ita-latn-10mb-after-ppt-Dp-100mb-ckpt500_seed3407/checkpoint-6000/trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.5954743946010321,
  "eval_steps": 3000,
  "global_step": 6000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 7.415488815307617,
      "epoch": 0.0004962286621675268,
      "grad_norm": 1.0390625,
      "learning_rate": 2e-06,
      "loss": 6.9797,
      "mean_token_accuracy": 0.08811815902590751,
      "num_tokens": 11448.0,
      "step": 5
    },
    {
      "entropy": 7.423927688598633,
      "epoch": 0.0009924573243350536,
      "grad_norm": 1.046875,
      "learning_rate": 4.5e-06,
      "loss": 6.9808,
      "mean_token_accuracy": 0.08794851526618004,
      "num_tokens": 22646.0,
      "step": 10
    },
    {
      "entropy": 7.419756555557251,
      "epoch": 0.0014886859865025804,
      "grad_norm": 1.0859375,
      "learning_rate": 7e-06,
      "loss": 6.9117,
      "mean_token_accuracy": 0.08956757485866547,
      "num_tokens": 34011.0,
      "step": 15
    },
    {
      "entropy": 7.414588403701782,
      "epoch": 0.001984914648670107,
      "grad_norm": 0.9296875,
      "learning_rate": 9.5e-06,
      "loss": 6.9364,
      "mean_token_accuracy": 0.09392186626791954,
      "num_tokens": 44566.0,
      "step": 20
    },
    {
      "entropy": 7.422558164596557,
      "epoch": 0.002481143310837634,
      "grad_norm": 0.94921875,
      "learning_rate": 1.2e-05,
      "loss": 7.0066,
      "mean_token_accuracy": 0.08173666819930077,
      "num_tokens": 56274.0,
      "step": 25
    },
    {
      "entropy": 7.404311037063598,
      "epoch": 0.0029773719730051607,
      "grad_norm": 0.95703125,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 6.9198,
      "mean_token_accuracy": 0.09323365837335587,
      "num_tokens": 66584.0,
      "step": 30
    },
    {
      "entropy": 7.403407335281372,
      "epoch": 0.0034736006351726877,
      "grad_norm": 0.98828125,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 6.9536,
      "mean_token_accuracy": 0.08776896521449089,
      "num_tokens": 76948.0,
      "step": 35
    },
    {
      "entropy": 7.424892377853394,
      "epoch": 0.003969829297340214,
      "grad_norm": 0.82421875,
      "learning_rate": 1.95e-05,
      "loss": 6.9625,
      "mean_token_accuracy": 0.09215026199817658,
      "num_tokens": 87253.0,
      "step": 40
    },
    {
      "entropy": 7.402211713790893,
      "epoch": 0.004466057959507741,
      "grad_norm": 0.95703125,
      "learning_rate": 2.2e-05,
      "loss": 6.9664,
      "mean_token_accuracy": 0.0926709569990635,
      "num_tokens": 98273.0,
      "step": 45
    },
    {
      "entropy": 7.390836954116821,
      "epoch": 0.004962286621675268,
      "grad_norm": 1.0390625,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 6.8805,
      "mean_token_accuracy": 0.09474824145436286,
      "num_tokens": 107788.0,
      "step": 50
    },
    {
      "entropy": 7.361630630493164,
      "epoch": 0.0054585152838427945,
      "grad_norm": 0.91015625,
      "learning_rate": 2.7e-05,
      "loss": 6.9007,
      "mean_token_accuracy": 0.08987915515899658,
      "num_tokens": 118632.0,
      "step": 55
    },
    {
      "entropy": 7.329354906082154,
      "epoch": 0.0059547439460103215,
      "grad_norm": 0.89453125,
      "learning_rate": 2.95e-05,
      "loss": 6.8369,
      "mean_token_accuracy": 0.09355762079358101,
      "num_tokens": 128649.0,
      "step": 60
    },
    {
      "entropy": 7.331360340118408,
      "epoch": 0.0064509726081778485,
      "grad_norm": 0.95703125,
      "learning_rate": 3.2e-05,
      "loss": 6.8725,
      "mean_token_accuracy": 0.09848466441035271,
      "num_tokens": 138958.0,
      "step": 65
    },
    {
      "entropy": 7.316051912307739,
      "epoch": 0.0069472012703453755,
      "grad_norm": 0.91796875,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 6.875,
      "mean_token_accuracy": 0.09506641924381257,
      "num_tokens": 151445.0,
      "step": 70
    },
    {
      "entropy": 7.357440185546875,
      "epoch": 0.007443429932512902,
      "grad_norm": 0.8515625,
      "learning_rate": 3.7e-05,
      "loss": 6.9168,
      "mean_token_accuracy": 0.0950050063431263,
      "num_tokens": 162239.0,
      "step": 75
    },
    {
      "entropy": 7.389816427230835,
      "epoch": 0.007939658594680429,
      "grad_norm": 1.0703125,
      "learning_rate": 3.95e-05,
      "loss": 6.8606,
      "mean_token_accuracy": 0.08998646065592766,
      "num_tokens": 173261.0,
      "step": 80
    },
    {
      "entropy": 7.374487590789795,
      "epoch": 0.008435887256847956,
      "grad_norm": 0.7578125,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 6.8612,
      "mean_token_accuracy": 0.08823171928524971,
      "num_tokens": 183857.0,
      "step": 85
    },
    {
      "entropy": 7.406818819046021,
      "epoch": 0.008932115919015483,
      "grad_norm": 0.98046875,
      "learning_rate": 4.45e-05,
      "loss": 7.0573,
      "mean_token_accuracy": 0.08514732420444489,
      "num_tokens": 196635.0,
      "step": 90
    },
    {
      "entropy": 7.393658542633057,
      "epoch": 0.00942834458118301,
      "grad_norm": 0.890625,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 7.0168,
      "mean_token_accuracy": 0.09471790939569473,
      "num_tokens": 208700.0,
      "step": 95
    },
    {
      "entropy": 7.416672229766846,
      "epoch": 0.009924573243350537,
      "grad_norm": 0.9609375,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 6.9349,
      "mean_token_accuracy": 0.08931025043129921,
      "num_tokens": 218915.0,
      "step": 100
    },
    {
      "entropy": 7.340483951568603,
      "epoch": 0.010420801905518062,
      "grad_norm": 0.99609375,
      "learning_rate": 5.2e-05,
      "loss": 6.8607,
      "mean_token_accuracy": 0.0964167907834053,
      "num_tokens": 230211.0,
      "step": 105
    },
    {
      "entropy": 7.323957920074463,
      "epoch": 0.010917030567685589,
      "grad_norm": 0.984375,
