hin-deva-10mb-after-ppt-Dp-100mb-ckpt500_seed455/checkpoint-6000/trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.5763688760806917,
  "eval_steps": 3000,
  "global_step": 6000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 7.227891731262207,
      "epoch": 0.0004803073967339097,
      "grad_norm": 1.140625,
      "learning_rate": 2e-06,
      "loss": 6.7116,
      "mean_token_accuracy": 0.12639277800917625,
      "num_tokens": 10821.0,
      "step": 5
    },
    {
      "entropy": 7.243266201019287,
      "epoch": 0.0009606147934678194,
      "grad_norm": 1.1640625,
      "learning_rate": 4.5e-06,
      "loss": 6.6481,
      "mean_token_accuracy": 0.12706964686512948,
      "num_tokens": 21838.0,
      "step": 10
    },
    {
      "entropy": 7.221618175506592,
      "epoch": 0.001440922190201729,
      "grad_norm": 1.0390625,
      "learning_rate": 7e-06,
      "loss": 6.7012,
      "mean_token_accuracy": 0.12022195979952813,
      "num_tokens": 33483.0,
      "step": 15
    },
    {
      "entropy": 7.250268363952637,
      "epoch": 0.0019212295869356388,
      "grad_norm": 1.0859375,
      "learning_rate": 9.5e-06,
      "loss": 6.7805,
      "mean_token_accuracy": 0.12316789031028748,
      "num_tokens": 45150.0,
      "step": 20
    },
    {
      "entropy": 7.207002735137939,
      "epoch": 0.0024015369836695487,
      "grad_norm": 1.109375,
      "learning_rate": 1.2e-05,
      "loss": 6.6792,
      "mean_token_accuracy": 0.11827812790870666,
      "num_tokens": 55862.0,
      "step": 25
    },
    {
      "entropy": 7.20290994644165,
      "epoch": 0.002881844380403458,
      "grad_norm": 0.9765625,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 6.8032,
      "mean_token_accuracy": 0.12267392575740814,
      "num_tokens": 66766.0,
      "step": 30
    },
    {
      "entropy": 7.144218921661377,
      "epoch": 0.0033621517771373678,
      "grad_norm": 1.125,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 6.5389,
      "mean_token_accuracy": 0.13368295580148698,
      "num_tokens": 77012.0,
      "step": 35
    },
    {
      "entropy": 7.158500909805298,
      "epoch": 0.0038424591738712775,
      "grad_norm": 0.9765625,
      "learning_rate": 1.95e-05,
      "loss": 6.7614,
      "mean_token_accuracy": 0.12391667440533638,
      "num_tokens": 89034.0,
      "step": 40
    },
    {
      "entropy": 7.117870426177978,
      "epoch": 0.004322766570605188,
      "grad_norm": 0.98828125,
      "learning_rate": 2.2e-05,
      "loss": 6.6403,
      "mean_token_accuracy": 0.12876288816332818,
      "num_tokens": 100341.0,
      "step": 45
    },
    {
      "entropy": 7.142163562774658,
      "epoch": 0.004803073967339097,
      "grad_norm": 1.015625,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 6.6103,
      "mean_token_accuracy": 0.12874003127217293,
      "num_tokens": 111997.0,
      "step": 50
    },
    {
      "entropy": 7.155911827087403,
      "epoch": 0.005283381364073006,
      "grad_norm": 1.0078125,
      "learning_rate": 2.7e-05,
      "loss": 6.6601,
      "mean_token_accuracy": 0.1270666964352131,
      "num_tokens": 123705.0,
      "step": 55
    },
    {
      "entropy": 7.139102792739868,
      "epoch": 0.005763688760806916,
      "grad_norm": 1.0625,
      "learning_rate": 2.95e-05,
      "loss": 6.5978,
      "mean_token_accuracy": 0.13043108433485032,
      "num_tokens": 134091.0,
      "step": 60
    },
    {
      "entropy": 7.0543046474456785,
      "epoch": 0.006243996157540826,
      "grad_norm": 0.99609375,
      "learning_rate": 3.2e-05,
      "loss": 6.7086,
      "mean_token_accuracy": 0.1285567693412304,
      "num_tokens": 145731.0,
      "step": 65
    },
    {
      "entropy": 7.078891086578369,
      "epoch": 0.0067243035542747355,
      "grad_norm": 1.0859375,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 6.6207,
      "mean_token_accuracy": 0.1347151793539524,
      "num_tokens": 156979.0,
      "step": 70
    },
    {
      "entropy": 7.112874746322632,
      "epoch": 0.007204610951008645,
      "grad_norm": 1.1328125,
      "learning_rate": 3.7e-05,
      "loss": 6.6131,
      "mean_token_accuracy": 0.1301887921988964,
      "num_tokens": 167021.0,
      "step": 75
    },
    {
      "entropy": 7.118173456192016,
      "epoch": 0.007684918347742555,
      "grad_norm": 1.125,
      "learning_rate": 3.95e-05,
      "loss": 6.6258,
      "mean_token_accuracy": 0.1291635274887085,
      "num_tokens": 179269.0,
      "step": 80
    },
    {
      "entropy": 7.088387060165405,
      "epoch": 0.008165225744476465,
      "grad_norm": 0.9765625,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 6.6018,
      "mean_token_accuracy": 0.13118657022714614,
      "num_tokens": 190328.0,
      "step": 85
    },
    {
      "entropy": 7.0138716220855715,
      "epoch": 0.008645533141210375,
      "grad_norm": 1.140625,
      "learning_rate": 4.45e-05,
      "loss": 6.4816,
      "mean_token_accuracy": 0.13541511297225953,
      "num_tokens": 201522.0,
      "step": 90
    },
    {
      "entropy": 7.05142731666565,
      "epoch": 0.009125840537944284,
      "grad_norm": 0.91015625,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 6.5891,
      "mean_token_accuracy": 0.13322217017412186,
      "num_tokens": 212963.0,
      "step": 95
    },
    {
      "entropy": 7.128700685501099,
      "epoch": 0.009606147934678195,
      "grad_norm": 0.984375,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 6.8104,
      "mean_token_accuracy": 0.11415100917220115,
      "num_tokens": 224446.0,
      "step": 100
    },
    {
      "entropy": 7.176903247833252,
      "epoch": 0.010086455331412104,
      "grad_norm": 1.1484375,
      "learning_rate": 5.2e-05,
      "loss": 6.6453,
      "mean_token_accuracy": 0.13400980979204177,
      "num_tokens": 235334.0,
      "step": 105
    },
    {
      "entropy": 7.227095699310302,
      "epoch": 0.010566762728146013,
      "grad_norm": 0.96875,
      "learning_rate": 5.45e-05,
      "loss": 6.712,
      "mean_token_accuracy": 0.11989362239837646,
      "num_tokens": 247024.0,
      "step": 110
    },
    {
      "entropy": 7.07423882484436,
      "epoch": 0.011047070124879923,
      "grad_norm": 1.0859375,
      "learning_rate": 5.7e-05,
      "loss": 6.6255,
      "mean_token_accuracy": 0.1259888932108879,
      "num_tokens": 258354.0,
      "step": 115
    },
    {
      "entropy": 6.989333200454712,
      "epoch": 0.011527377521613832,
      "grad_norm": 0.97265625,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 6.6793,
      "mean_token_accuracy": 0.12834338322281838,
      "num_tokens": 271202.0,
      "step": 120
    },
    {
      "entropy": 7.106875991821289,
      "epoch": 0.012007684918347743,
      "grad_norm": 1.1640625,
      "learning_rate": 6.2e-05,
      "loss": 6.6777,
      "mean_token_accuracy": 0.12354658171534538,
      "num_tokens": 282405.0,
      "step": 125
    },
    {
      "entropy": 7.062946128845215,
      "epoch": 0.012487992315081652,
      "grad_norm": 0.9609375,
      "learning_rate": 6.450000000000001e-05,
      "loss": 6.6353,
      "mean_token_accuracy": 0.12684911414980887,
      "num_tokens": 294211.0,
      "step": 130
    },
    {
      "entropy": 7.124184036254883,
      "epoch": 0.012968299711815562,
      "grad_norm": 1.1171875,
      "learning_rate": 6.7e-05,
      "loss": 6.6772,
      "mean_token_accuracy": 0.11896728649735451,
      "num_tokens": 305420.0,
      "step": 135
    },
    {
      "entropy": 7.085195732116699,
      "epoch": 0.013448607108549471,
      "grad_norm": 1.0078125,
      "learning_rate": 6.950000000000001e-05,
      "loss": 6.5818,
      "mean_token_accuracy": 0.1283716782927513,
      "num_tokens": 316430.0,
      "step": 140
    },
    {
      "entropy": 7.063561677932739,
      "epoch": 0.013928914505283382,
      "grad_norm": 1.0234375,
      "learning_rate": 7.2e-05,
      "loss": 6.6082,
      "mean_token_accuracy": 0.13248678371310235,
      "num_tokens": 328406.0,
      "step": 145
    },
    {
      "entropy": 7.072713708877563,
      "epoch": 0.01440922190201729,
      "grad_norm": 1.0703125,
      "learning_rate": 7.45e-05,
      "loss": 6.5784,
      "mean_token_accuracy": 0.13202109262347222,
      "num_tokens": 339051.0,
      "step": 150
    },
    {
      "entropy": 7.053579998016358,
      "epoch": 0.014889529298751201,
      "grad_norm": 1.0390625,
      "learning_rate": 7.7e-05,
      "loss": 6.5612,
      "mean_token_accuracy": 0.13085259720683098,
      "num_tokens": 350308.0,
      "step": 155
    },
    {
      "entropy": 7.130250310897827,
      "epoch": 0.01536983669548511,
      "grad_norm": 1.015625,
      "learning_rate": 7.950000000000001e-05,
      "loss": 6.5937,
      "mean_token_accuracy": 0.12741485759615898,
      "num_tokens": 362159.0,
      "step": 160
    },
    {
      "entropy": 7.091244792938232,
      "epoch": 0.01585014409221902,
      "grad_norm": 1.1171875,
      "learning_rate": 8.2e-05,
      "loss": 6.6467,
      "mean_token_accuracy": 0.12987646758556365,
      "num_tokens": 375157.0,
      "step": 165
    },
    {
      "entropy": 7.02424430847168,
      "epoch": 0.01633045148895293,
      "grad_norm": 1.0,
      "learning_rate": 8.450000000000001e-05,
      "loss": 6.4838,
      "mean_token_accuracy": 0.1344938911497593,
      "num_tokens": 387318.0,
      "step": 170
    },
    {
      "entropy": 6.999642038345337,
      "epoch": 0.01681075888568684,
      "grad_norm": 1.0625,
      "learning_rate": 8.7e-05,
      "loss": 6.5663,
      "mean_token_accuracy": 0.13669942170381547,
      "num_tokens": 398933.0,
      "step": 175
    },
    {
      "entropy": 7.019154787063599,
      "epoch": 0.01729106628242075,
      "grad_norm": 1.0390625,
      "learning_rate": 8.95e-05,
      "loss": 6.5536,
      "mean_token_accuracy": 0.1361289769411087,
      "num_tokens": 410460.0,
      "step": 180
    },
    {
      "entropy": 7.02680025100708,
      "epoch": 0.01777137367915466,
      "grad_norm": 1.0546875,
      "learning_rate": 9.2e-05,
      "loss": 6.5731,
      "mean_token_accuracy": 0.13559282198548317,
      "num_tokens": 421894.0,
      "step": 185
    },
    {
      "entropy": 7.089911031723022,
      "epoch": 0.01825168107588857,
      "grad_norm": 1.0859375,
      "learning_rate": 9.45e-05,
      "loss": 6.6183,
      "mean_token_accuracy": 0.12840212360024453,
      "num_tokens": 432976.0,
      "step": 190
    },
    {
      "entropy": 7.079966688156128,
      "epoch": 0.018731988472622477,
      "grad_norm": 1.1328125,
      "learning_rate": 9.7e-05,
      "loss": 6.6682,
      "mean_token_accuracy": 0.13189114406704902,
      "num_tokens": 444859.0,
      "step": 195
    },
    {
      "entropy": 7.1056396007537845,
      "epoch": 0.01921229586935639,
      "grad_norm": 1.0546875,
      "learning_rate": 9.95e-05,
      "loss": 6.582,
      "mean_token_accuracy": 0.1382762998342514,
      "num_tokens": 456508.0,
      "step": 200
    },
    {
      "entropy": 7.003632974624634,
      "epoch": 0.0196926032660903,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000102,
      "loss": 6.6044,
      "mean_token_accuracy": 0.12909564301371573,
      "num_tokens": 469064.0,
      "step": 205
    },
    {
      "entropy": 7.1269261837005615,
      "epoch": 0.020172910662824207,
      "grad_norm": 1.375,
      "learning_rate": 0.00010449999999999999,
      "loss": 6.5451,
      "mean_token_accuracy": 0.13477757573127747,
      "num_tokens": 480743.0,
      "step": 210
    },
    {
      "entropy": 7.030986785888672,
      "epoch": 0.020653218059558116,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000107,
      "loss": 6.5874,
      "mean_token_accuracy": 0.13419262766838075,
      "num_tokens": 493378.0,
      "step": 215
    },
    {
      "entropy": 7.058184576034546,
      "epoch": 0.021133525456292025,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0001095,
      "loss": 6.5813,
      "mean_token_accuracy": 0.12972840294241905,
      "num_tokens": 505641.0,
      "step": 220
    },
    {
      "entropy": 7.053190851211548,
      "epoch": 0.021613832853025938,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000112,
      "loss": 6.6461,
      "mean_token_accuracy": 0.1300312601029873,
      "num_tokens": 517783.0,
      "step": 225
    },
    {
      "entropy": 7.0220287322998045,
      "epoch": 0.022094140249759846,
      "grad_norm": 1.1875,
      "learning_rate": 0.0001145,
      "loss": 6.4915,
      "mean_token_accuracy": 0.1375163048505783,
      "num_tokens": 528407.0,
      "step": 230
    },
    {
      "entropy": 7.136337566375732,
      "epoch": 0.022574447646493755,
      "grad_norm": 1.140625,
      "learning_rate": 0.00011700000000000001,
      "loss": 6.6536,
      "mean_token_accuracy": 0.13205716013908386,
      "num_tokens": 539974.0,
      "step": 235
    },
    {
      "entropy": 6.980577278137207,
      "epoch": 0.023054755043227664,
      "grad_norm": 1.140625,
      "learning_rate": 0.00011949999999999999,
      "loss": 6.4986,
      "mean_token_accuracy": 0.13339756950736045,
      "num_tokens": 551127.0,
      "step": 240
    },
    {
      "entropy": 6.999716663360596,
      "epoch": 0.023535062439961577,
      "grad_norm": 1.125,
      "learning_rate": 0.000122,
      "loss": 6.5402,
      "mean_token_accuracy": 0.13888919427990915,
      "num_tokens": 561980.0,
      "step": 245
    },
    {
      "entropy": 7.093811845779419,
      "epoch": 0.024015369836695485,
      "grad_norm": 1.15625,
      "learning_rate": 0.0001245,
      "loss": 6.587,
      "mean_token_accuracy": 0.1365107201039791,
      "num_tokens": 573840.0,
      "step": 250
    },
    {
      "entropy": 6.957034015655518,
      "epoch": 0.024495677233429394,
      "grad_norm": 1.15625,
      "learning_rate": 0.000127,
      "loss": 6.5359,
      "mean_token_accuracy": 0.13650909885764123,
      "num_tokens": 585110.0,
      "step": 255
    },
    {
      "entropy": 6.963628101348877,
      "epoch": 0.024975984630163303,
      "grad_norm": 1.09375,
      "learning_rate": 0.0001295,
      "loss": 6.4158,
      "mean_token_accuracy": 0.13956196010112762,
      "num_tokens": 596902.0,
      "step": 260
    },
    {
      "entropy": 7.003592777252197,
      "epoch": 0.025456292026897216,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000132,
      "loss": 6.6084,
      "mean_token_accuracy": 0.13234634920954705,
      "num_tokens": 608738.0,
      "step": 265
    },
    {
      "entropy": 7.03656964302063,
      "epoch": 0.025936599423631124,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00013450000000000002,
      "loss": 6.527,
      "mean_token_accuracy": 0.13383049592375756,
      "num_tokens": 621078.0,
      "step": 270
    },
    {
      "entropy": 7.060523128509521,
      "epoch": 0.026416906820365033,
      "grad_norm": 1.21875,
      "learning_rate": 0.00013700000000000002,
      "loss": 6.6203,
      "mean_token_accuracy": 0.13373518884181976,
      "num_tokens": 632333.0,
      "step": 275
    },
    {
      "entropy": 6.994757175445557,
      "epoch": 0.026897214217098942,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0001395,
      "loss": 6.4486,
      "mean_token_accuracy": 0.13340532183647155,
      "num_tokens": 644474.0,
      "step": 280
    },
    {
      "entropy": 6.9266692161560055,
      "epoch": 0.027377521613832854,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00014199999999999998,
      "loss": 6.5437,
      "mean_token_accuracy": 0.13545908331871032,
      "num_tokens": 655608.0,
      "step": 285
    },
    {
      "entropy": 7.043833541870117,
      "epoch": 0.027857829010566763,
      "grad_norm": 1.09375,
      "learning_rate": 0.0001445,
      "loss": 6.5544,
      "mean_token_accuracy": 0.13327746465802193,
      "num_tokens": 667371.0,
      "step": 290
    },
    {
      "entropy": 7.025187253952026,
      "epoch": 0.028338136407300672,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000147,
      "loss": 6.5713,
      "mean_token_accuracy": 0.12935664281249046,
      "num_tokens": 679152.0,
      "step": 295
    },
    {
      "entropy": 6.939578628540039,
      "epoch": 0.02881844380403458,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0001495,
      "loss": 6.4936,
      "mean_token_accuracy": 0.1386971041560173,
      "num_tokens": 690359.0,
      "step": 300
    },
    {
      "entropy": 6.998729133605957,
      "epoch": 0.029298751200768493,
      "grad_norm": 1.203125,
      "learning_rate": 0.000152,
      "loss": 6.4871,
      "mean_token_accuracy": 0.13214991092681885,
      "num_tokens": 701819.0,
      "step": 305
    },
    {
      "entropy": 7.02758297920227,
      "epoch": 0.029779058597502402,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00015450000000000001,
      "loss": 6.5132,
      "mean_token_accuracy": 0.13222344741225242,
      "num_tokens": 711048.0,
      "step": 310
    },
    {
      "entropy": 6.924132680892944,
      "epoch": 0.03025936599423631,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000157,
      "loss": 6.4272,
      "mean_token_accuracy": 0.14279273226857186,
      "num_tokens": 721399.0,
      "step": 315
    },
    {
      "entropy": 6.975829458236694,
      "epoch": 0.03073967339097022,
      "grad_norm": 1.109375,
      "learning_rate": 0.0001595,
      "loss": 6.4467,
      "mean_token_accuracy": 0.14665683656930922,
      "num_tokens": 732144.0,
      "step": 320
    },
    {
      "entropy": 6.8750978946685795,
      "epoch": 0.03121998078770413,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000162,
      "loss": 6.439,
      "mean_token_accuracy": 0.13812113925814629,
      "num_tokens": 743962.0,
      "step": 325
    },
    {
      "entropy": 7.067961740493774,
      "epoch": 0.03170028818443804,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00016450000000000001,
      "loss": 6.5958,
      "mean_token_accuracy": 0.13596020564436911,
      "num_tokens": 755641.0,
      "step": 330
    },
    {
      "entropy": 6.8925072193145756,
      "epoch": 0.03218059558117195,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00016700000000000002,
      "loss": 6.4498,
      "mean_token_accuracy": 0.14140594229102135,
      "num_tokens": 766798.0,
      "step": 335
    },
    {
      "entropy": 6.949503374099732,
      "epoch": 0.03266090297790586,
      "grad_norm": 1.171875,
      "learning_rate": 0.00016950000000000003,
      "loss": 6.5306,
      "mean_token_accuracy": 0.1402709811925888,
      "num_tokens": 778097.0,
      "step": 340
    },
    {
      "entropy": 6.915549230575562,
      "epoch": 0.03314121037463977,
      "grad_norm": 1.3125,
      "learning_rate": 0.00017199999999999998,
      "loss": 6.458,
      "mean_token_accuracy": 0.14018686711788178,
      "num_tokens": 789775.0,
      "step": 345
    },
    {
      "entropy": 7.000381278991699,
      "epoch": 0.03362151777137368,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00017449999999999999,
      "loss": 6.5205,
      "mean_token_accuracy": 0.134874277561903,
      "num_tokens": 800155.0,
      "step": 350
    },
    {
      "entropy": 6.96249566078186,
      "epoch": 0.034101825168107586,
      "grad_norm": 1.171875,
      "learning_rate": 0.000177,
      "loss": 6.5313,
      "mean_token_accuracy": 0.13899640440940858,
      "num_tokens": 811213.0,
      "step": 355
    },
    {
      "entropy": 7.033797121047973,
      "epoch": 0.0345821325648415,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0001795,
      "loss": 6.5259,
      "mean_token_accuracy": 0.12905024066567422,
      "num_tokens": 822832.0,
      "step": 360
    },
    {
      "entropy": 6.906823778152466,
      "epoch": 0.03506243996157541,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000182,
      "loss": 6.5062,
      "mean_token_accuracy": 0.13670841753482818,
      "num_tokens": 834824.0,
      "step": 365
    },
    {
      "entropy": 6.941941642761231,
      "epoch": 0.03554274735830932,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0001845,
      "loss": 6.5528,
      "mean_token_accuracy": 0.1321589730679989,
      "num_tokens": 846969.0,
      "step": 370
    },
    {
      "entropy": 7.095604753494262,
      "epoch": 0.03602305475504323,
      "grad_norm": 1.1875,
      "learning_rate": 0.000187,
      "loss": 6.6279,
      "mean_token_accuracy": 0.12456684783101082,
      "num_tokens": 859271.0,
      "step": 375
    },
    {
      "entropy": 6.995700550079346,
      "epoch": 0.03650336215177714,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0001895,
      "loss": 6.5165,
      "mean_token_accuracy": 0.13592087998986244,
      "num_tokens": 869754.0,
      "step": 380
    },
    {
      "entropy": 6.893608093261719,
      "epoch": 0.036983669548511046,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000192,
      "loss": 6.465,
      "mean_token_accuracy": 0.1396843321621418,
      "num_tokens": 880519.0,
      "step": 385
    },
    {
      "entropy": 6.874870157241821,
      "epoch": 0.037463976945244955,
      "grad_norm": 1.34375,
      "learning_rate": 0.0001945,
      "loss": 6.4312,
      "mean_token_accuracy": 0.13749386817216874,
      "num_tokens": 890765.0,
      "step": 390
    },
    {
      "entropy": 6.942132997512817,
      "epoch": 0.037944284341978864,
      "grad_norm": 1.328125,
      "learning_rate": 0.00019700000000000002,
      "loss": 6.4429,
      "mean_token_accuracy": 0.1352140188217163,
      "num_tokens": 901717.0,
      "step": 395
    },
    {
      "entropy": 6.989188814163208,
      "epoch": 0.03842459173871278,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00019950000000000002,
      "loss": 6.527,
      "mean_token_accuracy": 0.13658951371908187,
      "num_tokens": 913334.0,
      "step": 400
    },
    {
      "entropy": 7.002504158020019,
      "epoch": 0.03890489913544669,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000202,
      "loss": 6.5367,
      "mean_token_accuracy": 0.13638701885938645,
      "num_tokens": 924753.0,
      "step": 405
    },
    {
      "entropy": 6.9172460556030275,
      "epoch": 0.0393852065321806,
      "grad_norm": 1.15625,
      "learning_rate": 0.00020449999999999998,
      "loss": 6.4758,
      "mean_token_accuracy": 0.13802975565195083,
      "num_tokens": 935692.0,
      "step": 410
    },
    {
      "entropy": 6.995597171783447,
      "epoch": 0.039865513928914506,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000207,
      "loss": 6.5407,
      "mean_token_accuracy": 0.13431550040841103,
      "num_tokens": 947902.0,
      "step": 415
    },
    {
      "entropy": 6.920178604125977,
      "epoch": 0.040345821325648415,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002095,
      "loss": 6.5029,
      "mean_token_accuracy": 0.14091430008411407,
      "num_tokens": 959437.0,
      "step": 420
    },
    {
      "entropy": 6.818842840194702,
      "epoch": 0.040826128722382324,
      "grad_norm": 1.140625,
      "learning_rate": 0.000212,
      "loss": 6.325,
      "mean_token_accuracy": 0.14433388337492942,
      "num_tokens": 970191.0,
      "step": 425
    },
    {
      "entropy": 6.920823621749878,
      "epoch": 0.04130643611911623,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002145,
      "loss": 6.4527,
      "mean_token_accuracy": 0.13763294816017152,
      "num_tokens": 981811.0,
      "step": 430
    },
    {
      "entropy": 6.8866934299469,
      "epoch": 0.04178674351585014,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00021700000000000002,
      "loss": 6.5686,
      "mean_token_accuracy": 0.13008847534656526,
      "num_tokens": 994034.0,
      "step": 435
    },
    {
      "entropy": 6.982378435134888,
      "epoch": 0.04226705091258405,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002195,
      "loss": 6.5232,
      "mean_token_accuracy": 0.1318051353096962,
      "num_tokens": 1005506.0,
      "step": 440
    },
    {
      "entropy": 6.865565204620362,
      "epoch": 0.042747358309317966,
      "grad_norm": 1.34375,
      "learning_rate": 0.000222,
      "loss": 6.4801,
      "mean_token_accuracy": 0.12741892486810685,
      "num_tokens": 1017103.0,
      "step": 445
    },
    {
      "entropy": 6.912385082244873,
      "epoch": 0.043227665706051875,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002245,
      "loss": 6.5681,
      "mean_token_accuracy": 0.13292499631643295,
      "num_tokens": 1029606.0,
      "step": 450
    },
    {
      "entropy": 6.989826202392578,
      "epoch": 0.043707973102785784,
      "grad_norm": 1.328125,
      "learning_rate": 0.00022700000000000002,
      "loss": 6.5356,
      "mean_token_accuracy": 0.12916895225644112,
      "num_tokens": 1041899.0,
      "step": 455
    },
    {
      "entropy": 6.912504577636719,
      "epoch": 0.04418828049951969,
      "grad_norm": 1.203125,
      "learning_rate": 0.00022950000000000002,
      "loss": 6.4378,
      "mean_token_accuracy": 0.14383152276277542,
      "num_tokens": 1053577.0,
      "step": 460
    },
    {
      "entropy": 6.913370943069458,
      "epoch": 0.0446685878962536,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00023200000000000003,
      "loss": 6.4929,
      "mean_token_accuracy": 0.13895317614078523,
      "num_tokens": 1065541.0,
      "step": 465
    },
    {
      "entropy": 6.934657716751099,
      "epoch": 0.04514889529298751,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00023449999999999998,
      "loss": 6.4416,
      "mean_token_accuracy": 0.137308207154274,
      "num_tokens": 1076138.0,
      "step": 470
    },
    {
      "entropy": 6.862859153747559,
      "epoch": 0.04562920268972142,
      "grad_norm": 1.1875,
      "learning_rate": 0.000237,
      "loss": 6.5041,
      "mean_token_accuracy": 0.13841527476906776,
      "num_tokens": 1087620.0,
      "step": 475
    },
    {
      "entropy": 6.862703037261963,
      "epoch": 0.04610951008645533,
      "grad_norm": 1.25,
      "learning_rate": 0.0002395,
      "loss": 6.3943,
      "mean_token_accuracy": 0.142742570489645,
      "num_tokens": 1099493.0,
      "step": 480
    },
    {
      "entropy": 6.874656057357788,
      "epoch": 0.046589817483189244,
      "grad_norm": 1.1875,
      "learning_rate": 0.000242,
      "loss": 6.4578,
      "mean_token_accuracy": 0.13684100583195685,
      "num_tokens": 1110559.0,
      "step": 485
    },
    {
      "entropy": 6.856756639480591,
      "epoch": 0.04707012487992315,
      "grad_norm": 1.125,
      "learning_rate": 0.0002445,
      "loss": 6.4327,
      "mean_token_accuracy": 0.13060991689562798,
      "num_tokens": 1121051.0,
      "step": 490
    },
    {
      "entropy": 6.838518524169922,
      "epoch": 0.04755043227665706,
      "grad_norm": 1.140625,
      "learning_rate": 0.000247,
      "loss": 6.3931,
      "mean_token_accuracy": 0.14155778735876084,
      "num_tokens": 1132305.0,
      "step": 495
    },
    {
      "entropy": 6.8174115180969235,
      "epoch": 0.04803073967339097,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002495,
      "loss": 6.4016,
      "mean_token_accuracy": 0.1437266267836094,
      "num_tokens": 1142979.0,
      "step": 500
    },
    {
      "entropy": 6.89087495803833,
      "epoch": 0.04851104707012488,
      "grad_norm": 1.34375,
      "learning_rate": 0.000252,
      "loss": 6.411,
      "mean_token_accuracy": 0.1463697850704193,
      "num_tokens": 1153889.0,
      "step": 505
    },
    {
      "entropy": 6.875075244903565,
      "epoch": 0.04899135446685879,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002545,
      "loss": 6.4292,
      "mean_token_accuracy": 0.13723334297537804,
      "num_tokens": 1166030.0,
      "step": 510
    },
    {
      "entropy": 6.966767072677612,
      "epoch": 0.0494716618635927,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000257,
      "loss": 6.6142,
      "mean_token_accuracy": 0.1286202512681484,
      "num_tokens": 1177545.0,
      "step": 515
    },
    {
      "entropy": 6.8341772079467775,
      "epoch": 0.049951969260326606,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002595,
      "loss": 6.4176,
      "mean_token_accuracy": 0.1422701135277748,
      "num_tokens": 1188944.0,
      "step": 520
    },
    {
      "entropy": 6.770974016189575,
      "epoch": 0.05043227665706052,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000262,
      "loss": 6.4883,
      "mean_token_accuracy": 0.13371295407414435,
      "num_tokens": 1199517.0,
      "step": 525
    },
    {
      "entropy": 6.8674828052520756,
      "epoch": 0.05091258405379443,
      "grad_norm": 1.078125,
      "learning_rate": 0.00026450000000000003,
      "loss": 6.4133,
      "mean_token_accuracy": 0.13357224985957145,
      "num_tokens": 1211536.0,
      "step": 530
    },
    {
      "entropy": 6.761517667770386,
      "epoch": 0.05139289145052834,
      "grad_norm": 1.09375,
      "learning_rate": 0.00026700000000000004,
      "loss": 6.3223,
      "mean_token_accuracy": 0.14621875584125518,
      "num_tokens": 1224159.0,
      "step": 535
    },
    {
      "entropy": 6.808783054351807,
      "epoch": 0.05187319884726225,
      "grad_norm": 1.109375,
      "learning_rate": 0.00026950000000000005,
      "loss": 6.4059,
      "mean_token_accuracy": 0.14924411177635194,
      "num_tokens": 1235226.0,
      "step": 540
    },
    {
      "entropy": 6.741241598129273,
      "epoch": 0.05235350624399616,
      "grad_norm": 1.1875,
      "learning_rate": 0.00027200000000000005,
      "loss": 6.3238,
      "mean_token_accuracy": 0.1498336486518383,
      "num_tokens": 1245862.0,
      "step": 545
    },
    {
      "entropy": 6.785985374450684,
      "epoch": 0.052833813640730067,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002745,
      "loss": 6.3347,
      "mean_token_accuracy": 0.14524050951004028,
      "num_tokens": 1257512.0,
      "step": 550
    },
    {
      "entropy": 6.758766126632691,
      "epoch": 0.053314121037463975,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000277,
      "loss": 6.4508,
      "mean_token_accuracy": 0.13895726203918457,
      "num_tokens": 1269317.0,
      "step": 555
    },
    {
      "entropy": 6.7143903255462645,
      "epoch": 0.053794428434197884,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002795,
      "loss": 6.3328,
      "mean_token_accuracy": 0.13423674702644348,
      "num_tokens": 1281878.0,
      "step": 560
    },
    {
      "entropy": 6.754738235473633,
      "epoch": 0.05427473583093179,
      "grad_norm": 1.015625,
      "learning_rate": 0.00028199999999999997,
      "loss": 6.3475,
      "mean_token_accuracy": 0.14375446885824203,
      "num_tokens": 1293178.0,
      "step": 565
    },
    {
      "entropy": 6.717558813095093,
      "epoch": 0.05475504322766571,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002845,
      "loss": 6.3412,
      "mean_token_accuracy": 0.14117253720760345,
      "num_tokens": 1304969.0,
      "step": 570
    },
    {
      "entropy": 6.835646152496338,
      "epoch": 0.05523535062439962,
      "grad_norm": 1.15625,
      "learning_rate": 0.000287,
      "loss": 6.3772,
      "mean_token_accuracy": 0.1407823756337166,
      "num_tokens": 1317301.0,
      "step": 575
    },
    {
      "entropy": 6.74691801071167,
      "epoch": 0.05571565802113353,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002895,
      "loss": 6.363,
      "mean_token_accuracy": 0.1412644602358341,
      "num_tokens": 1327688.0,
      "step": 580
    },
    {
      "entropy": 6.738905239105224,
      "epoch": 0.056195965417867436,
      "grad_norm": 1.4375,
      "learning_rate": 0.000292,
      "loss": 6.3153,
      "mean_token_accuracy": 0.13932914063334464,
      "num_tokens": 1338750.0,
      "step": 585
    },
    {
      "entropy": 6.783087301254272,
      "epoch": 0.056676272814601344,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002945,
      "loss": 6.46,
      "mean_token_accuracy": 0.13216890692710875,
      "num_tokens": 1350899.0,
      "step": 590
    },
    {
      "entropy": 6.860762405395508,
      "epoch": 0.05715658021133525,
      "grad_norm": 1.21875,
      "learning_rate": 0.000297,
      "loss": 6.3102,
      "mean_token_accuracy": 0.14398529529571533,
      "num_tokens": 1363184.0,
      "step": 595
    },
    {
      "entropy": 6.663176107406616,
      "epoch": 0.05763688760806916,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002995,
      "loss": 6.2972,
      "mean_token_accuracy": 0.1460181102156639,
      "num_tokens": 1374192.0,
      "step": 600
    },
    {
      "entropy": 6.650053453445435,
      "epoch": 0.05811719500480307,
      "grad_norm": 1.25,
      "learning_rate": 0.000302,
      "loss": 6.327,
      "mean_token_accuracy": 0.14458349198102952,
      "num_tokens": 1384888.0,
      "step": 605
    },
    {
      "entropy": 6.893434619903564,
      "epoch": 0.05859750240153699,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003045,
      "loss": 6.4132,
      "mean_token_accuracy": 0.14370108693838118,
      "num_tokens": 1395139.0,
      "step": 610
    },
    {
      "entropy": 6.52349820137024,
      "epoch": 0.059077809798270896,
      "grad_norm": 1.25,
      "learning_rate": 0.000307,
      "loss": 6.2221,
      "mean_token_accuracy": 0.14337074458599092,
      "num_tokens": 1407219.0,
      "step": 615
    },
    {
      "entropy": 6.780193042755127,
      "epoch": 0.059558117195004805,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003095,
      "loss": 6.3341,
      "mean_token_accuracy": 0.14062620624899863,
      "num_tokens": 1418917.0,
      "step": 620
    },
    {
      "entropy": 6.667874813079834,
      "epoch": 0.060038424591738714,
      "grad_norm": 1.15625,
      "learning_rate": 0.000312,
      "loss": 6.3008,
      "mean_token_accuracy": 0.14137485176324843,
      "num_tokens": 1430792.0,
      "step": 625
    },
    {
      "entropy": 6.647404527664184,
      "epoch": 0.06051873198847262,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003145,
      "loss": 6.3181,
      "mean_token_accuracy": 0.14391812160611153,
      "num_tokens": 1442825.0,
      "step": 630
    },
    {
      "entropy": 6.798049831390381,
      "epoch": 0.06099903938520653,
      "grad_norm": 1.171875,
      "learning_rate": 0.000317,
      "loss": 6.4112,
      "mean_token_accuracy": 0.14235697835683822,
      "num_tokens": 1454078.0,
      "step": 635
    },
    {
      "entropy": 6.718965768814087,
      "epoch": 0.06147934678194044,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003195,
      "loss": 6.3395,
      "mean_token_accuracy": 0.14023290127515792,
      "num_tokens": 1465164.0,
      "step": 640
    },
    {
      "entropy": 6.73060793876648,
      "epoch": 0.06195965417867435,
      "grad_norm": 1.15625,
      "learning_rate": 0.000322,
      "loss": 6.3445,
      "mean_token_accuracy": 0.14530042111873626,
      "num_tokens": 1476813.0,
      "step": 645
    },
    {
      "entropy": 6.674618053436279,
      "epoch": 0.06243996157540826,
      "grad_norm": 1.25,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.3688,
      "mean_token_accuracy": 0.14241051748394967,
      "num_tokens": 1488547.0,
      "step": 650
    },
    {
      "entropy": 6.862033843994141,
      "epoch": 0.06292026897214217,
      "grad_norm": 1.125,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.366,
      "mean_token_accuracy": 0.13735537007451057,
      "num_tokens": 1500714.0,
      "step": 655
    },
    {
      "entropy": 6.628754711151123,
      "epoch": 0.06340057636887608,
      "grad_norm": 1.234375,
      "learning_rate": 0.00032950000000000004,
      "loss": 6.3567,
      "mean_token_accuracy": 0.1390194483101368,
      "num_tokens": 1512212.0,
      "step": 660
    },
    {
      "entropy": 6.675348711013794,
      "epoch": 0.06388088376560999,
      "grad_norm": 1.28125,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.3282,
      "mean_token_accuracy": 0.1440582610666752,
      "num_tokens": 1523610.0,
      "step": 665
    },
    {
      "entropy": 6.641611242294312,
      "epoch": 0.0643611911623439,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00033450000000000005,
      "loss": 6.2809,
      "mean_token_accuracy": 0.1418041206896305,
      "num_tokens": 1535543.0,
      "step": 670
    },
    {
      "entropy": 6.843907260894776,
      "epoch": 0.06484149855907781,
      "grad_norm": 1.15625,
      "learning_rate": 0.000337,
      "loss": 6.506,
      "mean_token_accuracy": 0.13386593759059906,
      "num_tokens": 1547183.0,
      "step": 675
    },
    {
      "entropy": 6.615519618988037,
      "epoch": 0.06532180595581172,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003395,
      "loss": 6.2559,
      "mean_token_accuracy": 0.14652696400880813,
      "num_tokens": 1559265.0,
      "step": 680
    },
    {
      "entropy": 6.744142818450928,
      "epoch": 0.06580211335254563,
      "grad_norm": 1.25,
      "learning_rate": 0.000342,
      "loss": 6.2822,
      "mean_token_accuracy": 0.141647919267416,
      "num_tokens": 1569774.0,
      "step": 685
    },
    {
      "entropy": 6.5239664077758786,
      "epoch": 0.06628242074927954,
      "grad_norm": 1.328125,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.2239,
      "mean_token_accuracy": 0.14382586181163787,
      "num_tokens": 1579408.0,
      "step": 690
    },
    {
      "entropy": 6.593590784072876,
      "epoch": 0.06676272814601344,
      "grad_norm": 1.171875,
      "learning_rate": 0.000347,
      "loss": 6.2627,
      "mean_token_accuracy": 0.1462230786681175,
      "num_tokens": 1591849.0,
      "step": 695
    },
    {
      "entropy": 6.682480907440185,
      "epoch": 0.06724303554274735,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003495,
      "loss": 6.3308,
      "mean_token_accuracy": 0.14186072424054147,
      "num_tokens": 1603269.0,
      "step": 700
    },
    {
      "entropy": 6.747530841827393,
      "epoch": 0.06772334293948126,
      "grad_norm": 1.1875,
      "learning_rate": 0.000352,
      "loss": 6.3629,
      "mean_token_accuracy": 0.1359557554125786,
      "num_tokens": 1613639.0,
      "step": 705
    },
    {
      "entropy": 6.725496768951416,
      "epoch": 0.06820365033621517,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003545,
      "loss": 6.3855,
      "mean_token_accuracy": 0.13674816712737084,
      "num_tokens": 1625393.0,
      "step": 710
    },
    {
      "entropy": 6.64970293045044,
      "epoch": 0.0686839577329491,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000357,
      "loss": 6.2215,
      "mean_token_accuracy": 0.14604979306459426,
      "num_tokens": 1636355.0,
      "step": 715
    },
    {
      "entropy": 6.653625917434693,
      "epoch": 0.069164265129683,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003595,
      "loss": 6.2686,
      "mean_token_accuracy": 0.14605910181999207,
      "num_tokens": 1646157.0,
      "step": 720
    },
    {
      "entropy": 6.578489160537719,
      "epoch": 0.06964457252641691,
      "grad_norm": 1.125,
      "learning_rate": 0.000362,
      "loss": 6.2573,
      "mean_token_accuracy": 0.13798193708062173,
      "num_tokens": 1657827.0,
      "step": 725
    },
    {
      "entropy": 6.660614967346191,
      "epoch": 0.07012487992315082,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003645,
      "loss": 6.3351,
      "mean_token_accuracy": 0.14547959491610526,
      "num_tokens": 1669023.0,
      "step": 730
    },
    {
      "entropy": 6.684601163864135,
      "epoch": 0.07060518731988473,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000367,
      "loss": 6.2928,
      "mean_token_accuracy": 0.14130547940731047,
      "num_tokens": 1680048.0,
      "step": 735
    },
    {
      "entropy": 6.750500679016113,
      "epoch": 0.07108549471661864,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003695,
      "loss": 6.4804,
      "mean_token_accuracy": 0.1312839813530445,
      "num_tokens": 1690907.0,
      "step": 740
    },
    {
      "entropy": 6.736449241638184,
      "epoch": 0.07156580211335255,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000372,
      "loss": 6.3008,
      "mean_token_accuracy": 0.14280984848737716,
      "num_tokens": 1702337.0,
      "step": 745
    },
    {
      "entropy": 6.587835311889648,
      "epoch": 0.07204610951008646,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003745,
      "loss": 6.2527,
      "mean_token_accuracy": 0.1450270913541317,
      "num_tokens": 1713242.0,
      "step": 750
    },
    {
      "entropy": 6.592082738876343,
      "epoch": 0.07252641690682037,
      "grad_norm": 1.390625,
      "learning_rate": 0.000377,
      "loss": 6.2289,
      "mean_token_accuracy": 0.1451099433004856,
      "num_tokens": 1724939.0,
      "step": 755
    },
    {
      "entropy": 6.562298393249511,
      "epoch": 0.07300672430355427,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003795,
      "loss": 6.1847,
      "mean_token_accuracy": 0.1562918320298195,
      "num_tokens": 1737311.0,
      "step": 760
    },
    {
      "entropy": 6.704523372650146,
      "epoch": 0.07348703170028818,
      "grad_norm": 1.15625,
      "learning_rate": 0.000382,
      "loss": 6.3322,
      "mean_token_accuracy": 0.14251347556710242,
      "num_tokens": 1749528.0,
      "step": 765
    },
    {
      "entropy": 6.5230649471282955,
