.19609452933073043,
      "num_tokens": 26199003.0,
      "step": 14240
    },
    {
      "entropy": 5.3465320587158205,
      "epoch": 1.1898596725693285,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048632293042712545,
      "loss": 4.9876,
      "mean_token_accuracy": 0.19727510809898377,
      "num_tokens": 26209001.0,
      "step": 14245
    },
    {
      "entropy": 5.259861421585083,
      "epoch": 1.1902773137320415,
      "grad_norm": 2.0,
      "learning_rate": 0.00048631270750334674,
      "loss": 5.0346,
      "mean_token_accuracy": 0.19983138740062714,
      "num_tokens": 26217834.0,
      "step": 14250
    },
    {
      "entropy": 5.296390008926392,
      "epoch": 1.1906949548947545,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00048630248088028474,
      "loss": 5.0814,
      "mean_token_accuracy": 0.198896424472332,
      "num_tokens": 26226935.0,
      "step": 14255
    },
    {
      "entropy": 5.273751354217529,
      "epoch": 1.1911125960574673,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004862922505581185,
      "loss": 5.0748,
      "mean_token_accuracy": 0.19478984773159028,
      "num_tokens": 26236874.0,
      "step": 14260
    },
    {
      "entropy": 5.286432695388794,
      "epoch": 1.1915302372201804,
      "grad_norm": 1.4609375,
      "learning_rate": 0.000486282016537027,
      "loss": 5.0167,
      "mean_token_accuracy": 0.20358933061361312,
      "num_tokens": 26246361.0,
      "step": 14265
    },
    {
      "entropy": 5.1630449295043945,
      "epoch": 1.1919478783828934,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048627177881718954,
      "loss": 4.9614,
      "mean_token_accuracy": 0.20389124900102615,
      "num_tokens": 26255316.0,
      "step": 14270
    },
    {
      "entropy": 5.332671308517456,
      "epoch": 1.1923655195456064,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004862615373987853,
      "loss": 5.1057,
      "mean_token_accuracy": 0.19028071612119674,
      "num_tokens": 26265250.0,
      "step": 14275
    },
    {
      "entropy": 5.184759044647217,
      "epoch": 1.1927831607083195,
      "grad_norm": 2.53125,
      "learning_rate": 0.0004862512922819935,
      "loss": 4.8892,
      "mean_token_accuracy": 0.20815268903970718,
      "num_tokens": 26273858.0,
      "step": 14280
    },
    {
      "entropy": 5.295985984802246,
      "epoch": 1.1932008018710325,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004862410434669936,
      "loss": 5.1348,
      "mean_token_accuracy": 0.19313061088323594,
      "num_tokens": 26283536.0,
      "step": 14285
    },
    {
      "entropy": 5.25338830947876,
      "epoch": 1.1936184430337453,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048623079095396496,
      "loss": 4.9637,
      "mean_token_accuracy": 0.20267538875341415,
      "num_tokens": 26291761.0,
      "step": 14290
    },
    {
      "entropy": 5.219271087646485,
      "epoch": 1.1940360841964583,
      "grad_norm": 1.640625,
      "learning_rate": 0.000486220534743087,
      "loss": 4.9893,
      "mean_token_accuracy": 0.19486655592918395,
      "num_tokens": 26301082.0,
      "step": 14295
    },
    {
      "entropy": 5.1791120052337645,
      "epoch": 1.1944537253591714,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004862102748345394,
      "loss": 5.0746,
      "mean_token_accuracy": 0.1926334634423256,
      "num_tokens": 26310424.0,
      "step": 14300
    },
    {
      "entropy": 5.334938192367554,
      "epoch": 1.1948713665218844,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048620001122850164,
      "loss": 5.1347,
      "mean_token_accuracy": 0.19323744028806686,
      "num_tokens": 26319442.0,
      "step": 14305
    },
    {
      "entropy": 5.221412086486817,
      "epoch": 1.1952890076845974,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004861897439251535,
      "loss": 4.9385,
      "mean_token_accuracy": 0.2043040007352829,
      "num_tokens": 26327757.0,
      "step": 14310
    },
    {
      "entropy": 5.232567405700683,
      "epoch": 1.1957066488473105,
      "grad_norm": 1.40625,
      "learning_rate": 0.00048617947292467473,
      "loss": 5.0058,
      "mean_token_accuracy": 0.19671137779951095,
      "num_tokens": 26337327.0,
      "step": 14315
    },
    {
      "entropy": 5.265955829620362,
      "epoch": 1.1961242900100233,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00048616919822724506,
      "loss": 5.0879,
      "mean_token_accuracy": 0.19598034471273423,
      "num_tokens": 26345821.0,
      "step": 14320
    },
    {
      "entropy": 5.3195405960083,
      "epoch": 1.1965419311727363,
      "grad_norm": 1.578125,
      "learning_rate": 0.00048615891983304436,
      "loss": 5.0565,
      "mean_token_accuracy": 0.19755673706531524,
      "num_tokens": 26355423.0,
      "step": 14325
    },
    {
      "entropy": 5.207876825332642,
      "epoch": 1.1969595723354494,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004861486377422526,
      "loss": 4.9293,
      "mean_token_accuracy": 0.20696480125188826,
      "num_tokens": 26363770.0,
      "step": 14330
    },
    {
      "entropy": 5.173492097854615,
      "epoch": 1.1973772134981624,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004861383519550498,
      "loss": 4.9758,
      "mean_token_accuracy": 0.20179972797632217,
      "num_tokens": 26373530.0,
      "step": 14335
    },
    {
      "entropy": 5.3210162162780765,
      "epoch": 1.1977948546608754,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048612806247161594,
      "loss": 5.1457,
      "mean_token_accuracy": 0.18498797863721847,
      "num_tokens": 26382960.0,
      "step": 14340
    },
    {
      "entropy": 5.18615403175354,
      "epoch": 1.1982124958235885,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004861177692921313,
      "loss": 4.9231,
      "mean_token_accuracy": 0.20626206398010255,
      "num_tokens": 26392105.0,
      "step": 14345
    },
    {
      "entropy": 5.267859983444214,
      "epoch": 1.1986301369863013,
      "grad_norm": 1.546875,
      "learning_rate": 0.00048610747241677595,
      "loss": 5.067,
      "mean_token_accuracy": 0.1969328537583351,
      "num_tokens": 26401780.0,
      "step": 14350
    },
    {
      "entropy": 5.315733909606934,
      "epoch": 1.1990477781490143,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004860971718457302,
      "loss": 5.0282,
      "mean_token_accuracy": 0.19934705197811126,
      "num_tokens": 26410290.0,
      "step": 14355
    },
    {
      "entropy": 5.284405708312988,
      "epoch": 1.1994654193117273,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004860868675791743,
      "loss": 5.0691,
      "mean_token_accuracy": 0.20041172206401825,
      "num_tokens": 26420591.0,
      "step": 14360
    },
    {
      "entropy": 5.2831847190856935,
      "epoch": 1.1998830604744404,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00048607655961728876,
      "loss": 5.0567,
      "mean_token_accuracy": 0.19233104437589646,
      "num_tokens": 26429453.0,
      "step": 14365
    },
    {
      "entropy": 5.336340379714966,
      "epoch": 1.2003007016371534,
      "grad_norm": 1.75,
      "learning_rate": 0.000486066247960254,
      "loss": 5.0583,
      "mean_token_accuracy": 0.1985298439860344,
      "num_tokens": 26439111.0,
      "step": 14370
    },
    {
      "entropy": 5.230135250091553,
      "epoch": 1.2007183427998664,
      "grad_norm": 1.6875,
      "learning_rate": 0.00048605593260825044,
      "loss": 4.9772,
      "mean_token_accuracy": 0.2071394681930542,
      "num_tokens": 26448071.0,
      "step": 14375
    },
    {
      "entropy": 5.346202945709228,
      "epoch": 1.2011359839625793,
      "grad_norm": 1.640625,
      "learning_rate": 0.00048604561356145875,
      "loss": 5.1603,
      "mean_token_accuracy": 0.19762424379587173,
      "num_tokens": 26457226.0,
      "step": 14380
    },
    {
      "entropy": 5.329900598526001,
      "epoch": 1.2015536251252923,
      "grad_norm": 1.640625,
      "learning_rate": 0.00048603529082005955,
      "loss": 5.0445,
      "mean_token_accuracy": 0.20417539030313492,
      "num_tokens": 26466052.0,
      "step": 14385
    },
    {
      "entropy": 5.283273220062256,
      "epoch": 1.2019712662880053,
      "grad_norm": 1.84375,
      "learning_rate": 0.00048602496438423354,
      "loss": 5.0607,
      "mean_token_accuracy": 0.1960564061999321,
      "num_tokens": 26476102.0,
      "step": 14390
    },
    {
      "entropy": 5.2628437042236325,
      "epoch": 1.2023889074507184,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004860146342541616,
      "loss": 5.0512,
      "mean_token_accuracy": 0.19589365422725677,
      "num_tokens": 26485775.0,
      "step": 14395
    },
    {
      "entropy": 5.27186689376831,
      "epoch": 1.2028065486134314,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004860043004300244,
      "loss": 5.0049,
      "mean_token_accuracy": 0.19548733681440353,
      "num_tokens": 26495268.0,
      "step": 14400
    },
    {
      "entropy": 5.280046796798706,
      "epoch": 1.2032241897761444,
      "grad_norm": 1.75,
      "learning_rate": 0.000485993962912003,
      "loss": 5.0676,
      "mean_token_accuracy": 0.1996158704161644,
      "num_tokens": 26505253.0,
      "step": 14405
    },
    {
      "entropy": 5.252045440673828,
      "epoch": 1.2036418309388572,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004859836217002783,
      "loss": 5.0191,
      "mean_token_accuracy": 0.19658263027668,
      "num_tokens": 26513609.0,
      "step": 14410
    },
    {
      "entropy": 5.294670963287354,
      "epoch": 1.2040594721015703,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004859732767950313,
      "loss": 5.0913,
      "mean_token_accuracy": 0.19001831710338593,
      "num_tokens": 26523565.0,
      "step": 14415
    },
    {
      "entropy": 5.24935884475708,
      "epoch": 1.2044771132642833,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048596292819644315,
      "loss": 4.9764,
      "mean_token_accuracy": 0.20162741243839263,
      "num_tokens": 26532554.0,
      "step": 14420
    },
    {
      "entropy": 5.285177898406983,
      "epoch": 1.2048947544269963,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000485952575904695,
      "loss": 5.1114,
      "mean_token_accuracy": 0.1942128524184227,
      "num_tokens": 26541474.0,
      "step": 14425
    },
    {
      "entropy": 5.192542695999146,
      "epoch": 1.2053123955897094,
      "grad_norm": 2.265625,
      "learning_rate": 0.00048594221991996814,
      "loss": 4.9873,
      "mean_token_accuracy": 0.19751390367746352,
      "num_tokens": 26551095.0,
      "step": 14430
    },
    {
      "entropy": 5.2169252872467045,
      "epoch": 1.2057300367524224,
      "grad_norm": 1.71875,
      "learning_rate": 0.00048593186024244373,
      "loss": 4.9889,
      "mean_token_accuracy": 0.20066019594669343,
      "num_tokens": 26560144.0,
      "step": 14435
    },
    {
      "entropy": 5.219994401931762,
      "epoch": 1.2061476779151352,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004859214968723033,
      "loss": 4.9061,
      "mean_token_accuracy": 0.20562017560005189,
      "num_tokens": 26568983.0,
      "step": 14440
    },
    {
      "entropy": 5.211545515060425,
      "epoch": 1.2065653190778483,
      "grad_norm": 2.046875,
      "learning_rate": 0.00048591112980972813,
      "loss": 4.9511,
      "mean_token_accuracy": 0.20544698536396028,
      "num_tokens": 26577229.0,
      "step": 14445
    },
    {
      "entropy": 5.2767199039459225,
      "epoch": 1.2069829602405613,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004859007590548997,
      "loss": 5.0145,
      "mean_token_accuracy": 0.1980360820889473,
      "num_tokens": 26585567.0,
      "step": 14450
    },
    {
      "entropy": 5.2383928298950195,
      "epoch": 1.2074006014032743,
      "grad_norm": 2.625,
      "learning_rate": 0.00048589038460799974,
      "loss": 5.0553,
      "mean_token_accuracy": 0.19946762621402742,
      "num_tokens": 26594656.0,
      "step": 14455
    },
    {
      "entropy": 5.222034358978272,
      "epoch": 1.2078182425659874,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004858800064692097,
      "loss": 5.0357,
      "mean_token_accuracy": 0.18815708309412002,
      "num_tokens": 26604002.0,
      "step": 14460
    },
    {
      "entropy": 5.302719020843506,
      "epoch": 1.2082358837287004,
      "grad_norm": 1.75,
      "learning_rate": 0.00048586962463871124,
      "loss": 5.0332,
      "mean_token_accuracy": 0.19918496757745743,
      "num_tokens": 26613376.0,
      "step": 14465
    },
    {
      "entropy": 5.454264116287232,
      "epoch": 1.2086535248914134,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048585923911668625,
      "loss": 5.1787,
      "mean_token_accuracy": 0.19431419670581818,
      "num_tokens": 26623117.0,
      "step": 14470
    },
    {
      "entropy": 5.270261955261231,
      "epoch": 1.2090711660541262,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048584884990331644,
      "loss": 5.0861,
      "mean_token_accuracy": 0.19490847140550613,
      "num_tokens": 26632855.0,
      "step": 14475
    },
    {
      "entropy": 5.126923990249634,
      "epoch": 1.2094888072168393,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00048583845699878375,
      "loss": 4.8432,
      "mean_token_accuracy": 0.21043019145727157,
      "num_tokens": 26641753.0,
      "step": 14480
    },
    {
      "entropy": 5.204107141494751,
      "epoch": 1.2099064483795523,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004858280604032701,
      "loss": 4.9648,
      "mean_token_accuracy": 0.2037617802619934,
      "num_tokens": 26650278.0,
      "step": 14485
    },
    {
      "entropy": 5.30127444267273,
      "epoch": 1.2103240895422653,
      "grad_norm": 1.515625,
      "learning_rate": 0.00048581766011695745,
      "loss": 5.0417,
      "mean_token_accuracy": 0.20112627148628234,
      "num_tokens": 26658891.0,
      "step": 14490
    },
    {
      "entropy": 5.276368856430054,
      "epoch": 1.2107417307049784,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048580725614002785,
      "loss": 5.0765,
      "mean_token_accuracy": 0.19140923917293548,
      "num_tokens": 26668197.0,
      "step": 14495
    },
    {
      "entropy": 5.311325311660767,
      "epoch": 1.2111593718676912,
      "grad_norm": 2.21875,
      "learning_rate": 0.00048579684847266353,
      "loss": 5.1097,
      "mean_token_accuracy": 0.1925310179591179,
      "num_tokens": 26676796.0,
      "step": 14500
    },
    {
      "entropy": 5.224845218658447,
      "epoch": 1.2115770130304042,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004857864371150467,
      "loss": 4.9909,
      "mean_token_accuracy": 0.19985577166080476,
      "num_tokens": 26686237.0,
      "step": 14505
    },
    {
      "entropy": 5.189446878433228,
      "epoch": 1.2119946541931172,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00048577602206735943,
      "loss": 4.9889,
      "mean_token_accuracy": 0.20008378326892853,
      "num_tokens": 26695931.0,
      "step": 14510
    },
    {
      "entropy": 5.177141094207764,
      "epoch": 1.2124122953558303,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004857656033297843,
      "loss": 4.963,
      "mean_token_accuracy": 0.20287638753652573,
      "num_tokens": 26704326.0,
      "step": 14515
    },
    {
      "entropy": 5.2049376487731935,
      "epoch": 1.2128299365185433,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048575518090250347,
      "loss": 4.8932,
      "mean_token_accuracy": 0.2081867516040802,
      "num_tokens": 26713000.0,
      "step": 14520
    },
    {
      "entropy": 5.251620483398438,
      "epoch": 1.2132475776812564,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004857447547856996,
      "loss": 5.0272,
      "mean_token_accuracy": 0.1953184887766838,
      "num_tokens": 26722014.0,
      "step": 14525
    },
    {
      "entropy": 5.255597114562988,
      "epoch": 1.2136652188439694,
      "grad_norm": 2.203125,
      "learning_rate": 0.0004857343249795552,
      "loss": 4.9974,
      "mean_token_accuracy": 0.20104152113199233,
      "num_tokens": 26730975.0,
      "step": 14530
    },
    {
      "entropy": 5.173093938827515,
      "epoch": 1.2140828600066822,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004857238914842526,
      "loss": 5.0949,
      "mean_token_accuracy": 0.19209624230861663,
      "num_tokens": 26740087.0,
      "step": 14535
    },
    {
      "entropy": 5.227970123291016,
      "epoch": 1.2145005011693952,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00048571345429997473,
      "loss": 5.0488,
      "mean_token_accuracy": 0.1932145982980728,
      "num_tokens": 26749820.0,
      "step": 14540
    },
    {
      "entropy": 5.332805633544922,
      "epoch": 1.2149181423321083,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048570301342690416,
      "loss": 5.1236,
      "mean_token_accuracy": 0.19495101273059845,
      "num_tokens": 26759850.0,
      "step": 14545
    },
    {
      "entropy": 5.173327445983887,
      "epoch": 1.2153357834948213,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004856925688652238,
      "loss": 4.8751,
      "mean_token_accuracy": 0.20684108287096023,
      "num_tokens": 26769519.0,
      "step": 14550
    },
    {
      "entropy": 5.28520450592041,
      "epoch": 1.2157534246575343,
      "grad_norm": 1.9140625,
      "learning_rate": 0.00048568212061511634,
      "loss": 5.0735,
      "mean_token_accuracy": 0.1874800518155098,
      "num_tokens": 26779999.0,
      "step": 14555
    },
    {
      "entropy": 5.267046737670898,
      "epoch": 1.2161710658202471,
      "grad_norm": 1.8359375,
      "learning_rate": 0.00048567166867676475,
      "loss": 5.0244,
      "mean_token_accuracy": 0.19029269069433213,
      "num_tokens": 26788903.0,
      "step": 14560
    },
    {
      "entropy": 5.329600286483765,
      "epoch": 1.2165887069829602,
      "grad_norm": 1.78125,
      "learning_rate": 0.00048566121305035207,
      "loss": 5.1302,
      "mean_token_accuracy": 0.1884729191660881,
      "num_tokens": 26799161.0,
      "step": 14565
    },
    {
      "entropy": 5.240751695632935,
      "epoch": 1.2170063481456732,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00048565075373606126,
      "loss": 4.957,
      "mean_token_accuracy": 0.20473628640174865,
      "num_tokens": 26807887.0,
      "step": 14570
    },
    {
      "entropy": 5.195491981506348,
      "epoch": 1.2174239893083862,
      "grad_norm": 1.453125,
      "learning_rate": 0.00048564029073407545,
      "loss": 4.9491,
      "mean_token_accuracy": 0.20670594722032548,
      "num_tokens": 26816466.0,
      "step": 14575
    },
    {
      "entropy": 5.237527084350586,
      "epoch": 1.2178416304710993,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004856298240445778,
      "loss": 5.1147,
      "mean_token_accuracy": 0.19479836970567704,
      "num_tokens": 26825847.0,
      "step": 14580
    },
    {
      "entropy": 5.336928415298462,
      "epoch": 1.2182592716338123,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004856193536677515,
      "loss": 5.0718,
      "mean_token_accuracy": 0.18687526732683182,
      "num_tokens": 26835610.0,
      "step": 14585
    },
    {
      "entropy": 5.297607517242431,
      "epoch": 1.2186769127965253,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00048560887960377995,
      "loss": 4.9967,
      "mean_token_accuracy": 0.1963230311870575,
      "num_tokens": 26844319.0,
      "step": 14590
    },
    {
      "entropy": 5.225645971298218,
      "epoch": 1.2190945539592382,
      "grad_norm": 1.5,
      "learning_rate": 0.00048559840185284647,
      "loss": 5.0869,
      "mean_token_accuracy": 0.194437675178051,
      "num_tokens": 26853508.0,
      "step": 14595
    },
    {
      "entropy": 5.204660224914551,
      "epoch": 1.2195121951219512,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00048558792041513446,
      "loss": 5.0164,
      "mean_token_accuracy": 0.19990974217653273,
      "num_tokens": 26863268.0,
      "step": 14600
    },
    {
      "entropy": 5.263304948806763,
      "epoch": 1.2199298362846642,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004855774352908273,
      "loss": 5.0116,
      "mean_token_accuracy": 0.20002769231796264,
      "num_tokens": 26871879.0,
      "step": 14605
    },
    {
      "entropy": 5.256465148925781,
      "epoch": 1.2203474774473773,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004855669464801088,
      "loss": 5.1303,
      "mean_token_accuracy": 0.1939929485321045,
      "num_tokens": 26880920.0,
      "step": 14610
    },
    {
      "entropy": 5.301485872268676,
      "epoch": 1.2207651186100903,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00048555645398316237,
      "loss": 5.0955,
      "mean_token_accuracy": 0.19757025986909865,
      "num_tokens": 26890020.0,
      "step": 14615
    },
    {
      "entropy": 5.309092807769775,
      "epoch": 1.221182759772803,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0004855459578001718,
      "loss": 5.0423,
      "mean_token_accuracy": 0.18931961357593535,
      "num_tokens": 26899486.0,
      "step": 14620
    },
    {
      "entropy": 5.2220399379730225,
      "epoch": 1.2216004009355161,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004855354579313208,
      "loss": 5.0435,
      "mean_token_accuracy": 0.19913164228200914,
      "num_tokens": 26908298.0,
      "step": 14625
    },
    {
      "entropy": 5.343021106719971,
      "epoch": 1.2220180420982292,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004855249543767932,
      "loss": 5.0554,
      "mean_token_accuracy": 0.196397040784359,
      "num_tokens": 26917601.0,
      "step": 14630
    },
    {
      "entropy": 5.158879613876342,
      "epoch": 1.2224356832609422,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048551444713677285,
      "loss": 4.9244,
      "mean_token_accuracy": 0.21192069947719575,
      "num_tokens": 26926368.0,
      "step": 14635
    },
    {
      "entropy": 5.2148185729980465,
      "epoch": 1.2228533244236552,
      "grad_norm": 1.65625,
      "learning_rate": 0.00048550393621144374,
      "loss": 5.0681,
      "mean_token_accuracy": 0.19163237065076827,
      "num_tokens": 26935424.0,
      "step": 14640
    },
    {
      "entropy": 5.192430114746093,
      "epoch": 1.2232709655863683,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004854934216009898,
      "loss": 4.9471,
      "mean_token_accuracy": 0.2084062561392784,
      "num_tokens": 26944906.0,
      "step": 14645
    },
    {
      "entropy": 5.118523263931275,
      "epoch": 1.2236886067490813,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004854829033055951,
      "loss": 4.9144,
      "mean_token_accuracy": 0.20900702178478242,
      "num_tokens": 26953709.0,
      "step": 14650
    },
    {
      "entropy": 5.132123947143555,
      "epoch": 1.2241062479117941,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004854723813254439,
      "loss": 4.9962,
      "mean_token_accuracy": 0.1944580078125,
      "num_tokens": 26962726.0,
      "step": 14655
    },
    {
      "entropy": 5.249856042861938,
      "epoch": 1.2245238890745072,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0004854618556607203,
      "loss": 5.0989,
      "mean_token_accuracy": 0.19041819870471954,
      "num_tokens": 26972519.0,
      "step": 14660
    },
    {
      "entropy": 5.216292095184326,
      "epoch": 1.2249415302372202,
      "grad_norm": 1.5,
      "learning_rate": 0.00048545132631160856,
      "loss": 4.9898,
      "mean_token_accuracy": 0.19517933577299118,
      "num_tokens": 26981251.0,
      "step": 14665
    },
    {
      "entropy": 5.226133108139038,
      "epoch": 1.2253591713999332,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004854407932782931,
      "loss": 4.987,
      "mean_token_accuracy": 0.19706180393695832,
      "num_tokens": 26990707.0,
      "step": 14670
    },
    {
      "entropy": 5.21843843460083,
      "epoch": 1.2257768125626463,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004854302565609582,
      "loss": 4.9781,
      "mean_token_accuracy": 0.19611961394548416,
      "num_tokens": 26999364.0,
      "step": 14675
    },
    {
      "entropy": 5.264277410507202,
      "epoch": 1.226194453725359,
      "grad_norm": 2.203125,
      "learning_rate": 0.00048541971615978833,
      "loss": 5.0883,
      "mean_token_accuracy": 0.1883155643939972,
      "num_tokens": 27008757.0,
      "step": 14680
    },
    {
      "entropy": 5.214770936965943,
      "epoch": 1.226612094888072,
      "grad_norm": 1.796875,
      "learning_rate": 0.00048540917207496806,
      "loss": 4.9683,
      "mean_token_accuracy": 0.2043226957321167,
      "num_tokens": 27017244.0,
      "step": 14685
    },
    {
      "entropy": 5.249055814743042,
      "epoch": 1.2270297360507851,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00048539862430668204,
      "loss": 5.0567,
      "mean_token_accuracy": 0.19698273688554763,
      "num_tokens": 27026527.0,
      "step": 14690
    },
    {
      "entropy": 5.241149091720581,
      "epoch": 1.2274473772134982,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00048538807285511474,
      "loss": 4.9801,
      "mean_token_accuracy": 0.20637072771787643,
      "num_tokens": 27034890.0,
      "step": 14695
    },
    {
      "entropy": 5.217625617980957,
      "epoch": 1.2278650183762112,
      "grad_norm": 2.328125,
      "learning_rate": 0.00048537751772045103,
      "loss": 4.9711,
      "mean_token_accuracy": 0.20424241572618484,
      "num_tokens": 27044122.0,
      "step": 14700
    },
    {
      "entropy": 5.228095197677613,
      "epoch": 1.2282826595389242,
      "grad_norm": 2.265625,
      "learning_rate": 0.0004853669589028757,
      "loss": 5.0237,
      "mean_token_accuracy": 0.19509965777397156,
      "num_tokens": 27053929.0,
      "step": 14705
    },
    {
      "entropy": 5.207892227172851,
      "epoch": 1.2287003007016373,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004853563964025734,
      "loss": 4.9875,
      "mean_token_accuracy": 0.19774501621723176,
      "num_tokens": 27062914.0,
      "step": 14710
    },
    {
      "entropy": 5.186336088180542,
      "epoch": 1.22911794186435,
      "grad_norm": 1.921875,
      "learning_rate": 0.00048534583021972926,
      "loss": 5.0391,
      "mean_token_accuracy": 0.19481796473264695,
      "num_tokens": 27072121.0,
      "step": 14715
    },
    {
      "entropy": 5.152697896957397,
      "epoch": 1.2295355830270631,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004853352603545281,
      "loss": 4.984,
      "mean_token_accuracy": 0.195831735432148,
      "num_tokens": 27081753.0,
      "step": 14720
    },
    {
      "entropy": 5.238401079177857,
      "epoch": 1.2299532241897762,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004853246868071551,
      "loss": 5.0387,
      "mean_token_accuracy": 0.19170365035533904,
      "num_tokens": 27090942.0,
      "step": 14725
    },
    {
      "entropy": 5.297724437713623,
      "epoch": 1.2303708653524892,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048531410957779533,
      "loss": 5.0563,
      "mean_token_accuracy": 0.19707200825214385,
      "num_tokens": 27100620.0,
      "step": 14730
    },
    {
      "entropy": 5.344259119033813,
      "epoch": 1.2307885065152022,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004853035286666338,
      "loss": 5.0829,
      "mean_token_accuracy": 0.19380706399679185,
      "num_tokens": 27110370.0,
      "step": 14735
    },
    {
      "entropy": 5.163407039642334,
      "epoch": 1.231206147677915,
      "grad_norm": 1.8359375,
      "learning_rate": 0.000485292944073856,
      "loss": 4.9972,
      "mean_token_accuracy": 0.20362750589847564,
      "num_tokens": 27119163.0,
      "step": 14740
    },
    {
      "entropy": 5.165553760528565,
      "epoch": 1.231623788840628,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000485282355799647,
      "loss": 4.949,
      "mean_token_accuracy": 0.20361098200082778,
      "num_tokens": 27127855.0,
      "step": 14745
    },
    {
      "entropy": 5.275580787658692,
      "epoch": 1.232041430003341,
      "grad_norm": 1.25,
      "learning_rate": 0.00048527176384419217,
      "loss": 5.0151,
      "mean_token_accuracy": 0.1959287479519844,
      "num_tokens": 27136932.0,
      "step": 14750
    },
    {
      "entropy": 5.192290925979615,
      "epoch": 1.2324590711660541,
      "grad_norm": 1.859375,
      "learning_rate": 0.00048526116820767713,
      "loss": 4.9991,
      "mean_token_accuracy": 0.19824511259794236,
      "num_tokens": 27146371.0,
      "step": 14755
    },
    {
      "entropy": 5.19253191947937,
      "epoch": 1.2328767123287672,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004852505688902872,
      "loss": 5.0163,
      "mean_token_accuracy": 0.19734574109315872,
      "num_tokens": 27156473.0,
      "step": 14760
    },
    {
      "entropy": 5.329074192047119,
      "epoch": 1.2332943534914802,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004852399658922079,
      "loss": 5.1386,
      "mean_token_accuracy": 0.19202239215373992,
      "num_tokens": 27165915.0,
      "step": 14765
    },
    {
      "entropy": 5.201338863372802,
      "epoch": 1.2337119946541932,
      "grad_norm": 1.625,
      "learning_rate": 0.0004852293592136251,
      "loss": 4.9416,
      "mean_token_accuracy": 0.20602401196956635,
      "num_tokens": 27174792.0,
      "step": 14770
    },
    {
      "entropy": 5.238711023330689,
      "epoch": 1.234129635816906,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004852187488547242,
      "loss": 5.0087,
      "mean_token_accuracy": 0.20378874391317367,
      "num_tokens": 27183785.0,
      "step": 14775
    },
    {
      "entropy": 5.296892690658569,
      "epoch": 1.234547276979619,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00048520813481569116,
      "loss": 5.1732,
      "mean_token_accuracy": 0.19059200435876847,
      "num_tokens": 27193599.0,
      "step": 14780
    },
    {
      "entropy": 5.313228416442871,
      "epoch": 1.2349649181423321,
      "grad_norm": 2.28125,
      "learning_rate": 0.0004851975170967116,
      "loss": 5.1062,
      "mean_token_accuracy": 0.19528853446245192,
      "num_tokens": 27202191.0,
      "step": 14785
    },
    {
      "entropy": 5.236551427841187,
      "epoch": 1.2353825593050451,
      "grad_norm": 3.328125,
      "learning_rate": 0.0004851868956979715,
      "loss": 5.0646,
      "mean_token_accuracy": 0.19559020251035691,
      "num_tokens": 27210833.0,
      "step": 14790
    },
    {
      "entropy": 5.178174829483032,
      "epoch": 1.2358002004677582,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004851762706196568,
      "loss": 4.9879,
      "mean_token_accuracy": 0.19414181262254715,
      "num_tokens": 27220053.0,
      "step": 14795
    },
    {
      "entropy": 5.312853002548218,
      "epoch": 1.236217841630471,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048516564186195353,
      "loss": 5.103,
      "mean_token_accuracy": 0.1941021278500557,
      "num_tokens": 27229370.0,
      "step": 14800
    },
    {
      "entropy": 5.135332250595093,
      "epoch": 1.236635482793184,
      "grad_norm": 1.25,
      "learning_rate": 0.00048515500942504763,
      "loss": 4.9359,
      "mean_token_accuracy": 0.21032259464263917,
      "num_tokens": 27238923.0,
      "step": 14805
    },
    {
      "entropy": 5.243703079223633,
      "epoch": 1.237053123955897,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00048514437330912544,
      "loss": 5.0205,
      "mean_token_accuracy": 0.2015480801463127,
      "num_tokens": 27247929.0,
      "step": 14810
    },
    {
      "entropy": 5.2593605518341064,
      "epoch": 1.23747076511861,
      "grad_norm": 1.75,
      "learning_rate": 0.00048513373351437294,
      "loss": 5.0108,
      "mean_token_accuracy": 0.2016209989786148,
      "num_tokens": 27257455.0,
      "step": 14815
    },
    {
      "entropy": 5.215169048309326,
      "epoch": 1.2378884062813231,
      "grad_norm": 1.515625,
      "learning_rate": 0.00048512309004097653,
      "loss": 5.054,
      "mean_token_accuracy": 0.19398955702781678,
      "num_tokens": 27267718.0,
      "step": 14820
    },
    {
      "entropy": 5.19853401184082,
      "epoch": 1.2383060474440362,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004851124428891226,
      "loss": 4.9676,
      "mean_token_accuracy": 0.203886042535305,
      "num_tokens": 27276706.0,
      "step": 14825
    },
    {
      "entropy": 5.153223133087158,
      "epoch": 1.2387236886067492,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004851017920589973,
      "loss": 4.885,
      "mean_token_accuracy": 0.20217990279197692,
      "num_tokens": 27285389.0,
      "step": 14830
    },
    {
      "entropy": 5.247469043731689,
      "epoch": 1.239141329769462,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004850911375507873,
      "loss": 5.0665,
      "mean_token_accuracy": 0.2003770127892494,
      "num_tokens": 27295529.0,
      "step": 14835
    },
    {
      "entropy": 5.27239203453064,
      "epoch": 1.239558970932175,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00048508047936467907,
      "loss": 5.1181,
      "mean_token_accuracy": 0.19664446413516998,
      "num_tokens": 27305358.0,
      "step": 14840
    },
    {
      "entropy": 5.238768625259399,
      "epoch": 1.239976612094888,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048506981750085914,
      "loss": 4.9596,
      "mean_token_accuracy": 0.19923219531774522,
      "num_tokens": 27314765.0,
      "step": 14845
    },
    {
      "entropy": 5.253695821762085,
      "epoch": 1.240394253257601,
      "grad_norm": 1.515625,
      "learning_rate": 0.00048505915195951423,
      "loss": 5.004,
      "mean_token_accuracy": 0.19793737828731536,
      "num_tokens": 27324227.0,
      "step": 14850
    },
    {
      "entropy": 5.236152219772339,
      "epoch": 1.2408118944203141,
      "grad_norm": 1.515625,
      "learning_rate": 0.00048504848274083095,
      "loss": 5.0211,
      "mean_token_accuracy": 0.19453481286764146,
      "num_tokens": 27332890.0,
      "step": 14855
    },
    {
      "entropy": 5.1573714256286625,
      "epoch": 1.241229535583027,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00048503780984499625,
      "loss": 4.9448,
      "mean_token_accuracy": 0.20944852083921434,
      "num_tokens": 27342411.0,
      "step": 14860
    },
    {
      "entropy": 5.144931602478027,
      "epoch": 1.24164717674574,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004850271332721968,
      "loss": 5.0224,
      "mean_token_accuracy": 0.19907289892435073,
      "num_tokens": 27351649.0,
      "step": 14865
    },
    {
      "entropy": 5.176861095428467,
      "epoch": 1.242064817908453,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004850164530226196,
      "loss": 4.9138,
      "mean_token_accuracy": 0.20485368221998215,
      "num_tokens": 27360093.0,
      "step": 14870
    },
    {
      "entropy": 5.181537628173828,
      "epoch": 1.242482459071166,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004850057690964515,
      "loss": 4.9694,
      "mean_token_accuracy": 0.1979668214917183,
      "num_tokens": 27368593.0,
      "step": 14875
    },
    {
      "entropy": 5.1648963451385494,
      "epoch": 1.242900100233879,
      "grad_norm": 1.671875,
      "learning_rate": 0.00048499508149387973,
      "loss": 5.0576,
      "mean_token_accuracy": 0.18946534693241118,
      "num_tokens": 27378126.0,
      "step": 14880
    },
    {
      "entropy": 5.1938666820526125,
      "epoch": 1.2433177413965921,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00048498439021509124,
      "loss": 4.908,
      "mean_token_accuracy": 0.19969954192638398,
      "num_tokens": 27386225.0,
      "step": 14885
    },
    {
      "entropy": 5.307432079315186,
      "epoch": 1.2437353825593052,
      "grad_norm": 1.375,
      "learning_rate": 0.0004849736952602732,
      "loss": 5.0349,
      "mean_token_accuracy": 0.1974995717406273,
      "num_tokens": 27395629.0,
      "step": 14890
    },
    {
      "entropy": 5.250179672241211,
      "epoch": 1.244153023722018,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000484962996629613,
      "loss": 5.0501,
      "mean_token_accuracy": 0.2036457121372223,
      "num_tokens": 27404431.0,
      "step": 14895
    },
    {
      "entropy": 5.204356861114502,
      "epoch": 1.244570664884731,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00048495229432329775,
      "loss": 5.0277,
      "mean_token_accuracy": 0.20316305607557297,
      "num_tokens": 27412944.0,
      "step": 14900
    },
    {
      "entropy": 5.187578582763672,
      "epoch": 1.244988306047444,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00048494158834151485,
      "loss": 4.9506,
      "mean_token_accuracy": 0.1994410678744316,
      "num_tokens": 27422006.0,
      "step": 14905
    },
    {
      "entropy": 5.264703750610352,
      "epoch": 1.245405947210157,
      "grad_norm": 1.453125,
      "learning_rate": 0.00048493087868445176,
      "loss": 5.0348,
      "mean_token_accuracy": 0.2009270444512367,
      "num_tokens": 27431568.0,
      "step": 14910
    },
    {
      "entropy": 5.268335247039795,
      "epoch": 1.24582358837287,
      "grad_norm": 1.53125,
      "learning_rate": 0.00048492016535229594,
      "loss": 4.9967,
      "mean_token_accuracy": 0.19801871478557587,
      "num_tokens": 27440471.0,
      "step": 14915
    },
    {
      "entropy": 5.135608434677124,
      "epoch": 1.246241229535583,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004849094483452349,
      "loss": 4.944,
      "mean_token_accuracy": 0.2067255988717079,
      "num_tokens": 27450209.0,
      "step": 14920
    },
    {
      "entropy": 5.257896947860718,
      "epoch": 1.246658870698296,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004848987276634564,
      "loss": 5.1386,
      "mean_token_accuracy": 0.19279585033655167,
      "num_tokens": 27458876.0,
      "step": 14925
    },
    {
      "entropy": 5.329184579849243,
      "epoch": 1.247076511861009,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00048488800330714795,
      "loss": 5.0456,
      "mean_token_accuracy": 0.20091873407363892,
      "num_tokens": 27467526.0,
      "step": 14930
    },
    {
      "entropy": 5.1964329242706295,
      "epoch": 1.247494153023722,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00048487727527649733,
      "loss": 4.9551,
      "mean_token_accuracy": 0.1998325154185295,
      "num_tokens": 27477063.0,
      "step": 14935
    },
    {
      "entropy": 5.205545139312744,
      "epoch": 1.247911794186435,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00048486654357169246,
      "loss": 5.0541,
      "mean_token_accuracy": 0.19696759283542634,
      "num_tokens": 27486243.0,
      "step": 14940
    },
    {
      "entropy": 5.2557032108306885,
      "epoch": 1.248329435349148,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004848558081929211,
      "loss": 5.0519,
      "mean_token_accuracy": 0.19146557301282882,
      "num_tokens": 27495491.0,
      "step": 14945
    },
    {
      "entropy": 5.207174301147461,
      "epoch": 1.2487470765118611,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004848450691403712,
      "loss": 5.0338,
      "mean_token_accuracy": 0.19944650530815125,
      "num_tokens": 27505105.0,
      "step": 14950
    },
    {
      "entropy": 5.234648942947388,
      "epoch": 1.249164717674574,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004848343264142307,
      "loss": 4.9555,
      "mean_token_accuracy": 0.19827852845191957,
      "num_tokens": 27514457.0,
      "step": 14955
    },
    {
      "entropy": 5.310210514068603,
      "epoch": 1.249582358837287,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004848235800146879,
      "loss": 5.1199,
      "mean_token_accuracy": 0.1971853956580162,
      "num_tokens": 27524073.0,
      "step": 14960
    },
    {
      "entropy": 5.188165521621704,
      "epoch": 1.25,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004848128299419306,
      "loss": 4.9812,
      "mean_token_accuracy": 0.19790156185626984,
      "num_tokens": 27532840.0,
      "step": 14965
    },
    {
      "entropy": 5.29079041481018,
      "epoch": 1.250417641162713,
      "grad_norm": 1.578125,
      "learning_rate": 0.00048480207619614726,
      "loss": 5.1059,
      "mean_token_accuracy": 0.19027802497148513,
      "num_tokens": 27541775.0,
      "step": 14970
    },
    {
      "entropy": 5.282812595367432,
      "epoch": 1.250835282325426,
      "grad_norm": 1.390625,
      "learning_rate": 0.000484791318777526,
      "loss": 5.0172,
      "mean_token_accuracy": 0.20341935008764267,
      "num_tokens": 27550668.0,
      "step": 14975
    },
    {
      "entropy": 5.245224142074585,
      "epoch": 1.2512529234881389,
      "grad_norm": 1.59375,
      "learning_rate": 0.00048478055768625517,
      "loss": 5.0625,
      "mean_token_accuracy": 0.19392803758382798,
      "num_tokens": 27560255.0,
      "step": 14980
    },
    {
      "entropy": 5.188960790634155,
      "epoch": 1.251670564650852,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00048476979292252323,
      "loss": 4.9862,
      "mean_token_accuracy": 0.2017424076795578,
      "num_tokens": 27569028.0,
      "step": 14985
    },
    {
      "entropy": 5.213275194168091,
      "epoch": 1.252088205813565,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004847590244865185,
      "loss": 4.8933,
      "mean_token_accuracy": 0.20390485525131224,
      "num_tokens": 27577396.0,
      "step": 14990
    },
    {
      "entropy": 5.202198266983032,
      "epoch": 1.252505846976278,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00048474825237842964,
      "loss": 5.0668,
      "mean_token_accuracy": 0.19324394911527634,
      "num_tokens": 27586710.0,
      "step": 14995
    },
    {
      "entropy": 5.143872022628784,
      "epoch": 1.252923488138991,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000484737476598445,
      "loss": 4.9149,
      "mean_token_accuracy": 0.20065829604864122,
      "num_tokens": 27594842.0,
      "step": 15000
    },
    {
      "epoch": 1.252923488138991,
      "eval_entropy": 4.96080788796394,
      "eval_loss": 5.052180290222168,
      "eval_mean_token_accuracy": 0.20725265234389995,
      "eval_num_tokens": 27594842.0,
      "eval_runtime": 27.3356,
      "eval_samples_per_second": 1381.457,
      "eval_steps_per_second": 172.705,
      "step": 15000
    },
    {
      "entropy": 5.059685802459716,
      "epoch": 1.253341129301704,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048472669714675357,
      "loss": 4.9249,
      "mean_token_accuracy": 0.20108293145895004,
      "num_tokens": 27604661.0,
      "step": 15005
    },
    {
      "entropy": 5.15630202293396,
      "epoch": 1.253758770464417,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048471591402354374,
      "loss": 4.8571,
      "mean_token_accuracy": 0.21023858040571214,
      "num_tokens": 27614015.0,
      "step": 15010
    },
    {
      "entropy": 5.279444646835327,
      "epoch": 1.25417641162713,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048470512722900434,
      "loss": 5.1185,
      "mean_token_accuracy": 0.19190314561128616,
      "num_tokens": 27624164.0,
      "step": 15015
    },
    {
      "entropy": 5.179542684555054,
      "epoch": 1.254594052789843,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00048469433676332434,
      "loss": 4.9826,
      "mean_token_accuracy": 0.2063438966870308,
      "num_tokens": 27633993.0,
      "step": 15020
    },
    {
      "entropy": 5.237859439849854,
      "epoch": 1.255011693952556,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004846835426266926,
      "loss": 5.018,
      "mean_token_accuracy": 0.20208667367696762,
      "num_tokens": 27643772.0,
      "step": 15025
    },
    {
      "entropy": 5.228401184082031,
      "epoch": 1.255429335115269,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0004846727448192981,
      "loss": 4.9374,
      "mean_token_accuracy": 0.2009034648537636,
      "num_tokens": 27652511.0,
      "step": 15030
    },
    {
      "entropy": 5.18633770942688,
      "epoch": 1.255846976277982,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048466194334132964,
      "loss": 4.938,
      "mean_token_accuracy": 0.205113485455513,
      "num_tokens": 27661589.0,
      "step": 15035
    },
    {
      "entropy": 5.162937927246094,
      "epoch": 1.2562646174406948,
      "grad_norm": 1.421875,
      "learning_rate": 0.00048465113819297654,
      "loss": 5.006,
      "mean_token_accuracy": 0.20219608694314956,
      "num_tokens": 27670299.0,
      "step": 15040
    },
    {
      "entropy": 5.328812503814698,
      "epoch": 1.2566822586034079,
      "grad_norm": 1.53125,
      "learning_rate": 0.000484640329374428,
      "loss": 5.1339,
      "mean_token_accuracy": 0.19211086481809617,
      "num_tokens": 27679375.0,
      "step": 15045
    },
    {
      "entropy": 5.200551176071167,
      "epoch": 1.257099899766121,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00048462951688587317,
      "loss": 4.9737,
      "mean_token_accuracy": 0.21068527847528457,
      "num_tokens": 27688754.0,
      "step": 15050
    },
    {
      "entropy": 5.159639596939087,
      "epoch": 1.257517540928834,
      "grad_norm": 1.546875,
      "learning_rate": 0.00048461870072750117,
      "loss": 4.9953,
      "mean_token_accuracy": 0.1924261838197708,
      "num_tokens": 27697799.0,
      "step": 15055
    },
    {
      "entropy": 5.16965274810791,
      "epoch": 1.257935182091547,
      "grad_norm": 1.421875,
      "learning_rate": 0.00048460788089950164,
      "loss": 4.9364,
      "mean_token_accuracy": 0.20338577181100845,
      "num_tokens": 27707239.0,
      "step": 15060
    },
    {
      "entropy": 5.235404014587402,
      "epoch": 1.25835282325426,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004845970574020637,
      "loss": 4.9725,
      "mean_token_accuracy": 0.20111896544694902,
      "num_tokens": 27716027.0,
      "step": 15065
    },
    {
      "entropy": 5.185185527801513,
      "epoch": 1.258770464416973,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004845862302353771,
      "loss": 4.865,
      "mean_token_accuracy": 0.21508326828479768,
      "num_tokens": 27724849.0,
      "step": 15070
    },
    {
      "entropy": 5.176218700408936,
      "epoch": 1.2591881055796859,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00048457539939963126,
      "loss": 4.9104,
      "mean_token_accuracy": 0.20485732406377793,
      "num_tokens": 27733899.0,
      "step": 15075
    },
    {
      "entropy": 5.185537099838257,
      "epoch": 1.259605746742399,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048456456489501574,
      "loss": 5.0464,
      "mean_token_accuracy": 0.19719459414482116,
      "num_tokens": 27743967.0,