      "learning_rate": 5.45e-05,
      "loss": 6.8819,
      "mean_token_accuracy": 0.10094936266541481,
      "num_tokens": 241174.0,
      "step": 110
    },
    {
      "entropy": 7.3142293930053714,
      "epoch": 0.011413259229853116,
      "grad_norm": 0.9375,
      "learning_rate": 5.7e-05,
      "loss": 6.8898,
      "mean_token_accuracy": 0.09185363873839378,
      "num_tokens": 251708.0,
      "step": 115
    },
    {
      "entropy": 7.415872716903687,
      "epoch": 0.011909487892020643,
      "grad_norm": 0.84765625,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 7.0541,
      "mean_token_accuracy": 0.08325114101171494,
      "num_tokens": 263287.0,
      "step": 120
    },
    {
      "entropy": 7.43096809387207,
      "epoch": 0.01240571655418817,
      "grad_norm": 0.828125,
      "learning_rate": 6.2e-05,
      "loss": 6.9198,
      "mean_token_accuracy": 0.0861751176416874,
      "num_tokens": 275179.0,
      "step": 125
    },
    {
      "entropy": 7.31543345451355,
      "epoch": 0.012901945216355697,
      "grad_norm": 0.87890625,
      "learning_rate": 6.450000000000001e-05,
      "loss": 6.7886,
      "mean_token_accuracy": 0.09708081930875778,
      "num_tokens": 287223.0,
      "step": 130
    },
    {
      "entropy": 7.303066396713257,
      "epoch": 0.013398173878523224,
      "grad_norm": 1.296875,
      "learning_rate": 6.7e-05,
      "loss": 6.8086,
      "mean_token_accuracy": 0.09785129725933076,
      "num_tokens": 297868.0,
      "step": 135
    },
    {
      "entropy": 7.2979652881622314,
      "epoch": 0.013894402540690751,
      "grad_norm": 0.78515625,
      "learning_rate": 6.950000000000001e-05,
      "loss": 6.8927,
      "mean_token_accuracy": 0.09260747209191322,
      "num_tokens": 309702.0,
      "step": 140
    },
    {
      "entropy": 7.4375659942626955,
      "epoch": 0.014390631202858276,
      "grad_norm": 0.89453125,
      "learning_rate": 7.2e-05,
      "loss": 6.906,
      "mean_token_accuracy": 0.09100511819124221,
      "num_tokens": 321613.0,
      "step": 145
    },
    {
      "entropy": 7.383455657958985,
      "epoch": 0.014886859865025803,
      "grad_norm": 0.921875,
      "learning_rate": 7.45e-05,
      "loss": 6.8772,
      "mean_token_accuracy": 0.09483638480305671,
      "num_tokens": 332668.0,
      "step": 150
    },
    {
      "entropy": 7.347703647613526,
      "epoch": 0.01538308852719333,
      "grad_norm": 0.9609375,
      "learning_rate": 7.7e-05,
      "loss": 6.9471,
      "mean_token_accuracy": 0.09541060924530029,
      "num_tokens": 342579.0,
      "step": 155
    },
    {
      "entropy": 7.395112085342407,
      "epoch": 0.015879317189360857,
      "grad_norm": 1.0546875,
      "learning_rate": 7.950000000000001e-05,
      "loss": 6.7956,
      "mean_token_accuracy": 0.09388362988829613,
      "num_tokens": 352271.0,
      "step": 160
    },
    {
      "entropy": 7.331162643432617,
      "epoch": 0.016375545851528384,
      "grad_norm": 0.9140625,
      "learning_rate": 8.2e-05,
      "loss": 6.932,
      "mean_token_accuracy": 0.09907099306583404,
      "num_tokens": 364033.0,
      "step": 165
    },
    {
      "entropy": 7.286223697662353,
      "epoch": 0.01687177451369591,
      "grad_norm": 0.8671875,
      "learning_rate": 8.450000000000001e-05,
      "loss": 6.8812,
      "mean_token_accuracy": 0.09280342161655426,
      "num_tokens": 373788.0,
      "step": 170
    },
    {
      "entropy": 7.398203802108765,
      "epoch": 0.01736800317586344,
      "grad_norm": 0.98046875,
      "learning_rate": 8.7e-05,
      "loss": 6.8526,
      "mean_token_accuracy": 0.09731495082378387,
      "num_tokens": 384883.0,
      "step": 175
    },
    {
      "entropy": 7.3108381748199465,
      "epoch": 0.017864231838030965,
      "grad_norm": 1.046875,
      "learning_rate": 8.95e-05,
      "loss": 6.7825,
      "mean_token_accuracy": 0.09594601169228553,
      "num_tokens": 396044.0,
      "step": 180
    },
    {
      "entropy": 7.364286279678344,
      "epoch": 0.018360460500198492,
      "grad_norm": 1.125,
      "learning_rate": 9.2e-05,
      "loss": 6.9289,
      "mean_token_accuracy": 0.09398710206151009,
      "num_tokens": 406995.0,
      "step": 185
    },
    {
      "entropy": 7.368616962432862,
      "epoch": 0.01885668916236602,
      "grad_norm": 1.03125,
      "learning_rate": 9.45e-05,
      "loss": 6.8242,
      "mean_token_accuracy": 0.09675540551543235,
      "num_tokens": 417238.0,
      "step": 190
    },
    {
      "entropy": 7.337496566772461,
      "epoch": 0.019352917824533546,
      "grad_norm": 0.9609375,
      "learning_rate": 9.7e-05,
      "loss": 6.869,
      "mean_token_accuracy": 0.09491769820451737,
      "num_tokens": 427263.0,
      "step": 195
    },
    {
      "entropy": 7.282687520980835,
      "epoch": 0.019849146486701073,
      "grad_norm": 0.8515625,
      "learning_rate": 9.95e-05,
      "loss": 6.7887,
      "mean_token_accuracy": 0.1014923520386219,
      "num_tokens": 438357.0,
      "step": 200
    },
    {
      "entropy": 7.343814468383789,
      "epoch": 0.0203453751488686,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000102,
      "loss": 6.8331,
      "mean_token_accuracy": 0.09941279143095016,
      "num_tokens": 448738.0,
      "step": 205
    },
    {
      "entropy": 7.334179782867432,
      "epoch": 0.020841603811036124,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00010449999999999999,
      "loss": 6.8794,
      "mean_token_accuracy": 0.09967483356595039,
      "num_tokens": 457873.0,
      "step": 210
    },
    {
      "entropy": 7.292327451705932,
      "epoch": 0.02133783247320365,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000107,
      "loss": 6.8026,
      "mean_token_accuracy": 0.09620003700256348,
      "num_tokens": 468640.0,
      "step": 215
    },
    {
      "entropy": 7.303198051452637,
      "epoch": 0.021834061135371178,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0001095,
      "loss": 6.8446,