      "epoch": 0.07396733909702209,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003845,
      "loss": 6.2065,
      "mean_token_accuracy": 0.15109387934207916,
      "num_tokens": 1760304.0,
      "step": 770
    },
    {
      "entropy": 6.557912874221802,
      "epoch": 0.074447646493756,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.19,
      "mean_token_accuracy": 0.14943440705537797,
      "num_tokens": 1772014.0,
      "step": 775
    },
    {
      "entropy": 6.584811496734619,
      "epoch": 0.07492795389048991,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.2511,
      "mean_token_accuracy": 0.13997316658496856,
      "num_tokens": 1784585.0,
      "step": 780
    },
    {
      "entropy": 6.545004224777221,
      "epoch": 0.07540826128722382,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.1812,
      "mean_token_accuracy": 0.1486555442214012,
      "num_tokens": 1795246.0,
      "step": 785
    },
    {
      "entropy": 6.489949035644531,
      "epoch": 0.07588856868395773,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.1709,
      "mean_token_accuracy": 0.15014512687921525,
      "num_tokens": 1806826.0,
      "step": 790
    },
    {
      "entropy": 6.623802709579468,
      "epoch": 0.07636887608069164,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.3092,
      "mean_token_accuracy": 0.14633476734161377,
      "num_tokens": 1818766.0,
      "step": 795
    },
    {
      "entropy": 6.531846284866333,
      "epoch": 0.07684918347742556,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003995,
      "loss": 6.16,
      "mean_token_accuracy": 0.15384283810853958,
      "num_tokens": 1830824.0,
      "step": 800
    },
    {
      "entropy": 6.457365274429321,
      "epoch": 0.07732949087415947,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000402,
      "loss": 6.2466,
      "mean_token_accuracy": 0.14449433609843254,
      "num_tokens": 1841792.0,
      "step": 805
    },
    {
      "entropy": 6.584504079818726,
      "epoch": 0.07780979827089338,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004045,
      "loss": 6.2006,
      "mean_token_accuracy": 0.14591929987072944,
      "num_tokens": 1854683.0,
      "step": 810
    },
    {
      "entropy": 6.556228065490723,
      "epoch": 0.07829010566762729,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.2938,
      "mean_token_accuracy": 0.1405360147356987,
      "num_tokens": 1865826.0,
      "step": 815
    },
    {
      "entropy": 6.604849815368652,
      "epoch": 0.0787704130643612,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004095,
      "loss": 6.2419,
      "mean_token_accuracy": 0.14705629646778107,
      "num_tokens": 1876759.0,
      "step": 820
    },
    {
      "entropy": 6.513918781280518,
      "epoch": 0.0792507204610951,
      "grad_norm": 1.109375,
      "learning_rate": 0.000412,
      "loss": 6.2797,
      "mean_token_accuracy": 0.14388945400714875,
      "num_tokens": 1888013.0,
      "step": 825
    },
    {
      "entropy": 6.682355451583862,
      "epoch": 0.07973102785782901,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004145,
      "loss": 6.2876,
      "mean_token_accuracy": 0.14284019619226457,
      "num_tokens": 1898964.0,
      "step": 830
    },
    {
      "entropy": 6.492662382125855,
      "epoch": 0.08021133525456292,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000417,
      "loss": 6.1948,
      "mean_token_accuracy": 0.15035689398646354,
      "num_tokens": 1909728.0,
      "step": 835
    },
    {
      "entropy": 6.471567058563233,
      "epoch": 0.08069164265129683,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004195,
      "loss": 6.1331,
      "mean_token_accuracy": 0.15249202400445938,
      "num_tokens": 1919895.0,
      "step": 840
    },
    {
      "entropy": 6.498814487457276,
      "epoch": 0.08117195004803074,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000422,
      "loss": 6.0557,
      "mean_token_accuracy": 0.1550162822008133,
      "num_tokens": 1930839.0,
      "step": 845
    },
    {
      "entropy": 6.456363201141357,
      "epoch": 0.08165225744476465,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004245,
      "loss": 6.2616,
      "mean_token_accuracy": 0.14432529211044312,
      "num_tokens": 1942605.0,
      "step": 850
    },
    {
      "entropy": 6.504123783111572,
      "epoch": 0.08213256484149856,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000427,
      "loss": 6.2458,
      "mean_token_accuracy": 0.1410786345601082,
      "num_tokens": 1954863.0,
      "step": 855
    },
    {
      "entropy": 6.629303932189941,
      "epoch": 0.08261287223823247,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004295,
      "loss": 6.3239,
      "mean_token_accuracy": 0.14509388506412507,
      "num_tokens": 1966088.0,
      "step": 860
    },
    {
      "entropy": 6.504941749572754,
      "epoch": 0.08309317963496637,
      "grad_norm": 1.078125,
      "learning_rate": 0.000432,
      "loss": 6.1511,
      "mean_token_accuracy": 0.148190413415432,
      "num_tokens": 1978081.0,
      "step": 865
    },
    {
      "entropy": 6.559056520462036,
      "epoch": 0.08357348703170028,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004345,
      "loss": 6.2224,
      "mean_token_accuracy": 0.13891531452536582,
      "num_tokens": 1991212.0,
      "step": 870
    },
    {
      "entropy": 6.578066205978393,
      "epoch": 0.08405379442843419,
      "grad_norm": 1.28125,
      "learning_rate": 0.000437,
      "loss": 6.2084,
      "mean_token_accuracy": 0.1503169298171997,
      "num_tokens": 2002074.0,
      "step": 875
    },
    {
      "entropy": 6.426190519332886,
      "epoch": 0.0845341018251681,
      "grad_norm": 1.25,
      "learning_rate": 0.0004395,
      "loss": 6.1402,
      "mean_token_accuracy": 0.15407513976097106,
      "num_tokens": 2013039.0,
      "step": 880
    },
    {
      "entropy": 6.500976419448852,
      "epoch": 0.08501440922190202,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000442,
      "loss": 6.2778,
      "mean_token_accuracy": 0.14543393701314927,
      "num_tokens": 2024744.0,
      "step": 885
    },
    {
      "entropy": 6.499779176712036,
      "epoch": 0.08549471661863593,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004445,
      "loss": 6.1422,
      "mean_token_accuracy": 0.151559966057539,
      "num_tokens": 2035587.0,
      "step": 890
    },
    {
      "entropy": 6.507718563079834,
      "epoch": 0.08597502401536984,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000447,
      "loss": 6.2209,
      "mean_token_accuracy": 0.14204468578100204,
      "num_tokens": 2047924.0,
      "step": 895
    },
    {
      "entropy": 6.472938013076782,
      "epoch": 0.08645533141210375,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.1499,
      "mean_token_accuracy": 0.1465149611234665,
      "num_tokens": 2060161.0,
      "step": 900
    },
    {
      "entropy": 6.479418230056763,
      "epoch": 0.08693563880883766,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.1678,
      "mean_token_accuracy": 0.14912493750452996,
      "num_tokens": 2071674.0,
      "step": 905
    },
    {
      "entropy": 6.49160099029541,
      "epoch": 0.08741594620557157,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.2996,
      "mean_token_accuracy": 0.14318522065877914,
      "num_tokens": 2084025.0,
      "step": 910
    },
    {
      "entropy": 6.563802194595337,
      "epoch": 0.08789625360230548,
      "grad_norm": 1.125,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.2327,
      "mean_token_accuracy": 0.14436741322278976,
      "num_tokens": 2094501.0,
      "step": 915
    },
    {
      "entropy": 6.505239629745484,
      "epoch": 0.08837656099903939,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.2167,
      "mean_token_accuracy": 0.15397917181253434,
      "num_tokens": 2106207.0,
      "step": 920
    },
    {
      "entropy": 6.44503755569458,
      "epoch": 0.0888568683957733,
      "grad_norm": 1.34375,
      "learning_rate": 0.000462,
      "loss": 6.1276,
      "mean_token_accuracy": 0.15853458568453788,
      "num_tokens": 2116879.0,
      "step": 925
    },
    {
      "entropy": 6.479341220855713,
      "epoch": 0.0893371757925072,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004645,
      "loss": 6.1848,
      "mean_token_accuracy": 0.1470620773732662,
      "num_tokens": 2128884.0,
      "step": 930
    },
    {
      "entropy": 6.402610540390015,
      "epoch": 0.08981748318924111,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000467,
      "loss": 6.1305,
      "mean_token_accuracy": 0.14514548182487488,
      "num_tokens": 2140810.0,
      "step": 935
    },
    {
      "entropy": 6.536736297607422,
      "epoch": 0.09029779058597502,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004695,
      "loss": 6.1911,
      "mean_token_accuracy": 0.14236362800002098,
      "num_tokens": 2152312.0,
      "step": 940
    },
    {
      "entropy": 6.351864624023437,
      "epoch": 0.09077809798270893,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000472,
      "loss": 6.1298,
      "mean_token_accuracy": 0.144212955981493,
      "num_tokens": 2164408.0,
      "step": 945
    },
    {
      "entropy": 6.436456298828125,
      "epoch": 0.09125840537944284,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004745,
      "loss": 6.1547,
      "mean_token_accuracy": 0.15071666985750198,
      "num_tokens": 2176050.0,
      "step": 950
    },
    {
      "entropy": 6.357456064224243,
      "epoch": 0.09173871277617675,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000477,
      "loss": 6.0357,
      "mean_token_accuracy": 0.1549798734486103,
      "num_tokens": 2187354.0,
      "step": 955
    },
    {
      "entropy": 6.294004631042481,
      "epoch": 0.09221902017291066,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004795,
      "loss": 6.0707,
      "mean_token_accuracy": 0.14669691622257233,
      "num_tokens": 2198686.0,
      "step": 960
    },
    {
      "entropy": 6.354834318161011,
      "epoch": 0.09269932756964457,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000482,
      "loss": 6.1001,
      "mean_token_accuracy": 0.14863732010126113,
      "num_tokens": 2209780.0,
      "step": 965
    },
    {
      "entropy": 6.404127836227417,
      "epoch": 0.09317963496637849,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004845,
      "loss": 6.1929,
      "mean_token_accuracy": 0.15277946293354033,
      "num_tokens": 2221760.0,
      "step": 970
    },
    {
      "entropy": 6.414715957641602,
      "epoch": 0.0936599423631124,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000487,
      "loss": 6.2016,
      "mean_token_accuracy": 0.1440224066376686,
      "num_tokens": 2233767.0,
      "step": 975
    },
    {
      "entropy": 6.447878742218018,
      "epoch": 0.0941402497598463,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004895,
      "loss": 6.1245,
      "mean_token_accuracy": 0.1480201542377472,
      "num_tokens": 2244841.0,
      "step": 980
    },
    {
      "entropy": 6.362892484664917,
      "epoch": 0.09462055715658022,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000492,
      "loss": 5.9922,
      "mean_token_accuracy": 0.15667022168636321,
      "num_tokens": 2255620.0,
      "step": 985
    },
    {
      "entropy": 6.3917787075042725,
      "epoch": 0.09510086455331412,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004945,
      "loss": 6.1894,
      "mean_token_accuracy": 0.15466750115156175,
      "num_tokens": 2266345.0,
      "step": 990
    },
    {
      "entropy": 6.318432140350342,
      "epoch": 0.09558117195004803,
      "grad_norm": 1.328125,
      "learning_rate": 0.000497,
      "loss": 6.1248,
      "mean_token_accuracy": 0.15534795224666595,
      "num_tokens": 2277190.0,
      "step": 995
    },
    {
      "entropy": 6.3766883373260494,
      "epoch": 0.09606147934678194,
      "grad_norm": 1.125,
      "learning_rate": 0.0004995,
      "loss": 5.9979,
      "mean_token_accuracy": 0.15411072820425034,
      "num_tokens": 2288241.0,
      "step": 1000
    },
    {
      "entropy": 6.409312772750854,
      "epoch": 0.09654178674351585,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999999983283737,
      "loss": 6.1597,
      "mean_token_accuracy": 0.14942025095224382,
      "num_tokens": 2301092.0,
      "step": 1005
    },
    {
      "entropy": 6.327887964248657,
      "epoch": 0.09702209414024976,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999999915373924,
      "loss": 6.0533,
      "mean_token_accuracy": 0.14707459956407548,
      "num_tokens": 2313622.0,
      "step": 1010
    },
    {
      "entropy": 6.297283601760864,
      "epoch": 0.09750240153698367,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999999795225793,
      "loss": 6.0189,
      "mean_token_accuracy": 0.1562737464904785,
      "num_tokens": 2325699.0,
      "step": 1015
    },
    {
      "entropy": 6.347620439529419,
      "epoch": 0.09798270893371758,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999999622839347,
      "loss": 6.0973,
      "mean_token_accuracy": 0.15067264884710313,
      "num_tokens": 2337816.0,
      "step": 1020
    },
    {
      "entropy": 6.497613906860352,
      "epoch": 0.09846301633045149,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999999398214593,
      "loss": 6.2175,
      "mean_token_accuracy": 0.1453450456261635,
      "num_tokens": 2348793.0,
      "step": 1025
    },
    {
      "entropy": 6.378514766693115,
      "epoch": 0.0989433237271854,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999999121351532,
      "loss": 6.1278,
      "mean_token_accuracy": 0.14934940189123153,
      "num_tokens": 2360209.0,
      "step": 1030
    },
    {
      "entropy": 6.435119342803955,
      "epoch": 0.0994236311239193,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999998792250173,
      "loss": 6.1277,
      "mean_token_accuracy": 0.1459425002336502,
      "num_tokens": 2370597.0,
      "step": 1035
    },
    {
      "entropy": 6.408862972259522,
      "epoch": 0.09990393852065321,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004999998410910524,
      "loss": 6.1716,
      "mean_token_accuracy": 0.14220923334360122,
      "num_tokens": 2381466.0,
      "step": 1040
    },
    {
      "entropy": 6.488006258010865,
      "epoch": 0.10038424591738712,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004999997977332592,
      "loss": 6.2657,
      "mean_token_accuracy": 0.14426247775554657,
      "num_tokens": 2394507.0,
      "step": 1045
    },
    {
      "entropy": 6.375166702270508,
      "epoch": 0.10086455331412104,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999997491516389,
      "loss": 6.1158,
      "mean_token_accuracy": 0.14977992326021194,
      "num_tokens": 2406367.0,
      "step": 1050
    },
    {
      "entropy": 6.362482166290283,
      "epoch": 0.10134486071085495,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004999996953461925,
      "loss": 6.0791,
      "mean_token_accuracy": 0.15088748335838317,
      "num_tokens": 2416962.0,
      "step": 1055
    },
    {
      "entropy": 6.437332439422607,
      "epoch": 0.10182516810758886,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999996363169212,
      "loss": 6.1431,
      "mean_token_accuracy": 0.14360424429178237,
      "num_tokens": 2429441.0,
      "step": 1060
    },
    {
      "entropy": 6.327810049057007,
      "epoch": 0.10230547550432277,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999995720638266,
      "loss": 6.0265,
      "mean_token_accuracy": 0.15362217277288437,
      "num_tokens": 2441360.0,
      "step": 1065
    },
    {
      "entropy": 6.267141437530517,
      "epoch": 0.10278578290105668,
      "grad_norm": 1.1875,
      "learning_rate": 0.00049999950258691,
      "loss": 6.0527,
      "mean_token_accuracy": 0.14622927457094193,
      "num_tokens": 2452928.0,
      "step": 1070
    },
    {
      "entropy": 6.3083587169647215,
      "epoch": 0.10326609029779059,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999994278861731,
      "loss": 5.9817,
      "mean_token_accuracy": 0.15528771579265593,
      "num_tokens": 2464218.0,
      "step": 1075
    },
    {
      "entropy": 6.301430654525757,
      "epoch": 0.1037463976945245,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999993479616175,
      "loss": 6.0697,
      "mean_token_accuracy": 0.15502351075410842,
      "num_tokens": 2475084.0,
      "step": 1080
    },
    {
      "entropy": 6.31117525100708,
      "epoch": 0.1042267050912584,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004999992628132451,
      "loss": 6.0718,
      "mean_token_accuracy": 0.14931347519159316,
      "num_tokens": 2486845.0,
      "step": 1085
    },
    {
      "entropy": 6.348042297363281,
      "epoch": 0.10470701248799232,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999991724410582,
      "loss": 6.1471,
      "mean_token_accuracy": 0.14322768971323968,
      "num_tokens": 2498049.0,
      "step": 1090
    },
    {
      "entropy": 6.270689201354981,
      "epoch": 0.10518731988472622,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999990768450583,
      "loss": 6.081,
      "mean_token_accuracy": 0.15236250758171083,
      "num_tokens": 2509511.0,
      "step": 1095
    },
    {
      "entropy": 6.438291072845459,
      "epoch": 0.10566762728146013,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999989760252482,
      "loss": 6.0404,
      "mean_token_accuracy": 0.1563153713941574,
      "num_tokens": 2521518.0,
      "step": 1100
    },
    {
      "entropy": 6.131235933303833,
      "epoch": 0.10614793467819404,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999988699816299,
      "loss": 5.8906,
      "mean_token_accuracy": 0.15999584943056105,
      "num_tokens": 2533141.0,
      "step": 1105
    },
    {
      "entropy": 6.186205959320068,
      "epoch": 0.10662824207492795,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004999987587142058,
      "loss": 5.8889,
      "mean_token_accuracy": 0.15622006207704545,
      "num_tokens": 2542907.0,
      "step": 1110
    },
    {
      "entropy": 6.262308979034424,
      "epoch": 0.10710854947166186,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999986422229789,
      "loss": 6.0278,
      "mean_token_accuracy": 0.1462012603878975,
      "num_tokens": 2554497.0,
      "step": 1115
    },
    {
      "entropy": 6.411641597747803,
      "epoch": 0.10758885686839577,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999985205079514,
      "loss": 6.1524,
      "mean_token_accuracy": 0.1471574440598488,
      "num_tokens": 2567035.0,
      "step": 1120
    },
    {
      "entropy": 6.178546047210693,
      "epoch": 0.10806916426512968,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999983935691265,
      "loss": 5.9343,
      "mean_token_accuracy": 0.16074804067611695,
      "num_tokens": 2578003.0,
      "step": 1125
    },
    {
      "entropy": 6.447769737243652,
      "epoch": 0.10854947166186359,
      "grad_norm": 1.171875,
      "learning_rate": 0.000499998261406507,
      "loss": 6.2329,
      "mean_token_accuracy": 0.1447854146361351,
      "num_tokens": 2590272.0,
      "step": 1130
    },
    {
      "entropy": 6.202875328063965,
      "epoch": 0.10902977905859751,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999981240200958,
      "loss": 6.0603,
      "mean_token_accuracy": 0.14931458830833436,
      "num_tokens": 2602874.0,
      "step": 1135
    },
    {
      "entropy": 6.29072413444519,
      "epoch": 0.10951008645533142,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999979814098966,
      "loss": 6.0127,
      "mean_token_accuracy": 0.15842576026916505,
      "num_tokens": 2613835.0,
      "step": 1140
    },
    {
      "entropy": 6.339123868942261,
      "epoch": 0.10999039385206533,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999978335759121,
      "loss": 6.1187,
      "mean_token_accuracy": 0.1503593236207962,
      "num_tokens": 2624986.0,
      "step": 1145
    },
    {
      "entropy": 6.3845244407653805,
      "epoch": 0.11047070124879924,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999976805181461,
      "loss": 6.1242,
      "mean_token_accuracy": 0.14067715033888817,
      "num_tokens": 2637654.0,
      "step": 1150
    },
    {
      "entropy": 6.253936910629273,
      "epoch": 0.11095100864553314,
      "grad_norm": 1.0,
      "learning_rate": 0.000499997522236602,
      "loss": 6.0342,
      "mean_token_accuracy": 0.15293926745653152,
      "num_tokens": 2648947.0,
      "step": 1155
    },
    {
      "entropy": 6.260978364944458,
      "epoch": 0.11143131604226705,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999973587312837,
      "loss": 6.031,
      "mean_token_accuracy": 0.151402048766613,
      "num_tokens": 2659573.0,
      "step": 1160
    },
    {
      "entropy": 6.191074228286743,
      "epoch": 0.11191162343900096,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999971900021947,
      "loss": 6.0048,
      "mean_token_accuracy": 0.15867563784122468,
      "num_tokens": 2671491.0,
      "step": 1165
    },
    {
      "entropy": 6.292398309707641,
      "epoch": 0.11239193083573487,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999970160493391,
      "loss": 5.9964,
      "mean_token_accuracy": 0.1500651329755783,
      "num_tokens": 2682705.0,
      "step": 1170
    },
    {
      "entropy": 6.1503791332244875,
      "epoch": 0.11287223823246878,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999968368727209,
      "loss": 5.9329,
      "mean_token_accuracy": 0.15546464920043945,
      "num_tokens": 2693539.0,
      "step": 1175
    },
    {
      "entropy": 6.208331108093262,
      "epoch": 0.11335254562920269,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999966524723442,
      "loss": 5.9794,
      "mean_token_accuracy": 0.1552668482065201,
      "num_tokens": 2705403.0,
      "step": 1180
    },
    {
      "entropy": 6.37642993927002,
      "epoch": 0.1138328530259366,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999964628482135,
      "loss": 6.0783,
      "mean_token_accuracy": 0.15211726650595664,
      "num_tokens": 2717320.0,
      "step": 1185
    },
    {
      "entropy": 6.221129989624023,
      "epoch": 0.1143131604226705,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999962680003328,
      "loss": 6.0448,
      "mean_token_accuracy": 0.14739547073841094,
      "num_tokens": 2730056.0,
      "step": 1190
    },
    {
      "entropy": 6.283902359008789,
      "epoch": 0.11479346781940442,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000499996067928707,
      "loss": 5.922,
      "mean_token_accuracy": 0.15959976464509965,
      "num_tokens": 2741399.0,
      "step": 1195
    },
    {
      "entropy": 6.31879243850708,
      "epoch": 0.11527377521613832,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999958626333406,
      "loss": 6.155,
      "mean_token_accuracy": 0.1457771949470043,
      "num_tokens": 2752277.0,
      "step": 1200
    },
    {
      "entropy": 6.301510620117187,
      "epoch": 0.11575408261287223,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999956521142383,
      "loss": 6.0392,
      "mean_token_accuracy": 0.15469708293676376,
      "num_tokens": 2763172.0,
      "step": 1205
    },
    {
      "entropy": 6.277292966842651,
      "epoch": 0.11623439000960614,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999954363714051,
      "loss": 6.054,
      "mean_token_accuracy": 0.15491107255220413,
      "num_tokens": 2775096.0,
      "step": 1210
    },
    {
      "entropy": 6.317543888092041,
      "epoch": 0.11671469740634005,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999952154048459,
      "loss": 6.1321,
      "mean_token_accuracy": 0.1500074803829193,
      "num_tokens": 2786007.0,
      "step": 1215
    },
    {
      "entropy": 6.216952800750732,
      "epoch": 0.11719500480307397,
      "grad_norm": 1.125,
      "learning_rate": 0.000499994989214566,
      "loss": 5.9516,
      "mean_token_accuracy": 0.163367360830307,
      "num_tokens": 2797265.0,
      "step": 1220
    },
    {
      "entropy": 6.273515462875366,
      "epoch": 0.11767531219980788,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999947578005705,
      "loss": 6.0993,
      "mean_token_accuracy": 0.15390106439590454,
      "num_tokens": 2809427.0,
      "step": 1225
    },
    {
      "entropy": 6.245223760604858,
      "epoch": 0.11815561959654179,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999945211628648,
      "loss": 6.0295,
      "mean_token_accuracy": 0.15515164881944657,
      "num_tokens": 2822880.0,
      "step": 1230
    },
    {
      "entropy": 6.3243988990783695,
      "epoch": 0.1186359269932757,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999942793014544,
      "loss": 6.1043,
      "mean_token_accuracy": 0.14880725666880606,
      "num_tokens": 2834760.0,
      "step": 1235
    },
    {
      "entropy": 6.2938329696655275,
      "epoch": 0.11911623439000961,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499994032216345,
      "loss": 6.0577,
      "mean_token_accuracy": 0.14785425141453742,
      "num_tokens": 2846517.0,
      "step": 1240
    },
    {
      "entropy": 6.187609577178955,
      "epoch": 0.11959654178674352,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999937799075422,
      "loss": 5.9381,
      "mean_token_accuracy": 0.1608589246869087,
      "num_tokens": 2857280.0,
      "step": 1245
    },
    {
      "entropy": 6.250182056427002,
      "epoch": 0.12007684918347743,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499993522375052,
      "loss": 5.9878,
      "mean_token_accuracy": 0.15561182498931886,
      "num_tokens": 2869735.0,
      "step": 1250
    },
    {
      "entropy": 6.238501167297363,
      "epoch": 0.12055715658021134,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999932596188802,
      "loss": 5.9806,
      "mean_token_accuracy": 0.15750712007284165,
      "num_tokens": 2881681.0,
      "step": 1255
    },
    {
      "entropy": 6.122741842269898,
      "epoch": 0.12103746397694524,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999929916390331,
      "loss": 5.9974,
      "mean_token_accuracy": 0.15890538841485977,
      "num_tokens": 2893828.0,
      "step": 1260
    },
    {
      "entropy": 6.121092176437378,
      "epoch": 0.12151777137367915,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999927184355169,
      "loss": 5.9548,
      "mean_token_accuracy": 0.15334596037864684,
      "num_tokens": 2905804.0,
      "step": 1265
    },
    {
      "entropy": 6.387607383728027,
      "epoch": 0.12199807877041306,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999924400083377,
      "loss": 6.0998,
      "mean_token_accuracy": 0.14984989315271377,
      "num_tokens": 2918709.0,
      "step": 1270
    },
    {
      "entropy": 6.2461841106414795,
      "epoch": 0.12247838616714697,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999921563575022,
      "loss": 6.0112,
      "mean_token_accuracy": 0.15625457540154458,
      "num_tokens": 2930589.0,
      "step": 1275
    },
    {
      "entropy": 6.19944920539856,
      "epoch": 0.12295869356388088,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999918674830169,
      "loss": 6.0664,
      "mean_token_accuracy": 0.1594749942421913,
      "num_tokens": 2940635.0,
      "step": 1280
    },
    {
      "entropy": 6.171869897842408,
      "epoch": 0.12343900096061479,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999915733848886,
      "loss": 5.9354,
      "mean_token_accuracy": 0.16207429766654968,
      "num_tokens": 2952731.0,
      "step": 1285
    },
    {
      "entropy": 6.2040815353393555,
      "epoch": 0.1239193083573487,
      "grad_norm": 1.171875,
      "learning_rate": 0.000499991274063124,
      "loss": 5.9461,
      "mean_token_accuracy": 0.15291942954063414,
      "num_tokens": 2964544.0,
      "step": 1290
    },
    {
      "entropy": 6.162799215316772,
      "epoch": 0.12439961575408261,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999909695177301,
      "loss": 5.8843,
      "mean_token_accuracy": 0.1602574333548546,
      "num_tokens": 2974996.0,
      "step": 1295
    },
    {
      "entropy": 6.114618587493896,
      "epoch": 0.12487992315081652,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000499990659748714,
      "loss": 5.8875,
      "mean_token_accuracy": 0.15879584103822708,
      "num_tokens": 2986496.0,
      "step": 1300
    },
    {
      "entropy": 6.153219079971313,
      "epoch": 0.12536023054755044,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999903447560828,
      "loss": 5.9601,
      "mean_token_accuracy": 0.1546933814883232,
      "num_tokens": 2997693.0,
      "step": 1305
    },
    {
      "entropy": 6.1417900085449215,
      "epoch": 0.12584053794428435,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999900245398439,
      "loss": 5.9379,
      "mean_token_accuracy": 0.159236279129982,
      "num_tokens": 3009155.0,
      "step": 1310
    },
    {
      "entropy": 6.1472837924957275,
      "epoch": 0.12632084534101826,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999896991000047,
      "loss": 5.8526,
      "mean_token_accuracy": 0.16511083841323854,
      "num_tokens": 3020944.0,
      "step": 1315
    },
    {
      "entropy": 6.132077980041504,
      "epoch": 0.12680115273775217,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999893684365729,
      "loss": 5.9347,
      "mean_token_accuracy": 0.15706215500831605,
      "num_tokens": 3032893.0,
      "step": 1320
    },
    {
      "entropy": 6.187384700775146,
      "epoch": 0.12728146013448607,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999890325495559,
      "loss": 5.9891,
      "mean_token_accuracy": 0.15579385310411453,
      "num_tokens": 3044937.0,
      "step": 1325
    },
    {
      "entropy": 6.1738560676574705,
      "epoch": 0.12776176753121998,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999886914389617,
      "loss": 5.8937,
      "mean_token_accuracy": 0.1595969796180725,
      "num_tokens": 3056071.0,
      "step": 1330
    },
    {
      "entropy": 6.156427192687988,
      "epoch": 0.1282420749279539,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004999883451047981,
      "loss": 6.0543,
      "mean_token_accuracy": 0.15404095351696015,
      "num_tokens": 3067724.0,
      "step": 1335
    },
    {
      "entropy": 6.2366071224212645,
      "epoch": 0.1287223823246878,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999879935470733,
      "loss": 5.9563,
      "mean_token_accuracy": 0.1608007699251175,
      "num_tokens": 3080779.0,
      "step": 1340
    },
    {
      "entropy": 6.148285913467407,
      "epoch": 0.1292026897214217,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999876367657954,
      "loss": 5.9385,
      "mean_token_accuracy": 0.15236368924379348,
      "num_tokens": 3091991.0,
      "step": 1345
    },
    {
      "entropy": 6.128611755371094,
      "epoch": 0.12968299711815562,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999872747609725,
      "loss": 5.9266,
      "mean_token_accuracy": 0.15997803062200547,
      "num_tokens": 3103713.0,
      "step": 1350
    },
    {
      "entropy": 6.109859609603882,
      "epoch": 0.13016330451488953,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999869075326132,
      "loss": 5.9504,
      "mean_token_accuracy": 0.15759862959384918,
      "num_tokens": 3115891.0,
      "step": 1355
    },
    {
      "entropy": 6.09540114402771,
      "epoch": 0.13064361191162344,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499986535080726,
      "loss": 5.8154,
      "mean_token_accuracy": 0.16952185183763505,
      "num_tokens": 3126223.0,
      "step": 1360
    },
    {
      "entropy": 6.197498035430908,
      "epoch": 0.13112391930835735,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999861574053196,
      "loss": 5.9901,
      "mean_token_accuracy": 0.1543188914656639,
      "num_tokens": 3137537.0,
      "step": 1365
    },
    {
      "entropy": 6.249709987640381,
      "epoch": 0.13160422670509125,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999857745064027,
      "loss": 6.0021,
      "mean_token_accuracy": 0.15779454112052918,
      "num_tokens": 3150126.0,
      "step": 1370
    },
    {
      "entropy": 6.143283319473267,
      "epoch": 0.13208453410182516,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499985386383984,
      "loss": 5.9765,
      "mean_token_accuracy": 0.1616181269288063,
      "num_tokens": 3161684.0,
      "step": 1375
    },
    {
      "entropy": 6.0805628299713135,
      "epoch": 0.13256484149855907,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999849930380729,
      "loss": 5.8674,
      "mean_token_accuracy": 0.16023531705141067,
      "num_tokens": 3172342.0,
      "step": 1380
    },
    {
      "entropy": 6.236843204498291,
      "epoch": 0.13304514889529298,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999845944686781,
      "loss": 6.0028,
      "mean_token_accuracy": 0.16236832588911057,
      "num_tokens": 3183681.0,
      "step": 1385
    },
    {
      "entropy": 6.0158531188964846,
      "epoch": 0.1335254562920269,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999841906758093,
      "loss": 5.8018,
      "mean_token_accuracy": 0.16931295841932298,
      "num_tokens": 3194122.0,
      "step": 1390
    },
    {
      "entropy": 6.141476392745972,
      "epoch": 0.1340057636887608,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999837816594757,
      "loss": 5.8629,
      "mean_token_accuracy": 0.16289830058813096,
      "num_tokens": 3205719.0,
      "step": 1395
    },
    {
      "entropy": 6.072001695632935,
      "epoch": 0.1344860710854947,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999833674196865,
      "loss": 5.8931,
      "mean_token_accuracy": 0.1596195340156555,
      "num_tokens": 3216262.0,
      "step": 1400
    },
    {
      "entropy": 6.219841957092285,
      "epoch": 0.13496637848222862,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999829479564518,
      "loss": 5.9352,
      "mean_token_accuracy": 0.1537438616156578,
      "num_tokens": 3228035.0,
      "step": 1405
    },
    {
      "entropy": 6.063015413284302,
      "epoch": 0.13544668587896252,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499982523269781,
      "loss": 5.813,
      "mean_token_accuracy": 0.16350635886192322,
      "num_tokens": 3239643.0,
      "step": 1410
    },
    {
      "entropy": 6.081555032730103,
      "epoch": 0.13592699327569643,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999820933596842,
      "loss": 5.8878,
      "mean_token_accuracy": 0.1632865384221077,
      "num_tokens": 3250293.0,
      "step": 1415
    },
    {
      "entropy": 6.233055973052979,
      "epoch": 0.13640730067243034,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499981658226171,
      "loss": 6.0138,
      "mean_token_accuracy": 0.14861609488725663,
      "num_tokens": 3262448.0,
      "step": 1420
    },
    {
      "entropy": 6.011493682861328,
      "epoch": 0.13688760806916425,
      "grad_norm": 1.140625,
      "learning_rate": 0.000499981217869252,
      "loss": 5.8787,
      "mean_token_accuracy": 0.15858830362558365,
      "num_tokens": 3274041.0,
      "step": 1425
    },
    {
      "entropy": 6.152698564529419,
      "epoch": 0.1373679154658982,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499980772288937,
      "loss": 5.9147,
      "mean_token_accuracy": 0.15459389835596085,
      "num_tokens": 3285063.0,
      "step": 1430
    },
    {
      "entropy": 6.150080776214599,
      "epoch": 0.1378482228626321,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999803214852367,
      "loss": 5.9305,
      "mean_token_accuracy": 0.16173136979341507,
      "num_tokens": 3297031.0,
      "step": 1435
    },
    {
      "entropy": 6.131770992279053,
      "epoch": 0.138328530259366,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999798654581613,
      "loss": 5.9157,
      "mean_token_accuracy": 0.156547848880291,
      "num_tokens": 3308187.0,
      "step": 1440
    },
    {
      "entropy": 6.07386417388916,
      "epoch": 0.13880883765609991,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999794042077214,
      "loss": 5.903,
      "mean_token_accuracy": 0.16056945770978928,
      "num_tokens": 3319439.0,
      "step": 1445
    },
    {
      "entropy": 6.1171587944030765,
      "epoch": 0.13928914505283382,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999789377339279,
      "loss": 5.9105,
      "mean_token_accuracy": 0.15625833347439766,
      "num_tokens": 3331715.0,
      "step": 1450
    },
    {
      "entropy": 6.126484775543213,
      "epoch": 0.13976945244956773,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999784660367915,
      "loss": 5.8657,
      "mean_token_accuracy": 0.16759593337774276,
      "num_tokens": 3342281.0,
      "step": 1455
    },
    {
      "entropy": 6.233757495880127,
      "epoch": 0.14024975984630164,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999779891163231,
      "loss": 5.9932,
      "mean_token_accuracy": 0.1566716194152832,
      "num_tokens": 3353530.0,
      "step": 1460
    },
    {
      "entropy": 6.097651147842408,
      "epoch": 0.14073006724303555,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999775069725339,
      "loss": 5.9691,
      "mean_token_accuracy": 0.15415023267269135,
      "num_tokens": 3365083.0,
      "step": 1465
    },
    {
      "entropy": 6.1879171371459964,
      "epoch": 0.14121037463976946,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000499977019605435,
      "loss": 5.9113,
      "mean_token_accuracy": 0.15917533934116362,
      "num_tokens": 3378143.0,
      "step": 1470
    },
    {
      "entropy": 6.168327140808105,
      "epoch": 0.14169068203650337,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999765270150378,
      "loss": 5.8914,
      "mean_token_accuracy": 0.15909557938575744,
      "num_tokens": 3388438.0,
      "step": 1475
    },
    {
      "entropy": 6.007050275802612,
      "epoch": 0.14217098943323728,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999760292013536,
      "loss": 5.8862,
      "mean_token_accuracy": 0.15672476589679718,
      "num_tokens": 3400150.0,
      "step": 1480
    },
    {
      "entropy": 6.030299091339112,
      "epoch": 0.14265129682997119,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999755261643941,
      "loss": 5.7668,
      "mean_token_accuracy": 0.16843343377113343,
      "num_tokens": 3411073.0,
      "step": 1485
    },
    {
      "entropy": 6.067905569076538,
      "epoch": 0.1431316042267051,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999750179041709,
      "loss": 5.8339,
      "mean_token_accuracy": 0.17172128707170486,
      "num_tokens": 3422196.0,
      "step": 1490
    },
    {
      "entropy": 6.1059250831604,
      "epoch": 0.143611911623439,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999745044206959,
      "loss": 5.8685,
      "mean_token_accuracy": 0.15575093105435373,
      "num_tokens": 3432568.0,
      "step": 1495
    },
    {
      "entropy": 6.113522005081177,
      "epoch": 0.1440922190201729,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999739857139809,
      "loss": 5.8986,
      "mean_token_accuracy": 0.15804980993270873,
      "num_tokens": 3444096.0,
      "step": 1500
    },
    {
      "entropy": 5.95184359550476,
      "epoch": 0.14457252641690682,
      "grad_norm": 1.0,
      "learning_rate": 0.000499973461784038,
      "loss": 5.7744,
      "mean_token_accuracy": 0.16874846071004868,
      "num_tokens": 3455502.0,
      "step": 1505
    },
    {
      "entropy": 6.154258155822754,
      "epoch": 0.14505283381364073,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999729326308792,
      "loss": 5.9156,
      "mean_token_accuracy": 0.15503724962472915,
      "num_tokens": 3467753.0,
      "step": 1510
    },
    {
      "entropy": 6.040020275115967,
      "epoch": 0.14553314121037464,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000499972398254517,
      "loss": 5.7886,
      "mean_token_accuracy": 0.1608596622943878,
      "num_tokens": 3479206.0,
      "step": 1515
    },
    {
      "entropy": 6.036813592910766,
      "epoch": 0.14601344860710855,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000499971858654964,
      "loss": 5.822,
      "mean_token_accuracy": 0.16021780669689178,
      "num_tokens": 3490914.0,
      "step": 1520
    },
    {
      "entropy": 6.045322847366333,
      "epoch": 0.14649375600384246,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999713138322321,
      "loss": 5.8726,
      "mean_token_accuracy": 0.15877723693847656,
      "num_tokens": 3502932.0,
      "step": 1525
    },
    {
      "entropy": 5.995212125778198,
      "epoch": 0.14697406340057637,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999707637863346,
      "loss": 5.801,
      "mean_token_accuracy": 0.16770601868629456,
      "num_tokens": 3515071.0,
      "step": 1530
    },
    {
      "entropy": 6.137019300460816,