      "step": 15080
    },
    {
      "entropy": 5.217226028442383,
      "epoch": 1.260023387905112,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004845537267217202,
      "loss": 4.9665,
      "mean_token_accuracy": 0.19820189327001572,
      "num_tokens": 27753105.0,
      "step": 15085
    },
    {
      "entropy": 5.3193634986877445,
      "epoch": 1.260441029067825,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004845428848799346,
      "loss": 5.0686,
      "mean_token_accuracy": 0.2004306674003601,
      "num_tokens": 27762771.0,
      "step": 15090
    },
    {
      "entropy": 5.196981573104859,
      "epoch": 1.260858670230538,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004845320393698484,
      "loss": 4.9515,
      "mean_token_accuracy": 0.20931101441383362,
      "num_tokens": 27771653.0,
      "step": 15095
    },
    {
      "entropy": 5.155093812942505,
      "epoch": 1.2612763113932508,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004845211901916517,
      "loss": 5.0255,
      "mean_token_accuracy": 0.19518386274576188,
      "num_tokens": 27780585.0,
      "step": 15100
    },
    {
      "entropy": 5.17912917137146,
      "epoch": 1.2616939525559638,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00048451033734553444,
      "loss": 4.951,
      "mean_token_accuracy": 0.19944812804460527,
      "num_tokens": 27790625.0,
      "step": 15105
    },
    {
      "entropy": 5.430654048919678,
      "epoch": 1.2621115937186769,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004844994808316864,
      "loss": 5.2594,
      "mean_token_accuracy": 0.18385605067014693,
      "num_tokens": 27800948.0,
      "step": 15110
    },
    {
      "entropy": 5.342331409454346,
      "epoch": 1.26252923488139,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004844886206502979,
      "loss": 5.1009,
      "mean_token_accuracy": 0.19922924637794495,
      "num_tokens": 27809973.0,
      "step": 15115
    },
    {
      "entropy": 5.237545299530029,
      "epoch": 1.262946876044103,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048447775680155884,
      "loss": 4.9984,
      "mean_token_accuracy": 0.19870254695415496,
      "num_tokens": 27818738.0,
      "step": 15120
    },
    {
      "entropy": 5.176899671554565,
      "epoch": 1.263364517206816,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00048446688928565956,
      "loss": 4.98,
      "mean_token_accuracy": 0.21035105884075164,
      "num_tokens": 27827537.0,
      "step": 15125
    },
    {
      "entropy": 5.387529373168945,
      "epoch": 1.263782158369529,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004844560181027903,
      "loss": 5.1602,
      "mean_token_accuracy": 0.18868450820446014,
      "num_tokens": 27837498.0,
      "step": 15130
    },
    {
      "entropy": 5.2928235054016115,
      "epoch": 1.2641997995322418,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004844451432531412,
      "loss": 5.0505,
      "mean_token_accuracy": 0.1947311729192734,
      "num_tokens": 27847547.0,
      "step": 15135
    },
    {
      "entropy": 5.215535974502563,
      "epoch": 1.2646174406949549,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004844342647369028,
      "loss": 4.9915,
      "mean_token_accuracy": 0.20601839572191238,
      "num_tokens": 27856468.0,
      "step": 15140
    },
    {
      "entropy": 5.276471281051636,
      "epoch": 1.2650350818576679,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048442338255426555,
      "loss": 5.137,
      "mean_token_accuracy": 0.19670648425817489,
      "num_tokens": 27865324.0,
      "step": 15145
    },
    {
      "entropy": 5.214488410949707,
      "epoch": 1.265452723020381,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004844124967054199,
      "loss": 5.0237,
      "mean_token_accuracy": 0.19938111454248428,
      "num_tokens": 27875364.0,
      "step": 15150
    },
    {
      "entropy": 5.275440216064453,
      "epoch": 1.265870364183094,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004844016071905564,
      "loss": 4.9696,
      "mean_token_accuracy": 0.2087443247437477,
      "num_tokens": 27884739.0,
      "step": 15155
    },
    {
      "entropy": 5.125004243850708,
      "epoch": 1.2662880053458068,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004843907140098656,
      "loss": 4.8768,
      "mean_token_accuracy": 0.21263511776924132,
      "num_tokens": 27893235.0,
      "step": 15160
    },
    {
      "entropy": 5.115106296539307,
      "epoch": 1.2667056465085198,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004843798171635384,
      "loss": 5.028,
      "mean_token_accuracy": 0.19916052520275115,
      "num_tokens": 27902090.0,
      "step": 15165
    },
    {
      "entropy": 5.218580293655395,
      "epoch": 1.2671232876712328,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004843689166517655,
      "loss": 4.9765,
      "mean_token_accuracy": 0.1990986004471779,
      "num_tokens": 27910702.0,
      "step": 15170
    },
    {
      "entropy": 5.212385082244873,
      "epoch": 1.2675409288339459,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004843580124747377,
      "loss": 5.0085,
      "mean_token_accuracy": 0.20192957073450088,
      "num_tokens": 27919296.0,
      "step": 15175
    },
    {
      "entropy": 5.173358821868897,
      "epoch": 1.267958569996659,
      "grad_norm": 2.015625,
      "learning_rate": 0.00048434710463264585,
      "loss": 4.9963,
      "mean_token_accuracy": 0.20351825952529906,
      "num_tokens": 27929074.0,
      "step": 15180
    },
    {
      "entropy": 5.2973016738891605,
      "epoch": 1.268376211159372,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048433619312568095,
      "loss": 5.0663,
      "mean_token_accuracy": 0.19595036059617996,
      "num_tokens": 27938227.0,
      "step": 15185
    },
    {
      "entropy": 5.306293678283692,
      "epoch": 1.268793852322085,
      "grad_norm": 1.8046875,
      "learning_rate": 0.00048432527795403405,
      "loss": 5.0618,
      "mean_token_accuracy": 0.19511586129665376,
      "num_tokens": 27948050.0,
      "step": 15190
    },
    {
      "entropy": 5.178899049758911,
      "epoch": 1.2692114934847978,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004843143591178962,
      "loss": 4.9373,
      "mean_token_accuracy": 0.19839531183242798,
      "num_tokens": 27957702.0,
      "step": 15195
    },
    {
      "entropy": 5.264004755020141,
      "epoch": 1.2696291346475108,
      "grad_norm": 1.8515625,
      "learning_rate": 0.00048430343661745844,
      "loss": 5.0941,
      "mean_token_accuracy": 0.18878461569547653,
      "num_tokens": 27966085.0,
      "step": 15200
    },
    {
      "entropy": 5.2644538402557375,
      "epoch": 1.2700467758102238,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004842925104529122,
      "loss": 5.0726,
      "mean_token_accuracy": 0.1936160534620285,
      "num_tokens": 27976228.0,
      "step": 15205
    },
    {
      "entropy": 5.213149738311768,
      "epoch": 1.2704644169729369,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048428158062444857,
      "loss": 5.0061,
      "mean_token_accuracy": 0.20230662375688552,
      "num_tokens": 27984910.0,
      "step": 15210
    },
    {
      "entropy": 5.219276046752929,
      "epoch": 1.27088205813565,
      "grad_norm": 1.515625,
      "learning_rate": 0.00048427064713225894,
      "loss": 4.9842,
      "mean_token_accuracy": 0.20226027816534042,
      "num_tokens": 27993767.0,
      "step": 15215
    },
    {
      "entropy": 5.262640285491943,
      "epoch": 1.2712996992983627,
      "grad_norm": 1.453125,
      "learning_rate": 0.00048425970997653477,
      "loss": 5.0231,
      "mean_token_accuracy": 0.1960621878504753,
      "num_tokens": 28002587.0,
      "step": 15220
    },
    {
      "entropy": 5.234699964523315,
      "epoch": 1.2717173404610758,
      "grad_norm": 1.75,
      "learning_rate": 0.0004842487691574675,
      "loss": 5.0684,
      "mean_token_accuracy": 0.19308072775602342,
      "num_tokens": 28011598.0,
      "step": 15225
    },
    {
      "entropy": 5.297529554367065,
      "epoch": 1.2721349816237888,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048423782467524865,
      "loss": 5.1284,
      "mean_token_accuracy": 0.1936042055487633,
      "num_tokens": 28021321.0,
      "step": 15230
    },
    {
      "entropy": 5.249519920349121,
      "epoch": 1.2725526227865018,
      "grad_norm": 1.375,
      "learning_rate": 0.0004842268765300698,
      "loss": 5.035,
      "mean_token_accuracy": 0.20422664731740953,
      "num_tokens": 28031009.0,
      "step": 15235
    },
    {
      "entropy": 5.231356859207153,
      "epoch": 1.2729702639492149,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00048421592472212263,
      "loss": 5.0833,
      "mean_token_accuracy": 0.19535848796367644,
      "num_tokens": 28040330.0,
      "step": 15240
    },
    {
      "entropy": 5.220077085494995,
      "epoch": 1.273387905111928,
      "grad_norm": 1.9765625,
      "learning_rate": 0.00048420496925159886,
      "loss": 5.0121,
      "mean_token_accuracy": 0.19672103375196456,
      "num_tokens": 28048952.0,
      "step": 15245
    },
    {
      "entropy": 5.226531171798706,
      "epoch": 1.273805546274641,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004841940101186903,
      "loss": 5.0109,
      "mean_token_accuracy": 0.19336650669574737,
      "num_tokens": 28058803.0,
      "step": 15250
    },
    {
      "entropy": 5.253669118881225,
      "epoch": 1.2742231874373537,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004841830473235888,
      "loss": 5.0154,
      "mean_token_accuracy": 0.20033739805221557,
      "num_tokens": 28067978.0,
      "step": 15255
    },
    {
      "entropy": 5.130385828018189,
      "epoch": 1.2746408286000668,
      "grad_norm": 1.375,
      "learning_rate": 0.00048417208086648625,
      "loss": 4.9684,
      "mean_token_accuracy": 0.20443760752677917,
      "num_tokens": 28077229.0,
      "step": 15260
    },
    {
      "entropy": 5.190932130813598,
      "epoch": 1.2750584697627798,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004841611107475746,
      "loss": 5.0015,
      "mean_token_accuracy": 0.2048376277089119,
      "num_tokens": 28086316.0,
      "step": 15265
    },
    {
      "entropy": 5.365811109542847,
      "epoch": 1.2754761109254928,
      "grad_norm": 1.875,
      "learning_rate": 0.000484150136967046,
      "loss": 5.1924,
      "mean_token_accuracy": 0.18817194998264314,
      "num_tokens": 28096914.0,
      "step": 15270
    },
    {
      "entropy": 5.266621065139771,
      "epoch": 1.2758937520882059,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004841391595250925,
      "loss": 5.0583,
      "mean_token_accuracy": 0.1926616609096527,
      "num_tokens": 28106532.0,
      "step": 15275
    },
    {
      "entropy": 5.245172929763794,
      "epoch": 1.2763113932509187,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004841281784219063,
      "loss": 5.018,
      "mean_token_accuracy": 0.1983381450176239,
      "num_tokens": 28115052.0,
      "step": 15280
    },
    {
      "entropy": 5.199034404754639,
      "epoch": 1.2767290344136317,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048411719365767947,
      "loss": 4.9453,
      "mean_token_accuracy": 0.20577378571033478,
      "num_tokens": 28123571.0,
      "step": 15285
    },
    {
      "entropy": 5.123162031173706,
      "epoch": 1.2771466755763448,
      "grad_norm": 2.46875,
      "learning_rate": 0.00048410620523260455,
      "loss": 4.894,
      "mean_token_accuracy": 0.20883288830518723,
      "num_tokens": 28131973.0,
      "step": 15290
    },
    {
      "entropy": 5.2371073246002195,
      "epoch": 1.2775643167390578,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048409521314687383,
      "loss": 5.1089,
      "mean_token_accuracy": 0.19296378493309022,
      "num_tokens": 28140965.0,
      "step": 15295
    },
    {
      "entropy": 5.210053825378418,
      "epoch": 1.2779819579017708,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004840842174006797,
      "loss": 5.0797,
      "mean_token_accuracy": 0.19681747257709503,
      "num_tokens": 28150652.0,
      "step": 15300
    },
    {
      "entropy": 5.341313743591309,
      "epoch": 1.2783995990644839,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004840732179942146,
      "loss": 5.2412,
      "mean_token_accuracy": 0.18816760033369065,
      "num_tokens": 28160830.0,
      "step": 15305
    },
    {
      "entropy": 5.293534564971924,
      "epoch": 1.278817240227197,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004840622149276712,
      "loss": 4.8916,
      "mean_token_accuracy": 0.20480582863092422,
      "num_tokens": 28169423.0,
      "step": 15310
    },
    {
      "entropy": 5.188752746582031,
      "epoch": 1.2792348813899097,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004840512082012421,
      "loss": 4.9846,
      "mean_token_accuracy": 0.1988743007183075,
      "num_tokens": 28178997.0,
      "step": 15315
    },
    {
      "entropy": 5.1722527027130125,
      "epoch": 1.2796525225526227,
      "grad_norm": 1.84375,
      "learning_rate": 0.00048404019781511995,
      "loss": 4.9918,
      "mean_token_accuracy": 0.20217713713645935,
      "num_tokens": 28188014.0,
      "step": 15320
    },
    {
      "entropy": 5.237066125869751,
      "epoch": 1.2800701637153358,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00048402918376949753,
      "loss": 5.0466,
      "mean_token_accuracy": 0.19202245473861695,
      "num_tokens": 28197800.0,
      "step": 15325
    },
    {
      "entropy": 5.1956517696380615,
      "epoch": 1.2804878048780488,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004840181660645676,
      "loss": 4.9452,
      "mean_token_accuracy": 0.20003938525915146,
      "num_tokens": 28207388.0,
      "step": 15330
    },
    {
      "entropy": 5.198174476623535,
      "epoch": 1.2809054460407618,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004840071447005231,
      "loss": 5.0272,
      "mean_token_accuracy": 0.20530240386724471,
      "num_tokens": 28216774.0,
      "step": 15335
    },
    {
      "entropy": 5.247184133529663,
      "epoch": 1.2813230872034747,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000483996119677557,
      "loss": 5.0462,
      "mean_token_accuracy": 0.20051053315401077,
      "num_tokens": 28226402.0,
      "step": 15340
    },
    {
      "entropy": 5.181778287887573,
      "epoch": 1.281740728366188,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00048398509099586225,
      "loss": 5.0331,
      "mean_token_accuracy": 0.19129574596881865,
      "num_tokens": 28235890.0,
      "step": 15345
    },
    {
      "entropy": 5.190192985534668,
      "epoch": 1.2821583695289007,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048397405865563186,
      "loss": 4.973,
      "mean_token_accuracy": 0.20517520159482955,
      "num_tokens": 28244121.0,
      "step": 15350
    },
    {
      "entropy": 5.35064172744751,
      "epoch": 1.2825760106916138,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004839630226570591,
      "loss": 5.0698,
      "mean_token_accuracy": 0.19810592979192734,
      "num_tokens": 28252259.0,
      "step": 15355
    },
    {
      "entropy": 5.221536874771118,
      "epoch": 1.2829936518543268,
      "grad_norm": 1.78125,
      "learning_rate": 0.000483951983000337,
      "loss": 4.9359,
      "mean_token_accuracy": 0.20008476674556733,
      "num_tokens": 28262199.0,
      "step": 15360
    },
    {
      "entropy": 5.233777570724487,
      "epoch": 1.2834112930170398,
      "grad_norm": 1.8671875,
      "learning_rate": 0.000483940939685659,
      "loss": 5.0069,
      "mean_token_accuracy": 0.20171757340431212,
      "num_tokens": 28271120.0,
      "step": 15365
    },
    {
      "entropy": 5.23984375,
      "epoch": 1.2838289341797529,
      "grad_norm": 1.5,
      "learning_rate": 0.00048392989271321836,
      "loss": 5.0072,
      "mean_token_accuracy": 0.1996024653315544,
      "num_tokens": 28279766.0,
      "step": 15370
    },
    {
      "entropy": 5.2340188980102536,
      "epoch": 1.2842465753424657,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004839188420832084,
      "loss": 5.0136,
      "mean_token_accuracy": 0.20721420347690583,
      "num_tokens": 28288872.0,
      "step": 15375
    },
    {
      "entropy": 5.142161798477173,
      "epoch": 1.2846642165051787,
      "grad_norm": 1.375,
      "learning_rate": 0.00048390778779582266,
      "loss": 4.9616,
      "mean_token_accuracy": 0.19897125512361527,
      "num_tokens": 28297719.0,
      "step": 15380
    },
    {
      "entropy": 5.244450044631958,
      "epoch": 1.2850818576678917,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048389672985125467,
      "loss": 5.0628,
      "mean_token_accuracy": 0.19416426569223405,
      "num_tokens": 28306547.0,
      "step": 15385
    },
    {
      "entropy": 5.201297569274902,
      "epoch": 1.2854994988306048,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004838856682496979,
      "loss": 5.0237,
      "mean_token_accuracy": 0.20241621881723404,
      "num_tokens": 28316387.0,
      "step": 15390
    },
    {
      "entropy": 5.2879303932189945,
      "epoch": 1.2859171399933178,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004838746029913461,
      "loss": 5.0853,
      "mean_token_accuracy": 0.18849803656339645,
      "num_tokens": 28326642.0,
      "step": 15395
    },
    {
      "entropy": 5.236890459060669,
      "epoch": 1.2863347811560306,
      "grad_norm": 1.5390625,
      "learning_rate": 0.000483863534076393,
      "loss": 4.9825,
      "mean_token_accuracy": 0.20447689592838286,
      "num_tokens": 28335647.0,
      "step": 15400
    },
    {
      "entropy": 5.1674751281738285,
      "epoch": 1.2867524223187439,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004838524615050323,
      "loss": 5.0168,
      "mean_token_accuracy": 0.20066645592451096,
      "num_tokens": 28344536.0,
      "step": 15405
    },
    {
      "entropy": 5.247422075271606,
      "epoch": 1.2871700634814567,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004838413852774579,
      "loss": 5.0218,
      "mean_token_accuracy": 0.1975288763642311,
      "num_tokens": 28354121.0,
      "step": 15410
    },
    {
      "entropy": 5.1713231086730955,
      "epoch": 1.2875877046441697,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048383030539386355,
      "loss": 4.9804,
      "mean_token_accuracy": 0.2052040085196495,
      "num_tokens": 28362215.0,
      "step": 15415
    },
    {
      "entropy": 5.241284132003784,
      "epoch": 1.2880053458068828,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004838192218544434,
      "loss": 5.0381,
      "mean_token_accuracy": 0.20080064982175827,
      "num_tokens": 28371323.0,
      "step": 15420
    },
    {
      "entropy": 5.307083511352539,
      "epoch": 1.2884229869695958,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004838081346593915,
      "loss": 5.1017,
      "mean_token_accuracy": 0.19051435142755507,
      "num_tokens": 28380333.0,
      "step": 15425
    },
    {
      "entropy": 5.235315990447998,
      "epoch": 1.2888406281323088,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048379704380890177,
      "loss": 5.0102,
      "mean_token_accuracy": 0.1954706624150276,
      "num_tokens": 28389021.0,
      "step": 15430
    },
    {
      "entropy": 5.182530164718628,
      "epoch": 1.2892582692950216,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004837859493031685,
      "loss": 4.9367,
      "mean_token_accuracy": 0.19978894740343095,
      "num_tokens": 28398309.0,
      "step": 15435
    },
    {
      "entropy": 5.142619562149048,
      "epoch": 1.2896759104577347,
      "grad_norm": 1.421875,
      "learning_rate": 0.00048377485114238576,
      "loss": 4.8569,
      "mean_token_accuracy": 0.2065632611513138,
      "num_tokens": 28406946.0,
      "step": 15440
    },
    {
      "entropy": 5.189148378372193,
      "epoch": 1.2900935516204477,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004837637493267481,
      "loss": 5.0102,
      "mean_token_accuracy": 0.20140050500631332,
      "num_tokens": 28416561.0,
      "step": 15445
    },
    {
      "entropy": 5.177145528793335,
      "epoch": 1.2905111927831607,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004837526438564496,
      "loss": 4.8518,
      "mean_token_accuracy": 0.2204034373164177,
      "num_tokens": 28425133.0,
      "step": 15450
    },
    {
      "entropy": 5.17250428199768,
      "epoch": 1.2909288339458738,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004837415347316849,
      "loss": 5.0765,
      "mean_token_accuracy": 0.18984198421239853,
      "num_tokens": 28434282.0,
      "step": 15455
    },
    {
      "entropy": 5.2347253322601315,
      "epoch": 1.2913464751085866,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00048373042195264836,
      "loss": 5.0392,
      "mean_token_accuracy": 0.19670073240995406,
      "num_tokens": 28443149.0,
      "step": 15460
    },
    {
      "entropy": 5.304716920852661,
      "epoch": 1.2917641162712998,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004837193055195344,
      "loss": 5.1051,
      "mean_token_accuracy": 0.18857945799827575,
      "num_tokens": 28453294.0,
      "step": 15465
    },
    {
      "entropy": 5.225760841369629,
      "epoch": 1.2921817574340126,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048370818543253793,
      "loss": 4.9515,
      "mean_token_accuracy": 0.20724797248840332,
      "num_tokens": 28462518.0,
      "step": 15470
    },
    {
      "entropy": 5.170163774490357,
      "epoch": 1.2925993985967257,
      "grad_norm": 1.40625,
      "learning_rate": 0.00048369706169185334,
      "loss": 5.012,
      "mean_token_accuracy": 0.1949041083455086,
      "num_tokens": 28470674.0,
      "step": 15475
    },
    {
      "entropy": 5.217066097259521,
      "epoch": 1.2930170397594387,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048368593429767555,
      "loss": 5.0393,
      "mean_token_accuracy": 0.19544828981161116,
      "num_tokens": 28480088.0,
      "step": 15480
    },
    {
      "entropy": 5.249969625473023,
      "epoch": 1.2934346809221517,
      "grad_norm": 1.578125,
      "learning_rate": 0.00048367480325019923,
      "loss": 5.0716,
      "mean_token_accuracy": 0.19443043321371078,
      "num_tokens": 28489533.0,
      "step": 15485
    },
    {
      "entropy": 5.266088771820068,
      "epoch": 1.2938523220848648,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00048366366854961937,
      "loss": 5.0382,
      "mean_token_accuracy": 0.19626215547323228,
      "num_tokens": 28498758.0,
      "step": 15490
    },
    {
      "entropy": 5.170021295547485,
      "epoch": 1.2942699632475776,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004836525301961307,
      "loss": 4.9912,
      "mean_token_accuracy": 0.20285607129335403,
      "num_tokens": 28507633.0,
      "step": 15495
    },
    {
      "entropy": 5.133869457244873,
      "epoch": 1.2946876044102906,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004836413881899285,
      "loss": 4.9771,
      "mean_token_accuracy": 0.19845286756753922,
      "num_tokens": 28516905.0,
      "step": 15500
    },
    {
      "entropy": 5.2793288230896,
      "epoch": 1.2951052455730037,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004836302425312075,
      "loss": 5.1036,
      "mean_token_accuracy": 0.19056063890457153,
      "num_tokens": 28527419.0,
      "step": 15505
    },
    {
      "entropy": 5.1505725383758545,
      "epoch": 1.2955228867357167,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004836190932201631,
      "loss": 4.8445,
      "mean_token_accuracy": 0.21223642379045488,
      "num_tokens": 28535950.0,
      "step": 15510
    },
    {
      "entropy": 5.187836742401123,
      "epoch": 1.2959405278984297,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00048360794025699026,
      "loss": 4.9393,
      "mean_token_accuracy": 0.20249527245759963,
      "num_tokens": 28544341.0,
      "step": 15515
    },
    {
      "entropy": 5.245123720169067,
      "epoch": 1.2963581690611425,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004835967836418844,
      "loss": 5.0649,
      "mean_token_accuracy": 0.19246201813220978,
      "num_tokens": 28553805.0,
      "step": 15520
    },
    {
      "entropy": 5.220954179763794,
      "epoch": 1.2967758102238558,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004835856233750407,
      "loss": 4.9523,
      "mean_token_accuracy": 0.20517682284116745,
      "num_tokens": 28562224.0,
      "step": 15525
    },
    {
      "entropy": 5.157376384735107,
      "epoch": 1.2971934513865686,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048357445945665465,
      "loss": 4.8842,
      "mean_token_accuracy": 0.20892985612154008,
      "num_tokens": 28570771.0,
      "step": 15530
    },
    {
      "entropy": 5.25351300239563,
      "epoch": 1.2976110925492816,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004835632918869215,
      "loss": 5.1031,
      "mean_token_accuracy": 0.19254579544067382,
      "num_tokens": 28580447.0,
      "step": 15535
    },
    {
      "entropy": 5.349343109130859,
      "epoch": 1.2980287337119947,
      "grad_norm": 1.203125,
      "learning_rate": 0.000483552120666037,
      "loss": 5.1467,
      "mean_token_accuracy": 0.1970434680581093,
      "num_tokens": 28590472.0,
      "step": 15540
    },
    {
      "entropy": 5.2403614044189455,
      "epoch": 1.2984463748747077,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00048354094579419656,
      "loss": 4.9824,
      "mean_token_accuracy": 0.20229056775569915,
      "num_tokens": 28599201.0,
      "step": 15545
    },
    {
      "entropy": 5.181227397918701,
      "epoch": 1.2988640160374207,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004835297672715958,
      "loss": 5.0235,
      "mean_token_accuracy": 0.19573376178741456,
      "num_tokens": 28608499.0,
      "step": 15550
    },
    {
      "entropy": 5.240633916854859,
      "epoch": 1.2992816572001336,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048351858509843054,
      "loss": 5.0008,
      "mean_token_accuracy": 0.2075998678803444,
      "num_tokens": 28618310.0,
      "step": 15555
    },
    {
      "entropy": 5.270214033126831,
      "epoch": 1.2996992983628466,
      "grad_norm": 1.625,
      "learning_rate": 0.0004835073992748964,
      "loss": 5.0262,
      "mean_token_accuracy": 0.19287423342466353,
      "num_tokens": 28627374.0,
      "step": 15560
    },
    {
      "entropy": 5.200198650360107,
      "epoch": 1.3001169395255596,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048349620980118926,
      "loss": 4.9817,
      "mean_token_accuracy": 0.19596179723739623,
      "num_tokens": 28636331.0,
      "step": 15565
    },
    {
      "entropy": 5.206793785095215,
      "epoch": 1.3005345806882727,
      "grad_norm": 1.4375,
      "learning_rate": 0.000483485016677505,
      "loss": 4.9707,
      "mean_token_accuracy": 0.20109914541244506,
      "num_tokens": 28645006.0,
      "step": 15570
    },
    {
      "entropy": 5.203479146957397,
      "epoch": 1.3009522218509857,
      "grad_norm": 1.453125,
      "learning_rate": 0.00048347381990403957,
      "loss": 5.0115,
      "mean_token_accuracy": 0.20233114063739777,
      "num_tokens": 28653202.0,
      "step": 15575
    },
    {
      "entropy": 5.192722940444947,
      "epoch": 1.3013698630136985,
      "grad_norm": 2.046875,
      "learning_rate": 0.00048346261948098895,
      "loss": 5.012,
      "mean_token_accuracy": 0.20175158083438874,
      "num_tokens": 28662701.0,
      "step": 15580
    },
    {
      "entropy": 5.192215824127198,
      "epoch": 1.3017875041764118,
      "grad_norm": 1.5625,
      "learning_rate": 0.00048345141540854933,
      "loss": 4.9997,
      "mean_token_accuracy": 0.19889582991600036,
      "num_tokens": 28670967.0,
      "step": 15585
    },
    {
      "entropy": 5.289653730392456,
      "epoch": 1.3022051453391246,
      "grad_norm": 1.59375,
      "learning_rate": 0.00048344020768691673,
      "loss": 5.0688,
      "mean_token_accuracy": 0.1998969778418541,
      "num_tokens": 28679458.0,
      "step": 15590
    },
    {
      "entropy": 5.185567378997803,
      "epoch": 1.3026227865018376,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004834289963162874,
      "loss": 4.9602,
      "mean_token_accuracy": 0.20575502812862395,
      "num_tokens": 28688497.0,
      "step": 15595
    },
    {
      "entropy": 5.169763803482056,
      "epoch": 1.3030404276645506,
      "grad_norm": 1.625,
      "learning_rate": 0.0004834177812968576,
      "loss": 4.9928,
      "mean_token_accuracy": 0.19898176789283753,
      "num_tokens": 28697729.0,
      "step": 15600
    },
    {
      "entropy": 5.344694662094116,
      "epoch": 1.3034580688272637,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048340656262882366,
      "loss": 5.1034,
      "mean_token_accuracy": 0.19226397573947906,
      "num_tokens": 28707778.0,
      "step": 15605
    },
    {
      "entropy": 5.264089345932007,
      "epoch": 1.3038757099899767,
      "grad_norm": 1.484375,
      "learning_rate": 0.00048339534031238197,
      "loss": 5.0729,
      "mean_token_accuracy": 0.19521638602018357,
      "num_tokens": 28718010.0,
      "step": 15610
    },
    {
      "entropy": 5.2537562370300295,
      "epoch": 1.3042933511526895,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000483384114347729,
      "loss": 4.9823,
      "mean_token_accuracy": 0.2011805310845375,
      "num_tokens": 28727279.0,
      "step": 15615
    },
    {
      "entropy": 5.221255779266357,
      "epoch": 1.3047109923154026,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004833728847350613,
      "loss": 5.054,
      "mean_token_accuracy": 0.2042199432849884,
      "num_tokens": 28736510.0,
      "step": 15620
    },
    {
      "entropy": 5.288490581512451,
      "epoch": 1.3051286334781156,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048336165147457537,
      "loss": 5.0283,
      "mean_token_accuracy": 0.1996855154633522,
      "num_tokens": 28745226.0,
      "step": 15625
    },
    {
      "entropy": 5.253872108459473,
      "epoch": 1.3055462746408286,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004833504145664681,
      "loss": 4.9862,
      "mean_token_accuracy": 0.20105521380901337,
      "num_tokens": 28754052.0,
      "step": 15630
    },
    {
      "entropy": 5.189203596115112,
      "epoch": 1.3059639158035417,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00048333917401093587,
      "loss": 4.9619,
      "mean_token_accuracy": 0.20972601920366288,
      "num_tokens": 28763580.0,
      "step": 15635
    },
    {
      "entropy": 5.2106421947479244,
      "epoch": 1.3063815569662545,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00048332792980817563,
      "loss": 4.9684,
      "mean_token_accuracy": 0.20083021372556686,
      "num_tokens": 28772540.0,
      "step": 15640
    },
    {
      "entropy": 5.259501934051514,
      "epoch": 1.3067991981289677,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004833166819583843,
      "loss": 4.9551,
      "mean_token_accuracy": 0.2028495654463768,
      "num_tokens": 28782027.0,
      "step": 15645
    },
    {
      "entropy": 5.2267162799835205,
      "epoch": 1.3072168392916805,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004833054304617586,
      "loss": 4.9913,
      "mean_token_accuracy": 0.19609942883253098,
      "num_tokens": 28791724.0,
      "step": 15650
    },
    {
      "entropy": 5.14919810295105,
      "epoch": 1.3076344804543936,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00048329417531849567,
      "loss": 5.0254,
      "mean_token_accuracy": 0.19889941960573196,
      "num_tokens": 28800681.0,
      "step": 15655
    },
    {
      "entropy": 5.204528141021728,
      "epoch": 1.3080521216171066,
      "grad_norm": 1.765625,
      "learning_rate": 0.00048328291652879245,
      "loss": 5.0925,
      "mean_token_accuracy": 0.1998867318034172,
      "num_tokens": 28810121.0,
      "step": 15660
    },
    {
      "entropy": 5.214687061309815,
      "epoch": 1.3084697627798196,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00048327165409284614,
      "loss": 5.0848,
      "mean_token_accuracy": 0.19992616772651672,
      "num_tokens": 28819287.0,
      "step": 15665
    },
    {
      "entropy": 5.343918132781982,
      "epoch": 1.3088874039425327,
      "grad_norm": 1.5,
      "learning_rate": 0.0004832603880108537,
      "loss": 5.0765,
      "mean_token_accuracy": 0.19178273528814316,
      "num_tokens": 28828907.0,
      "step": 15670
    },
    {
      "entropy": 5.255748701095581,
      "epoch": 1.3093050451052455,
      "grad_norm": 1.40625,
      "learning_rate": 0.00048324911828301256,
      "loss": 5.0317,
      "mean_token_accuracy": 0.19255685657262803,
      "num_tokens": 28839087.0,
      "step": 15675
    },
    {
      "entropy": 5.265013122558594,
      "epoch": 1.3097226862679585,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004832378449095199,
      "loss": 5.1041,
      "mean_token_accuracy": 0.19614139944314957,
      "num_tokens": 28848956.0,
      "step": 15680
    },
    {
      "entropy": 5.246604633331299,
      "epoch": 1.3101403274306715,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004832265678905731,
      "loss": 5.0104,
      "mean_token_accuracy": 0.20048763900995253,
      "num_tokens": 28858105.0,
      "step": 15685
    },
    {
      "entropy": 5.27581524848938,
      "epoch": 1.3105579685933846,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004832152872263697,
      "loss": 5.1251,
      "mean_token_accuracy": 0.18931763172149657,
      "num_tokens": 28867539.0,
      "step": 15690
    },
    {
      "entropy": 5.201244306564331,
      "epoch": 1.3109756097560976,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004832040029171069,
      "loss": 4.9461,
      "mean_token_accuracy": 0.20217668116092682,
      "num_tokens": 28876536.0,
      "step": 15695
    },
    {
      "entropy": 5.2905357837677,
      "epoch": 1.3113932509188104,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048319271496298255,
      "loss": 5.0655,
      "mean_token_accuracy": 0.19102517664432525,
      "num_tokens": 28885354.0,
      "step": 15700
    },
    {
      "entropy": 5.240811586380005,
      "epoch": 1.3118108920815237,
      "grad_norm": 2.296875,
      "learning_rate": 0.00048318142336419413,
      "loss": 4.9864,
      "mean_token_accuracy": 0.20040543079376222,
      "num_tokens": 28894181.0,
      "step": 15705
    },
    {
      "entropy": 5.193211936950684,
      "epoch": 1.3122285332442365,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00048317012812093926,
      "loss": 4.9675,
      "mean_token_accuracy": 0.20521135330200196,
      "num_tokens": 28903803.0,
      "step": 15710
    },
    {
      "entropy": 5.2481475353240965,
      "epoch": 1.3126461744069495,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004831588292334157,
      "loss": 5.0691,
      "mean_token_accuracy": 0.2011494517326355,
      "num_tokens": 28913192.0,
      "step": 15715
    },
    {
      "entropy": 5.186801862716675,
      "epoch": 1.3130638155696626,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004831475267018213,
      "loss": 4.9692,
      "mean_token_accuracy": 0.20635924935340882,
      "num_tokens": 28922413.0,
      "step": 15720
    },
    {
      "entropy": 5.226707696914673,
      "epoch": 1.3134814567323756,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004831362205263538,
      "loss": 5.0418,
      "mean_token_accuracy": 0.18867288827896117,
      "num_tokens": 28933589.0,
      "step": 15725
    },
    {
      "entropy": 5.276617622375488,
      "epoch": 1.3138990978950886,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004831249107072114,
      "loss": 5.0227,
      "mean_token_accuracy": 0.20073601603507996,
      "num_tokens": 28942458.0,
      "step": 15730
    },
    {
      "entropy": 5.222417163848877,
      "epoch": 1.3143167390578014,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00048311359724459175,
      "loss": 4.9912,
      "mean_token_accuracy": 0.20586139410734178,
      "num_tokens": 28951834.0,
      "step": 15735
    },
    {
      "entropy": 5.1966859817504885,
      "epoch": 1.3147343802205145,
      "grad_norm": 2.03125,
      "learning_rate": 0.00048310228013869323,
      "loss": 5.0142,
      "mean_token_accuracy": 0.2013590857386589,
      "num_tokens": 28960885.0,
      "step": 15740
    },
    {
      "entropy": 5.243048810958863,
      "epoch": 1.3151520213832275,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004830909593897137,
      "loss": 5.0034,
      "mean_token_accuracy": 0.20456335842609405,
      "num_tokens": 28970922.0,
      "step": 15745
    },
    {
      "entropy": 5.261454296112061,
      "epoch": 1.3155696625459405,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004830796349978515,
      "loss": 5.0952,
      "mean_token_accuracy": 0.1940053731203079,
      "num_tokens": 28980873.0,
      "step": 15750
    },
    {
      "entropy": 5.276323127746582,
      "epoch": 1.3159873037086536,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004830683069633048,
      "loss": 5.0899,
      "mean_token_accuracy": 0.19063400477170944,
      "num_tokens": 28989876.0,
      "step": 15755
    },
    {
      "entropy": 5.269247674942017,
      "epoch": 1.3164049448713664,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004830569752862719,
      "loss": 5.0314,
      "mean_token_accuracy": 0.19650322198867798,
      "num_tokens": 28999480.0,
      "step": 15760
    },
    {
      "entropy": 5.2504490375518795,
      "epoch": 1.3168225860340796,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048304563996695127,
      "loss": 5.1008,
      "mean_token_accuracy": 0.19407769590616225,
      "num_tokens": 29009855.0,
      "step": 15765
    },
    {
      "entropy": 5.231771755218506,
      "epoch": 1.3172402271967925,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048303430100554126,
      "loss": 4.9194,
      "mean_token_accuracy": 0.20427149534225464,
      "num_tokens": 29018506.0,
      "step": 15770
    },
    {
      "entropy": 5.25968370437622,
      "epoch": 1.3176578683595055,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004830229584022404,
      "loss": 4.9935,
      "mean_token_accuracy": 0.1990010216832161,
      "num_tokens": 29027560.0,
      "step": 15775
    },
    {
      "entropy": 5.12415657043457,
      "epoch": 1.3180755095222185,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004830116121572472,
      "loss": 4.8947,
      "mean_token_accuracy": 0.20214923620223998,
      "num_tokens": 29036038.0,
      "step": 15780
    },
    {
      "entropy": 5.214746952056885,
      "epoch": 1.3184931506849316,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004830002622707604,
      "loss": 4.9672,
      "mean_token_accuracy": 0.20405053347349167,
      "num_tokens": 29046074.0,
      "step": 15785
    },
    {
      "entropy": 5.260791492462158,
      "epoch": 1.3189107918476446,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048298890874297866,
      "loss": 4.9723,
      "mean_token_accuracy": 0.19591133147478104,
      "num_tokens": 29055750.0,
      "step": 15790
    },
    {
      "entropy": 5.199095582962036,
      "epoch": 1.3193284330103574,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004829775515741007,
      "loss": 4.9469,
      "mean_token_accuracy": 0.20178987830877304,
      "num_tokens": 29065012.0,
      "step": 15795
    },
    {
      "entropy": 5.192213726043701,
      "epoch": 1.3197460741730704,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004829661907643253,
      "loss": 5.1727,
      "mean_token_accuracy": 0.18113635927438737,
      "num_tokens": 29074780.0,
      "step": 15800
    },
    {
      "entropy": 5.286743879318237,
      "epoch": 1.3201637153357835,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004829548263138515,
      "loss": 5.1149,
      "mean_token_accuracy": 0.195651812851429,
      "num_tokens": 29083321.0,
      "step": 15805
    },
    {
      "entropy": 5.212601900100708,
      "epoch": 1.3205813564984965,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004829434582228781,
      "loss": 4.9887,
      "mean_token_accuracy": 0.18942296206951142,
      "num_tokens": 29092110.0,
      "step": 15810
    },
    {
      "entropy": 5.131677722930908,
      "epoch": 1.3209989976612095,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004829320864916042,
      "loss": 4.9214,
      "mean_token_accuracy": 0.20175640434026718,
      "num_tokens": 29100451.0,
      "step": 15815
    },
    {
      "entropy": 5.306771564483642,
      "epoch": 1.3214166388239224,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048292071112022886,
      "loss": 5.0108,
      "mean_token_accuracy": 0.19495311826467515,
      "num_tokens": 29110468.0,
      "step": 15820
    },
    {
      "entropy": 5.237381076812744,
      "epoch": 1.3218342799866356,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004829093321089511,
      "loss": 5.059,
      "mean_token_accuracy": 0.1987290620803833,
      "num_tokens": 29119660.0,
      "step": 15825
    },
    {
      "entropy": 5.3036243438720705,
      "epoch": 1.3222519211493484,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004828979494579703,
      "loss": 5.1045,
      "mean_token_accuracy": 0.1915759935975075,
      "num_tokens": 29129145.0,
      "step": 15830
    },
    {
      "entropy": 5.2685870170593265,
      "epoch": 1.3226695623120615,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048288656316748554,
      "loss": 5.078,
      "mean_token_accuracy": 0.20158542692661285,
      "num_tokens": 29137980.0,
      "step": 15835
    },
    {
      "entropy": 5.208188247680664,
      "epoch": 1.3230872034747745,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00048287517323769634,
      "loss": 4.9585,
      "mean_token_accuracy": 0.20565597862005233,
      "num_tokens": 29147535.0,
      "step": 15840
    },
    {
      "entropy": 5.125730371475219,
      "epoch": 1.3235048446374875,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048286377966880206,
      "loss": 4.9646,
      "mean_token_accuracy": 0.19823764860630036,
      "num_tokens": 29157162.0,
      "step": 15845
    },
    {
      "entropy": 5.180935430526733,
      "epoch": 1.3239224858002006,
      "grad_norm": 1.609375,
      "learning_rate": 0.00048285238246100203,
      "loss": 4.9624,
      "mean_token_accuracy": 0.2021440401673317,
      "num_tokens": 29166542.0,
      "step": 15850
    },
    {
      "entropy": 5.2579535961151125,
      "epoch": 1.3243401269629134,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004828409816144959,
      "loss": 5.0397,
      "mean_token_accuracy": 0.19748064279556274,
      "num_tokens": 29176040.0,
      "step": 15855
    },
    {
      "entropy": 5.303733682632446,
      "epoch": 1.3247577681256264,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004828295771294831,
      "loss": 5.1144,
      "mean_token_accuracy": 0.19356649219989777,
      "num_tokens": 29184730.0,
      "step": 15860
    },
    {
      "entropy": 5.237729263305664,
      "epoch": 1.3251754092883394,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048281816900616354,
      "loss": 4.9924,
      "mean_token_accuracy": 0.20224183350801467,
      "num_tokens": 29193671.0,
      "step": 15865
    },
    {
      "entropy": 5.252899789810181,
      "epoch": 1.3255930504510525,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004828067572447367,
      "loss": 5.0259,
      "mean_token_accuracy": 0.19586314409971237,
      "num_tokens": 29203094.0,
      "step": 15870
    },
    {
      "entropy": 5.176011753082276,
      "epoch": 1.3260106916137655,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00048279534184540246,
      "loss": 5.0089,
      "mean_token_accuracy": 0.20544503778219222,
      "num_tokens": 29211678.0,
      "step": 15875
    },
    {
      "entropy": 5.2925036430358885,
      "epoch": 1.3264283327764783,
      "grad_norm": 1.625,
      "learning_rate": 0.0004827839228083606,
      "loss": 5.1556,
      "mean_token_accuracy": 0.19102041721343993,
      "num_tokens": 29220993.0,
      "step": 15880
    },
    {
      "entropy": 5.321775197982788,
      "epoch": 1.3268459739391916,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004827725001338111,
      "loss": 4.9859,
      "mean_token_accuracy": 0.20050719678401946,
      "num_tokens": 29230006.0,
      "step": 15885
    },
    {
      "entropy": 5.229358768463134,
      "epoch": 1.3272636151019044,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048276107382195393,
      "loss": 4.9541,
      "mean_token_accuracy": 0.20503713488578795,
      "num_tokens": 29238906.0,
      "step": 15890
    },
    {
      "entropy": 5.255975151062012,
      "epoch": 1.3276812562646174,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048274964387298903,
      "loss": 5.0719,
      "mean_token_accuracy": 0.2013029083609581,
      "num_tokens": 29248383.0,
      "step": 15895
    },
    {
      "entropy": 5.192051887512207,
      "epoch": 1.3280988974273304,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004827382102871166,
      "loss": 4.976,
      "mean_token_accuracy": 0.20066201239824294,
      "num_tokens": 29257999.0,
      "step": 15900
    },
    {
      "entropy": 5.205760478973389,
      "epoch": 1.3285165385900435,
      "grad_norm": 1.40625,
      "learning_rate": 0.00048272677306453665,
      "loss": 4.9498,
      "mean_token_accuracy": 0.20336118042469026,
      "num_tokens": 29267143.0,
      "step": 15905
    },
    {
      "entropy": 5.234712600708008,
      "epoch": 1.3289341797527565,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048271533220544955,
      "loss": 5.1072,
      "mean_token_accuracy": 0.1952838644385338,
      "num_tokens": 29276730.0,
      "step": 15910
    },
    {
      "entropy": 5.272205924987793,
      "epoch": 1.3293518209154693,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048270388771005555,
      "loss": 5.0953,
      "mean_token_accuracy": 0.18987479507923127,
      "num_tokens": 29285480.0,
      "step": 15915
    },
    {
      "entropy": 5.315062046051025,
      "epoch": 1.3297694620781824,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00048269243957855493,
      "loss": 5.0869,
      "mean_token_accuracy": 0.19630948901176454,
      "num_tokens": 29294927.0,
      "step": 15920
    },
    {
      "entropy": 5.257504892349243,
      "epoch": 1.3301871032408954,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004826809878111481,
      "loss": 5.0004,
      "mean_token_accuracy": 0.2019957810640335,
      "num_tokens": 29304567.0,
      "step": 15925
    },
    {
      "entropy": 5.144045495986939,
      "epoch": 1.3306047444036084,