      "mean_token_accuracy": 0.09441228732466697,
      "num_tokens": 479922.0,
      "step": 220
    },
    {
      "entropy": 7.285147094726563,
      "epoch": 0.022330289797538705,
      "grad_norm": 0.984375,
      "learning_rate": 0.000112,
      "loss": 6.7865,
      "mean_token_accuracy": 0.098848607391119,
      "num_tokens": 491056.0,
      "step": 225
    },
    {
      "entropy": 7.386207151412964,
      "epoch": 0.022826518459706232,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0001145,
      "loss": 6.9114,
      "mean_token_accuracy": 0.09412190616130829,
      "num_tokens": 504365.0,
      "step": 230
    },
    {
      "entropy": 7.306986141204834,
      "epoch": 0.02332274712187376,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00011700000000000001,
      "loss": 6.7885,
      "mean_token_accuracy": 0.09399791210889816,
      "num_tokens": 514347.0,
      "step": 235
    },
    {
      "entropy": 7.284820985794068,
      "epoch": 0.023818975784041286,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00011949999999999999,
      "loss": 6.8463,
      "mean_token_accuracy": 0.09885015189647675,
      "num_tokens": 525912.0,
      "step": 240
    },
    {
      "entropy": 7.264670991897583,
      "epoch": 0.024315204446208813,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000122,
      "loss": 6.722,
      "mean_token_accuracy": 0.09802954122424126,
      "num_tokens": 535566.0,
      "step": 245
    },
    {
      "entropy": 7.293220567703247,
      "epoch": 0.02481143310837634,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0001245,
      "loss": 6.7569,
      "mean_token_accuracy": 0.09652359411120415,
      "num_tokens": 546474.0,
      "step": 250
    },
    {
      "entropy": 7.343952226638794,
      "epoch": 0.025307661770543867,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000127,
      "loss": 6.9085,
      "mean_token_accuracy": 0.09467220306396484,
      "num_tokens": 558132.0,
      "step": 255
    },
    {
      "entropy": 7.391662740707398,
      "epoch": 0.025803890432711394,
      "grad_norm": 1.25,
      "learning_rate": 0.0001295,
      "loss": 6.8717,
      "mean_token_accuracy": 0.08968425244092941,
      "num_tokens": 568309.0,
      "step": 260
    },
    {
      "entropy": 7.242248058319092,
      "epoch": 0.02630011909487892,
      "grad_norm": 1.109375,
      "learning_rate": 0.000132,
      "loss": 6.7985,
      "mean_token_accuracy": 0.09653588980436326,
      "num_tokens": 579701.0,
      "step": 265
    },
    {
      "entropy": 7.3309746265411375,
      "epoch": 0.026796347757046448,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00013450000000000002,
      "loss": 6.7519,
      "mean_token_accuracy": 0.09935276582837105,
      "num_tokens": 590832.0,
      "step": 270
    },
    {
      "entropy": 7.18915901184082,
      "epoch": 0.027292576419213975,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00013700000000000002,
      "loss": 6.7155,
      "mean_token_accuracy": 0.09612485915422439,
      "num_tokens": 601094.0,
      "step": 275
    },
    {
      "entropy": 7.255670118331909,
      "epoch": 0.027788805081381502,
      "grad_norm": 1.09375,
      "learning_rate": 0.0001395,
      "loss": 6.7341,
      "mean_token_accuracy": 0.1019403874874115,
      "num_tokens": 612027.0,
      "step": 280
    },
    {
      "entropy": 7.241375875473023,
      "epoch": 0.02828503374354903,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00014199999999999998,
      "loss": 6.7852,
      "mean_token_accuracy": 0.0944296196103096,
      "num_tokens": 623698.0,
      "step": 285
    },
    {
      "entropy": 7.326890134811402,
      "epoch": 0.028781262405716553,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0001445,
      "loss": 6.8125,
      "mean_token_accuracy": 0.09790188670158387,
      "num_tokens": 635631.0,
      "step": 290
    },
    {
      "entropy": 7.2972441673278805,
      "epoch": 0.02927749106788408,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000147,
      "loss": 6.8098,
      "mean_token_accuracy": 0.09671351835131645,
      "num_tokens": 647663.0,
      "step": 295
    },
    {
      "entropy": 7.226529216766357,
      "epoch": 0.029773719730051607,
      "grad_norm": 1.0,
      "learning_rate": 0.0001495,
      "loss": 6.7771,
      "mean_token_accuracy": 0.0953554093837738,
      "num_tokens": 659352.0,
      "step": 300
    },
    {
      "entropy": 7.249944972991943,
      "epoch": 0.030269948392219134,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000152,
      "loss": 6.7277,
      "mean_token_accuracy": 0.10546225905418397,
      "num_tokens": 669379.0,
      "step": 305
    },
    {
      "entropy": 7.287822151184082,
      "epoch": 0.03076617705438666,
      "grad_norm": 1.171875,
      "learning_rate": 0.00015450000000000001,
      "loss": 6.7385,
      "mean_token_accuracy": 0.10098289772868156,
      "num_tokens": 680144.0,
      "step": 310
    },
    {
      "entropy": 7.2293524742126465,
      "epoch": 0.03126240571655419,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000157,
      "loss": 6.6844,
      "mean_token_accuracy": 0.09755392372608185,
      "num_tokens": 691165.0,
      "step": 315
    },
    {
      "entropy": 7.23696551322937,
      "epoch": 0.031758634378721715,
      "grad_norm": 1.171875,
      "learning_rate": 0.0001595,
      "loss": 6.8136,
      "mean_token_accuracy": 0.09727144762873649,
      "num_tokens": 703767.0,
      "step": 320
    },
    {
      "entropy": 7.2687766551971436,
      "epoch": 0.032254863040889245,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000162,
      "loss": 6.7885,
      "mean_token_accuracy": 0.10152007564902306,
      "num_tokens": 714298.0,
      "step": 325
    },
    {
      "entropy": 7.237318611145019,
      "epoch": 0.03275109170305677,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00016450000000000001,
      "loss": 6.7841,
      "mean_token_accuracy": 0.10548736825585366,