      "epoch": 0.14745437079731027,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004999702085172838,
      "loss": 5.9119,
      "mean_token_accuracy": 0.15909691900014877,
      "num_tokens": 3526845.0,
      "step": 1535
    },
    {
      "entropy": 6.070224285125732,
      "epoch": 0.14793467819404418,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004999696480250929,
      "loss": 5.8893,
      "mean_token_accuracy": 0.15973274111747743,
      "num_tokens": 3538810.0,
      "step": 1540
    },
    {
      "entropy": 6.158443641662598,
      "epoch": 0.1484149855907781,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999690823097747,
      "loss": 5.8598,
      "mean_token_accuracy": 0.16933712512254714,
      "num_tokens": 3549570.0,
      "step": 1545
    },
    {
      "entropy": 5.8632913589477536,
      "epoch": 0.148895292987512,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999685113713426,
      "loss": 5.7738,
      "mean_token_accuracy": 0.1719530537724495,
      "num_tokens": 3560937.0,
      "step": 1550
    },
    {
      "entropy": 6.06872525215149,
      "epoch": 0.1493756003842459,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999679352098096,
      "loss": 5.8384,
      "mean_token_accuracy": 0.15944236814975737,
      "num_tokens": 3573603.0,
      "step": 1555
    },
    {
      "entropy": 5.972690153121948,
      "epoch": 0.14985590778097982,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999673538251891,
      "loss": 5.8561,
      "mean_token_accuracy": 0.15244044810533525,
      "num_tokens": 3586034.0,
      "step": 1560
    },
    {
      "entropy": 6.002362155914307,
      "epoch": 0.15033621517771373,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999667672174947,
      "loss": 5.7819,
      "mean_token_accuracy": 0.16955426633358,
      "num_tokens": 3597305.0,
      "step": 1565
    },
    {
      "entropy": 6.131285715103149,
      "epoch": 0.15081652257444764,
      "grad_norm": 1.046875,
      "learning_rate": 0.00049996617538674,
      "loss": 5.9327,
      "mean_token_accuracy": 0.15085813403129578,
      "num_tokens": 3608758.0,
      "step": 1570
    },
    {
      "entropy": 6.118503046035767,
      "epoch": 0.15129682997118155,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999655783329386,
      "loss": 5.9296,
      "mean_token_accuracy": 0.15356628447771073,
      "num_tokens": 3621508.0,
      "step": 1575
    },
    {
      "entropy": 6.0756830215454105,
      "epoch": 0.15177713736791545,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999649760561046,
      "loss": 5.891,
      "mean_token_accuracy": 0.1630493327975273,
      "num_tokens": 3634826.0,
      "step": 1580
    },
    {
      "entropy": 5.9437319278717045,
      "epoch": 0.15225744476464936,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999643685562519,
      "loss": 5.7229,
      "mean_token_accuracy": 0.17993323653936386,
      "num_tokens": 3646427.0,
      "step": 1585
    },
    {
      "entropy": 5.94010214805603,
      "epoch": 0.15273775216138327,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999637558333945,
      "loss": 5.8497,
      "mean_token_accuracy": 0.16093539148569108,
      "num_tokens": 3658157.0,
      "step": 1590
    },
    {
      "entropy": 6.139387130737305,
      "epoch": 0.15321805955811718,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999631378875467,
      "loss": 5.8908,
      "mean_token_accuracy": 0.1610818549990654,
      "num_tokens": 3670695.0,
      "step": 1595
    },
    {
      "entropy": 5.97409348487854,
      "epoch": 0.15369836695485112,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999625147187228,
      "loss": 5.809,
      "mean_token_accuracy": 0.15726363956928252,
      "num_tokens": 3683379.0,
      "step": 1600
    },
    {
      "entropy": 5.979640197753906,
      "epoch": 0.15417867435158503,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999618863269373,
      "loss": 5.7012,
      "mean_token_accuracy": 0.17374149709939957,
      "num_tokens": 3694411.0,
      "step": 1605
    },
    {
      "entropy": 5.935902690887451,
      "epoch": 0.15465898174831894,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004999612527122049,
      "loss": 5.762,
      "mean_token_accuracy": 0.1684124767780304,
      "num_tokens": 3703833.0,
      "step": 1610
    },
    {
      "entropy": 6.048168516159057,
      "epoch": 0.15513928914505284,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999606138745402,
      "loss": 5.8406,
      "mean_token_accuracy": 0.1746752843260765,
      "num_tokens": 3714702.0,
      "step": 1615
    },
    {
      "entropy": 6.089325952529907,
      "epoch": 0.15561959654178675,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999599698139581,
      "loss": 5.9244,
      "mean_token_accuracy": 0.1633257105946541,
      "num_tokens": 3726000.0,
      "step": 1620
    },
    {
      "entropy": 6.040888738632202,
      "epoch": 0.15609990393852066,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999593205304734,
      "loss": 5.7988,
      "mean_token_accuracy": 0.15986983478069305,
      "num_tokens": 3737681.0,
      "step": 1625
    },
    {
      "entropy": 6.123997163772583,
      "epoch": 0.15658021133525457,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999586660241012,
      "loss": 5.9075,
      "mean_token_accuracy": 0.16165615245699883,
      "num_tokens": 3749477.0,
      "step": 1630
    },
    {
      "entropy": 6.087912082672119,
      "epoch": 0.15706051873198848,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999580062948569,
      "loss": 5.8909,
      "mean_token_accuracy": 0.16025447547435762,
      "num_tokens": 3759946.0,
      "step": 1635
    },
    {
      "entropy": 5.989850997924805,
      "epoch": 0.1575408261287224,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999573413427556,
      "loss": 5.7826,
      "mean_token_accuracy": 0.16652001142501832,
      "num_tokens": 3770147.0,
      "step": 1640
    },
    {
      "entropy": 5.966318368911743,
      "epoch": 0.1580211335254563,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999566711678128,
      "loss": 5.7355,
      "mean_token_accuracy": 0.16983530521392823,
      "num_tokens": 3781028.0,
      "step": 1645
    },
    {
      "entropy": 6.011458206176758,
      "epoch": 0.1585014409221902,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999559957700442,
      "loss": 5.8065,
      "mean_token_accuracy": 0.15967805832624435,
      "num_tokens": 3792529.0,
      "step": 1650
    },
    {
      "entropy": 6.144451379776001,
      "epoch": 0.15898174831892412,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999553151494653,
      "loss": 5.9237,
      "mean_token_accuracy": 0.16285151839256287,
      "num_tokens": 3804630.0,
      "step": 1655
    },
    {
      "entropy": 5.864691495895386,
      "epoch": 0.15946205571565802,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999546293060919,
      "loss": 5.6858,
      "mean_token_accuracy": 0.1718939796090126,
      "num_tokens": 3815019.0,
      "step": 1660
    },
    {
      "entropy": 5.973699855804443,
      "epoch": 0.15994236311239193,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00049995393823994,
      "loss": 5.7936,
      "mean_token_accuracy": 0.16477740556001663,
      "num_tokens": 3825648.0,
      "step": 1665
    },
    {
      "entropy": 6.009506607055664,
      "epoch": 0.16042267050912584,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999532419510255,
      "loss": 5.7815,
      "mean_token_accuracy": 0.16592843681573868,
      "num_tokens": 3836779.0,
      "step": 1670
    },
    {
      "entropy": 5.950329399108886,
      "epoch": 0.16090297790585975,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499952540439365,
      "loss": 5.8272,
      "mean_token_accuracy": 0.16188324987888336,
      "num_tokens": 3848350.0,
      "step": 1675
    },
    {
      "entropy": 6.050455141067505,
      "epoch": 0.16138328530259366,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999518337049743,
      "loss": 5.8283,
      "mean_token_accuracy": 0.16736475825309755,
      "num_tokens": 3859792.0,
      "step": 1680
    },
    {
      "entropy": 5.972486734390259,
      "epoch": 0.16186359269932757,
      "grad_norm": 1.0,
      "learning_rate": 0.00049995112174787,
      "loss": 5.7611,
      "mean_token_accuracy": 0.16538213193416595,
      "num_tokens": 3871539.0,
      "step": 1685
    },
    {
      "entropy": 5.93617525100708,
      "epoch": 0.16234390009606148,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999504045680687,
      "loss": 5.7937,
      "mean_token_accuracy": 0.16570636183023452,
      "num_tokens": 3885025.0,
      "step": 1690
    },
    {
      "entropy": 6.037606048583984,
      "epoch": 0.1628242074927954,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999496821655869,
      "loss": 5.9112,
      "mean_token_accuracy": 0.15885820835828782,
      "num_tokens": 3896519.0,
      "step": 1695
    },
    {
      "entropy": 5.961896991729736,
      "epoch": 0.1633045148895293,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999489545404414,
      "loss": 5.6758,
      "mean_token_accuracy": 0.16855804473161698,
      "num_tokens": 3907636.0,
      "step": 1700
    },
    {
      "entropy": 5.917847633361816,
      "epoch": 0.1637848222862632,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999482216926493,
      "loss": 5.8151,
      "mean_token_accuracy": 0.16696375161409377,
      "num_tokens": 3917746.0,
      "step": 1705
    },
    {
      "entropy": 5.967260789871216,
      "epoch": 0.1642651296829971,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999474836222273,
      "loss": 5.7212,
      "mean_token_accuracy": 0.17350309789180757,
      "num_tokens": 3928901.0,
      "step": 1710
    },
    {
      "entropy": 5.922950553894043,
      "epoch": 0.16474543707973102,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999467403291928,
      "loss": 5.8411,
      "mean_token_accuracy": 0.17054857611656188,
      "num_tokens": 3940094.0,
      "step": 1715
    },
    {
      "entropy": 5.958613348007202,
      "epoch": 0.16522574447646493,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999459918135628,
      "loss": 5.7596,
      "mean_token_accuracy": 0.16999174505472184,
      "num_tokens": 3952714.0,
      "step": 1720
    },
    {
      "entropy": 5.936737680435181,
      "epoch": 0.16570605187319884,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000499945238075355,
      "loss": 5.7413,
      "mean_token_accuracy": 0.17201962023973466,
      "num_tokens": 3963107.0,
      "step": 1725
    },
    {
      "entropy": 5.932821226119995,
      "epoch": 0.16618635926993275,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999444791145865,
      "loss": 5.6878,
      "mean_token_accuracy": 0.17226921916007995,
      "num_tokens": 3973454.0,
      "step": 1730
    },
    {
      "entropy": 5.937837934494018,
      "epoch": 0.16666666666666666,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004999437149312754,
      "loss": 5.8047,
      "mean_token_accuracy": 0.16127632111310958,
      "num_tokens": 3985352.0,
      "step": 1735
    },
    {
      "entropy": 6.028000020980835,
      "epoch": 0.16714697406340057,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499942945525439,
      "loss": 5.769,
      "mean_token_accuracy": 0.1692960739135742,
      "num_tokens": 3997088.0,
      "step": 1740
    },
    {
      "entropy": 5.931283950805664,
      "epoch": 0.16762728146013448,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999421708970954,
      "loss": 5.7969,
      "mean_token_accuracy": 0.16781435459852218,
      "num_tokens": 4008551.0,
      "step": 1745
    },
    {
      "entropy": 5.948990726470948,
      "epoch": 0.16810758885686838,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999413910462625,
      "loss": 5.8268,
      "mean_token_accuracy": 0.16457579284906387,
      "num_tokens": 4020886.0,
      "step": 1750
    },
    {
      "entropy": 6.006660127639771,
      "epoch": 0.1685878962536023,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999406059729586,
      "loss": 5.7652,
      "mean_token_accuracy": 0.1626640737056732,
      "num_tokens": 4032811.0,
      "step": 1755
    },
    {
      "entropy": 5.967091226577759,
      "epoch": 0.1690682036503362,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999398156772016,
      "loss": 5.747,
      "mean_token_accuracy": 0.17119821906089783,
      "num_tokens": 4044102.0,
      "step": 1760
    },
    {
      "entropy": 5.9318421363830565,
      "epoch": 0.16954851104707014,
      "grad_norm": 1.03125,
      "learning_rate": 0.00049993902015901,
      "loss": 5.7835,
      "mean_token_accuracy": 0.17211457341909409,
      "num_tokens": 4056849.0,
      "step": 1765
    },
    {
      "entropy": 5.943634080886841,
      "epoch": 0.17002881844380405,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999382194184023,
      "loss": 5.8128,
      "mean_token_accuracy": 0.16628441363573074,
      "num_tokens": 4068392.0,
      "step": 1770
    },
    {
      "entropy": 5.970344209671021,
      "epoch": 0.17050912584053796,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999374134553972,
      "loss": 5.6582,
      "mean_token_accuracy": 0.17245645970106124,
      "num_tokens": 4078949.0,
      "step": 1775
    },
    {
      "entropy": 5.9032188892364506,
      "epoch": 0.17098943323727187,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999366022700131,
      "loss": 5.7916,
      "mean_token_accuracy": 0.16414022147655488,
      "num_tokens": 4091432.0,
      "step": 1780
    },
    {
      "entropy": 5.925055408477784,
      "epoch": 0.17146974063400577,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999357858622691,
      "loss": 5.6565,
      "mean_token_accuracy": 0.17122484147548675,
      "num_tokens": 4103001.0,
      "step": 1785
    },
    {
      "entropy": 5.929124546051026,
      "epoch": 0.17195004803073968,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999349642321842,
      "loss": 5.7667,
      "mean_token_accuracy": 0.1666597157716751,
      "num_tokens": 4113570.0,
      "step": 1790
    },
    {
      "entropy": 5.980472898483276,
      "epoch": 0.1724303554274736,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999341373797772,
      "loss": 5.8077,
      "mean_token_accuracy": 0.16480496674776077,
      "num_tokens": 4125658.0,
      "step": 1795
    },
    {
      "entropy": 5.8866361618042,
      "epoch": 0.1729106628242075,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999333053050675,
      "loss": 5.6989,
      "mean_token_accuracy": 0.16817544847726823,
      "num_tokens": 4137424.0,
      "step": 1800
    },
    {
      "entropy": 5.941749000549317,
      "epoch": 0.1733909702209414,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999324680080744,
      "loss": 5.7376,
      "mean_token_accuracy": 0.16792946606874465,
      "num_tokens": 4149772.0,
      "step": 1805
    },
    {
      "entropy": 5.8657555103302,
      "epoch": 0.17387127761767532,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999316254888172,
      "loss": 5.7335,
      "mean_token_accuracy": 0.17179457992315292,
      "num_tokens": 4162246.0,
      "step": 1810
    },
    {
      "entropy": 5.883719253540039,
      "epoch": 0.17435158501440923,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999307777473157,
      "loss": 5.701,
      "mean_token_accuracy": 0.17882005274295806,
      "num_tokens": 4173449.0,
      "step": 1815
    },
    {
      "entropy": 5.866553688049317,
      "epoch": 0.17483189241114314,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999299247835893,
      "loss": 5.7518,
      "mean_token_accuracy": 0.16288111954927445,
      "num_tokens": 4185560.0,
      "step": 1820
    },
    {
      "entropy": 5.952811813354492,
      "epoch": 0.17531219980787704,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499929066597658,
      "loss": 5.8043,
      "mean_token_accuracy": 0.15593840181827545,
      "num_tokens": 4197843.0,
      "step": 1825
    },
    {
      "entropy": 5.971073150634766,
      "epoch": 0.17579250720461095,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999282031895418,
      "loss": 5.6586,
      "mean_token_accuracy": 0.16830604374408722,
      "num_tokens": 4210009.0,
      "step": 1830
    },
    {
      "entropy": 5.914163827896118,
      "epoch": 0.17627281460134486,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999273345592604,
      "loss": 5.8931,
      "mean_token_accuracy": 0.15920523107051848,
      "num_tokens": 4222213.0,
      "step": 1835
    },
    {
      "entropy": 5.962227725982666,
      "epoch": 0.17675312199807877,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999264607068343,
      "loss": 5.7543,
      "mean_token_accuracy": 0.16375723034143447,
      "num_tokens": 4234159.0,
      "step": 1840
    },
    {
      "entropy": 5.948575210571289,
      "epoch": 0.17723342939481268,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999255816322837,
      "loss": 5.8124,
      "mean_token_accuracy": 0.1582082837820053,
      "num_tokens": 4246358.0,
      "step": 1845
    },
    {
      "entropy": 6.0558061599731445,
      "epoch": 0.1777137367915466,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499924697335629,
      "loss": 5.8779,
      "mean_token_accuracy": 0.16516841500997542,
      "num_tokens": 4256698.0,
      "step": 1850
    },
    {
      "entropy": 5.875409936904907,
      "epoch": 0.1781940441882805,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999238078168906,
      "loss": 5.6677,
      "mean_token_accuracy": 0.18177317082881927,
      "num_tokens": 4268007.0,
      "step": 1855
    },
    {
      "entropy": 5.906381893157959,
      "epoch": 0.1786743515850144,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999229130760894,
      "loss": 5.7785,
      "mean_token_accuracy": 0.16978150457143784,
      "num_tokens": 4278583.0,
      "step": 1860
    },
    {
      "entropy": 6.020100021362305,
      "epoch": 0.17915465898174832,
      "grad_norm": 1.0,
      "learning_rate": 0.000499922013113246,
      "loss": 5.8081,
      "mean_token_accuracy": 0.1636409193277359,
      "num_tokens": 4290267.0,
      "step": 1865
    },
    {
      "entropy": 5.916716623306274,
      "epoch": 0.17963496637848222,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999211079283814,
      "loss": 5.7066,
      "mean_token_accuracy": 0.17464031428098678,
      "num_tokens": 4302177.0,
      "step": 1870
    },
    {
      "entropy": 5.793494367599488,
      "epoch": 0.18011527377521613,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004999201975215164,
      "loss": 5.6668,
      "mean_token_accuracy": 0.17644754350185393,
      "num_tokens": 4312420.0,
      "step": 1875
    },
    {
      "entropy": 5.989563703536987,
      "epoch": 0.18059558117195004,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999192818926725,
      "loss": 5.7904,
      "mean_token_accuracy": 0.1659339353442192,
      "num_tokens": 4324050.0,
      "step": 1880
    },
    {
      "entropy": 5.9014684677124025,
      "epoch": 0.18107588856868395,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999183610418706,
      "loss": 5.6492,
      "mean_token_accuracy": 0.17169519364833832,
      "num_tokens": 4335336.0,
      "step": 1885
    },
    {
      "entropy": 5.792817640304565,
      "epoch": 0.18155619596541786,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999174349691322,
      "loss": 5.6589,
      "mean_token_accuracy": 0.17302853614091873,
      "num_tokens": 4346239.0,
      "step": 1890
    },
    {
      "entropy": 5.912134981155395,
      "epoch": 0.18203650336215177,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999165036744788,
      "loss": 5.7356,
      "mean_token_accuracy": 0.16781715154647828,
      "num_tokens": 4357195.0,
      "step": 1895
    },
    {
      "entropy": 5.846279335021973,
      "epoch": 0.18251681075888568,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999155671579322,
      "loss": 5.6776,
      "mean_token_accuracy": 0.17427498400211333,
      "num_tokens": 4368905.0,
      "step": 1900
    },
    {
      "entropy": 5.962751770019532,
      "epoch": 0.1829971181556196,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499914625419514,
      "loss": 5.7554,
      "mean_token_accuracy": 0.16602100878953935,
      "num_tokens": 4381708.0,
      "step": 1905
    },
    {
      "entropy": 5.798335075378418,
      "epoch": 0.1834774255523535,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999136784592459,
      "loss": 5.6004,
      "mean_token_accuracy": 0.18304169028997422,
      "num_tokens": 4393199.0,
      "step": 1910
    },
    {
      "entropy": 5.928949308395386,
      "epoch": 0.1839577329490874,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999127262771502,
      "loss": 5.8242,
      "mean_token_accuracy": 0.16348036080598832,
      "num_tokens": 4403986.0,
      "step": 1915
    },
    {
      "entropy": 5.974527883529663,
      "epoch": 0.1844380403458213,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999117688732487,
      "loss": 5.677,
      "mean_token_accuracy": 0.1677529901266098,
      "num_tokens": 4415574.0,
      "step": 1920
    },
    {
      "entropy": 5.840640020370484,
      "epoch": 0.18491834774255522,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999108062475638,
      "loss": 5.7085,
      "mean_token_accuracy": 0.17418631464242934,
      "num_tokens": 4427579.0,
      "step": 1925
    },
    {
      "entropy": 5.920491027832031,
      "epoch": 0.18539865513928913,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499909838400118,
      "loss": 5.7006,
      "mean_token_accuracy": 0.1702801316976547,
      "num_tokens": 4440045.0,
      "step": 1930
    },
    {
      "entropy": 5.875643730163574,
      "epoch": 0.18587896253602307,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999088653309334,
      "loss": 5.7865,
      "mean_token_accuracy": 0.16631472557783128,
      "num_tokens": 4450129.0,
      "step": 1935
    },
    {
      "entropy": 5.843389463424683,
      "epoch": 0.18635926993275698,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999078870400329,
      "loss": 5.6529,
      "mean_token_accuracy": 0.1726601541042328,
      "num_tokens": 4461378.0,
      "step": 1940
    },
    {
      "entropy": 5.898785829544067,
      "epoch": 0.18683957732949089,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999069035274391,
      "loss": 5.6755,
      "mean_token_accuracy": 0.17404537796974182,
      "num_tokens": 4472249.0,
      "step": 1945
    },
    {
      "entropy": 5.829293203353882,
      "epoch": 0.1873198847262248,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999059147931747,
      "loss": 5.7345,
      "mean_token_accuracy": 0.16566374748945237,
      "num_tokens": 4484243.0,
      "step": 1950
    },
    {
      "entropy": 5.932005405426025,
      "epoch": 0.1878001921229587,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999049208372629,
      "loss": 5.6561,
      "mean_token_accuracy": 0.1711133375763893,
      "num_tokens": 4495776.0,
      "step": 1955
    },
    {
      "entropy": 5.815832805633545,
      "epoch": 0.1882804995196926,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999039216597267,
      "loss": 5.731,
      "mean_token_accuracy": 0.17108235955238343,
      "num_tokens": 4508581.0,
      "step": 1960
    },
    {
      "entropy": 5.915356016159057,
      "epoch": 0.18876080691642652,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999029172605892,
      "loss": 5.6688,
      "mean_token_accuracy": 0.16956983059644698,
      "num_tokens": 4519753.0,
      "step": 1965
    },
    {
      "entropy": 5.870398330688476,
      "epoch": 0.18924111431316043,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999019076398738,
      "loss": 5.7318,
      "mean_token_accuracy": 0.1767261579632759,
      "num_tokens": 4529957.0,
      "step": 1970
    },
    {
      "entropy": 5.917528676986694,
      "epoch": 0.18972142170989434,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499900892797604,
      "loss": 5.7002,
      "mean_token_accuracy": 0.17024850398302077,
      "num_tokens": 4541181.0,
      "step": 1975
    },
    {
      "entropy": 5.877421092987061,
      "epoch": 0.19020172910662825,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998998727338031,
      "loss": 5.6726,
      "mean_token_accuracy": 0.17142467349767684,
      "num_tokens": 4552235.0,
      "step": 1980
    },
    {
      "entropy": 5.848327493667602,
      "epoch": 0.19068203650336216,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998988474484952,
      "loss": 5.6808,
      "mean_token_accuracy": 0.17754830718040465,
      "num_tokens": 4564045.0,
      "step": 1985
    },
    {
      "entropy": 5.832021617889405,
      "epoch": 0.19116234390009607,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998978169417038,
      "loss": 5.587,
      "mean_token_accuracy": 0.18072767406702042,
      "num_tokens": 4575141.0,
      "step": 1990
    },
    {
      "entropy": 5.814238882064819,
      "epoch": 0.19164265129682997,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998967812134529,
      "loss": 5.6878,
      "mean_token_accuracy": 0.1720028355717659,
      "num_tokens": 4586491.0,
      "step": 1995
    },
    {
      "entropy": 5.894739389419556,
      "epoch": 0.19212295869356388,
      "grad_norm": 1.125,
      "learning_rate": 0.0004998957402637664,
      "loss": 5.6897,
      "mean_token_accuracy": 0.17228867560625077,
      "num_tokens": 4598421.0,
      "step": 2000
    },
    {
      "entropy": 5.837182331085205,
      "epoch": 0.1926032660902978,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004998946940926687,
      "loss": 5.6102,
      "mean_token_accuracy": 0.18361353129148483,
      "num_tokens": 4609688.0,
      "step": 2005
    },
    {
      "entropy": 5.837462520599365,
      "epoch": 0.1930835734870317,
      "grad_norm": 1.0,
      "learning_rate": 0.000499893642700184,
      "loss": 5.7614,
      "mean_token_accuracy": 0.1679168865084648,
      "num_tokens": 4621169.0,
      "step": 2010
    },
    {
      "entropy": 5.882701253890991,
      "epoch": 0.1935638808837656,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998925860863368,
      "loss": 5.6982,
      "mean_token_accuracy": 0.1753264084458351,
      "num_tokens": 4632512.0,
      "step": 2015
    },
    {
      "entropy": 5.8722984313964846,
      "epoch": 0.19404418828049952,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998915242511516,
      "loss": 5.7468,
      "mean_token_accuracy": 0.16945238560438156,
      "num_tokens": 4644356.0,
      "step": 2020
    },
    {
      "entropy": 5.9419224739074705,
      "epoch": 0.19452449567723343,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004998904571946528,
      "loss": 5.7306,
      "mean_token_accuracy": 0.16765568107366563,
      "num_tokens": 4655601.0,
      "step": 2025
    },
    {
      "entropy": 5.848211431503296,
      "epoch": 0.19500480307396734,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004998893849168655,
      "loss": 5.7448,
      "mean_token_accuracy": 0.16884764283895493,
      "num_tokens": 4666813.0,
      "step": 2030
    },
    {
      "entropy": 5.902902412414551,
      "epoch": 0.19548511047070125,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998883074178144,
      "loss": 5.7389,
      "mean_token_accuracy": 0.16699665933847427,
      "num_tokens": 4679123.0,
      "step": 2035
    },
    {
      "entropy": 5.874791812896729,
      "epoch": 0.19596541786743515,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004998872246975247,
      "loss": 5.6311,
      "mean_token_accuracy": 0.17366175949573517,
      "num_tokens": 4690485.0,
      "step": 2040
    },
    {
      "entropy": 5.7398899555206295,
      "epoch": 0.19644572526416906,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998861367560213,
      "loss": 5.6471,
      "mean_token_accuracy": 0.17550230622291565,
      "num_tokens": 4702573.0,
      "step": 2045
    },
    {
      "entropy": 5.889365386962891,
      "epoch": 0.19692603266090297,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004998850435933296,
      "loss": 5.6907,
      "mean_token_accuracy": 0.16575469076633453,
      "num_tokens": 4715653.0,
      "step": 2050
    },
    {
      "entropy": 5.819795274734497,
      "epoch": 0.19740634005763688,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998839452094749,
      "loss": 5.6751,
      "mean_token_accuracy": 0.1743623748421669,
      "num_tokens": 4726051.0,
      "step": 2055
    },
    {
      "entropy": 5.900219297409057,
      "epoch": 0.1978866474543708,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998828416044829,
      "loss": 5.8183,
      "mean_token_accuracy": 0.16225597262382507,
      "num_tokens": 4738176.0,
      "step": 2060
    },
    {
      "entropy": 5.914799547195434,
      "epoch": 0.1983669548511047,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499881732778379,
      "loss": 5.7246,
      "mean_token_accuracy": 0.1708556368947029,
      "num_tokens": 4749547.0,
      "step": 2065
    },
    {
      "entropy": 5.816590356826782,
      "epoch": 0.1988472622478386,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499880618731189,
      "loss": 5.6392,
      "mean_token_accuracy": 0.17196280360221863,
      "num_tokens": 4761967.0,
      "step": 2070
    },
    {
      "entropy": 5.815549898147583,
      "epoch": 0.19932756964457252,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998794994629388,
      "loss": 5.6193,
      "mean_token_accuracy": 0.17873595058918,
      "num_tokens": 4772847.0,
      "step": 2075
    },
    {
      "entropy": 5.846785545349121,
      "epoch": 0.19980787704130643,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998783749736545,
      "loss": 5.6845,
      "mean_token_accuracy": 0.17165183573961257,
      "num_tokens": 4785049.0,
      "step": 2080
    },
    {
      "entropy": 5.838068294525146,
      "epoch": 0.20028818443804033,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998772452633619,
      "loss": 5.7118,
      "mean_token_accuracy": 0.16461516320705413,
      "num_tokens": 4797484.0,
      "step": 2085
    },
    {
      "entropy": 5.804093217849731,
      "epoch": 0.20076849183477424,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998761103320876,
      "loss": 5.6499,
      "mean_token_accuracy": 0.16633011251688004,
      "num_tokens": 4809422.0,
      "step": 2090
    },
    {
      "entropy": 5.889832544326782,
      "epoch": 0.20124879923150815,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998749701798577,
      "loss": 5.8256,
      "mean_token_accuracy": 0.1649285450577736,
      "num_tokens": 4822340.0,
      "step": 2095
    },
    {
      "entropy": 5.832509899139405,
      "epoch": 0.2017291066282421,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998738248066986,
      "loss": 5.6039,
      "mean_token_accuracy": 0.17457860857248306,
      "num_tokens": 4833382.0,
      "step": 2100
    },
    {
      "entropy": 5.753271865844726,
      "epoch": 0.202209414024976,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998726742126372,
      "loss": 5.6169,
      "mean_token_accuracy": 0.17748918682336806,
      "num_tokens": 4843916.0,
      "step": 2105
    },
    {
      "entropy": 5.860424470901489,
      "epoch": 0.2026897214217099,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998715183976999,
      "loss": 5.6176,
      "mean_token_accuracy": 0.17942588478326799,
      "num_tokens": 4855420.0,
      "step": 2110
    },
    {
      "entropy": 5.8845704078674315,
      "epoch": 0.20317002881844382,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998703573619137,
      "loss": 5.6902,
      "mean_token_accuracy": 0.16863269060850145,
      "num_tokens": 4866540.0,
      "step": 2115
    },
    {
      "entropy": 5.844192361831665,
      "epoch": 0.20365033621517772,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998691911053056,
      "loss": 5.6375,
      "mean_token_accuracy": 0.17058192938566208,
      "num_tokens": 4878860.0,
      "step": 2120
    },
    {
      "entropy": 5.910974168777466,
      "epoch": 0.20413064361191163,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998680196279026,
      "loss": 5.764,
      "mean_token_accuracy": 0.1690535604953766,
      "num_tokens": 4889692.0,
      "step": 2125
    },
    {
      "entropy": 5.937790060043335,
      "epoch": 0.20461095100864554,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998668429297319,
      "loss": 5.7047,
      "mean_token_accuracy": 0.1704765021800995,
      "num_tokens": 4901337.0,
      "step": 2130
    },
    {
      "entropy": 5.702803325653076,
      "epoch": 0.20509125840537945,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998656610108208,
      "loss": 5.6129,
      "mean_token_accuracy": 0.17906004935503006,
      "num_tokens": 4912340.0,
      "step": 2135
    },
    {
      "entropy": 5.85637321472168,
      "epoch": 0.20557156580211336,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998644738711969,
      "loss": 5.5885,
      "mean_token_accuracy": 0.1815409779548645,
      "num_tokens": 4923429.0,
      "step": 2140
    },
    {
      "entropy": 5.828443336486816,
      "epoch": 0.20605187319884727,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998632815108874,
      "loss": 5.6722,
      "mean_token_accuracy": 0.17124899476766586,
      "num_tokens": 4934826.0,
      "step": 2145
    },
    {
      "entropy": 5.8796604633331295,
      "epoch": 0.20653218059558118,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998620839299203,
      "loss": 5.7412,
      "mean_token_accuracy": 0.17052844166755676,
      "num_tokens": 4945189.0,
      "step": 2150
    },
    {
      "entropy": 5.883285331726074,
      "epoch": 0.2070124879923151,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998608811283233,
      "loss": 5.6215,
      "mean_token_accuracy": 0.17871106863021852,
      "num_tokens": 4956060.0,
      "step": 2155
    },
    {
      "entropy": 5.798979902267456,
      "epoch": 0.207492795389049,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998596731061244,
      "loss": 5.6482,
      "mean_token_accuracy": 0.17223062068223954,
      "num_tokens": 4967710.0,
      "step": 2160
    },
    {
      "entropy": 5.831351280212402,
      "epoch": 0.2079731027857829,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998584598633516,
      "loss": 5.6595,
      "mean_token_accuracy": 0.1707835465669632,
      "num_tokens": 4978716.0,
      "step": 2165
    },
    {
      "entropy": 5.831381130218506,
      "epoch": 0.2084534101825168,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998572414000329,
      "loss": 5.7381,
      "mean_token_accuracy": 0.16959594637155534,
      "num_tokens": 4990742.0,
      "step": 2170
    },
    {
      "entropy": 5.919949722290039,
      "epoch": 0.20893371757925072,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004998560177161969,
      "loss": 5.6786,
      "mean_token_accuracy": 0.17193486988544465,
      "num_tokens": 5001217.0,
      "step": 2175
    },
    {
      "entropy": 5.807414817810058,
      "epoch": 0.20941402497598463,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998547888118718,
      "loss": 5.6645,
      "mean_token_accuracy": 0.17050760686397554,
      "num_tokens": 5012750.0,
      "step": 2180
    },
    {
      "entropy": 5.836668300628662,
      "epoch": 0.20989433237271854,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998535546870862,
      "loss": 5.6711,
      "mean_token_accuracy": 0.18007341921329498,
      "num_tokens": 5023818.0,
      "step": 2185
    },
    {
      "entropy": 5.917913913726807,
      "epoch": 0.21037463976945245,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998523153418687,
      "loss": 5.6925,
      "mean_token_accuracy": 0.1711179181933403,
      "num_tokens": 5035060.0,
      "step": 2190
    },
    {
      "entropy": 5.815863227844238,
      "epoch": 0.21085494716618636,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998510707762481,
      "loss": 5.733,
      "mean_token_accuracy": 0.16495421826839446,
      "num_tokens": 5046780.0,
      "step": 2195
    },
    {
      "entropy": 5.850568675994873,
      "epoch": 0.21133525456292027,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998498209902533,
      "loss": 5.6982,
      "mean_token_accuracy": 0.16920250058174133,
      "num_tokens": 5059152.0,
      "step": 2200
    },
    {
      "entropy": 5.817641115188598,
      "epoch": 0.21181556195965417,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998485659839134,
      "loss": 5.654,
      "mean_token_accuracy": 0.17348035871982576,
      "num_tokens": 5069971.0,
      "step": 2205
    },
    {
      "entropy": 5.805312061309815,
      "epoch": 0.21229586935638808,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998473057572575,
      "loss": 5.6181,
      "mean_token_accuracy": 0.17668834328651428,
      "num_tokens": 5080640.0,
      "step": 2210
    },
    {
      "entropy": 5.8015978813171385,
      "epoch": 0.212776176753122,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998460403103146,
      "loss": 5.6379,
      "mean_token_accuracy": 0.17368897497653962,
      "num_tokens": 5091240.0,
      "step": 2215
    },
    {
      "entropy": 5.724646091461182,
      "epoch": 0.2132564841498559,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998447696431146,
      "loss": 5.6556,
      "mean_token_accuracy": 0.17121760100126265,
      "num_tokens": 5102920.0,
      "step": 2220
    },
    {
      "entropy": 5.8558906555175785,
      "epoch": 0.2137367915465898,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998434937556865,
      "loss": 5.6784,
      "mean_token_accuracy": 0.16950192004442216,
      "num_tokens": 5114240.0,
      "step": 2225
    },
    {
      "entropy": 5.941771984100342,
      "epoch": 0.21421709894332372,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998422126480602,
      "loss": 5.8965,
      "mean_token_accuracy": 0.16069666296243668,
      "num_tokens": 5126653.0,
      "step": 2230
    },
    {
      "entropy": 5.9492322444915775,
      "epoch": 0.21469740634005763,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998409263202653,
      "loss": 5.7472,
      "mean_token_accuracy": 0.1665404573082924,
      "num_tokens": 5138778.0,
      "step": 2235
    },
    {
      "entropy": 5.845252323150635,
      "epoch": 0.21517771373679154,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998396347723318,
      "loss": 5.5971,
      "mean_token_accuracy": 0.17824865579605104,
      "num_tokens": 5148670.0,
      "step": 2240
    },
    {
      "entropy": 5.624603366851806,
      "epoch": 0.21565802113352545,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998383380042895,
      "loss": 5.6491,
      "mean_token_accuracy": 0.17016488909721375,
      "num_tokens": 5159617.0,
      "step": 2245
    },
    {
      "entropy": 5.844017457962036,
      "epoch": 0.21613832853025935,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998370360161688,
      "loss": 5.6985,
      "mean_token_accuracy": 0.17508894503116607,
      "num_tokens": 5171225.0,
      "step": 2250
    },
    {
      "entropy": 5.878959703445434,
      "epoch": 0.21661863592699326,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998357288079996,
      "loss": 5.7061,
      "mean_token_accuracy": 0.16998043954372405,
      "num_tokens": 5184586.0,
      "step": 2255
    },
    {
      "entropy": 5.834930086135865,
      "epoch": 0.21709894332372717,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998344163798125,
      "loss": 5.6633,
      "mean_token_accuracy": 0.16968356519937516,
      "num_tokens": 5196089.0,
      "step": 2260
    },
    {
      "entropy": 5.834982538223267,
      "epoch": 0.21757925072046108,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998330987316379,
      "loss": 5.6957,
      "mean_token_accuracy": 0.17092847377061843,
      "num_tokens": 5207581.0,
      "step": 2265
    },
    {
      "entropy": 5.78864893913269,
      "epoch": 0.21805955811719502,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004998317758635062,
      "loss": 5.6454,
      "mean_token_accuracy": 0.17239997535943985,
      "num_tokens": 5219108.0,
      "step": 2270
    },
    {
      "entropy": 5.828027296066284,
      "epoch": 0.21853986551392893,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998304477754484,
      "loss": 5.5916,
      "mean_token_accuracy": 0.17449767142534256,
      "num_tokens": 5231137.0,
      "step": 2275
    },
    {
      "entropy": 5.926539087295533,
      "epoch": 0.21902017291066284,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998291144674952,
      "loss": 5.6633,
      "mean_token_accuracy": 0.1777758538722992,
      "num_tokens": 5243528.0,
      "step": 2280
    },
    {
      "entropy": 5.6308393478393555,
      "epoch": 0.21950048030739674,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998277759396776,