      "grad_norm": 1.890625,
      "learning_rate": 0.00048266953240803576,
      "loss": 4.9675,
      "mean_token_accuracy": 0.20738932490348816,
      "num_tokens": 29313302.0,
      "step": 15930
    },
    {
      "entropy": 5.227485942840576,
      "epoch": 1.3310223855663215,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004826580733694182,
      "loss": 4.9804,
      "mean_token_accuracy": 0.20081841796636582,
      "num_tokens": 29322195.0,
      "step": 15935
    },
    {
      "entropy": 5.179180383682251,
      "epoch": 1.3314400267290343,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004826466106954961,
      "loss": 5.0939,
      "mean_token_accuracy": 0.19329103827476501,
      "num_tokens": 29331416.0,
      "step": 15940
    },
    {
      "entropy": 5.245913982391357,
      "epoch": 1.3318576678917475,
      "grad_norm": 1.53125,
      "learning_rate": 0.00048263514438647015,
      "loss": 4.9996,
      "mean_token_accuracy": 0.20941993594169617,
      "num_tokens": 29340497.0,
      "step": 15945
    },
    {
      "entropy": 5.1805655002594,
      "epoch": 1.3322753090544603,
      "grad_norm": 1.328125,
      "learning_rate": 0.000482623674442541,
      "loss": 4.9555,
      "mean_token_accuracy": 0.2032453551888466,
      "num_tokens": 29349306.0,
      "step": 15950
    },
    {
      "entropy": 5.24085545539856,
      "epoch": 1.3326929502171734,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004826122008639096,
      "loss": 5.0112,
      "mean_token_accuracy": 0.19845830798149108,
      "num_tokens": 29358452.0,
      "step": 15955
    },
    {
      "entropy": 5.256392908096314,
      "epoch": 1.3331105913798864,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00048260072365077675,
      "loss": 5.0747,
      "mean_token_accuracy": 0.19256907403469087,
      "num_tokens": 29367640.0,
      "step": 15960
    },
    {
      "entropy": 5.322294139862061,
      "epoch": 1.3335282325425994,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004825892428033432,
      "loss": 4.9923,
      "mean_token_accuracy": 0.1914020523428917,
      "num_tokens": 29377352.0,
      "step": 15965
    },
    {
      "entropy": 5.201574277877808,
      "epoch": 1.3339458737053125,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004825777583218102,
      "loss": 4.997,
      "mean_token_accuracy": 0.20690932273864746,
      "num_tokens": 29386802.0,
      "step": 15970
    },
    {
      "entropy": 5.096928882598877,
      "epoch": 1.3343635148680253,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004825662702063787,
      "loss": 4.8623,
      "mean_token_accuracy": 0.2130153000354767,
      "num_tokens": 29395078.0,
      "step": 15975
    },
    {
      "entropy": 5.266151475906372,
      "epoch": 1.3347811560307383,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004825547784572497,
      "loss": 4.9706,
      "mean_token_accuracy": 0.20320275127887727,
      "num_tokens": 29403837.0,
      "step": 15980
    },
    {
      "entropy": 5.2246458530426025,
      "epoch": 1.3351987971934514,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004825432830746245,
      "loss": 4.9987,
      "mean_token_accuracy": 0.19886497259140015,
      "num_tokens": 29412222.0,
      "step": 15985
    },
    {
      "entropy": 5.199143791198731,
      "epoch": 1.3356164383561644,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004825317840587044,
      "loss": 4.9202,
      "mean_token_accuracy": 0.20685300529003142,
      "num_tokens": 29421632.0,
      "step": 15990
    },
    {
      "entropy": 5.255343961715698,
      "epoch": 1.3360340795188774,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004825202814096905,
      "loss": 5.0753,
      "mean_token_accuracy": 0.19958332031965256,
      "num_tokens": 29430768.0,
      "step": 15995
    },
    {
      "entropy": 5.383226823806763,
      "epoch": 1.3364517206815905,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004825087751277843,
      "loss": 5.0982,
      "mean_token_accuracy": 0.19339663833379744,
      "num_tokens": 29440280.0,
      "step": 16000
    },
    {
      "entropy": 5.3482208251953125,
      "epoch": 1.3368693618443035,
      "grad_norm": 1.53125,
      "learning_rate": 0.00048249726521318716,
      "loss": 5.0155,
      "mean_token_accuracy": 0.19871012568473817,
      "num_tokens": 29449850.0,
      "step": 16005
    },
    {
      "entropy": 5.22251615524292,
      "epoch": 1.3372870030070163,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004824857516661008,
      "loss": 4.9584,
      "mean_token_accuracy": 0.20878487080335617,
      "num_tokens": 29459233.0,
      "step": 16010
    },
    {
      "entropy": 5.28769474029541,
      "epoch": 1.3377046441697293,
      "grad_norm": 1.40625,
      "learning_rate": 0.00048247423448672647,
      "loss": 5.0876,
      "mean_token_accuracy": 0.1893463686108589,
      "num_tokens": 29469351.0,
      "step": 16015
    },
    {
      "entropy": 5.226628828048706,
      "epoch": 1.3381222853324424,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004824627136752659,
      "loss": 5.0542,
      "mean_token_accuracy": 0.19566083550453187,
      "num_tokens": 29477653.0,
      "step": 16020
    },
    {
      "entropy": 5.211916065216064,
      "epoch": 1.3385399264951554,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004824511892319209,
      "loss": 5.0049,
      "mean_token_accuracy": 0.20014063715934755,
      "num_tokens": 29486944.0,
      "step": 16025
    },
    {
      "entropy": 5.156609535217285,
      "epoch": 1.3389575676578684,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048243966115689305,
      "loss": 4.9421,
      "mean_token_accuracy": 0.2102418452501297,
      "num_tokens": 29495791.0,
      "step": 16030
    },
    {
      "entropy": 5.332604169845581,
      "epoch": 1.3393752088205813,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004824281294503843,
      "loss": 5.211,
      "mean_token_accuracy": 0.1912085235118866,
      "num_tokens": 29505403.0,
      "step": 16035
    },
    {
      "entropy": 5.205655956268311,
      "epoch": 1.3397928499832943,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004824165941125964,
      "loss": 4.8922,
      "mean_token_accuracy": 0.20228234082460403,
      "num_tokens": 29515310.0,
      "step": 16040
    },
    {
      "entropy": 5.292618083953857,
      "epoch": 1.3402104911460073,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004824050551437315,
      "loss": 5.0746,
      "mean_token_accuracy": 0.1979966416954994,
      "num_tokens": 29524824.0,
      "step": 16045
    },
    {
      "entropy": 5.239114904403687,
      "epoch": 1.3406281323087204,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004823935125439912,
      "loss": 4.9791,
      "mean_token_accuracy": 0.1937490776181221,
      "num_tokens": 29533667.0,
      "step": 16050
    },
    {
      "entropy": 5.2572649955749515,
      "epoch": 1.3410457734714334,
      "grad_norm": 1.6796875,
      "learning_rate": 0.000482381966313578,
      "loss": 5.1094,
      "mean_token_accuracy": 0.1916586548089981,
      "num_tokens": 29541745.0,
      "step": 16055
    },
    {
      "entropy": 5.1982049465179445,
      "epoch": 1.3414634146341464,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048237041645269387,
      "loss": 4.9477,
      "mean_token_accuracy": 0.19926127642393113,
      "num_tokens": 29551004.0,
      "step": 16060
    },
    {
      "entropy": 5.126488590240479,
      "epoch": 1.3418810557968595,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004823588629615409,
      "loss": 4.9854,
      "mean_token_accuracy": 0.20674188435077667,
      "num_tokens": 29561061.0,
      "step": 16065
    },
    {
      "entropy": 5.121004867553711,
      "epoch": 1.3422986969595723,
      "grad_norm": 2.3125,
      "learning_rate": 0.00048234730584032143,
      "loss": 4.8167,
      "mean_token_accuracy": 0.2137200117111206,
      "num_tokens": 29569689.0,
      "step": 16070
    },
    {
      "entropy": 5.217213821411133,
      "epoch": 1.3427163381222853,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004823357450892379,
      "loss": 5.0672,
      "mean_token_accuracy": 0.19563427567481995,
      "num_tokens": 29579576.0,
      "step": 16075
    },
    {
      "entropy": 5.1925712585449215,
      "epoch": 1.3431339792849983,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004823241807084925,
      "loss": 4.9246,
      "mean_token_accuracy": 0.20981278270483017,
      "num_tokens": 29588531.0,
      "step": 16080
    },
    {
      "entropy": 5.272253370285034,
      "epoch": 1.3435516204477114,
      "grad_norm": 1.53125,
      "learning_rate": 0.00048231261269828776,
      "loss": 5.0278,
      "mean_token_accuracy": 0.1993756353855133,
      "num_tokens": 29596838.0,
      "step": 16085
    },
    {
      "entropy": 5.235178852081299,
      "epoch": 1.3439692616104244,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00048230104105882624,
      "loss": 5.0066,
      "mean_token_accuracy": 0.19892941266298295,
      "num_tokens": 29606769.0,
      "step": 16090
    },
    {
      "entropy": 5.2623756408691404,
      "epoch": 1.3443869027731372,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004822894657903104,
      "loss": 5.0907,
      "mean_token_accuracy": 0.1900889351963997,
      "num_tokens": 29616994.0,
      "step": 16095
    },
    {
      "entropy": 5.140792369842529,
      "epoch": 1.3448045439358502,
      "grad_norm": 1.453125,
      "learning_rate": 0.000482277886892943,
      "loss": 4.9664,
      "mean_token_accuracy": 0.20802156776189804,
      "num_tokens": 29626865.0,
      "step": 16100
    },
    {
      "entropy": 5.1955819606781,
      "epoch": 1.3452221850985633,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004822663043669267,
      "loss": 4.8983,
      "mean_token_accuracy": 0.2093164399266243,
      "num_tokens": 29636439.0,
      "step": 16105
    },
    {
      "entropy": 5.228407382965088,
      "epoch": 1.3456398262612763,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00048225471821246416,
      "loss": 4.9071,
      "mean_token_accuracy": 0.21215400993824005,
      "num_tokens": 29645019.0,
      "step": 16110
    },
    {
      "entropy": 5.232847261428833,
      "epoch": 1.3460574674239894,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004822431284297584,
      "loss": 5.1418,
      "mean_token_accuracy": 0.1980874791741371,
      "num_tokens": 29654660.0,
      "step": 16115
    },
    {
      "entropy": 5.212048101425171,
      "epoch": 1.3464751085867024,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004822315350190122,
      "loss": 5.0417,
      "mean_token_accuracy": 0.19158000200986863,
      "num_tokens": 29662880.0,
      "step": 16120
    },
    {
      "entropy": 5.324804878234863,
      "epoch": 1.3468927497494154,
      "grad_norm": 1.609375,
      "learning_rate": 0.00048221993798042856,
      "loss": 5.0447,
      "mean_token_accuracy": 0.19586261361837387,
      "num_tokens": 29672336.0,
      "step": 16125
    },
    {
      "entropy": 5.226585054397583,
      "epoch": 1.3473103909121282,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00048220833731421045,
      "loss": 4.9874,
      "mean_token_accuracy": 0.1956270456314087,
      "num_tokens": 29681495.0,
      "step": 16130
    },
    {
      "entropy": 5.240067052841186,
      "epoch": 1.3477280320748413,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004821967330205609,
      "loss": 4.9776,
      "mean_token_accuracy": 0.20323118269443513,
      "num_tokens": 29690214.0,
      "step": 16135
    },
    {
      "entropy": 5.245200061798096,
      "epoch": 1.3481456732375543,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004821851250996832,
      "loss": 5.036,
      "mean_token_accuracy": 0.194924795627594,
      "num_tokens": 29699309.0,
      "step": 16140
    },
    {
      "entropy": 5.185122776031494,
      "epoch": 1.3485633144002673,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004821735135517805,
      "loss": 4.9472,
      "mean_token_accuracy": 0.20413346141576766,
      "num_tokens": 29708741.0,
      "step": 16145
    },
    {
      "entropy": 5.179406309127808,
      "epoch": 1.3489809555629804,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004821618983770561,
      "loss": 5.0699,
      "mean_token_accuracy": 0.19628502279520035,
      "num_tokens": 29717616.0,
      "step": 16150
    },
    {
      "entropy": 5.229258298873901,
      "epoch": 1.3493985967256932,
      "grad_norm": 2.25,
      "learning_rate": 0.00048215027957571326,
      "loss": 4.9605,
      "mean_token_accuracy": 0.19894986003637313,
      "num_tokens": 29726601.0,
      "step": 16155
    },
    {
      "entropy": 5.297155857086182,
      "epoch": 1.3498162378884062,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004821386571479555,
      "loss": 5.0333,
      "mean_token_accuracy": 0.20128232091665268,
      "num_tokens": 29736930.0,
      "step": 16160
    },
    {
      "entropy": 5.1735546588897705,
      "epoch": 1.3502338790511192,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004821270310939861,
      "loss": 4.8995,
      "mean_token_accuracy": 0.20319953560829163,
      "num_tokens": 29745980.0,
      "step": 16165
    },
    {
      "entropy": 5.179254007339478,
      "epoch": 1.3506515202138323,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004821154014140088,
      "loss": 4.9979,
      "mean_token_accuracy": 0.19894501864910125,
      "num_tokens": 29755701.0,
      "step": 16170
    },
    {
      "entropy": 5.162038087844849,
      "epoch": 1.3510691613765453,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004821037681082271,
      "loss": 4.8815,
      "mean_token_accuracy": 0.21190578043460845,
      "num_tokens": 29763503.0,
      "step": 16175
    },
    {
      "entropy": 5.140215444564819,
      "epoch": 1.3514868025392583,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004820921311768446,
      "loss": 4.8631,
      "mean_token_accuracy": 0.20751822292804717,
      "num_tokens": 29771667.0,
      "step": 16180
    },
    {
      "entropy": 5.243862438201904,
      "epoch": 1.3519044437019714,
      "grad_norm": 1.5,
      "learning_rate": 0.00048208049062006505,
      "loss": 5.067,
      "mean_token_accuracy": 0.1934657260775566,
      "num_tokens": 29780836.0,
      "step": 16185
    },
    {
      "entropy": 5.233483600616455,
      "epoch": 1.3523220848646842,
      "grad_norm": 1.546875,
      "learning_rate": 0.00048206884643809224,
      "loss": 4.9768,
      "mean_token_accuracy": 0.1973920941352844,
      "num_tokens": 29790397.0,
      "step": 16190
    },
    {
      "entropy": 5.203599119186402,
      "epoch": 1.3527397260273972,
      "grad_norm": 1.59375,
      "learning_rate": 0.00048205719863113014,
      "loss": 5.0347,
      "mean_token_accuracy": 0.20532125383615493,
      "num_tokens": 29799227.0,
      "step": 16195
    },
    {
      "entropy": 5.219901609420776,
      "epoch": 1.3531573671901103,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004820455471993824,
      "loss": 5.0219,
      "mean_token_accuracy": 0.19354509115219115,
      "num_tokens": 29808182.0,
      "step": 16200
    },
    {
      "entropy": 5.2514519691467285,
      "epoch": 1.3535750083528233,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00048203389214305326,
      "loss": 5.067,
      "mean_token_accuracy": 0.19573459327220916,
      "num_tokens": 29817843.0,
      "step": 16205
    },
    {
      "entropy": 5.239224767684936,
      "epoch": 1.3539926495155363,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004820222334623465,
      "loss": 4.9782,
      "mean_token_accuracy": 0.20426790565252304,
      "num_tokens": 29827082.0,
      "step": 16210
    },
    {
      "entropy": 5.273222017288208,
      "epoch": 1.3544102906782491,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00048201057115746643,
      "loss": 4.9673,
      "mean_token_accuracy": 0.2034750759601593,
      "num_tokens": 29836176.0,
      "step": 16215
    },
    {
      "entropy": 5.2026291847229,
      "epoch": 1.3548279318409622,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004819989052286172,
      "loss": 4.9411,
      "mean_token_accuracy": 0.20735379457473754,
      "num_tokens": 29845252.0,
      "step": 16220
    },
    {
      "entropy": 5.203726196289063,
      "epoch": 1.3552455730036752,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004819872356760028,
      "loss": 4.9795,
      "mean_token_accuracy": 0.19833919405937195,
      "num_tokens": 29854115.0,
      "step": 16225
    },
    {
      "entropy": 5.209327840805054,
      "epoch": 1.3556632141663882,
      "grad_norm": 1.375,
      "learning_rate": 0.00048197556249982774,
      "loss": 5.0009,
      "mean_token_accuracy": 0.20923833250999452,
      "num_tokens": 29861918.0,
      "step": 16230
    },
    {
      "entropy": 5.2920228958129885,
      "epoch": 1.3560808553291013,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004819638857002965,
      "loss": 5.0069,
      "mean_token_accuracy": 0.2014613479375839,
      "num_tokens": 29871421.0,
      "step": 16235
    },
    {
      "entropy": 5.196942234039307,
      "epoch": 1.3564984964918143,
      "grad_norm": 2.375,
      "learning_rate": 0.00048195220527761306,
      "loss": 4.9369,
      "mean_token_accuracy": 0.20465838313102722,
      "num_tokens": 29880328.0,
      "step": 16240
    },
    {
      "entropy": 5.276680374145508,
      "epoch": 1.3569161376545273,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00048194052123198236,
      "loss": 5.1106,
      "mean_token_accuracy": 0.1928577870130539,
      "num_tokens": 29889777.0,
      "step": 16245
    },
    {
      "entropy": 5.1882469177246096,
      "epoch": 1.3573337788172402,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048192883356360867,
      "loss": 4.9719,
      "mean_token_accuracy": 0.20044808983802795,
      "num_tokens": 29898638.0,
      "step": 16250
    },
    {
      "entropy": 5.233463764190674,
      "epoch": 1.3577514199799532,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004819171422726966,
      "loss": 5.0012,
      "mean_token_accuracy": 0.20340635776519775,
      "num_tokens": 29907856.0,
      "step": 16255
    },
    {
      "entropy": 5.258271503448486,
      "epoch": 1.3581690611426662,
      "grad_norm": 1.5625,
      "learning_rate": 0.00048190544735945094,
      "loss": 5.0653,
      "mean_token_accuracy": 0.1984056055545807,
      "num_tokens": 29916992.0,
      "step": 16260
    },
    {
      "entropy": 5.241979360580444,
      "epoch": 1.3585867023053793,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004818937488240764,
      "loss": 5.0037,
      "mean_token_accuracy": 0.20135711878538132,
      "num_tokens": 29926323.0,
      "step": 16265
    },
    {
      "entropy": 5.2713963985443115,
      "epoch": 1.3590043434680923,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048188204666677775,
      "loss": 5.0277,
      "mean_token_accuracy": 0.1995244562625885,
      "num_tokens": 29935386.0,
      "step": 16270
    },
    {
      "entropy": 5.200611305236817,
      "epoch": 1.359421984630805,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004818703408877598,
      "loss": 5.0242,
      "mean_token_accuracy": 0.20079482346773148,
      "num_tokens": 29944741.0,
      "step": 16275
    },
    {
      "entropy": 5.241830158233642,
      "epoch": 1.3598396257935181,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004818586314872276,
      "loss": 4.9182,
      "mean_token_accuracy": 0.20589712262153625,
      "num_tokens": 29952407.0,
      "step": 16280
    },
    {
      "entropy": 5.218263721466064,
      "epoch": 1.3602572669562312,
      "grad_norm": 1.421875,
      "learning_rate": 0.00048184691846538596,
      "loss": 4.9654,
      "mean_token_accuracy": 0.19902905076742172,
      "num_tokens": 29962315.0,
      "step": 16285
    },
    {
      "entropy": 5.22742486000061,
      "epoch": 1.3606749081189442,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048183520182244006,
      "loss": 5.0341,
      "mean_token_accuracy": 0.20251507610082625,
      "num_tokens": 29972070.0,
      "step": 16290
    },
    {
      "entropy": 5.197206211090088,
      "epoch": 1.3610925492816572,
      "grad_norm": 1.6484375,
      "learning_rate": 0.000481823481558595,
      "loss": 4.9491,
      "mean_token_accuracy": 0.2020959198474884,
      "num_tokens": 29980416.0,
      "step": 16295
    },
    {
      "entropy": 5.2874839305877686,
      "epoch": 1.3615101904443703,
      "grad_norm": 1.71875,
      "learning_rate": 0.00048181175767405594,
      "loss": 5.0171,
      "mean_token_accuracy": 0.20511281192302705,
      "num_tokens": 29989436.0,
      "step": 16300
    },
    {
      "entropy": 5.261190605163574,
      "epoch": 1.3619278316070833,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004818000301690282,
      "loss": 4.9395,
      "mean_token_accuracy": 0.20289011895656586,
      "num_tokens": 29998295.0,
      "step": 16305
    },
    {
      "entropy": 5.172616100311279,
      "epoch": 1.3623454727697961,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004817882990437169,
      "loss": 4.9913,
      "mean_token_accuracy": 0.1942872852087021,
      "num_tokens": 30007892.0,
      "step": 16310
    },
    {
      "entropy": 5.163495492935181,
      "epoch": 1.3627631139325092,
      "grad_norm": 1.75,
      "learning_rate": 0.00048177656429832763,
      "loss": 4.9455,
      "mean_token_accuracy": 0.21036762446165086,
      "num_tokens": 30016477.0,
      "step": 16315
    },
    {
      "entropy": 5.085618877410889,
      "epoch": 1.3631807550952222,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00048176482593306557,
      "loss": 4.9456,
      "mean_token_accuracy": 0.2014121800661087,
      "num_tokens": 30025693.0,
      "step": 16320
    },
    {
      "entropy": 5.273138427734375,
      "epoch": 1.3635983962579352,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00048175308394813644,
      "loss": 4.9644,
      "mean_token_accuracy": 0.20253812074661254,
      "num_tokens": 30034552.0,
      "step": 16325
    },
    {
      "entropy": 5.175660562515259,
      "epoch": 1.3640160374206483,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004817413383437456,
      "loss": 4.9693,
      "mean_token_accuracy": 0.2088318273425102,
      "num_tokens": 30043584.0,
      "step": 16330
    },
    {
      "entropy": 5.268996524810791,
      "epoch": 1.364433678583361,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00048172958912009886,
      "loss": 5.0953,
      "mean_token_accuracy": 0.19376614689826965,
      "num_tokens": 30053189.0,
      "step": 16335
    },
    {
      "entropy": 5.1708635807037355,
      "epoch": 1.364851319746074,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00048171783627740176,
      "loss": 4.9378,
      "mean_token_accuracy": 0.1991383358836174,
      "num_tokens": 30062144.0,
      "step": 16340
    },
    {
      "entropy": 5.263130712509155,
      "epoch": 1.3652689609087871,
      "grad_norm": 1.609375,
      "learning_rate": 0.00048170607981586023,
      "loss": 4.9723,
      "mean_token_accuracy": 0.2073100298643112,
      "num_tokens": 30071646.0,
      "step": 16345
    },
    {
      "entropy": 5.303702163696289,
      "epoch": 1.3656866020715002,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004816943197356799,
      "loss": 5.0763,
      "mean_token_accuracy": 0.19640464186668397,
      "num_tokens": 30081322.0,
      "step": 16350
    },
    {
      "entropy": 5.145657253265381,
      "epoch": 1.3661042432342132,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004816825560370667,
      "loss": 4.9506,
      "mean_token_accuracy": 0.20959710329771042,
      "num_tokens": 30090059.0,
      "step": 16355
    },
    {
      "entropy": 5.292239761352539,
      "epoch": 1.3665218843969262,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00048167078872022657,
      "loss": 5.0749,
      "mean_token_accuracy": 0.19597577899694443,
      "num_tokens": 30100924.0,
      "step": 16360
    },
    {
      "entropy": 5.188052797317505,
      "epoch": 1.3669395255596393,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00048165901778536547,
      "loss": 4.9463,
      "mean_token_accuracy": 0.20262345671653748,
      "num_tokens": 30109972.0,
      "step": 16365
    },
    {
      "entropy": 5.239314365386963,
      "epoch": 1.367357166722352,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048164724323268956,
      "loss": 4.9784,
      "mean_token_accuracy": 0.19500905871391297,
      "num_tokens": 30118705.0,
      "step": 16370
    },
    {
      "entropy": 5.266728258132934,
      "epoch": 1.3677748078850651,
      "grad_norm": 1.453125,
      "learning_rate": 0.00048163546506240493,
      "loss": 5.072,
      "mean_token_accuracy": 0.19757989794015884,
      "num_tokens": 30127631.0,
      "step": 16375
    },
    {
      "entropy": 5.15447850227356,
      "epoch": 1.3681924490477781,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004816236832747177,
      "loss": 4.9278,
      "mean_token_accuracy": 0.212708280980587,
      "num_tokens": 30136127.0,
      "step": 16380
    },
    {
      "entropy": 5.22601203918457,
      "epoch": 1.3686100902104912,
      "grad_norm": 1.375,
      "learning_rate": 0.0004816118978698342,
      "loss": 4.941,
      "mean_token_accuracy": 0.20710660070180892,
      "num_tokens": 30145609.0,
      "step": 16385
    },
    {
      "entropy": 5.3749700546264645,
      "epoch": 1.3690277313732042,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004816001088479608,
      "loss": 5.0939,
      "mean_token_accuracy": 0.1873183861374855,
      "num_tokens": 30155705.0,
      "step": 16390
    },
    {
      "entropy": 5.205834579467774,
      "epoch": 1.369445372535917,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004815883162093037,
      "loss": 5.0426,
      "mean_token_accuracy": 0.19153282642364503,
      "num_tokens": 30165549.0,
      "step": 16395
    },
    {
      "entropy": 5.178115177154541,
      "epoch": 1.36986301369863,
      "grad_norm": 1.625,
      "learning_rate": 0.00048157651995406954,
      "loss": 4.9956,
      "mean_token_accuracy": 0.2050653487443924,
      "num_tokens": 30174716.0,
      "step": 16400
    },
    {
      "entropy": 5.219642877578735,
      "epoch": 1.370280654861343,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004815647200824647,
      "loss": 4.9563,
      "mean_token_accuracy": 0.20402771830558777,
      "num_tokens": 30184009.0,
      "step": 16405
    },
    {
      "entropy": 5.331546354293823,
      "epoch": 1.3706982960240561,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048155291659469587,
      "loss": 5.0469,
      "mean_token_accuracy": 0.19827071726322174,
      "num_tokens": 30193731.0,
      "step": 16410
    },
    {
      "entropy": 5.247933292388916,
      "epoch": 1.3711159371867692,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00048154110949096945,
      "loss": 5.151,
      "mean_token_accuracy": 0.19503606408834456,
      "num_tokens": 30203643.0,
      "step": 16415
    },
    {
      "entropy": 5.24662446975708,
      "epoch": 1.3715335783494822,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004815292987714924,
      "loss": 4.9967,
      "mean_token_accuracy": 0.19720692038536072,
      "num_tokens": 30212436.0,
      "step": 16420
    },
    {
      "entropy": 5.236334800720215,
      "epoch": 1.3719512195121952,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004815174844364714,
      "loss": 4.9315,
      "mean_token_accuracy": 0.20476281493902207,
      "num_tokens": 30221730.0,
      "step": 16425
    },
    {
      "entropy": 5.317168569564819,
      "epoch": 1.372368860674908,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004815056664861133,
      "loss": 5.0068,
      "mean_token_accuracy": 0.1987475261092186,
      "num_tokens": 30230645.0,
      "step": 16430
    },
    {
      "entropy": 5.168989086151123,
      "epoch": 1.372786501837621,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004814938449206248,
      "loss": 4.9788,
      "mean_token_accuracy": 0.20343690514564514,
      "num_tokens": 30239513.0,
      "step": 16435
    },
    {
      "entropy": 5.182251071929931,
      "epoch": 1.373204143000334,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048148201974021297,
      "loss": 4.9595,
      "mean_token_accuracy": 0.20142049342393875,
      "num_tokens": 30248920.0,
      "step": 16440
    },
    {
      "entropy": 5.197758913040161,
      "epoch": 1.3736217841630471,
      "grad_norm": 1.609375,
      "learning_rate": 0.000481470190945085,
      "loss": 4.9552,
      "mean_token_accuracy": 0.2039742276072502,
      "num_tokens": 30257851.0,
      "step": 16445
    },
    {
      "entropy": 5.143648099899292,
      "epoch": 1.3740394253257602,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004814583585354477,
      "loss": 4.982,
      "mean_token_accuracy": 0.21135611534118653,
      "num_tokens": 30267676.0,
      "step": 16450
    },
    {
      "entropy": 5.116444778442383,
      "epoch": 1.374457066488473,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00048144652251150836,
      "loss": 4.8768,
      "mean_token_accuracy": 0.20631337761878968,
      "num_tokens": 30277744.0,
      "step": 16455
    },
    {
      "entropy": 5.319468259811401,
      "epoch": 1.374874707651186,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000481434682873474,
      "loss": 5.0398,
      "mean_token_accuracy": 0.19118778854608537,
      "num_tokens": 30286578.0,
      "step": 16460
    },
    {
      "entropy": 5.283721256256103,
      "epoch": 1.375292348813899,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004814228396215522,
      "loss": 5.0149,
      "mean_token_accuracy": 0.20121330469846727,
      "num_tokens": 30295896.0,
      "step": 16465
    },
    {
      "entropy": 5.186836528778076,
      "epoch": 1.375709989976612,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004814109927559501,
      "loss": 5.0803,
      "mean_token_accuracy": 0.19709692746400834,
      "num_tokens": 30305397.0,
      "step": 16470
    },
    {
      "entropy": 5.249338674545288,
      "epoch": 1.3761276311393251,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004813991422768751,
      "loss": 5.0207,
      "mean_token_accuracy": 0.20541560649871826,
      "num_tokens": 30314025.0,
      "step": 16475
    },
    {
      "entropy": 5.243709850311279,
      "epoch": 1.3765452723020382,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00048138728818453456,
      "loss": 5.031,
      "mean_token_accuracy": 0.19735687226057053,
      "num_tokens": 30324384.0,
      "step": 16480
    },
    {
      "entropy": 5.228837633132935,
      "epoch": 1.3769629134647512,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048137543047913636,
      "loss": 5.0348,
      "mean_token_accuracy": 0.20136696845293045,
      "num_tokens": 30333663.0,
      "step": 16485
    },
    {
      "entropy": 5.249251031875611,
      "epoch": 1.377380554627464,
      "grad_norm": 1.25,
      "learning_rate": 0.00048136356916088764,
      "loss": 4.9832,
      "mean_token_accuracy": 0.19809741228818895,
      "num_tokens": 30343603.0,
      "step": 16490
    },
    {
      "entropy": 5.28144850730896,
      "epoch": 1.377798195790177,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048135170422999624,
      "loss": 5.0518,
      "mean_token_accuracy": 0.1968901351094246,
      "num_tokens": 30352680.0,
      "step": 16495
    },
    {
      "entropy": 5.211698246002197,
      "epoch": 1.37821583695289,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048133983568666994,
      "loss": 5.0183,
      "mean_token_accuracy": 0.20279511213302612,
      "num_tokens": 30362597.0,
      "step": 16500
    },
    {
      "entropy": 5.336991786956787,
      "epoch": 1.378633478115603,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004813279635311165,
      "loss": 5.1584,
      "mean_token_accuracy": 0.18890759199857712,
      "num_tokens": 30372037.0,
      "step": 16505
    },
    {
      "entropy": 5.306573581695557,
      "epoch": 1.3790511192783161,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004813160877635436,
      "loss": 5.015,
      "mean_token_accuracy": 0.20456478297710418,
      "num_tokens": 30380867.0,
      "step": 16510
    },
    {
      "entropy": 5.287637615203858,
      "epoch": 1.379468760441029,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048130420838415926,
      "loss": 5.0599,
      "mean_token_accuracy": 0.19350311309099197,
      "num_tokens": 30391181.0,
      "step": 16515
    },
    {
      "entropy": 5.247127342224121,
      "epoch": 1.379886401603742,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048129232539317147,
      "loss": 5.0607,
      "mean_token_accuracy": 0.19978141486644746,
      "num_tokens": 30400519.0,
      "step": 16520
    },
    {
      "entropy": 5.200404310226441,
      "epoch": 1.380304042766455,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048128043879078816,
      "loss": 4.9615,
      "mean_token_accuracy": 0.20181944370269775,
      "num_tokens": 30409570.0,
      "step": 16525
    },
    {
      "entropy": 5.29139666557312,
      "epoch": 1.380721683929168,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004812685485772175,
      "loss": 4.9837,
      "mean_token_accuracy": 0.20188671052455903,
      "num_tokens": 30418458.0,
      "step": 16530
    },
    {
      "entropy": 5.265335655212402,
      "epoch": 1.381139325091881,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048125665475266756,
      "loss": 4.997,
      "mean_token_accuracy": 0.2030320554971695,
      "num_tokens": 30427933.0,
      "step": 16535
    },
    {
      "entropy": 5.327639436721801,
      "epoch": 1.3815569662545941,
      "grad_norm": 1.25,
      "learning_rate": 0.00048124475731734667,
      "loss": 5.0756,
      "mean_token_accuracy": 0.20148151367902756,
      "num_tokens": 30436825.0,
      "step": 16540
    },
    {
      "entropy": 5.295147705078125,
      "epoch": 1.3819746074173072,
      "grad_norm": 1.703125,
      "learning_rate": 0.000481232856271463,
      "loss": 4.9902,
      "mean_token_accuracy": 0.19475983083248138,
      "num_tokens": 30445367.0,
      "step": 16545
    },
    {
      "entropy": 5.318212413787842,
      "epoch": 1.38239224858002,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048122095161522496,
      "loss": 5.121,
      "mean_token_accuracy": 0.19134945571422576,
      "num_tokens": 30455236.0,
      "step": 16550
    },
    {
      "entropy": 5.18565411567688,
      "epoch": 1.382809889742733,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004812090433488409,
      "loss": 4.9712,
      "mean_token_accuracy": 0.1960343137383461,
      "num_tokens": 30465003.0,
      "step": 16555
    },
    {
      "entropy": 5.2159703254699705,
      "epoch": 1.383227530905446,
      "grad_norm": 1.375,
      "learning_rate": 0.00048119713147251944,
      "loss": 4.9559,
      "mean_token_accuracy": 0.20289413034915924,
      "num_tokens": 30473686.0,
      "step": 16560
    },
    {
      "entropy": 5.225674247741699,
      "epoch": 1.383645172068159,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048118521598646886,
      "loss": 5.0605,
      "mean_token_accuracy": 0.2006628394126892,
      "num_tokens": 30482684.0,
      "step": 16565
    },
    {
      "entropy": 5.246957445144654,
      "epoch": 1.384062813230872,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048117329689089793,
      "loss": 4.933,
      "mean_token_accuracy": 0.1984789028763771,
      "num_tokens": 30491356.0,
      "step": 16570
    },
    {
      "entropy": 5.229938650131226,
      "epoch": 1.384480454393585,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00048116137418601526,
      "loss": 4.9578,
      "mean_token_accuracy": 0.19915591329336166,
      "num_tokens": 30500829.0,
      "step": 16575
    },
    {
      "entropy": 5.223511362075806,
      "epoch": 1.384898095556298,
      "grad_norm": 1.671875,
      "learning_rate": 0.00048114944787202955,
      "loss": 5.1195,
      "mean_token_accuracy": 0.19205650240182875,
      "num_tokens": 30509759.0,
      "step": 16580
    },
    {
      "entropy": 5.194095802307129,
      "epoch": 1.385315736719011,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00048113751794914974,
      "loss": 4.9161,
      "mean_token_accuracy": 0.20769768059253693,
      "num_tokens": 30518638.0,
      "step": 16585
    },
    {
      "entropy": 5.207504940032959,
      "epoch": 1.385733377881724,
      "grad_norm": 1.8515625,
      "learning_rate": 0.00048112558441758437,
      "loss": 4.9759,
      "mean_token_accuracy": 0.20056079477071762,
      "num_tokens": 30527973.0,
      "step": 16590
    },
    {
      "entropy": 5.20620493888855,
      "epoch": 1.386151019044437,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004811136472775426,
      "loss": 4.9109,
      "mean_token_accuracy": 0.20595538169145583,
      "num_tokens": 30537120.0,
      "step": 16595
    },
    {
      "entropy": 5.271022987365723,
      "epoch": 1.38656866020715,
      "grad_norm": 2.203125,
      "learning_rate": 0.0004811017065292334,
      "loss": 4.9984,
      "mean_token_accuracy": 0.19897040724754333,
      "num_tokens": 30546372.0,
      "step": 16600
    },
    {
      "entropy": 5.305221700668335,
      "epoch": 1.3869863013698631,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048108976217286576,
      "loss": 5.0214,
      "mean_token_accuracy": 0.19598299860954285,
      "num_tokens": 30555377.0,
      "step": 16605
    },
    {
      "entropy": 5.185356855392456,
      "epoch": 1.387403942532576,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004810778142086487,
      "loss": 4.963,
      "mean_token_accuracy": 0.20411504209041595,
      "num_tokens": 30565208.0,
      "step": 16610
    },
    {
      "entropy": 5.175542879104614,
      "epoch": 1.387821583695289,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00048106586263679146,
      "loss": 5.0223,
      "mean_token_accuracy": 0.20476298928260803,
      "num_tokens": 30573230.0,
      "step": 16615
    },
    {
      "entropy": 5.207584524154663,
      "epoch": 1.388239224858002,
      "grad_norm": 1.5625,
      "learning_rate": 0.00048105390745750324,
      "loss": 4.9076,
      "mean_token_accuracy": 0.2044145330786705,
      "num_tokens": 30581570.0,
      "step": 16620
    },
    {
      "entropy": 5.214545488357544,
      "epoch": 1.388656866020715,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00048104194867099336,
      "loss": 4.9886,
      "mean_token_accuracy": 0.20459677577018737,
      "num_tokens": 30591091.0,
      "step": 16625
    },
    {
      "entropy": 5.195149660110474,
      "epoch": 1.389074507183428,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048102998627747116,
      "loss": 4.9874,
      "mean_token_accuracy": 0.20406362116336824,
      "num_tokens": 30601545.0,
      "step": 16630
    },
    {
      "entropy": 5.2948870182037355,
      "epoch": 1.3894921483461409,
      "grad_norm": 1.5625,
      "learning_rate": 0.000481018020277146,
      "loss": 5.0378,
      "mean_token_accuracy": 0.19523000419139863,
      "num_tokens": 30610630.0,
      "step": 16635
    },
    {
      "entropy": 5.206798934936524,
      "epoch": 1.389909789508854,
      "grad_norm": 1.578125,
      "learning_rate": 0.00048100605067022753,
      "loss": 4.9275,
      "mean_token_accuracy": 0.2073330894112587,
      "num_tokens": 30620015.0,
      "step": 16640
    },
    {
      "entropy": 5.25678277015686,
      "epoch": 1.390327430671567,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048099407745692504,
      "loss": 4.9365,
      "mean_token_accuracy": 0.20815445035696029,
      "num_tokens": 30628590.0,
      "step": 16645
    },
    {
      "entropy": 5.137222003936768,
      "epoch": 1.39074507183428,
      "grad_norm": 1.703125,
      "learning_rate": 0.00048098210063744827,
      "loss": 4.9858,
      "mean_token_accuracy": 0.20285354405641556,
      "num_tokens": 30638624.0,
      "step": 16650
    },
    {
      "entropy": 5.225707769393921,
      "epoch": 1.391162712996993,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00048097012021200696,
      "loss": 5.0377,
      "mean_token_accuracy": 0.19517669975757598,
      "num_tokens": 30648247.0,
      "step": 16655
    },
    {
      "entropy": 5.340005016326904,
      "epoch": 1.391580354159706,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004809581361808108,
      "loss": 5.1415,
      "mean_token_accuracy": 0.20029016137123107,
      "num_tokens": 30659045.0,
      "step": 16660
    },
    {
      "entropy": 5.270746469497681,
      "epoch": 1.391997995322419,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004809461485440694,
      "loss": 5.0165,
      "mean_token_accuracy": 0.1988289549946785,
      "num_tokens": 30668948.0,
      "step": 16665
    },
    {
      "entropy": 5.305721664428711,
      "epoch": 1.392415636485132,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004809341573019929,
      "loss": 4.9404,
      "mean_token_accuracy": 0.20942579954862595,
      "num_tokens": 30678818.0,
      "step": 16670
    },
    {
      "entropy": 5.1248640537261965,
      "epoch": 1.392833277647845,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004809221624547911,
      "loss": 4.8862,
      "mean_token_accuracy": 0.2093770444393158,
      "num_tokens": 30687731.0,
      "step": 16675
    },
    {
      "entropy": 5.194254970550537,
      "epoch": 1.393250918810558,
      "grad_norm": 1.5625,
      "learning_rate": 0.000480910164002674,
      "loss": 4.9841,
      "mean_token_accuracy": 0.2011880025267601,
      "num_tokens": 30696880.0,
      "step": 16680
    },
    {
      "entropy": 5.26388521194458,
      "epoch": 1.393668559973271,
      "grad_norm": 1.578125,
      "learning_rate": 0.00048089816194585153,
      "loss": 4.9782,
      "mean_token_accuracy": 0.1969760090112686,
      "num_tokens": 30706274.0,
      "step": 16685
    },
    {
      "entropy": 5.1839902877807615,
      "epoch": 1.394086201135984,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004808861562845339,
      "loss": 4.9722,
      "mean_token_accuracy": 0.20056352615356446,
      "num_tokens": 30715184.0,
      "step": 16690
    },
    {
      "entropy": 5.226570558547974,
      "epoch": 1.3945038422986968,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004808741470189313,
      "loss": 4.9248,
      "mean_token_accuracy": 0.21125609427690506,
      "num_tokens": 30724501.0,
      "step": 16695
    },
    {
      "entropy": 5.166960906982422,
      "epoch": 1.39492148346141,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004808621341492541,
      "loss": 4.9271,
      "mean_token_accuracy": 0.20682960897684097,
      "num_tokens": 30733496.0,
      "step": 16700
    },
    {
      "entropy": 5.114538955688476,
      "epoch": 1.395339124624123,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004808501176757122,
      "loss": 4.9174,
      "mean_token_accuracy": 0.21072573512792586,
      "num_tokens": 30742096.0,
      "step": 16705
    },
    {
      "entropy": 5.141414499282837,
      "epoch": 1.395756765786836,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004808380975985163,
      "loss": 4.8843,
      "mean_token_accuracy": 0.20665495693683625,
      "num_tokens": 30751452.0,
      "step": 16710
    },
    {
      "entropy": 5.220099782943725,
      "epoch": 1.396174406949549,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004808260739178768,
      "loss": 5.0244,
      "mean_token_accuracy": 0.19996500611305237,
      "num_tokens": 30760686.0,
      "step": 16715
    },
    {
      "entropy": 5.218145656585693,
      "epoch": 1.396592048112262,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0004808140466340041,
      "loss": 4.9925,
      "mean_token_accuracy": 0.20100715458393098,
      "num_tokens": 30769374.0,
      "step": 16720
    },
    {
      "entropy": 5.215980911254883,
      "epoch": 1.397009689274975,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004808020157471087,
      "loss": 5.0316,
      "mean_token_accuracy": 0.20226773321628572,
      "num_tokens": 30778910.0,
      "step": 16725
    },
    {
      "entropy": 5.212262105941773,
      "epoch": 1.3974273304376879,
      "grad_norm": 1.484375,
      "learning_rate": 0.00048078998125740136,
      "loss": 4.979,
      "mean_token_accuracy": 0.19979380667209626,
      "num_tokens": 30787844.0,
      "step": 16730
    },
    {
      "entropy": 5.251246643066406,
      "epoch": 1.3978449716004009,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004807779431650927,
      "loss": 5.044,
      "mean_token_accuracy": 0.20198030024766922,
      "num_tokens": 30797415.0,
      "step": 16735
    },
    {
      "entropy": 5.276937532424927,
      "epoch": 1.398262612763114,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004807659014703934,
      "loss": 5.0163,
      "mean_token_accuracy": 0.20407470166683198,
      "num_tokens": 30805595.0,
      "step": 16740
    },
    {
      "entropy": 5.233874797821045,
      "epoch": 1.398680253925827,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004807538561735143,
      "loss": 4.9186,
      "mean_token_accuracy": 0.2104392573237419,
      "num_tokens": 30813840.0,
      "step": 16745
    },
    {
      "entropy": 5.236031007766724,
      "epoch": 1.39909789508854,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00048074180727466636,
      "loss": 4.9265,
      "mean_token_accuracy": 0.20302237719297409,
      "num_tokens": 30822208.0,
      "step": 16750
    },
    {
      "entropy": 5.23362979888916,
      "epoch": 1.3995155362512528,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00048072975477406036,
      "loss": 5.1327,
      "mean_token_accuracy": 0.19141052514314652,
      "num_tokens": 30831804.0,
      "step": 16755
    },
    {
      "entropy": 5.236087846755981,
      "epoch": 1.399933177413966,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004807176986719074,
      "loss": 4.9844,
      "mean_token_accuracy": 0.20507715493440629,
      "num_tokens": 30841450.0,
      "step": 16760
    },
    {
      "entropy": 5.235486364364624,
      "epoch": 1.4003508185766789,
      "grad_norm": 2.265625,
      "learning_rate": 0.00048070563896841843,
      "loss": 5.0208,
      "mean_token_accuracy": 0.19853376597166061,
      "num_tokens": 30849906.0,
      "step": 16765
    },
    {
      "entropy": 5.177568483352661,
      "epoch": 1.400768459739392,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048069357566380474,
      "loss": 5.0187,
      "mean_token_accuracy": 0.19746333062648774,
      "num_tokens": 30859643.0,
      "step": 16770
    },
    {
      "entropy": 5.155114650726318,
      "epoch": 1.401186100902105,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00048068150875827737,
      "loss": 4.8766,
      "mean_token_accuracy": 0.2066615790128708,