      "num_tokens": 724135.0,
      "step": 330
    },
    {
      "entropy": 7.179241037368774,
      "epoch": 0.03324732036522429,
      "grad_norm": 0.9375,
      "learning_rate": 0.00016700000000000002,
      "loss": 6.8555,
      "mean_token_accuracy": 0.09413279443979264,
      "num_tokens": 735050.0,
      "step": 335
    },
    {
      "entropy": 7.315862798690796,
      "epoch": 0.03374354902739182,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00016950000000000003,
      "loss": 6.7585,
      "mean_token_accuracy": 0.09567313492298127,
      "num_tokens": 746073.0,
      "step": 340
    },
    {
      "entropy": 7.213669204711914,
      "epoch": 0.034239777689559346,
      "grad_norm": 1.015625,
      "learning_rate": 0.00017199999999999998,
      "loss": 6.7686,
      "mean_token_accuracy": 0.09640354067087173,
      "num_tokens": 758998.0,
      "step": 345
    },
    {
      "entropy": 7.203896570205688,
      "epoch": 0.03473600635172688,
      "grad_norm": 1.078125,
      "learning_rate": 0.00017449999999999999,
      "loss": 6.6867,
      "mean_token_accuracy": 0.09490930512547494,
      "num_tokens": 771608.0,
      "step": 350
    },
    {
      "entropy": 7.220716094970703,
      "epoch": 0.0352322350138944,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000177,
      "loss": 6.7711,
      "mean_token_accuracy": 0.09604462832212449,
      "num_tokens": 782394.0,
      "step": 355
    },
    {
      "entropy": 7.213645648956299,
      "epoch": 0.03572846367606193,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0001795,
      "loss": 6.6827,
      "mean_token_accuracy": 0.10222365930676461,
      "num_tokens": 792389.0,
      "step": 360
    },
    {
      "entropy": 7.195154285430908,
      "epoch": 0.036224692338229454,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000182,
      "loss": 6.7383,
      "mean_token_accuracy": 0.09877654984593391,
      "num_tokens": 803219.0,
      "step": 365
    },
    {
      "entropy": 7.26901273727417,
      "epoch": 0.036720921000396985,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0001845,
      "loss": 6.8117,
      "mean_token_accuracy": 0.09464164227247238,
      "num_tokens": 814249.0,
      "step": 370
    },
    {
      "entropy": 7.248525381088257,
      "epoch": 0.03721714966256451,
      "grad_norm": 0.984375,
      "learning_rate": 0.000187,
      "loss": 6.7002,
      "mean_token_accuracy": 0.10740549117326736,
      "num_tokens": 825138.0,
      "step": 375
    },
    {
      "entropy": 7.151234674453735,
      "epoch": 0.03771337832473204,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0001895,
      "loss": 6.7529,
      "mean_token_accuracy": 0.10170127227902412,
      "num_tokens": 836024.0,
      "step": 380
    },
    {
      "entropy": 7.221024084091186,
      "epoch": 0.03820960698689956,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000192,
      "loss": 6.7426,
      "mean_token_accuracy": 0.0970191240310669,
      "num_tokens": 847765.0,
      "step": 385
    },
    {
      "entropy": 7.190954208374023,
      "epoch": 0.03870583564906709,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0001945,
      "loss": 6.7344,
      "mean_token_accuracy": 0.0974167302250862,
      "num_tokens": 859980.0,
      "step": 390
    },
    {
      "entropy": 7.299342060089112,
      "epoch": 0.039202064311234616,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00019700000000000002,
      "loss": 6.7494,
      "mean_token_accuracy": 0.09645001664757728,
      "num_tokens": 871514.0,
      "step": 395
    },
    {
      "entropy": 7.188382196426391,
      "epoch": 0.03969829297340215,
      "grad_norm": 1.046875,
      "learning_rate": 0.00019950000000000002,
      "loss": 6.7508,
      "mean_token_accuracy": 0.09433658048510551,
      "num_tokens": 881984.0,
      "step": 400
    },
    {
      "entropy": 7.196045541763306,
      "epoch": 0.04019452163556967,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000202,
      "loss": 6.7063,
      "mean_token_accuracy": 0.10188237726688384,
      "num_tokens": 894198.0,
      "step": 405
    },
    {
      "entropy": 7.18197021484375,
      "epoch": 0.0406907502977372,
      "grad_norm": 1.0,
      "learning_rate": 0.00020449999999999998,
      "loss": 6.719,
      "mean_token_accuracy": 0.09976514428853989,
      "num_tokens": 905394.0,
      "step": 410
    },
    {
      "entropy": 7.248855495452881,
      "epoch": 0.041186978959904724,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000207,
      "loss": 6.7958,
      "mean_token_accuracy": 0.09452574029564857,
      "num_tokens": 917010.0,
      "step": 415
    },
    {
      "entropy": 7.155883693695069,
      "epoch": 0.04168320762207225,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002095,
      "loss": 6.6121,
      "mean_token_accuracy": 0.10424820482730865,
      "num_tokens": 926761.0,
      "step": 420
    },
    {
      "entropy": 7.060562229156494,
      "epoch": 0.04217943628423978,
      "grad_norm": 1.0625,
      "learning_rate": 0.000212,
      "loss": 6.6519,
      "mean_token_accuracy": 0.10032859295606614,
      "num_tokens": 938385.0,
      "step": 425
    },
    {
      "entropy": 7.134511184692383,
      "epoch": 0.0426756649464073,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002145,
      "loss": 6.6897,
      "mean_token_accuracy": 0.0993834190070629,
      "num_tokens": 949709.0,
      "step": 430
    },
    {
      "entropy": 7.11284236907959,
      "epoch": 0.04317189360857483,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00021700000000000002,
      "loss": 6.7173,
      "mean_token_accuracy": 0.09899322465062141,
      "num_tokens": 960837.0,
      "step": 435
    },
    {
      "entropy": 7.13794903755188,
      "epoch": 0.043668122270742356,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0002195,
      "loss": 6.5545,
      "mean_token_accuracy": 0.10581014305353165,
      "num_tokens": 971036.0,