      "loss": 5.6212,
      "mean_token_accuracy": 0.17822325676679612,
      "num_tokens": 5254869.0,
      "step": 2285
    },
    {
      "entropy": 5.7623279094696045,
      "epoch": 0.21998078770413065,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998264321920265,
      "loss": 5.6039,
      "mean_token_accuracy": 0.1737493708729744,
      "num_tokens": 5265370.0,
      "step": 2290
    },
    {
      "entropy": 5.752946853637695,
      "epoch": 0.22046109510086456,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998250832245734,
      "loss": 5.5688,
      "mean_token_accuracy": 0.17435249984264373,
      "num_tokens": 5277812.0,
      "step": 2295
    },
    {
      "entropy": 5.7676074504852295,
      "epoch": 0.22094140249759847,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998237290373494,
      "loss": 5.5722,
      "mean_token_accuracy": 0.1777767837047577,
      "num_tokens": 5289107.0,
      "step": 2300
    },
    {
      "entropy": 5.785541439056397,
      "epoch": 0.22142170989433238,
      "grad_norm": 1.125,
      "learning_rate": 0.000499822369630386,
      "loss": 5.6088,
      "mean_token_accuracy": 0.16769287288188933,
      "num_tokens": 5301489.0,
      "step": 2305
    },
    {
      "entropy": 5.68263053894043,
      "epoch": 0.2219020172910663,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998210050037148,
      "loss": 5.5856,
      "mean_token_accuracy": 0.1790359228849411,
      "num_tokens": 5313273.0,
      "step": 2310
    },
    {
      "entropy": 5.8171937465667725,
      "epoch": 0.2223823246878002,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998196351573674,
      "loss": 5.6136,
      "mean_token_accuracy": 0.17444271743297576,
      "num_tokens": 5324227.0,
      "step": 2315
    },
    {
      "entropy": 5.752804708480835,
      "epoch": 0.2228626320845341,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998182600913757,
      "loss": 5.5654,
      "mean_token_accuracy": 0.17911654710769653,
      "num_tokens": 5335138.0,
      "step": 2320
    },
    {
      "entropy": 5.734329175949097,
      "epoch": 0.22334293948126802,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998168798057715,
      "loss": 5.635,
      "mean_token_accuracy": 0.17262459695339202,
      "num_tokens": 5345652.0,
      "step": 2325
    },
    {
      "entropy": 5.882148027420044,
      "epoch": 0.22382324687800192,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499815494300587,
      "loss": 5.7368,
      "mean_token_accuracy": 0.16915619671344756,
      "num_tokens": 5357999.0,
      "step": 2330
    },
    {
      "entropy": 5.817476558685303,
      "epoch": 0.22430355427473583,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998141035758542,
      "loss": 5.6377,
      "mean_token_accuracy": 0.17220452576875686,
      "num_tokens": 5369032.0,
      "step": 2335
    },
    {
      "entropy": 5.754540586471558,
      "epoch": 0.22478386167146974,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998127076316054,
      "loss": 5.5625,
      "mean_token_accuracy": 0.17929039448499678,
      "num_tokens": 5380692.0,
      "step": 2340
    },
    {
      "entropy": 5.843355941772461,
      "epoch": 0.22526416906820365,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998113064678734,
      "loss": 5.6892,
      "mean_token_accuracy": 0.16779275834560395,
      "num_tokens": 5392320.0,
      "step": 2345
    },
    {
      "entropy": 5.837249851226806,
      "epoch": 0.22574447646493756,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004998099000846901,
      "loss": 5.6634,
      "mean_token_accuracy": 0.17556915134191514,
      "num_tokens": 5404374.0,
      "step": 2350
    },
    {
      "entropy": 5.825764131546021,
      "epoch": 0.22622478386167147,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998084884820887,
      "loss": 5.6715,
      "mean_token_accuracy": 0.17142031490802764,
      "num_tokens": 5415471.0,
      "step": 2355
    },
    {
      "entropy": 5.676828861236572,
      "epoch": 0.22670509125840538,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004998070716601016,
      "loss": 5.5019,
      "mean_token_accuracy": 0.1765056163072586,
      "num_tokens": 5425971.0,
      "step": 2360
    },
    {
      "entropy": 5.79780216217041,
      "epoch": 0.2271853986551393,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998056496187618,
      "loss": 5.6161,
      "mean_token_accuracy": 0.1789500296115875,
      "num_tokens": 5437638.0,
      "step": 2365
    },
    {
      "entropy": 5.821310758590698,
      "epoch": 0.2276657060518732,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998042223581025,
      "loss": 5.6528,
      "mean_token_accuracy": 0.1725280314683914,
      "num_tokens": 5449463.0,
      "step": 2370
    },
    {
      "entropy": 5.799275064468384,
      "epoch": 0.2281460134486071,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004998027898781565,
      "loss": 5.7219,
      "mean_token_accuracy": 0.17100997567176818,
      "num_tokens": 5460972.0,
      "step": 2375
    },
    {
      "entropy": 5.948947763442993,
      "epoch": 0.228626320845341,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998013521789574,
      "loss": 5.7269,
      "mean_token_accuracy": 0.1655080497264862,
      "num_tokens": 5471575.0,
      "step": 2380
    },
    {
      "entropy": 5.700678491592408,
      "epoch": 0.22910662824207492,
      "grad_norm": 1.125,
      "learning_rate": 0.0004997999092605384,
      "loss": 5.5654,
      "mean_token_accuracy": 0.17265488654375077,
      "num_tokens": 5483041.0,
      "step": 2385
    },
    {
      "entropy": 5.7677713394165036,
      "epoch": 0.22958693563880883,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499798461122933,
      "loss": 5.6049,
      "mean_token_accuracy": 0.17628835141658783,
      "num_tokens": 5494040.0,
      "step": 2390
    },
    {
      "entropy": 5.69358229637146,
      "epoch": 0.23006724303554274,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997970077661748,
      "loss": 5.505,
      "mean_token_accuracy": 0.18428827673196793,
      "num_tokens": 5504217.0,
      "step": 2395
    },
    {
      "entropy": 5.819785404205322,
      "epoch": 0.23054755043227665,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997955491902977,
      "loss": 5.6779,
      "mean_token_accuracy": 0.17342882454395295,
      "num_tokens": 5517180.0,
      "step": 2400
    },
    {
      "entropy": 5.78384485244751,
      "epoch": 0.23102785782901056,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997940853953354,
      "loss": 5.5816,
      "mean_token_accuracy": 0.17583905160427094,
      "num_tokens": 5529626.0,
      "step": 2405
    },
    {
      "entropy": 5.810919857025146,
      "epoch": 0.23150816522574447,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499792616381322,
      "loss": 5.6114,
      "mean_token_accuracy": 0.1800064906477928,
      "num_tokens": 5540844.0,
      "step": 2410
    },
    {
      "entropy": 5.803003025054932,
      "epoch": 0.23198847262247838,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997911421482914,
      "loss": 5.5387,
      "mean_token_accuracy": 0.17990489304065704,
      "num_tokens": 5551732.0,
      "step": 2415
    },
    {
      "entropy": 5.806057739257812,
      "epoch": 0.23246878001921228,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499789662696278,
      "loss": 5.6818,
      "mean_token_accuracy": 0.17363432347774505,
      "num_tokens": 5562692.0,
      "step": 2420
    },
    {
      "entropy": 5.837599229812622,
      "epoch": 0.2329490874159462,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004997881780253162,
      "loss": 5.6658,
      "mean_token_accuracy": 0.17378875017166137,
      "num_tokens": 5574884.0,
      "step": 2425
    },
    {
      "entropy": 5.689996194839478,
      "epoch": 0.2334293948126801,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004997866881354403,
      "loss": 5.5141,
      "mean_token_accuracy": 0.1813263177871704,
      "num_tokens": 5585335.0,
      "step": 2430
    },
    {
      "entropy": 5.813428974151611,
      "epoch": 0.23390970220941404,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499785193026685,
      "loss": 5.666,
      "mean_token_accuracy": 0.1711764082312584,
      "num_tokens": 5597621.0,
      "step": 2435
    },
    {
      "entropy": 5.675594329833984,
      "epoch": 0.23439000960614795,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997836926990851,
      "loss": 5.5147,
      "mean_token_accuracy": 0.1748025432229042,
      "num_tokens": 5609795.0,
      "step": 2440
    },
    {
      "entropy": 5.757413482666015,
      "epoch": 0.23487031700288186,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997821871526752,
      "loss": 5.6207,
      "mean_token_accuracy": 0.17631727755069732,
      "num_tokens": 5621473.0,
      "step": 2445
    },
    {
      "entropy": 5.846551418304443,
      "epoch": 0.23535062439961577,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997806763874905,
      "loss": 5.7071,
      "mean_token_accuracy": 0.1744229331612587,
      "num_tokens": 5635064.0,
      "step": 2450
    },
    {
      "entropy": 5.882826089859009,
      "epoch": 0.23583093179634967,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997791604035659,
      "loss": 5.7063,
      "mean_token_accuracy": 0.1706729456782341,
      "num_tokens": 5647382.0,
      "step": 2455
    },
    {
      "entropy": 5.769627428054809,
      "epoch": 0.23631123919308358,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997776392009366,
      "loss": 5.5864,
      "mean_token_accuracy": 0.1790638193488121,
      "num_tokens": 5658417.0,
      "step": 2460
    },
    {
      "entropy": 5.74236798286438,
      "epoch": 0.2367915465898175,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997761127796381,
      "loss": 5.5629,
      "mean_token_accuracy": 0.1808903470635414,
      "num_tokens": 5672190.0,
      "step": 2465
    },
    {
      "entropy": 5.736099815368652,
      "epoch": 0.2372718539865514,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997745811397056,
      "loss": 5.5734,
      "mean_token_accuracy": 0.1817570447921753,
      "num_tokens": 5684259.0,
      "step": 2470
    },
    {
      "entropy": 5.679139423370361,
      "epoch": 0.2377521613832853,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997730442811748,
      "loss": 5.6394,
      "mean_token_accuracy": 0.17837475389242172,
      "num_tokens": 5696505.0,
      "step": 2475
    },
    {
      "entropy": 5.720158481597901,
      "epoch": 0.23823246878001922,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997715022040814,
      "loss": 5.5226,
      "mean_token_accuracy": 0.1807715639472008,
      "num_tokens": 5707638.0,
      "step": 2480
    },
    {
      "entropy": 5.720221424102784,
      "epoch": 0.23871277617675313,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499769954908461,
      "loss": 5.5071,
      "mean_token_accuracy": 0.18572808802127838,
      "num_tokens": 5719098.0,
      "step": 2485
    },
    {
      "entropy": 5.92160997390747,
      "epoch": 0.23919308357348704,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997684023943498,
      "loss": 5.742,
      "mean_token_accuracy": 0.15980200469493866,
      "num_tokens": 5731345.0,
      "step": 2490
    },
    {
      "entropy": 5.600750160217285,
      "epoch": 0.23967339097022095,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004997668446617837,
      "loss": 5.4909,
      "mean_token_accuracy": 0.17794736325740815,
      "num_tokens": 5742604.0,
      "step": 2495
    },
    {
      "entropy": 5.677820205688477,
      "epoch": 0.24015369836695485,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997652817107989,
      "loss": 5.5214,
      "mean_token_accuracy": 0.18534122705459594,
      "num_tokens": 5754620.0,
      "step": 2500
    },
    {
      "entropy": 5.7748322010040285,
      "epoch": 0.24063400576368876,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997637135414315,
      "loss": 5.6133,
      "mean_token_accuracy": 0.17856863886117935,
      "num_tokens": 5764764.0,
      "step": 2505
    },
    {
      "entropy": 5.605795955657959,
      "epoch": 0.24111431316042267,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997621401537183,
      "loss": 5.4975,
      "mean_token_accuracy": 0.18817870914936066,
      "num_tokens": 5776341.0,
      "step": 2510
    },
    {
      "entropy": 5.8177571296691895,
      "epoch": 0.24159462055715658,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997605615476955,
      "loss": 5.717,
      "mean_token_accuracy": 0.1655554562807083,
      "num_tokens": 5788604.0,
      "step": 2515
    },
    {
      "entropy": 5.782369232177734,
      "epoch": 0.2420749279538905,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004997589777234,
      "loss": 5.6014,
      "mean_token_accuracy": 0.17745051085948943,
      "num_tokens": 5800194.0,
      "step": 2520
    },
    {
      "entropy": 5.723162698745727,
      "epoch": 0.2425552353506244,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997573886808684,
      "loss": 5.5993,
      "mean_token_accuracy": 0.1841670349240303,
      "num_tokens": 5812978.0,
      "step": 2525
    },
    {
      "entropy": 5.698678493499756,
      "epoch": 0.2430355427473583,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997557944201375,
      "loss": 5.4984,
      "mean_token_accuracy": 0.18675288707017898,
      "num_tokens": 5824591.0,
      "step": 2530
    },
    {
      "entropy": 5.7060136795043945,
      "epoch": 0.24351585014409222,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997541949412445,
      "loss": 5.6051,
      "mean_token_accuracy": 0.17518853545188903,
      "num_tokens": 5835970.0,
      "step": 2535
    },
    {
      "entropy": 5.779909896850586,
      "epoch": 0.24399615754082613,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997525902442266,
      "loss": 5.6478,
      "mean_token_accuracy": 0.17169417887926103,
      "num_tokens": 5848687.0,
      "step": 2540
    },
    {
      "entropy": 5.723712730407715,
      "epoch": 0.24447646493756003,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997509803291207,
      "loss": 5.6057,
      "mean_token_accuracy": 0.1737314537167549,
      "num_tokens": 5860820.0,
      "step": 2545
    },
    {
      "entropy": 5.723750114440918,
      "epoch": 0.24495677233429394,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997493651959647,
      "loss": 5.5573,
      "mean_token_accuracy": 0.18182809501886368,
      "num_tokens": 5872172.0,
      "step": 2550
    },
    {
      "entropy": 5.743210744857788,
      "epoch": 0.24543707973102785,
      "grad_norm": 1.125,
      "learning_rate": 0.0004997477448447955,
      "loss": 5.5897,
      "mean_token_accuracy": 0.1766476571559906,
      "num_tokens": 5883574.0,
      "step": 2555
    },
    {
      "entropy": 5.840282154083252,
      "epoch": 0.24591738712776176,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004997461192756512,
      "loss": 5.6828,
      "mean_token_accuracy": 0.17689221501350402,
      "num_tokens": 5895223.0,
      "step": 2560
    },
    {
      "entropy": 5.701527786254883,
      "epoch": 0.24639769452449567,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997444884885694,
      "loss": 5.5116,
      "mean_token_accuracy": 0.178897462785244,
      "num_tokens": 5907448.0,
      "step": 2565
    },
    {
      "entropy": 5.728735780715942,
      "epoch": 0.24687800192122958,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997428524835879,
      "loss": 5.6146,
      "mean_token_accuracy": 0.17883744686841965,
      "num_tokens": 5918906.0,
      "step": 2570
    },
    {
      "entropy": 5.740764522552491,
      "epoch": 0.2473583093179635,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997412112607446,
      "loss": 5.4864,
      "mean_token_accuracy": 0.18324243873357773,
      "num_tokens": 5930619.0,
      "step": 2575
    },
    {
      "entropy": 5.808829116821289,
      "epoch": 0.2478386167146974,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997395648200778,
      "loss": 5.6809,
      "mean_token_accuracy": 0.1694455027580261,
      "num_tokens": 5941554.0,
      "step": 2580
    },
    {
      "entropy": 5.767109441757202,
      "epoch": 0.2483189241114313,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997379131616257,
      "loss": 5.6457,
      "mean_token_accuracy": 0.18047072738409042,
      "num_tokens": 5953309.0,
      "step": 2585
    },
    {
      "entropy": 5.707102966308594,
      "epoch": 0.24879923150816521,
      "grad_norm": 1.125,
      "learning_rate": 0.0004997362562854266,
      "loss": 5.5915,
      "mean_token_accuracy": 0.17505357414484024,
      "num_tokens": 5965286.0,
      "step": 2590
    },
    {
      "entropy": 5.772854042053223,
      "epoch": 0.24927953890489912,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997345941915187,
      "loss": 5.5786,
      "mean_token_accuracy": 0.18021601289510727,
      "num_tokens": 5976558.0,
      "step": 2595
    },
    {
      "entropy": 5.757956790924072,
      "epoch": 0.24975984630163303,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997329268799412,
      "loss": 5.622,
      "mean_token_accuracy": 0.16917790919542314,
      "num_tokens": 5988793.0,
      "step": 2600
    },
    {
      "entropy": 5.767512655258178,
      "epoch": 0.25024015369836694,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997312543507322,
      "loss": 5.6539,
      "mean_token_accuracy": 0.16811801493167877,
      "num_tokens": 5999825.0,
      "step": 2605
    },
    {
      "entropy": 5.757353258132935,
      "epoch": 0.2507204610951009,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997295766039309,
      "loss": 5.6145,
      "mean_token_accuracy": 0.17652266919612886,
      "num_tokens": 6010408.0,
      "step": 2610
    },
    {
      "entropy": 5.77374758720398,
      "epoch": 0.25120076849183476,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997278936395761,
      "loss": 5.6168,
      "mean_token_accuracy": 0.17355270236730574,
      "num_tokens": 6021161.0,
      "step": 2615
    },
    {
      "entropy": 5.654427337646484,
      "epoch": 0.2516810758885687,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997262054577071,
      "loss": 5.4818,
      "mean_token_accuracy": 0.18697400987148285,
      "num_tokens": 6032646.0,
      "step": 2620
    },
    {
      "entropy": 5.742709732055664,
      "epoch": 0.2521613832853026,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997245120583627,
      "loss": 5.6445,
      "mean_token_accuracy": 0.17707106322050095,
      "num_tokens": 6045019.0,
      "step": 2625
    },
    {
      "entropy": 5.6496696949005125,
      "epoch": 0.2526416906820365,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997228134415825,
      "loss": 5.5044,
      "mean_token_accuracy": 0.18557969927787782,
      "num_tokens": 6055644.0,
      "step": 2630
    },
    {
      "entropy": 5.749445962905884,
      "epoch": 0.2531219980787704,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004997211096074059,
      "loss": 5.5559,
      "mean_token_accuracy": 0.1822674036026001,
      "num_tokens": 6066488.0,
      "step": 2635
    },
    {
      "entropy": 5.750200700759888,
      "epoch": 0.25360230547550433,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997194005558722,
      "loss": 5.5778,
      "mean_token_accuracy": 0.1778818905353546,
      "num_tokens": 6078300.0,
      "step": 2640
    },
    {
      "entropy": 5.687224864959717,
      "epoch": 0.2540826128722382,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997176862870216,
      "loss": 5.5639,
      "mean_token_accuracy": 0.17298152297735214,
      "num_tokens": 6089180.0,
      "step": 2645
    },
    {
      "entropy": 5.736628150939941,
      "epoch": 0.25456292026897215,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997159668008933,
      "loss": 5.5897,
      "mean_token_accuracy": 0.18671621829271318,
      "num_tokens": 6099093.0,
      "step": 2650
    },
    {
      "entropy": 5.749474334716797,
      "epoch": 0.25504322766570603,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004997142420975277,
      "loss": 5.6091,
      "mean_token_accuracy": 0.17849962413311005,
      "num_tokens": 6109161.0,
      "step": 2655
    },
    {
      "entropy": 5.838236284255982,
      "epoch": 0.25552353506243997,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997125121769647,
      "loss": 5.7148,
      "mean_token_accuracy": 0.16754066944122314,
      "num_tokens": 6120031.0,
      "step": 2660
    },
    {
      "entropy": 5.6778631687164305,
      "epoch": 0.25600384245917385,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997107770392444,
      "loss": 5.5715,
      "mean_token_accuracy": 0.17843878418207168,
      "num_tokens": 6132398.0,
      "step": 2665
    },
    {
      "entropy": 5.6774248600006105,
      "epoch": 0.2564841498559078,
      "grad_norm": 1.15625,
      "learning_rate": 0.000499709036684407,
      "loss": 5.5453,
      "mean_token_accuracy": 0.17298439741134644,
      "num_tokens": 6143099.0,
      "step": 2670
    },
    {
      "entropy": 5.722945070266723,
      "epoch": 0.25696445725264166,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997072911124932,
      "loss": 5.5244,
      "mean_token_accuracy": 0.1837367057800293,
      "num_tokens": 6154815.0,
      "step": 2675
    },
    {
      "entropy": 5.701107454299927,
      "epoch": 0.2574447646493756,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997055403235432,
      "loss": 5.5811,
      "mean_token_accuracy": 0.18124920427799224,
      "num_tokens": 6165696.0,
      "step": 2680
    },
    {
      "entropy": 5.690740966796875,
      "epoch": 0.2579250720461095,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997037843175978,
      "loss": 5.6098,
      "mean_token_accuracy": 0.17476513981819153,
      "num_tokens": 6177042.0,
      "step": 2685
    },
    {
      "entropy": 5.786804866790772,
      "epoch": 0.2584053794428434,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997020230946978,
      "loss": 5.5954,
      "mean_token_accuracy": 0.17359965294599533,
      "num_tokens": 6188691.0,
      "step": 2690
    },
    {
      "entropy": 5.580425882339478,
      "epoch": 0.25888568683957736,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004997002566548841,
      "loss": 5.4196,
      "mean_token_accuracy": 0.18495422154664992,
      "num_tokens": 6199931.0,
      "step": 2695
    },
    {
      "entropy": 5.656063985824585,
      "epoch": 0.25936599423631124,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996984849981976,
      "loss": 5.5481,
      "mean_token_accuracy": 0.18747731745243074,
      "num_tokens": 6211493.0,
      "step": 2700
    },
    {
      "entropy": 5.758135843276977,
      "epoch": 0.2598463016330452,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996967081246794,
      "loss": 5.4869,
      "mean_token_accuracy": 0.18370969891548156,
      "num_tokens": 6221505.0,
      "step": 2705
    },
    {
      "entropy": 5.605699253082276,
      "epoch": 0.26032660902977905,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996949260343711,
      "loss": 5.4631,
      "mean_token_accuracy": 0.18868651241064072,
      "num_tokens": 6233058.0,
      "step": 2710
    },
    {
      "entropy": 5.705110359191894,
      "epoch": 0.260806916426513,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996931387273137,
      "loss": 5.6101,
      "mean_token_accuracy": 0.1747105076909065,
      "num_tokens": 6244713.0,
      "step": 2715
    },
    {
      "entropy": 5.716080856323242,
      "epoch": 0.2612872238232469,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004996913462035487,
      "loss": 5.5604,
      "mean_token_accuracy": 0.17616876363754272,
      "num_tokens": 6256624.0,
      "step": 2720
    },
    {
      "entropy": 5.65197606086731,
      "epoch": 0.2617675312199808,
      "grad_norm": 0.875,
      "learning_rate": 0.000499689548463118,
      "loss": 5.5627,
      "mean_token_accuracy": 0.17669420540332795,
      "num_tokens": 6271134.0,
      "step": 2725
    },
    {
      "entropy": 5.815130949020386,
      "epoch": 0.2622478386167147,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996877455060631,
      "loss": 5.6592,
      "mean_token_accuracy": 0.1788387343287468,
      "num_tokens": 6282082.0,
      "step": 2730
    },
    {
      "entropy": 5.766822576522827,
      "epoch": 0.2627281460134486,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004996859373324259,
      "loss": 5.6059,
      "mean_token_accuracy": 0.17674000561237335,
      "num_tokens": 6293869.0,
      "step": 2735
    },
    {
      "entropy": 5.676133012771606,
      "epoch": 0.2632084534101825,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004996841239422485,
      "loss": 5.6232,
      "mean_token_accuracy": 0.17271844148635865,
      "num_tokens": 6304858.0,
      "step": 2740
    },
    {
      "entropy": 5.713190507888794,
      "epoch": 0.26368876080691644,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004996823053355729,
      "loss": 5.5515,
      "mean_token_accuracy": 0.18129074424505234,
      "num_tokens": 6316738.0,
      "step": 2745
    },
    {
      "entropy": 5.679355955123901,
      "epoch": 0.2641690682036503,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004996804815124413,
      "loss": 5.5169,
      "mean_token_accuracy": 0.17863013297319413,
      "num_tokens": 6327691.0,
      "step": 2750
    },
    {
      "entropy": 5.720383548736573,
      "epoch": 0.26464937560038426,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004996786524728962,
      "loss": 5.5227,
      "mean_token_accuracy": 0.1802159935235977,
      "num_tokens": 6337964.0,
      "step": 2755
    },
    {
      "entropy": 5.613131427764893,
      "epoch": 0.26512968299711814,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996768182169797,
      "loss": 5.5725,
      "mean_token_accuracy": 0.17915034890174866,
      "num_tokens": 6349764.0,
      "step": 2760
    },
    {
      "entropy": 5.7847541809082035,
      "epoch": 0.2656099903938521,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004996749787447349,
      "loss": 5.5324,
      "mean_token_accuracy": 0.18092670440673828,
      "num_tokens": 6360538.0,
      "step": 2765
    },
    {
      "entropy": 5.733683729171753,
      "epoch": 0.26609029779058596,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000499673134056204,
      "loss": 5.5823,
      "mean_token_accuracy": 0.1780352845788002,
      "num_tokens": 6372489.0,
      "step": 2770
    },
    {
      "entropy": 5.526462173461914,
      "epoch": 0.2665706051873199,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996712841514303,
      "loss": 5.4429,
      "mean_token_accuracy": 0.18955876380205156,
      "num_tokens": 6383080.0,
      "step": 2775
    },
    {
      "entropy": 5.735677242279053,
      "epoch": 0.2670509125840538,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996694290304563,
      "loss": 5.5797,
      "mean_token_accuracy": 0.17862945348024367,
      "num_tokens": 6394033.0,
      "step": 2780
    },
    {
      "entropy": 5.6970155239105225,
      "epoch": 0.2675312199807877,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996675686933255,
      "loss": 5.5801,
      "mean_token_accuracy": 0.17881778180599212,
      "num_tokens": 6404990.0,
      "step": 2785
    },
    {
      "entropy": 5.6988551139831545,
      "epoch": 0.2680115273775216,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996657031400807,
      "loss": 5.4424,
      "mean_token_accuracy": 0.18443361967802047,
      "num_tokens": 6416337.0,
      "step": 2790
    },
    {
      "entropy": 5.689300441741944,
      "epoch": 0.26849183477425553,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996638323707655,
      "loss": 5.6479,
      "mean_token_accuracy": 0.17969592064619064,
      "num_tokens": 6427892.0,
      "step": 2795
    },
    {
      "entropy": 5.668305397033691,
      "epoch": 0.2689721421709894,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996619563854232,
      "loss": 5.5535,
      "mean_token_accuracy": 0.1804644450545311,
      "num_tokens": 6440300.0,
      "step": 2800
    },
    {
      "entropy": 5.702674102783203,
      "epoch": 0.26945244956772335,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996600751840974,
      "loss": 5.4618,
      "mean_token_accuracy": 0.184764464199543,
      "num_tokens": 6451499.0,
      "step": 2805
    },
    {
      "entropy": 5.6231475353240965,
      "epoch": 0.26993275696445723,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996581887668317,
      "loss": 5.5463,
      "mean_token_accuracy": 0.18376342356204986,
      "num_tokens": 6462882.0,
      "step": 2810
    },
    {
      "entropy": 5.676126384735108,
      "epoch": 0.27041306436119117,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00049965629713367,
      "loss": 5.5311,
      "mean_token_accuracy": 0.1827904239296913,
      "num_tokens": 6475042.0,
      "step": 2815
    },
    {
      "entropy": 5.673922348022461,
      "epoch": 0.27089337175792505,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996544002846561,
      "loss": 5.5345,
      "mean_token_accuracy": 0.18013176023960115,
      "num_tokens": 6486245.0,
      "step": 2820
    },
    {
      "entropy": 5.65201621055603,
      "epoch": 0.271373679154659,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996524982198343,
      "loss": 5.4305,
      "mean_token_accuracy": 0.18405921012163162,
      "num_tokens": 6497705.0,
      "step": 2825
    },
    {
      "entropy": 5.646395635604859,
      "epoch": 0.27185398655139287,
      "grad_norm": 1.125,
      "learning_rate": 0.0004996505909392485,
      "loss": 5.4717,
      "mean_token_accuracy": 0.18815743029117585,
      "num_tokens": 6508193.0,
      "step": 2830
    },
    {
      "entropy": 5.656166219711304,
      "epoch": 0.2723342939481268,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004996486784429429,
      "loss": 5.5878,
      "mean_token_accuracy": 0.17944951206445695,
      "num_tokens": 6519228.0,
      "step": 2835
    },
    {
      "entropy": 5.720212316513061,
      "epoch": 0.2728146013448607,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004996467607309622,
      "loss": 5.5327,
      "mean_token_accuracy": 0.1800275057554245,
      "num_tokens": 6530743.0,
      "step": 2840
    },
    {
      "entropy": 5.637170839309692,
      "epoch": 0.2732949087415946,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996448378033507,
      "loss": 5.4248,
      "mean_token_accuracy": 0.17780370265245438,
      "num_tokens": 6542588.0,
      "step": 2845
    },
    {
      "entropy": 5.768726253509522,
      "epoch": 0.2737752161383285,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996429096601532,
      "loss": 5.6334,
      "mean_token_accuracy": 0.1724066972732544,
      "num_tokens": 6554002.0,
      "step": 2850
    },
    {
      "entropy": 5.740063047409057,
      "epoch": 0.27425552353506244,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004996409763014144,
      "loss": 5.6518,
      "mean_token_accuracy": 0.1790720894932747,
      "num_tokens": 6563358.0,
      "step": 2855
    },
    {
      "entropy": 5.677773904800415,
      "epoch": 0.2747358309317964,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996390377271791,
      "loss": 5.5653,
      "mean_token_accuracy": 0.17528493255376815,
      "num_tokens": 6574986.0,
      "step": 2860
    },
    {
      "entropy": 5.6705268859863285,
      "epoch": 0.27521613832853026,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996370939374924,
      "loss": 5.5155,
      "mean_token_accuracy": 0.18047979474067688,
      "num_tokens": 6587407.0,
      "step": 2865
    },
    {
      "entropy": 5.591242885589599,
      "epoch": 0.2756964457252642,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004996351449323994,
      "loss": 5.4615,
      "mean_token_accuracy": 0.18611672520637512,
      "num_tokens": 6598483.0,
      "step": 2870
    },
    {
      "entropy": 5.6495672225952145,
      "epoch": 0.2761767531219981,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996331907119455,
      "loss": 5.5689,
      "mean_token_accuracy": 0.17973967641592026,
      "num_tokens": 6610749.0,
      "step": 2875
    },
    {
      "entropy": 5.750370597839355,
      "epoch": 0.276657060518732,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996312312761758,
      "loss": 5.5389,
      "mean_token_accuracy": 0.17384600937366484,
      "num_tokens": 6621762.0,
      "step": 2880
    },
    {
      "entropy": 5.708437776565551,
      "epoch": 0.2771373679154659,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499629266625136,
      "loss": 5.5945,
      "mean_token_accuracy": 0.17520614564418793,
      "num_tokens": 6632296.0,
      "step": 2885
    },
    {
      "entropy": 5.702244853973388,
      "epoch": 0.27761767531219983,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996272967588715,
      "loss": 5.5447,
      "mean_token_accuracy": 0.17878043949604033,
      "num_tokens": 6644413.0,
      "step": 2890
    },
    {
      "entropy": 5.613515996932984,
      "epoch": 0.2780979827089337,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996253216774283,
      "loss": 5.4922,
      "mean_token_accuracy": 0.18331679552793503,
      "num_tokens": 6655634.0,
      "step": 2895
    },
    {
      "entropy": 5.636519432067871,
      "epoch": 0.27857829010566765,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004996233413808521,
      "loss": 5.439,
      "mean_token_accuracy": 0.18576697260141373,
      "num_tokens": 6665977.0,
      "step": 2900
    },
    {
      "entropy": 5.716982173919678,
      "epoch": 0.27905859750240153,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996213558691889,
      "loss": 5.5078,
      "mean_token_accuracy": 0.17724172174930572,
      "num_tokens": 6677165.0,
      "step": 2905
    },
    {
      "entropy": 5.662445211410523,
      "epoch": 0.27953890489913547,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996193651424848,
      "loss": 5.6107,
      "mean_token_accuracy": 0.17485882341861725,
      "num_tokens": 6688182.0,
      "step": 2910
    },
    {
      "entropy": 5.6889976978302,
      "epoch": 0.28001921229586935,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000499617369200786,
      "loss": 5.526,
      "mean_token_accuracy": 0.17932689487934111,
      "num_tokens": 6699259.0,
      "step": 2915
    },
    {
      "entropy": 5.670577478408814,
      "epoch": 0.2804995196926033,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996153680441389,
      "loss": 5.5305,
      "mean_token_accuracy": 0.1791873812675476,
      "num_tokens": 6710071.0,
      "step": 2920
    },
    {
      "entropy": 5.654937314987182,
      "epoch": 0.28097982708933716,
      "grad_norm": 1.078125,
      "learning_rate": 0.00049961336167259,
      "loss": 5.5611,
      "mean_token_accuracy": 0.18469414710998536,
      "num_tokens": 6720267.0,
      "step": 2925
    },
    {
      "entropy": 5.754828691482544,
      "epoch": 0.2814601344860711,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004996113500861857,
      "loss": 5.555,
      "mean_token_accuracy": 0.1751690164208412,
      "num_tokens": 6730919.0,
      "step": 2930
    },
    {
      "entropy": 5.632857036590576,
      "epoch": 0.281940441882805,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004996093332849729,
      "loss": 5.546,
      "mean_token_accuracy": 0.18120535612106323,
      "num_tokens": 6741480.0,
      "step": 2935
    },
    {
      "entropy": 5.709432744979859,
      "epoch": 0.2824207492795389,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004996073112689983,
      "loss": 5.4782,
      "mean_token_accuracy": 0.18390992283821106,
      "num_tokens": 6751816.0,
      "step": 2940
    },
    {
      "entropy": 5.61135721206665,
      "epoch": 0.2829010566762728,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004996052840383088,
      "loss": 5.5091,
      "mean_token_accuracy": 0.17860084921121597,
      "num_tokens": 6763645.0,
      "step": 2945
    },
    {
      "entropy": 5.656038093566894,
      "epoch": 0.28338136407300674,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996032515929516,
      "loss": 5.4791,
      "mean_token_accuracy": 0.18455523252487183,
      "num_tokens": 6774795.0,
      "step": 2950
    },
    {
      "entropy": 5.625377082824707,
      "epoch": 0.2838616714697406,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996012139329738,
      "loss": 5.3998,
      "mean_token_accuracy": 0.18535638302564622,
      "num_tokens": 6786833.0,
      "step": 2955
    },
    {
      "entropy": 5.528808164596557,
      "epoch": 0.28434197886647455,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004995991710584228,
      "loss": 5.4166,
      "mean_token_accuracy": 0.18870054185390472,
      "num_tokens": 6798134.0,
      "step": 2960
    },
    {
      "entropy": 5.686023616790772,
      "epoch": 0.28482228626320844,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995971229693459,
      "loss": 5.528,
      "mean_token_accuracy": 0.18367817848920823,
      "num_tokens": 6807918.0,
      "step": 2965
    },
    {
      "entropy": 5.678937149047852,
      "epoch": 0.28530259365994237,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995950696657909,
      "loss": 5.6145,
      "mean_token_accuracy": 0.17956380695104598,
      "num_tokens": 6818689.0,
      "step": 2970
    },
    {
      "entropy": 5.626307201385498,
      "epoch": 0.28578290105667625,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004995930111478051,
      "loss": 5.472,
      "mean_token_accuracy": 0.18589290529489516,
      "num_tokens": 6830871.0,
      "step": 2975
    },
    {
      "entropy": 5.633310222625733,
      "epoch": 0.2862632084534102,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995909474154365,
      "loss": 5.4663,
      "mean_token_accuracy": 0.1886886775493622,
      "num_tokens": 6842857.0,
      "step": 2980
    },
    {
      "entropy": 5.628792810440063,
      "epoch": 0.28674351585014407,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004995888784687331,
      "loss": 5.4875,
      "mean_token_accuracy": 0.18284399956464767,
      "num_tokens": 6853727.0,
      "step": 2985
    },
    {
      "entropy": 5.7017065525054935,
      "epoch": 0.287223823246878,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995868043077428,
      "loss": 5.5282,
      "mean_token_accuracy": 0.18200663477182388,
      "num_tokens": 6864728.0,
      "step": 2990
    },
    {
      "entropy": 5.6810667514801025,
      "epoch": 0.2877041306436119,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004995847249325137,
      "loss": 5.5912,
      "mean_token_accuracy": 0.1790914699435234,
      "num_tokens": 6875001.0,
      "step": 2995
    },
    {
      "entropy": 5.5915452003479,
      "epoch": 0.2881844380403458,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995826403430942,
      "loss": 5.47,
      "mean_token_accuracy": 0.19052988588809966,
      "num_tokens": 6887026.0,
      "step": 3000
    },
    {
      "epoch": 0.2881844380403458,
      "eval_entropy": 5.464870153898614,
      "eval_loss": 5.527435779571533,
      "eval_mean_token_accuracy": 0.18901096377476345,
      "eval_num_tokens": 6887026.0,
      "eval_runtime": 20.7043,
      "eval_samples_per_second": 1584.936,
      "eval_steps_per_second": 198.123,
      "step": 3000
    },
    {
      "entropy": 5.818291330337525,
      "epoch": 0.2886647454370797,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004995805505395328,
      "loss": 5.5829,
      "mean_token_accuracy": 0.18093512058258057,
      "num_tokens": 6900118.0,
      "step": 3005
    },
    {
      "entropy": 5.656552839279175,
      "epoch": 0.28914505283381364,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004995784555218778,
      "loss": 5.496,
      "mean_token_accuracy": 0.1875130459666252,
      "num_tokens": 6912135.0,
      "step": 3010
    },
    {
      "entropy": 5.590249824523926,
      "epoch": 0.2896253602305475,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995763552901779,
      "loss": 5.4133,
      "mean_token_accuracy": 0.19095088839530944,
      "num_tokens": 6923781.0,
      "step": 3015
    },
    {
      "entropy": 5.679035711288452,
      "epoch": 0.29010566762728146,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004995742498444818,
      "loss": 5.5001,
      "mean_token_accuracy": 0.17695826888084412,
      "num_tokens": 6935804.0,
      "step": 3020
    },
    {
      "entropy": 5.605193185806274,
      "epoch": 0.2905859750240154,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004995721391848387,
      "loss": 5.5272,
      "mean_token_accuracy": 0.1858009621500969,
      "num_tokens": 6948289.0,
      "step": 3025
    },
    {
      "entropy": 5.670139503479004,
      "epoch": 0.2910662824207493,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004995700233112972,