      "num_tokens": 30868728.0,
      "step": 16775
    },
    {
      "entropy": 5.200304460525513,
      "epoch": 1.401603742064818,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048066943825204764,
      "loss": 4.9531,
      "mean_token_accuracy": 0.2015573039650917,
      "num_tokens": 30877253.0,
      "step": 16780
    },
    {
      "entropy": 5.2137978076934814,
      "epoch": 1.402021383227531,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004806573641453268,
      "loss": 4.9942,
      "mean_token_accuracy": 0.1997855305671692,
      "num_tokens": 30886118.0,
      "step": 16785
    },
    {
      "entropy": 5.209441757202148,
      "epoch": 1.4024390243902438,
      "grad_norm": 1.453125,
      "learning_rate": 0.00048064528643832627,
      "loss": 4.9787,
      "mean_token_accuracy": 0.20340603590011597,
      "num_tokens": 30894533.0,
      "step": 16790
    },
    {
      "entropy": 5.172437477111816,
      "epoch": 1.4028566655529568,
      "grad_norm": 1.71875,
      "learning_rate": 0.00048063320513125747,
      "loss": 4.8811,
      "mean_token_accuracy": 0.20516163855791092,
      "num_tokens": 30904934.0,
      "step": 16795
    },
    {
      "entropy": 5.135058736801147,
      "epoch": 1.4032743067156699,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004806211202243319,
      "loss": 4.9267,
      "mean_token_accuracy": 0.2102593019604683,
      "num_tokens": 30913641.0,
      "step": 16800
    },
    {
      "entropy": 5.192354726791382,
      "epoch": 1.403691947878383,
      "grad_norm": 1.40625,
      "learning_rate": 0.00048060903171776113,
      "loss": 4.8851,
      "mean_token_accuracy": 0.2052180528640747,
      "num_tokens": 30922529.0,
      "step": 16805
    },
    {
      "entropy": 5.205151844024658,
      "epoch": 1.404109589041096,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004805969396117567,
      "loss": 4.9785,
      "mean_token_accuracy": 0.20157587230205537,
      "num_tokens": 30932059.0,
      "step": 16810
    },
    {
      "entropy": 5.266339206695557,
      "epoch": 1.4045272302038088,
      "grad_norm": 1.53125,
      "learning_rate": 0.00048058484390653037,
      "loss": 5.1048,
      "mean_token_accuracy": 0.19591250717639924,
      "num_tokens": 30941141.0,
      "step": 16815
    },
    {
      "entropy": 5.304641485214233,
      "epoch": 1.404944871366522,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004805727446022939,
      "loss": 5.0453,
      "mean_token_accuracy": 0.202554327249527,
      "num_tokens": 30949785.0,
      "step": 16820
    },
    {
      "entropy": 5.150935363769531,
      "epoch": 1.4053625125292348,
      "grad_norm": 1.421875,
      "learning_rate": 0.00048056064169925904,
      "loss": 4.9258,
      "mean_token_accuracy": 0.2062472090125084,
      "num_tokens": 30958421.0,
      "step": 16825
    },
    {
      "entropy": 5.101469326019287,
      "epoch": 1.4057801536919479,
      "grad_norm": 1.375,
      "learning_rate": 0.00048054853519763786,
      "loss": 4.9091,
      "mean_token_accuracy": 0.20670131593942642,
      "num_tokens": 30968177.0,
      "step": 16830
    },
    {
      "entropy": 5.284583902359008,
      "epoch": 1.406197794854661,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004805364250976421,
      "loss": 4.9845,
      "mean_token_accuracy": 0.19776400923728943,
      "num_tokens": 30978048.0,
      "step": 16835
    },
    {
      "entropy": 5.179466342926025,
      "epoch": 1.406615436017374,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004805243113994837,
      "loss": 4.9864,
      "mean_token_accuracy": 0.20070878267288209,
      "num_tokens": 30988588.0,
      "step": 16840
    },
    {
      "entropy": 5.23879108428955,
      "epoch": 1.407033077180087,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004805121941033749,
      "loss": 4.9763,
      "mean_token_accuracy": 0.19767540246248244,
      "num_tokens": 30998000.0,
      "step": 16845
    },
    {
      "entropy": 5.1828655242919925,
      "epoch": 1.4074507183427998,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004805000732095278,
      "loss": 5.0104,
      "mean_token_accuracy": 0.20006529241800308,
      "num_tokens": 31007103.0,
      "step": 16850
    },
    {
      "entropy": 5.220724439620971,
      "epoch": 1.4078683595055128,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004804879487181546,
      "loss": 4.995,
      "mean_token_accuracy": 0.2055037036538124,
      "num_tokens": 31016658.0,
      "step": 16855
    },
    {
      "entropy": 5.330377435684204,
      "epoch": 1.4082860006682258,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004804758206294675,
      "loss": 5.119,
      "mean_token_accuracy": 0.1892683133482933,
      "num_tokens": 31026968.0,
      "step": 16860
    },
    {
      "entropy": 5.154836940765381,
      "epoch": 1.4087036418309389,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004804636889436787,
      "loss": 4.9115,
      "mean_token_accuracy": 0.20680506229400636,
      "num_tokens": 31035890.0,
      "step": 16865
    },
    {
      "entropy": 5.274804162979126,
      "epoch": 1.409121282993652,
      "grad_norm": 1.25,
      "learning_rate": 0.00048045155366100085,
      "loss": 5.0699,
      "mean_token_accuracy": 0.19586092084646226,
      "num_tokens": 31045203.0,
      "step": 16870
    },
    {
      "entropy": 5.256712436676025,
      "epoch": 1.4095389241563647,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004804394147816462,
      "loss": 5.0607,
      "mean_token_accuracy": 0.1967946320772171,
      "num_tokens": 31055802.0,
      "step": 16875
    },
    {
      "entropy": 5.210637903213501,
      "epoch": 1.409956565319078,
      "grad_norm": 2.03125,
      "learning_rate": 0.00048042727230582726,
      "loss": 4.9538,
      "mean_token_accuracy": 0.20798544138669967,
      "num_tokens": 31065434.0,
      "step": 16880
    },
    {
      "entropy": 5.215636396408081,
      "epoch": 1.4103742064817908,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004804151262337568,
      "loss": 4.863,
      "mean_token_accuracy": 0.21311477124691008,
      "num_tokens": 31073931.0,
      "step": 16885
    },
    {
      "entropy": 5.255144357681274,
      "epoch": 1.4107918476445038,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004804029765656471,
      "loss": 5.0768,
      "mean_token_accuracy": 0.19199151992797853,
      "num_tokens": 31083068.0,
      "step": 16890
    },
    {
      "entropy": 5.1387745380401615,
      "epoch": 1.4112094888072169,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004803908233017111,
      "loss": 4.8829,
      "mean_token_accuracy": 0.21092920303344725,
      "num_tokens": 31092306.0,
      "step": 16895
    },
    {
      "entropy": 5.180415487289428,
      "epoch": 1.41162712996993,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004803786664421616,
      "loss": 4.9137,
      "mean_token_accuracy": 0.2129834458231926,
      "num_tokens": 31100540.0,
      "step": 16900
    },
    {
      "entropy": 5.203063249588013,
      "epoch": 1.412044771132643,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048036650598721125,
      "loss": 4.9418,
      "mean_token_accuracy": 0.20263179242610932,
      "num_tokens": 31110337.0,
      "step": 16905
    },
    {
      "entropy": 5.0729021549224855,
      "epoch": 1.4124624122953557,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048035434193707306,
      "loss": 4.8517,
      "mean_token_accuracy": 0.20870340168476104,
      "num_tokens": 31119985.0,
      "step": 16910
    },
    {
      "entropy": 5.083775854110717,
      "epoch": 1.4128800534580688,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048034217429195986,
      "loss": 4.8166,
      "mean_token_accuracy": 0.20770271569490434,
      "num_tokens": 31128147.0,
      "step": 16915
    },
    {
      "entropy": 5.29504075050354,
      "epoch": 1.4132976946207818,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00048033000305208485,
      "loss": 5.0387,
      "mean_token_accuracy": 0.19874367713928223,
      "num_tokens": 31137150.0,
      "step": 16920
    },
    {
      "entropy": 5.218954610824585,
      "epoch": 1.4137153357834948,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048031782821766086,
      "loss": 4.9125,
      "mean_token_accuracy": 0.20586813539266585,
      "num_tokens": 31145847.0,
      "step": 16925
    },
    {
      "entropy": 5.1928599834442135,
      "epoch": 1.4141329769462079,
      "grad_norm": 1.453125,
      "learning_rate": 0.00048030564978890117,
      "loss": 5.0289,
      "mean_token_accuracy": 0.20259684771299363,
      "num_tokens": 31154313.0,
      "step": 16930
    },
    {
      "entropy": 5.24595537185669,
      "epoch": 1.4145506181089207,
      "grad_norm": 1.765625,
      "learning_rate": 0.000480293467766019,
      "loss": 5.0914,
      "mean_token_accuracy": 0.1919054552912712,
      "num_tokens": 31164104.0,
      "step": 16935
    },
    {
      "entropy": 5.18181734085083,
      "epoch": 1.414968259271634,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004802812821492276,
      "loss": 4.8616,
      "mean_token_accuracy": 0.21005224585533142,
      "num_tokens": 31172833.0,
      "step": 16940
    },
    {
      "entropy": 5.202144479751587,
      "epoch": 1.4153859004343468,
      "grad_norm": 1.375,
      "learning_rate": 0.00048026909293874016,
      "loss": 5.0353,
      "mean_token_accuracy": 0.19639531075954436,
      "num_tokens": 31182557.0,
      "step": 16945
    },
    {
      "entropy": 5.2007850170135494,
      "epoch": 1.4158035415970598,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004802569001347702,
      "loss": 4.992,
      "mean_token_accuracy": 0.20406373590230942,
      "num_tokens": 31191641.0,
      "step": 16950
    },
    {
      "entropy": 5.158174228668213,
      "epoch": 1.4162211827597728,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00048024470373753114,
      "loss": 4.9175,
      "mean_token_accuracy": 0.2022813305258751,
      "num_tokens": 31200689.0,
      "step": 16955
    },
    {
      "entropy": 5.268483114242554,
      "epoch": 1.4166388239224859,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00048023250374723656,
      "loss": 5.0846,
      "mean_token_accuracy": 0.19307434558868408,
      "num_tokens": 31209745.0,
      "step": 16960
    },
    {
      "entropy": 5.219144058227539,
      "epoch": 1.417056465085199,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004802203001640998,
      "loss": 4.9398,
      "mean_token_accuracy": 0.2028568997979164,
      "num_tokens": 31218802.0,
      "step": 16965
    },
    {
      "entropy": 5.190017747879028,
      "epoch": 1.4174741062479117,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004802080929883348,
      "loss": 4.9227,
      "mean_token_accuracy": 0.20173587054014205,
      "num_tokens": 31227650.0,
      "step": 16970
    },
    {
      "entropy": 5.151567029953003,
      "epoch": 1.4178917474106247,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004801958822201551,
      "loss": 4.9181,
      "mean_token_accuracy": 0.20610949546098709,
      "num_tokens": 31236490.0,
      "step": 16975
    },
    {
      "entropy": 5.233379983901978,
      "epoch": 1.4183093885733378,
      "grad_norm": 1.71875,
      "learning_rate": 0.00048018366785977445,
      "loss": 4.9567,
      "mean_token_accuracy": 0.2048879712820053,
      "num_tokens": 31245557.0,
      "step": 16980
    },
    {
      "entropy": 5.126082801818848,
      "epoch": 1.4187270297360508,
      "grad_norm": 1.375,
      "learning_rate": 0.00048017144990740673,
      "loss": 4.8425,
      "mean_token_accuracy": 0.21044390201568602,
      "num_tokens": 31253804.0,
      "step": 16985
    },
    {
      "entropy": 5.1709764957427975,
      "epoch": 1.4191446708987638,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004801592283632658,
      "loss": 4.9852,
      "mean_token_accuracy": 0.1990163266658783,
      "num_tokens": 31262913.0,
      "step": 16990
    },
    {
      "entropy": 5.208834791183472,
      "epoch": 1.4195623120614766,
      "grad_norm": 1.375,
      "learning_rate": 0.00048014700322756567,
      "loss": 4.9409,
      "mean_token_accuracy": 0.2026805117726326,
      "num_tokens": 31271274.0,
      "step": 16995
    },
    {
      "entropy": 5.203345012664795,
      "epoch": 1.41997995322419,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004801347745005203,
      "loss": 4.9239,
      "mean_token_accuracy": 0.20175321996212006,
      "num_tokens": 31280599.0,
      "step": 17000
    },
    {
      "entropy": 5.2335536003112795,
      "epoch": 1.4203975943869027,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004801225421823437,
      "loss": 4.9568,
      "mean_token_accuracy": 0.2074551075696945,
      "num_tokens": 31288986.0,
      "step": 17005
    },
    {
      "entropy": 5.26753339767456,
      "epoch": 1.4208152355496158,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00048011030627325015,
      "loss": 4.9962,
      "mean_token_accuracy": 0.20334589779376983,
      "num_tokens": 31298979.0,
      "step": 17010
    },
    {
      "entropy": 5.092473077774048,
      "epoch": 1.4212328767123288,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00048009806677345394,
      "loss": 4.9229,
      "mean_token_accuracy": 0.20718373507261276,
      "num_tokens": 31307502.0,
      "step": 17015
    },
    {
      "entropy": 5.082223558425904,
      "epoch": 1.4216505178750418,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000480085823683169,
      "loss": 4.896,
      "mean_token_accuracy": 0.20844393074512482,
      "num_tokens": 31316146.0,
      "step": 17020
    },
    {
      "entropy": 5.176376247406006,
      "epoch": 1.4220681590377549,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004800735770026099,
      "loss": 4.9886,
      "mean_token_accuracy": 0.19831094294786453,
      "num_tokens": 31325542.0,
      "step": 17025
    },
    {
      "entropy": 5.232873344421387,
      "epoch": 1.4224858002004677,
      "grad_norm": 1.6875,
      "learning_rate": 0.000480061326731991,
      "loss": 5.0077,
      "mean_token_accuracy": 0.1996428430080414,
      "num_tokens": 31334718.0,
      "step": 17030
    },
    {
      "entropy": 5.246603393554688,
      "epoch": 1.4229034413631807,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004800490728715267,
      "loss": 5.0003,
      "mean_token_accuracy": 0.20089405924081802,
      "num_tokens": 31343635.0,
      "step": 17035
    },
    {
      "entropy": 5.177912187576294,
      "epoch": 1.4233210825258937,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004800368154214317,
      "loss": 4.9609,
      "mean_token_accuracy": 0.2061464697122574,
      "num_tokens": 31352488.0,
      "step": 17040
    },
    {
      "entropy": 5.216895294189453,
      "epoch": 1.4237387236886068,
      "grad_norm": 2.34375,
      "learning_rate": 0.0004800245543819203,
      "loss": 5.0174,
      "mean_token_accuracy": 0.19882187396287918,
      "num_tokens": 31361788.0,
      "step": 17045
    },
    {
      "entropy": 5.181366777420044,
      "epoch": 1.4241563648513198,
      "grad_norm": 2.75,
      "learning_rate": 0.00048001228975320736,
      "loss": 5.0298,
      "mean_token_accuracy": 0.19416781961917878,
      "num_tokens": 31371190.0,
      "step": 17050
    },
    {
      "entropy": 5.225235319137573,
      "epoch": 1.4245740060140326,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004800000215355075,
      "loss": 5.0328,
      "mean_token_accuracy": 0.20008387863636018,
      "num_tokens": 31380696.0,
      "step": 17055
    },
    {
      "entropy": 5.324580383300781,
      "epoch": 1.4249916471767459,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004799877497290355,
      "loss": 5.0502,
      "mean_token_accuracy": 0.1929371178150177,
      "num_tokens": 31390129.0,
      "step": 17060
    },
    {
      "entropy": 5.228053379058838,
      "epoch": 1.4254092883394587,
      "grad_norm": 1.8203125,
      "learning_rate": 0.00047997547433400615,
      "loss": 4.9674,
      "mean_token_accuracy": 0.20307117253541945,
      "num_tokens": 31399697.0,
      "step": 17065
    },
    {
      "entropy": 5.2029108047485355,
      "epoch": 1.4258269295021717,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004799631953506345,
      "loss": 4.9874,
      "mean_token_accuracy": 0.20282674282789231,
      "num_tokens": 31408873.0,
      "step": 17070
    },
    {
      "entropy": 5.17996563911438,
      "epoch": 1.4262445706648847,
      "grad_norm": 1.703125,
      "learning_rate": 0.00047995091277913535,
      "loss": 4.8672,
      "mean_token_accuracy": 0.20975452810525894,
      "num_tokens": 31416980.0,
      "step": 17075
    },
    {
      "entropy": 5.326786375045776,
      "epoch": 1.4266622118275978,
      "grad_norm": 1.421875,
      "learning_rate": 0.00047993862661972373,
      "loss": 5.0634,
      "mean_token_accuracy": 0.19427034854888917,
      "num_tokens": 31425759.0,
      "step": 17080
    },
    {
      "entropy": 5.294229650497437,
      "epoch": 1.4270798529903108,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004799263368726148,
      "loss": 5.0914,
      "mean_token_accuracy": 0.1982121378183365,
      "num_tokens": 31435481.0,
      "step": 17085
    },
    {
      "entropy": 5.213776445388794,
      "epoch": 1.4274974941530236,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004799140435380237,
      "loss": 4.9196,
      "mean_token_accuracy": 0.19824998527765275,
      "num_tokens": 31443924.0,
      "step": 17090
    },
    {
      "entropy": 5.21834282875061,
      "epoch": 1.4279151353157367,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047990174661616556,
      "loss": 5.068,
      "mean_token_accuracy": 0.1962093934416771,
      "num_tokens": 31453982.0,
      "step": 17095
    },
    {
      "entropy": 5.2091597557067875,
      "epoch": 1.4283327764784497,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047988944610725574,
      "loss": 5.0088,
      "mean_token_accuracy": 0.2025838553905487,
      "num_tokens": 31463292.0,
      "step": 17100
    },
    {
      "entropy": 5.21344518661499,
      "epoch": 1.4287504176411627,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004798771420115095,
      "loss": 4.9766,
      "mean_token_accuracy": 0.20063523948192596,
      "num_tokens": 31472314.0,
      "step": 17105
    },
    {
      "entropy": 5.214603853225708,
      "epoch": 1.4291680588038758,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004798648343291423,
      "loss": 4.9879,
      "mean_token_accuracy": 0.20083611607551574,
      "num_tokens": 31482231.0,
      "step": 17110
    },
    {
      "entropy": 5.122223043441773,
      "epoch": 1.4295856999665886,
      "grad_norm": 1.40625,
      "learning_rate": 0.00047985252306036966,
      "loss": 4.8929,
      "mean_token_accuracy": 0.20760913640260698,
      "num_tokens": 31491405.0,
      "step": 17115
    },
    {
      "entropy": 5.257683992385864,
      "epoch": 1.4300033411293018,
      "grad_norm": 4.125,
      "learning_rate": 0.00047984020820540695,
      "loss": 5.0514,
      "mean_token_accuracy": 0.19425065219402313,
      "num_tokens": 31501662.0,
      "step": 17120
    },
    {
      "entropy": 5.284451484680176,
      "epoch": 1.4304209822920146,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047982788976446984,
      "loss": 5.0162,
      "mean_token_accuracy": 0.20174674391746522,
      "num_tokens": 31510671.0,
      "step": 17125
    },
    {
      "entropy": 5.225034189224243,
      "epoch": 1.4308386234547277,
      "grad_norm": 1.6171875,
      "learning_rate": 0.000479815567737774,
      "loss": 4.909,
      "mean_token_accuracy": 0.21071716099977494,
      "num_tokens": 31519309.0,
      "step": 17130
    },
    {
      "entropy": 5.202015733718872,
      "epoch": 1.4312562646174407,
      "grad_norm": 1.5,
      "learning_rate": 0.000479803242125535,
      "loss": 4.9541,
      "mean_token_accuracy": 0.20001513957977296,
      "num_tokens": 31528540.0,
      "step": 17135
    },
    {
      "entropy": 5.228109359741211,
      "epoch": 1.4316739057801537,
      "grad_norm": 1.453125,
      "learning_rate": 0.00047979091292796886,
      "loss": 5.0038,
      "mean_token_accuracy": 0.20628512054681777,
      "num_tokens": 31538035.0,
      "step": 17140
    },
    {
      "entropy": 5.229824018478394,
      "epoch": 1.4320915469428668,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004797785801452911,
      "loss": 5.0491,
      "mean_token_accuracy": 0.20120821595191957,
      "num_tokens": 31548197.0,
      "step": 17145
    },
    {
      "entropy": 5.286081695556641,
      "epoch": 1.4325091881055796,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004797662437777179,
      "loss": 5.0062,
      "mean_token_accuracy": 0.20379914939403534,
      "num_tokens": 31557268.0,
      "step": 17150
    },
    {
      "entropy": 5.251896858215332,
      "epoch": 1.4329268292682926,
      "grad_norm": 1.453125,
      "learning_rate": 0.00047975390382546523,
      "loss": 5.0287,
      "mean_token_accuracy": 0.19902937710285187,
      "num_tokens": 31566503.0,
      "step": 17155
    },
    {
      "entropy": 5.182585716247559,
      "epoch": 1.4333444704310057,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00047974156028874894,
      "loss": 5.0058,
      "mean_token_accuracy": 0.19718464314937592,
      "num_tokens": 31575754.0,
      "step": 17160
    },
    {
      "entropy": 5.219321012496948,
      "epoch": 1.4337621115937187,
      "grad_norm": 1.375,
      "learning_rate": 0.0004797292131677851,
      "loss": 5.045,
      "mean_token_accuracy": 0.1930239111185074,
      "num_tokens": 31585172.0,
      "step": 17165
    },
    {
      "entropy": 5.132456588745117,
      "epoch": 1.4341797527564317,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00047971686246279014,
      "loss": 5.0454,
      "mean_token_accuracy": 0.2010883167386055,
      "num_tokens": 31594011.0,
      "step": 17170
    },
    {
      "entropy": 5.189423656463623,
      "epoch": 1.4345973939191445,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004797045081739799,
      "loss": 4.8743,
      "mean_token_accuracy": 0.2083294630050659,
      "num_tokens": 31602598.0,
      "step": 17175
    },
    {
      "entropy": 5.244989252090454,
      "epoch": 1.4350150350818578,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004796921503015709,
      "loss": 5.0013,
      "mean_token_accuracy": 0.20149212777614595,
      "num_tokens": 31611589.0,
      "step": 17180
    },
    {
      "entropy": 5.199524784088135,
      "epoch": 1.4354326762445706,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00047967978884577944,
      "loss": 4.9332,
      "mean_token_accuracy": 0.2007725641131401,
      "num_tokens": 31620502.0,
      "step": 17185
    },
    {
      "entropy": 5.244508695602417,
      "epoch": 1.4358503174072836,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004796674238068219,
      "loss": 5.064,
      "mean_token_accuracy": 0.1874406471848488,
      "num_tokens": 31630209.0,
      "step": 17190
    },
    {
      "entropy": 5.241157865524292,
      "epoch": 1.4362679585699967,
      "grad_norm": 1.4375,
      "learning_rate": 0.00047965505518491474,
      "loss": 4.9905,
      "mean_token_accuracy": 0.19980623126029967,
      "num_tokens": 31639346.0,
      "step": 17195
    },
    {
      "entropy": 5.25229811668396,
      "epoch": 1.4366855997327097,
      "grad_norm": 1.8828125,
      "learning_rate": 0.00047964268298027456,
      "loss": 5.005,
      "mean_token_accuracy": 0.19447385966777803,
      "num_tokens": 31647482.0,
      "step": 17200
    },
    {
      "entropy": 5.183177137374878,
      "epoch": 1.4371032408954227,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047963030719311777,
      "loss": 4.9552,
      "mean_token_accuracy": 0.2121869996190071,
      "num_tokens": 31656218.0,
      "step": 17205
    },
    {
      "entropy": 5.241290616989136,
      "epoch": 1.4375208820581356,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004796179278236612,
      "loss": 5.0011,
      "mean_token_accuracy": 0.19283705949783325,
      "num_tokens": 31665306.0,
      "step": 17210
    },
    {
      "entropy": 5.152054119110107,
      "epoch": 1.4379385232208486,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004796055448721215,
      "loss": 4.8843,
      "mean_token_accuracy": 0.20450772047042848,
      "num_tokens": 31673689.0,
      "step": 17215
    },
    {
      "entropy": 5.201492595672607,
      "epoch": 1.4383561643835616,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00047959315833871546,
      "loss": 5.0843,
      "mean_token_accuracy": 0.19507393836975098,
      "num_tokens": 31683742.0,
      "step": 17220
    },
    {
      "entropy": 5.28100323677063,
      "epoch": 1.4387738055462747,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00047958076822365994,
      "loss": 5.032,
      "mean_token_accuracy": 0.1965644598007202,
      "num_tokens": 31693721.0,
      "step": 17225
    },
    {
      "entropy": 5.238499736785888,
      "epoch": 1.4391914467089877,
      "grad_norm": 1.46875,
      "learning_rate": 0.00047956837452717185,
      "loss": 4.9394,
      "mean_token_accuracy": 0.2086552321910858,
      "num_tokens": 31703103.0,
      "step": 17230
    },
    {
      "entropy": 5.124771785736084,
      "epoch": 1.4396090878717005,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047955597724946803,
      "loss": 4.8794,
      "mean_token_accuracy": 0.2074646160006523,
      "num_tokens": 31711718.0,
      "step": 17235
    },
    {
      "entropy": 5.160788440704346,
      "epoch": 1.4400267290344138,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004795435763907656,
      "loss": 5.0177,
      "mean_token_accuracy": 0.20585647225379944,
      "num_tokens": 31721248.0,
      "step": 17240
    },
    {
      "entropy": 5.184292316436768,
      "epoch": 1.4404443701971266,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004795311719512817,
      "loss": 4.9922,
      "mean_token_accuracy": 0.2003171682357788,
      "num_tokens": 31730109.0,
      "step": 17245
    },
    {
      "entropy": 5.364536762237549,
      "epoch": 1.4408620113598396,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047951876393123344,
      "loss": 5.2181,
      "mean_token_accuracy": 0.18434448540210724,
      "num_tokens": 31739768.0,
      "step": 17250
    },
    {
      "entropy": 5.293281316757202,
      "epoch": 1.4412796525225526,
      "grad_norm": 1.46875,
      "learning_rate": 0.00047950635233083805,
      "loss": 5.1434,
      "mean_token_accuracy": 0.19666717797517777,
      "num_tokens": 31749965.0,
      "step": 17255
    },
    {
      "entropy": 5.230672550201416,
      "epoch": 1.4416972936852657,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00047949393715031275,
      "loss": 4.9005,
      "mean_token_accuracy": 0.21367616653442384,
      "num_tokens": 31759112.0,
      "step": 17260
    },
    {
      "entropy": 5.193408441543579,
      "epoch": 1.4421149348479787,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047948151838987505,
      "loss": 4.9279,
      "mean_token_accuracy": 0.20653024315834045,
      "num_tokens": 31768535.0,
      "step": 17265
    },
    {
      "entropy": 5.124992990493775,
      "epoch": 1.4425325760106915,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00047946909604974205,
      "loss": 4.9325,
      "mean_token_accuracy": 0.2074887380003929,
      "num_tokens": 31777088.0,
      "step": 17270
    },
    {
      "entropy": 5.1616363525390625,
      "epoch": 1.4429502171734045,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004794566701301315,
      "loss": 4.8737,
      "mean_token_accuracy": 0.20246465802192687,
      "num_tokens": 31785895.0,
      "step": 17275
    },
    {
      "entropy": 5.2473108768463135,
      "epoch": 1.4433678583361176,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047944424063126083,
      "loss": 5.0359,
      "mean_token_accuracy": 0.20124829560518265,
      "num_tokens": 31795404.0,
      "step": 17280
    },
    {
      "entropy": 5.208041524887085,
      "epoch": 1.4437854994988306,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004794318075533477,
      "loss": 5.0413,
      "mean_token_accuracy": 0.19252155274152755,
      "num_tokens": 31804899.0,
      "step": 17285
    },
    {
      "entropy": 5.1714130401611325,
      "epoch": 1.4442031406615436,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004794193708966096,
      "loss": 4.9336,
      "mean_token_accuracy": 0.20391966551542282,
      "num_tokens": 31814055.0,
      "step": 17290
    },
    {
      "entropy": 5.173014974594116,
      "epoch": 1.4446207818242565,
      "grad_norm": 1.453125,
      "learning_rate": 0.00047940693066126445,
      "loss": 4.9509,
      "mean_token_accuracy": 0.21178096979856492,
      "num_tokens": 31823193.0,
      "step": 17295
    },
    {
      "entropy": 5.220968008041382,
      "epoch": 1.4450384229869697,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004793944868475299,
      "loss": 4.9769,
      "mean_token_accuracy": 0.2043525755405426,
      "num_tokens": 31832359.0,
      "step": 17300
    },
    {
      "entropy": 5.170807647705078,
      "epoch": 1.4454560641496825,
      "grad_norm": 1.625,
      "learning_rate": 0.0004793820394556238,
      "loss": 4.9406,
      "mean_token_accuracy": 0.2016233026981354,
      "num_tokens": 31841477.0,
      "step": 17305
    },
    {
      "entropy": 5.2088179111480715,
      "epoch": 1.4458737053123956,
      "grad_norm": 5.71875,
      "learning_rate": 0.0004793695884857641,
      "loss": 5.0294,
      "mean_token_accuracy": 0.19429323375225066,
      "num_tokens": 31851068.0,
      "step": 17310
    },
    {
      "entropy": 5.159500503540039,
      "epoch": 1.4462913464751086,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004793571339381687,
      "loss": 4.9111,
      "mean_token_accuracy": 0.20449647605419158,
      "num_tokens": 31860026.0,
      "step": 17315
    },
    {
      "entropy": 5.268568229675293,
      "epoch": 1.4467089876378216,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004793446758130558,
      "loss": 4.9585,
      "mean_token_accuracy": 0.20257394015789032,
      "num_tokens": 31868654.0,
      "step": 17320
    },
    {
      "entropy": 5.173357582092285,
      "epoch": 1.4471266288005347,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00047933221411064334,
      "loss": 4.9301,
      "mean_token_accuracy": 0.21179575026035308,
      "num_tokens": 31877801.0,
      "step": 17325
    },
    {
      "entropy": 5.1579560279846195,
      "epoch": 1.4475442699632475,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004793197488311494,
      "loss": 4.9804,
      "mean_token_accuracy": 0.19419656991958617,
      "num_tokens": 31887266.0,
      "step": 17330
    },
    {
      "entropy": 5.199007511138916,
      "epoch": 1.4479619111259605,
      "grad_norm": 1.546875,
      "learning_rate": 0.00047930727997479253,
      "loss": 4.9057,
      "mean_token_accuracy": 0.2117537423968315,
      "num_tokens": 31895957.0,
      "step": 17335
    },
    {
      "entropy": 5.225063514709473,
      "epoch": 1.4483795522886735,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004792948075417906,
      "loss": 5.0547,
      "mean_token_accuracy": 0.19289455711841583,
      "num_tokens": 31904443.0,
      "step": 17340
    },
    {
      "entropy": 5.201251792907715,
      "epoch": 1.4487971934513866,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004792823315323622,
      "loss": 4.9391,
      "mean_token_accuracy": 0.19901740550994873,
      "num_tokens": 31913534.0,
      "step": 17345
    },
    {
      "entropy": 5.266697645187378,
      "epoch": 1.4492148346140996,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004792698519467258,
      "loss": 5.0256,
      "mean_token_accuracy": 0.2009579211473465,
      "num_tokens": 31922905.0,
      "step": 17350
    },
    {
      "entropy": 5.256600666046142,
      "epoch": 1.4496324757768126,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004792573687850997,
      "loss": 5.0335,
      "mean_token_accuracy": 0.19662017226219178,
      "num_tokens": 31932861.0,
      "step": 17355
    },
    {
      "entropy": 5.282006406784058,
      "epoch": 1.4500501169395257,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004792448820477025,
      "loss": 4.9958,
      "mean_token_accuracy": 0.19949227422475815,
      "num_tokens": 31941100.0,
      "step": 17360
    },
    {
      "entropy": 5.175850582122803,
      "epoch": 1.4504677581022385,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004792323917347528,
      "loss": 4.9712,
      "mean_token_accuracy": 0.20323652476072313,
      "num_tokens": 31949860.0,
      "step": 17365
    },
    {
      "entropy": 5.123331594467163,
      "epoch": 1.4508853992649515,
      "grad_norm": 1.75,
      "learning_rate": 0.0004792198978464692,
      "loss": 4.8575,
      "mean_token_accuracy": 0.20322991758584977,
      "num_tokens": 31958216.0,
      "step": 17370
    },
    {
      "entropy": 5.2102151870727536,
      "epoch": 1.4513030404276646,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004792074003830705,
      "loss": 4.9056,
      "mean_token_accuracy": 0.21306131184101104,
      "num_tokens": 31966967.0,
      "step": 17375
    },
    {
      "entropy": 5.190389776229859,
      "epoch": 1.4517206815903776,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004791948993447754,
      "loss": 4.9699,
      "mean_token_accuracy": 0.2055326968431473,
      "num_tokens": 31976852.0,
      "step": 17380
    },
    {
      "entropy": 5.170652675628662,
      "epoch": 1.4521383227530906,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004791823947318029,
      "loss": 4.9133,
      "mean_token_accuracy": 0.20908245742321013,
      "num_tokens": 31985966.0,
      "step": 17385
    },
    {
      "entropy": 5.1745726585388185,
      "epoch": 1.4525559639158034,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004791698865443717,
      "loss": 4.9196,
      "mean_token_accuracy": 0.2125571459531784,
      "num_tokens": 31994897.0,
      "step": 17390
    },
    {
      "entropy": 5.160941505432129,
      "epoch": 1.4529736050785165,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047915737478270093,
      "loss": 4.9548,
      "mean_token_accuracy": 0.20056630671024323,
      "num_tokens": 32004504.0,
      "step": 17395
    },
    {
      "entropy": 5.203639364242553,
      "epoch": 1.4533912462412295,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004791448594470095,
      "loss": 4.9464,
      "mean_token_accuracy": 0.20237859338521957,
      "num_tokens": 32012907.0,
      "step": 17400
    },
    {
      "entropy": 5.200242280960083,
      "epoch": 1.4538088874039425,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004791323405375167,
      "loss": 4.878,
      "mean_token_accuracy": 0.20787356793880463,
      "num_tokens": 32021778.0,
      "step": 17405
    },
    {
      "entropy": 5.254102659225464,
      "epoch": 1.4542265285666556,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004791198180544415,
      "loss": 4.9832,
      "mean_token_accuracy": 0.20261708199977874,
      "num_tokens": 32030471.0,
      "step": 17410
    },
    {
      "entropy": 5.261766242980957,
      "epoch": 1.4546441697293686,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004791072919980032,
      "loss": 5.0774,
      "mean_token_accuracy": 0.19860900640487672,
      "num_tokens": 32039747.0,
      "step": 17415
    },
    {
      "entropy": 5.2372959613800045,
      "epoch": 1.4550618108920816,
      "grad_norm": 1.40625,
      "learning_rate": 0.00047909476236842105,
      "loss": 5.0059,
      "mean_token_accuracy": 0.1988307237625122,
      "num_tokens": 32048911.0,
      "step": 17420
    },
    {
      "entropy": 5.232951259613037,
      "epoch": 1.4554794520547945,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004790822291659144,
      "loss": 5.0289,
      "mean_token_accuracy": 0.1905501589179039,
      "num_tokens": 32058942.0,
      "step": 17425
    },
    {
      "entropy": 5.232475709915161,
      "epoch": 1.4558970932175075,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047906969239070276,
      "loss": 5.0734,
      "mean_token_accuracy": 0.19302746504545212,
      "num_tokens": 32067828.0,
      "step": 17430
    },
    {
      "entropy": 5.181407070159912,
      "epoch": 1.4563147343802205,
      "grad_norm": 1.375,
      "learning_rate": 0.00047905715204300543,
      "loss": 4.9248,
      "mean_token_accuracy": 0.20554703176021577,
      "num_tokens": 32077765.0,
      "step": 17435
    },
    {
      "entropy": 5.1658882141113285,
      "epoch": 1.4567323755429336,
      "grad_norm": 1.7265625,
      "learning_rate": 0.000479044608123042,
      "loss": 4.9731,
      "mean_token_accuracy": 0.20378677248954774,
      "num_tokens": 32087525.0,
      "step": 17440
    },
    {
      "entropy": 5.29058952331543,
      "epoch": 1.4571500167056466,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004790320606310322,
      "loss": 5.0976,
      "mean_token_accuracy": 0.19415424317121505,
      "num_tokens": 32097005.0,
      "step": 17445
    },
    {
      "entropy": 5.3188224792480465,
      "epoch": 1.4575676578683594,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004790195095671956,
      "loss": 5.0521,
      "mean_token_accuracy": 0.19000907689332963,
      "num_tokens": 32106761.0,
      "step": 17450
    },
    {
      "entropy": 5.270943880081177,
      "epoch": 1.4579852990310724,
      "grad_norm": 1.625,
      "learning_rate": 0.0004790069549317518,
      "loss": 4.9593,
      "mean_token_accuracy": 0.196079520881176,
      "num_tokens": 32115803.0,
      "step": 17455
    },
    {
      "entropy": 5.026858949661255,
      "epoch": 1.4584029401937855,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004789943967249207,
      "loss": 4.8749,
      "mean_token_accuracy": 0.21214470863342286,
      "num_tokens": 32124889.0,
      "step": 17460
    },
    {
      "entropy": 5.100525569915772,
      "epoch": 1.4588205813564985,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004789818349469222,
      "loss": 4.9562,
      "mean_token_accuracy": 0.20787112563848495,
      "num_tokens": 32133765.0,
      "step": 17465
    },
    {
      "entropy": 5.287478351593018,
      "epoch": 1.4592382225192115,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047896926959797615,
      "loss": 5.0073,
      "mean_token_accuracy": 0.20020412504673005,
      "num_tokens": 32142890.0,
      "step": 17470
    },
    {
      "entropy": 5.224126672744751,
      "epoch": 1.4596558636819246,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004789567006783025,
      "loss": 4.9429,
      "mean_token_accuracy": 0.20328955501317977,
      "num_tokens": 32151791.0,
      "step": 17475
    },
    {
      "entropy": 5.132331800460816,
      "epoch": 1.4600735048446376,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004789441281881213,
      "loss": 4.9947,
      "mean_token_accuracy": 0.2075645163655281,
      "num_tokens": 32160313.0,
      "step": 17480
    },
    {
      "entropy": 5.201134252548218,
      "epoch": 1.4604911460073504,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004789315521276526,
      "loss": 4.9665,
      "mean_token_accuracy": 0.2013493612408638,
      "num_tokens": 32170044.0,
      "step": 17485
    },
    {
      "entropy": 5.300574111938476,
      "epoch": 1.4609087871700634,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004789189724971167,
      "loss": 5.0193,
      "mean_token_accuracy": 0.19656626135110855,
      "num_tokens": 32178907.0,
      "step": 17490
    },
    {
      "entropy": 5.189722394943237,
      "epoch": 1.4613264283327765,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00047890638929673366,
      "loss": 4.9482,
      "mean_token_accuracy": 0.19827573448419572,
      "num_tokens": 32188197.0,
      "step": 17495
    },
    {
      "entropy": 5.222075271606445,
      "epoch": 1.4617440694954895,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004788938025267238,
      "loss": 5.0144,
      "mean_token_accuracy": 0.1993713229894638,
      "num_tokens": 32198206.0,
      "step": 17500
    },
    {
      "entropy": 5.344067764282227,
      "epoch": 1.4621617106582026,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047888121218730756,
      "loss": 5.1133,
      "mean_token_accuracy": 0.2009824275970459,
      "num_tokens": 32207306.0,
      "step": 17505
    },
    {
      "entropy": 5.207397031784057,
      "epoch": 1.4625793518209154,
      "grad_norm": 1.421875,
      "learning_rate": 0.00047886861827870533,
      "loss": 4.9387,
      "mean_token_accuracy": 0.20389721989631654,
      "num_tokens": 32215945.0,
      "step": 17510
    },
    {
      "entropy": 5.120438909530639,
      "epoch": 1.4629969929836284,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004788560208011375,
      "loss": 4.8864,
      "mean_token_accuracy": 0.20538726300001145,
      "num_tokens": 32224601.0,
      "step": 17515
    },
    {
      "entropy": 5.084963655471801,
      "epoch": 1.4634146341463414,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004788434197548246,
      "loss": 4.857,
      "mean_token_accuracy": 0.21329005807638168,
      "num_tokens": 32233265.0,
      "step": 17520
    },
    {
      "entropy": 5.168479108810425,
      "epoch": 1.4638322753090545,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004788308151399873,
      "loss": 4.8719,
      "mean_token_accuracy": 0.21333863139152526,
      "num_tokens": 32242749.0,
      "step": 17525
    },
    {
      "entropy": 5.246595621109009,
      "epoch": 1.4642499164717675,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00047881820695684624,
      "loss": 4.996,
      "mean_token_accuracy": 0.2016911193728447,
      "num_tokens": 32251295.0,
      "step": 17530
    },
    {
      "entropy": 5.16205415725708,
      "epoch": 1.4646675576344805,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004788055952056221,
      "loss": 4.9603,
      "mean_token_accuracy": 0.20636319518089294,
      "num_tokens": 32261099.0,
      "step": 17535
    },
    {
      "entropy": 5.191459798812867,
      "epoch": 1.4650851987971936,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00047879297988653576,
      "loss": 4.9991,
      "mean_token_accuracy": 0.20415739715099335,
      "num_tokens": 32269420.0,
      "step": 17540
    },
    {
      "entropy": 5.1766094207763675,
      "epoch": 1.4655028399599064,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047878036099980795,
      "loss": 4.8934,
      "mean_token_accuracy": 0.20163794606924057,
      "num_tokens": 32278455.0,
      "step": 17545
    },
    {
      "entropy": 5.200012111663819,
      "epoch": 1.4659204811226194,
      "grad_norm": 3.34375,
      "learning_rate": 0.00047876773854565964,
      "loss": 5.0164,
      "mean_token_accuracy": 0.1969103157520294,
      "num_tokens": 32287292.0,
      "step": 17550
    },
    {
      "entropy": 5.265354824066162,
      "epoch": 1.4663381222853324,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004787551125243118,
      "loss": 5.0459,
      "mean_token_accuracy": 0.19428304135799407,
      "num_tokens": 32296384.0,
      "step": 17555
    },
    {
      "entropy": 5.133001756668091,
      "epoch": 1.4667557634480455,
      "grad_norm": 1.25,
      "learning_rate": 0.0004787424829359854,
      "loss": 4.9554,
      "mean_token_accuracy": 0.2087532490491867,
      "num_tokens": 32304908.0,
      "step": 17560
    },
    {
      "entropy": 5.222298574447632,
      "epoch": 1.4671734046107585,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004787298497809017,
      "loss": 4.9814,
      "mean_token_accuracy": 0.20500562936067582,
      "num_tokens": 32314843.0,
      "step": 17565
    },
    {
      "entropy": 5.211683320999145,
      "epoch": 1.4675910457734713,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004787172130592817,
      "loss": 4.8712,
      "mean_token_accuracy": 0.2102135717868805,
      "num_tokens": 32324101.0,
      "step": 17570
    },
    {
      "entropy": 5.128016710281372,
      "epoch": 1.4680086869361844,
      "grad_norm": 1.5625,
      "learning_rate": 0.00047870457277134667,
      "loss": 4.9648,
      "mean_token_accuracy": 0.20789317637681962,
      "num_tokens": 32333736.0,
      "step": 17575
    },
    {
      "entropy": 5.193919849395752,
      "epoch": 1.4684263280988974,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004786919289173179,
      "loss": 4.9913,
      "mean_token_accuracy": 0.2018069714307785,
      "num_tokens": 32342369.0,
      "step": 17580
    },
    {
      "entropy": 5.238475513458252,
      "epoch": 1.4688439692616104,
      "grad_norm": 1.546875,
      "learning_rate": 0.00047867928149741666,
      "loss": 4.9584,
      "mean_token_accuracy": 0.20656982958316802,
      "num_tokens": 32351576.0,
      "step": 17585
    },
    {
      "entropy": 5.3105370044708256,
      "epoch": 1.4692616104243235,
      "grad_norm": 3.78125,
      "learning_rate": 0.00047866663051186444,
      "loss": 5.0166,
      "mean_token_accuracy": 0.20221993178129197,
      "num_tokens": 32360355.0,
      "step": 17590
    },
    {
      "entropy": 5.222168731689453,
      "epoch": 1.4696792515870365,
      "grad_norm": 1.5,
      "learning_rate": 0.0004786539759608828,
      "loss": 5.0055,
      "mean_token_accuracy": 0.20137722045183182,
      "num_tokens": 32368611.0,
      "step": 17595
    },
    {
      "entropy": 5.253585147857666,
      "epoch": 1.4700968927497495,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00047864131784469314,
      "loss": 5.041,
      "mean_token_accuracy": 0.19858877062797547,
      "num_tokens": 32377753.0,
      "step": 17600
    },
    {
      "entropy": 5.2596118450164795,
      "epoch": 1.4705145339124623,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000478628656163517,
      "loss": 5.1045,
      "mean_token_accuracy": 0.1905035436153412,
      "num_tokens": 32388135.0,
      "step": 17605
    },
    {
      "entropy": 5.198094367980957,
      "epoch": 1.4709321750751754,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004786159909175762,
      "loss": 4.9795,
      "mean_token_accuracy": 0.20993900001049043,
      "num_tokens": 32397149.0,
      "step": 17610
    },
    {
      "entropy": 5.147052001953125,
      "epoch": 1.4713498162378884,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004786033221070924,
      "loss": 4.9295,
      "mean_token_accuracy": 0.20904012620449067,
      "num_tokens": 32405811.0,
      "step": 17615
    },
    {
      "entropy": 5.268931913375854,
      "epoch": 1.4717674574006014,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004785906497322873,
      "loss": 4.9134,
      "mean_token_accuracy": 0.2125402495265007,
      "num_tokens": 32413800.0,
      "step": 17620
    },
    {
      "entropy": 5.225132179260254,