      "step": 440
    },
    {
      "entropy": 7.189764070510864,
      "epoch": 0.044164350932909886,
      "grad_norm": 1.078125,
      "learning_rate": 0.000222,
      "loss": 6.6988,
      "mean_token_accuracy": 0.09889093339443207,
      "num_tokens": 983057.0,
      "step": 445
    },
    {
      "entropy": 7.109373235702515,
      "epoch": 0.04466057959507741,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002245,
      "loss": 6.657,
      "mean_token_accuracy": 0.10294226482510567,
      "num_tokens": 994303.0,
      "step": 450
    },
    {
      "entropy": 7.106617975234985,
      "epoch": 0.04515680825724494,
      "grad_norm": 1.09375,
      "learning_rate": 0.00022700000000000002,
      "loss": 6.6683,
      "mean_token_accuracy": 0.1027058370411396,
      "num_tokens": 1005033.0,
      "step": 455
    },
    {
      "entropy": 7.166249561309814,
      "epoch": 0.045653036919412464,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00022950000000000002,
      "loss": 6.7313,
      "mean_token_accuracy": 0.10222981199622154,
      "num_tokens": 1015940.0,
      "step": 460
    },
    {
      "entropy": 7.1316859245300295,
      "epoch": 0.046149265581579994,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00023200000000000003,
      "loss": 6.6994,
      "mean_token_accuracy": 0.1031339094042778,
      "num_tokens": 1027582.0,
      "step": 465
    },
    {
      "entropy": 7.089751434326172,
      "epoch": 0.04664549424374752,
      "grad_norm": 0.9375,
      "learning_rate": 0.00023449999999999998,
      "loss": 6.6675,
      "mean_token_accuracy": 0.10097738355398178,
      "num_tokens": 1040046.0,
      "step": 470
    },
    {
      "entropy": 7.186102390289307,
      "epoch": 0.04714172290591505,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000237,
      "loss": 6.7444,
      "mean_token_accuracy": 0.09288590252399445,
      "num_tokens": 1050636.0,
      "step": 475
    },
    {
      "entropy": 7.142090320587158,
      "epoch": 0.04763795156808257,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002395,
      "loss": 6.6294,
      "mean_token_accuracy": 0.10634742677211761,
      "num_tokens": 1060748.0,
      "step": 480
    },
    {
      "entropy": 7.175977611541748,
      "epoch": 0.0481341802302501,
      "grad_norm": 0.875,
      "learning_rate": 0.000242,
      "loss": 6.6716,
      "mean_token_accuracy": 0.09640654549002647,
      "num_tokens": 1071940.0,
      "step": 485
    },
    {
      "entropy": 7.124125528335571,
      "epoch": 0.048630408892417626,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002445,
      "loss": 6.701,
      "mean_token_accuracy": 0.10127488672733306,
      "num_tokens": 1082962.0,
      "step": 490
    },
    {
      "entropy": 7.0944671630859375,
      "epoch": 0.04912663755458515,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000247,
      "loss": 6.6792,
      "mean_token_accuracy": 0.1065860316157341,
      "num_tokens": 1093082.0,
      "step": 495
    },
    {
      "entropy": 7.197715377807617,
      "epoch": 0.04962286621675268,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002495,
      "loss": 6.7682,
      "mean_token_accuracy": 0.10142254531383514,
      "num_tokens": 1103528.0,
      "step": 500
    },
    {
      "entropy": 7.073600006103516,
      "epoch": 0.0501190948789202,
      "grad_norm": 1.15625,
      "learning_rate": 0.000252,
      "loss": 6.6416,
      "mean_token_accuracy": 0.0960576593875885,
      "num_tokens": 1114813.0,
      "step": 505
    },
    {
      "entropy": 7.003829717636108,
      "epoch": 0.050615323541087734,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002545,
      "loss": 6.6403,
      "mean_token_accuracy": 0.09825952127575874,
      "num_tokens": 1125345.0,
      "step": 510
    },
    {
      "entropy": 7.108769464492798,
      "epoch": 0.05111155220325526,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000257,
      "loss": 6.6572,
      "mean_token_accuracy": 0.09955623000860214,
      "num_tokens": 1134985.0,
      "step": 515
    },
    {
      "entropy": 7.135790109634399,
      "epoch": 0.05160778086542279,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002595,
      "loss": 6.6783,
      "mean_token_accuracy": 0.10150780752301217,
      "num_tokens": 1145796.0,
      "step": 520
    },
    {
      "entropy": 7.187171220779419,
      "epoch": 0.05210400952759031,
      "grad_norm": 1.125,
      "learning_rate": 0.000262,
      "loss": 6.8131,
      "mean_token_accuracy": 0.09044179245829582,
      "num_tokens": 1156936.0,
      "step": 525
    },
    {
      "entropy": 7.090089416503906,
      "epoch": 0.05260023818975784,
      "grad_norm": 1.046875,
      "learning_rate": 0.00026450000000000003,
      "loss": 6.6959,
      "mean_token_accuracy": 0.10054284557700158,
      "num_tokens": 1167526.0,
      "step": 530
    },
    {
      "entropy": 7.1854321479797365,
      "epoch": 0.053096466851925365,
      "grad_norm": 1.25,
      "learning_rate": 0.00026700000000000004,
      "loss": 6.7306,
      "mean_token_accuracy": 0.09534744620323181,
      "num_tokens": 1177647.0,
      "step": 535
    },
    {
      "entropy": 7.077204322814941,
      "epoch": 0.053592695514092896,
      "grad_norm": 1.140625,
      "learning_rate": 0.00026950000000000005,
      "loss": 6.6923,
      "mean_token_accuracy": 0.1065493330359459,
      "num_tokens": 1188255.0,
      "step": 540
    },
    {
      "entropy": 7.047231197357178,
      "epoch": 0.05408892417626042,
      "grad_norm": 0.984375,
      "learning_rate": 0.00027200000000000005,
      "loss": 6.6221,
      "mean_token_accuracy": 0.11043694391846656,
      "num_tokens": 1199433.0,
      "step": 545
    },
    {
      "entropy": 7.008746385574341,
      "epoch": 0.05458515283842795,
      "grad_norm": 0.890625,
      "learning_rate": 0.0002745,
      "loss": 6.7064,
      "mean_token_accuracy": 0.09511427283287048,
      "num_tokens": 1210067.0,
      "step": 550