      "loss": 5.4705,
      "mean_token_accuracy": 0.1871022552251816,
      "num_tokens": 6958719.0,
      "step": 3030
    },
    {
      "entropy": 5.644929981231689,
      "epoch": 0.2915465898174832,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995679022239066,
      "loss": 5.4987,
      "mean_token_accuracy": 0.1887220710515976,
      "num_tokens": 6969041.0,
      "step": 3035
    },
    {
      "entropy": 5.657832288742066,
      "epoch": 0.2920268972142171,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995657759227162,
      "loss": 5.5892,
      "mean_token_accuracy": 0.18469344228506088,
      "num_tokens": 6981206.0,
      "step": 3040
    },
    {
      "entropy": 5.698363971710205,
      "epoch": 0.29250720461095103,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995636444077751,
      "loss": 5.4718,
      "mean_token_accuracy": 0.18664478808641433,
      "num_tokens": 6993169.0,
      "step": 3045
    },
    {
      "entropy": 5.6591535091400145,
      "epoch": 0.2929875120076849,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004995615076791333,
      "loss": 5.5453,
      "mean_token_accuracy": 0.18127015978097916,
      "num_tokens": 7005175.0,
      "step": 3050
    },
    {
      "entropy": 5.68058557510376,
      "epoch": 0.29346781940441885,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004995593657368399,
      "loss": 5.5523,
      "mean_token_accuracy": 0.18219427466392518,
      "num_tokens": 7016738.0,
      "step": 3055
    },
    {
      "entropy": 5.672103595733643,
      "epoch": 0.29394812680115273,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499557218580945,
      "loss": 5.4567,
      "mean_token_accuracy": 0.1948094740509987,
      "num_tokens": 7028399.0,
      "step": 3060
    },
    {
      "entropy": 5.6687592506408695,
      "epoch": 0.29442843419788667,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004995550662114981,
      "loss": 5.6016,
      "mean_token_accuracy": 0.17741256803274155,
      "num_tokens": 7039528.0,
      "step": 3065
    },
    {
      "entropy": 5.513007926940918,
      "epoch": 0.29490874159462055,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995529086285495,
      "loss": 5.384,
      "mean_token_accuracy": 0.19410449415445327,
      "num_tokens": 7051399.0,
      "step": 3070
    },
    {
      "entropy": 5.523861217498779,
      "epoch": 0.2953890489913545,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499550745832149,
      "loss": 5.3799,
      "mean_token_accuracy": 0.19034833759069442,
      "num_tokens": 7062452.0,
      "step": 3075
    },
    {
      "entropy": 5.623450851440429,
      "epoch": 0.29586935638808837,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995485778223471,
      "loss": 5.52,
      "mean_token_accuracy": 0.18511104583740234,
      "num_tokens": 7074052.0,
      "step": 3080
    },
    {
      "entropy": 5.718336009979248,
      "epoch": 0.2963496637848223,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995464045991939,
      "loss": 5.5609,
      "mean_token_accuracy": 0.1748232737183571,
      "num_tokens": 7086692.0,
      "step": 3085
    },
    {
      "entropy": 5.661276388168335,
      "epoch": 0.2968299711815562,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00049954422616274,
      "loss": 5.5216,
      "mean_token_accuracy": 0.18554221391677855,
      "num_tokens": 7098221.0,
      "step": 3090
    },
    {
      "entropy": 5.593517160415649,
      "epoch": 0.2973102785782901,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995420425130359,
      "loss": 5.4816,
      "mean_token_accuracy": 0.19004761874675752,
      "num_tokens": 7110129.0,
      "step": 3095
    },
    {
      "entropy": 5.6908856391906735,
      "epoch": 0.297790585975024,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004995398536501324,
      "loss": 5.5565,
      "mean_token_accuracy": 0.17893462777137756,
      "num_tokens": 7121523.0,
      "step": 3100
    },
    {
      "entropy": 5.648267221450806,
      "epoch": 0.29827089337175794,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004995376595740801,
      "loss": 5.4882,
      "mean_token_accuracy": 0.18737834542989731,
      "num_tokens": 7133482.0,
      "step": 3105
    },
    {
      "entropy": 5.525942325592041,
      "epoch": 0.2987512007684918,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995354602849302,
      "loss": 5.4693,
      "mean_token_accuracy": 0.18503279387950897,
      "num_tokens": 7144031.0,
      "step": 3110
    },
    {
      "entropy": 5.697982311248779,
      "epoch": 0.29923150816522576,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004995332557827337,
      "loss": 5.545,
      "mean_token_accuracy": 0.17582330256700515,
      "num_tokens": 7155469.0,
      "step": 3115
    },
    {
      "entropy": 5.702816772460937,
      "epoch": 0.29971181556195964,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995310460675416,
      "loss": 5.484,
      "mean_token_accuracy": 0.1801610305905342,
      "num_tokens": 7167077.0,
      "step": 3120
    },
    {
      "entropy": 5.588912391662598,
      "epoch": 0.3001921229586936,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995288311394053,
      "loss": 5.4241,
      "mean_token_accuracy": 0.19050415456295014,
      "num_tokens": 7178917.0,
      "step": 3125
    },
    {
      "entropy": 5.690155410766602,
      "epoch": 0.30067243035542746,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004995266109983764,
      "loss": 5.5274,
      "mean_token_accuracy": 0.17950878590345382,
      "num_tokens": 7190284.0,
      "step": 3130
    },
    {
      "entropy": 5.59369478225708,
      "epoch": 0.3011527377521614,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995243856445062,
      "loss": 5.5229,
      "mean_token_accuracy": 0.18678916692733766,
      "num_tokens": 7202391.0,
      "step": 3135
    },
    {
      "entropy": 5.611648178100586,
      "epoch": 0.3016330451488953,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995221550778466,
      "loss": 5.4388,
      "mean_token_accuracy": 0.18906817734241485,
      "num_tokens": 7212882.0,
      "step": 3140
    },
    {
      "entropy": 5.572828865051269,
      "epoch": 0.3021133525456292,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004995199192984491,
      "loss": 5.4207,
      "mean_token_accuracy": 0.19024200290441512,
      "num_tokens": 7224102.0,
      "step": 3145
    },
    {
      "entropy": 5.622887372970581,
      "epoch": 0.3025936599423631,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004995176783063657,
      "loss": 5.3831,
      "mean_token_accuracy": 0.19056126922369004,
      "num_tokens": 7234579.0,
      "step": 3150
    },
    {
      "entropy": 5.591602516174317,
      "epoch": 0.30307396733909703,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004995154321016487,
      "loss": 5.5268,
      "mean_token_accuracy": 0.178172567486763,
      "num_tokens": 7246065.0,
      "step": 3155
    },
    {
      "entropy": 5.68428406715393,
      "epoch": 0.3035542747358309,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995131806843499,
      "loss": 5.4869,
      "mean_token_accuracy": 0.181064872443676,
      "num_tokens": 7258899.0,
      "step": 3160
    },
    {
      "entropy": 5.673531150817871,
      "epoch": 0.30403458213256485,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004995109240545218,
      "loss": 5.5653,
      "mean_token_accuracy": 0.1694480761885643,
      "num_tokens": 7271887.0,
      "step": 3165
    },
    {
      "entropy": 5.638287353515625,
      "epoch": 0.3045148895292987,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004995086622122167,
      "loss": 5.5158,
      "mean_token_accuracy": 0.1824269562959671,
      "num_tokens": 7282773.0,
      "step": 3170
    },
    {
      "entropy": 5.595272827148437,
      "epoch": 0.30499519692603266,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995063951574871,
      "loss": 5.3575,
      "mean_token_accuracy": 0.1881730765104294,
      "num_tokens": 7294460.0,
      "step": 3175
    },
    {
      "entropy": 5.5052574634552,
      "epoch": 0.30547550432276654,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004995041228903856,
      "loss": 5.4105,
      "mean_token_accuracy": 0.19128011167049408,
      "num_tokens": 7305125.0,
      "step": 3180
    },
    {
      "entropy": 5.513477230072022,
      "epoch": 0.3059558117195005,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000499501845410965,
      "loss": 5.4231,
      "mean_token_accuracy": 0.18600183427333833,
      "num_tokens": 7317738.0,
      "step": 3185
    },
    {
      "entropy": 5.626646995544434,
      "epoch": 0.30643611911623436,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004994995627192781,
      "loss": 5.4402,
      "mean_token_accuracy": 0.18957698047161103,
      "num_tokens": 7329820.0,
      "step": 3190
    },
    {
      "entropy": 5.631344318389893,
      "epoch": 0.3069164265129683,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004994972748153781,
      "loss": 5.4908,
      "mean_token_accuracy": 0.1821371302008629,
      "num_tokens": 7341471.0,
      "step": 3195
    },
    {
      "entropy": 5.59535813331604,
      "epoch": 0.30739673390970224,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499494981699318,
      "loss": 5.6001,
      "mean_token_accuracy": 0.18577772974967957,
      "num_tokens": 7352459.0,
      "step": 3200
    },
    {
      "entropy": 5.653695297241211,
      "epoch": 0.3078770413064361,
      "grad_norm": 1.078125,
      "learning_rate": 0.000499492683371151,
      "loss": 5.3938,
      "mean_token_accuracy": 0.19221991002559663,
      "num_tokens": 7363129.0,
      "step": 3205
    },
    {
      "entropy": 5.646122694015503,
      "epoch": 0.30835734870317005,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004994903798309306,
      "loss": 5.4662,
      "mean_token_accuracy": 0.18413919508457183,
      "num_tokens": 7373726.0,
      "step": 3210
    },
    {
      "entropy": 5.582058477401733,
      "epoch": 0.30883765609990393,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004994880710787102,
      "loss": 5.5275,
      "mean_token_accuracy": 0.18591587394475936,
      "num_tokens": 7383847.0,
      "step": 3215
    },
    {
      "entropy": 5.663880491256714,
      "epoch": 0.30931796349663787,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004994857571145432,
      "loss": 5.5101,
      "mean_token_accuracy": 0.1802189826965332,
      "num_tokens": 7396425.0,
      "step": 3220
    },
    {
      "entropy": 5.7047515392303465,
      "epoch": 0.30979827089337175,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994834379384837,
      "loss": 5.5725,
      "mean_token_accuracy": 0.18061236888170243,
      "num_tokens": 7407591.0,
      "step": 3225
    },
    {
      "entropy": 5.576733589172363,
      "epoch": 0.3102785782901057,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004994811135505851,
      "loss": 5.4386,
      "mean_token_accuracy": 0.18614037483930587,
      "num_tokens": 7419668.0,
      "step": 3230
    },
    {
      "entropy": 5.710690593719482,
      "epoch": 0.31075888568683957,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004994787839509018,
      "loss": 5.5376,
      "mean_token_accuracy": 0.18019334673881532,
      "num_tokens": 7431969.0,
      "step": 3235
    },
    {
      "entropy": 5.7016349792480465,
      "epoch": 0.3112391930835735,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994764491394876,
      "loss": 5.536,
      "mean_token_accuracy": 0.1807459458708763,
      "num_tokens": 7443467.0,
      "step": 3240
    },
    {
      "entropy": 5.650354623794556,
      "epoch": 0.3117195004803074,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004994741091163969,
      "loss": 5.5093,
      "mean_token_accuracy": 0.1855812057852745,
      "num_tokens": 7455220.0,
      "step": 3245
    },
    {
      "entropy": 5.63733320236206,
      "epoch": 0.3121998078770413,
      "grad_norm": 1.078125,
      "learning_rate": 0.000499471763881684,
      "loss": 5.4346,
      "mean_token_accuracy": 0.19234927892684936,
      "num_tokens": 7466452.0,
      "step": 3250
    },
    {
      "entropy": 5.601991462707519,
      "epoch": 0.3126801152737752,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004994694134354031,
      "loss": 5.5011,
      "mean_token_accuracy": 0.19211043864488603,
      "num_tokens": 7478798.0,
      "step": 3255
    },
    {
      "entropy": 5.621982431411743,
      "epoch": 0.31316042267050914,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499467057777609,
      "loss": 5.5647,
      "mean_token_accuracy": 0.18470945805311204,
      "num_tokens": 7491039.0,
      "step": 3260
    },
    {
      "entropy": 5.654377794265747,
      "epoch": 0.313640730067243,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004994646969083565,
      "loss": 5.4301,
      "mean_token_accuracy": 0.1862166404724121,
      "num_tokens": 7501455.0,
      "step": 3265
    },
    {
      "entropy": 5.600139474868774,
      "epoch": 0.31412103746397696,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004994623308277002,
      "loss": 5.4911,
      "mean_token_accuracy": 0.18477616012096404,
      "num_tokens": 7513484.0,
      "step": 3270
    },
    {
      "entropy": 5.630734634399414,
      "epoch": 0.31460134486071084,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000499459959535695,
      "loss": 5.4646,
      "mean_token_accuracy": 0.1820991650223732,
      "num_tokens": 7524868.0,
      "step": 3275
    },
    {
      "entropy": 5.5333410739898685,
      "epoch": 0.3150816522574448,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004994575830323962,
      "loss": 5.3374,
      "mean_token_accuracy": 0.20163633078336715,
      "num_tokens": 7536290.0,
      "step": 3280
    },
    {
      "entropy": 5.530055189132691,
      "epoch": 0.31556195965417866,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004994552013178586,
      "loss": 5.4453,
      "mean_token_accuracy": 0.19109649062156678,
      "num_tokens": 7546994.0,
      "step": 3285
    },
    {
      "entropy": 5.628282976150513,
      "epoch": 0.3160422670509126,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499452814392138,
      "loss": 5.4386,
      "mean_token_accuracy": 0.18418767750263215,
      "num_tokens": 7559231.0,
      "step": 3290
    },
    {
      "entropy": 5.555571508407593,
      "epoch": 0.3165225744476465,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004994504222552894,
      "loss": 5.42,
      "mean_token_accuracy": 0.18502341955900192,
      "num_tokens": 7570894.0,
      "step": 3295
    },
    {
      "entropy": 5.638952684402466,
      "epoch": 0.3170028818443804,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004994480249073684,
      "loss": 5.5369,
      "mean_token_accuracy": 0.17960984706878663,
      "num_tokens": 7581765.0,
      "step": 3300
    },
    {
      "entropy": 5.6019188404083256,
      "epoch": 0.3174831892411143,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004994456223484308,
      "loss": 5.3676,
      "mean_token_accuracy": 0.19097954630851746,
      "num_tokens": 7591828.0,
      "step": 3305
    },
    {
      "entropy": 5.5234428405761715,
      "epoch": 0.31796349663784823,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004994432145785323,
      "loss": 5.4598,
      "mean_token_accuracy": 0.18659110516309738,
      "num_tokens": 7604230.0,
      "step": 3310
    },
    {
      "entropy": 5.552060508728028,
      "epoch": 0.3184438040345821,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004994408015977288,
      "loss": 5.4194,
      "mean_token_accuracy": 0.1864725023508072,
      "num_tokens": 7615780.0,
      "step": 3315
    },
    {
      "entropy": 5.630988740921021,
      "epoch": 0.31892411143131605,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004994383834060764,
      "loss": 5.3843,
      "mean_token_accuracy": 0.18607487082481383,
      "num_tokens": 7627633.0,
      "step": 3320
    },
    {
      "entropy": 5.587767124176025,
      "epoch": 0.31940441882804993,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004994359600036311,
      "loss": 5.492,
      "mean_token_accuracy": 0.18158925622701644,
      "num_tokens": 7639487.0,
      "step": 3325
    },
    {
      "entropy": 5.546722078323365,
      "epoch": 0.31988472622478387,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004994335313904493,
      "loss": 5.4084,
      "mean_token_accuracy": 0.1924412354826927,
      "num_tokens": 7649976.0,
      "step": 3330
    },
    {
      "entropy": 5.717032480239868,
      "epoch": 0.32036503362151775,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004994310975665873,
      "loss": 5.596,
      "mean_token_accuracy": 0.17611310482025147,
      "num_tokens": 7662555.0,
      "step": 3335
    },
    {
      "entropy": 5.594285297393799,
      "epoch": 0.3208453410182517,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004994286585321017,
      "loss": 5.385,
      "mean_token_accuracy": 0.19134295880794525,
      "num_tokens": 7673869.0,
      "step": 3340
    },
    {
      "entropy": 5.514962387084961,
      "epoch": 0.32132564841498557,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499426214287049,
      "loss": 5.4885,
      "mean_token_accuracy": 0.1779414877295494,
      "num_tokens": 7687149.0,
      "step": 3345
    },
    {
      "entropy": 5.687413120269776,
      "epoch": 0.3218059558117195,
      "grad_norm": 1.0,
      "learning_rate": 0.0004994237648314862,
      "loss": 5.5271,
      "mean_token_accuracy": 0.17661560773849488,
      "num_tokens": 7698352.0,
      "step": 3350
    },
    {
      "entropy": 5.554401206970215,
      "epoch": 0.3222862632084534,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994213101654697,
      "loss": 5.3484,
      "mean_token_accuracy": 0.19577597379684447,
      "num_tokens": 7709117.0,
      "step": 3355
    },
    {
      "entropy": 5.548784303665161,
      "epoch": 0.3227665706051873,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000499418850289057,
      "loss": 5.388,
      "mean_token_accuracy": 0.1938525527715683,
      "num_tokens": 7721731.0,
      "step": 3360
    },
    {
      "entropy": 5.504498910903931,
      "epoch": 0.32324687800192126,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004994163852023048,
      "loss": 5.3551,
      "mean_token_accuracy": 0.18785073310136796,
      "num_tokens": 7733010.0,
      "step": 3365
    },
    {
      "entropy": 5.544020366668701,
      "epoch": 0.32372718539865514,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004994139149052706,
      "loss": 5.3405,
      "mean_token_accuracy": 0.19931638836860657,
      "num_tokens": 7743905.0,
      "step": 3370
    },
    {
      "entropy": 5.582752513885498,
      "epoch": 0.3242074927953891,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004994114393980117,
      "loss": 5.3837,
      "mean_token_accuracy": 0.1933474823832512,
      "num_tokens": 7754891.0,
      "step": 3375
    },
    {
      "entropy": 5.638752889633179,
      "epoch": 0.32468780019212296,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004994089586805856,
      "loss": 5.4998,
      "mean_token_accuracy": 0.17981112450361253,
      "num_tokens": 7766831.0,
      "step": 3380
    },
    {
      "entropy": 5.589061307907104,
      "epoch": 0.3251681075888569,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004994064727530496,
      "loss": 5.5713,
      "mean_token_accuracy": 0.1798534110188484,
      "num_tokens": 7778239.0,
      "step": 3385
    },
    {
      "entropy": 5.627358675003052,
      "epoch": 0.3256484149855908,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004994039816154618,
      "loss": 5.3839,
      "mean_token_accuracy": 0.19744640588760376,
      "num_tokens": 7789449.0,
      "step": 3390
    },
    {
      "entropy": 5.633282423019409,
      "epoch": 0.3261287223823247,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00049940148526788,
      "loss": 5.477,
      "mean_token_accuracy": 0.18499864339828492,
      "num_tokens": 7799507.0,
      "step": 3395
    },
    {
      "entropy": 5.601848602294922,
      "epoch": 0.3266090297790586,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004993989837103618,
      "loss": 5.4662,
      "mean_token_accuracy": 0.1881308600306511,
      "num_tokens": 7811730.0,
      "step": 3400
    },
    {
      "entropy": 5.526321029663086,
      "epoch": 0.3270893371757925,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004993964769429657,
      "loss": 5.3345,
      "mean_token_accuracy": 0.19526016414165498,
      "num_tokens": 7823108.0,
      "step": 3405
    },
    {
      "entropy": 5.560944032669068,
      "epoch": 0.3275696445725264,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004993939649657498,
      "loss": 5.4105,
      "mean_token_accuracy": 0.19024797976017,
      "num_tokens": 7835384.0,
      "step": 3410
    },
    {
      "entropy": 5.577944231033325,
      "epoch": 0.32804995196926034,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004993914477787721,
      "loss": 5.4485,
      "mean_token_accuracy": 0.18951481580734253,
      "num_tokens": 7845860.0,
      "step": 3415
    },
    {
      "entropy": 5.608705806732178,
      "epoch": 0.3285302593659942,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004993889253820915,
      "loss": 5.5016,
      "mean_token_accuracy": 0.18188290894031525,
      "num_tokens": 7856861.0,
      "step": 3420
    },
    {
      "entropy": 5.60044264793396,
      "epoch": 0.32901056676272816,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004993863977757663,
      "loss": 5.4575,
      "mean_token_accuracy": 0.19013751596212386,
      "num_tokens": 7868902.0,
      "step": 3425
    },
    {
      "entropy": 5.652436304092407,
      "epoch": 0.32949087415946204,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004993838649598552,
      "loss": 5.4713,
      "mean_token_accuracy": 0.18285695016384124,
      "num_tokens": 7880369.0,
      "step": 3430
    },
    {
      "entropy": 5.612339353561401,
      "epoch": 0.329971181556196,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004993813269344171,
      "loss": 5.465,
      "mean_token_accuracy": 0.18173497468233107,
      "num_tokens": 7892041.0,
      "step": 3435
    },
    {
      "entropy": 5.53320951461792,
      "epoch": 0.33045148895292986,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004993787836995108,
      "loss": 5.4159,
      "mean_token_accuracy": 0.18637658953666686,
      "num_tokens": 7903644.0,
      "step": 3440
    },
    {
      "entropy": 5.604093980789185,
      "epoch": 0.3309317963496638,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004993762352551954,
      "loss": 5.4706,
      "mean_token_accuracy": 0.18366012424230577,
      "num_tokens": 7915679.0,
      "step": 3445
    },
    {
      "entropy": 5.606694078445434,
      "epoch": 0.3314121037463977,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004993736816015301,
      "loss": 5.4939,
      "mean_token_accuracy": 0.18642788231372834,
      "num_tokens": 7928020.0,
      "step": 3450
    },
    {
      "entropy": 5.679842901229859,
      "epoch": 0.3318924111431316,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004993711227385742,
      "loss": 5.5504,
      "mean_token_accuracy": 0.1830623283982277,
      "num_tokens": 7940521.0,
      "step": 3455
    },
    {
      "entropy": 5.619497394561767,
      "epoch": 0.3323727185398655,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993685586663871,
      "loss": 5.4811,
      "mean_token_accuracy": 0.18078036606311798,
      "num_tokens": 7952234.0,
      "step": 3460
    },
    {
      "entropy": 5.625652265548706,
      "epoch": 0.33285302593659943,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004993659893850281,
      "loss": 5.5652,
      "mean_token_accuracy": 0.18391605764627456,
      "num_tokens": 7964900.0,
      "step": 3465
    },
    {
      "entropy": 5.690730333328247,
      "epoch": 0.3333333333333333,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004993634148945573,
      "loss": 5.5465,
      "mean_token_accuracy": 0.17793019860982895,
      "num_tokens": 7976661.0,
      "step": 3470
    },
    {
      "entropy": 5.56783537864685,
      "epoch": 0.33381364073006725,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004993608351950341,
      "loss": 5.392,
      "mean_token_accuracy": 0.18778552561998368,
      "num_tokens": 7987605.0,
      "step": 3475
    },
    {
      "entropy": 5.541170978546143,
      "epoch": 0.33429394812680113,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004993582502865185,
      "loss": 5.4486,
      "mean_token_accuracy": 0.1894877225160599,
      "num_tokens": 7998879.0,
      "step": 3480
    },
    {
      "entropy": 5.54890513420105,
      "epoch": 0.33477425552353507,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004993556601690706,
      "loss": 5.4338,
      "mean_token_accuracy": 0.18976636379957199,
      "num_tokens": 8010634.0,
      "step": 3485
    },
    {
      "entropy": 5.658605289459229,
      "epoch": 0.33525456292026895,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004993530648427505,
      "loss": 5.4464,
      "mean_token_accuracy": 0.18427447527647017,
      "num_tokens": 8022134.0,
      "step": 3490
    },
    {
      "entropy": 5.554767370223999,
      "epoch": 0.3357348703170029,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004993504643076184,
      "loss": 5.4025,
      "mean_token_accuracy": 0.1867258295416832,
      "num_tokens": 8034454.0,
      "step": 3495
    },
    {
      "entropy": 5.602069759368897,
      "epoch": 0.33621517771373677,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004993478585637347,
      "loss": 5.477,
      "mean_token_accuracy": 0.18074701130390167,
      "num_tokens": 8046175.0,
      "step": 3500
    },
    {
      "entropy": 5.712220001220703,
      "epoch": 0.3366954851104707,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004993452476111599,
      "loss": 5.5252,
      "mean_token_accuracy": 0.18209043890237808,
      "num_tokens": 8057335.0,
      "step": 3505
    },
    {
      "entropy": 5.547671604156494,
      "epoch": 0.3371757925072046,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004993426314499546,
      "loss": 5.358,
      "mean_token_accuracy": 0.1893295481801033,
      "num_tokens": 8068860.0,
      "step": 3510
    },
    {
      "entropy": 5.57526421546936,
      "epoch": 0.3376560999039385,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004993400100801796,
      "loss": 5.4186,
      "mean_token_accuracy": 0.1862942934036255,
      "num_tokens": 8078957.0,
      "step": 3515
    },
    {
      "entropy": 5.643449354171753,
      "epoch": 0.3381364073006724,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004993373835018956,
      "loss": 5.5344,
      "mean_token_accuracy": 0.18162660151720048,
      "num_tokens": 8090143.0,
      "step": 3520
    },
    {
      "entropy": 5.491999244689941,
      "epoch": 0.33861671469740634,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004993347517151638,
      "loss": 5.3432,
      "mean_token_accuracy": 0.19063944220542908,
      "num_tokens": 8101961.0,
      "step": 3525
    },
    {
      "entropy": 5.457345819473266,
      "epoch": 0.3390970220941403,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004993321147200452,
      "loss": 5.2787,
      "mean_token_accuracy": 0.1959090545773506,
      "num_tokens": 8112635.0,
      "step": 3530
    },
    {
      "entropy": 5.586727905273437,
      "epoch": 0.33957732949087416,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499329472516601,
      "loss": 5.4553,
      "mean_token_accuracy": 0.18357420414686204,
      "num_tokens": 8123730.0,
      "step": 3535
    },
    {
      "entropy": 5.6162670135498045,
      "epoch": 0.3400576368876081,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004993268251048925,
      "loss": 5.4708,
      "mean_token_accuracy": 0.18503204733133316,
      "num_tokens": 8135258.0,
      "step": 3540
    },
    {
      "entropy": 5.516633415222168,
      "epoch": 0.340537944284342,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004993241724849814,
      "loss": 5.3375,
      "mean_token_accuracy": 0.19189947992563247,
      "num_tokens": 8145946.0,
      "step": 3545
    },
    {
      "entropy": 5.623859071731568,
      "epoch": 0.3410182516810759,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499321514656929,
      "loss": 5.5398,
      "mean_token_accuracy": 0.18308452218770982,
      "num_tokens": 8158422.0,
      "step": 3550
    },
    {
      "entropy": 5.592836666107178,
      "epoch": 0.3414985590778098,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004993188516207972,
      "loss": 5.4729,
      "mean_token_accuracy": 0.1765248253941536,
      "num_tokens": 8170849.0,
      "step": 3555
    },
    {
      "entropy": 5.522468757629395,
      "epoch": 0.34197886647454373,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004993161833766478,
      "loss": 5.4053,
      "mean_token_accuracy": 0.18717055767774582,
      "num_tokens": 8180907.0,
      "step": 3560
    },
    {
      "entropy": 5.654421377182007,
      "epoch": 0.3424591738712776,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993135099245426,
      "loss": 5.492,
      "mean_token_accuracy": 0.1786191776394844,
      "num_tokens": 8193418.0,
      "step": 3565
    },
    {
      "entropy": 5.612684154510498,
      "epoch": 0.34293948126801155,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993108312645438,
      "loss": 5.5456,
      "mean_token_accuracy": 0.1864428073167801,
      "num_tokens": 8203873.0,
      "step": 3570
    },
    {
      "entropy": 5.698003482818604,
      "epoch": 0.34341978866474543,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993081473967135,
      "loss": 5.5527,
      "mean_token_accuracy": 0.18066151887178422,
      "num_tokens": 8215150.0,
      "step": 3575
    },
    {
      "entropy": 5.683527517318725,
      "epoch": 0.34390009606147937,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004993054583211143,
      "loss": 5.5062,
      "mean_token_accuracy": 0.18355498164892198,
      "num_tokens": 8225732.0,
      "step": 3580
    },
    {
      "entropy": 5.606461095809936,
      "epoch": 0.34438040345821325,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004993027640378081,
      "loss": 5.4642,
      "mean_token_accuracy": 0.18439879417419433,
      "num_tokens": 8237523.0,
      "step": 3585
    },
    {
      "entropy": 5.5978563785552975,
      "epoch": 0.3448607108549472,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499300064546858,
      "loss": 5.4489,
      "mean_token_accuracy": 0.19081379026174544,
      "num_tokens": 8249660.0,
      "step": 3590
    },
    {
      "entropy": 5.511026430130005,
      "epoch": 0.34534101825168106,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004992973598483264,
      "loss": 5.2816,
      "mean_token_accuracy": 0.1942738652229309,
      "num_tokens": 8261224.0,
      "step": 3595
    },
    {
      "entropy": 5.485535192489624,
      "epoch": 0.345821325648415,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499294649942276,
      "loss": 5.511,
      "mean_token_accuracy": 0.18905575275421144,
      "num_tokens": 8272177.0,
      "step": 3600
    },
    {
      "entropy": 5.594182538986206,
      "epoch": 0.3463016330451489,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004992919348287699,
      "loss": 5.4197,
      "mean_token_accuracy": 0.19150247275829316,
      "num_tokens": 8284449.0,
      "step": 3605
    },
    {
      "entropy": 5.594610643386841,
      "epoch": 0.3467819404418828,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004992892145078711,
      "loss": 5.468,
      "mean_token_accuracy": 0.18804367482662201,
      "num_tokens": 8297028.0,
      "step": 3610
    },
    {
      "entropy": 5.546618127822876,
      "epoch": 0.3472622478386167,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004992864889796427,
      "loss": 5.3444,
      "mean_token_accuracy": 0.19881392419338226,
      "num_tokens": 8308348.0,
      "step": 3615
    },
    {
      "entropy": 5.498028326034546,
      "epoch": 0.34774255523535064,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004992837582441481,
      "loss": 5.3611,
      "mean_token_accuracy": 0.19773702919483185,
      "num_tokens": 8318951.0,
      "step": 3620
    },
    {
      "entropy": 5.5819737911224365,
      "epoch": 0.3482228626320845,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004992810223014506,
      "loss": 5.4765,
      "mean_token_accuracy": 0.18301982432603836,
      "num_tokens": 8329979.0,
      "step": 3625
    },
    {
      "entropy": 5.614229679107666,
      "epoch": 0.34870317002881845,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004992782811516137,
      "loss": 5.413,
      "mean_token_accuracy": 0.18938301205635072,
      "num_tokens": 8340332.0,
      "step": 3630
    },
    {
      "entropy": 5.55254979133606,
      "epoch": 0.34918347742555234,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004992755347947011,
      "loss": 5.4196,
      "mean_token_accuracy": 0.18862021565437317,
      "num_tokens": 8350376.0,
      "step": 3635
    },
    {
      "entropy": 5.557257032394409,
      "epoch": 0.34966378482228627,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004992727832307766,
      "loss": 5.3572,
      "mean_token_accuracy": 0.19571560025215148,
      "num_tokens": 8361884.0,
      "step": 3640
    },
    {
      "entropy": 5.468594741821289,
      "epoch": 0.35014409221902015,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004992700264599039,
      "loss": 5.2926,
      "mean_token_accuracy": 0.19994410872459412,
      "num_tokens": 8372914.0,
      "step": 3645
    },
    {
      "entropy": 5.571347856521607,
      "epoch": 0.3506243996157541,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004992672644821473,
      "loss": 5.4608,
      "mean_token_accuracy": 0.18287917524576186,
      "num_tokens": 8385800.0,
      "step": 3650
    },
    {
      "entropy": 5.54053807258606,
      "epoch": 0.35110470701248797,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004992644972975707,
      "loss": 5.3491,
      "mean_token_accuracy": 0.19611871093511582,
      "num_tokens": 8397783.0,
      "step": 3655
    },
    {
      "entropy": 5.529743576049805,
      "epoch": 0.3515850144092219,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004992617249062383,
      "loss": 5.4164,
      "mean_token_accuracy": 0.18574022948741914,
      "num_tokens": 8409096.0,
      "step": 3660
    },
    {
      "entropy": 5.502672433853149,
      "epoch": 0.3520653218059558,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004992589473082147,
      "loss": 5.3163,
      "mean_token_accuracy": 0.1920155480504036,
      "num_tokens": 8420833.0,
      "step": 3665
    },
    {
      "entropy": 5.545916223526001,
      "epoch": 0.3525456292026897,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004992561645035641,
      "loss": 5.414,
      "mean_token_accuracy": 0.18519209176301957,
      "num_tokens": 8431848.0,
      "step": 3670
    },
    {
      "entropy": 5.649065732955933,
      "epoch": 0.3530259365994236,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004992533764923515,
      "loss": 5.4835,
      "mean_token_accuracy": 0.18125005215406417,
      "num_tokens": 8443620.0,
      "step": 3675
    },
    {
      "entropy": 5.60196213722229,
      "epoch": 0.35350624399615754,
      "grad_norm": 1.0,
      "learning_rate": 0.0004992505832746412,
      "loss": 5.4801,
      "mean_token_accuracy": 0.18716504275798798,
      "num_tokens": 8455339.0,
      "step": 3680
    },
    {
      "entropy": 5.492714786529541,
      "epoch": 0.3539865513928914,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004992477848504983,
      "loss": 5.3294,
      "mean_token_accuracy": 0.19695633202791213,
      "num_tokens": 8466342.0,
      "step": 3685
    },
    {
      "entropy": 5.551093006134034,
      "epoch": 0.35446685878962536,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004992449812199877,
      "loss": 5.3557,
      "mean_token_accuracy": 0.18981612622737884,
      "num_tokens": 8476955.0,
      "step": 3690
    },
    {
      "entropy": 5.432433032989502,
      "epoch": 0.3549471661863593,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004992421723831745,
      "loss": 5.3381,
      "mean_token_accuracy": 0.18711488842964172,
      "num_tokens": 8488148.0,
      "step": 3695
    },
    {
      "entropy": 5.5809650897979735,
      "epoch": 0.3554274735830932,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004992393583401239,
      "loss": 5.3562,
      "mean_token_accuracy": 0.192018586397171,
      "num_tokens": 8499014.0,
      "step": 3700
    },
    {
      "entropy": 5.576114749908447,
      "epoch": 0.3559077809798271,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004992365390909014,
      "loss": 5.4037,
      "mean_token_accuracy": 0.18824678510427476,
      "num_tokens": 8510191.0,
      "step": 3705
    },
    {
      "entropy": 5.506738519668579,
      "epoch": 0.356388088376561,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004992337146355721,
      "loss": 5.3057,
      "mean_token_accuracy": 0.19836191236972808,
      "num_tokens": 8522368.0,
      "step": 3710
    },
    {
      "entropy": 5.510123682022095,
      "epoch": 0.35686839577329493,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004992308849742019,
      "loss": 5.3163,
      "mean_token_accuracy": 0.19676650166511536,
      "num_tokens": 8533692.0,
      "step": 3715
    },
    {
      "entropy": 5.541636180877686,
      "epoch": 0.3573487031700288,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004992280501068563,
      "loss": 5.4076,
      "mean_token_accuracy": 0.1902167245745659,
      "num_tokens": 8545283.0,
      "step": 3720
    },
    {
      "entropy": 5.564352035522461,
      "epoch": 0.35782901056676275,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004992252100336012,
      "loss": 5.3982,
      "mean_token_accuracy": 0.19354566633701326,
      "num_tokens": 8555094.0,
      "step": 3725
    },
    {
      "entropy": 5.505967330932617,
      "epoch": 0.35830931796349663,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004992223647545027,
      "loss": 5.4186,
      "mean_token_accuracy": 0.18728039860725404,
      "num_tokens": 8565179.0,
      "step": 3730
    },
    {
      "entropy": 5.492090654373169,
      "epoch": 0.35878962536023057,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004992195142696266,
      "loss": 5.3341,
      "mean_token_accuracy": 0.18886168897151948,
      "num_tokens": 8575764.0,
      "step": 3735
    },
    {
      "entropy": 5.420212936401367,
      "epoch": 0.35926993275696445,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004992166585790391,
      "loss": 5.2794,
      "mean_token_accuracy": 0.1977664574980736,
      "num_tokens": 8586052.0,
      "step": 3740
    },
    {
      "entropy": 5.560855150222778,
      "epoch": 0.3597502401536984,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004992137976828067,
      "loss": 5.4331,
      "mean_token_accuracy": 0.19117721170186996,
      "num_tokens": 8597832.0,
      "step": 3745
    },
    {
      "entropy": 5.553453636169434,
      "epoch": 0.36023054755043227,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992109315809955,
      "loss": 5.4173,
      "mean_token_accuracy": 0.19055771678686143,
      "num_tokens": 8609917.0,
      "step": 3750
    },
    {
      "entropy": 5.66266098022461,
      "epoch": 0.3607108549471662,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004992080602736725,
      "loss": 5.4398,
      "mean_token_accuracy": 0.18691930472850798,
      "num_tokens": 8620137.0,
      "step": 3755
    },
    {
      "entropy": 5.530855512619018,
      "epoch": 0.3611911623439001,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004992051837609039,
      "loss": 5.4414,
      "mean_token_accuracy": 0.19409410506486893,
      "num_tokens": 8630620.0,
      "step": 3760
    },
    {
      "entropy": 5.511130523681641,
      "epoch": 0.361671469740634,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004992023020427568,
      "loss": 5.3249,
      "mean_token_accuracy": 0.19436410814523697,
      "num_tokens": 8642820.0,
      "step": 3765
    },
    {
      "entropy": 5.561899900436401,
      "epoch": 0.3621517771373679,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004991994151192979,
      "loss": 5.3095,
      "mean_token_accuracy": 0.20017425268888472,
      "num_tokens": 8654081.0,
      "step": 3770
    },
    {
      "entropy": 5.5231773376464846,
      "epoch": 0.36263208453410184,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004991965229905943,
      "loss": 5.341,
      "mean_token_accuracy": 0.19354602843523025,
      "num_tokens": 8665097.0,
      "step": 3775
    },
    {
      "entropy": 5.462382698059082,