      "epoch": 1.4721850985633145,
      "grad_norm": 1.421875,
      "learning_rate": 0.00047857797379338284,
      "loss": 4.9684,
      "mean_token_accuracy": 0.21165940016508103,
      "num_tokens": 32422462.0,
      "step": 17625
    },
    {
      "entropy": 5.196895551681519,
      "epoch": 1.4726027397260273,
      "grad_norm": 1.421875,
      "learning_rate": 0.00047856529429060097,
      "loss": 5.0504,
      "mean_token_accuracy": 0.20006458163261415,
      "num_tokens": 32433149.0,
      "step": 17630
    },
    {
      "entropy": 5.136654901504516,
      "epoch": 1.4730203808887403,
      "grad_norm": 1.375,
      "learning_rate": 0.00047855261122416355,
      "loss": 4.9128,
      "mean_token_accuracy": 0.20536228716373445,
      "num_tokens": 32442033.0,
      "step": 17635
    },
    {
      "entropy": 5.321806335449219,
      "epoch": 1.4734380220514534,
      "grad_norm": 1.5,
      "learning_rate": 0.0004785399245942926,
      "loss": 5.0201,
      "mean_token_accuracy": 0.19595760852098465,
      "num_tokens": 32451189.0,
      "step": 17640
    },
    {
      "entropy": 5.250296688079834,
      "epoch": 1.4738556632141664,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004785272344012104,
      "loss": 4.9477,
      "mean_token_accuracy": 0.20348060578107835,
      "num_tokens": 32459929.0,
      "step": 17645
    },
    {
      "entropy": 5.121863603591919,
      "epoch": 1.4742733043768794,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004785145406451389,
      "loss": 5.0093,
      "mean_token_accuracy": 0.2009739950299263,
      "num_tokens": 32468275.0,
      "step": 17650
    },
    {
      "entropy": 5.221024465560913,
      "epoch": 1.4746909455395925,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004785018433263004,
      "loss": 4.9626,
      "mean_token_accuracy": 0.2075847029685974,
      "num_tokens": 32477833.0,
      "step": 17655
    },
    {
      "entropy": 5.139748001098633,
      "epoch": 1.4751085867023055,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047848914244491715,
      "loss": 4.884,
      "mean_token_accuracy": 0.2064445957541466,
      "num_tokens": 32486458.0,
      "step": 17660
    },
    {
      "entropy": 5.166080904006958,
      "epoch": 1.4755262278650183,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004784764380012115,
      "loss": 4.9331,
      "mean_token_accuracy": 0.20899076163768768,
      "num_tokens": 32495355.0,
      "step": 17665
    },
    {
      "entropy": 5.185784482955933,
      "epoch": 1.4759438690277313,
      "grad_norm": 1.453125,
      "learning_rate": 0.00047846372999540596,
      "loss": 4.9118,
      "mean_token_accuracy": 0.20672701001167298,
      "num_tokens": 32504083.0,
      "step": 17670
    },
    {
      "entropy": 5.168741989135742,
      "epoch": 1.4763615101904444,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047845101842772285,
      "loss": 5.036,
      "mean_token_accuracy": 0.20185963064432144,
      "num_tokens": 32513303.0,
      "step": 17675
    },
    {
      "entropy": 5.226506233215332,
      "epoch": 1.4767791513531574,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047843830329838475,
      "loss": 5.009,
      "mean_token_accuracy": 0.19830494970083237,
      "num_tokens": 32522183.0,
      "step": 17680
    },
    {
      "entropy": 5.246686935424805,
      "epoch": 1.4771967925158704,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047842558460761437,
      "loss": 5.0261,
      "mean_token_accuracy": 0.20101239085197448,
      "num_tokens": 32531443.0,
      "step": 17685
    },
    {
      "entropy": 5.241864490509033,
      "epoch": 1.4776144336785833,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00047841286235563417,
      "loss": 4.927,
      "mean_token_accuracy": 0.2079019159078598,
      "num_tokens": 32539435.0,
      "step": 17690
    },
    {
      "entropy": 5.0933818340301515,
      "epoch": 1.4780320748412963,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00047840013654266707,
      "loss": 4.9516,
      "mean_token_accuracy": 0.20341772139072417,
      "num_tokens": 32548465.0,
      "step": 17695
    },
    {
      "entropy": 5.180377960205078,
      "epoch": 1.4784497160040093,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004783874071689357,
      "loss": 4.9388,
      "mean_token_accuracy": 0.21006034463644027,
      "num_tokens": 32556511.0,
      "step": 17700
    },
    {
      "entropy": 5.270375490188599,
      "epoch": 1.4788673571667224,
      "grad_norm": 1.5,
      "learning_rate": 0.00047837467423466284,
      "loss": 5.1224,
      "mean_token_accuracy": 0.19496878385543823,
      "num_tokens": 32566077.0,
      "step": 17705
    },
    {
      "entropy": 5.300290060043335,
      "epoch": 1.4792849983294354,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047836193774007164,
      "loss": 5.0568,
      "mean_token_accuracy": 0.2033578112721443,
      "num_tokens": 32575942.0,
      "step": 17710
    },
    {
      "entropy": 5.160777473449707,
      "epoch": 1.4797026394921484,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004783491976853849,
      "loss": 4.9297,
      "mean_token_accuracy": 0.21448773443698882,
      "num_tokens": 32584595.0,
      "step": 17715
    },
    {
      "entropy": 5.1718686580657955,
      "epoch": 1.4801202806548615,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047833645407082564,
      "loss": 4.9266,
      "mean_token_accuracy": 0.20040156543254853,
      "num_tokens": 32594048.0,
      "step": 17720
    },
    {
      "entropy": 5.111905336380005,
      "epoch": 1.4805379218175743,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004783237068966171,
      "loss": 4.8184,
      "mean_token_accuracy": 0.204721699655056,
      "num_tokens": 32603197.0,
      "step": 17725
    },
    {
      "entropy": 5.1329621315002445,
      "epoch": 1.4809555629802873,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00047831095616298225,
      "loss": 4.9434,
      "mean_token_accuracy": 0.2012620151042938,
      "num_tokens": 32612882.0,
      "step": 17730
    },
    {
      "entropy": 5.205880260467529,
      "epoch": 1.4813732041430003,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004782982018701444,
      "loss": 4.9598,
      "mean_token_accuracy": 0.2072032406926155,
      "num_tokens": 32622425.0,
      "step": 17735
    },
    {
      "entropy": 5.348980522155761,
      "epoch": 1.4817908453057134,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004782854440183268,
      "loss": 5.0581,
      "mean_token_accuracy": 0.19735749661922455,
      "num_tokens": 32631295.0,
      "step": 17740
    },
    {
      "entropy": 5.333075380325317,
      "epoch": 1.4822084864684264,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004782726826077528,
      "loss": 5.0411,
      "mean_token_accuracy": 0.19946993738412858,
      "num_tokens": 32639476.0,
      "step": 17745
    },
    {
      "entropy": 5.180958604812622,
      "epoch": 1.4826261276311392,
      "grad_norm": 1.5625,
      "learning_rate": 0.00047825991763864587,
      "loss": 4.9597,
      "mean_token_accuracy": 0.2020549312233925,
      "num_tokens": 32648704.0,
      "step": 17750
    },
    {
      "entropy": 5.232555246353149,
      "epoch": 1.4830437687938522,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047824714911122934,
      "loss": 5.0605,
      "mean_token_accuracy": 0.20192982107400895,
      "num_tokens": 32658258.0,
      "step": 17755
    },
    {
      "entropy": 5.233055448532104,
      "epoch": 1.4834614099565653,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047823437702572684,
      "loss": 5.0114,
      "mean_token_accuracy": 0.19933263659477235,
      "num_tokens": 32668002.0,
      "step": 17760
    },
    {
      "entropy": 5.235143709182739,
      "epoch": 1.4838790511192783,
      "grad_norm": 1.4609375,
      "learning_rate": 0.000478221601382362,
      "loss": 5.0497,
      "mean_token_accuracy": 0.19815564304590225,
      "num_tokens": 32677961.0,
      "step": 17765
    },
    {
      "entropy": 5.227993106842041,
      "epoch": 1.4842966922819913,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047820882218135835,
      "loss": 4.9297,
      "mean_token_accuracy": 0.20839090943336486,
      "num_tokens": 32687523.0,
      "step": 17770
    },
    {
      "entropy": 5.1344880104064945,
      "epoch": 1.4847143334447044,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047819603942293964,
      "loss": 4.9162,
      "mean_token_accuracy": 0.20584795027971267,
      "num_tokens": 32696563.0,
      "step": 17775
    },
    {
      "entropy": 5.270570278167725,
      "epoch": 1.4851319746074174,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004781832531073297,
      "loss": 4.9941,
      "mean_token_accuracy": 0.19855001866817473,
      "num_tokens": 32704928.0,
      "step": 17780
    },
    {
      "entropy": 5.272962522506714,
      "epoch": 1.4855496157701302,
      "grad_norm": 1.6875,
      "learning_rate": 0.00047817046323475246,
      "loss": 5.0238,
      "mean_token_accuracy": 0.19902665317058563,
      "num_tokens": 32714228.0,
      "step": 17785
    },
    {
      "entropy": 5.229912281036377,
      "epoch": 1.4859672569328433,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004781576698054316,
      "loss": 4.9842,
      "mean_token_accuracy": 0.19830787032842637,
      "num_tokens": 32724164.0,
      "step": 17790
    },
    {
      "entropy": 5.172384452819824,
      "epoch": 1.4863848980955563,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004781448728195912,
      "loss": 4.9809,
      "mean_token_accuracy": 0.20358891487121583,
      "num_tokens": 32732516.0,
      "step": 17795
    },
    {
      "entropy": 5.273459482192993,
      "epoch": 1.4868025392582693,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047813207227745527,
      "loss": 5.1338,
      "mean_token_accuracy": 0.1911216303706169,
      "num_tokens": 32741819.0,
      "step": 17800
    },
    {
      "entropy": 5.22867784500122,
      "epoch": 1.4872201804209824,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047811926817924785,
      "loss": 4.912,
      "mean_token_accuracy": 0.21443323493003846,
      "num_tokens": 32751255.0,
      "step": 17805
    },
    {
      "entropy": 5.32355489730835,
      "epoch": 1.4876378215836952,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047810646052519324,
      "loss": 5.0654,
      "mean_token_accuracy": 0.2000792756676674,
      "num_tokens": 32760275.0,
      "step": 17810
    },
    {
      "entropy": 5.239915609359741,
      "epoch": 1.4880554627464082,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004780936493155155,
      "loss": 4.9704,
      "mean_token_accuracy": 0.20597456395626068,
      "num_tokens": 32769146.0,
      "step": 17815
    },
    {
      "entropy": 5.20129976272583,
      "epoch": 1.4884731039091212,
      "grad_norm": 1.6875,
      "learning_rate": 0.00047808083455043893,
      "loss": 5.0809,
      "mean_token_accuracy": 0.19879241436719894,
      "num_tokens": 32778523.0,
      "step": 17820
    },
    {
      "entropy": 5.176807165145874,
      "epoch": 1.4888907450718343,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047806801623018804,
      "loss": 4.8454,
      "mean_token_accuracy": 0.21314631551504135,
      "num_tokens": 32786746.0,
      "step": 17825
    },
    {
      "entropy": 5.220636034011841,
      "epoch": 1.4893083862345473,
      "grad_norm": 1.421875,
      "learning_rate": 0.000478055194354987,
      "loss": 4.9577,
      "mean_token_accuracy": 0.20959661155939102,
      "num_tokens": 32796396.0,
      "step": 17830
    },
    {
      "entropy": 5.1357948780059814,
      "epoch": 1.4897260273972603,
      "grad_norm": 1.25,
      "learning_rate": 0.00047804236892506034,
      "loss": 4.9966,
      "mean_token_accuracy": 0.20616236478090286,
      "num_tokens": 32806329.0,
      "step": 17835
    },
    {
      "entropy": 5.227914047241211,
      "epoch": 1.4901436685599734,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047802953994063266,
      "loss": 4.9634,
      "mean_token_accuracy": 0.2039125770330429,
      "num_tokens": 32815959.0,
      "step": 17840
    },
    {
      "entropy": 5.221332454681397,
      "epoch": 1.4905613097226862,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047801670740192843,
      "loss": 4.9393,
      "mean_token_accuracy": 0.2065960183739662,
      "num_tokens": 32824762.0,
      "step": 17845
    },
    {
      "entropy": 5.084002685546875,
      "epoch": 1.4909789508853992,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047800387130917245,
      "loss": 4.8737,
      "mean_token_accuracy": 0.20891590118408204,
      "num_tokens": 32833949.0,
      "step": 17850
    },
    {
      "entropy": 5.251035118103028,
      "epoch": 1.4913965920481123,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047799103166258926,
      "loss": 5.0199,
      "mean_token_accuracy": 0.19309116899967194,
      "num_tokens": 32843428.0,
      "step": 17855
    },
    {
      "entropy": 5.398384523391724,
      "epoch": 1.4918142332108253,
      "grad_norm": 1.578125,
      "learning_rate": 0.00047797818846240383,
      "loss": 5.1427,
      "mean_token_accuracy": 0.19500296860933303,
      "num_tokens": 32853878.0,
      "step": 17860
    },
    {
      "entropy": 5.184971952438355,
      "epoch": 1.4922318743735383,
      "grad_norm": 1.375,
      "learning_rate": 0.0004779653417088408,
      "loss": 4.9717,
      "mean_token_accuracy": 0.19595676362514497,
      "num_tokens": 32863011.0,
      "step": 17865
    },
    {
      "entropy": 5.186335229873658,
      "epoch": 1.4926495155362511,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00047795249140212523,
      "loss": 4.9808,
      "mean_token_accuracy": 0.20183979570865632,
      "num_tokens": 32872238.0,
      "step": 17870
    },
    {
      "entropy": 5.226812934875488,
      "epoch": 1.4930671566989642,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047793963754248195,
      "loss": 4.9683,
      "mean_token_accuracy": 0.20124614089727402,
      "num_tokens": 32880382.0,
      "step": 17875
    },
    {
      "entropy": 5.254161643981933,
      "epoch": 1.4934847978616772,
      "grad_norm": 1.328125,
      "learning_rate": 0.000477926780130136,
      "loss": 4.992,
      "mean_token_accuracy": 0.2012491837143898,
      "num_tokens": 32889199.0,
      "step": 17880
    },
    {
      "entropy": 5.2795463562011715,
      "epoch": 1.4939024390243902,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047791391916531255,
      "loss": 5.1189,
      "mean_token_accuracy": 0.1915426105260849,
      "num_tokens": 32898867.0,
      "step": 17885
    },
    {
      "entropy": 5.32041301727295,
      "epoch": 1.4943200801871033,
      "grad_norm": 2.125,
      "learning_rate": 0.0004779010546482367,
      "loss": 5.114,
      "mean_token_accuracy": 0.18929582089185715,
      "num_tokens": 32907947.0,
      "step": 17890
    },
    {
      "entropy": 5.281715726852417,
      "epoch": 1.4947377213498163,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004778881865791337,
      "loss": 4.9345,
      "mean_token_accuracy": 0.2038005471229553,
      "num_tokens": 32917045.0,
      "step": 17895
    },
    {
      "entropy": 5.185208415985107,
      "epoch": 1.4951553625125293,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047787531495822866,
      "loss": 4.917,
      "mean_token_accuracy": 0.20727527290582656,
      "num_tokens": 32926644.0,
      "step": 17900
    },
    {
      "entropy": 5.18152174949646,
      "epoch": 1.4955730036752422,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00047786243978574706,
      "loss": 4.999,
      "mean_token_accuracy": 0.2007044658064842,
      "num_tokens": 32935920.0,
      "step": 17905
    },
    {
      "entropy": 5.223235464096069,
      "epoch": 1.4959906448379552,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047784956106191435,
      "loss": 4.9816,
      "mean_token_accuracy": 0.19770140200853348,
      "num_tokens": 32945561.0,
      "step": 17910
    },
    {
      "entropy": 5.215451669692993,
      "epoch": 1.4964082860006682,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004778366787869559,
      "loss": 4.9312,
      "mean_token_accuracy": 0.2064352288842201,
      "num_tokens": 32955531.0,
      "step": 17915
    },
    {
      "entropy": 5.093587350845337,
      "epoch": 1.4968259271633813,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004778237929610971,
      "loss": 4.7838,
      "mean_token_accuracy": 0.20969012677669524,
      "num_tokens": 32964001.0,
      "step": 17920
    },
    {
      "entropy": 5.228625297546387,
      "epoch": 1.4972435683260943,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004778109035845638,
      "loss": 5.0478,
      "mean_token_accuracy": 0.20487563461065292,
      "num_tokens": 32973707.0,
      "step": 17925
    },
    {
      "entropy": 5.226443862915039,
      "epoch": 1.497661209488807,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047779801065758146,
      "loss": 4.9937,
      "mean_token_accuracy": 0.19956338554620742,
      "num_tokens": 32983029.0,
      "step": 17930
    },
    {
      "entropy": 5.218592882156372,
      "epoch": 1.4980788506515201,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004777851141803758,
      "loss": 4.9611,
      "mean_token_accuracy": 0.20367655605077745,
      "num_tokens": 32992351.0,
      "step": 17935
    },
    {
      "entropy": 5.126899003982544,
      "epoch": 1.4984964918142332,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004777722141531727,
      "loss": 4.8133,
      "mean_token_accuracy": 0.2178340047597885,
      "num_tokens": 33001090.0,
      "step": 17940
    },
    {
      "entropy": 5.2349278926849365,
      "epoch": 1.4989141329769462,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047775931057619786,
      "loss": 5.0358,
      "mean_token_accuracy": 0.19449341744184495,
      "num_tokens": 33010294.0,
      "step": 17945
    },
    {
      "entropy": 5.19752721786499,
      "epoch": 1.4993317741396592,
      "grad_norm": 1.375,
      "learning_rate": 0.0004777464034496773,
      "loss": 4.9878,
      "mean_token_accuracy": 0.1985456347465515,
      "num_tokens": 33018633.0,
      "step": 17950
    },
    {
      "entropy": 5.222539186477661,
      "epoch": 1.4997494153023723,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004777334927738368,
      "loss": 5.035,
      "mean_token_accuracy": 0.19977982193231583,
      "num_tokens": 33028291.0,
      "step": 17955
    },
    {
      "entropy": 5.260934686660766,
      "epoch": 1.5001670564650853,
      "grad_norm": 1.46875,
      "learning_rate": 0.00047772057854890247,
      "loss": 4.9387,
      "mean_token_accuracy": 0.20507076978683472,
      "num_tokens": 33036603.0,
      "step": 17960
    },
    {
      "entropy": 5.211347675323486,
      "epoch": 1.5005846976277981,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047770766077510045,
      "loss": 4.9711,
      "mean_token_accuracy": 0.20392703115940095,
      "num_tokens": 33045329.0,
      "step": 17965
    },
    {
      "entropy": 5.129393148422241,
      "epoch": 1.5010023387905111,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047769473945265685,
      "loss": 4.9285,
      "mean_token_accuracy": 0.20685894936323165,
      "num_tokens": 33054753.0,
      "step": 17970
    },
    {
      "entropy": 5.159753370285034,
      "epoch": 1.5014199799532242,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004776818145817979,
      "loss": 4.9434,
      "mean_token_accuracy": 0.21154507249593735,
      "num_tokens": 33063469.0,
      "step": 17975
    },
    {
      "entropy": 5.20007848739624,
      "epoch": 1.5018376211159372,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047766888616274973,
      "loss": 5.0121,
      "mean_token_accuracy": 0.19947468042373656,
      "num_tokens": 33072247.0,
      "step": 17980
    },
    {
      "entropy": 5.232151031494141,
      "epoch": 1.5022552622786502,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047765595419573876,
      "loss": 5.0232,
      "mean_token_accuracy": 0.1999536782503128,
      "num_tokens": 33083505.0,
      "step": 17985
    },
    {
      "entropy": 5.198408746719361,
      "epoch": 1.502672903441363,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047764301868099136,
      "loss": 5.0111,
      "mean_token_accuracy": 0.20019261687994003,
      "num_tokens": 33094104.0,
      "step": 17990
    },
    {
      "entropy": 5.232410335540772,
      "epoch": 1.5030905446040763,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047763007961873407,
      "loss": 4.9357,
      "mean_token_accuracy": 0.21364189982414244,
      "num_tokens": 33102936.0,
      "step": 17995
    },
    {
      "entropy": 5.256219530105591,
      "epoch": 1.5035081857667891,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004776171370091933,
      "loss": 4.946,
      "mean_token_accuracy": 0.20755905359983445,
      "num_tokens": 33111648.0,
      "step": 18000
    },
    {
      "epoch": 1.5035081857667891,
      "eval_entropy": 4.990027291419723,
      "eval_loss": 4.989250183105469,
      "eval_mean_token_accuracy": 0.21310557920254533,
      "eval_num_tokens": 33111648.0,
      "eval_runtime": 27.5072,
      "eval_samples_per_second": 1372.839,
      "eval_steps_per_second": 171.628,
      "step": 18000
    },
    {
      "entropy": 5.158056211471558,
      "epoch": 1.5039258269295022,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004776041908525956,
      "loss": 4.9859,
      "mean_token_accuracy": 0.19459817707538604,
      "num_tokens": 33120235.0,
      "step": 18005
    },
    {
      "entropy": 5.114632606506348,
      "epoch": 1.5043434680922152,
      "grad_norm": 1.375,
      "learning_rate": 0.0004775912411491677,
      "loss": 4.9456,
      "mean_token_accuracy": 0.20728246122598648,
      "num_tokens": 33129152.0,
      "step": 18010
    },
    {
      "entropy": 5.140415906906128,
      "epoch": 1.504761109254928,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004775782878991363,
      "loss": 4.8822,
      "mean_token_accuracy": 0.2084478422999382,
      "num_tokens": 33137907.0,
      "step": 18015
    },
    {
      "entropy": 5.177781248092652,
      "epoch": 1.5051787504176413,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047756533110272813,
      "loss": 4.9473,
      "mean_token_accuracy": 0.2040211334824562,
      "num_tokens": 33146001.0,
      "step": 18020
    },
    {
      "entropy": 5.235043287277222,
      "epoch": 1.505596391580354,
      "grad_norm": 1.515625,
      "learning_rate": 0.00047755237076016997,
      "loss": 4.9432,
      "mean_token_accuracy": 0.19963305592536926,
      "num_tokens": 33155389.0,
      "step": 18025
    },
    {
      "entropy": 5.2042053699493405,
      "epoch": 1.506014032743067,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047753940687168876,
      "loss": 5.0219,
      "mean_token_accuracy": 0.20460870414972304,
      "num_tokens": 33164965.0,
      "step": 18030
    },
    {
      "entropy": 5.242608594894409,
      "epoch": 1.5064316739057801,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004775264394375114,
      "loss": 5.0583,
      "mean_token_accuracy": 0.19594694525003434,
      "num_tokens": 33174590.0,
      "step": 18035
    },
    {
      "entropy": 5.232761192321777,
      "epoch": 1.5068493150684932,
      "grad_norm": 1.34375,
      "learning_rate": 0.000477513468457865,
      "loss": 4.9377,
      "mean_token_accuracy": 0.2029197782278061,
      "num_tokens": 33183063.0,
      "step": 18040
    },
    {
      "entropy": 5.275362968444824,
      "epoch": 1.5072669562312062,
      "grad_norm": 1.734375,
      "learning_rate": 0.00047750049393297656,
      "loss": 5.0747,
      "mean_token_accuracy": 0.19670417308807372,
      "num_tokens": 33192198.0,
      "step": 18045
    },
    {
      "entropy": 5.316382503509521,
      "epoch": 1.507684597393919,
      "grad_norm": 1.8359375,
      "learning_rate": 0.00047748751586307314,
      "loss": 5.0888,
      "mean_token_accuracy": 0.19510869681835175,
      "num_tokens": 33202445.0,
      "step": 18050
    },
    {
      "entropy": 5.241377925872802,
      "epoch": 1.5081022385566323,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00047747453424838216,
      "loss": 4.9897,
      "mean_token_accuracy": 0.20355637967586518,
      "num_tokens": 33211017.0,
      "step": 18055
    },
    {
      "entropy": 5.112583875656128,
      "epoch": 1.508519879719345,
      "grad_norm": 1.4375,
      "learning_rate": 0.00047746154908913065,
      "loss": 4.942,
      "mean_token_accuracy": 0.21322618573904037,
      "num_tokens": 33220345.0,
      "step": 18060
    },
    {
      "entropy": 5.247761678695679,
      "epoch": 1.5089375208820581,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000477448560385546,
      "loss": 4.9723,
      "mean_token_accuracy": 0.20848355144262315,
      "num_tokens": 33229849.0,
      "step": 18065
    },
    {
      "entropy": 5.179928207397461,
      "epoch": 1.5093551620447712,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004774355681378557,
      "loss": 4.9179,
      "mean_token_accuracy": 0.20710987150669097,
      "num_tokens": 33239079.0,
      "step": 18070
    },
    {
      "entropy": 5.127002811431884,
      "epoch": 1.509772803207484,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047742257234628703,
      "loss": 4.8797,
      "mean_token_accuracy": 0.21095765382051468,
      "num_tokens": 33248048.0,
      "step": 18075
    },
    {
      "entropy": 5.123681735992432,
      "epoch": 1.5101904443701972,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047740957301106756,
      "loss": 4.9152,
      "mean_token_accuracy": 0.21148792952299117,
      "num_tokens": 33257173.0,
      "step": 18080
    },
    {
      "entropy": 5.187163639068603,
      "epoch": 1.51060808553291,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000477396570132425,
      "loss": 5.0032,
      "mean_token_accuracy": 0.2005190968513489,
      "num_tokens": 33266800.0,
      "step": 18085
    },
    {
      "entropy": 5.2402643203735355,
      "epoch": 1.511025726695623,
      "grad_norm": 1.421875,
      "learning_rate": 0.00047738356371058673,
      "loss": 5.0566,
      "mean_token_accuracy": 0.195840522646904,
      "num_tokens": 33276061.0,
      "step": 18090
    },
    {
      "entropy": 5.179122495651245,
      "epoch": 1.511443367858336,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004773705537457806,
      "loss": 4.8757,
      "mean_token_accuracy": 0.209619876742363,
      "num_tokens": 33284786.0,
      "step": 18095
    },
    {
      "entropy": 5.134295654296875,
      "epoch": 1.5118610090210491,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00047735754023823433,
      "loss": 4.9471,
      "mean_token_accuracy": 0.20300881266593934,
      "num_tokens": 33293849.0,
      "step": 18100
    },
    {
      "entropy": 5.123455810546875,
      "epoch": 1.5122786501837622,
      "grad_norm": 1.5,
      "learning_rate": 0.0004773445231881757,
      "loss": 4.8518,
      "mean_token_accuracy": 0.211598140001297,
      "num_tokens": 33303387.0,
      "step": 18105
    },
    {
      "entropy": 5.176509046554566,
      "epoch": 1.512696291346475,
      "grad_norm": 1.75,
      "learning_rate": 0.00047733150259583274,
      "loss": 4.8915,
      "mean_token_accuracy": 0.2068120464682579,
      "num_tokens": 33312502.0,
      "step": 18110
    },
    {
      "entropy": 5.111059188842773,
      "epoch": 1.5131139325091882,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047731847846143325,
      "loss": 4.9096,
      "mean_token_accuracy": 0.21217695325613023,
      "num_tokens": 33321909.0,
      "step": 18115
    },
    {
      "entropy": 5.09513292312622,
      "epoch": 1.513531573671901,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004773054507852052,
      "loss": 4.89,
      "mean_token_accuracy": 0.20228530168533326,
      "num_tokens": 33331249.0,
      "step": 18120
    },
    {
      "entropy": 5.202250814437866,
      "epoch": 1.513949214834614,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004772924195673768,
      "loss": 4.9448,
      "mean_token_accuracy": 0.2073888823390007,
      "num_tokens": 33340850.0,
      "step": 18125
    },
    {
      "entropy": 5.241228246688843,
      "epoch": 1.5143668559973271,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047727938480817604,
      "loss": 4.9989,
      "mean_token_accuracy": 0.19972241967916488,
      "num_tokens": 33350429.0,
      "step": 18130
    },
    {
      "entropy": 5.229430627822876,
      "epoch": 1.51478449716004,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00047726634650783117,
      "loss": 4.9752,
      "mean_token_accuracy": 0.20204821974039078,
      "num_tokens": 33359687.0,
      "step": 18135
    },
    {
      "entropy": 5.199322891235352,
      "epoch": 1.5152021383227532,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004772533046665705,
      "loss": 5.0146,
      "mean_token_accuracy": 0.19649224132299423,
      "num_tokens": 33368177.0,
      "step": 18140
    },
    {
      "entropy": 5.137014532089234,
      "epoch": 1.515619779485466,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004772402592846222,
      "loss": 4.9945,
      "mean_token_accuracy": 0.20210421532392503,
      "num_tokens": 33376604.0,
      "step": 18145
    },
    {
      "entropy": 5.179344177246094,
      "epoch": 1.516037420648179,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047722721036221467,
      "loss": 4.9045,
      "mean_token_accuracy": 0.20410909205675126,
      "num_tokens": 33385488.0,
      "step": 18150
    },
    {
      "entropy": 5.244606256484985,
      "epoch": 1.516455061810892,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004772141578995765,
      "loss": 5.0355,
      "mean_token_accuracy": 0.1931261032819748,
      "num_tokens": 33394524.0,
      "step": 18155
    },
    {
      "entropy": 5.22447509765625,
      "epoch": 1.516872702973605,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00047720110189693605,
      "loss": 5.0044,
      "mean_token_accuracy": 0.2051696926355362,
      "num_tokens": 33404300.0,
      "step": 18160
    },
    {
      "entropy": 5.178873062133789,
      "epoch": 1.5172903441363181,
      "grad_norm": 1.40625,
      "learning_rate": 0.00047718804235452193,
      "loss": 4.9362,
      "mean_token_accuracy": 0.2078246667981148,
      "num_tokens": 33414050.0,
      "step": 18165
    },
    {
      "entropy": 5.202832460403442,
      "epoch": 1.517707985299031,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004771749792725628,
      "loss": 5.0514,
      "mean_token_accuracy": 0.20441612601280212,
      "num_tokens": 33423981.0,
      "step": 18170
    },
    {
      "entropy": 5.1415407180786135,
      "epoch": 1.5181256264617442,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004771619126512873,
      "loss": 4.7952,
      "mean_token_accuracy": 0.20838343352079391,
      "num_tokens": 33433104.0,
      "step": 18175
    },
    {
      "entropy": 5.197199964523316,
      "epoch": 1.518543267624457,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00047714884249092413,
      "loss": 5.0254,
      "mean_token_accuracy": 0.19836369156837463,
      "num_tokens": 33442034.0,
      "step": 18180
    },
    {
      "entropy": 5.091469717025757,
      "epoch": 1.51896090878717,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004771357687917021,
      "loss": 4.9366,
      "mean_token_accuracy": 0.20607532262802125,
      "num_tokens": 33450431.0,
      "step": 18185
    },
    {
      "entropy": 5.254856109619141,
      "epoch": 1.519378549949883,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00047712269155385017,
      "loss": 5.072,
      "mean_token_accuracy": 0.20015305578708648,
      "num_tokens": 33459790.0,
      "step": 18190
    },
    {
      "entropy": 5.292317199707031,
      "epoch": 1.519796191112596,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047710961077759736,
      "loss": 5.0518,
      "mean_token_accuracy": 0.20105894804000854,
      "num_tokens": 33468802.0,
      "step": 18195
    },
    {
      "entropy": 5.247967290878296,
      "epoch": 1.5202138322753092,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047709652646317236,
      "loss": 4.9156,
      "mean_token_accuracy": 0.2099500060081482,
      "num_tokens": 33477756.0,
      "step": 18200
    },
    {
      "entropy": 5.180398893356323,
      "epoch": 1.520631473438022,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047708343861080447,
      "loss": 5.0348,
      "mean_token_accuracy": 0.19799203872680665,
      "num_tokens": 33487410.0,
      "step": 18205
    },
    {
      "entropy": 5.215941715240478,
      "epoch": 1.521049114600735,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004770703472207227,
      "loss": 4.9652,
      "mean_token_accuracy": 0.2026677832007408,
      "num_tokens": 33496682.0,
      "step": 18210
    },
    {
      "entropy": 5.333031845092774,
      "epoch": 1.521466755763448,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004770572522931563,
      "loss": 5.1147,
      "mean_token_accuracy": 0.1901180252432823,
      "num_tokens": 33505754.0,
      "step": 18215
    },
    {
      "entropy": 5.238153886795044,
      "epoch": 1.521884396926161,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004770441538283345,
      "loss": 4.9807,
      "mean_token_accuracy": 0.20599044412374495,
      "num_tokens": 33514904.0,
      "step": 18220
    },
    {
      "entropy": 5.168804931640625,
      "epoch": 1.522302038088874,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047703105182648656,
      "loss": 4.9388,
      "mean_token_accuracy": 0.20628890246152878,
      "num_tokens": 33523684.0,
      "step": 18225
    },
    {
      "entropy": 5.187350511550903,
      "epoch": 1.522719679251587,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004770179462878419,
      "loss": 5.0028,
      "mean_token_accuracy": 0.20178761929273606,
      "num_tokens": 33533066.0,
      "step": 18230
    },
    {
      "entropy": 5.2130934715271,
      "epoch": 1.5231373204143002,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004770048372126299,
      "loss": 4.9522,
      "mean_token_accuracy": 0.19797794371843339,
      "num_tokens": 33542048.0,
      "step": 18235
    },
    {
      "entropy": 5.178755617141723,
      "epoch": 1.523554961577013,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004769917246010801,
      "loss": 4.8599,
      "mean_token_accuracy": 0.21011954545974731,
      "num_tokens": 33550785.0,
      "step": 18240
    },
    {
      "entropy": 5.170193767547607,
      "epoch": 1.523972602739726,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004769786084534219,
      "loss": 4.9503,
      "mean_token_accuracy": 0.19236876219511032,
      "num_tokens": 33560141.0,
      "step": 18245
    },
    {
      "entropy": 5.216722583770752,
      "epoch": 1.524390243902439,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00047696548876988517,
      "loss": 4.9901,
      "mean_token_accuracy": 0.20250598788261415,
      "num_tokens": 33568866.0,
      "step": 18250
    },
    {
      "entropy": 5.187884569168091,
      "epoch": 1.5248078850651519,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004769523655506994,
      "loss": 4.9783,
      "mean_token_accuracy": 0.2027960851788521,
      "num_tokens": 33579217.0,
      "step": 18255
    },
    {
      "entropy": 5.24456934928894,
      "epoch": 1.5252255262278651,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004769392387960944,
      "loss": 5.0647,
      "mean_token_accuracy": 0.18902273327112198,
      "num_tokens": 33589608.0,
      "step": 18260
    },
    {
      "entropy": 5.234592914581299,
      "epoch": 1.525643167390578,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004769261085063,
      "loss": 4.9941,
      "mean_token_accuracy": 0.1982766017317772,
      "num_tokens": 33599263.0,
      "step": 18265
    },
    {
      "entropy": 5.11523175239563,
      "epoch": 1.526060808553291,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047691297468154594,
      "loss": 4.9711,
      "mean_token_accuracy": 0.20779789686203004,
      "num_tokens": 33607946.0,
      "step": 18270
    },
    {
      "entropy": 5.244549226760864,
      "epoch": 1.526478449716004,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004768998373220622,
      "loss": 5.0524,
      "mean_token_accuracy": 0.2006062999367714,
      "num_tokens": 33617194.0,
      "step": 18275
    },
    {
      "entropy": 5.245501375198364,
      "epoch": 1.526896090878717,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004768866964280789,
      "loss": 5.0008,
      "mean_token_accuracy": 0.20275562703609468,
      "num_tokens": 33626796.0,
      "step": 18280
    },
    {
      "entropy": 5.198130178451538,
      "epoch": 1.52731373204143,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047687355199982584,
      "loss": 4.9452,
      "mean_token_accuracy": 0.2008989632129669,
      "num_tokens": 33636290.0,
      "step": 18285
    },
    {
      "entropy": 5.139774131774902,
      "epoch": 1.5277313732041429,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004768604040375333,
      "loss": 4.9607,
      "mean_token_accuracy": 0.21182930171489717,
      "num_tokens": 33644489.0,
      "step": 18290
    },
    {
      "entropy": 5.137392234802246,
      "epoch": 1.5281490143668561,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047684725254143137,
      "loss": 4.8586,
      "mean_token_accuracy": 0.21153387129306794,
      "num_tokens": 33653677.0,
      "step": 18295
    },
    {
      "entropy": 5.106986045837402,
      "epoch": 1.528566655529569,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004768340975117504,
      "loss": 4.9215,
      "mean_token_accuracy": 0.21095658391714095,
      "num_tokens": 33662787.0,
      "step": 18300
    },
    {
      "entropy": 5.170689058303833,
      "epoch": 1.528984296692282,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004768209389487205,
      "loss": 4.9887,
      "mean_token_accuracy": 0.21428881585597992,
      "num_tokens": 33671167.0,
      "step": 18305
    },
    {
      "entropy": 5.191334199905396,
      "epoch": 1.529401937854995,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004768077768525722,
      "loss": 4.863,
      "mean_token_accuracy": 0.21616191864013673,
      "num_tokens": 33679253.0,
      "step": 18310
    },
    {
      "entropy": 5.239003849029541,
      "epoch": 1.5298195790177078,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004767946112235359,
      "loss": 4.9762,
      "mean_token_accuracy": 0.1992676541209221,
      "num_tokens": 33688863.0,
      "step": 18315
    },
    {
      "entropy": 5.160636520385742,
      "epoch": 1.530237220180421,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000476781442061842,
      "loss": 4.957,
      "mean_token_accuracy": 0.2011109247803688,
      "num_tokens": 33697201.0,
      "step": 18320
    },
    {
      "entropy": 5.249578142166138,
      "epoch": 1.5306548613431339,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047676826936772097,
      "loss": 5.0391,
      "mean_token_accuracy": 0.20177465081214904,
      "num_tokens": 33706075.0,
      "step": 18325
    },
    {
      "entropy": 5.229749631881714,
      "epoch": 1.531072502505847,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00047675509314140365,
      "loss": 5.0416,
      "mean_token_accuracy": 0.19490955621004105,
      "num_tokens": 33715290.0,
      "step": 18330
    },
    {
      "entropy": 5.141003704071045,
      "epoch": 1.53149014366856,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004767419133831206,
      "loss": 4.8956,
      "mean_token_accuracy": 0.21412067264318466,
      "num_tokens": 33724508.0,
      "step": 18335
    },
    {
      "entropy": 5.262171888351441,
      "epoch": 1.531907784831273,
      "grad_norm": 1.375,
      "learning_rate": 0.00047672873009310234,
      "loss": 5.0333,
      "mean_token_accuracy": 0.2014237403869629,
      "num_tokens": 33733333.0,
      "step": 18340
    },
    {
      "entropy": 5.145748329162598,
      "epoch": 1.532325425993986,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047671554327157996,
      "loss": 4.8829,
      "mean_token_accuracy": 0.20640962123870848,
      "num_tokens": 33741780.0,
      "step": 18345
    },
    {
      "entropy": 5.133035182952881,
      "epoch": 1.5327430671566988,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047670235291878417,
      "loss": 4.8831,
      "mean_token_accuracy": 0.20646077692508696,
      "num_tokens": 33750259.0,
      "step": 18350
    },
    {
      "entropy": 5.18579626083374,
      "epoch": 1.533160708319412,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004766891590349459,
      "loss": 4.9863,
      "mean_token_accuracy": 0.19920340031385422,
      "num_tokens": 33760383.0,
      "step": 18355
    },
    {
      "entropy": 5.230004405975341,
      "epoch": 1.533578349482125,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004766759616202961,
      "loss": 4.9067,
      "mean_token_accuracy": 0.2047243595123291,
      "num_tokens": 33768524.0,
      "step": 18360
    },
    {
      "entropy": 5.1465507507324215,
      "epoch": 1.533995990644838,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004766627606750659,
      "loss": 4.9275,
      "mean_token_accuracy": 0.2091185212135315,
      "num_tokens": 33777448.0,
      "step": 18365
    },
    {
      "entropy": 5.22729640007019,
      "epoch": 1.534413631807551,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004766495561994863,
      "loss": 5.012,
      "mean_token_accuracy": 0.20222467482089995,
      "num_tokens": 33787316.0,
      "step": 18370
    },
    {
      "entropy": 5.229552412033081,
      "epoch": 1.534831272970264,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004766363481937885,
      "loss": 4.9133,
      "mean_token_accuracy": 0.20583610385656356,
      "num_tokens": 33796945.0,
      "step": 18375
    },
    {
      "entropy": 5.237408638000488,
      "epoch": 1.535248914132977,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00047662313665820365,
      "loss": 5.036,
      "mean_token_accuracy": 0.19435563683509827,
      "num_tokens": 33807866.0,
      "step": 18380
    },
    {
      "entropy": 5.150338649749756,
      "epoch": 1.5356665552956899,
      "grad_norm": 1.375,
      "learning_rate": 0.0004766099215929632,
      "loss": 4.9583,
      "mean_token_accuracy": 0.20171258002519607,
      "num_tokens": 33817298.0,
      "step": 18385
    },
    {
      "entropy": 5.159289979934693,
      "epoch": 1.536084196458403,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047659670299829833,
      "loss": 4.9145,
      "mean_token_accuracy": 0.20904300659894942,
      "num_tokens": 33826552.0,
      "step": 18390
    },
    {
      "entropy": 5.159009790420532,
      "epoch": 1.536501837621116,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004765834808744405,
      "loss": 4.9677,
      "mean_token_accuracy": 0.20548245459794998,
      "num_tokens": 33836800.0,
      "step": 18395
    },
    {
      "entropy": 5.195048046112061,
      "epoch": 1.536919478783829,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00047657025522162123,
      "loss": 4.9257,
      "mean_token_accuracy": 0.20621144026517868,
      "num_tokens": 33846245.0,
      "step": 18400
    },
    {
      "entropy": 5.160901594161987,
      "epoch": 1.537337119946542,
      "grad_norm": 1.40625,
      "learning_rate": 0.000476557026040072,
      "loss": 4.8764,
      "mean_token_accuracy": 0.21300604343414306,
      "num_tokens": 33855100.0,
      "step": 18405
    },
    {
      "entropy": 5.075445032119751,
      "epoch": 1.5377547611092548,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047654379333002446,
      "loss": 4.8455,
      "mean_token_accuracy": 0.20802339911460876,
      "num_tokens": 33863307.0,
      "step": 18410
    },
    {
      "entropy": 5.156131076812744,
      "epoch": 1.538172402271968,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047653055709171013,
      "loss": 4.9911,
      "mean_token_accuracy": 0.19981231689453124,
      "num_tokens": 33871891.0,
      "step": 18415
    },
    {
      "entropy": 5.162611579895019,
      "epoch": 1.5385900434346809,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00047651731732536084,
      "loss": 4.9602,
      "mean_token_accuracy": 0.21149032264947892,
      "num_tokens": 33880913.0,
      "step": 18420
    },
    {
      "entropy": 5.333642911911011,
      "epoch": 1.539007684597394,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00047650407403120845,
      "loss": 5.0088,
      "mean_token_accuracy": 0.1996071606874466,
      "num_tokens": 33889838.0,
      "step": 18425
    },
    {
      "entropy": 5.28747787475586,
      "epoch": 1.539425325760107,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00047649082720948455,
      "loss": 5.0185,
      "mean_token_accuracy": 0.2020004376769066,
      "num_tokens": 33898989.0,
      "step": 18430
    },
    {
      "entropy": 5.2050680160522464,
      "epoch": 1.53984296692282,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047647757686042125,
      "loss": 4.9745,
      "mean_token_accuracy": 0.20091331899166107,
      "num_tokens": 33907884.0,
      "step": 18435
    },
    {
      "entropy": 5.210937547683716,
      "epoch": 1.540260608085533,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047646432298425047,
      "loss": 4.9799,
      "mean_token_accuracy": 0.20254916697740555,
      "num_tokens": 33916762.0,
      "step": 18440
    },
    {
      "entropy": 5.158131217956543,
      "epoch": 1.5406782492482458,
      "grad_norm": 1.375,
      "learning_rate": 0.00047645106558120417,
      "loss": 4.8785,
      "mean_token_accuracy": 0.2075303629040718,
      "num_tokens": 33925413.0,
      "step": 18445
    },
    {
      "entropy": 5.127823829650879,
      "epoch": 1.541095890410959,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004764378046515145,
      "loss": 4.9354,
      "mean_token_accuracy": 0.20416170954704285,
      "num_tokens": 33934247.0,
      "step": 18450
    },
    {
      "entropy": 5.133670377731323,
      "epoch": 1.5415135315736719,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004764245401954136,
      "loss": 4.8829,
      "mean_token_accuracy": 0.2149432361125946,
      "num_tokens": 33943610.0,
      "step": 18455
    },
    {
      "entropy": 5.3318031311035154,
      "epoch": 1.541931172736385,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004764112722131337,
      "loss": 5.1939,
      "mean_token_accuracy": 0.18520619571208954,
      "num_tokens": 33953447.0,
      "step": 18460
    },
    {
      "entropy": 5.272027158737183,
      "epoch": 1.542348813899098,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00047639800070490695,
      "loss": 4.9923,
      "mean_token_accuracy": 0.20291330218315123,