    },
    {
      "entropy": 7.019316959381103,
      "epoch": 0.055081381500595473,
      "grad_norm": 1.015625,
      "learning_rate": 0.000277,
      "loss": 6.5836,
      "mean_token_accuracy": 0.1026071272790432,
      "num_tokens": 1220639.0,
      "step": 555
    },
    {
      "entropy": 7.124681997299194,
      "epoch": 0.055577610162763004,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002795,
      "loss": 6.6339,
      "mean_token_accuracy": 0.10223110392689705,
      "num_tokens": 1233158.0,
      "step": 560
    },
    {
      "entropy": 7.022218942642212,
      "epoch": 0.05607383882493053,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00028199999999999997,
      "loss": 6.6128,
      "mean_token_accuracy": 0.10005637407302856,
      "num_tokens": 1244552.0,
      "step": 565
    },
    {
      "entropy": 7.1609320640563965,
      "epoch": 0.05657006748709806,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002845,
      "loss": 6.7327,
      "mean_token_accuracy": 0.09698689877986907,
      "num_tokens": 1255478.0,
      "step": 570
    },
    {
      "entropy": 6.899557256698609,
      "epoch": 0.05706629614926558,
      "grad_norm": 1.015625,
      "learning_rate": 0.000287,
      "loss": 6.5604,
      "mean_token_accuracy": 0.10638210996985435,
      "num_tokens": 1267507.0,
      "step": 575
    },
    {
      "entropy": 7.014608669281006,
      "epoch": 0.057562524811433105,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0002895,
      "loss": 6.5649,
      "mean_token_accuracy": 0.09933194518089294,
      "num_tokens": 1279377.0,
      "step": 580
    },
    {
      "entropy": 7.112942886352539,
      "epoch": 0.058058753473600636,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000292,
      "loss": 6.6557,
      "mean_token_accuracy": 0.10565379559993744,
      "num_tokens": 1290671.0,
      "step": 585
    },
    {
      "entropy": 6.903086757659912,
      "epoch": 0.05855498213576816,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002945,
      "loss": 6.6323,
      "mean_token_accuracy": 0.10672209039330482,
      "num_tokens": 1302792.0,
      "step": 590
    },
    {
      "entropy": 7.101827001571655,
      "epoch": 0.05905121079793569,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000297,
      "loss": 6.5868,
      "mean_token_accuracy": 0.09854365140199661,
      "num_tokens": 1313315.0,
      "step": 595
    },
    {
      "entropy": 6.9200047016143795,
      "epoch": 0.05954743946010321,
      "grad_norm": 1.125,
      "learning_rate": 0.0002995,
      "loss": 6.6504,
      "mean_token_accuracy": 0.09922990500926972,
      "num_tokens": 1326358.0,
      "step": 600
    },
    {
      "entropy": 7.008511209487915,
      "epoch": 0.060043668122270744,
      "grad_norm": 1.0625,
      "learning_rate": 0.000302,
      "loss": 6.5085,
      "mean_token_accuracy": 0.1003392606973648,
      "num_tokens": 1336266.0,
      "step": 605
    },
    {
      "entropy": 6.907269525527954,
      "epoch": 0.06053989678443827,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003045,
      "loss": 6.4734,
      "mean_token_accuracy": 0.10566542968153954,
      "num_tokens": 1347610.0,
      "step": 610
    },
    {
      "entropy": 6.930313205718994,
      "epoch": 0.0610361254466058,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000307,
      "loss": 6.4398,
      "mean_token_accuracy": 0.10840422213077545,
      "num_tokens": 1358317.0,
      "step": 615
    },
    {
      "entropy": 6.901958322525024,
      "epoch": 0.06153235410877332,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003095,
      "loss": 6.5551,
      "mean_token_accuracy": 0.1008565068244934,
      "num_tokens": 1370126.0,
      "step": 620
    },
    {
      "entropy": 6.962115097045898,
      "epoch": 0.06202858277094085,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000312,
      "loss": 6.5148,
      "mean_token_accuracy": 0.10650759786367417,
      "num_tokens": 1381571.0,
      "step": 625
    },
    {
      "entropy": 6.91578860282898,
      "epoch": 0.06252481143310838,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003145,
      "loss": 6.5516,
      "mean_token_accuracy": 0.10437883734703064,
      "num_tokens": 1392486.0,
      "step": 630
    },
    {
      "entropy": 6.933821678161621,
      "epoch": 0.0630210400952759,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000317,
      "loss": 6.5571,
      "mean_token_accuracy": 0.10577636584639549,
      "num_tokens": 1403086.0,
      "step": 635
    },
    {
      "entropy": 6.9292055606842045,
      "epoch": 0.06351726875744343,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003195,
      "loss": 6.5243,
      "mean_token_accuracy": 0.10535438284277916,
      "num_tokens": 1414537.0,
      "step": 640
    },
    {
      "entropy": 6.871757984161377,
      "epoch": 0.06401349741961096,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000322,
      "loss": 6.5485,
      "mean_token_accuracy": 0.10128815025091172,
      "num_tokens": 1426998.0,
      "step": 645
    },
    {
      "entropy": 6.936267471313476,
      "epoch": 0.06450972608177849,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.5753,
      "mean_token_accuracy": 0.10118880569934845,
      "num_tokens": 1437667.0,
      "step": 650
    },
    {
      "entropy": 6.886267423629761,
      "epoch": 0.065005954743946,
      "grad_norm": 1.0,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.4528,
      "mean_token_accuracy": 0.10906932204961776,
      "num_tokens": 1448280.0,
      "step": 655
    },
    {
      "entropy": 6.921428108215332,
      "epoch": 0.06550218340611354,
      "grad_norm": 1.234375,
      "learning_rate": 0.00032950000000000004,
      "loss": 6.5105,
      "mean_token_accuracy": 0.10349921584129333,
      "num_tokens": 1458608.0,
      "step": 660
    },
    {
      "entropy": 6.980496501922607,
      "epoch": 0.06599841206828107,