      "epoch": 0.3631123919308357,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004991936256567133,
      "loss": 5.3391,
      "mean_token_accuracy": 0.19357120245695114,
      "num_tokens": 8676576.0,
      "step": 3780
    },
    {
      "entropy": 5.477186632156372,
      "epoch": 0.36359269932756966,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499190723117722,
      "loss": 5.336,
      "mean_token_accuracy": 0.18812832236289978,
      "num_tokens": 8689217.0,
      "step": 3785
    },
    {
      "entropy": 5.620044994354248,
      "epoch": 0.36407300672430354,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004991878153736877,
      "loss": 5.4715,
      "mean_token_accuracy": 0.18229800015687941,
      "num_tokens": 8701101.0,
      "step": 3790
    },
    {
      "entropy": 5.547843313217163,
      "epoch": 0.3645533141210375,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004991849024246781,
      "loss": 5.3998,
      "mean_token_accuracy": 0.19204504489898683,
      "num_tokens": 8712769.0,
      "step": 3795
    },
    {
      "entropy": 5.443006086349487,
      "epoch": 0.36503362151777136,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004991819842707608,
      "loss": 5.3291,
      "mean_token_accuracy": 0.19806933552026748,
      "num_tokens": 8724691.0,
      "step": 3800
    },
    {
      "entropy": 5.450334262847901,
      "epoch": 0.3655139289145053,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004991790609120035,
      "loss": 5.2922,
      "mean_token_accuracy": 0.19550720751285552,
      "num_tokens": 8738008.0,
      "step": 3805
    },
    {
      "entropy": 5.618185710906983,
      "epoch": 0.3659942363112392,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499176132348474,
      "loss": 5.499,
      "mean_token_accuracy": 0.18377866148948668,
      "num_tokens": 8750353.0,
      "step": 3810
    },
    {
      "entropy": 5.5274430274963375,
      "epoch": 0.3664745437079731,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004991731985802405,
      "loss": 5.297,
      "mean_token_accuracy": 0.1962455078959465,
      "num_tokens": 8761319.0,
      "step": 3815
    },
    {
      "entropy": 5.489836931228638,
      "epoch": 0.366954851104707,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004991702596073708,
      "loss": 5.3662,
      "mean_token_accuracy": 0.18967041075229646,
      "num_tokens": 8772722.0,
      "step": 3820
    },
    {
      "entropy": 5.604290866851807,
      "epoch": 0.36743515850144093,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004991673154299335,
      "loss": 5.4603,
      "mean_token_accuracy": 0.1876854345202446,
      "num_tokens": 8783886.0,
      "step": 3825
    },
    {
      "entropy": 5.547819948196411,
      "epoch": 0.3679154658981748,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004991643660479967,
      "loss": 5.369,
      "mean_token_accuracy": 0.19597966372966766,
      "num_tokens": 8795161.0,
      "step": 3830
    },
    {
      "entropy": 5.429578685760498,
      "epoch": 0.36839577329490875,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004991614114616289,
      "loss": 5.3451,
      "mean_token_accuracy": 0.19433749616146087,
      "num_tokens": 8805417.0,
      "step": 3835
    },
    {
      "entropy": 5.505033826828003,
      "epoch": 0.3688760806916426,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004991584516708988,
      "loss": 5.412,
      "mean_token_accuracy": 0.18523598313331605,
      "num_tokens": 8817980.0,
      "step": 3840
    },
    {
      "entropy": 5.420059394836426,
      "epoch": 0.36935638808837656,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004991554866758751,
      "loss": 5.2082,
      "mean_token_accuracy": 0.20497371703386308,
      "num_tokens": 8829814.0,
      "step": 3845
    },
    {
      "entropy": 5.418878698348999,
      "epoch": 0.36983669548511044,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004991525164766265,
      "loss": 5.3158,
      "mean_token_accuracy": 0.18982076942920684,
      "num_tokens": 8840220.0,
      "step": 3850
    },
    {
      "entropy": 5.548164701461792,
      "epoch": 0.3703170028818444,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004991495410732222,
      "loss": 5.4399,
      "mean_token_accuracy": 0.1889181837439537,
      "num_tokens": 8852156.0,
      "step": 3855
    },
    {
      "entropy": 5.589278888702393,
      "epoch": 0.37079731027857826,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004991465604657311,
      "loss": 5.3641,
      "mean_token_accuracy": 0.19025969952344896,
      "num_tokens": 8863362.0,
      "step": 3860
    },
    {
      "entropy": 5.473960542678833,
      "epoch": 0.3712776176753122,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004991435746542224,
      "loss": 5.396,
      "mean_token_accuracy": 0.19214082211256028,
      "num_tokens": 8874957.0,
      "step": 3865
    },
    {
      "entropy": 5.453449058532715,
      "epoch": 0.37175792507204614,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004991405836387655,
      "loss": 5.3135,
      "mean_token_accuracy": 0.19935648441314696,
      "num_tokens": 8886171.0,
      "step": 3870
    },
    {
      "entropy": 5.563224935531617,
      "epoch": 0.37223823246878,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004991375874194298,
      "loss": 5.4337,
      "mean_token_accuracy": 0.187838639318943,
      "num_tokens": 8898103.0,
      "step": 3875
    },
    {
      "entropy": 5.5430069923400875,
      "epoch": 0.37271853986551395,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499134585996285,
      "loss": 5.398,
      "mean_token_accuracy": 0.19007300585508347,
      "num_tokens": 8910195.0,
      "step": 3880
    },
    {
      "entropy": 5.4910304069519045,
      "epoch": 0.37319884726224783,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004991315793694004,
      "loss": 5.3636,
      "mean_token_accuracy": 0.19342513233423234,
      "num_tokens": 8922865.0,
      "step": 3885
    },
    {
      "entropy": 5.46483850479126,
      "epoch": 0.37367915465898177,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004991285675388463,
      "loss": 5.4029,
      "mean_token_accuracy": 0.18625415563583375,
      "num_tokens": 8933913.0,
      "step": 3890
    },
    {
      "entropy": 5.497545957565308,
      "epoch": 0.37415946205571565,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004991255505046922,
      "loss": 5.3197,
      "mean_token_accuracy": 0.19304455518722535,
      "num_tokens": 8945749.0,
      "step": 3895
    },
    {
      "entropy": 5.52088508605957,
      "epoch": 0.3746397694524496,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004991225282670083,
      "loss": 5.3163,
      "mean_token_accuracy": 0.1921261578798294,
      "num_tokens": 8957454.0,
      "step": 3900
    },
    {
      "entropy": 5.522637987136841,
      "epoch": 0.37512007684918347,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499119500825865,
      "loss": 5.3717,
      "mean_token_accuracy": 0.19198500365018845,
      "num_tokens": 8968173.0,
      "step": 3905
    },
    {
      "entropy": 5.474001359939575,
      "epoch": 0.3756003842459174,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004991164681813323,
      "loss": 5.387,
      "mean_token_accuracy": 0.18982407748699187,
      "num_tokens": 8979187.0,
      "step": 3910
    },
    {
      "entropy": 5.494542980194092,
      "epoch": 0.3760806916426513,
      "grad_norm": 1.0,
      "learning_rate": 0.0004991134303334807,
      "loss": 5.2957,
      "mean_token_accuracy": 0.20396983474493027,
      "num_tokens": 8990007.0,
      "step": 3915
    },
    {
      "entropy": 5.530715227127075,
      "epoch": 0.3765609990393852,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004991103872823807,
      "loss": 5.4387,
      "mean_token_accuracy": 0.1851722002029419,
      "num_tokens": 9001108.0,
      "step": 3920
    },
    {
      "entropy": 5.57948408126831,
      "epoch": 0.3770413064361191,
      "grad_norm": 1.09375,
      "learning_rate": 0.000499107339028103,
      "loss": 5.4431,
      "mean_token_accuracy": 0.19176366329193115,
      "num_tokens": 9012704.0,
      "step": 3925
    },
    {
      "entropy": 5.530899286270142,
      "epoch": 0.37752161383285304,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004991042855707184,
      "loss": 5.3752,
      "mean_token_accuracy": 0.19565950483083724,
      "num_tokens": 9024351.0,
      "step": 3930
    },
    {
      "entropy": 5.4647791385650635,
      "epoch": 0.3780019212295869,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004991012269102977,
      "loss": 5.2973,
      "mean_token_accuracy": 0.1938745141029358,
      "num_tokens": 9035099.0,
      "step": 3935
    },
    {
      "entropy": 5.512908458709717,
      "epoch": 0.37848222862632086,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004990981630469119,
      "loss": 5.3336,
      "mean_token_accuracy": 0.19299072474241258,
      "num_tokens": 9045675.0,
      "step": 3940
    },
    {
      "entropy": 5.50911808013916,
      "epoch": 0.37896253602305474,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990950939806323,
      "loss": 5.3123,
      "mean_token_accuracy": 0.19613827466964723,
      "num_tokens": 9057133.0,
      "step": 3945
    },
    {
      "entropy": 5.506547594070435,
      "epoch": 0.3794428434197887,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00049909201971153,
      "loss": 5.3139,
      "mean_token_accuracy": 0.19197983741760255,
      "num_tokens": 9067806.0,
      "step": 3950
    },
    {
      "entropy": 5.408311080932617,
      "epoch": 0.37992315081652256,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004990889402396763,
      "loss": 5.3266,
      "mean_token_accuracy": 0.19067565053701402,
      "num_tokens": 9079893.0,
      "step": 3955
    },
    {
      "entropy": 5.501181411743164,
      "epoch": 0.3804034582132565,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004990858555651431,
      "loss": 5.3461,
      "mean_token_accuracy": 0.1902176633477211,
      "num_tokens": 9091583.0,
      "step": 3960
    },
    {
      "entropy": 5.5375447273254395,
      "epoch": 0.3808837656099904,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004990827656880015,
      "loss": 5.3492,
      "mean_token_accuracy": 0.19774834513664247,
      "num_tokens": 9102919.0,
      "step": 3965
    },
    {
      "entropy": 5.504542636871338,
      "epoch": 0.3813640730067243,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004990796706083235,
      "loss": 5.3098,
      "mean_token_accuracy": 0.19810621440410614,
      "num_tokens": 9113876.0,
      "step": 3970
    },
    {
      "entropy": 5.5456907749176025,
      "epoch": 0.3818443804034582,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004990765703261809,
      "loss": 5.4696,
      "mean_token_accuracy": 0.18376226276159285,
      "num_tokens": 9125630.0,
      "step": 3975
    },
    {
      "entropy": 5.5273786067962645,
      "epoch": 0.38232468780019213,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004990734648416458,
      "loss": 5.3966,
      "mean_token_accuracy": 0.1941981002688408,
      "num_tokens": 9137302.0,
      "step": 3980
    },
    {
      "entropy": 5.505181312561035,
      "epoch": 0.382804995196926,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004990703541547901,
      "loss": 5.3504,
      "mean_token_accuracy": 0.19993013739585877,
      "num_tokens": 9148048.0,
      "step": 3985
    },
    {
      "entropy": 5.537240314483642,
      "epoch": 0.38328530259365995,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004990672382656863,
      "loss": 5.4085,
      "mean_token_accuracy": 0.18771413266658782,
      "num_tokens": 9158656.0,
      "step": 3990
    },
    {
      "entropy": 5.527101278305054,
      "epoch": 0.38376560999039383,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004990641171744064,
      "loss": 5.2938,
      "mean_token_accuracy": 0.19386786669492723,
      "num_tokens": 9169324.0,
      "step": 3995
    },
    {
      "entropy": 5.407563257217407,
      "epoch": 0.38424591738712777,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004990609908810231,
      "loss": 5.262,
      "mean_token_accuracy": 0.19891532510519028,
      "num_tokens": 9180238.0,
      "step": 4000
    },
    {
      "entropy": 5.340078401565552,
      "epoch": 0.38472622478386165,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004990578593856089,
      "loss": 5.1258,
      "mean_token_accuracy": 0.2009473592042923,
      "num_tokens": 9190766.0,
      "step": 4005
    },
    {
      "entropy": 5.422877788543701,
      "epoch": 0.3852065321805956,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004990547226882366,
      "loss": 5.2753,
      "mean_token_accuracy": 0.20038543045520782,
      "num_tokens": 9202445.0,
      "step": 4010
    },
    {
      "entropy": 5.488477802276611,
      "epoch": 0.38568683957732947,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004990515807889788,
      "loss": 5.3471,
      "mean_token_accuracy": 0.19519307762384414,
      "num_tokens": 9212383.0,
      "step": 4015
    },
    {
      "entropy": 5.466127920150757,
      "epoch": 0.3861671469740634,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004990484336879087,
      "loss": 5.2721,
      "mean_token_accuracy": 0.19732844680547715,
      "num_tokens": 9223619.0,
      "step": 4020
    },
    {
      "entropy": 5.503373289108277,
      "epoch": 0.3866474543707973,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004990452813850992,
      "loss": 5.4139,
      "mean_token_accuracy": 0.18991868942975998,
      "num_tokens": 9234137.0,
      "step": 4025
    },
    {
      "entropy": 5.5534507751464846,
      "epoch": 0.3871277617675312,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004990421238806236,
      "loss": 5.3405,
      "mean_token_accuracy": 0.19380167722702027,
      "num_tokens": 9246093.0,
      "step": 4030
    },
    {
      "entropy": 5.569162893295288,
      "epoch": 0.38760806916426516,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004990389611745551,
      "loss": 5.438,
      "mean_token_accuracy": 0.18329400569200516,
      "num_tokens": 9258021.0,
      "step": 4035
    },
    {
      "entropy": 5.515481615066529,
      "epoch": 0.38808837656099904,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990357932669672,
      "loss": 5.4778,
      "mean_token_accuracy": 0.19151371717453003,
      "num_tokens": 9267668.0,
      "step": 4040
    },
    {
      "entropy": 5.529144048690796,
      "epoch": 0.388568683957733,
      "grad_norm": 1.0,
      "learning_rate": 0.0004990326201579335,
      "loss": 5.3662,
      "mean_token_accuracy": 0.19302817881107331,
      "num_tokens": 9279381.0,
      "step": 4045
    },
    {
      "entropy": 5.539267539978027,
      "epoch": 0.38904899135446686,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004990294418475274,
      "loss": 5.4709,
      "mean_token_accuracy": 0.17897016704082488,
      "num_tokens": 9292578.0,
      "step": 4050
    },
    {
      "entropy": 5.494290828704834,
      "epoch": 0.3895292987512008,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004990262583358231,
      "loss": 5.229,
      "mean_token_accuracy": 0.19708636105060579,
      "num_tokens": 9304116.0,
      "step": 4055
    },
    {
      "entropy": 5.544553422927857,
      "epoch": 0.3900096061479347,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004990230696228943,
      "loss": 5.459,
      "mean_token_accuracy": 0.18975809663534166,
      "num_tokens": 9315877.0,
      "step": 4060
    },
    {
      "entropy": 5.448103952407837,
      "epoch": 0.3904899135446686,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004990198757088149,
      "loss": 5.3419,
      "mean_token_accuracy": 0.19630881249904633,
      "num_tokens": 9326859.0,
      "step": 4065
    },
    {
      "entropy": 5.528612184524536,
      "epoch": 0.3909702209414025,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004990166765936593,
      "loss": 5.4155,
      "mean_token_accuracy": 0.1852904587984085,
      "num_tokens": 9338067.0,
      "step": 4070
    },
    {
      "entropy": 5.513800716400146,
      "epoch": 0.3914505283381364,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004990134722775016,
      "loss": 5.4513,
      "mean_token_accuracy": 0.19027821272611617,
      "num_tokens": 9349564.0,
      "step": 4075
    },
    {
      "entropy": 5.407103776931763,
      "epoch": 0.3919308357348703,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004990102627604162,
      "loss": 5.2413,
      "mean_token_accuracy": 0.1956426590681076,
      "num_tokens": 9359962.0,
      "step": 4080
    },
    {
      "entropy": 5.451745700836182,
      "epoch": 0.39241114313160425,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004990070480424778,
      "loss": 5.2618,
      "mean_token_accuracy": 0.20054119378328322,
      "num_tokens": 9370251.0,
      "step": 4085
    },
    {
      "entropy": 5.51818585395813,
      "epoch": 0.3928914505283381,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004990038281237608,
      "loss": 5.3917,
      "mean_token_accuracy": 0.18839506059885025,
      "num_tokens": 9381848.0,
      "step": 4090
    },
    {
      "entropy": 5.449072074890137,
      "epoch": 0.39337175792507206,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004990006030043401,
      "loss": 5.3092,
      "mean_token_accuracy": 0.1983541339635849,
      "num_tokens": 9392475.0,
      "step": 4095
    },
    {
      "entropy": 5.569762277603149,
      "epoch": 0.39385206532180594,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004989973726842906,
      "loss": 5.444,
      "mean_token_accuracy": 0.18511533886194229,
      "num_tokens": 9404014.0,
      "step": 4100
    },
    {
      "entropy": 5.449309587478638,
      "epoch": 0.3943323727185399,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004989941371636872,
      "loss": 5.3045,
      "mean_token_accuracy": 0.19107208400964737,
      "num_tokens": 9415590.0,
      "step": 4105
    },
    {
      "entropy": 5.562420129776001,
      "epoch": 0.39481268011527376,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004989908964426051,
      "loss": 5.4767,
      "mean_token_accuracy": 0.1867026135325432,
      "num_tokens": 9427486.0,
      "step": 4110
    },
    {
      "entropy": 5.526501893997192,
      "epoch": 0.3952929875120077,
      "grad_norm": 1.0,
      "learning_rate": 0.0004989876505211194,
      "loss": 5.3531,
      "mean_token_accuracy": 0.1961198627948761,
      "num_tokens": 9438212.0,
      "step": 4115
    },
    {
      "entropy": 5.4781585216522215,
      "epoch": 0.3957732949087416,
      "grad_norm": 1.0,
      "learning_rate": 0.0004989843993993056,
      "loss": 5.2989,
      "mean_token_accuracy": 0.20287738740444183,
      "num_tokens": 9450922.0,
      "step": 4120
    },
    {
      "entropy": 5.404292345046997,
      "epoch": 0.3962536023054755,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004989811430772392,
      "loss": 5.2786,
      "mean_token_accuracy": 0.19236334413290024,
      "num_tokens": 9461676.0,
      "step": 4125
    },
    {
      "entropy": 5.347488355636597,
      "epoch": 0.3967339097022094,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004989778815549957,
      "loss": 5.24,
      "mean_token_accuracy": 0.20357639640569686,
      "num_tokens": 9473517.0,
      "step": 4130
    },
    {
      "entropy": 5.565899229049682,
      "epoch": 0.39721421709894333,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004989746148326508,
      "loss": 5.3851,
      "mean_token_accuracy": 0.18520036488771438,
      "num_tokens": 9485660.0,
      "step": 4135
    },
    {
      "entropy": 5.5331018447875975,
      "epoch": 0.3976945244956772,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004989713429102805,
      "loss": 5.3907,
      "mean_token_accuracy": 0.19673049300909043,
      "num_tokens": 9495983.0,
      "step": 4140
    },
    {
      "entropy": 5.484323978424072,
      "epoch": 0.39817483189241115,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004989680657879607,
      "loss": 5.3414,
      "mean_token_accuracy": 0.18187149465084076,
      "num_tokens": 9507438.0,
      "step": 4145
    },
    {
      "entropy": 5.417265224456787,
      "epoch": 0.39865513928914503,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004989647834657675,
      "loss": 5.2306,
      "mean_token_accuracy": 0.19936590641736984,
      "num_tokens": 9517508.0,
      "step": 4150
    },
    {
      "entropy": 5.404279184341431,
      "epoch": 0.39913544668587897,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000498961495943777,
      "loss": 5.2593,
      "mean_token_accuracy": 0.20006999522447586,
      "num_tokens": 9527641.0,
      "step": 4155
    },
    {
      "entropy": 5.517825746536255,
      "epoch": 0.39961575408261285,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004989582032220656,
      "loss": 5.3594,
      "mean_token_accuracy": 0.19267345517873763,
      "num_tokens": 9538762.0,
      "step": 4160
    },
    {
      "entropy": 5.507742118835449,
      "epoch": 0.4000960614793468,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004989549053007096,
      "loss": 5.3795,
      "mean_token_accuracy": 0.1895456463098526,
      "num_tokens": 9550246.0,
      "step": 4165
    },
    {
      "entropy": 5.497794342041016,
      "epoch": 0.40057636887608067,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004989516021797858,
      "loss": 5.3223,
      "mean_token_accuracy": 0.1911580443382263,
      "num_tokens": 9562341.0,
      "step": 4170
    },
    {
      "entropy": 5.520681953430175,
      "epoch": 0.4010566762728146,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000498948293859371,
      "loss": 5.4028,
      "mean_token_accuracy": 0.190832157433033,
      "num_tokens": 9574484.0,
      "step": 4175
    },
    {
      "entropy": 5.501823234558105,
      "epoch": 0.4015369836695485,
      "grad_norm": 1.125,
      "learning_rate": 0.0004989449803395415,
      "loss": 5.3477,
      "mean_token_accuracy": 0.19757452756166458,
      "num_tokens": 9585241.0,
      "step": 4180
    },
    {
      "entropy": 5.497673177719117,
      "epoch": 0.4020172910662824,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004989416616203747,
      "loss": 5.3505,
      "mean_token_accuracy": 0.1926446497440338,
      "num_tokens": 9598683.0,
      "step": 4185
    },
    {
      "entropy": 5.506463527679443,
      "epoch": 0.4024975984630163,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004989383377019476,
      "loss": 5.377,
      "mean_token_accuracy": 0.19254742115736007,
      "num_tokens": 9610127.0,
      "step": 4190
    },
    {
      "entropy": 5.474005556106567,
      "epoch": 0.40297790585975024,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004989350085843371,
      "loss": 5.277,
      "mean_token_accuracy": 0.20008898973464967,
      "num_tokens": 9620009.0,
      "step": 4195
    },
    {
      "entropy": 5.456943798065185,
      "epoch": 0.4034582132564842,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004989316742676207,
      "loss": 5.3699,
      "mean_token_accuracy": 0.197726047039032,
      "num_tokens": 9632479.0,
      "step": 4200
    },
    {
      "entropy": 5.550093126296997,
      "epoch": 0.40393852065321806,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004989283347518757,
      "loss": 5.4215,
      "mean_token_accuracy": 0.1978586345911026,
      "num_tokens": 9644447.0,
      "step": 4205
    },
    {
      "entropy": 5.522815990447998,
      "epoch": 0.404418828049952,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004989249900371797,
      "loss": 5.3277,
      "mean_token_accuracy": 0.1889297991991043,
      "num_tokens": 9656886.0,
      "step": 4210
    },
    {
      "entropy": 5.524968194961548,
      "epoch": 0.4048991354466859,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004989216401236103,
      "loss": 5.3587,
      "mean_token_accuracy": 0.19858562350273132,
      "num_tokens": 9667079.0,
      "step": 4215
    },
    {
      "entropy": 5.5138630867004395,
      "epoch": 0.4053794428434198,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004989182850112455,
      "loss": 5.4155,
      "mean_token_accuracy": 0.19114071875810623,
      "num_tokens": 9678237.0,
      "step": 4220
    },
    {
      "entropy": 5.472230434417725,
      "epoch": 0.4058597502401537,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004989149247001629,
      "loss": 5.2796,
      "mean_token_accuracy": 0.19770650267601014,
      "num_tokens": 9688809.0,
      "step": 4225
    },
    {
      "entropy": 5.550781154632569,
      "epoch": 0.40634005763688763,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004989115591904407,
      "loss": 5.3776,
      "mean_token_accuracy": 0.19769087433815002,
      "num_tokens": 9700460.0,
      "step": 4230
    },
    {
      "entropy": 5.434881353378296,
      "epoch": 0.4068203650336215,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004989081884821569,
      "loss": 5.3034,
      "mean_token_accuracy": 0.1907295361161232,
      "num_tokens": 9711379.0,
      "step": 4235
    },
    {
      "entropy": 5.438016271591186,
      "epoch": 0.40730067243035545,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004989048125753899,
      "loss": 5.2464,
      "mean_token_accuracy": 0.19751791059970855,
      "num_tokens": 9722792.0,
      "step": 4240
    },
    {
      "entropy": 5.459333801269532,
      "epoch": 0.40778097982708933,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000498901431470218,
      "loss": 5.3005,
      "mean_token_accuracy": 0.20453376471996307,
      "num_tokens": 9733524.0,
      "step": 4245
    },
    {
      "entropy": 5.497627925872803,
      "epoch": 0.40826128722382327,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004988980451667198,
      "loss": 5.365,
      "mean_token_accuracy": 0.19118060022592545,
      "num_tokens": 9745430.0,
      "step": 4250
    },
    {
      "entropy": 5.4426673412322994,
      "epoch": 0.40874159462055715,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004988946536649737,
      "loss": 5.2733,
      "mean_token_accuracy": 0.19136472195386886,
      "num_tokens": 9756927.0,
      "step": 4255
    },
    {
      "entropy": 5.4981752872467045,
      "epoch": 0.4092219020172911,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004988912569650585,
      "loss": 5.3144,
      "mean_token_accuracy": 0.1946692317724228,
      "num_tokens": 9768350.0,
      "step": 4260
    },
    {
      "entropy": 5.391961908340454,
      "epoch": 0.40970220941402496,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004988878550670533,
      "loss": 5.2514,
      "mean_token_accuracy": 0.19954414069652557,
      "num_tokens": 9779238.0,
      "step": 4265
    },
    {
      "entropy": 5.539362478256225,
      "epoch": 0.4101825168107589,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004988844479710369,
      "loss": 5.4227,
      "mean_token_accuracy": 0.18671586662530898,
      "num_tokens": 9789262.0,
      "step": 4270
    },
    {
      "entropy": 5.55619854927063,
      "epoch": 0.4106628242074928,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004988810356770884,
      "loss": 5.3967,
      "mean_token_accuracy": 0.18707060366868972,
      "num_tokens": 9800909.0,
      "step": 4275
    },
    {
      "entropy": 5.474341297149659,
      "epoch": 0.4111431316042267,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000498877618185287,
      "loss": 5.3564,
      "mean_token_accuracy": 0.19191308319568634,
      "num_tokens": 9812538.0,
      "step": 4280
    },
    {
      "entropy": 5.504796314239502,
      "epoch": 0.4116234390009606,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004988741954957121,
      "loss": 5.3438,
      "mean_token_accuracy": 0.19248106777668,
      "num_tokens": 9824071.0,
      "step": 4285
    },
    {
      "entropy": 5.6006495475769045,
      "epoch": 0.41210374639769454,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004988707676084432,
      "loss": 5.4576,
      "mean_token_accuracy": 0.18734299689531325,
      "num_tokens": 9836713.0,
      "step": 4290
    },
    {
      "entropy": 5.442639112472534,
      "epoch": 0.4125840537944284,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004988673345235597,
      "loss": 5.2816,
      "mean_token_accuracy": 0.19828231036663055,
      "num_tokens": 9847336.0,
      "step": 4295
    },
    {
      "entropy": 5.3893345355987545,
      "epoch": 0.41306436119116235,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004988638962411416,
      "loss": 5.2755,
      "mean_token_accuracy": 0.19688578248023986,
      "num_tokens": 9858567.0,
      "step": 4300
    },
    {
      "entropy": 5.549181509017944,
      "epoch": 0.41354466858789624,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004988604527612685,
      "loss": 5.3433,
      "mean_token_accuracy": 0.188631808757782,
      "num_tokens": 9869254.0,
      "step": 4305
    },
    {
      "entropy": 5.480014610290527,
      "epoch": 0.4140249759846302,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004988570040840205,
      "loss": 5.2878,
      "mean_token_accuracy": 0.1991993695497513,
      "num_tokens": 9879635.0,
      "step": 4310
    },
    {
      "entropy": 5.4838696956634525,
      "epoch": 0.41450528338136405,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004988535502094774,
      "loss": 5.3499,
      "mean_token_accuracy": 0.1955707401037216,
      "num_tokens": 9891728.0,
      "step": 4315
    },
    {
      "entropy": 5.380111885070801,
      "epoch": 0.414985590778098,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004988500911377198,
      "loss": 5.1665,
      "mean_token_accuracy": 0.20012396574020386,
      "num_tokens": 9903140.0,
      "step": 4320
    },
    {
      "entropy": 5.461152172088623,
      "epoch": 0.41546589817483187,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004988466268688276,
      "loss": 5.3057,
      "mean_token_accuracy": 0.19505317509174347,
      "num_tokens": 9915562.0,
      "step": 4325
    },
    {
      "entropy": 5.452009868621826,
      "epoch": 0.4159462055715658,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004988431574028814,
      "loss": 5.2535,
      "mean_token_accuracy": 0.1924479991197586,
      "num_tokens": 9925720.0,
      "step": 4330
    },
    {
      "entropy": 5.45178861618042,
      "epoch": 0.4164265129682997,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004988396827399618,
      "loss": 5.3253,
      "mean_token_accuracy": 0.1864607259631157,
      "num_tokens": 9936492.0,
      "step": 4335
    },
    {
      "entropy": 5.456461381912232,
      "epoch": 0.4169068203650336,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004988362028801495,
      "loss": 5.2863,
      "mean_token_accuracy": 0.19337694197893143,
      "num_tokens": 9948658.0,
      "step": 4340
    },
    {
      "entropy": 5.352551698684692,
      "epoch": 0.4173871277617675,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004988327178235253,
      "loss": 5.2053,
      "mean_token_accuracy": 0.20683322846889496,
      "num_tokens": 9959710.0,
      "step": 4345
    },
    {
      "entropy": 5.393221235275268,
      "epoch": 0.41786743515850144,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004988292275701699,
      "loss": 5.2125,
      "mean_token_accuracy": 0.19954641163349152,
      "num_tokens": 9970302.0,
      "step": 4350
    },
    {
      "entropy": 5.471257591247559,
      "epoch": 0.4183477425552353,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004988257321201646,
      "loss": 5.3299,
      "mean_token_accuracy": 0.19249966144561767,
      "num_tokens": 9981776.0,
      "step": 4355
    },
    {
      "entropy": 5.5098655223846436,
      "epoch": 0.41882804995196926,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004988222314735902,
      "loss": 5.3041,
      "mean_token_accuracy": 0.19173195511102675,
      "num_tokens": 9993471.0,
      "step": 4360
    },
    {
      "entropy": 5.464787006378174,
      "epoch": 0.41930835734870314,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004988187256305284,
      "loss": 5.3324,
      "mean_token_accuracy": 0.19311019480228425,
      "num_tokens": 10003818.0,
      "step": 4365
    },
    {
      "entropy": 5.452540111541748,
      "epoch": 0.4197886647454371,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004988152145910603,
      "loss": 5.3958,
      "mean_token_accuracy": 0.1840040236711502,
      "num_tokens": 10015507.0,
      "step": 4370
    },
    {
      "entropy": 5.454175996780395,
      "epoch": 0.420268972142171,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004988116983552675,
      "loss": 5.24,
      "mean_token_accuracy": 0.19901308417320251,
      "num_tokens": 10026994.0,
      "step": 4375
    },
    {
      "entropy": 5.51434555053711,
      "epoch": 0.4207492795389049,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004988081769232317,
      "loss": 5.3841,
      "mean_token_accuracy": 0.18538453578948974,
      "num_tokens": 10039547.0,
      "step": 4380
    },
    {
      "entropy": 5.497226095199585,
      "epoch": 0.42122958693563883,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004988046502950346,
      "loss": 5.3996,
      "mean_token_accuracy": 0.18713475316762923,
      "num_tokens": 10050568.0,
      "step": 4385
    },
    {
      "entropy": 5.465748262405396,
      "epoch": 0.4217098943323727,
      "grad_norm": 1.03125,
      "learning_rate": 0.000498801118470758,
      "loss": 5.3053,
      "mean_token_accuracy": 0.19666541069746019,
      "num_tokens": 10061780.0,
      "step": 4390
    },
    {
      "entropy": 5.418157529830933,
      "epoch": 0.42219020172910665,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000498797581450484,
      "loss": 5.2778,
      "mean_token_accuracy": 0.19300197064876556,
      "num_tokens": 10073101.0,
      "step": 4395
    },
    {
      "entropy": 5.395181512832641,
      "epoch": 0.42267050912584053,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004987940392342948,
      "loss": 5.2187,
      "mean_token_accuracy": 0.2072927638888359,
      "num_tokens": 10084397.0,
      "step": 4400
    },
    {
      "entropy": 5.450293445587159,
      "epoch": 0.42315081652257447,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004987904918222726,
      "loss": 5.2945,
      "mean_token_accuracy": 0.2025788187980652,
      "num_tokens": 10095635.0,
      "step": 4405
    },
    {
      "entropy": 5.453633403778076,
      "epoch": 0.42363112391930835,
      "grad_norm": 1.0,
      "learning_rate": 0.0004987869392144996,
      "loss": 5.3574,
      "mean_token_accuracy": 0.1989255353808403,
      "num_tokens": 10107744.0,
      "step": 4410
    },
    {
      "entropy": 5.467885255813599,
      "epoch": 0.4241114313160423,
      "grad_norm": 1.0,
      "learning_rate": 0.0004987833814110584,
      "loss": 5.2895,
      "mean_token_accuracy": 0.19859737455844878,
      "num_tokens": 10118552.0,
      "step": 4415
    },
    {
      "entropy": 5.5184972286224365,
      "epoch": 0.42459173871277617,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004987798184120316,
      "loss": 5.4094,
      "mean_token_accuracy": 0.18422331511974335,
      "num_tokens": 10130813.0,
      "step": 4420
    },
    {
      "entropy": 5.478092527389526,
      "epoch": 0.4250720461095101,
      "grad_norm": 1.0,
      "learning_rate": 0.0004987762502175018,
      "loss": 5.4202,
      "mean_token_accuracy": 0.1944909319281578,
      "num_tokens": 10142196.0,
      "step": 4425
    },
    {
      "entropy": 5.4415602684021,
      "epoch": 0.425552353506244,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000498772676827552,
      "loss": 5.3258,
      "mean_token_accuracy": 0.20208674818277358,
      "num_tokens": 10155012.0,
      "step": 4430
    },
    {
      "entropy": 5.535742902755738,
      "epoch": 0.4260326609029779,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004987690982422652,
      "loss": 5.3964,
      "mean_token_accuracy": 0.1853977084159851,
      "num_tokens": 10166160.0,
      "step": 4435
    },
    {
      "entropy": 5.494490146636963,
      "epoch": 0.4265129682997118,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004987655144617243,
      "loss": 5.1972,
      "mean_token_accuracy": 0.20469662696123123,
      "num_tokens": 10177663.0,
      "step": 4440
    },
    {
      "entropy": 5.355368185043335,
      "epoch": 0.42699327569644574,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004987619254860126,
      "loss": 5.2806,
      "mean_token_accuracy": 0.19842516481876374,
      "num_tokens": 10189127.0,
      "step": 4445
    },
    {
      "entropy": 5.589107513427734,
      "epoch": 0.4274735830931796,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004987583313152134,
      "loss": 5.4552,
      "mean_token_accuracy": 0.18255480378866196,
      "num_tokens": 10200765.0,
      "step": 4450
    },
    {
      "entropy": 5.511512899398804,
      "epoch": 0.42795389048991356,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004987547319494104,
      "loss": 5.3025,
      "mean_token_accuracy": 0.19883785098791124,
      "num_tokens": 10213047.0,
      "step": 4455
    },
    {
      "entropy": 5.406325197219848,
      "epoch": 0.42843419788664744,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004987511273886867,
      "loss": 5.2606,
      "mean_token_accuracy": 0.19982735961675643,
      "num_tokens": 10223790.0,
      "step": 4460
    },
    {
      "entropy": 5.44874906539917,
      "epoch": 0.4289145052833814,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004987475176331263,
      "loss": 5.2505,
      "mean_token_accuracy": 0.19563947319984437,
      "num_tokens": 10235077.0,
      "step": 4465
    },
    {
      "entropy": 5.48102707862854,
      "epoch": 0.42939481268011526,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004987439026828129,
      "loss": 5.3316,
      "mean_token_accuracy": 0.1842127338051796,
      "num_tokens": 10248844.0,
      "step": 4470
    },
    {
      "entropy": 5.429107141494751,
      "epoch": 0.4298751200768492,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004987402825378305,
      "loss": 5.2619,
      "mean_token_accuracy": 0.20477957427501678,
      "num_tokens": 10259013.0,
      "step": 4475
    },
    {
      "entropy": 5.377386236190796,
      "epoch": 0.4303554274735831,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004987366571982631,
      "loss": 5.2802,
      "mean_token_accuracy": 0.19595861285924912,
      "num_tokens": 10271415.0,
      "step": 4480
    },
    {
      "entropy": 5.398539686203003,
      "epoch": 0.430835734870317,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004987330266641948,
      "loss": 5.2841,
      "mean_token_accuracy": 0.198825865983963,
      "num_tokens": 10283204.0,
      "step": 4485
    },
    {
      "entropy": 5.583152008056641,
      "epoch": 0.4313160422670509,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004987293909357101,
      "loss": 5.4546,
      "mean_token_accuracy": 0.18715879917144776,
      "num_tokens": 10294111.0,
      "step": 4490
    },
    {
      "entropy": 5.563319492340088,
      "epoch": 0.43179634966378483,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004987257500128933,
      "loss": 5.3408,
      "mean_token_accuracy": 0.19044281542301178,
      "num_tokens": 10305417.0,
      "step": 4495
    },
    {
      "entropy": 5.380277633666992,
      "epoch": 0.4322766570605187,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004987221038958288,
      "loss": 5.2229,
      "mean_token_accuracy": 0.19826748967170715,
      "num_tokens": 10316416.0,
      "step": 4500
    },
    {
      "entropy": 5.474719905853272,
      "epoch": 0.43275696445725265,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004987184525846015,
      "loss": 5.3903,
      "mean_token_accuracy": 0.19537886679172517,
      "num_tokens": 10328251.0,
      "step": 4505
    },
    {
      "entropy": 5.545657396316528,
      "epoch": 0.4332372718539865,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004987147960792958,
      "loss": 5.3519,
      "mean_token_accuracy": 0.20158660411834717,
      "num_tokens": 10340189.0,
      "step": 4510
    },
    {
      "entropy": 5.494429349899292,
      "epoch": 0.43371757925072046,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004987111343799971,
      "loss": 5.2836,
      "mean_token_accuracy": 0.197966331243515,
      "num_tokens": 10351660.0,
      "step": 4515
    },
    {
      "entropy": 5.421718645095825,
      "epoch": 0.43419788664745435,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00049870746748679,
      "loss": 5.3319,
      "mean_token_accuracy": 0.19888116866350175,
      "num_tokens": 10362240.0,
      "step": 4520
    },
    {
      "entropy": 5.557341003417969,
      "epoch": 0.4346781940441883,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004987037953997598,