      "num_tokens": 33962702.0,
      "step": 18465
    },
    {
      "entropy": 5.271010303497315,
      "epoch": 1.5427664550618108,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004763847256709658,
      "loss": 5.0015,
      "mean_token_accuracy": 0.19439080655574797,
      "num_tokens": 33972604.0,
      "step": 18470
    },
    {
      "entropy": 5.216072273254395,
      "epoch": 1.543184096224524,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004763714471115427,
      "loss": 5.071,
      "mean_token_accuracy": 0.19482940584421157,
      "num_tokens": 33982443.0,
      "step": 18475
    },
    {
      "entropy": 5.212449026107788,
      "epoch": 1.5436017373872368,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004763581650268699,
      "loss": 4.9572,
      "mean_token_accuracy": 0.21042489856481553,
      "num_tokens": 33990688.0,
      "step": 18480
    },
    {
      "entropy": 5.235848903656006,
      "epoch": 1.5440193785499499,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047634487941718016,
      "loss": 5.0082,
      "mean_token_accuracy": 0.1993616595864296,
      "num_tokens": 33999657.0,
      "step": 18485
    },
    {
      "entropy": 5.257433938980102,
      "epoch": 1.544437019712663,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004763315902827059,
      "loss": 5.0727,
      "mean_token_accuracy": 0.1985306113958359,
      "num_tokens": 34009842.0,
      "step": 18490
    },
    {
      "entropy": 5.355138206481934,
      "epoch": 1.544854660875376,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047631829762367976,
      "loss": 5.0779,
      "mean_token_accuracy": 0.1861989676952362,
      "num_tokens": 34019600.0,
      "step": 18495
    },
    {
      "entropy": 5.224793577194214,
      "epoch": 1.545272302038089,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004763050014403346,
      "loss": 5.0539,
      "mean_token_accuracy": 0.19741141051054,
      "num_tokens": 34028256.0,
      "step": 18500
    },
    {
      "entropy": 5.21249942779541,
      "epoch": 1.5456899432008018,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047629170173290296,
      "loss": 5.0056,
      "mean_token_accuracy": 0.20040290355682372,
      "num_tokens": 34037117.0,
      "step": 18505
    },
    {
      "entropy": 5.211995077133179,
      "epoch": 1.546107584363515,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047627839850161784,
      "loss": 4.9603,
      "mean_token_accuracy": 0.20206488519906998,
      "num_tokens": 34045697.0,
      "step": 18510
    },
    {
      "entropy": 5.288525390625,
      "epoch": 1.5465252255262278,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00047626509174671214,
      "loss": 5.0957,
      "mean_token_accuracy": 0.1965210348367691,
      "num_tokens": 34055735.0,
      "step": 18515
    },
    {
      "entropy": 5.28881425857544,
      "epoch": 1.5469428666889409,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004762517814684187,
      "loss": 5.0401,
      "mean_token_accuracy": 0.19762174934148788,
      "num_tokens": 34065780.0,
      "step": 18520
    },
    {
      "entropy": 5.180434942245483,
      "epoch": 1.547360507851654,
      "grad_norm": 1.375,
      "learning_rate": 0.00047623846766697056,
      "loss": 4.9598,
      "mean_token_accuracy": 0.2046153262257576,
      "num_tokens": 34075314.0,
      "step": 18525
    },
    {
      "entropy": 5.267871856689453,
      "epoch": 1.5477781490143667,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00047622515034260077,
      "loss": 5.0405,
      "mean_token_accuracy": 0.2064945325255394,
      "num_tokens": 34085475.0,
      "step": 18530
    },
    {
      "entropy": 5.299253129959107,
      "epoch": 1.54819579017708,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004762118294955426,
      "loss": 5.011,
      "mean_token_accuracy": 0.19382331967353822,
      "num_tokens": 34094758.0,
      "step": 18535
    },
    {
      "entropy": 5.259174251556397,
      "epoch": 1.5486134313397928,
      "grad_norm": 1.25,
      "learning_rate": 0.0004761985051260291,
      "loss": 5.0568,
      "mean_token_accuracy": 0.19635215848684312,
      "num_tokens": 34103605.0,
      "step": 18540
    },
    {
      "entropy": 5.174782085418701,
      "epoch": 1.5490310725025058,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047618517723429366,
      "loss": 4.895,
      "mean_token_accuracy": 0.20346198379993438,
      "num_tokens": 34112751.0,
      "step": 18545
    },
    {
      "entropy": 5.193311643600464,
      "epoch": 1.5494487136652189,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047617184582056946,
      "loss": 4.9842,
      "mean_token_accuracy": 0.20435038357973098,
      "num_tokens": 34121814.0,
      "step": 18550
    },
    {
      "entropy": 5.17559289932251,
      "epoch": 1.549866354827932,
      "grad_norm": 1.421875,
      "learning_rate": 0.00047615851088508995,
      "loss": 4.9175,
      "mean_token_accuracy": 0.20891180336475373,
      "num_tokens": 34130297.0,
      "step": 18555
    },
    {
      "entropy": 5.154672431945801,
      "epoch": 1.550283995990645,
      "grad_norm": 1.484375,
      "learning_rate": 0.00047614517242808866,
      "loss": 4.9068,
      "mean_token_accuracy": 0.20748199820518493,
      "num_tokens": 34138525.0,
      "step": 18560
    },
    {
      "entropy": 5.097889947891235,
      "epoch": 1.5507016371533577,
      "grad_norm": 1.53125,
      "learning_rate": 0.00047613183044979885,
      "loss": 4.8476,
      "mean_token_accuracy": 0.21640335023403168,
      "num_tokens": 34147029.0,
      "step": 18565
    },
    {
      "entropy": 5.200020694732666,
      "epoch": 1.551119278316071,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047611848495045434,
      "loss": 4.9839,
      "mean_token_accuracy": 0.20047767162323,
      "num_tokens": 34155617.0,
      "step": 18570
    },
    {
      "entropy": 5.269506645202637,
      "epoch": 1.5515369194787838,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047610513593028867,
      "loss": 5.003,
      "mean_token_accuracy": 0.19974423348903655,
      "num_tokens": 34164693.0,
      "step": 18575
    },
    {
      "entropy": 5.193707847595215,
      "epoch": 1.5519545606414968,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004760917833895355,
      "loss": 4.9555,
      "mean_token_accuracy": 0.2019449070096016,
      "num_tokens": 34173394.0,
      "step": 18580
    },
    {
      "entropy": 5.180975675582886,
      "epoch": 1.5523722018042099,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004760784273284286,
      "loss": 4.9305,
      "mean_token_accuracy": 0.21133212447166444,
      "num_tokens": 34182522.0,
      "step": 18585
    },
    {
      "entropy": 5.186946964263916,
      "epoch": 1.5527898429669227,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004760650677472018,
      "loss": 4.9445,
      "mean_token_accuracy": 0.2032308906316757,
      "num_tokens": 34192456.0,
      "step": 18590
    },
    {
      "entropy": 5.3095499038696286,
      "epoch": 1.553207484129636,
      "grad_norm": 1.515625,
      "learning_rate": 0.000476051704646089,
      "loss": 5.1042,
      "mean_token_accuracy": 0.194295971095562,
      "num_tokens": 34202360.0,
      "step": 18595
    },
    {
      "entropy": 5.245389032363891,
      "epoch": 1.5536251252923488,
      "grad_norm": 1.421875,
      "learning_rate": 0.000476038338025324,
      "loss": 4.9718,
      "mean_token_accuracy": 0.20338471680879594,
      "num_tokens": 34211892.0,
      "step": 18600
    },
    {
      "entropy": 5.237710046768188,
      "epoch": 1.5540427664550618,
      "grad_norm": 1.453125,
      "learning_rate": 0.00047602496788514103,
      "loss": 4.9859,
      "mean_token_accuracy": 0.203501196205616,
      "num_tokens": 34221776.0,
      "step": 18605
    },
    {
      "entropy": 5.1633096694946286,
      "epoch": 1.5544604076177748,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004760115942257739,
      "loss": 4.9397,
      "mean_token_accuracy": 0.20960572510957717,
      "num_tokens": 34230564.0,
      "step": 18610
    },
    {
      "entropy": 5.073726844787598,
      "epoch": 1.5548780487804879,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00047599821704745693,
      "loss": 4.9407,
      "mean_token_accuracy": 0.20572277158498764,
      "num_tokens": 34239871.0,
      "step": 18615
    },
    {
      "entropy": 5.192074966430664,
      "epoch": 1.5552956899432009,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004759848363504243,
      "loss": 4.9737,
      "mean_token_accuracy": 0.2092226430773735,
      "num_tokens": 34248881.0,
      "step": 18620
    },
    {
      "entropy": 5.213341760635376,
      "epoch": 1.5557133311059137,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004759714521349101,
      "loss": 4.9517,
      "mean_token_accuracy": 0.20838092714548112,
      "num_tokens": 34257596.0,
      "step": 18625
    },
    {
      "entropy": 5.1645208358764645,
      "epoch": 1.556130972268627,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004759580644011487,
      "loss": 4.9522,
      "mean_token_accuracy": 0.20490102469921112,
      "num_tokens": 34267030.0,
      "step": 18630
    },
    {
      "entropy": 5.221986293792725,
      "epoch": 1.5565486134313398,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004759446731493745,
      "loss": 4.989,
      "mean_token_accuracy": 0.20296455025672913,
      "num_tokens": 34276269.0,
      "step": 18635
    },
    {
      "entropy": 5.123061943054199,
      "epoch": 1.5569662545940528,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000475931278379822,
      "loss": 4.9237,
      "mean_token_accuracy": 0.20616275072097778,
      "num_tokens": 34285193.0,
      "step": 18640
    },
    {
      "entropy": 5.107967138290405,
      "epoch": 1.5573838957567658,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004759178800927255,
      "loss": 4.83,
      "mean_token_accuracy": 0.21516538560390472,
      "num_tokens": 34294216.0,
      "step": 18645
    },
    {
      "entropy": 5.262439250946045,
      "epoch": 1.5578015369194786,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004759044782883198,
      "loss": 5.0616,
      "mean_token_accuracy": 0.21008716076612471,
      "num_tokens": 34302546.0,
      "step": 18650
    },
    {
      "entropy": 5.177655124664307,
      "epoch": 1.558219178082192,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004758910729668392,
      "loss": 4.8621,
      "mean_token_accuracy": 0.21396249383687974,
      "num_tokens": 34312176.0,
      "step": 18655
    },
    {
      "entropy": 5.247171926498413,
      "epoch": 1.5586368192449047,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004758776641285187,
      "loss": 4.9834,
      "mean_token_accuracy": 0.20473207533359528,
      "num_tokens": 34321836.0,
      "step": 18660
    },
    {
      "entropy": 5.2046167850494385,
      "epoch": 1.5590544604076177,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004758642517735928,
      "loss": 4.9722,
      "mean_token_accuracy": 0.21328698247671127,
      "num_tokens": 34330624.0,
      "step": 18665
    },
    {
      "entropy": 5.153477954864502,
      "epoch": 1.5594721015703308,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047585083590229643,
      "loss": 4.9028,
      "mean_token_accuracy": 0.21002527475357055,
      "num_tokens": 34339239.0,
      "step": 18670
    },
    {
      "entropy": 5.236139583587646,
      "epoch": 1.5598897427330438,
      "grad_norm": 1.25,
      "learning_rate": 0.00047583741651486443,
      "loss": 5.0191,
      "mean_token_accuracy": 0.19978638589382172,
      "num_tokens": 34348823.0,
      "step": 18675
    },
    {
      "entropy": 5.092165756225586,
      "epoch": 1.5603073838957568,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004758239936115317,
      "loss": 4.8481,
      "mean_token_accuracy": 0.2098802626132965,
      "num_tokens": 34357844.0,
      "step": 18680
    },
    {
      "entropy": 5.18106894493103,
      "epoch": 1.5607250250584697,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047581056719253324,
      "loss": 4.9262,
      "mean_token_accuracy": 0.20668885260820388,
      "num_tokens": 34367055.0,
      "step": 18685
    },
    {
      "entropy": 5.32228569984436,
      "epoch": 1.561142666221183,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004757971372581041,
      "loss": 5.114,
      "mean_token_accuracy": 0.1975152537226677,
      "num_tokens": 34377329.0,
      "step": 18690
    },
    {
      "entropy": 5.220578718185425,
      "epoch": 1.5615603073838957,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047578370380847935,
      "loss": 5.0906,
      "mean_token_accuracy": 0.19448876082897187,
      "num_tokens": 34386508.0,
      "step": 18695
    },
    {
      "entropy": 5.206160545349121,
      "epoch": 1.5619779485466088,
      "grad_norm": 1.71875,
      "learning_rate": 0.00047577026684389413,
      "loss": 4.9685,
      "mean_token_accuracy": 0.19835634231567384,
      "num_tokens": 34396075.0,
      "step": 18700
    },
    {
      "entropy": 5.280202102661133,
      "epoch": 1.5623955897093218,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004757568263645838,
      "loss": 5.0148,
      "mean_token_accuracy": 0.2054536297917366,
      "num_tokens": 34405409.0,
      "step": 18705
    },
    {
      "entropy": 5.241973924636841,
      "epoch": 1.5628132308720346,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004757433823707835,
      "loss": 4.9807,
      "mean_token_accuracy": 0.20118271559476852,
      "num_tokens": 34415559.0,
      "step": 18710
    },
    {
      "entropy": 5.260473299026489,
      "epoch": 1.5632308720347479,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004757299348627287,
      "loss": 4.9883,
      "mean_token_accuracy": 0.20950516760349275,
      "num_tokens": 34424137.0,
      "step": 18715
    },
    {
      "entropy": 5.153454732894898,
      "epoch": 1.5636485131974607,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047571648384065477,
      "loss": 4.979,
      "mean_token_accuracy": 0.2053193897008896,
      "num_tokens": 34433498.0,
      "step": 18720
    },
    {
      "entropy": 5.203670263290405,
      "epoch": 1.5640661543601737,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004757030293047972,
      "loss": 4.9354,
      "mean_token_accuracy": 0.21138562709093095,
      "num_tokens": 34443218.0,
      "step": 18725
    },
    {
      "entropy": 5.169069051742554,
      "epoch": 1.5644837955228867,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004756895712553915,
      "loss": 4.9475,
      "mean_token_accuracy": 0.2037816107273102,
      "num_tokens": 34452981.0,
      "step": 18730
    },
    {
      "entropy": 5.196348094940186,
      "epoch": 1.5649014366855998,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047567610969267326,
      "loss": 4.984,
      "mean_token_accuracy": 0.1968775659799576,
      "num_tokens": 34462763.0,
      "step": 18735
    },
    {
      "entropy": 5.150794267654419,
      "epoch": 1.5653190778483128,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004756626446168782,
      "loss": 4.8743,
      "mean_token_accuracy": 0.21604453176259994,
      "num_tokens": 34471841.0,
      "step": 18740
    },
    {
      "entropy": 5.310590076446533,
      "epoch": 1.5657367190110256,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047564917602824203,
      "loss": 5.1019,
      "mean_token_accuracy": 0.1986771285533905,
      "num_tokens": 34482280.0,
      "step": 18745
    },
    {
      "entropy": 5.383828353881836,
      "epoch": 1.5661543601737389,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004756357039270004,
      "loss": 5.1414,
      "mean_token_accuracy": 0.19505720138549804,
      "num_tokens": 34492774.0,
      "step": 18750
    },
    {
      "entropy": 5.185140609741211,
      "epoch": 1.5665720013364517,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047562222831338933,
      "loss": 4.898,
      "mean_token_accuracy": 0.2078563615679741,
      "num_tokens": 34501626.0,
      "step": 18755
    },
    {
      "entropy": 5.143675708770752,
      "epoch": 1.5669896424991647,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047560874918764473,
      "loss": 4.9823,
      "mean_token_accuracy": 0.2007572129368782,
      "num_tokens": 34510264.0,
      "step": 18760
    },
    {
      "entropy": 5.294533014297485,
      "epoch": 1.5674072836618778,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047559526655000244,
      "loss": 5.0445,
      "mean_token_accuracy": 0.19215273559093476,
      "num_tokens": 34520540.0,
      "step": 18765
    },
    {
      "entropy": 5.248611164093018,
      "epoch": 1.5678249248245906,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004755817804006986,
      "loss": 4.9941,
      "mean_token_accuracy": 0.2032113030552864,
      "num_tokens": 34529669.0,
      "step": 18770
    },
    {
      "entropy": 5.182350492477417,
      "epoch": 1.5682425659873038,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004755682907399692,
      "loss": 4.9897,
      "mean_token_accuracy": 0.20342108607292175,
      "num_tokens": 34538802.0,
      "step": 18775
    },
    {
      "entropy": 5.253236532211304,
      "epoch": 1.5686602071500166,
      "grad_norm": 1.421875,
      "learning_rate": 0.00047555479756805055,
      "loss": 5.0367,
      "mean_token_accuracy": 0.20248427540063857,
      "num_tokens": 34547583.0,
      "step": 18780
    },
    {
      "entropy": 5.193507099151612,
      "epoch": 1.5690778483127297,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004755413008851786,
      "loss": 4.9307,
      "mean_token_accuracy": 0.2106501817703247,
      "num_tokens": 34556045.0,
      "step": 18785
    },
    {
      "entropy": 5.155056476593018,
      "epoch": 1.5694954894754427,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004755278006915899,
      "loss": 4.813,
      "mean_token_accuracy": 0.21558625549077987,
      "num_tokens": 34564910.0,
      "step": 18790
    },
    {
      "entropy": 5.1626549243927,
      "epoch": 1.5699131306381557,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00047551429698752064,
      "loss": 5.0233,
      "mean_token_accuracy": 0.1997707724571228,
      "num_tokens": 34573274.0,
      "step": 18795
    },
    {
      "entropy": 5.163156414031983,
      "epoch": 1.5703307718008688,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00047550078977320724,
      "loss": 4.9152,
      "mean_token_accuracy": 0.20510024428367615,
      "num_tokens": 34583015.0,
      "step": 18800
    },
    {
      "entropy": 5.3007080078125,
      "epoch": 1.5707484129635816,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004754872790488862,
      "loss": 5.0575,
      "mean_token_accuracy": 0.19582573771476747,
      "num_tokens": 34592311.0,
      "step": 18805
    },
    {
      "entropy": 5.2488336086273195,
      "epoch": 1.5711660541262948,
      "grad_norm": 1.8359375,
      "learning_rate": 0.000475473764814794,
      "loss": 4.9281,
      "mean_token_accuracy": 0.21208879649639129,
      "num_tokens": 34600806.0,
      "step": 18810
    },
    {
      "entropy": 5.182951354980469,
      "epoch": 1.5715836952890077,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047546024707116727,
      "loss": 4.9008,
      "mean_token_accuracy": 0.20462341755628585,
      "num_tokens": 34609830.0,
      "step": 18815
    },
    {
      "entropy": 5.20071005821228,
      "epoch": 1.5720013364517207,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00047544672581824256,
      "loss": 4.9332,
      "mean_token_accuracy": 0.20291958302259444,
      "num_tokens": 34618870.0,
      "step": 18820
    },
    {
      "entropy": 5.100830745697022,
      "epoch": 1.5724189776144337,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004754332010562567,
      "loss": 4.8661,
      "mean_token_accuracy": 0.20691495686769484,
      "num_tokens": 34628349.0,
      "step": 18825
    },
    {
      "entropy": 5.206101894378662,
      "epoch": 1.5728366187771465,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004754196727854463,
      "loss": 5.0095,
      "mean_token_accuracy": 0.19823441952466964,
      "num_tokens": 34638492.0,
      "step": 18830
    },
    {
      "entropy": 5.278979110717773,
      "epoch": 1.5732542599398598,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047540614100604836,
      "loss": 5.0793,
      "mean_token_accuracy": 0.2039191395044327,
      "num_tokens": 34649084.0,
      "step": 18835
    },
    {
      "entropy": 5.20273642539978,
      "epoch": 1.5736719011025726,
      "grad_norm": 1.375,
      "learning_rate": 0.0004753926057182997,
      "loss": 4.9654,
      "mean_token_accuracy": 0.20383328646421434,
      "num_tokens": 34658392.0,
      "step": 18840
    },
    {
      "entropy": 5.1891885757446286,
      "epoch": 1.5740895422652856,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047537906692243716,
      "loss": 5.0075,
      "mean_token_accuracy": 0.19744484275579452,
      "num_tokens": 34667727.0,
      "step": 18845
    },
    {
      "entropy": 5.210335826873779,
      "epoch": 1.5745071834279987,
      "grad_norm": 1.3125,
      "learning_rate": 0.000475365524618698,
      "loss": 4.9601,
      "mean_token_accuracy": 0.20081791877746583,
      "num_tokens": 34676260.0,
      "step": 18850
    },
    {
      "entropy": 5.298014354705811,
      "epoch": 1.5749248245907117,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004753519788073191,
      "loss": 5.1052,
      "mean_token_accuracy": 0.1918730542063713,
      "num_tokens": 34685145.0,
      "step": 18855
    },
    {
      "entropy": 5.169727802276611,
      "epoch": 1.5753424657534247,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004753384294885376,
      "loss": 4.8606,
      "mean_token_accuracy": 0.22122335135936738,
      "num_tokens": 34694035.0,
      "step": 18860
    },
    {
      "entropy": 5.188487148284912,
      "epoch": 1.5757601069161375,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004753248766625908,
      "loss": 4.9576,
      "mean_token_accuracy": 0.20321147590875627,
      "num_tokens": 34702634.0,
      "step": 18865
    },
    {
      "entropy": 5.167422962188721,
      "epoch": 1.5761777480788508,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004753113203297159,
      "loss": 4.9483,
      "mean_token_accuracy": 0.19966938197612763,
      "num_tokens": 34711925.0,
      "step": 18870
    },
    {
      "entropy": 5.101533365249634,
      "epoch": 1.5765953892415636,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004752977604901502,
      "loss": 4.8838,
      "mean_token_accuracy": 0.21762792021036148,
      "num_tokens": 34720131.0,
      "step": 18875
    },
    {
      "entropy": 5.1734216690063475,
      "epoch": 1.5770130304042767,
      "grad_norm": 1.4375,
      "learning_rate": 0.00047528419714413107,
      "loss": 5.0083,
      "mean_token_accuracy": 0.1988901287317276,
      "num_tokens": 34728869.0,
      "step": 18880
    },
    {
      "entropy": 5.163319826126099,
      "epoch": 1.5774306715669897,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047527063029189606,
      "loss": 4.85,
      "mean_token_accuracy": 0.2179366260766983,
      "num_tokens": 34737343.0,
      "step": 18885
    },
    {
      "entropy": 5.227366733551025,
      "epoch": 1.5778483127297025,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004752570599336825,
      "loss": 5.0509,
      "mean_token_accuracy": 0.19371123015880584,
      "num_tokens": 34746752.0,
      "step": 18890
    },
    {
      "entropy": 5.236007881164551,
      "epoch": 1.5782659538924158,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047524348606972817,
      "loss": 4.9862,
      "mean_token_accuracy": 0.20699062198400497,
      "num_tokens": 34755672.0,
      "step": 18895
    },
    {
      "entropy": 5.203511524200439,
      "epoch": 1.5786835950551286,
      "grad_norm": 1.421875,
      "learning_rate": 0.00047522990870027055,
      "loss": 4.9986,
      "mean_token_accuracy": 0.2004362165927887,
      "num_tokens": 34764558.0,
      "step": 18900
    },
    {
      "entropy": 5.206935405731201,
      "epoch": 1.5791012362178416,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004752163278255473,
      "loss": 4.928,
      "mean_token_accuracy": 0.19966644048690796,
      "num_tokens": 34773663.0,
      "step": 18905
    },
    {
      "entropy": 5.0801615715026855,
      "epoch": 1.5795188773805546,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004752027434457963,
      "loss": 4.8761,
      "mean_token_accuracy": 0.2106747716665268,
      "num_tokens": 34782297.0,
      "step": 18910
    },
    {
      "entropy": 5.271767568588257,
      "epoch": 1.5799365185432677,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004751891555612552,
      "loss": 5.1339,
      "mean_token_accuracy": 0.19416941702365875,
      "num_tokens": 34791734.0,
      "step": 18915
    },
    {
      "entropy": 5.221511316299439,
      "epoch": 1.5803541597059807,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004751755641721621,
      "loss": 4.9589,
      "mean_token_accuracy": 0.20598703026771545,
      "num_tokens": 34801187.0,
      "step": 18920
    },
    {
      "entropy": 5.186386489868164,
      "epoch": 1.5807718008686935,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047516196927875474,
      "loss": 4.9631,
      "mean_token_accuracy": 0.20633205771446228,
      "num_tokens": 34810084.0,
      "step": 18925
    },
    {
      "entropy": 5.3238136768341064,
      "epoch": 1.5811894420314068,
      "grad_norm": 1.8828125,
      "learning_rate": 0.00047514837088127117,
      "loss": 5.0625,
      "mean_token_accuracy": 0.20030854642391205,
      "num_tokens": 34819160.0,
      "step": 18930
    },
    {
      "entropy": 5.17928557395935,
      "epoch": 1.5816070831941196,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004751347689799494,
      "loss": 4.8605,
      "mean_token_accuracy": 0.22027141600847244,
      "num_tokens": 34828272.0,
      "step": 18935
    },
    {
      "entropy": 5.073698329925537,
      "epoch": 1.5820247243568326,
      "grad_norm": 1.515625,
      "learning_rate": 0.00047512116357502766,
      "loss": 4.9099,
      "mean_token_accuracy": 0.20748468935489656,
      "num_tokens": 34838030.0,
      "step": 18940
    },
    {
      "entropy": 5.163582611083984,
      "epoch": 1.5824423655195456,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047510755466674407,
      "loss": 4.9879,
      "mean_token_accuracy": 0.20462685972452163,
      "num_tokens": 34846489.0,
      "step": 18945
    },
    {
      "entropy": 5.2715559005737305,
      "epoch": 1.5828600066822585,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00047509394225533685,
      "loss": 4.9484,
      "mean_token_accuracy": 0.20432062745094298,
      "num_tokens": 34855926.0,
      "step": 18950
    },
    {
      "entropy": 5.23129186630249,
      "epoch": 1.5832776478449717,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004750803263410443,
      "loss": 4.9355,
      "mean_token_accuracy": 0.20649914294481278,
      "num_tokens": 34865158.0,
      "step": 18955
    },
    {
      "entropy": 5.156368207931519,
      "epoch": 1.5836952890076845,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004750667069241048,
      "loss": 5.0114,
      "mean_token_accuracy": 0.19714290648698807,
      "num_tokens": 34875429.0,
      "step": 18960
    },
    {
      "entropy": 5.235931444168091,
      "epoch": 1.5841129301703976,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004750530840047568,
      "loss": 5.0072,
      "mean_token_accuracy": 0.20724041163921356,
      "num_tokens": 34885663.0,
      "step": 18965
    },
    {
      "entropy": 5.21707820892334,
      "epoch": 1.5845305713331106,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047503945758323873,
      "loss": 4.9243,
      "mean_token_accuracy": 0.2070024862885475,
      "num_tokens": 34894976.0,
      "step": 18970
    },
    {
      "entropy": 5.225789213180542,
      "epoch": 1.5849482124958236,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004750258276597892,
      "loss": 4.9375,
      "mean_token_accuracy": 0.20253977477550505,
      "num_tokens": 34905630.0,
      "step": 18975
    },
    {
      "entropy": 5.157743978500366,
      "epoch": 1.5853658536585367,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047501219423464673,
      "loss": 4.8878,
      "mean_token_accuracy": 0.2093823090195656,
      "num_tokens": 34914906.0,
      "step": 18980
    },
    {
      "entropy": 5.125348329544067,
      "epoch": 1.5857834948212495,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004749985573080501,
      "loss": 4.853,
      "mean_token_accuracy": 0.20970769822597504,
      "num_tokens": 34923274.0,
      "step": 18985
    },
    {
      "entropy": 5.164556550979614,
      "epoch": 1.5862011359839627,
      "grad_norm": 1.609375,
      "learning_rate": 0.000474984916880238,
      "loss": 4.9262,
      "mean_token_accuracy": 0.2095055416226387,
      "num_tokens": 34932085.0,
      "step": 18990
    },
    {
      "entropy": 5.180087327957153,
      "epoch": 1.5866187771466755,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004749712729514491,
      "loss": 5.0075,
      "mean_token_accuracy": 0.19972948580980301,
      "num_tokens": 34940871.0,
      "step": 18995
    },
    {
      "entropy": 5.241435813903808,
      "epoch": 1.5870364183093886,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004749576255219225,
      "loss": 5.072,
      "mean_token_accuracy": 0.19720330834388733,
      "num_tokens": 34950703.0,
      "step": 19000
    },
    {
      "entropy": 5.285844182968139,
      "epoch": 1.5874540594721016,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004749439745918969,
      "loss": 5.0067,
      "mean_token_accuracy": 0.20023841559886932,
      "num_tokens": 34959988.0,
      "step": 19005
    },
    {
      "entropy": 5.233851671218872,
      "epoch": 1.5878717006348144,
      "grad_norm": 1.375,
      "learning_rate": 0.00047493032016161134,
      "loss": 5.0205,
      "mean_token_accuracy": 0.20130687355995178,
      "num_tokens": 34968608.0,
      "step": 19010
    },
    {
      "entropy": 5.28958158493042,
      "epoch": 1.5882893417975277,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004749166622313049,
      "loss": 5.0187,
      "mean_token_accuracy": 0.198570616543293,
      "num_tokens": 34977200.0,
      "step": 19015
    },
    {
      "entropy": 5.2539667129516605,
      "epoch": 1.5887069829602405,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047490300080121667,
      "loss": 4.9287,
      "mean_token_accuracy": 0.21528749018907548,
      "num_tokens": 34986403.0,
      "step": 19020
    },
    {
      "entropy": 5.176285457611084,
      "epoch": 1.5891246241229535,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047488933587158585,
      "loss": 5.041,
      "mean_token_accuracy": 0.19975536465644836,
      "num_tokens": 34995672.0,
      "step": 19025
    },
    {
      "entropy": 5.167521667480469,
      "epoch": 1.5895422652856666,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004748756674426514,
      "loss": 5.0632,
      "mean_token_accuracy": 0.20073603093624115,
      "num_tokens": 35004636.0,
      "step": 19030
    },
    {
      "entropy": 5.18248348236084,
      "epoch": 1.5899599064483796,
      "grad_norm": 1.203125,
      "learning_rate": 0.000474861995514653,
      "loss": 4.8719,
      "mean_token_accuracy": 0.209346804022789,
      "num_tokens": 35014023.0,
      "step": 19035
    },
    {
      "entropy": 5.313436841964721,
      "epoch": 1.5903775476110926,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004748483200878297,
      "loss": 5.0837,
      "mean_token_accuracy": 0.20305539667606354,
      "num_tokens": 35023762.0,
      "step": 19040
    },
    {
      "entropy": 5.227099180221558,
      "epoch": 1.5907951887738054,
      "grad_norm": 1.390625,
      "learning_rate": 0.000474834641162421,
      "loss": 4.9187,
      "mean_token_accuracy": 0.20950115621089935,
      "num_tokens": 35032485.0,
      "step": 19045
    },
    {
      "entropy": 5.207560062408447,
      "epoch": 1.5912128299365187,
      "grad_norm": 1.375,
      "learning_rate": 0.00047482095873866635,
      "loss": 5.0195,
      "mean_token_accuracy": 0.19574302136898042,
      "num_tokens": 35041416.0,
      "step": 19050
    },
    {
      "entropy": 5.25741662979126,
      "epoch": 1.5916304710992315,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004748072728168053,
      "loss": 4.9458,
      "mean_token_accuracy": 0.20360504239797592,
      "num_tokens": 35051223.0,
      "step": 19055
    },
    {
      "entropy": 5.164383172988892,
      "epoch": 1.5920481122619445,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004747935833970775,
      "loss": 5.0023,
      "mean_token_accuracy": 0.2050727501511574,
      "num_tokens": 35060935.0,
      "step": 19060
    },
    {
      "entropy": 5.129144620895386,
      "epoch": 1.5924657534246576,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004747798904797225,
      "loss": 4.8867,
      "mean_token_accuracy": 0.21382359713315963,
      "num_tokens": 35070249.0,
      "step": 19065
    },
    {
      "entropy": 5.22361798286438,
      "epoch": 1.5928833945873704,
      "grad_norm": 1.546875,
      "learning_rate": 0.00047476619406497996,
      "loss": 4.9071,
      "mean_token_accuracy": 0.20372487604618073,
      "num_tokens": 35079040.0,
      "step": 19070
    },
    {
      "entropy": 5.096227550506592,
      "epoch": 1.5933010357500836,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004747524941530898,
      "loss": 4.8707,
      "mean_token_accuracy": 0.20860759019851685,
      "num_tokens": 35087451.0,
      "step": 19075
    },
    {
      "entropy": 5.241660213470459,
      "epoch": 1.5937186769127965,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004747387907442918,
      "loss": 4.9852,
      "mean_token_accuracy": 0.20921770185232164,
      "num_tokens": 35096144.0,
      "step": 19080
    },
    {
      "entropy": 5.357588529586792,
      "epoch": 1.5941363180755095,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00047472508383882586,
      "loss": 4.993,
      "mean_token_accuracy": 0.2077196791768074,
      "num_tokens": 35105547.0,
      "step": 19085
    },
    {
      "entropy": 5.216818904876709,
      "epoch": 1.5945539592382225,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004747113734369319,
      "loss": 4.9456,
      "mean_token_accuracy": 0.20444457828998566,
      "num_tokens": 35114164.0,
      "step": 19090
    },
    {
      "entropy": 5.134893465042114,
      "epoch": 1.5949716004009356,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004746976595388499,
      "loss": 4.9932,
      "mean_token_accuracy": 0.19953957796096802,
      "num_tokens": 35125184.0,
      "step": 19095
    },
    {
      "entropy": 5.155754327774048,
      "epoch": 1.5953892415636486,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047468394214482,
      "loss": 4.8733,
      "mean_token_accuracy": 0.21295168101787568,
      "num_tokens": 35134207.0,
      "step": 19100
    },
    {
      "entropy": 5.1831199645996096,
      "epoch": 1.5958068827263614,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004746702212550824,
      "loss": 4.9413,
      "mean_token_accuracy": 0.21327533721923828,
      "num_tokens": 35143651.0,
      "step": 19105
    },
    {
      "entropy": 5.14981050491333,
      "epoch": 1.5962245238890747,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047465649686987717,
      "loss": 4.99,
      "mean_token_accuracy": 0.20494986921548844,
      "num_tokens": 35152907.0,
      "step": 19110
    },
    {
      "entropy": 5.166287088394165,
      "epoch": 1.5966421650517875,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004746427689894447,
      "loss": 4.8861,
      "mean_token_accuracy": 0.21617903858423232,
      "num_tokens": 35161927.0,
      "step": 19115
    },
    {
      "entropy": 5.234162473678589,
      "epoch": 1.5970598062145005,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004746290376140252,
      "loss": 4.9862,
      "mean_token_accuracy": 0.20618948936462403,
      "num_tokens": 35171303.0,
      "step": 19120
    },
    {
      "entropy": 5.282760572433472,
      "epoch": 1.5974774473772135,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004746153027438592,
      "loss": 5.0646,
      "mean_token_accuracy": 0.1980816438794136,
      "num_tokens": 35181160.0,
      "step": 19125
    },
    {
      "entropy": 5.195584011077881,
      "epoch": 1.5978950885399263,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004746015643791869,
      "loss": 5.012,
      "mean_token_accuracy": 0.19717658758163453,
      "num_tokens": 35190335.0,
      "step": 19130
    },
    {
      "entropy": 5.262140321731567,
      "epoch": 1.5983127297026396,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000474587822520249,
      "loss": 5.0215,
      "mean_token_accuracy": 0.20452949404716492,
      "num_tokens": 35200036.0,
      "step": 19135
    },
    {
      "entropy": 5.22915096282959,
      "epoch": 1.5987303708653524,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004745740771672861,
      "loss": 4.9273,
      "mean_token_accuracy": 0.21057048439979553,
      "num_tokens": 35208203.0,
      "step": 19140
    },
    {
      "entropy": 5.132527685165405,
      "epoch": 1.5991480120280654,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004745603283205387,
      "loss": 4.8793,
      "mean_token_accuracy": 0.213470695912838,
      "num_tokens": 35216963.0,
      "step": 19145
    },
    {
      "entropy": 5.221748208999633,
      "epoch": 1.5995656531907785,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047454657598024755,
      "loss": 4.9598,
      "mean_token_accuracy": 0.2028125062584877,
      "num_tokens": 35226170.0,
      "step": 19150
    },
    {
      "entropy": 5.098206472396851,
      "epoch": 1.5999832943534915,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047453282014665334,
      "loss": 4.8458,
      "mean_token_accuracy": 0.21259595304727555,
      "num_tokens": 35234383.0,
      "step": 19155
    },
    {
      "entropy": 5.131440544128418,
      "epoch": 1.6004009355162045,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00047451906081999695,
      "loss": 4.842,
      "mean_token_accuracy": 0.21737159341573714,
      "num_tokens": 35242771.0,
      "step": 19160
    },
    {
      "entropy": 5.1809317111969,
      "epoch": 1.6008185766789174,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004745052980005193,
      "loss": 4.9128,
      "mean_token_accuracy": 0.21655474305152894,
      "num_tokens": 35251333.0,
      "step": 19165
    },
    {
      "entropy": 5.205656814575195,
      "epoch": 1.6012362178416306,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004744915316884612,
      "loss": 4.9898,
      "mean_token_accuracy": 0.19419017285108567,
      "num_tokens": 35260423.0,
      "step": 19170
    },
    {
      "entropy": 5.238546085357666,
      "epoch": 1.6016538590043434,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004744777618840637,
      "loss": 4.9254,
      "mean_token_accuracy": 0.2144421935081482,
      "num_tokens": 35268838.0,
      "step": 19175
    },
    {
      "entropy": 5.2639039039611815,
      "epoch": 1.6020715001670565,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000474463988587568,
      "loss": 5.0481,
      "mean_token_accuracy": 0.19906053692102432,
      "num_tokens": 35279395.0,
      "step": 19180
    },
    {
      "entropy": 5.209416913986206,
      "epoch": 1.6024891413297695,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047445021179921495,
      "loss": 4.9744,
      "mean_token_accuracy": 0.2023167297244072,
      "num_tokens": 35289637.0,
      "step": 19185
    },
    {
      "entropy": 5.153041124343872,
      "epoch": 1.6029067824924823,
      "grad_norm": 1.5,
      "learning_rate": 0.0004744364315192459,
      "loss": 4.9632,
      "mean_token_accuracy": 0.2047106072306633,
      "num_tokens": 35299813.0,
      "step": 19190
    },
    {
      "entropy": 5.217156267166137,
      "epoch": 1.6033244236551956,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047442264774790203,
      "loss": 4.9591,
      "mean_token_accuracy": 0.20702714771032332,
      "num_tokens": 35308162.0,
      "step": 19195
    },
    {
      "entropy": 5.1592193126678465,
      "epoch": 1.6037420648179084,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004744088604854248,
      "loss": 4.9538,
      "mean_token_accuracy": 0.20360019207000732,
      "num_tokens": 35317232.0,
      "step": 19200
    },
    {
      "entropy": 5.226922178268433,
      "epoch": 1.6041597059806214,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047439506973205527,
      "loss": 4.9141,
      "mean_token_accuracy": 0.2078560099005699,
      "num_tokens": 35326011.0,
      "step": 19205
    },
    {
      "entropy": 5.0783045291900635,
      "epoch": 1.6045773471433344,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004743812754880351,
      "loss": 4.8347,
      "mean_token_accuracy": 0.21435738503932952,
      "num_tokens": 35333935.0,
      "step": 19210
    },
    {
      "entropy": 5.197670030593872,
      "epoch": 1.6049949883060475,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004743674777536057,
      "loss": 4.9648,
      "mean_token_accuracy": 0.20162280648946762,
      "num_tokens": 35343512.0,
      "step": 19215
    },
    {
      "entropy": 5.201079082489014,
      "epoch": 1.6054126294687605,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004743536765290087,
      "loss": 4.9649,
      "mean_token_accuracy": 0.2033594936132431,
      "num_tokens": 35353276.0,
      "step": 19220
    },
    {
      "entropy": 5.181528568267822,
      "epoch": 1.6058302706314733,
      "grad_norm": 1.546875,
      "learning_rate": 0.00047433987181448546,
      "loss": 4.9597,
      "mean_token_accuracy": 0.2116337925195694,
      "num_tokens": 35362101.0,
      "step": 19225
    },
    {
      "entropy": 5.130848360061646,
      "epoch": 1.6062479117941866,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004743260636102779,
      "loss": 4.9339,
      "mean_token_accuracy": 0.20939384996891022,
      "num_tokens": 35370998.0,
      "step": 19230
    },
    {
      "entropy": 5.319819831848145,
      "epoch": 1.6066655529568994,
      "grad_norm": 1.375,
      "learning_rate": 0.00047431225191662776,
      "loss": 5.0496,
      "mean_token_accuracy": 0.19944816529750825,
      "num_tokens": 35380347.0,
      "step": 19235
    },
    {
      "entropy": 5.2511204242706295,
      "epoch": 1.6070831941196124,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047429843673377664,
      "loss": 4.9532,
      "mean_token_accuracy": 0.2085131287574768,
      "num_tokens": 35388586.0,
      "step": 19240
    },
    {
      "entropy": 5.205608701705932,
      "epoch": 1.6075008352823255,
      "grad_norm": 1.921875,
      "learning_rate": 0.00047428461806196646,
      "loss": 5.0209,
      "mean_token_accuracy": 0.20404692143201827,
      "num_tokens": 35397649.0,
      "step": 19245
    },
    {
      "entropy": 5.182544803619384,
      "epoch": 1.6079184764450383,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004742707959014393,
      "loss": 4.9535,
      "mean_token_accuracy": 0.21204194575548171,
      "num_tokens": 35405790.0,
      "step": 19250
    },
    {
      "entropy": 5.1776103496551515,
      "epoch": 1.6083361176077515,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047425697025243693,
      "loss": 4.9771,
      "mean_token_accuracy": 0.20098508149385452,
      "num_tokens": 35414389.0,
      "step": 19255
    },
    {
      "entropy": 5.2164692878723145,
      "epoch": 1.6087537587704643,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047424314111520134,
      "loss": 4.976,
      "mean_token_accuracy": 0.20520440638065338,
      "num_tokens": 35424909.0,
      "step": 19260
    },
    {
      "entropy": 5.256686449050903,
      "epoch": 1.6091713999331774,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004742293084899749,
      "loss": 4.9439,
      "mean_token_accuracy": 0.2077789232134819,
      "num_tokens": 35434054.0,
      "step": 19265
    },
    {
      "entropy": 5.165187788009644,
      "epoch": 1.6095890410958904,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004742154723769995,
      "loss": 4.96,
      "mean_token_accuracy": 0.20782779306173324,
      "num_tokens": 35443722.0,
      "step": 19270
    },
    {
      "entropy": 5.248608827590942,
      "epoch": 1.6100066822586034,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004742016327765175,
      "loss": 4.9881,
      "mean_token_accuracy": 0.19932927340269088,
      "num_tokens": 35453067.0,
      "step": 19275
    },
    {
      "entropy": 5.127292823791504,
      "epoch": 1.6104243234213165,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004741877896887711,
      "loss": 4.8617,
      "mean_token_accuracy": 0.20645489543676376,
      "num_tokens": 35462425.0,
      "step": 19280
    },
    {
      "entropy": 5.166293621063232,
      "epoch": 1.6108419645840293,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004741739431140027,
      "loss": 4.9079,
      "mean_token_accuracy": 0.2059673175215721,
      "num_tokens": 35472037.0,
      "step": 19285
    },
    {
      "entropy": 5.237065029144287,
      "epoch": 1.6112596057467425,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047416009305245474,
      "loss": 4.9795,
      "mean_token_accuracy": 0.20431418418884278,
      "num_tokens": 35481411.0,
      "step": 19290
    },
    {
      "entropy": 5.106869888305664,
      "epoch": 1.6116772469094554,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004741462395043695,
      "loss": 4.8819,
      "mean_token_accuracy": 0.21397385150194168,
      "num_tokens": 35490019.0,
      "step": 19295
    },
    {
      "entropy": 5.128438949584961,
      "epoch": 1.6120948880721684,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00047413238246998975,
      "loss": 4.9741,
      "mean_token_accuracy": 0.21089595407247544,
      "num_tokens": 35499229.0,
      "step": 19300
    },
    {
      "entropy": 5.200277948379517,
      "epoch": 1.6125125292348814,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004741185219495579,
      "loss": 4.8858,
      "mean_token_accuracy": 0.20747314095497132,
      "num_tokens": 35508473.0,
      "step": 19305
    },
    {
      "entropy": 5.188142490386963,
      "epoch": 1.6129301703975942,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004741046579433166,
      "loss": 4.9573,
      "mean_token_accuracy": 0.20661916881799697,
      "num_tokens": 35517887.0,
      "step": 19310
    },
    {
      "entropy": 5.192843866348267,
      "epoch": 1.6133478115603075,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047409079045150865,