      "grad_norm": 1.203125,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.695,
      "mean_token_accuracy": 0.09470629766583442,
      "num_tokens": 1470435.0,
      "step": 665
    },
    {
      "entropy": 6.830832529067993,
      "epoch": 0.06649464073044858,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00033450000000000005,
      "loss": 6.5469,
      "mean_token_accuracy": 0.10282312780618667,
      "num_tokens": 1481959.0,
      "step": 670
    },
    {
      "entropy": 6.974913597106934,
      "epoch": 0.06699086939261611,
      "grad_norm": 1.015625,
      "learning_rate": 0.000337,
      "loss": 6.5976,
      "mean_token_accuracy": 0.09841747283935547,
      "num_tokens": 1493798.0,
      "step": 675
    },
    {
      "entropy": 6.887060308456421,
      "epoch": 0.06748709805478365,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003395,
      "loss": 6.403,
      "mean_token_accuracy": 0.10493207052350044,
      "num_tokens": 1502969.0,
      "step": 680
    },
    {
      "entropy": 6.897101402282715,
      "epoch": 0.06798332671695118,
      "grad_norm": 1.1875,
      "learning_rate": 0.000342,
      "loss": 6.5595,
      "mean_token_accuracy": 0.10253233015537262,
      "num_tokens": 1513825.0,
      "step": 685
    },
    {
      "entropy": 6.822127819061279,
      "epoch": 0.06847955537911869,
      "grad_norm": 0.953125,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.4575,
      "mean_token_accuracy": 0.10815273150801659,
      "num_tokens": 1525328.0,
      "step": 690
    },
    {
      "entropy": 6.891428852081299,
      "epoch": 0.06897578404128622,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000347,
      "loss": 6.4859,
      "mean_token_accuracy": 0.10564783737063407,
      "num_tokens": 1537383.0,
      "step": 695
    },
    {
      "entropy": 6.913006925582886,
      "epoch": 0.06947201270345375,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003495,
      "loss": 6.5826,
      "mean_token_accuracy": 0.10424128845334053,
      "num_tokens": 1549671.0,
      "step": 700
    },
    {
      "entropy": 6.7738439559936525,
      "epoch": 0.06996824136562128,
      "grad_norm": 1.171875,
      "learning_rate": 0.000352,
      "loss": 6.4562,
      "mean_token_accuracy": 0.1023649513721466,
      "num_tokens": 1559974.0,
      "step": 705
    },
    {
      "entropy": 6.948292827606201,
      "epoch": 0.0704644700277888,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003545,
      "loss": 6.5642,
      "mean_token_accuracy": 0.09899581745266914,
      "num_tokens": 1570890.0,
      "step": 710
    },
    {
      "entropy": 6.786388778686524,
      "epoch": 0.07096069868995633,
      "grad_norm": 1.15625,
      "learning_rate": 0.000357,
      "loss": 6.473,
      "mean_token_accuracy": 0.11216894015669823,
      "num_tokens": 1581434.0,
      "step": 715
    },
    {
      "entropy": 6.885098886489868,
      "epoch": 0.07145692735212386,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003595,
      "loss": 6.5411,
      "mean_token_accuracy": 0.1010577216744423,
      "num_tokens": 1593710.0,
      "step": 720
    },
    {
      "entropy": 6.914775133132935,
      "epoch": 0.07195315601429139,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000362,
      "loss": 6.5039,
      "mean_token_accuracy": 0.10822037607431412,
      "num_tokens": 1604635.0,
      "step": 725
    },
    {
      "entropy": 6.96195125579834,
      "epoch": 0.07244938467645891,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003645,
      "loss": 6.5621,
      "mean_token_accuracy": 0.10865710079669952,
      "num_tokens": 1615258.0,
      "step": 730
    },
    {
      "entropy": 6.818844604492187,
      "epoch": 0.07294561333862644,
      "grad_norm": 1.0,
      "learning_rate": 0.000367,
      "loss": 6.4657,
      "mean_token_accuracy": 0.1062205784022808,
      "num_tokens": 1625942.0,
      "step": 735
    },
    {
      "entropy": 6.773487854003906,
      "epoch": 0.07344184200079397,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003695,
      "loss": 6.4609,
      "mean_token_accuracy": 0.11281525269150734,
      "num_tokens": 1636270.0,
      "step": 740
    },
    {
      "entropy": 6.791012191772461,
      "epoch": 0.07393807066296149,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000372,
      "loss": 6.442,
      "mean_token_accuracy": 0.10923936069011689,
      "num_tokens": 1647029.0,
      "step": 745
    },
    {
      "entropy": 6.894232225418091,
      "epoch": 0.07443429932512902,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003745,
      "loss": 6.6043,
      "mean_token_accuracy": 0.10349742248654366,
      "num_tokens": 1659483.0,
      "step": 750
    },
    {
      "entropy": 6.763030767440796,
      "epoch": 0.07493052798729655,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000377,
      "loss": 6.4626,
      "mean_token_accuracy": 0.10359987020492553,
      "num_tokens": 1669425.0,
      "step": 755
    },
    {
      "entropy": 6.839314746856689,
      "epoch": 0.07542675664946408,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003795,
      "loss": 6.4669,
      "mean_token_accuracy": 0.10515316724777221,
      "num_tokens": 1680426.0,
      "step": 760
    },
    {
      "entropy": 6.7409745216369625,
      "epoch": 0.0759229853116316,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000382,
      "loss": 6.49,
      "mean_token_accuracy": 0.1022932469844818,
      "num_tokens": 1692803.0,
      "step": 765
    },
    {
      "entropy": 6.86904993057251,
      "epoch": 0.07641921397379912,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003845,
      "loss": 6.5127,
      "mean_token_accuracy": 0.10286033898591995,
      "num_tokens": 1703919.0,
      "step": 770
    },
    {
      "entropy": 6.803786706924439,
      "epoch": 0.07691544263596665,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.4902,
      "mean_token_accuracy": 0.11186280027031899,
      "num_tokens": 1714487.0,
      "step": 775
    },