      "loss": 5.3484,
      "mean_token_accuracy": 0.19845753610134126,
      "num_tokens": 10373646.0,
      "step": 4525
    },
    {
      "entropy": 5.476250886917114,
      "epoch": 0.43515850144092216,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004987001181189918,
      "loss": 5.3796,
      "mean_token_accuracy": 0.19115481227636338,
      "num_tokens": 10385540.0,
      "step": 4530
    },
    {
      "entropy": 5.437349319458008,
      "epoch": 0.4356388088376561,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004986964356445713,
      "loss": 5.3696,
      "mean_token_accuracy": 0.18966257721185684,
      "num_tokens": 10396641.0,
      "step": 4535
    },
    {
      "entropy": 5.506645393371582,
      "epoch": 0.43611911623439004,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004986927479765837,
      "loss": 5.3367,
      "mean_token_accuracy": 0.18645385652780533,
      "num_tokens": 10408165.0,
      "step": 4540
    },
    {
      "entropy": 5.495576858520508,
      "epoch": 0.4365994236311239,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004986890551151148,
      "loss": 5.2924,
      "mean_token_accuracy": 0.2013896584510803,
      "num_tokens": 10419646.0,
      "step": 4545
    },
    {
      "entropy": 5.476487255096435,
      "epoch": 0.43707973102785785,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004986853570602503,
      "loss": 5.3203,
      "mean_token_accuracy": 0.19487122297286988,
      "num_tokens": 10431051.0,
      "step": 4550
    },
    {
      "entropy": 5.51167311668396,
      "epoch": 0.43756003842459174,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004986816538120758,
      "loss": 5.3834,
      "mean_token_accuracy": 0.19041346460580827,
      "num_tokens": 10441833.0,
      "step": 4555
    },
    {
      "entropy": 5.485381746292115,
      "epoch": 0.43804034582132567,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004986779453706778,
      "loss": 5.34,
      "mean_token_accuracy": 0.1925880566239357,
      "num_tokens": 10453544.0,
      "step": 4560
    },
    {
      "entropy": 5.440112972259522,
      "epoch": 0.43852065321805955,
      "grad_norm": 1.0,
      "learning_rate": 0.0004986742317361419,
      "loss": 5.1994,
      "mean_token_accuracy": 0.20150742083787918,
      "num_tokens": 10464140.0,
      "step": 4565
    },
    {
      "entropy": 5.35095181465149,
      "epoch": 0.4390009606147935,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004986705129085546,
      "loss": 5.2498,
      "mean_token_accuracy": 0.199386166036129,
      "num_tokens": 10475739.0,
      "step": 4570
    },
    {
      "entropy": 5.374697923660278,
      "epoch": 0.43948126801152737,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004986667888880021,
      "loss": 5.306,
      "mean_token_accuracy": 0.19678469151258468,
      "num_tokens": 10487451.0,
      "step": 4575
    },
    {
      "entropy": 5.5155011177062985,
      "epoch": 0.4399615754082613,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004986630596745709,
      "loss": 5.3544,
      "mean_token_accuracy": 0.18970419615507125,
      "num_tokens": 10499367.0,
      "step": 4580
    },
    {
      "entropy": 5.475177049636841,
      "epoch": 0.4404418828049952,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004986593252683477,
      "loss": 5.1871,
      "mean_token_accuracy": 0.20306814312934876,
      "num_tokens": 10511192.0,
      "step": 4585
    },
    {
      "entropy": 5.327244138717651,
      "epoch": 0.4409221902017291,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004986555856694191,
      "loss": 5.2036,
      "mean_token_accuracy": 0.1945111259818077,
      "num_tokens": 10522754.0,
      "step": 4590
    },
    {
      "entropy": 5.475100040435791,
      "epoch": 0.441402497598463,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004986518408778718,
      "loss": 5.348,
      "mean_token_accuracy": 0.19128721654415132,
      "num_tokens": 10533439.0,
      "step": 4595
    },
    {
      "entropy": 5.515309047698975,
      "epoch": 0.44188280499519694,
      "grad_norm": 1.0,
      "learning_rate": 0.0004986480908937929,
      "loss": 5.3042,
      "mean_token_accuracy": 0.19626817256212234,
      "num_tokens": 10543654.0,
      "step": 4600
    },
    {
      "entropy": 5.387768936157227,
      "epoch": 0.4423631123919308,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004986443357172695,
      "loss": 5.2121,
      "mean_token_accuracy": 0.20078853964805604,
      "num_tokens": 10554936.0,
      "step": 4605
    },
    {
      "entropy": 5.446403360366821,
      "epoch": 0.44284341978866476,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004986405753483887,
      "loss": 5.2854,
      "mean_token_accuracy": 0.19537069499492646,
      "num_tokens": 10566410.0,
      "step": 4610
    },
    {
      "entropy": 5.405148506164551,
      "epoch": 0.44332372718539864,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004986368097872377,
      "loss": 5.2625,
      "mean_token_accuracy": 0.20326067954301835,
      "num_tokens": 10577258.0,
      "step": 4615
    },
    {
      "entropy": 5.4609393119812015,
      "epoch": 0.4438040345821326,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004986330390339042,
      "loss": 5.374,
      "mean_token_accuracy": 0.18600859493017197,
      "num_tokens": 10588783.0,
      "step": 4620
    },
    {
      "entropy": 5.450092601776123,
      "epoch": 0.44428434197886646,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004986292630884755,
      "loss": 5.2481,
      "mean_token_accuracy": 0.19656331688165665,
      "num_tokens": 10599778.0,
      "step": 4625
    },
    {
      "entropy": 5.446336889266968,
      "epoch": 0.4447646493756004,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004986254819510393,
      "loss": 5.2785,
      "mean_token_accuracy": 0.20218563824892044,
      "num_tokens": 10611211.0,
      "step": 4630
    },
    {
      "entropy": 5.3755090713500975,
      "epoch": 0.4452449567723343,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004986216956216835,
      "loss": 5.207,
      "mean_token_accuracy": 0.19774384051561356,
      "num_tokens": 10624535.0,
      "step": 4635
    },
    {
      "entropy": 5.404080009460449,
      "epoch": 0.4457252641690682,
      "grad_norm": 1.078125,
      "learning_rate": 0.000498617904100496,
      "loss": 5.2558,
      "mean_token_accuracy": 0.20761176943778992,
      "num_tokens": 10635206.0,
      "step": 4640
    },
    {
      "entropy": 5.434352111816406,
      "epoch": 0.4462055715658021,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004986141073875646,
      "loss": 5.2934,
      "mean_token_accuracy": 0.20206933170557023,
      "num_tokens": 10646993.0,
      "step": 4645
    },
    {
      "entropy": 5.472942113876343,
      "epoch": 0.44668587896253603,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004986103054829779,
      "loss": 5.238,
      "mean_token_accuracy": 0.20501162111759186,
      "num_tokens": 10657262.0,
      "step": 4650
    },
    {
      "entropy": 5.384912443161011,
      "epoch": 0.4471661863592699,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004986064983868237,
      "loss": 5.2686,
      "mean_token_accuracy": 0.19664817154407502,
      "num_tokens": 10668962.0,
      "step": 4655
    },
    {
      "entropy": 5.4284168720245365,
      "epoch": 0.44764649375600385,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004986026860991906,
      "loss": 5.2701,
      "mean_token_accuracy": 0.20509240478277208,
      "num_tokens": 10680183.0,
      "step": 4660
    },
    {
      "entropy": 5.4118485927581785,
      "epoch": 0.44812680115273773,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004985988686201672,
      "loss": 5.2222,
      "mean_token_accuracy": 0.20363913774490355,
      "num_tokens": 10691221.0,
      "step": 4665
    },
    {
      "entropy": 5.391137552261353,
      "epoch": 0.44860710854947167,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004985950459498419,
      "loss": 5.2516,
      "mean_token_accuracy": 0.19561606347560884,
      "num_tokens": 10703462.0,
      "step": 4670
    },
    {
      "entropy": 5.427238655090332,
      "epoch": 0.44908741594620555,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004985912180883037,
      "loss": 5.2361,
      "mean_token_accuracy": 0.20176161378622054,
      "num_tokens": 10713774.0,
      "step": 4675
    },
    {
      "entropy": 5.503265142440796,
      "epoch": 0.4495677233429395,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004985873850356411,
      "loss": 5.4299,
      "mean_token_accuracy": 0.18668080121278763,
      "num_tokens": 10725109.0,
      "step": 4680
    },
    {
      "entropy": 5.374135065078735,
      "epoch": 0.45004803073967337,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004985835467919436,
      "loss": 5.2556,
      "mean_token_accuracy": 0.20445990413427353,
      "num_tokens": 10735695.0,
      "step": 4685
    },
    {
      "entropy": 5.471426630020142,
      "epoch": 0.4505283381364073,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004985797033572999,
      "loss": 5.3379,
      "mean_token_accuracy": 0.19836002886295317,
      "num_tokens": 10747554.0,
      "step": 4690
    },
    {
      "entropy": 5.485050630569458,
      "epoch": 0.4510086455331412,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004985758547317994,
      "loss": 5.3506,
      "mean_token_accuracy": 0.19426705986261367,
      "num_tokens": 10759175.0,
      "step": 4695
    },
    {
      "entropy": 5.399711275100708,
      "epoch": 0.4514889529298751,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004985720009155315,
      "loss": 5.2068,
      "mean_token_accuracy": 0.20349736958742143,
      "num_tokens": 10770365.0,
      "step": 4700
    },
    {
      "entropy": 5.403173398971558,
      "epoch": 0.45196926032660906,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004985681419085856,
      "loss": 5.3162,
      "mean_token_accuracy": 0.19575918912887574,
      "num_tokens": 10783096.0,
      "step": 4705
    },
    {
      "entropy": 5.527654552459717,
      "epoch": 0.45244956772334294,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004985642777110513,
      "loss": 5.3945,
      "mean_token_accuracy": 0.18861877173185349,
      "num_tokens": 10795312.0,
      "step": 4710
    },
    {
      "entropy": 5.434057331085205,
      "epoch": 0.4529298751200769,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004985604083230183,
      "loss": 5.2325,
      "mean_token_accuracy": 0.20321549028158187,
      "num_tokens": 10806451.0,
      "step": 4715
    },
    {
      "entropy": 5.502018451690674,
      "epoch": 0.45341018251681076,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004985565337445765,
      "loss": 5.3758,
      "mean_token_accuracy": 0.19537254124879838,
      "num_tokens": 10818185.0,
      "step": 4720
    },
    {
      "entropy": 5.43809666633606,
      "epoch": 0.4538904899135447,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004985526539758158,
      "loss": 5.2414,
      "mean_token_accuracy": 0.207122240960598,
      "num_tokens": 10830203.0,
      "step": 4725
    },
    {
      "entropy": 5.4323817729949955,
      "epoch": 0.4543707973102786,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004985487690168263,
      "loss": 5.2848,
      "mean_token_accuracy": 0.18830326050519944,
      "num_tokens": 10842953.0,
      "step": 4730
    },
    {
      "entropy": 5.536609172821045,
      "epoch": 0.4548511047070125,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000498544878867698,
      "loss": 5.4359,
      "mean_token_accuracy": 0.1911514326930046,
      "num_tokens": 10853322.0,
      "step": 4735
    },
    {
      "entropy": 5.4663468360900875,
      "epoch": 0.4553314121037464,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004985409835285215,
      "loss": 5.3258,
      "mean_token_accuracy": 0.19328247606754304,
      "num_tokens": 10865480.0,
      "step": 4740
    },
    {
      "entropy": 5.419348192214966,
      "epoch": 0.45581171950048033,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004985370829993873,
      "loss": 5.1782,
      "mean_token_accuracy": 0.20896158069372178,
      "num_tokens": 10875700.0,
      "step": 4745
    },
    {
      "entropy": 5.381450128555298,
      "epoch": 0.4562920268972142,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004985331772803857,
      "loss": 5.2798,
      "mean_token_accuracy": 0.20026755034923555,
      "num_tokens": 10887350.0,
      "step": 4750
    },
    {
      "entropy": 5.4131402492523195,
      "epoch": 0.45677233429394815,
      "grad_norm": 1.0,
      "learning_rate": 0.0004985292663716074,
      "loss": 5.2119,
      "mean_token_accuracy": 0.20269576758146285,
      "num_tokens": 10899355.0,
      "step": 4755
    },
    {
      "entropy": 5.435826301574707,
      "epoch": 0.457252641690682,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004985253502731435,
      "loss": 5.2401,
      "mean_token_accuracy": 0.20178792178630828,
      "num_tokens": 10911521.0,
      "step": 4760
    },
    {
      "entropy": 5.380969095230102,
      "epoch": 0.45773294908741596,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004985214289850845,
      "loss": 5.3019,
      "mean_token_accuracy": 0.19617623090744019,
      "num_tokens": 10923363.0,
      "step": 4765
    },
    {
      "entropy": 5.430521678924561,
      "epoch": 0.45821325648414984,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004985175025075217,
      "loss": 5.2856,
      "mean_token_accuracy": 0.20007649809122086,
      "num_tokens": 10934965.0,
      "step": 4770
    },
    {
      "entropy": 5.434817981719971,
      "epoch": 0.4586935638808838,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004985135708405462,
      "loss": 5.289,
      "mean_token_accuracy": 0.19134962558746338,
      "num_tokens": 10946126.0,
      "step": 4775
    },
    {
      "entropy": 5.448909282684326,
      "epoch": 0.45917387127761766,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004985096339842493,
      "loss": 5.3693,
      "mean_token_accuracy": 0.19597102999687194,
      "num_tokens": 10957667.0,
      "step": 4780
    },
    {
      "entropy": 5.520076704025269,
      "epoch": 0.4596541786743516,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004985056919387224,
      "loss": 5.4392,
      "mean_token_accuracy": 0.18832276463508607,
      "num_tokens": 10969999.0,
      "step": 4785
    },
    {
      "entropy": 5.416590023040771,
      "epoch": 0.4601344860710855,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004985017447040569,
      "loss": 5.1346,
      "mean_token_accuracy": 0.20553265661001205,
      "num_tokens": 10980672.0,
      "step": 4790
    },
    {
      "entropy": 5.495411825180054,
      "epoch": 0.4606147934678194,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004984977922803447,
      "loss": 5.2499,
      "mean_token_accuracy": 0.198009392619133,
      "num_tokens": 10991471.0,
      "step": 4795
    },
    {
      "entropy": 5.4282470703125,
      "epoch": 0.4610951008645533,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004984938346676772,
      "loss": 5.3584,
      "mean_token_accuracy": 0.19561775922775268,
      "num_tokens": 11001960.0,
      "step": 4800
    },
    {
      "entropy": 5.459453105926514,
      "epoch": 0.46157540826128723,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004984898718661468,
      "loss": 5.214,
      "mean_token_accuracy": 0.19864818453788757,
      "num_tokens": 11013274.0,
      "step": 4805
    },
    {
      "entropy": 5.433310556411743,
      "epoch": 0.4620557156580211,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004984859038758451,
      "loss": 5.2316,
      "mean_token_accuracy": 0.20722649693489076,
      "num_tokens": 11024635.0,
      "step": 4810
    },
    {
      "entropy": 5.39674940109253,
      "epoch": 0.46253602305475505,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004984819306968642,
      "loss": 5.2562,
      "mean_token_accuracy": 0.19551037847995759,
      "num_tokens": 11036317.0,
      "step": 4815
    },
    {
      "entropy": 5.399358463287354,
      "epoch": 0.46301633045148893,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004984779523292966,
      "loss": 5.2454,
      "mean_token_accuracy": 0.20224211364984512,
      "num_tokens": 11047776.0,
      "step": 4820
    },
    {
      "entropy": 5.45464768409729,
      "epoch": 0.46349663784822287,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004984739687732345,
      "loss": 5.2706,
      "mean_token_accuracy": 0.1989618420600891,
      "num_tokens": 11058746.0,
      "step": 4825
    },
    {
      "entropy": 5.489168930053711,
      "epoch": 0.46397694524495675,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004984699800287705,
      "loss": 5.3085,
      "mean_token_accuracy": 0.19539135098457336,
      "num_tokens": 11068892.0,
      "step": 4830
    },
    {
      "entropy": 5.417141962051391,
      "epoch": 0.4644572526416907,
      "grad_norm": 0.9375,
      "learning_rate": 0.000498465986095997,
      "loss": 5.2951,
      "mean_token_accuracy": 0.19863324016332626,
      "num_tokens": 11080038.0,
      "step": 4835
    },
    {
      "entropy": 5.4287285804748535,
      "epoch": 0.46493756003842457,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004984619869750069,
      "loss": 5.2336,
      "mean_token_accuracy": 0.19549733251333237,
      "num_tokens": 11091261.0,
      "step": 4840
    },
    {
      "entropy": 5.365886974334717,
      "epoch": 0.4654178674351585,
      "grad_norm": 0.84375,
      "learning_rate": 0.000498457982665893,
      "loss": 5.1959,
      "mean_token_accuracy": 0.20959132462739943,
      "num_tokens": 11103015.0,
      "step": 4845
    },
    {
      "entropy": 5.398194742202759,
      "epoch": 0.4658981748318924,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004984539731687483,
      "loss": 5.1875,
      "mean_token_accuracy": 0.20573316812515258,
      "num_tokens": 11115075.0,
      "step": 4850
    },
    {
      "entropy": 5.422974729537964,
      "epoch": 0.4663784822286263,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004984499584836659,
      "loss": 5.1639,
      "mean_token_accuracy": 0.19655676186084747,
      "num_tokens": 11125937.0,
      "step": 4855
    },
    {
      "entropy": 5.357757711410523,
      "epoch": 0.4668587896253602,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000498445938610739,
      "loss": 5.261,
      "mean_token_accuracy": 0.1969615936279297,
      "num_tokens": 11136863.0,
      "step": 4860
    },
    {
      "entropy": 5.46019606590271,
      "epoch": 0.46733909702209414,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004984419135500608,
      "loss": 5.3345,
      "mean_token_accuracy": 0.1884562835097313,
      "num_tokens": 11149934.0,
      "step": 4865
    },
    {
      "entropy": 5.426164960861206,
      "epoch": 0.4678194044188281,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004984378833017249,
      "loss": 5.2615,
      "mean_token_accuracy": 0.19857032001018524,
      "num_tokens": 11162208.0,
      "step": 4870
    },
    {
      "entropy": 5.373647451400757,
      "epoch": 0.46829971181556196,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004984338478658248,
      "loss": 5.3099,
      "mean_token_accuracy": 0.19991184175014495,
      "num_tokens": 11172892.0,
      "step": 4875
    },
    {
      "entropy": 5.405699920654297,
      "epoch": 0.4687800192122959,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004984298072424542,
      "loss": 5.2609,
      "mean_token_accuracy": 0.20024777799844742,
      "num_tokens": 11184339.0,
      "step": 4880
    },
    {
      "entropy": 5.495201969146729,
      "epoch": 0.4692603266090298,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000498425761431707,
      "loss": 5.3249,
      "mean_token_accuracy": 0.19282843321561813,
      "num_tokens": 11195681.0,
      "step": 4885
    },
    {
      "entropy": 5.488913202285767,
      "epoch": 0.4697406340057637,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000498421710433677,
      "loss": 5.2586,
      "mean_token_accuracy": 0.1919654995203018,
      "num_tokens": 11207982.0,
      "step": 4890
    },
    {
      "entropy": 5.44200701713562,
      "epoch": 0.4702209414024976,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004984176542484584,
      "loss": 5.3191,
      "mean_token_accuracy": 0.1956244334578514,
      "num_tokens": 11218901.0,
      "step": 4895
    },
    {
      "entropy": 5.394407844543457,
      "epoch": 0.47070124879923153,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004984135928761452,
      "loss": 5.2852,
      "mean_token_accuracy": 0.20009127259254456,
      "num_tokens": 11230764.0,
      "step": 4900
    },
    {
      "entropy": 5.3779762268066404,
      "epoch": 0.4711815561959654,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004984095263168317,
      "loss": 5.201,
      "mean_token_accuracy": 0.20654135346412658,
      "num_tokens": 11241489.0,
      "step": 4905
    },
    {
      "entropy": 5.292277431488037,
      "epoch": 0.47166186359269935,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004984054545706124,
      "loss": 5.184,
      "mean_token_accuracy": 0.20778871178627015,
      "num_tokens": 11254500.0,
      "step": 4910
    },
    {
      "entropy": 5.515752506256104,
      "epoch": 0.47214217098943323,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000498401377637582,
      "loss": 5.3205,
      "mean_token_accuracy": 0.19741638749837875,
      "num_tokens": 11266235.0,
      "step": 4915
    },
    {
      "entropy": 5.452426195144653,
      "epoch": 0.47262247838616717,
      "grad_norm": 0.921875,
      "learning_rate": 0.000498397295517835,
      "loss": 5.2729,
      "mean_token_accuracy": 0.2045632541179657,
      "num_tokens": 11277471.0,
      "step": 4920
    },
    {
      "entropy": 5.277326488494873,
      "epoch": 0.47310278578290105,
      "grad_norm": 0.875,
      "learning_rate": 0.0004983932082114659,
      "loss": 5.191,
      "mean_token_accuracy": 0.19845269620418549,
      "num_tokens": 11290501.0,
      "step": 4925
    },
    {
      "entropy": 5.3753087520599365,
      "epoch": 0.473583093179635,
      "grad_norm": 1.0,
      "learning_rate": 0.0004983891157185699,
      "loss": 5.2264,
      "mean_token_accuracy": 0.20023567825555802,
      "num_tokens": 11301933.0,
      "step": 4930
    },
    {
      "entropy": 5.497870826721192,
      "epoch": 0.47406340057636887,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004983850180392421,
      "loss": 5.2835,
      "mean_token_accuracy": 0.2010246366262436,
      "num_tokens": 11314539.0,
      "step": 4935
    },
    {
      "entropy": 5.398234462738037,
      "epoch": 0.4745437079731028,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004983809151735775,
      "loss": 5.2049,
      "mean_token_accuracy": 0.20689275115728378,
      "num_tokens": 11326641.0,
      "step": 4940
    },
    {
      "entropy": 5.431383752822876,
      "epoch": 0.4750240153698367,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004983768071216713,
      "loss": 5.2949,
      "mean_token_accuracy": 0.19610632658004762,
      "num_tokens": 11337214.0,
      "step": 4945
    },
    {
      "entropy": 5.4590826511383055,
      "epoch": 0.4755043227665706,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004983726938836189,
      "loss": 5.2547,
      "mean_token_accuracy": 0.19909323155879974,
      "num_tokens": 11349540.0,
      "step": 4950
    },
    {
      "entropy": 5.427983903884888,
      "epoch": 0.4759846301633045,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004983685754595159,
      "loss": 5.2455,
      "mean_token_accuracy": 0.1953454926609993,
      "num_tokens": 11359817.0,
      "step": 4955
    },
    {
      "entropy": 5.492042064666748,
      "epoch": 0.47646493756003844,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004983644518494578,
      "loss": 5.3042,
      "mean_token_accuracy": 0.1945435732603073,
      "num_tokens": 11371120.0,
      "step": 4960
    },
    {
      "entropy": 5.423617649078369,
      "epoch": 0.4769452449567723,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004983603230535403,
      "loss": 5.2801,
      "mean_token_accuracy": 0.19471257477998732,
      "num_tokens": 11382071.0,
      "step": 4965
    },
    {
      "entropy": 5.477252626419068,
      "epoch": 0.47742555235350626,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004983561890718594,
      "loss": 5.2822,
      "mean_token_accuracy": 0.19555846601724625,
      "num_tokens": 11392938.0,
      "step": 4970
    },
    {
      "entropy": 5.416876697540284,
      "epoch": 0.47790585975024014,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000498352049904511,
      "loss": 5.2883,
      "mean_token_accuracy": 0.20221930146217346,
      "num_tokens": 11405416.0,
      "step": 4975
    },
    {
      "entropy": 5.47752423286438,
      "epoch": 0.4783861671469741,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004983479055515914,
      "loss": 5.2636,
      "mean_token_accuracy": 0.20490847080945968,
      "num_tokens": 11417008.0,
      "step": 4980
    },
    {
      "entropy": 5.486589956283569,
      "epoch": 0.47886647454370795,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004983437560131964,
      "loss": 5.2787,
      "mean_token_accuracy": 0.20015133023262024,
      "num_tokens": 11428154.0,
      "step": 4985
    },
    {
      "entropy": 5.464849948883057,
      "epoch": 0.4793467819404419,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004983396012894228,
      "loss": 5.2545,
      "mean_token_accuracy": 0.19617437869310378,
      "num_tokens": 11438559.0,
      "step": 4990
    },
    {
      "entropy": 5.512053728103638,
      "epoch": 0.47982708933717577,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004983354413803666,
      "loss": 5.3521,
      "mean_token_accuracy": 0.19373709857463836,
      "num_tokens": 11449501.0,
      "step": 4995
    },
    {
      "entropy": 5.457212734222412,
      "epoch": 0.4803073967339097,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004983312762861248,
      "loss": 5.3397,
      "mean_token_accuracy": 0.19626328051090242,
      "num_tokens": 11461841.0,
      "step": 5000
    },
    {
      "entropy": 5.437505531311035,
      "epoch": 0.4807877041306436,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004983271060067939,
      "loss": 5.2489,
      "mean_token_accuracy": 0.19419465363025665,
      "num_tokens": 11472876.0,
      "step": 5005
    },
    {
      "entropy": 5.393140459060669,
      "epoch": 0.4812680115273775,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004983229305424707,
      "loss": 5.3018,
      "mean_token_accuracy": 0.19776208847761154,
      "num_tokens": 11484541.0,
      "step": 5010
    },
    {
      "entropy": 5.28839340209961,
      "epoch": 0.4817483189241114,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004983187498932522,
      "loss": 5.1502,
      "mean_token_accuracy": 0.2065964624285698,
      "num_tokens": 11496071.0,
      "step": 5015
    },
    {
      "entropy": 5.437427282333374,
      "epoch": 0.48222862632084534,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004983145640592354,
      "loss": 5.218,
      "mean_token_accuracy": 0.20344727039337157,
      "num_tokens": 11507574.0,
      "step": 5020
    },
    {
      "entropy": 5.407801675796509,
      "epoch": 0.4827089337175792,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004983103730405176,
      "loss": 5.2709,
      "mean_token_accuracy": 0.1997988373041153,
      "num_tokens": 11519738.0,
      "step": 5025
    },
    {
      "entropy": 5.411714172363281,
      "epoch": 0.48318924111431316,
      "grad_norm": 0.953125,
      "learning_rate": 0.000498306176837196,
      "loss": 5.27,
      "mean_token_accuracy": 0.2023049235343933,
      "num_tokens": 11530669.0,
      "step": 5030
    },
    {
      "entropy": 5.318043279647827,
      "epoch": 0.48366954851104704,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004983019754493681,
      "loss": 5.1812,
      "mean_token_accuracy": 0.20663500875234603,
      "num_tokens": 11541680.0,
      "step": 5035
    },
    {
      "entropy": 5.36067156791687,
      "epoch": 0.484149855907781,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004982977688771314,
      "loss": 5.2454,
      "mean_token_accuracy": 0.19821488112211227,
      "num_tokens": 11553204.0,
      "step": 5040
    },
    {
      "entropy": 5.446121311187744,
      "epoch": 0.4846301633045149,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004982935571205835,
      "loss": 5.2645,
      "mean_token_accuracy": 0.1924535945057869,
      "num_tokens": 11564841.0,
      "step": 5045
    },
    {
      "entropy": 5.396951055526733,
      "epoch": 0.4851104707012488,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004982893401798223,
      "loss": 5.2627,
      "mean_token_accuracy": 0.19767125099897384,
      "num_tokens": 11576723.0,
      "step": 5050
    },
    {
      "entropy": 5.46646819114685,
      "epoch": 0.48559077809798273,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004982851180549456,
      "loss": 5.2362,
      "mean_token_accuracy": 0.20364539921283722,
      "num_tokens": 11587727.0,
      "step": 5055
    },
    {
      "entropy": 5.388969612121582,
      "epoch": 0.4860710854947166,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004982808907460515,
      "loss": 5.2251,
      "mean_token_accuracy": 0.19947829693555832,
      "num_tokens": 11598508.0,
      "step": 5060
    },
    {
      "entropy": 5.335644006729126,
      "epoch": 0.48655139289145055,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004982766582532382,
      "loss": 5.1567,
      "mean_token_accuracy": 0.2114589810371399,
      "num_tokens": 11609684.0,
      "step": 5065
    },
    {
      "entropy": 5.422902297973633,
      "epoch": 0.48703170028818443,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004982724205766038,
      "loss": 5.2904,
      "mean_token_accuracy": 0.1893261805176735,
      "num_tokens": 11621642.0,
      "step": 5070
    },
    {
      "entropy": 5.429874610900879,
      "epoch": 0.48751200768491837,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004982681777162468,
      "loss": 5.2261,
      "mean_token_accuracy": 0.19831710159778596,
      "num_tokens": 11632778.0,
      "step": 5075
    },
    {
      "entropy": 5.369401168823242,
      "epoch": 0.48799231508165225,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004982639296722657,
      "loss": 5.2064,
      "mean_token_accuracy": 0.20496241748332977,
      "num_tokens": 11643581.0,
      "step": 5080
    },
    {
      "entropy": 5.504480600357056,
      "epoch": 0.4884726224783862,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004982596764447591,
      "loss": 5.3841,
      "mean_token_accuracy": 0.18862680196762086,
      "num_tokens": 11656126.0,
      "step": 5085
    },
    {
      "entropy": 5.528008937835693,
      "epoch": 0.48895292987512007,
      "grad_norm": 0.875,
      "learning_rate": 0.0004982554180338258,
      "loss": 5.3009,
      "mean_token_accuracy": 0.20304699689149858,
      "num_tokens": 11667765.0,
      "step": 5090
    },
    {
      "entropy": 5.426607084274292,
      "epoch": 0.489433237271854,
      "grad_norm": 1.0,
      "learning_rate": 0.0004982511544395646,
      "loss": 5.3048,
      "mean_token_accuracy": 0.19115327447652816,
      "num_tokens": 11679268.0,
      "step": 5095
    },
    {
      "entropy": 5.386170196533203,
      "epoch": 0.4899135446685879,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004982468856620745,
      "loss": 5.2334,
      "mean_token_accuracy": 0.1988862708210945,
      "num_tokens": 11690049.0,
      "step": 5100
    },
    {
      "entropy": 5.4239543914794925,
      "epoch": 0.4903938520653218,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004982426117014545,
      "loss": 5.2161,
      "mean_token_accuracy": 0.20996261686086654,
      "num_tokens": 11701782.0,
      "step": 5105
    },
    {
      "entropy": 5.367910099029541,
      "epoch": 0.4908741594620557,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004982383325578041,
      "loss": 5.2343,
      "mean_token_accuracy": 0.21035679429769516,
      "num_tokens": 11712488.0,
      "step": 5110
    },
    {
      "entropy": 5.433872318267822,
      "epoch": 0.49135446685878964,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004982340482312226,
      "loss": 5.261,
      "mean_token_accuracy": 0.2035107880830765,
      "num_tokens": 11724036.0,
      "step": 5115
    },
    {
      "entropy": 5.382973289489746,
      "epoch": 0.4918347742555235,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004982297587218092,
      "loss": 5.2333,
      "mean_token_accuracy": 0.20306237488985063,
      "num_tokens": 11736461.0,
      "step": 5120
    },
    {
      "entropy": 5.289829730987549,
      "epoch": 0.49231508165225746,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004982254640296637,
      "loss": 5.1882,
      "mean_token_accuracy": 0.20668814331293106,
      "num_tokens": 11747549.0,
      "step": 5125
    },
    {
      "entropy": 5.506736087799072,
      "epoch": 0.49279538904899134,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004982211641548857,
      "loss": 5.3402,
      "mean_token_accuracy": 0.19071976840496063,
      "num_tokens": 11759012.0,
      "step": 5130
    },
    {
      "entropy": 5.494192266464234,
      "epoch": 0.4932756964457253,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004982168590975752,
      "loss": 5.2512,
      "mean_token_accuracy": 0.20154621452093124,
      "num_tokens": 11770535.0,
      "step": 5135
    },
    {
      "entropy": 5.384505939483643,
      "epoch": 0.49375600384245916,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004982125488578321,
      "loss": 5.2968,
      "mean_token_accuracy": 0.2019633173942566,
      "num_tokens": 11783708.0,
      "step": 5140
    },
    {
      "entropy": 5.41214337348938,
      "epoch": 0.4942363112391931,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004982082334357563,
      "loss": 5.2518,
      "mean_token_accuracy": 0.2000403508543968,
      "num_tokens": 11794565.0,
      "step": 5145
    },
    {
      "entropy": 5.483220005035401,
      "epoch": 0.494716618635927,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004982039128314481,
      "loss": 5.2635,
      "mean_token_accuracy": 0.19790053218603135,
      "num_tokens": 11805631.0,
      "step": 5150
    },
    {
      "entropy": 5.455584239959717,
      "epoch": 0.4951969260326609,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004981995870450079,
      "loss": 5.2846,
      "mean_token_accuracy": 0.1977648615837097,
      "num_tokens": 11817041.0,
      "step": 5155
    },
    {
      "entropy": 5.433424806594848,
      "epoch": 0.4956772334293948,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004981952560765361,
      "loss": 5.2311,
      "mean_token_accuracy": 0.2010979473590851,
      "num_tokens": 11829090.0,
      "step": 5160
    },
    {
      "entropy": 5.551053524017334,
      "epoch": 0.49615754082612873,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004981909199261331,
      "loss": 5.3545,
      "mean_token_accuracy": 0.19073701798915862,
      "num_tokens": 11839980.0,
      "step": 5165
    },
    {
      "entropy": 5.461454248428344,
      "epoch": 0.4966378482228626,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004981865785938998,
      "loss": 5.2544,
      "mean_token_accuracy": 0.2027146428823471,
      "num_tokens": 11851217.0,
      "step": 5170
    },
    {
      "entropy": 5.42721152305603,
      "epoch": 0.49711815561959655,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004981822320799367,
      "loss": 5.2859,
      "mean_token_accuracy": 0.20002590417861937,
      "num_tokens": 11862506.0,
      "step": 5175
    },
    {
      "entropy": 5.434733438491821,
      "epoch": 0.49759846301633043,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004981778803843449,
      "loss": 5.3059,
      "mean_token_accuracy": 0.1936892732977867,
      "num_tokens": 11874844.0,
      "step": 5180
    },
    {
      "entropy": 5.353609848022461,
      "epoch": 0.49807877041306436,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004981735235072256,
      "loss": 5.2012,
      "mean_token_accuracy": 0.21021807044744492,
      "num_tokens": 11885231.0,
      "step": 5185
    },
    {
      "entropy": 5.419971656799317,
      "epoch": 0.49855907780979825,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004981691614486796,
      "loss": 5.2696,
      "mean_token_accuracy": 0.19656101614236832,
      "num_tokens": 11897222.0,
      "step": 5190
    },
    {
      "entropy": 5.41405143737793,
      "epoch": 0.4990393852065322,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004981647942088084,
      "loss": 5.2097,
      "mean_token_accuracy": 0.20165806859731675,
      "num_tokens": 11909493.0,
      "step": 5195
    },
    {
      "entropy": 5.442890644073486,
      "epoch": 0.49951969260326606,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004981604217877135,
      "loss": 5.2297,
      "mean_token_accuracy": 0.1997319132089615,
      "num_tokens": 11920833.0,
      "step": 5200
    },
    {
      "entropy": 5.326638078689575,
      "epoch": 0.5,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000498156044185496,
      "loss": 5.1187,
      "mean_token_accuracy": 0.21121429800987243,
      "num_tokens": 11932910.0,
      "step": 5205
    },
    {
      "entropy": 5.45737657546997,
      "epoch": 0.5004803073967339,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004981516614022579,
      "loss": 5.2552,
      "mean_token_accuracy": 0.21273988783359526,
      "num_tokens": 11944536.0,
      "step": 5210
    },
    {
      "entropy": 5.361939764022827,
      "epoch": 0.5009606147934679,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004981472734381008,
      "loss": 5.2433,
      "mean_token_accuracy": 0.20870155096054077,
      "num_tokens": 11955724.0,
      "step": 5215
    },
    {
      "entropy": 5.352823257446289,
      "epoch": 0.5014409221902018,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004981428802931267,
      "loss": 5.2209,
      "mean_token_accuracy": 0.20748957693576814,
      "num_tokens": 11965921.0,
      "step": 5220
    },
    {
      "entropy": 5.401951789855957,
      "epoch": 0.5019212295869356,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004981384819674375,
      "loss": 5.1873,
      "mean_token_accuracy": 0.20220513939857482,
      "num_tokens": 11976356.0,
      "step": 5225
    },
    {
      "entropy": 5.387565755844117,
      "epoch": 0.5024015369836695,
      "grad_norm": 0.875,
      "learning_rate": 0.0004981340784611354,
      "loss": 5.2356,
      "mean_token_accuracy": 0.19975537657737732,
      "num_tokens": 11987959.0,
      "step": 5230
    },
    {
      "entropy": 5.358628940582276,
      "epoch": 0.5028818443804035,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004981296697743224,
      "loss": 5.2545,
      "mean_token_accuracy": 0.1979578971862793,
      "num_tokens": 12000089.0,
      "step": 5235
    },
    {
      "entropy": 5.416265630722046,
      "epoch": 0.5033621517771374,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004981252559071012,
      "loss": 5.3043,
      "mean_token_accuracy": 0.1985698774456978,
      "num_tokens": 12011286.0,
      "step": 5240
    },
    {
      "entropy": 5.387771987915039,
      "epoch": 0.5038424591738713,
      "grad_norm": 0.875,
      "learning_rate": 0.0004981208368595739,
      "loss": 5.1867,
      "mean_token_accuracy": 0.2020393803715706,
      "num_tokens": 12022809.0,
      "step": 5245
    },
    {
      "entropy": 5.369564867019653,
      "epoch": 0.5043227665706052,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004981164126318435,
      "loss": 5.1773,
      "mean_token_accuracy": 0.20869052708148955,
      "num_tokens": 12034311.0,
      "step": 5250
    },
    {
      "entropy": 5.341700887680053,
      "epoch": 0.5048030739673391,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004981119832240124,
      "loss": 5.1814,
      "mean_token_accuracy": 0.20985617637634277,
      "num_tokens": 12046405.0,
      "step": 5255
    },
    {
      "entropy": 5.383070611953736,
      "epoch": 0.505283381364073,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004981075486361837,
      "loss": 5.2498,
      "mean_token_accuracy": 0.20150127559900283,
      "num_tokens": 12059022.0,
      "step": 5260
    },
    {
      "entropy": 5.412682151794433,
      "epoch": 0.5057636887608069,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004981031088684601,
      "loss": 5.2446,
      "mean_token_accuracy": 0.20172821581363679,
      "num_tokens": 12070666.0,
      "step": 5265
    },
    {
      "entropy": 5.348124551773071,