      "loss": 4.945,
      "mean_token_accuracy": 0.2061976358294487,
      "num_tokens": 35527470.0,
      "step": 19315
    },
    {
      "entropy": 5.1686248779296875,
      "epoch": 1.6137654527230203,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047407691947437677,
      "loss": 5.01,
      "mean_token_accuracy": 0.20499294698238374,
      "num_tokens": 35537670.0,
      "step": 19320
    },
    {
      "entropy": 5.140740585327149,
      "epoch": 1.6141830938857333,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047406304501216375,
      "loss": 4.9406,
      "mean_token_accuracy": 0.2043554589152336,
      "num_tokens": 35546903.0,
      "step": 19325
    },
    {
      "entropy": 5.216354465484619,
      "epoch": 1.6146007350484464,
      "grad_norm": 1.5625,
      "learning_rate": 0.00047404916706511255,
      "loss": 4.9272,
      "mean_token_accuracy": 0.21052003502845765,
      "num_tokens": 35556198.0,
      "step": 19330
    },
    {
      "entropy": 5.295907878875733,
      "epoch": 1.6150183762111594,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047403528563346603,
      "loss": 5.078,
      "mean_token_accuracy": 0.19259276539087294,
      "num_tokens": 35566381.0,
      "step": 19335
    },
    {
      "entropy": 5.1755311489105225,
      "epoch": 1.6154360173738724,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047402140071746743,
      "loss": 4.9257,
      "mean_token_accuracy": 0.21266453862190246,
      "num_tokens": 35575692.0,
      "step": 19340
    },
    {
      "entropy": 5.175512599945068,
      "epoch": 1.6158536585365852,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047400751231735946,
      "loss": 4.9239,
      "mean_token_accuracy": 0.20384102910757065,
      "num_tokens": 35584783.0,
      "step": 19345
    },
    {
      "entropy": 5.216027784347534,
      "epoch": 1.6162712996992985,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047399362043338547,
      "loss": 4.9955,
      "mean_token_accuracy": 0.20020997524261475,
      "num_tokens": 35595410.0,
      "step": 19350
    },
    {
      "entropy": 5.1843479633331295,
      "epoch": 1.6166889408620113,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004739797250657887,
      "loss": 4.8873,
      "mean_token_accuracy": 0.20507861524820328,
      "num_tokens": 35604950.0,
      "step": 19355
    },
    {
      "entropy": 5.231988573074341,
      "epoch": 1.6171065820247243,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047396582621481224,
      "loss": 4.9709,
      "mean_token_accuracy": 0.1977037027478218,
      "num_tokens": 35614227.0,
      "step": 19360
    },
    {
      "entropy": 5.151505804061889,
      "epoch": 1.6175242231874374,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047395192388069955,
      "loss": 4.9713,
      "mean_token_accuracy": 0.2016235738992691,
      "num_tokens": 35623349.0,
      "step": 19365
    },
    {
      "entropy": 5.139121198654175,
      "epoch": 1.6179418643501502,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047393801806369396,
      "loss": 4.7908,
      "mean_token_accuracy": 0.22234184443950653,
      "num_tokens": 35632169.0,
      "step": 19370
    },
    {
      "entropy": 5.208208703994751,
      "epoch": 1.6183595055128634,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00047392410876403887,
      "loss": 5.0454,
      "mean_token_accuracy": 0.19969044923782348,
      "num_tokens": 35642121.0,
      "step": 19375
    },
    {
      "entropy": 5.305715656280517,
      "epoch": 1.6187771466755763,
      "grad_norm": 1.4375,
      "learning_rate": 0.00047391019598197785,
      "loss": 4.9887,
      "mean_token_accuracy": 0.20363927334547044,
      "num_tokens": 35651771.0,
      "step": 19380
    },
    {
      "entropy": 5.201372146606445,
      "epoch": 1.6191947878382893,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004738962797177544,
      "loss": 4.9539,
      "mean_token_accuracy": 0.2031026542186737,
      "num_tokens": 35661422.0,
      "step": 19385
    },
    {
      "entropy": 5.235545492172241,
      "epoch": 1.6196124290010023,
      "grad_norm": 1.25,
      "learning_rate": 0.0004738823599716121,
      "loss": 5.0596,
      "mean_token_accuracy": 0.19618463814258574,
      "num_tokens": 35672245.0,
      "step": 19390
    },
    {
      "entropy": 5.185998201370239,
      "epoch": 1.6200300701637154,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047386843674379484,
      "loss": 4.993,
      "mean_token_accuracy": 0.19686597287654878,
      "num_tokens": 35681690.0,
      "step": 19395
    },
    {
      "entropy": 5.161135339736939,
      "epoch": 1.6204477113264284,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004738545100345461,
      "loss": 4.8385,
      "mean_token_accuracy": 0.22171805799007416,
      "num_tokens": 35690265.0,
      "step": 19400
    },
    {
      "entropy": 5.239273023605347,
      "epoch": 1.6208653524891412,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047384057984410985,
      "loss": 4.9949,
      "mean_token_accuracy": 0.20334905982017518,
      "num_tokens": 35699768.0,
      "step": 19405
    },
    {
      "entropy": 5.168808555603027,
      "epoch": 1.6212829936518545,
      "grad_norm": 1.25,
      "learning_rate": 0.00047382664617272985,
      "loss": 4.9352,
      "mean_token_accuracy": 0.20614499300718309,
      "num_tokens": 35708785.0,
      "step": 19410
    },
    {
      "entropy": 5.185434436798095,
      "epoch": 1.6217006348145673,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047381270902065013,
      "loss": 5.0188,
      "mean_token_accuracy": 0.19920573085546495,
      "num_tokens": 35719224.0,
      "step": 19415
    },
    {
      "entropy": 5.19195704460144,
      "epoch": 1.6221182759772803,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047379876838811463,
      "loss": 4.999,
      "mean_token_accuracy": 0.198872971534729,
      "num_tokens": 35727830.0,
      "step": 19420
    },
    {
      "entropy": 5.245377159118652,
      "epoch": 1.6225359171399933,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004737848242753674,
      "loss": 4.9877,
      "mean_token_accuracy": 0.20162131488323212,
      "num_tokens": 35736818.0,
      "step": 19425
    },
    {
      "entropy": 5.195708799362182,
      "epoch": 1.6229535583027062,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004737708766826525,
      "loss": 4.8829,
      "mean_token_accuracy": 0.20886215269565583,
      "num_tokens": 35746009.0,
      "step": 19430
    },
    {
      "entropy": 5.147295236587524,
      "epoch": 1.6233711994654194,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004737569256102142,
      "loss": 4.8247,
      "mean_token_accuracy": 0.21653302013874054,
      "num_tokens": 35754012.0,
      "step": 19435
    },
    {
      "entropy": 5.166499423980713,
      "epoch": 1.6237888406281322,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047374297105829663,
      "loss": 4.9975,
      "mean_token_accuracy": 0.2034294605255127,
      "num_tokens": 35763709.0,
      "step": 19440
    },
    {
      "entropy": 5.202342891693116,
      "epoch": 1.6242064817908453,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047372901302714413,
      "loss": 5.0237,
      "mean_token_accuracy": 0.20530429631471633,
      "num_tokens": 35773457.0,
      "step": 19445
    },
    {
      "entropy": 5.1435449600219725,
      "epoch": 1.6246241229535583,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047371505151700105,
      "loss": 4.8807,
      "mean_token_accuracy": 0.21354353278875352,
      "num_tokens": 35781713.0,
      "step": 19450
    },
    {
      "entropy": 5.11921763420105,
      "epoch": 1.6250417641162713,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047370108652811186,
      "loss": 4.8377,
      "mean_token_accuracy": 0.21056436151266097,
      "num_tokens": 35790796.0,
      "step": 19455
    },
    {
      "entropy": 5.150867700576782,
      "epoch": 1.6254594052789844,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004736871180607209,
      "loss": 4.8676,
      "mean_token_accuracy": 0.20786465406417848,
      "num_tokens": 35799919.0,
      "step": 19460
    },
    {
      "entropy": 5.29546217918396,
      "epoch": 1.6258770464416972,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047367314611507276,
      "loss": 5.0585,
      "mean_token_accuracy": 0.2009742796421051,
      "num_tokens": 35809642.0,
      "step": 19465
    },
    {
      "entropy": 5.250559091567993,
      "epoch": 1.6262946876044104,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047365917069141215,
      "loss": 4.9796,
      "mean_token_accuracy": 0.2021600767970085,
      "num_tokens": 35819020.0,
      "step": 19470
    },
    {
      "entropy": 5.290408563613892,
      "epoch": 1.6267123287671232,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004736451917899835,
      "loss": 4.9741,
      "mean_token_accuracy": 0.20525989383459092,
      "num_tokens": 35827657.0,
      "step": 19475
    },
    {
      "entropy": 5.1203124046325685,
      "epoch": 1.6271299699298363,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004736312094110317,
      "loss": 4.9665,
      "mean_token_accuracy": 0.2041858732700348,
      "num_tokens": 35836727.0,
      "step": 19480
    },
    {
      "entropy": 5.12161021232605,
      "epoch": 1.6275476110925493,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004736172235548016,
      "loss": 4.9115,
      "mean_token_accuracy": 0.20274921804666518,
      "num_tokens": 35846251.0,
      "step": 19485
    },
    {
      "entropy": 5.186095905303955,
      "epoch": 1.6279652522552621,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004736032342215378,
      "loss": 4.8742,
      "mean_token_accuracy": 0.20608542263507842,
      "num_tokens": 35855736.0,
      "step": 19490
    },
    {
      "entropy": 5.177062606811523,
      "epoch": 1.6283828934179754,
      "grad_norm": 2.09375,
      "learning_rate": 0.00047358924141148543,
      "loss": 4.8956,
      "mean_token_accuracy": 0.2182978704571724,
      "num_tokens": 35864630.0,
      "step": 19495
    },
    {
      "entropy": 5.21374020576477,
      "epoch": 1.6288005345806882,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047357524512488925,
      "loss": 4.9804,
      "mean_token_accuracy": 0.20113199353218078,
      "num_tokens": 35874021.0,
      "step": 19500
    },
    {
      "entropy": 5.2005042552948,
      "epoch": 1.6292181757434012,
      "grad_norm": 1.25,
      "learning_rate": 0.0004735612453619944,
      "loss": 4.9473,
      "mean_token_accuracy": 0.20366003066301347,
      "num_tokens": 35882696.0,
      "step": 19505
    },
    {
      "entropy": 5.2903505802154545,
      "epoch": 1.6296358169061143,
      "grad_norm": 1.390625,
      "learning_rate": 0.000473547242123046,
      "loss": 5.0741,
      "mean_token_accuracy": 0.1956958293914795,
      "num_tokens": 35892887.0,
      "step": 19510
    },
    {
      "entropy": 5.15625,
      "epoch": 1.6300534580688273,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004735332354082891,
      "loss": 4.8779,
      "mean_token_accuracy": 0.21098068207502366,
      "num_tokens": 35901547.0,
      "step": 19515
    },
    {
      "entropy": 5.263314533233642,
      "epoch": 1.6304710992315403,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047351922521796894,
      "loss": 5.0654,
      "mean_token_accuracy": 0.20369016528129577,
      "num_tokens": 35910982.0,
      "step": 19520
    },
    {
      "entropy": 5.2380860328674315,
      "epoch": 1.6308887403942531,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004735052115523307,
      "loss": 4.9749,
      "mean_token_accuracy": 0.20522836297750474,
      "num_tokens": 35920548.0,
      "step": 19525
    },
    {
      "entropy": 5.377325391769409,
      "epoch": 1.6313063815569664,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004734911944116199,
      "loss": 4.9723,
      "mean_token_accuracy": 0.20220075249671937,
      "num_tokens": 35930371.0,
      "step": 19530
    },
    {
      "entropy": 5.198346757888794,
      "epoch": 1.6317240227196792,
      "grad_norm": 1.46875,
      "learning_rate": 0.00047347717379608184,
      "loss": 5.017,
      "mean_token_accuracy": 0.20199014097452164,
      "num_tokens": 35939724.0,
      "step": 19535
    },
    {
      "entropy": 5.276255798339844,
      "epoch": 1.6321416638823922,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00047346314970596185,
      "loss": 5.1423,
      "mean_token_accuracy": 0.18922709077596664,
      "num_tokens": 35949432.0,
      "step": 19540
    },
    {
      "entropy": 5.349049091339111,
      "epoch": 1.6325593050451053,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047344912214150565,
      "loss": 5.0793,
      "mean_token_accuracy": 0.1991352066397667,
      "num_tokens": 35958687.0,
      "step": 19545
    },
    {
      "entropy": 5.1891937255859375,
      "epoch": 1.632976946207818,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004734350911029586,
      "loss": 5.0119,
      "mean_token_accuracy": 0.19418214112520218,
      "num_tokens": 35967875.0,
      "step": 19550
    },
    {
      "entropy": 5.225782489776611,
      "epoch": 1.6333945873705313,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047342105659056645,
      "loss": 4.897,
      "mean_token_accuracy": 0.21332972049713134,
      "num_tokens": 35977561.0,
      "step": 19555
    },
    {
      "entropy": 5.189958572387695,
      "epoch": 1.6338122285332441,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004734070186045749,
      "loss": 4.9162,
      "mean_token_accuracy": 0.20849571079015733,
      "num_tokens": 35986733.0,
      "step": 19560
    },
    {
      "entropy": 5.099578714370727,
      "epoch": 1.6342298696959572,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004733929771452296,
      "loss": 4.9062,
      "mean_token_accuracy": 0.21374163925647735,
      "num_tokens": 35995779.0,
      "step": 19565
    },
    {
      "entropy": 5.202760696411133,
      "epoch": 1.6346475108586702,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004733789322127764,
      "loss": 5.0234,
      "mean_token_accuracy": 0.2042147174477577,
      "num_tokens": 36004601.0,
      "step": 19570
    },
    {
      "entropy": 5.2062811851501465,
      "epoch": 1.6350651520213833,
      "grad_norm": 1.375,
      "learning_rate": 0.0004733648838074613,
      "loss": 4.9039,
      "mean_token_accuracy": 0.20660743713378907,
      "num_tokens": 36014356.0,
      "step": 19575
    },
    {
      "entropy": 5.161878252029419,
      "epoch": 1.6354827931840963,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047335083192953,
      "loss": 4.8499,
      "mean_token_accuracy": 0.21470391899347305,
      "num_tokens": 36023084.0,
      "step": 19580
    },
    {
      "entropy": 5.16281361579895,
      "epoch": 1.635900434346809,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004733367765792288,
      "loss": 4.9103,
      "mean_token_accuracy": 0.2057546153664589,
      "num_tokens": 36031462.0,
      "step": 19585
    },
    {
      "entropy": 5.173369979858398,
      "epoch": 1.6363180755095224,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047332271775680346,
      "loss": 4.8926,
      "mean_token_accuracy": 0.21195039302110671,
      "num_tokens": 36041107.0,
      "step": 19590
    },
    {
      "entropy": 5.168484497070312,
      "epoch": 1.6367357166722352,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004733086554625004,
      "loss": 4.95,
      "mean_token_accuracy": 0.20458746403455735,
      "num_tokens": 36050026.0,
      "step": 19595
    },
    {
      "entropy": 5.156893110275268,
      "epoch": 1.6371533578349482,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004732945896965654,
      "loss": 4.9711,
      "mean_token_accuracy": 0.20600138157606124,
      "num_tokens": 36058858.0,
      "step": 19600
    },
    {
      "entropy": 5.257830476760864,
      "epoch": 1.6375709989976612,
      "grad_norm": 1.8828125,
      "learning_rate": 0.00047328052045924506,
      "loss": 5.0439,
      "mean_token_accuracy": 0.19607376903295518,
      "num_tokens": 36067995.0,
      "step": 19605
    },
    {
      "entropy": 5.17674298286438,
      "epoch": 1.637988640160374,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004732664477507855,
      "loss": 4.9177,
      "mean_token_accuracy": 0.20759281218051912,
      "num_tokens": 36077425.0,
      "step": 19610
    },
    {
      "entropy": 5.191337060928345,
      "epoch": 1.6384062813230873,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047325237157143314,
      "loss": 4.9352,
      "mean_token_accuracy": 0.20020273476839065,
      "num_tokens": 36086667.0,
      "step": 19615
    },
    {
      "entropy": 5.179776382446289,
      "epoch": 1.6388239224858,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00047323829192143436,
      "loss": 4.894,
      "mean_token_accuracy": 0.2136690855026245,
      "num_tokens": 36095199.0,
      "step": 19620
    },
    {
      "entropy": 5.137758398056031,
      "epoch": 1.6392415636485131,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047322420880103566,
      "loss": 4.8942,
      "mean_token_accuracy": 0.20340508371591567,
      "num_tokens": 36104587.0,
      "step": 19625
    },
    {
      "entropy": 5.244015121459961,
      "epoch": 1.6396592048112262,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004732101222104836,
      "loss": 4.9361,
      "mean_token_accuracy": 0.2066241294145584,
      "num_tokens": 36114017.0,
      "step": 19630
    },
    {
      "entropy": 5.184106063842774,
      "epoch": 1.6400768459739392,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004731960321500247,
      "loss": 4.9605,
      "mean_token_accuracy": 0.20201909244060517,
      "num_tokens": 36122301.0,
      "step": 19635
    },
    {
      "entropy": 5.125623512268066,
      "epoch": 1.6404944871366522,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004731819386199058,
      "loss": 4.9288,
      "mean_token_accuracy": 0.21494675278663636,
      "num_tokens": 36131538.0,
      "step": 19640
    },
    {
      "entropy": 5.215543985366821,
      "epoch": 1.640912128299365,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004731678416203734,
      "loss": 4.9331,
      "mean_token_accuracy": 0.2044935032725334,
      "num_tokens": 36140472.0,
      "step": 19645
    },
    {
      "entropy": 5.167410564422608,
      "epoch": 1.6413297694620783,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047315374115167446,
      "loss": 4.9178,
      "mean_token_accuracy": 0.20780009031295776,
      "num_tokens": 36148947.0,
      "step": 19650
    },
    {
      "entropy": 5.116166305541992,
      "epoch": 1.6417474106247911,
      "grad_norm": 1.375,
      "learning_rate": 0.0004731396372140558,
      "loss": 4.8801,
      "mean_token_accuracy": 0.20493905991315842,
      "num_tokens": 36157895.0,
      "step": 19655
    },
    {
      "entropy": 5.202142190933228,
      "epoch": 1.6421650517875042,
      "grad_norm": 1.375,
      "learning_rate": 0.0004731255298077642,
      "loss": 4.943,
      "mean_token_accuracy": 0.21433926671743392,
      "num_tokens": 36168170.0,
      "step": 19660
    },
    {
      "entropy": 5.202002143859863,
      "epoch": 1.6425826929502172,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00047311141893304675,
      "loss": 5.0056,
      "mean_token_accuracy": 0.2005702570080757,
      "num_tokens": 36176864.0,
      "step": 19665
    },
    {
      "entropy": 5.285471343994141,
      "epoch": 1.64300033411293,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004730973045901505,
      "loss": 4.9915,
      "mean_token_accuracy": 0.2068271443247795,
      "num_tokens": 36186334.0,
      "step": 19670
    },
    {
      "entropy": 5.256713151931763,
      "epoch": 1.6434179752756433,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047308318677932243,
      "loss": 4.9894,
      "mean_token_accuracy": 0.20809144228696824,
      "num_tokens": 36194688.0,
      "step": 19675
    },
    {
      "entropy": 5.088052129745483,
      "epoch": 1.643835616438356,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004730690655008097,
      "loss": 4.953,
      "mean_token_accuracy": 0.20112271159887313,
      "num_tokens": 36203962.0,
      "step": 19680
    },
    {
      "entropy": 5.218835783004761,
      "epoch": 1.644253257601069,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004730549407548597,
      "loss": 5.0019,
      "mean_token_accuracy": 0.19580864459276198,
      "num_tokens": 36212780.0,
      "step": 19685
    },
    {
      "entropy": 5.260392522811889,
      "epoch": 1.6446708987637821,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047304081254171944,
      "loss": 4.9718,
      "mean_token_accuracy": 0.20621494948863983,
      "num_tokens": 36221744.0,
      "step": 19690
    },
    {
      "entropy": 5.197296285629273,
      "epoch": 1.6450885399264952,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004730266808616364,
      "loss": 5.0661,
      "mean_token_accuracy": 0.20622545778751372,
      "num_tokens": 36231261.0,
      "step": 19695
    },
    {
      "entropy": 5.216507863998413,
      "epoch": 1.6455061810892082,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000473012545714858,
      "loss": 4.9639,
      "mean_token_accuracy": 0.20818955898284913,
      "num_tokens": 36240599.0,
      "step": 19700
    },
    {
      "entropy": 5.120757389068603,
      "epoch": 1.645923822251921,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047299840710163157,
      "loss": 4.8227,
      "mean_token_accuracy": 0.21667277216911315,
      "num_tokens": 36249402.0,
      "step": 19705
    },
    {
      "entropy": 5.19513168334961,
      "epoch": 1.6463414634146343,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004729842650222047,
      "loss": 4.878,
      "mean_token_accuracy": 0.21917334496974944,
      "num_tokens": 36258379.0,
      "step": 19710
    },
    {
      "entropy": 5.182104682922363,
      "epoch": 1.646759104577347,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047297011947682494,
      "loss": 4.9576,
      "mean_token_accuracy": 0.20872343480587005,
      "num_tokens": 36267393.0,
      "step": 19715
    },
    {
      "entropy": 5.23886456489563,
      "epoch": 1.6471767457400601,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047295597046574,
      "loss": 4.9851,
      "mean_token_accuracy": 0.20521213263273239,
      "num_tokens": 36277542.0,
      "step": 19720
    },
    {
      "entropy": 5.213271808624268,
      "epoch": 1.6475943869027732,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004729418179891975,
      "loss": 4.9888,
      "mean_token_accuracy": 0.20395977050065994,
      "num_tokens": 36287623.0,
      "step": 19725
    },
    {
      "entropy": 5.142971754074097,
      "epoch": 1.6480120280654862,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004729276620474452,
      "loss": 4.8912,
      "mean_token_accuracy": 0.21244652569293976,
      "num_tokens": 36295828.0,
      "step": 19730
    },
    {
      "entropy": 5.168543767929077,
      "epoch": 1.6484296692281992,
      "grad_norm": 1.75,
      "learning_rate": 0.00047291350264073095,
      "loss": 4.9481,
      "mean_token_accuracy": 0.2009446769952774,
      "num_tokens": 36304825.0,
      "step": 19735
    },
    {
      "entropy": 5.247984170913696,
      "epoch": 1.648847310390912,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047289933976930263,
      "loss": 4.9951,
      "mean_token_accuracy": 0.20509831011295318,
      "num_tokens": 36313691.0,
      "step": 19740
    },
    {
      "entropy": 5.211964464187622,
      "epoch": 1.6492649515536253,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004728851734334081,
      "loss": 4.951,
      "mean_token_accuracy": 0.20112333595752716,
      "num_tokens": 36323291.0,
      "step": 19745
    },
    {
      "entropy": 5.175315618515015,
      "epoch": 1.649682592716338,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004728710036332954,
      "loss": 4.9395,
      "mean_token_accuracy": 0.2036467343568802,
      "num_tokens": 36332859.0,
      "step": 19750
    },
    {
      "entropy": 5.209541511535645,
      "epoch": 1.6501002338790511,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047285683036921275,
      "loss": 4.9687,
      "mean_token_accuracy": 0.2085937350988388,
      "num_tokens": 36342049.0,
      "step": 19755
    },
    {
      "entropy": 5.110041284561158,
      "epoch": 1.6505178750417642,
      "grad_norm": 1.515625,
      "learning_rate": 0.00047284265364140794,
      "loss": 4.912,
      "mean_token_accuracy": 0.20162131637334824,
      "num_tokens": 36351512.0,
      "step": 19760
    },
    {
      "entropy": 5.220494413375855,
      "epoch": 1.650935516204477,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004728284734501295,
      "loss": 5.0058,
      "mean_token_accuracy": 0.20385658144950866,
      "num_tokens": 36361923.0,
      "step": 19765
    },
    {
      "entropy": 5.201858186721802,
      "epoch": 1.6513531573671902,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004728142897956253,
      "loss": 5.0051,
      "mean_token_accuracy": 0.201225408911705,
      "num_tokens": 36370543.0,
      "step": 19770
    },
    {
      "entropy": 5.115618181228638,
      "epoch": 1.651770798529903,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000472800102678144,
      "loss": 4.889,
      "mean_token_accuracy": 0.21700887531042098,
      "num_tokens": 36378465.0,
      "step": 19775
    },
    {
      "entropy": 5.238144969940185,
      "epoch": 1.652188439692616,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00047278591209793384,
      "loss": 4.9789,
      "mean_token_accuracy": 0.2057315304875374,
      "num_tokens": 36387540.0,
      "step": 19780
    },
    {
      "entropy": 5.216234111785889,
      "epoch": 1.6526060808553291,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004727717180552431,
      "loss": 4.9675,
      "mean_token_accuracy": 0.20264612436294555,
      "num_tokens": 36397859.0,
      "step": 19785
    },
    {
      "entropy": 5.148710870742798,
      "epoch": 1.6530237220180422,
      "grad_norm": 1.515625,
      "learning_rate": 0.00047275752055032043,
      "loss": 4.8728,
      "mean_token_accuracy": 0.2148661881685257,
      "num_tokens": 36406955.0,
      "step": 19790
    },
    {
      "entropy": 5.135104560852051,
      "epoch": 1.6534413631807552,
      "grad_norm": 1.640625,
      "learning_rate": 0.00047274331958341434,
      "loss": 4.8694,
      "mean_token_accuracy": 0.211592897772789,
      "num_tokens": 36415499.0,
      "step": 19795
    },
    {
      "entropy": 5.16286826133728,
      "epoch": 1.653859004343468,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004727291151547733,
      "loss": 4.9315,
      "mean_token_accuracy": 0.20528517812490463,
      "num_tokens": 36424348.0,
      "step": 19800
    },
    {
      "entropy": 5.14400086402893,
      "epoch": 1.6542766455061813,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004727149072646463,
      "loss": 4.874,
      "mean_token_accuracy": 0.2060215726494789,
      "num_tokens": 36433134.0,
      "step": 19805
    },
    {
      "entropy": 5.086054182052612,
      "epoch": 1.654694286668894,
      "grad_norm": 1.578125,
      "learning_rate": 0.00047270069591328175,
      "loss": 4.8431,
      "mean_token_accuracy": 0.214297117292881,
      "num_tokens": 36442300.0,
      "step": 19810
    },
    {
      "entropy": 5.233012104034424,
      "epoch": 1.655111927831607,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047268648110092857,
      "loss": 5.0279,
      "mean_token_accuracy": 0.19737976640462876,
      "num_tokens": 36452265.0,
      "step": 19815
    },
    {
      "entropy": 5.211163473129273,
      "epoch": 1.6555295689943201,
      "grad_norm": 1.375,
      "learning_rate": 0.00047267226282783555,
      "loss": 4.9469,
      "mean_token_accuracy": 0.20313335508108138,
      "num_tokens": 36462009.0,
      "step": 19820
    },
    {
      "entropy": 5.2558797836303714,
      "epoch": 1.655947210157033,
      "grad_norm": 1.375,
      "learning_rate": 0.0004726580410942516,
      "loss": 5.0673,
      "mean_token_accuracy": 0.19776800870895386,
      "num_tokens": 36470799.0,
      "step": 19825
    },
    {
      "entropy": 5.232189321517945,
      "epoch": 1.6563648513197462,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047264381590042575,
      "loss": 4.9495,
      "mean_token_accuracy": 0.20047251582145692,
      "num_tokens": 36479362.0,
      "step": 19830
    },
    {
      "entropy": 5.17759599685669,
      "epoch": 1.656782492482459,
      "grad_norm": 1.359375,
      "learning_rate": 0.000472629587246607,
      "loss": 4.8935,
      "mean_token_accuracy": 0.20568197667598725,
      "num_tokens": 36487773.0,
      "step": 19835
    },
    {
      "entropy": 5.086628770828247,
      "epoch": 1.657200133645172,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047261535513304446,
      "loss": 4.8499,
      "mean_token_accuracy": 0.21870521157979966,
      "num_tokens": 36496133.0,
      "step": 19840
    },
    {
      "entropy": 5.206830930709839,
      "epoch": 1.657617774807885,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047260111955998725,
      "loss": 5.0091,
      "mean_token_accuracy": 0.20386249721050262,
      "num_tokens": 36505392.0,
      "step": 19845
    },
    {
      "entropy": 5.240763187408447,
      "epoch": 1.6580354159705981,
      "grad_norm": 1.453125,
      "learning_rate": 0.00047258688052768456,
      "loss": 4.9836,
      "mean_token_accuracy": 0.2040731802582741,
      "num_tokens": 36514876.0,
      "step": 19850
    },
    {
      "entropy": 5.1919961929321286,
      "epoch": 1.6584530571333111,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004725726380363857,
      "loss": 4.8996,
      "mean_token_accuracy": 0.21279468834400178,
      "num_tokens": 36523828.0,
      "step": 19855
    },
    {
      "entropy": 5.167350435256958,
      "epoch": 1.658870698296024,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047255839208634005,
      "loss": 4.9736,
      "mean_token_accuracy": 0.20157965123653412,
      "num_tokens": 36532263.0,
      "step": 19860
    },
    {
      "entropy": 5.126400709152222,
      "epoch": 1.6592883394587372,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047254414267779686,
      "loss": 4.8698,
      "mean_token_accuracy": 0.21058623790740966,
      "num_tokens": 36541015.0,
      "step": 19865
    },
    {
      "entropy": 5.111027765274048,
      "epoch": 1.65970598062145,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047252988981100576,
      "loss": 4.9605,
      "mean_token_accuracy": 0.20256894528865815,
      "num_tokens": 36550254.0,
      "step": 19870
    },
    {
      "entropy": 5.173180484771729,
      "epoch": 1.660123621784163,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004725156334862161,
      "loss": 4.995,
      "mean_token_accuracy": 0.19718596786260606,
      "num_tokens": 36560941.0,
      "step": 19875
    },
    {
      "entropy": 5.292994356155395,
      "epoch": 1.660541262946876,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047250137370367755,
      "loss": 5.0055,
      "mean_token_accuracy": 0.19614981710910798,
      "num_tokens": 36570443.0,
      "step": 19880
    },
    {
      "entropy": 5.258962869644165,
      "epoch": 1.660958904109589,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047248711046363976,
      "loss": 4.9823,
      "mean_token_accuracy": 0.20191563069820403,
      "num_tokens": 36580532.0,
      "step": 19885
    },
    {
      "entropy": 5.204261922836304,
      "epoch": 1.6613765452723022,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004724728437663523,
      "loss": 4.9276,
      "mean_token_accuracy": 0.20584801584482193,
      "num_tokens": 36589359.0,
      "step": 19890
    },
    {
      "entropy": 5.235820579528808,
      "epoch": 1.661794186435015,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047245857361206506,
      "loss": 4.9603,
      "mean_token_accuracy": 0.20352103114128112,
      "num_tokens": 36598229.0,
      "step": 19895
    },
    {
      "entropy": 5.140442848205566,
      "epoch": 1.662211827597728,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047244430000102785,
      "loss": 4.9185,
      "mean_token_accuracy": 0.20831808745861052,
      "num_tokens": 36607655.0,
      "step": 19900
    },
    {
      "entropy": 5.244342184066772,
      "epoch": 1.662629468760441,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004724300229334905,
      "loss": 4.9952,
      "mean_token_accuracy": 0.1982486516237259,
      "num_tokens": 36618060.0,
      "step": 19905
    },
    {
      "entropy": 5.2188855648040775,
      "epoch": 1.663047109923154,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047241574240970284,
      "loss": 4.9292,
      "mean_token_accuracy": 0.20655286759138108,
      "num_tokens": 36627825.0,
      "step": 19910
    },
    {
      "entropy": 5.160020637512207,
      "epoch": 1.663464751085867,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047240145842991513,
      "loss": 4.9257,
      "mean_token_accuracy": 0.21406910717487335,
      "num_tokens": 36635971.0,
      "step": 19915
    },
    {
      "entropy": 5.106333923339844,
      "epoch": 1.66388239224858,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047238717099437713,
      "loss": 4.7917,
      "mean_token_accuracy": 0.2127847746014595,
      "num_tokens": 36644340.0,
      "step": 19920
    },
    {
      "entropy": 5.197990751266479,
      "epoch": 1.6643000334112932,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004723728801033392,
      "loss": 4.9441,
      "mean_token_accuracy": 0.20018717646598816,
      "num_tokens": 36654371.0,
      "step": 19925
    },
    {
      "entropy": 5.1067720413208,
      "epoch": 1.664717674574006,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004723585857570514,
      "loss": 4.8165,
      "mean_token_accuracy": 0.21989476680755615,
      "num_tokens": 36663633.0,
      "step": 19930
    },
    {
      "entropy": 5.091370964050293,
      "epoch": 1.665135315736719,
      "grad_norm": 1.28125,
      "learning_rate": 0.000472344287955764,
      "loss": 4.8324,
      "mean_token_accuracy": 0.21990165561437608,
      "num_tokens": 36673004.0,
      "step": 19935
    },
    {
      "entropy": 5.088794708251953,
      "epoch": 1.665552956899432,
      "grad_norm": 1.25,
      "learning_rate": 0.0004723299866997273,
      "loss": 4.8393,
      "mean_token_accuracy": 0.22104309648275375,
      "num_tokens": 36681323.0,
      "step": 19940
    },
    {
      "entropy": 5.2208168506622314,
      "epoch": 1.6659705980621449,
      "grad_norm": 1.375,
      "learning_rate": 0.0004723156819891917,
      "loss": 5.0311,
      "mean_token_accuracy": 0.19263727813959122,
      "num_tokens": 36691128.0,
      "step": 19945
    },
    {
      "entropy": 5.1351954460144045,
      "epoch": 1.6663882392248581,
      "grad_norm": 1.375,
      "learning_rate": 0.00047230137382440747,
      "loss": 4.9232,
      "mean_token_accuracy": 0.21018970161676406,
      "num_tokens": 36699759.0,
      "step": 19950
    },
    {
      "entropy": 5.130295944213867,
      "epoch": 1.666805880387571,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004722870622056253,
      "loss": 4.9559,
      "mean_token_accuracy": 0.20856220573186873,
      "num_tokens": 36708723.0,
      "step": 19955
    },
    {
      "entropy": 5.211036300659179,
      "epoch": 1.667223521550284,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004722727471330956,
      "loss": 4.8718,
      "mean_token_accuracy": 0.20452526658773423,
      "num_tokens": 36717488.0,
      "step": 19960
    },
    {
      "entropy": 5.085398197174072,
      "epoch": 1.667641162712997,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000472258428607069,
      "loss": 4.8497,
      "mean_token_accuracy": 0.20833572149276733,
      "num_tokens": 36726477.0,
      "step": 19965
    },
    {
      "entropy": 5.164642381668091,
      "epoch": 1.66805880387571,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004722441066277961,
      "loss": 4.8469,
      "mean_token_accuracy": 0.2083793967962265,
      "num_tokens": 36734676.0,
      "step": 19970
    },
    {
      "entropy": 5.179826498031616,
      "epoch": 1.668476445038423,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00047222978119552776,
      "loss": 4.9933,
      "mean_token_accuracy": 0.20257626473903656,
      "num_tokens": 36744425.0,
      "step": 19975
    },
    {
      "entropy": 5.204170989990234,
      "epoch": 1.6688940862011359,
      "grad_norm": 1.375,
      "learning_rate": 0.00047221545231051473,
      "loss": 4.9446,
      "mean_token_accuracy": 0.20276660174131395,
      "num_tokens": 36754113.0,
      "step": 19980
    },
    {
      "entropy": 5.151750755310059,
      "epoch": 1.6693117273638491,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004722011199730078,
      "loss": 4.8982,
      "mean_token_accuracy": 0.20922341495752333,
      "num_tokens": 36763013.0,
      "step": 19985
    },
    {
      "entropy": 5.147961711883545,
      "epoch": 1.669729368526562,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004721867841832579,
      "loss": 4.9571,
      "mean_token_accuracy": 0.20012480467557908,
      "num_tokens": 36772115.0,
      "step": 19990
    },
    {
      "entropy": 5.254347658157348,
      "epoch": 1.670147009689275,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004721724449415159,
      "loss": 4.9696,
      "mean_token_accuracy": 0.19898971319198608,
      "num_tokens": 36781210.0,
      "step": 19995
    },
    {
      "entropy": 5.27505784034729,
      "epoch": 1.670564650851988,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047215810224803304,
      "loss": 5.0325,
      "mean_token_accuracy": 0.1990812227129936,
      "num_tokens": 36791131.0,
      "step": 20000
    },
    {
      "entropy": 5.141687345504761,
      "epoch": 1.6709822920147008,
      "grad_norm": 1.25,
      "learning_rate": 0.00047214375610306016,
      "loss": 4.8123,
      "mean_token_accuracy": 0.21562213599681854,
      "num_tokens": 36800618.0,
      "step": 20005
    },
    {
      "entropy": 5.124288940429688,
      "epoch": 1.671399933177414,
      "grad_norm": 1.484375,
      "learning_rate": 0.00047212940650684853,
      "loss": 4.9937,
      "mean_token_accuracy": 0.2003643050789833,
      "num_tokens": 36809995.0,
      "step": 20010
    },
    {
      "entropy": 5.198476076126099,
      "epoch": 1.671817574340127,
      "grad_norm": 1.375,
      "learning_rate": 0.0004721150534596495,
      "loss": 4.9654,
      "mean_token_accuracy": 0.20180171728134155,
      "num_tokens": 36818492.0,
      "step": 20015
    },
    {
      "entropy": 5.212288284301758,
      "epoch": 1.67223521550284,
      "grad_norm": 1.234375,
      "learning_rate": 0.000472100696961714,
      "loss": 4.9167,
      "mean_token_accuracy": 0.2127817988395691,
      "num_tokens": 36827280.0,
      "step": 20020
    },
    {
      "entropy": 5.188753414154053,
      "epoch": 1.672652856665553,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004720863370132937,
      "loss": 4.976,
      "mean_token_accuracy": 0.2105330154299736,
      "num_tokens": 36837931.0,
      "step": 20025
    },
    {
      "entropy": 5.215947437286377,
      "epoch": 1.673070497828266,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004720719736146398,
      "loss": 5.0121,
      "mean_token_accuracy": 0.2027532562613487,
      "num_tokens": 36847581.0,
      "step": 20030
    },
    {
      "entropy": 5.164034223556518,
      "epoch": 1.673488138990979,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047205760676600377,
      "loss": 4.9221,
      "mean_token_accuracy": 0.20446307361125945,
      "num_tokens": 36857147.0,
      "step": 20035
    },
    {
      "entropy": 5.127425765991211,
      "epoch": 1.6739057801536918,
      "grad_norm": 1.25,
      "learning_rate": 0.0004720432364676372,
      "loss": 4.9221,
      "mean_token_accuracy": 0.20291653722524644,
      "num_tokens": 36866105.0,
      "step": 20040
    },
    {
      "entropy": 5.20668478012085,
      "epoch": 1.674323421316405,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004720288627197915,
      "loss": 4.8775,
      "mean_token_accuracy": 0.2151976391673088,
      "num_tokens": 36874631.0,
      "step": 20045
    },
    {
      "entropy": 5.149134874343872,
      "epoch": 1.674741062479118,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004720144855227185,
      "loss": 4.8535,
      "mean_token_accuracy": 0.2150028795003891,
      "num_tokens": 36884264.0,
      "step": 20050
    },
    {
      "entropy": 5.212637805938721,
      "epoch": 1.675158703641831,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004720001048766696,
      "loss": 4.9467,
      "mean_token_accuracy": 0.2070896953344345,
      "num_tokens": 36893143.0,
      "step": 20055
    },
    {
      "entropy": 5.282405185699463,
      "epoch": 1.675576344804544,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004719857207818969,
      "loss": 5.0255,
      "mean_token_accuracy": 0.20156907737255098,
      "num_tokens": 36903026.0,
      "step": 20060
    },
    {
      "entropy": 5.290762138366699,
      "epoch": 1.6759939859672568,
      "grad_norm": 1.328125,
      "learning_rate": 0.000471971333238652,
      "loss": 5.0318,
      "mean_token_accuracy": 0.20415476113557815,
      "num_tokens": 36912737.0,
      "step": 20065
    },
    {
      "entropy": 5.27071213722229,
      "epoch": 1.67641162712997,
      "grad_norm": 1.53125,
      "learning_rate": 0.00047195694224718684,
      "loss": 5.0087,
      "mean_token_accuracy": 0.20318905264139175,
      "num_tokens": 36921148.0,
      "step": 20070
    },
    {
      "entropy": 5.099691009521484,
      "epoch": 1.6768292682926829,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047194254780775325,
      "loss": 4.8279,
      "mean_token_accuracy": 0.213830304145813,
      "num_tokens": 36930043.0,
      "step": 20075
    },
    {
      "entropy": 5.229407882690429,
      "epoch": 1.677246909455396,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004719281499206034,
      "loss": 5.0088,
      "mean_token_accuracy": 0.200849249958992,
      "num_tokens": 36939124.0,
      "step": 20080
    },
    {
      "entropy": 5.307264375686645,
      "epoch": 1.677664550618109,
      "grad_norm": 1.40625,
      "learning_rate": 0.00047191374858598917,
      "loss": 5.0651,
      "mean_token_accuracy": 0.19725067764520646,
      "num_tokens": 36948392.0,
      "step": 20085
    },
    {
      "entropy": 5.2197747230529785,
      "epoch": 1.678082191780822,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004718993438041628,
      "loss": 4.9576,
      "mean_token_accuracy": 0.2045763462781906,
      "num_tokens": 36958019.0,
      "step": 20090
    },
    {
      "entropy": 5.156815719604492,
      "epoch": 1.678499832943535,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047188493557537634,
      "loss": 4.9022,
      "mean_token_accuracy": 0.20272983610630035,
      "num_tokens": 36966904.0,
      "step": 20095
    },
    {
      "entropy": 5.306857776641846,
      "epoch": 1.6789174741062478,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004718705238998822,
      "loss": 4.9775,
      "mean_token_accuracy": 0.20175060629844666,
      "num_tokens": 36976085.0,
      "step": 20100
    },
    {
      "entropy": 5.174426937103272,
      "epoch": 1.679335115268961,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047185610877793245,
      "loss": 4.8818,
      "mean_token_accuracy": 0.21443032771348952,
      "num_tokens": 36984345.0,
      "step": 20105
    },
    {
      "entropy": 5.086754369735718,
      "epoch": 1.6797527564316739,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047184169020977965,
      "loss": 4.9065,
      "mean_token_accuracy": 0.2139348104596138,
      "num_tokens": 36993133.0,
      "step": 20110
    },
    {
      "entropy": 5.179343605041504,
      "epoch": 1.680170397594387,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047182726819567603,
      "loss": 5.0005,
      "mean_token_accuracy": 0.20244463831186293,
      "num_tokens": 37002797.0,
      "step": 20115
    },
    {
      "entropy": 5.280191659927368,
      "epoch": 1.6805880387571,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047181284273587414,
      "loss": 4.8918,
      "mean_token_accuracy": 0.20694702565670015,
      "num_tokens": 37012598.0,
      "step": 20120
    },
    {
      "entropy": 5.1645283699035645,
      "epoch": 1.6810056799198128,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00047179841383062667,
      "loss": 4.921,
      "mean_token_accuracy": 0.2145317554473877,
      "num_tokens": 37022006.0,
      "step": 20125
    },
    {
      "entropy": 5.052135992050171,
      "epoch": 1.681423321082526,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047178398148018606,
      "loss": 4.8545,
      "mean_token_accuracy": 0.20530718713998794,
      "num_tokens": 37030713.0,
      "step": 20130
    },
    {
      "entropy": 5.155506610870361,
      "epoch": 1.6818409622452388,
      "grad_norm": 1.25,
      "learning_rate": 0.0004717695456848049,
      "loss": 4.9066,
      "mean_token_accuracy": 0.21597629487514497,
      "num_tokens": 37039347.0,
      "step": 20135
    },
    {
      "entropy": 5.276638984680176,
      "epoch": 1.6822586034079519,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004717551064447361,
      "loss": 4.9831,
      "mean_token_accuracy": 0.20095666795969008,
      "num_tokens": 37048010.0,
      "step": 20140
    },
    {
      "entropy": 5.1252110481262205,
      "epoch": 1.682676244570665,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004717406637602322,
      "loss": 4.8579,
      "mean_token_accuracy": 0.20889840573072432,
      "num_tokens": 37056517.0,
      "step": 20145
    },
    {
      "entropy": 5.14582257270813,
      "epoch": 1.683093885733378,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004717262176315462,
      "loss": 4.9096,
      "mean_token_accuracy": 0.20505275428295136,
      "num_tokens": 37065085.0,
      "step": 20150
    },
    {
      "entropy": 5.197108888626099,
      "epoch": 1.683511526896091,
      "grad_norm": 1.296875,
      "learning_rate": 0.000471711768058931,
      "loss": 4.8822,
      "mean_token_accuracy": 0.20972085744142532,
      "num_tokens": 37074483.0,
      "step": 20155
    },
    {
      "entropy": 5.241299486160278,
      "epoch": 1.6839291680588038,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004716973150426395,
      "loss": 4.9795,
      "mean_token_accuracy": 0.2036333814263344,
      "num_tokens": 37083679.0,
      "step": 20160
    },
    {
      "entropy": 5.134264659881592,
      "epoch": 1.684346809221517,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047168285858292477,