      "epoch": 0.5062439961575408,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004980986639209448,
      "loss": 5.2353,
      "mean_token_accuracy": 0.20279886573553085,
      "num_tokens": 12082702.0,
      "step": 5270
    },
    {
      "entropy": 5.403745651245117,
      "epoch": 0.5067243035542748,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000498094213793741,
      "loss": 5.2519,
      "mean_token_accuracy": 0.2026989296078682,
      "num_tokens": 12094034.0,
      "step": 5275
    },
    {
      "entropy": 5.477759599685669,
      "epoch": 0.5072046109510087,
      "grad_norm": 1.0,
      "learning_rate": 0.000498089758486952,
      "loss": 5.2416,
      "mean_token_accuracy": 0.19031255692243576,
      "num_tokens": 12104597.0,
      "step": 5280
    },
    {
      "entropy": 5.4051412582397464,
      "epoch": 0.5076849183477425,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004980852980006812,
      "loss": 5.2428,
      "mean_token_accuracy": 0.1945531353354454,
      "num_tokens": 12116577.0,
      "step": 5285
    },
    {
      "entropy": 5.400573635101319,
      "epoch": 0.5081652257444764,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004980808323350323,
      "loss": 5.1891,
      "mean_token_accuracy": 0.2012152597308159,
      "num_tokens": 12128005.0,
      "step": 5290
    },
    {
      "entropy": 5.446866655349732,
      "epoch": 0.5086455331412104,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004980763614901089,
      "loss": 5.2932,
      "mean_token_accuracy": 0.19632818996906282,
      "num_tokens": 12138591.0,
      "step": 5295
    },
    {
      "entropy": 5.370537042617798,
      "epoch": 0.5091258405379443,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004980718854660146,
      "loss": 5.2708,
      "mean_token_accuracy": 0.19851180166006088,
      "num_tokens": 12150010.0,
      "step": 5300
    },
    {
      "entropy": 5.425984525680542,
      "epoch": 0.5096061479346782,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004980674042628537,
      "loss": 5.2221,
      "mean_token_accuracy": 0.1974261075258255,
      "num_tokens": 12161268.0,
      "step": 5305
    },
    {
      "entropy": 5.356367588043213,
      "epoch": 0.5100864553314121,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00049806291788073,
      "loss": 5.191,
      "mean_token_accuracy": 0.2062757596373558,
      "num_tokens": 12172647.0,
      "step": 5310
    },
    {
      "entropy": 5.359721612930298,
      "epoch": 0.510566762728146,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004980584263197477,
      "loss": 5.2712,
      "mean_token_accuracy": 0.2021322503685951,
      "num_tokens": 12183521.0,
      "step": 5315
    },
    {
      "entropy": 5.5467798709869385,
      "epoch": 0.5110470701248799,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004980539295800111,
      "loss": 5.3181,
      "mean_token_accuracy": 0.19267420917749406,
      "num_tokens": 12193616.0,
      "step": 5320
    },
    {
      "entropy": 5.384074115753174,
      "epoch": 0.5115273775216138,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004980494276616246,
      "loss": 5.255,
      "mean_token_accuracy": 0.20038930028676988,
      "num_tokens": 12204817.0,
      "step": 5325
    },
    {
      "entropy": 5.4296790599823,
      "epoch": 0.5120076849183477,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004980449205646926,
      "loss": 5.2938,
      "mean_token_accuracy": 0.19566230475902557,
      "num_tokens": 12217281.0,
      "step": 5330
    },
    {
      "entropy": 5.4535449028015135,
      "epoch": 0.5124879923150817,
      "grad_norm": 0.9375,
      "learning_rate": 0.00049804040828932,
      "loss": 5.201,
      "mean_token_accuracy": 0.20819566696882247,
      "num_tokens": 12228705.0,
      "step": 5335
    },
    {
      "entropy": 5.407874345779419,
      "epoch": 0.5129682997118156,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004980358908356113,
      "loss": 5.3027,
      "mean_token_accuracy": 0.19639102071523667,
      "num_tokens": 12240010.0,
      "step": 5340
    },
    {
      "entropy": 5.46102237701416,
      "epoch": 0.5134486071085494,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004980313682036717,
      "loss": 5.3508,
      "mean_token_accuracy": 0.19489552974700927,
      "num_tokens": 12251384.0,
      "step": 5345
    },
    {
      "entropy": 5.442351341247559,
      "epoch": 0.5139289145052833,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004980268403936058,
      "loss": 5.1934,
      "mean_token_accuracy": 0.20403928607702254,
      "num_tokens": 12262548.0,
      "step": 5350
    },
    {
      "entropy": 5.337255620956421,
      "epoch": 0.5144092219020173,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004980223074055189,
      "loss": 5.1488,
      "mean_token_accuracy": 0.2125844120979309,
      "num_tokens": 12274022.0,
      "step": 5355
    },
    {
      "entropy": 5.32992377281189,
      "epoch": 0.5148895292987512,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004980177692395164,
      "loss": 5.1394,
      "mean_token_accuracy": 0.20436035841703415,
      "num_tokens": 12285000.0,
      "step": 5360
    },
    {
      "entropy": 5.344844722747803,
      "epoch": 0.5153698366954851,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004980132258957035,
      "loss": 5.1909,
      "mean_token_accuracy": 0.2062470480799675,
      "num_tokens": 12297472.0,
      "step": 5365
    },
    {
      "entropy": 5.438365650177002,
      "epoch": 0.515850144092219,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004980086773741856,
      "loss": 5.2418,
      "mean_token_accuracy": 0.20333612710237503,
      "num_tokens": 12309310.0,
      "step": 5370
    },
    {
      "entropy": 5.4867119789123535,
      "epoch": 0.516330451488953,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004980041236750685,
      "loss": 5.326,
      "mean_token_accuracy": 0.19241970926523208,
      "num_tokens": 12321897.0,
      "step": 5375
    },
    {
      "entropy": 5.317311429977417,
      "epoch": 0.5168107588856868,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004979995647984577,
      "loss": 5.1593,
      "mean_token_accuracy": 0.21523418128490449,
      "num_tokens": 12333484.0,
      "step": 5380
    },
    {
      "entropy": 5.347341156005859,
      "epoch": 0.5172910662824207,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004979950007444593,
      "loss": 5.2072,
      "mean_token_accuracy": 0.201828633248806,
      "num_tokens": 12345154.0,
      "step": 5385
    },
    {
      "entropy": 5.499171924591065,
      "epoch": 0.5177713736791547,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004979904315131792,
      "loss": 5.2714,
      "mean_token_accuracy": 0.20158810019493104,
      "num_tokens": 12354934.0,
      "step": 5390
    },
    {
      "entropy": 5.52605619430542,
      "epoch": 0.5182516810758886,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004979858571047233,
      "loss": 5.4253,
      "mean_token_accuracy": 0.18736765235662461,
      "num_tokens": 12367729.0,
      "step": 5395
    },
    {
      "entropy": 5.423756122589111,
      "epoch": 0.5187319884726225,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004979812775191979,
      "loss": 5.213,
      "mean_token_accuracy": 0.2047743871808052,
      "num_tokens": 12380366.0,
      "step": 5400
    },
    {
      "entropy": 5.342135381698609,
      "epoch": 0.5192122958693564,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004979766927567094,
      "loss": 5.2534,
      "mean_token_accuracy": 0.20312506705522537,
      "num_tokens": 12392338.0,
      "step": 5405
    },
    {
      "entropy": 5.413405084609986,
      "epoch": 0.5196926032660903,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004979721028173643,
      "loss": 5.2124,
      "mean_token_accuracy": 0.20402789860963821,
      "num_tokens": 12403280.0,
      "step": 5410
    },
    {
      "entropy": 5.391100168228149,
      "epoch": 0.5201729106628242,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000497967507701269,
      "loss": 5.1922,
      "mean_token_accuracy": 0.20790980011224747,
      "num_tokens": 12414279.0,
      "step": 5415
    },
    {
      "entropy": 5.280627202987671,
      "epoch": 0.5206532180595581,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004979629074085303,
      "loss": 5.1181,
      "mean_token_accuracy": 0.20512137711048126,
      "num_tokens": 12425486.0,
      "step": 5420
    },
    {
      "entropy": 5.3557658195495605,
      "epoch": 0.521133525456292,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004979583019392548,
      "loss": 5.2372,
      "mean_token_accuracy": 0.1978609561920166,
      "num_tokens": 12437295.0,
      "step": 5425
    },
    {
      "entropy": 5.4152984619140625,
      "epoch": 0.521613832853026,
      "grad_norm": 1.0,
      "learning_rate": 0.0004979536912935497,
      "loss": 5.191,
      "mean_token_accuracy": 0.20112977772951127,
      "num_tokens": 12448684.0,
      "step": 5430
    },
    {
      "entropy": 5.327011346817017,
      "epoch": 0.5220941402497599,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000497949075471522,
      "loss": 5.1345,
      "mean_token_accuracy": 0.21084612160921096,
      "num_tokens": 12461209.0,
      "step": 5435
    },
    {
      "entropy": 5.3874060153961185,
      "epoch": 0.5225744476464937,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004979444544732786,
      "loss": 5.2244,
      "mean_token_accuracy": 0.20195067375898362,
      "num_tokens": 12472863.0,
      "step": 5440
    },
    {
      "entropy": 5.43286771774292,
      "epoch": 0.5230547550432276,
      "grad_norm": 0.90625,
      "learning_rate": 0.000497939828298927,
      "loss": 5.2461,
      "mean_token_accuracy": 0.2011614888906479,
      "num_tokens": 12483510.0,
      "step": 5445
    },
    {
      "entropy": 5.438796520233154,
      "epoch": 0.5235350624399616,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004979351969485747,
      "loss": 5.258,
      "mean_token_accuracy": 0.2009929746389389,
      "num_tokens": 12495532.0,
      "step": 5450
    },
    {
      "entropy": 5.308896827697754,
      "epoch": 0.5240153698366955,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004979305604223291,
      "loss": 5.1792,
      "mean_token_accuracy": 0.2020716667175293,
      "num_tokens": 12508413.0,
      "step": 5455
    },
    {
      "entropy": 5.4415699481964115,
      "epoch": 0.5244956772334294,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004979259187202978,
      "loss": 5.279,
      "mean_token_accuracy": 0.19909130334854125,
      "num_tokens": 12520705.0,
      "step": 5460
    },
    {
      "entropy": 5.3851653099060055,
      "epoch": 0.5249759846301633,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004979212718425887,
      "loss": 5.2203,
      "mean_token_accuracy": 0.20675526857376098,
      "num_tokens": 12531560.0,
      "step": 5465
    },
    {
      "entropy": 5.3778948307037355,
      "epoch": 0.5254562920268973,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004979166197893096,
      "loss": 5.231,
      "mean_token_accuracy": 0.1981738477945328,
      "num_tokens": 12542539.0,
      "step": 5470
    },
    {
      "entropy": 5.45626540184021,
      "epoch": 0.5259365994236311,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004979119625605683,
      "loss": 5.2839,
      "mean_token_accuracy": 0.196984925866127,
      "num_tokens": 12555228.0,
      "step": 5475
    },
    {
      "entropy": 5.4724688053131105,
      "epoch": 0.526416906820365,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004979073001564734,
      "loss": 5.2011,
      "mean_token_accuracy": 0.20548761636018753,
      "num_tokens": 12565715.0,
      "step": 5480
    },
    {
      "entropy": 5.348662614822388,
      "epoch": 0.5268972142170989,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004979026325771328,
      "loss": 5.1535,
      "mean_token_accuracy": 0.20471411794424058,
      "num_tokens": 12576896.0,
      "step": 5485
    },
    {
      "entropy": 5.398907089233399,
      "epoch": 0.5273775216138329,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004978979598226549,
      "loss": 5.2109,
      "mean_token_accuracy": 0.20377436876296998,
      "num_tokens": 12588117.0,
      "step": 5490
    },
    {
      "entropy": 5.401965808868408,
      "epoch": 0.5278578290105668,
      "grad_norm": 1.0,
      "learning_rate": 0.0004978932818931483,
      "loss": 5.2141,
      "mean_token_accuracy": 0.20156438052654266,
      "num_tokens": 12599987.0,
      "step": 5495
    },
    {
      "entropy": 5.391234588623047,
      "epoch": 0.5283381364073007,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004978885987887216,
      "loss": 5.3052,
      "mean_token_accuracy": 0.19640427678823472,
      "num_tokens": 12611453.0,
      "step": 5500
    },
    {
      "entropy": 5.429474925994873,
      "epoch": 0.5288184438040345,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004978839105094833,
      "loss": 5.251,
      "mean_token_accuracy": 0.19685932248830795,
      "num_tokens": 12623068.0,
      "step": 5505
    },
    {
      "entropy": 5.4846560001373295,
      "epoch": 0.5292987512007685,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004978792170555426,
      "loss": 5.2881,
      "mean_token_accuracy": 0.2029639035463333,
      "num_tokens": 12634952.0,
      "step": 5510
    },
    {
      "entropy": 5.382822465896607,
      "epoch": 0.5297790585975024,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004978745184270083,
      "loss": 5.2231,
      "mean_token_accuracy": 0.19447317570447922,
      "num_tokens": 12647027.0,
      "step": 5515
    },
    {
      "entropy": 5.291022396087646,
      "epoch": 0.5302593659942363,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004978698146239893,
      "loss": 5.0893,
      "mean_token_accuracy": 0.21481227427721022,
      "num_tokens": 12657249.0,
      "step": 5520
    },
    {
      "entropy": 5.328450345993042,
      "epoch": 0.5307396733909702,
      "grad_norm": 1.0,
      "learning_rate": 0.0004978651056465952,
      "loss": 5.1225,
      "mean_token_accuracy": 0.21356747448444366,
      "num_tokens": 12666714.0,
      "step": 5525
    },
    {
      "entropy": 5.382013273239136,
      "epoch": 0.5312199807877042,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000497860391494935,
      "loss": 5.1978,
      "mean_token_accuracy": 0.20151051133871078,
      "num_tokens": 12678627.0,
      "step": 5530
    },
    {
      "entropy": 5.422570323944091,
      "epoch": 0.531700288184438,
      "grad_norm": 1.0,
      "learning_rate": 0.0004978556721691183,
      "loss": 5.249,
      "mean_token_accuracy": 0.20071519315242767,
      "num_tokens": 12691009.0,
      "step": 5535
    },
    {
      "entropy": 5.40853099822998,
      "epoch": 0.5321805955811719,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004978509476692547,
      "loss": 5.2104,
      "mean_token_accuracy": 0.20311132520437242,
      "num_tokens": 12702503.0,
      "step": 5540
    },
    {
      "entropy": 5.388434028625488,
      "epoch": 0.5326609029779059,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004978462179954538,
      "loss": 5.2706,
      "mean_token_accuracy": 0.19816106259822847,
      "num_tokens": 12714018.0,
      "step": 5545
    },
    {
      "entropy": 5.449985551834106,
      "epoch": 0.5331412103746398,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004978414831478253,
      "loss": 5.3655,
      "mean_token_accuracy": 0.19622621387243272,
      "num_tokens": 12726726.0,
      "step": 5550
    },
    {
      "entropy": 5.391891956329346,
      "epoch": 0.5336215177713737,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004978367431264794,
      "loss": 5.1971,
      "mean_token_accuracy": 0.2061743766069412,
      "num_tokens": 12737790.0,
      "step": 5555
    },
    {
      "entropy": 5.332040452957154,
      "epoch": 0.5341018251681076,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004978319979315261,
      "loss": 5.0995,
      "mean_token_accuracy": 0.20796922445297242,
      "num_tokens": 12749473.0,
      "step": 5560
    },
    {
      "entropy": 5.365551233291626,
      "epoch": 0.5345821325648416,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004978272475630752,
      "loss": 5.2445,
      "mean_token_accuracy": 0.20265517830848695,
      "num_tokens": 12762220.0,
      "step": 5565
    },
    {
      "entropy": 5.270161485671997,
      "epoch": 0.5350624399615754,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004978224920212374,
      "loss": 5.0454,
      "mean_token_accuracy": 0.21360960006713867,
      "num_tokens": 12772788.0,
      "step": 5570
    },
    {
      "entropy": 5.274341678619384,
      "epoch": 0.5355427473583093,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004978177313061232,
      "loss": 5.1385,
      "mean_token_accuracy": 0.20616633892059327,
      "num_tokens": 12784552.0,
      "step": 5575
    },
    {
      "entropy": 5.378845739364624,
      "epoch": 0.5360230547550432,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004978129654178426,
      "loss": 5.1903,
      "mean_token_accuracy": 0.20797884613275527,
      "num_tokens": 12797618.0,
      "step": 5580
    },
    {
      "entropy": 5.3385977268219,
      "epoch": 0.5365033621517772,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004978081943565067,
      "loss": 5.1231,
      "mean_token_accuracy": 0.2072676420211792,
      "num_tokens": 12807789.0,
      "step": 5585
    },
    {
      "entropy": 5.347647190093994,
      "epoch": 0.5369836695485111,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004978034181222261,
      "loss": 5.2511,
      "mean_token_accuracy": 0.20461219698190689,
      "num_tokens": 12819020.0,
      "step": 5590
    },
    {
      "entropy": 5.411481857299805,
      "epoch": 0.537463976945245,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004977986367151119,
      "loss": 5.2165,
      "mean_token_accuracy": 0.20503491014242173,
      "num_tokens": 12829527.0,
      "step": 5595
    },
    {
      "entropy": 5.452748870849609,
      "epoch": 0.5379442843419788,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004977938501352747,
      "loss": 5.2228,
      "mean_token_accuracy": 0.19553297907114028,
      "num_tokens": 12840821.0,
      "step": 5600
    },
    {
      "entropy": 5.387438583374023,
      "epoch": 0.5384245917387128,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004977890583828259,
      "loss": 5.2097,
      "mean_token_accuracy": 0.20679532140493392,
      "num_tokens": 12853314.0,
      "step": 5605
    },
    {
      "entropy": 5.363187074661255,
      "epoch": 0.5389048991354467,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004977842614578768,
      "loss": 5.2159,
      "mean_token_accuracy": 0.1987797886133194,
      "num_tokens": 12863507.0,
      "step": 5610
    },
    {
      "entropy": 5.410052537918091,
      "epoch": 0.5393852065321806,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004977794593605386,
      "loss": 5.2156,
      "mean_token_accuracy": 0.20640329867601395,
      "num_tokens": 12874142.0,
      "step": 5615
    },
    {
      "entropy": 5.407776117324829,
      "epoch": 0.5398655139289145,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000497774652090923,
      "loss": 5.2235,
      "mean_token_accuracy": 0.21270929723978044,
      "num_tokens": 12884548.0,
      "step": 5620
    },
    {
      "entropy": 5.4175914287567135,
      "epoch": 0.5403458213256485,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004977698396491414,
      "loss": 5.23,
      "mean_token_accuracy": 0.20302337259054185,
      "num_tokens": 12895562.0,
      "step": 5625
    },
    {
      "entropy": 5.364365816116333,
      "epoch": 0.5408261287223823,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004977650220353055,
      "loss": 5.1981,
      "mean_token_accuracy": 0.20721670091152192,
      "num_tokens": 12905308.0,
      "step": 5630
    },
    {
      "entropy": 5.408176565170288,
      "epoch": 0.5413064361191162,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004977601992495274,
      "loss": 5.241,
      "mean_token_accuracy": 0.20085862874984742,
      "num_tokens": 12916731.0,
      "step": 5635
    },
    {
      "entropy": 5.474975299835205,
      "epoch": 0.5417867435158501,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004977553712919189,
      "loss": 5.3039,
      "mean_token_accuracy": 0.19791211187839508,
      "num_tokens": 12927609.0,
      "step": 5640
    },
    {
      "entropy": 5.467264604568482,
      "epoch": 0.5422670509125841,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004977505381625921,
      "loss": 5.2218,
      "mean_token_accuracy": 0.20000761449337007,
      "num_tokens": 12939445.0,
      "step": 5645
    },
    {
      "entropy": 5.331275463104248,
      "epoch": 0.542747358309318,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004977456998616593,
      "loss": 5.2125,
      "mean_token_accuracy": 0.2033090338110924,
      "num_tokens": 12951638.0,
      "step": 5650
    },
    {
      "entropy": 5.3639062404632565,
      "epoch": 0.5432276657060519,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004977408563892327,
      "loss": 5.1944,
      "mean_token_accuracy": 0.20639608949422836,
      "num_tokens": 12963157.0,
      "step": 5655
    },
    {
      "entropy": 5.379507303237915,
      "epoch": 0.5437079731027857,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004977360077454249,
      "loss": 5.1691,
      "mean_token_accuracy": 0.20129819214344025,
      "num_tokens": 12974187.0,
      "step": 5660
    },
    {
      "entropy": 5.422531843185425,
      "epoch": 0.5441882804995197,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004977311539303483,
      "loss": 5.2565,
      "mean_token_accuracy": 0.2019005075097084,
      "num_tokens": 12986612.0,
      "step": 5665
    },
    {
      "entropy": 5.432423257827759,
      "epoch": 0.5446685878962536,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004977262949441158,
      "loss": 5.2706,
      "mean_token_accuracy": 0.19311508238315583,
      "num_tokens": 12998343.0,
      "step": 5670
    },
    {
      "entropy": 5.389460563659668,
      "epoch": 0.5451488952929875,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004977214307868399,
      "loss": 5.2414,
      "mean_token_accuracy": 0.1979389950633049,
      "num_tokens": 13009272.0,
      "step": 5675
    },
    {
      "entropy": 5.421024799346924,
      "epoch": 0.5456292026897214,
      "grad_norm": 1.015625,
      "learning_rate": 0.000497716561458634,
      "loss": 5.2222,
      "mean_token_accuracy": 0.20285080522298812,
      "num_tokens": 13019961.0,
      "step": 5680
    },
    {
      "entropy": 5.334949350357055,
      "epoch": 0.5461095100864554,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004977116869596107,
      "loss": 5.1698,
      "mean_token_accuracy": 0.2091596841812134,
      "num_tokens": 13030827.0,
      "step": 5685
    },
    {
      "entropy": 5.475653886795044,
      "epoch": 0.5465898174831892,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004977068072898834,
      "loss": 5.3864,
      "mean_token_accuracy": 0.19146479815244674,
      "num_tokens": 13043702.0,
      "step": 5690
    },
    {
      "entropy": 5.294312477111816,
      "epoch": 0.5470701248799231,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004977019224495652,
      "loss": 5.0635,
      "mean_token_accuracy": 0.21373049765825272,
      "num_tokens": 13054474.0,
      "step": 5695
    },
    {
      "entropy": 5.431678915023804,
      "epoch": 0.547550432276657,
      "grad_norm": 1.0,
      "learning_rate": 0.0004976970324387698,
      "loss": 5.1994,
      "mean_token_accuracy": 0.19995237439870833,
      "num_tokens": 13064920.0,
      "step": 5700
    },
    {
      "entropy": 5.422818613052368,
      "epoch": 0.548030739673391,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004976921372576104,
      "loss": 5.3061,
      "mean_token_accuracy": 0.19882712215185167,
      "num_tokens": 13077214.0,
      "step": 5705
    },
    {
      "entropy": 5.377505445480347,
      "epoch": 0.5485110470701249,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004976872369062011,
      "loss": 5.1852,
      "mean_token_accuracy": 0.2016940712928772,
      "num_tokens": 13089472.0,
      "step": 5710
    },
    {
      "entropy": 5.425291728973389,
      "epoch": 0.5489913544668588,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004976823313846552,
      "loss": 5.2664,
      "mean_token_accuracy": 0.20560273826122283,
      "num_tokens": 13100369.0,
      "step": 5715
    },
    {
      "entropy": 5.4309999465942385,
      "epoch": 0.5494716618635928,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004976774206930869,
      "loss": 5.216,
      "mean_token_accuracy": 0.19250478893518447,
      "num_tokens": 13110813.0,
      "step": 5720
    },
    {
      "entropy": 5.3887077331542965,
      "epoch": 0.5499519692603266,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004976725048316101,
      "loss": 5.226,
      "mean_token_accuracy": 0.20413192212581635,
      "num_tokens": 13121946.0,
      "step": 5725
    },
    {
      "entropy": 5.340195560455323,
      "epoch": 0.5504322766570605,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004976675838003388,
      "loss": 5.1624,
      "mean_token_accuracy": 0.20904316008090973,
      "num_tokens": 13133286.0,
      "step": 5730
    },
    {
      "entropy": 5.305120754241943,
      "epoch": 0.5509125840537944,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004976626575993877,
      "loss": 5.1538,
      "mean_token_accuracy": 0.21018180698156358,
      "num_tokens": 13145023.0,
      "step": 5735
    },
    {
      "entropy": 5.3617431163787845,
      "epoch": 0.5513928914505284,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004976577262288705,
      "loss": 5.1049,
      "mean_token_accuracy": 0.21021644175052642,
      "num_tokens": 13156379.0,
      "step": 5740
    },
    {
      "entropy": 5.336035060882568,
      "epoch": 0.5518731988472623,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004976527896889023,
      "loss": 5.1607,
      "mean_token_accuracy": 0.21135113686323165,
      "num_tokens": 13166313.0,
      "step": 5745
    },
    {
      "entropy": 5.434840059280395,
      "epoch": 0.5523535062439962,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004976478479795974,
      "loss": 5.2277,
      "mean_token_accuracy": 0.19773912727832793,
      "num_tokens": 13177968.0,
      "step": 5750
    },
    {
      "entropy": 5.449684190750122,
      "epoch": 0.55283381364073,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004976429011010706,
      "loss": 5.2258,
      "mean_token_accuracy": 0.2007650300860405,
      "num_tokens": 13189464.0,
      "step": 5755
    },
    {
      "entropy": 5.392709541320801,
      "epoch": 0.553314121037464,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004976379490534366,
      "loss": 5.2468,
      "mean_token_accuracy": 0.19604432582855225,
      "num_tokens": 13201716.0,
      "step": 5760
    },
    {
      "entropy": 5.307784986495972,
      "epoch": 0.5537944284341979,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004976329918368107,
      "loss": 5.0766,
      "mean_token_accuracy": 0.20370612442493438,
      "num_tokens": 13212976.0,
      "step": 5765
    },
    {
      "entropy": 5.254825067520142,
      "epoch": 0.5542747358309318,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004976280294513079,
      "loss": 5.1094,
      "mean_token_accuracy": 0.2174687936902046,
      "num_tokens": 13225265.0,
      "step": 5770
    },
    {
      "entropy": 5.518316793441772,
      "epoch": 0.5547550432276657,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004976230618970431,
      "loss": 5.3966,
      "mean_token_accuracy": 0.1890160024166107,
      "num_tokens": 13236443.0,
      "step": 5775
    },
    {
      "entropy": 5.4118640422821045,
      "epoch": 0.5552353506243997,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000497618089174132,
      "loss": 5.2616,
      "mean_token_accuracy": 0.20688981860876082,
      "num_tokens": 13248681.0,
      "step": 5780
    },
    {
      "entropy": 5.411721515655517,
      "epoch": 0.5557156580211335,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004976131112826898,
      "loss": 5.2184,
      "mean_token_accuracy": 0.20831214040517806,
      "num_tokens": 13258719.0,
      "step": 5785
    },
    {
      "entropy": 5.352180671691895,
      "epoch": 0.5561959654178674,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004976081282228323,
      "loss": 5.1901,
      "mean_token_accuracy": 0.20076449066400529,
      "num_tokens": 13270544.0,
      "step": 5790
    },
    {
      "entropy": 5.399480199813842,
      "epoch": 0.5566762728146013,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000497603139994675,
      "loss": 5.1475,
      "mean_token_accuracy": 0.20780245810747147,
      "num_tokens": 13280759.0,
      "step": 5795
    },
    {
      "entropy": 5.356031084060669,
      "epoch": 0.5571565802113353,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004975981465983338,
      "loss": 5.2436,
      "mean_token_accuracy": 0.20736255645751953,
      "num_tokens": 13292496.0,
      "step": 5800
    },
    {
      "entropy": 5.4005491733551025,
      "epoch": 0.5576368876080692,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004975931480339246,
      "loss": 5.2365,
      "mean_token_accuracy": 0.20336752086877824,
      "num_tokens": 13303003.0,
      "step": 5805
    },
    {
      "entropy": 5.409140586853027,
      "epoch": 0.5581171950048031,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004975881443015635,
      "loss": 5.212,
      "mean_token_accuracy": 0.19998831003904344,
      "num_tokens": 13315036.0,
      "step": 5810
    },
    {
      "entropy": 5.373931360244751,
      "epoch": 0.5585975024015369,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004975831354013667,
      "loss": 5.1846,
      "mean_token_accuracy": 0.20424798727035523,
      "num_tokens": 13327529.0,
      "step": 5815
    },
    {
      "entropy": 5.3569361686706545,
      "epoch": 0.5590778097982709,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004975781213334503,
      "loss": 5.1841,
      "mean_token_accuracy": 0.20125899463891983,
      "num_tokens": 13339494.0,
      "step": 5820
    },
    {
      "entropy": 5.355644989013672,
      "epoch": 0.5595581171950048,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004975731020979309,
      "loss": 5.1847,
      "mean_token_accuracy": 0.200406713783741,
      "num_tokens": 13351184.0,
      "step": 5825
    },
    {
      "entropy": 5.341763210296631,
      "epoch": 0.5600384245917387,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004975680776949249,
      "loss": 5.1111,
      "mean_token_accuracy": 0.20737645626068116,
      "num_tokens": 13362402.0,
      "step": 5830
    },
    {
      "entropy": 5.362391948699951,
      "epoch": 0.5605187319884726,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004975630481245492,
      "loss": 5.289,
      "mean_token_accuracy": 0.2005311831831932,
      "num_tokens": 13375257.0,
      "step": 5835
    },
    {
      "entropy": 5.396390819549561,
      "epoch": 0.5609990393852066,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004975580133869202,
      "loss": 5.2084,
      "mean_token_accuracy": 0.2017640694975853,
      "num_tokens": 13385917.0,
      "step": 5840
    },
    {
      "entropy": 5.461558437347412,
      "epoch": 0.5614793467819404,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004975529734821552,
      "loss": 5.2318,
      "mean_token_accuracy": 0.19770166873931885,
      "num_tokens": 13397962.0,
      "step": 5845
    },
    {
      "entropy": 5.353429841995239,
      "epoch": 0.5619596541786743,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004975479284103708,
      "loss": 5.1282,
      "mean_token_accuracy": 0.2092791184782982,
      "num_tokens": 13408654.0,
      "step": 5850
    },
    {
      "entropy": 5.339301633834839,
      "epoch": 0.5624399615754082,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004975428781716845,
      "loss": 5.227,
      "mean_token_accuracy": 0.19910691678524017,
      "num_tokens": 13419790.0,
      "step": 5855
    },
    {
      "entropy": 5.335329055786133,
      "epoch": 0.5629202689721422,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004975378227662134,
      "loss": 5.1426,
      "mean_token_accuracy": 0.20664088279008866,
      "num_tokens": 13430305.0,
      "step": 5860
    },
    {
      "entropy": 5.412826681137085,
      "epoch": 0.5634005763688761,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004975327621940746,
      "loss": 5.245,
      "mean_token_accuracy": 0.2021554946899414,
      "num_tokens": 13441730.0,
      "step": 5865
    },
    {
      "entropy": 5.428021860122681,
      "epoch": 0.56388088376561,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004975276964553861,
      "loss": 5.2463,
      "mean_token_accuracy": 0.20189877301454545,
      "num_tokens": 13452003.0,
      "step": 5870
    },
    {
      "entropy": 5.372456645965576,
      "epoch": 0.5643611911623438,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004975226255502651,
      "loss": 5.1215,
      "mean_token_accuracy": 0.21266588419675828,
      "num_tokens": 13463953.0,
      "step": 5875
    },
    {
      "entropy": 5.424715328216553,
      "epoch": 0.5648414985590778,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004975175494788297,
      "loss": 5.2641,
      "mean_token_accuracy": 0.20160273611545562,
      "num_tokens": 13473820.0,
      "step": 5880
    },
    {
      "entropy": 5.441176080703736,
      "epoch": 0.5653218059558117,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004975124682411974,
      "loss": 5.286,
      "mean_token_accuracy": 0.20114496052265168,
      "num_tokens": 13484181.0,
      "step": 5885
    },
    {
      "entropy": 5.387451887130737,
      "epoch": 0.5658021133525456,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004975073818374863,
      "loss": 5.228,
      "mean_token_accuracy": 0.19973448663949966,
      "num_tokens": 13494480.0,
      "step": 5890
    },
    {
      "entropy": 5.299653434753418,
      "epoch": 0.5662824207492796,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004975022902678145,
      "loss": 5.1666,
      "mean_token_accuracy": 0.2074673131108284,
      "num_tokens": 13506026.0,
      "step": 5895
    },
    {
      "entropy": 5.443895101547241,
      "epoch": 0.5667627281460135,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004974971935323003,
      "loss": 5.2673,
      "mean_token_accuracy": 0.20036189407110214,
      "num_tokens": 13516849.0,
      "step": 5900
    },
    {
      "entropy": 5.402905654907227,
      "epoch": 0.5672430355427474,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004974920916310619,
      "loss": 5.1512,
      "mean_token_accuracy": 0.20813584476709365,
      "num_tokens": 13527386.0,
      "step": 5905
    },
    {
      "entropy": 5.332938861846924,
      "epoch": 0.5677233429394812,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004974869845642178,
      "loss": 5.2928,
      "mean_token_accuracy": 0.1970230907201767,
      "num_tokens": 13540238.0,
      "step": 5910
    },
    {
      "entropy": 5.3412638187408445,
      "epoch": 0.5682036503362152,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004974818723318866,
      "loss": 5.1504,
      "mean_token_accuracy": 0.21104614138603212,
      "num_tokens": 13551808.0,
      "step": 5915
    },
    {
      "entropy": 5.365589952468872,
      "epoch": 0.5686839577329491,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004974767549341868,
      "loss": 5.1893,
      "mean_token_accuracy": 0.20336873829364777,
      "num_tokens": 13563795.0,
      "step": 5920
    },
    {
      "entropy": 5.380089664459229,
      "epoch": 0.569164265129683,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004974716323712376,
      "loss": 5.1722,
      "mean_token_accuracy": 0.2080070421099663,
      "num_tokens": 13575168.0,
      "step": 5925
    },
    {
      "entropy": 5.408396148681641,
      "epoch": 0.5696445725264169,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004974665046431576,
      "loss": 5.2607,
      "mean_token_accuracy": 0.20000140964984894,
      "num_tokens": 13587013.0,
      "step": 5930
    },
    {
      "entropy": 5.3197416305542,
      "epoch": 0.5701248799231509,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004974613717500659,
      "loss": 5.1139,
      "mean_token_accuracy": 0.20427340120077134,
      "num_tokens": 13598138.0,
      "step": 5935
    },
    {
      "entropy": 5.390833044052124,
      "epoch": 0.5706051873198847,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004974562336920818,
      "loss": 5.2304,
      "mean_token_accuracy": 0.20116340070962907,
      "num_tokens": 13609174.0,
      "step": 5940
    },
    {
      "entropy": 5.350306415557862,
      "epoch": 0.5710854947166186,
      "grad_norm": 1.125,
      "learning_rate": 0.0004974510904693245,
      "loss": 5.1381,
      "mean_token_accuracy": 0.2101511612534523,
      "num_tokens": 13619724.0,
      "step": 5945
    },
    {
      "entropy": 5.372778129577637,
      "epoch": 0.5715658021133525,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004974459420819134,
      "loss": 5.1858,
      "mean_token_accuracy": 0.20770729184150696,
      "num_tokens": 13631114.0,
      "step": 5950
    },
    {
      "entropy": 5.350826692581177,
      "epoch": 0.5720461095100865,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000497440788529968,
      "loss": 5.1166,
      "mean_token_accuracy": 0.21026804000139238,
      "num_tokens": 13643035.0,
      "step": 5955
    },
    {
      "entropy": 5.3847267627716064,
      "epoch": 0.5725264169068204,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004974356298136081,
      "loss": 5.191,
      "mean_token_accuracy": 0.20383514314889908,
      "num_tokens": 13656603.0,
      "step": 5960
    },
    {
      "entropy": 5.3011445045471195,
      "epoch": 0.5730067243035543,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004974304659329533,
      "loss": 5.2123,
      "mean_token_accuracy": 0.20558879226446153,
      "num_tokens": 13668420.0,
      "step": 5965
    },
    {
      "entropy": 5.403281784057617,
      "epoch": 0.5734870317002881,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004974252968881236,
      "loss": 5.2207,
      "mean_token_accuracy": 0.20342885106801986,
      "num_tokens": 13680521.0,
      "step": 5970
    },
    {
      "entropy": 5.330495548248291,
      "epoch": 0.5739673390970221,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000497420122679239,
      "loss": 5.1118,
      "mean_token_accuracy": 0.21299284994602202,
      "num_tokens": 13692503.0,
      "step": 5975
    },
    {
      "entropy": 5.244579029083252,
      "epoch": 0.574447646493756,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004974149433064196,
      "loss": 5.0048,
      "mean_token_accuracy": 0.2173138216137886,
      "num_tokens": 13703318.0,
      "step": 5980
    },
    {
      "entropy": 5.374129772186279,
      "epoch": 0.5749279538904899,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004974097587697856,
      "loss": 5.1934,
      "mean_token_accuracy": 0.20268867909908295,
      "num_tokens": 13713957.0,
      "step": 5985
    },
    {
      "entropy": 5.379579162597656,
      "epoch": 0.5754082612872238,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004974045690694575,
      "loss": 5.2935,
      "mean_token_accuracy": 0.19796367436647416,
      "num_tokens": 13725300.0,
      "step": 5990
    },
    {
      "entropy": 5.396798706054687,
      "epoch": 0.5758885686839578,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004973993742055557,
      "loss": 5.2496,
      "mean_token_accuracy": 0.20056764930486679,
      "num_tokens": 13735463.0,
      "step": 5995
    },
    {
      "entropy": 5.457717847824097,
      "epoch": 0.5763688760806917,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004973941741782007,
      "loss": 5.3304,
      "mean_token_accuracy": 0.1912854179739952,
      "num_tokens": 13749062.0,
      "step": 6000
    },
    {
      "epoch": 0.5763688760806917,
      "eval_entropy": 5.242419455703557,
      "eval_loss": 5.2162017822265625,
      "eval_mean_token_accuracy": 0.21053085668781715,
      "eval_num_tokens": 13749062.0,
      "eval_runtime": 20.5169,
      "eval_samples_per_second": 1599.412,
      "eval_steps_per_second": 199.933,
      "step": 6000
    }
  ],
  "logging_steps": 5,
  "max_steps": 104090,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 3000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3115039950766080.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}