      "loss": 4.914,
      "mean_token_accuracy": 0.2105153739452362,
      "num_tokens": 37092258.0,
      "step": 20165
    },
    {
      "entropy": 5.175561141967774,
      "epoch": 1.6847644503842298,
      "grad_norm": 1.53125,
      "learning_rate": 0.00047166839868003984,
      "loss": 5.0069,
      "mean_token_accuracy": 0.19648572951555252,
      "num_tokens": 37101100.0,
      "step": 20170
    },
    {
      "entropy": 5.302547216415405,
      "epoch": 1.6851820915469429,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047165393533423784,
      "loss": 5.0593,
      "mean_token_accuracy": 0.20240088701248168,
      "num_tokens": 37110876.0,
      "step": 20175
    },
    {
      "entropy": 5.236174821853638,
      "epoch": 1.685599732709656,
      "grad_norm": 1.671875,
      "learning_rate": 0.00047163946854577194,
      "loss": 4.9015,
      "mean_token_accuracy": 0.20927055180072784,
      "num_tokens": 37119770.0,
      "step": 20180
    },
    {
      "entropy": 5.128023147583008,
      "epoch": 1.6860173738723687,
      "grad_norm": 2.125,
      "learning_rate": 0.0004716249983148955,
      "loss": 4.9048,
      "mean_token_accuracy": 0.20661077052354812,
      "num_tokens": 37129173.0,
      "step": 20185
    },
    {
      "entropy": 5.1199873924255375,
      "epoch": 1.686435015035082,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004716105246418618,
      "loss": 4.864,
      "mean_token_accuracy": 0.21677151173353196,
      "num_tokens": 37138229.0,
      "step": 20190
    },
    {
      "entropy": 5.245790100097656,
      "epoch": 1.6868526561977948,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004715960475269242,
      "loss": 5.0127,
      "mean_token_accuracy": 0.20405316948890687,
      "num_tokens": 37147792.0,
      "step": 20195
    },
    {
      "entropy": 5.241026544570923,
      "epoch": 1.6872702973605078,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004715815669703362,
      "loss": 4.8881,
      "mean_token_accuracy": 0.21544091105461122,
      "num_tokens": 37156729.0,
      "step": 20200
    },
    {
      "entropy": 5.120996761322021,
      "epoch": 1.6876879385232209,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00047156708297235115,
      "loss": 4.924,
      "mean_token_accuracy": 0.20826505422592162,
      "num_tokens": 37165678.0,
      "step": 20205
    },
    {
      "entropy": 5.330936527252197,
      "epoch": 1.6881055796859339,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004715525955332227,
      "loss": 5.0455,
      "mean_token_accuracy": 0.20050025433301927,
      "num_tokens": 37174704.0,
      "step": 20210
    },
    {
      "entropy": 5.26752495765686,
      "epoch": 1.688523220848647,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004715381046532046,
      "loss": 4.9545,
      "mean_token_accuracy": 0.2040913686156273,
      "num_tokens": 37184548.0,
      "step": 20215
    },
    {
      "entropy": 5.124199104309082,
      "epoch": 1.6889408620113597,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004715236103325503,
      "loss": 4.9018,
      "mean_token_accuracy": 0.20464568436145783,
      "num_tokens": 37193668.0,
      "step": 20220
    },
    {
      "entropy": 5.125013113021851,
      "epoch": 1.689358503174073,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047150911257151364,
      "loss": 4.9346,
      "mean_token_accuracy": 0.20223360657691955,
      "num_tokens": 37202504.0,
      "step": 20225
    },
    {
      "entropy": 5.178895330429077,
      "epoch": 1.6897761443367858,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004714946113703484,
      "loss": 4.8635,
      "mean_token_accuracy": 0.21255605816841125,
      "num_tokens": 37212140.0,
      "step": 20230
    },
    {
      "entropy": 5.232712507247925,
      "epoch": 1.6901937854994988,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004714801067293086,
      "loss": 5.0693,
      "mean_token_accuracy": 0.2032686084508896,
      "num_tokens": 37221824.0,
      "step": 20235
    },
    {
      "entropy": 5.29395022392273,
      "epoch": 1.6906114266622119,
      "grad_norm": 1.8359375,
      "learning_rate": 0.00047146559864864787,
      "loss": 5.0167,
      "mean_token_accuracy": 0.19972566068172454,
      "num_tokens": 37231877.0,
      "step": 20240
    },
    {
      "entropy": 5.279973030090332,
      "epoch": 1.6910290678249247,
      "grad_norm": 1.8515625,
      "learning_rate": 0.00047145108712862046,
      "loss": 4.9908,
      "mean_token_accuracy": 0.2025890663266182,
      "num_tokens": 37240601.0,
      "step": 20245
    },
    {
      "entropy": 5.091606569290161,
      "epoch": 1.691446708987638,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004714365721694803,
      "loss": 4.8961,
      "mean_token_accuracy": 0.2045533090829849,
      "num_tokens": 37249645.0,
      "step": 20250
    },
    {
      "entropy": 5.1299560546875,
      "epoch": 1.6918643501503507,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00047142205377148145,
      "loss": 4.9051,
      "mean_token_accuracy": 0.20778035521507263,
      "num_tokens": 37258170.0,
      "step": 20255
    },
    {
      "entropy": 5.19208517074585,
      "epoch": 1.6922819913130638,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004714075319348781,
      "loss": 4.9716,
      "mean_token_accuracy": 0.20785680413246155,
      "num_tokens": 37268045.0,
      "step": 20260
    },
    {
      "entropy": 5.233912420272827,
      "epoch": 1.6926996324757768,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004713930066599246,
      "loss": 4.9634,
      "mean_token_accuracy": 0.20252827554941177,
      "num_tokens": 37278558.0,
      "step": 20265
    },
    {
      "entropy": 5.150757884979248,
      "epoch": 1.6931172736384899,
      "grad_norm": 1.515625,
      "learning_rate": 0.000471378477946875,
      "loss": 4.8179,
      "mean_token_accuracy": 0.21397656500339507,
      "num_tokens": 37287119.0,
      "step": 20270
    },
    {
      "entropy": 5.120070886611939,
      "epoch": 1.6935349148012029,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004713639457959838,
      "loss": 4.9393,
      "mean_token_accuracy": 0.2074027583003044,
      "num_tokens": 37296300.0,
      "step": 20275
    },
    {
      "entropy": 5.181284427642822,
      "epoch": 1.6939525559639157,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004713494102075053,
      "loss": 4.93,
      "mean_token_accuracy": 0.207780858874321,
      "num_tokens": 37305291.0,
      "step": 20280
    },
    {
      "entropy": 5.221661138534546,
      "epoch": 1.694370197126629,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004713348711816942,
      "loss": 4.9,
      "mean_token_accuracy": 0.20995957106351854,
      "num_tokens": 37314441.0,
      "step": 20285
    },
    {
      "entropy": 5.215374374389649,
      "epoch": 1.6947878382893418,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004713203287188047,
      "loss": 4.9966,
      "mean_token_accuracy": 0.20747386515140534,
      "num_tokens": 37323516.0,
      "step": 20290
    },
    {
      "entropy": 5.161603164672852,
      "epoch": 1.6952054794520548,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004713057828190916,
      "loss": 4.922,
      "mean_token_accuracy": 0.20976592004299163,
      "num_tokens": 37332651.0,
      "step": 20295
    },
    {
      "entropy": 5.1770429611206055,
      "epoch": 1.6956231206147678,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004712912334828094,
      "loss": 4.9613,
      "mean_token_accuracy": 0.20336879044771194,
      "num_tokens": 37341580.0,
      "step": 20300
    },
    {
      "entropy": 5.2496514320373535,
      "epoch": 1.6960407617774806,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047127668071021293,
      "loss": 5.017,
      "mean_token_accuracy": 0.20297275483608246,
      "num_tokens": 37351385.0,
      "step": 20305
    },
    {
      "entropy": 5.083111381530761,
      "epoch": 1.696458402940194,
      "grad_norm": 1.328125,
      "learning_rate": 0.000471262124501557,
      "loss": 4.7655,
      "mean_token_accuracy": 0.22087411731481552,
      "num_tokens": 37360893.0,
      "step": 20310
    },
    {
      "entropy": 5.146052408218384,
      "epoch": 1.6968760441029067,
      "grad_norm": 1.375,
      "learning_rate": 0.0004712475648570962,
      "loss": 4.8775,
      "mean_token_accuracy": 0.20894370675086976,
      "num_tokens": 37370301.0,
      "step": 20315
    },
    {
      "entropy": 5.182099962234497,
      "epoch": 1.6972936852656197,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047123300177708563,
      "loss": 4.9903,
      "mean_token_accuracy": 0.2001637950539589,
      "num_tokens": 37380881.0,
      "step": 20320
    },
    {
      "entropy": 5.124613332748413,
      "epoch": 1.6977113264283328,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004712184352617801,
      "loss": 4.8791,
      "mean_token_accuracy": 0.2156565323472023,
      "num_tokens": 37390167.0,
      "step": 20325
    },
    {
      "entropy": 5.268690729141236,
      "epoch": 1.6981289675910458,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047120386531143476,
      "loss": 5.1109,
      "mean_token_accuracy": 0.19192306101322174,
      "num_tokens": 37399163.0,
      "step": 20330
    },
    {
      "entropy": 5.286881017684936,
      "epoch": 1.6985466087537588,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047118929192630453,
      "loss": 5.0341,
      "mean_token_accuracy": 0.20058843791484832,
      "num_tokens": 37409047.0,
      "step": 20335
    },
    {
      "entropy": 5.197458744049072,
      "epoch": 1.6989642499164717,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004711747151066446,
      "loss": 4.8655,
      "mean_token_accuracy": 0.21277865171432495,
      "num_tokens": 37417936.0,
      "step": 20340
    },
    {
      "entropy": 5.0677268505096436,
      "epoch": 1.699381891079185,
      "grad_norm": 1.515625,
      "learning_rate": 0.00047116013485271014,
      "loss": 4.7762,
      "mean_token_accuracy": 0.21594499200582504,
      "num_tokens": 37426453.0,
      "step": 20345
    },
    {
      "entropy": 5.042849063873291,
      "epoch": 1.6997995322418977,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004711455511647564,
      "loss": 4.7412,
      "mean_token_accuracy": 0.22662240862846375,
      "num_tokens": 37435353.0,
      "step": 20350
    },
    {
      "entropy": 5.138701581954956,
      "epoch": 1.7002171734046108,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047113096404303875,
      "loss": 4.9746,
      "mean_token_accuracy": 0.20668933689594268,
      "num_tokens": 37444895.0,
      "step": 20355
    },
    {
      "entropy": 5.202768754959107,
      "epoch": 1.7006348145673238,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004711163734878124,
      "loss": 4.8539,
      "mean_token_accuracy": 0.20772019177675247,
      "num_tokens": 37454038.0,
      "step": 20360
    },
    {
      "entropy": 5.158268547058105,
      "epoch": 1.7010524557300366,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004711017794993329,
      "loss": 4.9465,
      "mean_token_accuracy": 0.20540921241045,
      "num_tokens": 37463316.0,
      "step": 20365
    },
    {
      "entropy": 5.1331957340240475,
      "epoch": 1.7014700968927499,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047108718207785575,
      "loss": 4.9374,
      "mean_token_accuracy": 0.20502015203237534,
      "num_tokens": 37473341.0,
      "step": 20370
    },
    {
      "entropy": 5.2631724834442135,
      "epoch": 1.7018877380554627,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004710725812236364,
      "loss": 4.9653,
      "mean_token_accuracy": 0.21041183471679686,
      "num_tokens": 37482601.0,
      "step": 20375
    },
    {
      "entropy": 5.16471300125122,
      "epoch": 1.7023053792181757,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004710579769369305,
      "loss": 4.8591,
      "mean_token_accuracy": 0.21454714834690095,
      "num_tokens": 37491304.0,
      "step": 20380
    },
    {
      "entropy": 5.063629388809204,
      "epoch": 1.7027230203808887,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004710433692179937,
      "loss": 4.8535,
      "mean_token_accuracy": 0.20766912251710892,
      "num_tokens": 37500848.0,
      "step": 20385
    },
    {
      "entropy": 5.174039506912232,
      "epoch": 1.7031406615436018,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004710287580670818,
      "loss": 4.8348,
      "mean_token_accuracy": 0.21593288332223892,
      "num_tokens": 37509649.0,
      "step": 20390
    },
    {
      "entropy": 5.193998670578003,
      "epoch": 1.7035583027063148,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047101414348445046,
      "loss": 4.9841,
      "mean_token_accuracy": 0.20056151151657103,
      "num_tokens": 37519207.0,
      "step": 20395
    },
    {
      "entropy": 5.238597822189331,
      "epoch": 1.7039759438690276,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047099952547035564,
      "loss": 4.983,
      "mean_token_accuracy": 0.19763025939464568,
      "num_tokens": 37528819.0,
      "step": 20400
    },
    {
      "entropy": 5.210116720199585,
      "epoch": 1.7043935850317409,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004709849040250532,
      "loss": 4.9084,
      "mean_token_accuracy": 0.20521895438432694,
      "num_tokens": 37538439.0,
      "step": 20405
    },
    {
      "entropy": 5.1068840503692625,
      "epoch": 1.7048112261944537,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004709702791487991,
      "loss": 4.8886,
      "mean_token_accuracy": 0.21186998784542083,
      "num_tokens": 37547758.0,
      "step": 20410
    },
    {
      "entropy": 5.143831443786621,
      "epoch": 1.7052288673571667,
      "grad_norm": 2.0,
      "learning_rate": 0.0004709556508418494,
      "loss": 4.9151,
      "mean_token_accuracy": 0.21122766733169557,
      "num_tokens": 37556693.0,
      "step": 20415
    },
    {
      "entropy": 5.159623432159424,
      "epoch": 1.7056465085198798,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004709410191044601,
      "loss": 4.9694,
      "mean_token_accuracy": 0.20892215222120286,
      "num_tokens": 37564631.0,
      "step": 20420
    },
    {
      "entropy": 5.168721437454224,
      "epoch": 1.7060641496825926,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004709263839368875,
      "loss": 4.8318,
      "mean_token_accuracy": 0.21594227701425553,
      "num_tokens": 37573687.0,
      "step": 20425
    },
    {
      "entropy": 5.113115549087524,
      "epoch": 1.7064817908453058,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004709117453393876,
      "loss": 4.9066,
      "mean_token_accuracy": 0.2054399996995926,
      "num_tokens": 37582652.0,
      "step": 20430
    },
    {
      "entropy": 5.07117223739624,
      "epoch": 1.7068994320080186,
      "grad_norm": 1.53125,
      "learning_rate": 0.00047089710331221687,
      "loss": 4.7891,
      "mean_token_accuracy": 0.21822351962327957,
      "num_tokens": 37591242.0,
      "step": 20435
    },
    {
      "entropy": 5.2920609474182125,
      "epoch": 1.7073170731707317,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004708824578556315,
      "loss": 5.1255,
      "mean_token_accuracy": 0.1958068773150444,
      "num_tokens": 37601527.0,
      "step": 20440
    },
    {
      "entropy": 5.267307901382447,
      "epoch": 1.7077347143334447,
      "grad_norm": 1.25,
      "learning_rate": 0.00047086780896988796,
      "loss": 4.9983,
      "mean_token_accuracy": 0.20463395416736602,
      "num_tokens": 37611100.0,
      "step": 20445
    },
    {
      "entropy": 5.176217889785766,
      "epoch": 1.7081523554961577,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00047085315665524257,
      "loss": 4.9228,
      "mean_token_accuracy": 0.21291374415159225,
      "num_tokens": 37619386.0,
      "step": 20450
    },
    {
      "entropy": 5.217793989181518,
      "epoch": 1.7085699966588708,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000470838500911952,
      "loss": 5.0222,
      "mean_token_accuracy": 0.20441450476646422,
      "num_tokens": 37628695.0,
      "step": 20455
    },
    {
      "entropy": 5.25447416305542,
      "epoch": 1.7089876378215836,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004708238417402728,
      "loss": 4.964,
      "mean_token_accuracy": 0.2042546048760414,
      "num_tokens": 37637588.0,
      "step": 20460
    },
    {
      "entropy": 5.154146003723144,
      "epoch": 1.7094052789842968,
      "grad_norm": 3.703125,
      "learning_rate": 0.00047080917914046146,
      "loss": 4.809,
      "mean_token_accuracy": 0.21868397295475006,
      "num_tokens": 37646114.0,
      "step": 20465
    },
    {
      "entropy": 5.2630850315094,
      "epoch": 1.7098229201470097,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047079451311277473,
      "loss": 5.075,
      "mean_token_accuracy": 0.19540415853261947,
      "num_tokens": 37654987.0,
      "step": 20470
    },
    {
      "entropy": 5.169364309310913,
      "epoch": 1.7102405613097227,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004707798436574694,
      "loss": 4.9383,
      "mean_token_accuracy": 0.21131244748830796,
      "num_tokens": 37664715.0,
      "step": 20475
    },
    {
      "entropy": 5.217259740829467,
      "epoch": 1.7106582024724357,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00047076517077480223,
      "loss": 5.0263,
      "mean_token_accuracy": 0.20485977977514266,
      "num_tokens": 37675016.0,
      "step": 20480
    },
    {
      "entropy": 5.249435091018677,
      "epoch": 1.7110758436351485,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004707504944650301,
      "loss": 4.941,
      "mean_token_accuracy": 0.20866553038358687,
      "num_tokens": 37683577.0,
      "step": 20485
    },
    {
      "entropy": 5.25882363319397,
      "epoch": 1.7114934847978618,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047073581472840997,
      "loss": 4.9268,
      "mean_token_accuracy": 0.20102616548538207,
      "num_tokens": 37692681.0,
      "step": 20490
    },
    {
      "entropy": 5.084091424942017,
      "epoch": 1.7119111259605746,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004707211315651988,
      "loss": 4.8578,
      "mean_token_accuracy": 0.21331736892461778,
      "num_tokens": 37701698.0,
      "step": 20495
    },
    {
      "entropy": 5.1129309177398685,
      "epoch": 1.7123287671232876,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004707064449756536,
      "loss": 4.9472,
      "mean_token_accuracy": 0.202505923807621,
      "num_tokens": 37710450.0,
      "step": 20500
    },
    {
      "entropy": 5.1810835838317875,
      "epoch": 1.7127464082860007,
      "grad_norm": 1.703125,
      "learning_rate": 0.00047069175496003147,
      "loss": 4.9669,
      "mean_token_accuracy": 0.20894842445850373,
      "num_tokens": 37719201.0,
      "step": 20505
    },
    {
      "entropy": 5.296409463882446,
      "epoch": 1.7131640494487137,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047067706151858967,
      "loss": 4.9917,
      "mean_token_accuracy": 0.20837360322475434,
      "num_tokens": 37728671.0,
      "step": 20510
    },
    {
      "entropy": 5.198051500320434,
      "epoch": 1.7135816906114267,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004706623646515854,
      "loss": 4.8915,
      "mean_token_accuracy": 0.20982447564601897,
      "num_tokens": 37737675.0,
      "step": 20515
    },
    {
      "entropy": 5.108592414855957,
      "epoch": 1.7139993317741395,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047064766435927585,
      "loss": 4.9154,
      "mean_token_accuracy": 0.2171660214662552,
      "num_tokens": 37746299.0,
      "step": 20520
    },
    {
      "entropy": 5.1394799709320065,
      "epoch": 1.7144169729368528,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00047063296064191846,
      "loss": 4.8004,
      "mean_token_accuracy": 0.20718855261802674,
      "num_tokens": 37755770.0,
      "step": 20525
    },
    {
      "entropy": 5.137127447128296,
      "epoch": 1.7148346140995656,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00047061825349977067,
      "loss": 4.9452,
      "mean_token_accuracy": 0.20878740549087524,
      "num_tokens": 37765217.0,
      "step": 20530
    },
    {
      "entropy": 5.168634128570557,
      "epoch": 1.7152522552622786,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00047060354293308976,
      "loss": 4.9194,
      "mean_token_accuracy": 0.21355330646038057,
      "num_tokens": 37773891.0,
      "step": 20535
    },
    {
      "entropy": 5.173954153060913,
      "epoch": 1.7156698964249917,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004705888289421335,
      "loss": 4.9315,
      "mean_token_accuracy": 0.20872323215007782,
      "num_tokens": 37783323.0,
      "step": 20540
    },
    {
      "entropy": 5.248117399215698,
      "epoch": 1.7160875375877045,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004705741115271593,
      "loss": 5.0155,
      "mean_token_accuracy": 0.20095303952693938,
      "num_tokens": 37792436.0,
      "step": 20545
    },
    {
      "entropy": 5.182192659378051,
      "epoch": 1.7165051787504177,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004705593906884249,
      "loss": 4.9154,
      "mean_token_accuracy": 0.20771195441484452,
      "num_tokens": 37802061.0,
      "step": 20550
    },
    {
      "entropy": 5.129126787185669,
      "epoch": 1.7169228199131306,
      "grad_norm": 1.484375,
      "learning_rate": 0.00047054466642618793,
      "loss": 4.8433,
      "mean_token_accuracy": 0.21134544610977174,
      "num_tokens": 37811101.0,
      "step": 20555
    },
    {
      "entropy": 5.241861248016358,
      "epoch": 1.7173404610758436,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047052993874070623,
      "loss": 4.9964,
      "mean_token_accuracy": 0.20834031999111174,
      "num_tokens": 37820783.0,
      "step": 20560
    },
    {
      "entropy": 5.153719043731689,
      "epoch": 1.7177581022385566,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047051520763223754,
      "loss": 4.8916,
      "mean_token_accuracy": 0.20406176745891572,
      "num_tokens": 37830075.0,
      "step": 20565
    },
    {
      "entropy": 5.059848499298096,
      "epoch": 1.7181757434012697,
      "grad_norm": 2.234375,
      "learning_rate": 0.00047050047310103984,
      "loss": 4.7922,
      "mean_token_accuracy": 0.21144964247941972,
      "num_tokens": 37838344.0,
      "step": 20570
    },
    {
      "entropy": 5.120484924316406,
      "epoch": 1.7185933845639827,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00047048573514737107,
      "loss": 4.9064,
      "mean_token_accuracy": 0.20225743800401688,
      "num_tokens": 37847911.0,
      "step": 20575
    },
    {
      "entropy": 5.095972299575806,
      "epoch": 1.7190110257266955,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00047047099377148916,
      "loss": 4.8415,
      "mean_token_accuracy": 0.21828738301992417,
      "num_tokens": 37857044.0,
      "step": 20580
    },
    {
      "entropy": 5.235614871978759,
      "epoch": 1.7194286668894088,
      "grad_norm": 1.984375,
      "learning_rate": 0.0004704562489736522,
      "loss": 4.928,
      "mean_token_accuracy": 0.20942811369895936,
      "num_tokens": 37865912.0,
      "step": 20585
    },
    {
      "entropy": 5.221312761306763,
      "epoch": 1.7198463080521216,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004704415007541184,
      "loss": 5.0031,
      "mean_token_accuracy": 0.20476872026920317,
      "num_tokens": 37875118.0,
      "step": 20590
    },
    {
      "entropy": 5.1713385581970215,
      "epoch": 1.7202639492148346,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004704267491131458,
      "loss": 4.8905,
      "mean_token_accuracy": 0.2132647901773453,
      "num_tokens": 37882891.0,
      "step": 20595
    },
    {
      "entropy": 5.128694534301758,
      "epoch": 1.7206815903775476,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00047041199405099273,
      "loss": 4.9176,
      "mean_token_accuracy": 0.21680514067411422,
      "num_tokens": 37892274.0,
      "step": 20600
    },
    {
      "entropy": 5.104113149642944,
      "epoch": 1.7210992315402605,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047039723556791744,
      "loss": 4.7996,
      "mean_token_accuracy": 0.22019950151443482,
      "num_tokens": 37901061.0,
      "step": 20605
    },
    {
      "entropy": 5.145769214630127,
      "epoch": 1.7215168727029737,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004703824736641785,
      "loss": 4.7896,
      "mean_token_accuracy": 0.21267174929380417,
      "num_tokens": 37910253.0,
      "step": 20610
    },
    {
      "entropy": 5.1598076820373535,
      "epoch": 1.7219345138656865,
      "grad_norm": 1.6875,
      "learning_rate": 0.00047036770834003405,
      "loss": 4.921,
      "mean_token_accuracy": 0.21076069325208663,
      "num_tokens": 37919453.0,
      "step": 20615
    },
    {
      "entropy": 5.094369554519654,
      "epoch": 1.7223521550283996,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00047035293959574273,
      "loss": 4.8031,
      "mean_token_accuracy": 0.21682598143815995,
      "num_tokens": 37928462.0,
      "step": 20620
    },
    {
      "entropy": 5.189353179931641,
      "epoch": 1.7227697961911126,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000470338167431563,
      "loss": 4.9314,
      "mean_token_accuracy": 0.2130808025598526,
      "num_tokens": 37938631.0,
      "step": 20625
    },
    {
      "entropy": 5.275042533874512,
      "epoch": 1.7231874373538256,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00047032339184775366,
      "loss": 4.9755,
      "mean_token_accuracy": 0.20193572491407394,
      "num_tokens": 37948490.0,
      "step": 20630
    },
    {
      "entropy": 5.045252180099487,
      "epoch": 1.7236050785165387,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004703086128445731,
      "loss": 4.7992,
      "mean_token_accuracy": 0.22184834033250808,
      "num_tokens": 37957861.0,
      "step": 20635
    },
    {
      "entropy": 5.098060035705567,
      "epoch": 1.7240227196792515,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004702938304222803,
      "loss": 4.8955,
      "mean_token_accuracy": 0.21130026876926422,
      "num_tokens": 37966587.0,
      "step": 20640
    },
    {
      "entropy": 5.182022047042847,
      "epoch": 1.7244403608419647,
      "grad_norm": 1.46875,
      "learning_rate": 0.00047027904458113386,
      "loss": 4.9522,
      "mean_token_accuracy": 0.20536856353282928,
      "num_tokens": 37976295.0,
      "step": 20645
    },
    {
      "entropy": 5.300833654403687,
      "epoch": 1.7248580020046775,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004702642553213927,
      "loss": 5.1563,
      "mean_token_accuracy": 0.1928934171795845,
      "num_tokens": 37985525.0,
      "step": 20650
    },
    {
      "entropy": 5.335783863067627,
      "epoch": 1.7252756431673906,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047024946264331564,
      "loss": 4.9885,
      "mean_token_accuracy": 0.1995474860072136,
      "num_tokens": 37995656.0,
      "step": 20655
    },
    {
      "entropy": 5.189170598983765,
      "epoch": 1.7256932843301036,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004702346665471618,
      "loss": 4.9272,
      "mean_token_accuracy": 0.2070934370160103,
      "num_tokens": 38004790.0,
      "step": 20660
    },
    {
      "entropy": 5.224855232238769,
      "epoch": 1.7261109254928164,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004702198670331901,
      "loss": 5.0627,
      "mean_token_accuracy": 0.19827278554439545,
      "num_tokens": 38014244.0,
      "step": 20665
    },
    {
      "entropy": 5.142083549499512,
      "epoch": 1.7265285666555297,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00047020506410165965,
      "loss": 4.912,
      "mean_token_accuracy": 0.2139233887195587,
      "num_tokens": 38023756.0,
      "step": 20670
    },
    {
      "entropy": 5.119731426239014,
      "epoch": 1.7269462078182425,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047019025775282955,
      "loss": 4.8669,
      "mean_token_accuracy": 0.21270580291748048,
      "num_tokens": 38033402.0,
      "step": 20675
    },
    {
      "entropy": 5.198072528839111,
      "epoch": 1.7273638489809555,
      "grad_norm": 1.328125,
      "learning_rate": 0.000470175447986959,
      "loss": 4.9212,
      "mean_token_accuracy": 0.20417163223028184,
      "num_tokens": 38042580.0,
      "step": 20680
    },
    {
      "entropy": 5.2142902374267575,
      "epoch": 1.7277814901436686,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004701606348043074,
      "loss": 5.0173,
      "mean_token_accuracy": 0.20417283475399017,
      "num_tokens": 38051809.0,
      "step": 20685
    },
    {
      "entropy": 5.174270391464233,
      "epoch": 1.7281991313063816,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004701458182051339,
      "loss": 4.8153,
      "mean_token_accuracy": 0.21492623835802077,
      "num_tokens": 38061053.0,
      "step": 20690
    },
    {
      "entropy": 5.108365154266357,
      "epoch": 1.7286167724690946,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047013099818969797,
      "loss": 4.8198,
      "mean_token_accuracy": 0.20815867483615874,
      "num_tokens": 38070190.0,
      "step": 20695
    },
    {
      "entropy": 5.138682889938354,
      "epoch": 1.7290344136318074,
      "grad_norm": 1.359375,
      "learning_rate": 0.000470116174758259,
      "loss": 5.0044,
      "mean_token_accuracy": 0.2047281578183174,
      "num_tokens": 38078049.0,
      "step": 20700
    },
    {
      "entropy": 5.264612913131714,
      "epoch": 1.7294520547945207,
      "grad_norm": 1.796875,
      "learning_rate": 0.00047010134791107663,
      "loss": 4.9487,
      "mean_token_accuracy": 0.20426188707351683,
      "num_tokens": 38089130.0,
      "step": 20705
    },
    {
      "entropy": 5.192202854156494,
      "epoch": 1.7298696959572335,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047008651764841034,
      "loss": 4.8428,
      "mean_token_accuracy": 0.21003013104200363,
      "num_tokens": 38098484.0,
      "step": 20710
    },
    {
      "entropy": 5.18324384689331,
      "epoch": 1.7302873371199465,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047007168397051956,
      "loss": 4.9255,
      "mean_token_accuracy": 0.21641747653484344,
      "num_tokens": 38107982.0,
      "step": 20715
    },
    {
      "entropy": 5.172736740112304,
      "epoch": 1.7307049782826596,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047005684687766424,
      "loss": 4.924,
      "mean_token_accuracy": 0.21194266825914382,
      "num_tokens": 38116816.0,
      "step": 20720
    },
    {
      "entropy": 5.128597164154053,
      "epoch": 1.7311226194453724,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004700420063701041,
      "loss": 4.8529,
      "mean_token_accuracy": 0.21590989530086518,
      "num_tokens": 38125752.0,
      "step": 20725
    },
    {
      "entropy": 5.159169483184814,
      "epoch": 1.7315402606080856,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004700271624480988,
      "loss": 4.9505,
      "mean_token_accuracy": 0.21118825525045395,
      "num_tokens": 38135192.0,
      "step": 20730
    },
    {
      "entropy": 5.246472263336182,
      "epoch": 1.7319579017707984,
      "grad_norm": 1.375,
      "learning_rate": 0.00047001231511190826,
      "loss": 4.9683,
      "mean_token_accuracy": 0.1992129996418953,
      "num_tokens": 38144730.0,
      "step": 20735
    },
    {
      "entropy": 5.107661342620849,
      "epoch": 1.7323755429335115,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004699974643617924,
      "loss": 4.8435,
      "mean_token_accuracy": 0.21439514607191085,
      "num_tokens": 38153319.0,
      "step": 20740
    },
    {
      "entropy": 5.137152671813965,
      "epoch": 1.7327931840962245,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004699826101980112,
      "loss": 4.8783,
      "mean_token_accuracy": 0.21171682626008986,
      "num_tokens": 38162067.0,
      "step": 20745
    },
    {
      "entropy": 5.258701467514038,
      "epoch": 1.7332108252589375,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004699677526208248,
      "loss": 5.0329,
      "mean_token_accuracy": 0.20030689239501953,
      "num_tokens": 38170901.0,
      "step": 20750
    },
    {
      "entropy": 5.1413161277771,
      "epoch": 1.7336284664216506,
      "grad_norm": 1.359375,
      "learning_rate": 0.00046995289163049317,
      "loss": 4.9311,
      "mean_token_accuracy": 0.206833553314209,
      "num_tokens": 38179815.0,
      "step": 20755
    },
    {
      "entropy": 5.2077795505523685,
      "epoch": 1.7340461075843634,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046993802722727646,
      "loss": 4.9387,
      "mean_token_accuracy": 0.20938231348991393,
      "num_tokens": 38188816.0,
      "step": 20760
    },
    {
      "entropy": 5.186800527572632,
      "epoch": 1.7344637487470767,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004699231594114349,
      "loss": 4.8852,
      "mean_token_accuracy": 0.2071742296218872,
      "num_tokens": 38198447.0,
      "step": 20765
    },
    {
      "entropy": 5.251977920532227,
      "epoch": 1.7348813899097895,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004699082881832289,
      "loss": 4.9766,
      "mean_token_accuracy": 0.20436772853136062,
      "num_tokens": 38207075.0,
      "step": 20770
    },
    {
      "entropy": 5.199755430221558,
      "epoch": 1.7352990310725025,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046989341354291864,
      "loss": 4.9643,
      "mean_token_accuracy": 0.19833430647850037,
      "num_tokens": 38216297.0,
      "step": 20775
    },
    {
      "entropy": 5.310128688812256,
      "epoch": 1.7357166722352155,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004698785354907646,
      "loss": 5.0738,
      "mean_token_accuracy": 0.1966646984219551,
      "num_tokens": 38225581.0,
      "step": 20780
    },
    {
      "entropy": 5.142106914520264,
      "epoch": 1.7361343133979283,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004698636540270272,
      "loss": 4.8063,
      "mean_token_accuracy": 0.21949632316827775,
      "num_tokens": 38234214.0,
      "step": 20785
    },
    {
      "entropy": 5.235606241226196,
      "epoch": 1.7365519545606416,
      "grad_norm": 1.25,
      "learning_rate": 0.0004698487691519671,
      "loss": 5.022,
      "mean_token_accuracy": 0.19622836261987686,
      "num_tokens": 38243859.0,
      "step": 20790
    },
    {
      "entropy": 5.200755214691162,
      "epoch": 1.7369695957233544,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00046983388086584463,
      "loss": 4.8417,
      "mean_token_accuracy": 0.211751352250576,
      "num_tokens": 38252667.0,
      "step": 20795
    },
    {
      "entropy": 5.040313625335694,
      "epoch": 1.7373872368860674,
      "grad_norm": 1.375,
      "learning_rate": 0.00046981898916892054,
      "loss": 4.7816,
      "mean_token_accuracy": 0.21426222026348113,
      "num_tokens": 38262216.0,
      "step": 20800
    },
    {
      "entropy": 5.158411121368408,
      "epoch": 1.7378048780487805,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00046980409406145555,
      "loss": 4.9517,
      "mean_token_accuracy": 0.2045581728219986,
      "num_tokens": 38271672.0,
      "step": 20805
    },
    {
      "entropy": 5.085760593414307,
      "epoch": 1.7382225192114935,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004697891955437105,
      "loss": 4.8415,
      "mean_token_accuracy": 0.22146798372268678,
      "num_tokens": 38279907.0,
      "step": 20810
    },
    {
      "entropy": 5.134995317459106,
      "epoch": 1.7386401603742065,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00046977429361594596,
      "loss": 4.8522,
      "mean_token_accuracy": 0.21601584702730178,
      "num_tokens": 38289099.0,
      "step": 20815
    },
    {
      "entropy": 5.0701751708984375,
      "epoch": 1.7390578015369194,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004697593882784231,
      "loss": 4.8561,
      "mean_token_accuracy": 0.21134332418441773,
      "num_tokens": 38297754.0,
      "step": 20820
    },
    {
      "entropy": 5.210834646224976,
      "epoch": 1.7394754426996326,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004697444795314027,
      "loss": 4.9097,
      "mean_token_accuracy": 0.21009967625141143,
      "num_tokens": 38306429.0,
      "step": 20825
    },
    {
      "entropy": 5.2233319759368895,
      "epoch": 1.7398930838623454,
      "grad_norm": 1.359375,
      "learning_rate": 0.00046972956737514573,
      "loss": 4.9325,
      "mean_token_accuracy": 0.20812920928001405,
      "num_tokens": 38315505.0,
      "step": 20830
    },
    {
      "entropy": 5.156542587280273,
      "epoch": 1.7403107250250585,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004697146518099133,
      "loss": 4.9492,
      "mean_token_accuracy": 0.20844421535730362,
      "num_tokens": 38324439.0,
      "step": 20835
    },
    {
      "entropy": 5.200738525390625,
      "epoch": 1.7407283661877715,
      "grad_norm": 1.328125,
      "learning_rate": 0.00046969973283596657,
      "loss": 4.9002,
      "mean_token_accuracy": 0.2121814399957657,
      "num_tokens": 38334145.0,
      "step": 20840
    },
    {
      "entropy": 5.180101346969605,
      "epoch": 1.7411460073504843,
      "grad_norm": 1.53125,
      "learning_rate": 0.00046968481045356656,
      "loss": 4.9241,
      "mean_token_accuracy": 0.210992793738842,
      "num_tokens": 38342811.0,
      "step": 20845
    },
    {
      "entropy": 5.121627950668335,
      "epoch": 1.7415636485131976,
      "grad_norm": 3.03125,
      "learning_rate": 0.00046966988466297464,
      "loss": 4.9025,
      "mean_token_accuracy": 0.21210924088954924,
      "num_tokens": 38351072.0,
      "step": 20850
    },
    {
      "entropy": 5.179961729049682,
      "epoch": 1.7419812896759104,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00046965495546445204,
      "loss": 4.8685,
      "mean_token_accuracy": 0.21614393591880798,
      "num_tokens": 38360127.0,
      "step": 20855
    },
    {
      "entropy": 5.153711128234863,
      "epoch": 1.7423989308386234,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00046964002285826016,
      "loss": 4.8824,
      "mean_token_accuracy": 0.20009496212005615,
      "num_tokens": 38368270.0,
      "step": 20860
    },
    {
      "entropy": 5.172607040405273,
      "epoch": 1.7428165720013364,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004696250868446604,
      "loss": 4.9574,
      "mean_token_accuracy": 0.20273905545473098,
      "num_tokens": 38376970.0,
      "step": 20865
    },
    {
      "entropy": 5.1857551574707035,
      "epoch": 1.7432342131640495,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046961014742391425,
      "loss": 4.9881,
      "mean_token_accuracy": 0.20191019922494888,
      "num_tokens": 38387103.0,
      "step": 20870
    },
    {
      "entropy": 5.274353981018066,
      "epoch": 1.7436518543267625,
      "grad_norm": 1.375,
      "learning_rate": 0.00046959520459628317,
      "loss": 4.9322,
      "mean_token_accuracy": 0.20863609611988068,
      "num_tokens": 38395890.0,
      "step": 20875
    },
    {
      "entropy": 5.213463306427002,
      "epoch": 1.7440694954894753,
      "grad_norm": 1.4375,
      "learning_rate": 0.00046958025836202886,
      "loss": 5.0156,
      "mean_token_accuracy": 0.20192147344350814,
      "num_tokens": 38405868.0,
      "step": 20880
    },
    {
      "entropy": 5.160831308364868,
      "epoch": 1.7444871366521886,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004695653087214128,
      "loss": 5.0145,
      "mean_token_accuracy": 0.20280708372592926,
      "num_tokens": 38414854.0,
      "step": 20885
    },
    {
      "entropy": 5.169708156585694,
      "epoch": 1.7449047778149014,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046955035567469695,
      "loss": 4.8517,
      "mean_token_accuracy": 0.20817867219448088,
      "num_tokens": 38424158.0,
      "step": 20890
    },
    {
      "entropy": 5.245330333709717,
      "epoch": 1.7453224189776144,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046953539922214285,
      "loss": 4.9297,
      "mean_token_accuracy": 0.20381746143102647,
      "num_tokens": 38433200.0,
      "step": 20895
    },
    {
      "entropy": 5.200853109359741,
      "epoch": 1.7457400601403275,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004695204393640125,
      "loss": 4.8947,
      "mean_token_accuracy": 0.20379038006067277,
      "num_tokens": 38442183.0,
      "step": 20900
    },
    {
      "entropy": 5.134096002578735,
      "epoch": 1.7461577013030403,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046950547610056766,
      "loss": 4.8827,
      "mean_token_accuracy": 0.2052141547203064,
      "num_tokens": 38451451.0,
      "step": 20905
    },
    {
      "entropy": 5.202793025970459,
      "epoch": 1.7465753424657535,
      "grad_norm": 1.25,
      "learning_rate": 0.0004694905094320703,
      "loss": 4.9423,
      "mean_token_accuracy": 0.20757792443037032,
      "num_tokens": 38460043.0,
      "step": 20910
    },
    {
      "entropy": 5.200147533416748,
      "epoch": 1.7469929836284663,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046947553935878253,
      "loss": 4.9634,
      "mean_token_accuracy": 0.20355763882398606,
      "num_tokens": 38470714.0,
      "step": 20915
    },
    {
      "entropy": 5.221422481536865,
      "epoch": 1.7474106247911794,
      "grad_norm": 1.5,
      "learning_rate": 0.0004694605658809663,
      "loss": 4.9514,
      "mean_token_accuracy": 0.20843559503555298,
      "num_tokens": 38480528.0,
      "step": 20920
    },
    {
      "entropy": 5.129616641998291,
      "epoch": 1.7478282659538924,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004694455889988838,
      "loss": 4.8498,
      "mean_token_accuracy": 0.20705964416265488,
      "num_tokens": 38490101.0,
      "step": 20925
    },
    {
      "entropy": 5.214071750640869,
      "epoch": 1.7482459071166054,
      "grad_norm": 1.40625,
      "learning_rate": 0.00046943060871279725,
      "loss": 5.0416,
      "mean_token_accuracy": 0.1991410195827484,
      "num_tokens": 38498624.0,
      "step": 20930
    },
    {
      "entropy": 5.143876934051514,
      "epoch": 1.7486635482793185,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00046941562502296886,
      "loss": 4.8208,
      "mean_token_accuracy": 0.21500442773103715,
      "num_tokens": 38506646.0,
      "step": 20935
    },
    {
      "entropy": 5.116863346099853,
      "epoch": 1.7490811894420313,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004694006379296609,
      "loss": 4.8611,
      "mean_token_accuracy": 0.2125670701265335,
      "num_tokens": 38516092.0,
      "step": 20940
    },
    {
      "entropy": 5.111257314682007,
      "epoch": 1.7494988306047445,
      "grad_norm": 1.25,
      "learning_rate": 0.00046938564743313576,
      "loss": 4.8555,
      "mean_token_accuracy": 0.20599322617053986,
      "num_tokens": 38526486.0,
      "step": 20945
    },
    {
      "entropy": 5.150310134887695,
      "epoch": 1.7499164717674573,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00046937065353365594,
      "loss": 4.899,
      "mean_token_accuracy": 0.20603618770837784,
      "num_tokens": 38535474.0,
      "step": 20950
    },
    {
      "entropy": 5.205947065353394,
      "epoch": 1.7503341129301704,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00046935565623148383,
      "loss": 4.94,
      "mean_token_accuracy": 0.20744649469852447,
      "num_tokens": 38544598.0,
      "step": 20955
    },
    {
      "entropy": 5.170606470108032,
      "epoch": 1.7507517540928834,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00046934065552688196,
      "loss": 4.9661,
      "mean_token_accuracy": 0.21013639271259307,
      "num_tokens": 38553569.0,
      "step": 20960
    },
    {
      "entropy": 5.236491632461548,
      "epoch": 1.7511693952555962,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004693256514201131,
      "loss": 4.9895,
      "mean_token_accuracy": 0.21254801601171494,
      "num_tokens": 38563369.0,
      "step": 20965
    },
    {
      "entropy": 5.245092153549194,
      "epoch": 1.7515870364183095,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046931064391143976,
      "loss": 4.9433,
      "mean_token_accuracy": 0.20296305567026138,
      "num_tokens": 38573187.0,
      "step": 20970
    },
    {
      "entropy": 5.129215478897095,
      "epoch": 1.7520046775810223,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046929563300112466,
      "loss": 4.7797,
      "mean_token_accuracy": 0.23131679892539977,
      "num_tokens": 38581014.0,
      "step": 20975
    },
    {
      "entropy": 5.0871435642242435,
      "epoch": 1.7524223187437353,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046928061868943067,
      "loss": 4.8592,
      "mean_token_accuracy": 0.20607406497001649,
      "num_tokens": 38590445.0,
      "step": 20980
    },
    {
      "entropy": 5.110336685180664,
      "epoch": 1.7528399599064484,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004692656009766206,
      "loss": 4.8509,
      "mean_token_accuracy": 0.21550438553094864,
      "num_tokens": 38599372.0,
      "step": 20985
    },
    {
      "entropy": 5.27717604637146,
      "epoch": 1.7532576010691614,
      "grad_norm": 1.296875,
      "learning_rate": 0.00046925057986295734,
      "loss": 5.0566,
      "mean_token_accuracy": 0.19833142161369324,
      "num_tokens": 38609000.0,
      "step": 20990
    },
    {
      "entropy": 5.194234561920166,
      "epoch": 1.7536752422318744,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046923555534870385,
      "loss": 4.9045,
      "mean_token_accuracy": 0.20556726157665253,
      "num_tokens": 38618420.0,
      "step": 20995
    },
    {
      "entropy": 5.094934511184692,
      "epoch": 1.7540928833945872,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004692205274341232,
      "loss": 4.814,
      "mean_token_accuracy": 0.223762808740139,
      "num_tokens": 38627577.0,
      "step": 21000
    },
    {
      "epoch": 1.7540928833945872,
      "eval_entropy": 4.948788805347306,
      "eval_loss": 4.937788009643555,
      "eval_mean_token_accuracy": 0.21806599070203428,
      "eval_num_tokens": 38627577.0,
      "eval_runtime": 27.5368,
      "eval_samples_per_second": 1371.367,
      "eval_steps_per_second": 171.444,
      "step": 21000
    }
  ],
  "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": 5.7181744355328e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}