    {
      "entropy": 6.829148006439209,
      "epoch": 0.07741167129813419,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.5375,
      "mean_token_accuracy": 0.103410904109478,
      "num_tokens": 1726601.0,
      "step": 780
    },
    {
      "entropy": 6.77627649307251,
      "epoch": 0.0779078999603017,
      "grad_norm": 0.96875,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.4293,
      "mean_token_accuracy": 0.10090705007314682,
      "num_tokens": 1738743.0,
      "step": 785
    },
    {
      "entropy": 6.75617151260376,
      "epoch": 0.07840412862246923,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.509,
      "mean_token_accuracy": 0.10762425512075424,
      "num_tokens": 1748286.0,
      "step": 790
    },
    {
      "entropy": 6.78504114151001,
      "epoch": 0.07890035728463676,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.5497,
      "mean_token_accuracy": 0.10279078707098961,
      "num_tokens": 1758917.0,
      "step": 795
    },
    {
      "entropy": 6.834147977828979,
      "epoch": 0.0793965859468043,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003995,
      "loss": 6.3768,
      "mean_token_accuracy": 0.11014361456036567,
      "num_tokens": 1769179.0,
      "step": 800
    },
    {
      "entropy": 6.686023187637329,
      "epoch": 0.07989281460897181,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000402,
      "loss": 6.5004,
      "mean_token_accuracy": 0.09990919530391693,
      "num_tokens": 1782369.0,
      "step": 805
    },
    {
      "entropy": 6.789707326889038,
      "epoch": 0.08038904327113934,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004045,
      "loss": 6.5087,
      "mean_token_accuracy": 0.11087186112999917,
      "num_tokens": 1795992.0,
      "step": 810
    },
    {
      "entropy": 6.767456674575806,
      "epoch": 0.08088527193330687,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.4063,
      "mean_token_accuracy": 0.11133067905902863,
      "num_tokens": 1806938.0,
      "step": 815
    },
    {
      "entropy": 6.6453076839447025,
      "epoch": 0.0813815005954744,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004095,
      "loss": 6.3335,
      "mean_token_accuracy": 0.1083066463470459,
      "num_tokens": 1817663.0,
      "step": 820
    },
    {
      "entropy": 6.804880285263062,
      "epoch": 0.08187772925764192,
      "grad_norm": 0.984375,
      "learning_rate": 0.000412,
      "loss": 6.4937,
      "mean_token_accuracy": 0.10573754906654358,
      "num_tokens": 1828254.0,
      "step": 825
    },
    {
      "entropy": 6.744491100311279,
      "epoch": 0.08237395791980945,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004145,
      "loss": 6.4121,
      "mean_token_accuracy": 0.10808726996183396,
      "num_tokens": 1839001.0,
      "step": 830
    },
    {
      "entropy": 6.780965948104859,
      "epoch": 0.08287018658197698,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000417,
      "loss": 6.5072,
      "mean_token_accuracy": 0.1064543254673481,
      "num_tokens": 1849906.0,
      "step": 835
    },
    {
      "entropy": 6.638266897201538,
      "epoch": 0.0833664152441445,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004195,
      "loss": 6.4146,
      "mean_token_accuracy": 0.10410640761256218,
      "num_tokens": 1860982.0,
      "step": 840
    },
    {
      "entropy": 6.7845916748046875,
      "epoch": 0.08386264390631203,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000422,
      "loss": 6.3668,
      "mean_token_accuracy": 0.10832294374704361,
      "num_tokens": 1872177.0,
      "step": 845
    },
    {
      "entropy": 6.6455484390258786,
      "epoch": 0.08435887256847956,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004245,
      "loss": 6.345,
      "mean_token_accuracy": 0.10717036500573159,
      "num_tokens": 1882783.0,
      "step": 850
    },
    {
      "entropy": 6.82792387008667,
      "epoch": 0.08485510123064709,
      "grad_norm": 1.09375,
      "learning_rate": 0.000427,
      "loss": 6.521,
      "mean_token_accuracy": 0.0995104432106018,
      "num_tokens": 1893706.0,
      "step": 855
    },
    {
      "entropy": 6.6573139190673825,
      "epoch": 0.0853513298928146,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004295,
      "loss": 6.399,
      "mean_token_accuracy": 0.10572489872574806,
      "num_tokens": 1903852.0,
      "step": 860
    },
    {
      "entropy": 6.720736408233643,
      "epoch": 0.08584755855498213,
      "grad_norm": 1.0625,
      "learning_rate": 0.000432,
      "loss": 6.4272,
      "mean_token_accuracy": 0.10979009121656418,
      "num_tokens": 1914157.0,
      "step": 865
    },
    {
      "entropy": 6.719715213775634,
      "epoch": 0.08634378721714966,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004345,
      "loss": 6.3646,
      "mean_token_accuracy": 0.10477298423647881,
      "num_tokens": 1924684.0,
      "step": 870
    },
    {
      "entropy": 6.765842390060425,
      "epoch": 0.0868400158793172,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000437,
      "loss": 6.5126,
      "mean_token_accuracy": 0.1022205650806427,
      "num_tokens": 1935657.0,
      "step": 875
    },
    {
      "entropy": 6.674496650695801,
      "epoch": 0.08733624454148471,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004395,
      "loss": 6.3866,
      "mean_token_accuracy": 0.10951956436038017,
      "num_tokens": 1946735.0,
      "step": 880
    },
    {
      "entropy": 6.771207475662232,
      "epoch": 0.08783247320365224,
      "grad_norm": 1.046875,
      "learning_rate": 0.000442,
      "loss": 6.558,
      "mean_token_accuracy": 0.10081006437540055,
      "num_tokens": 1957871.0,
      "step": 885
    },
    {
      "entropy": 6.6469238758087155,
      "epoch": 0.08832870186581977,
      "grad_norm": 1.125,
      "learning_rate": 0.0004445,
      "loss": 6.4001,
      "mean_token_accuracy": 0.10841071009635925,
      "num_tokens": 1969032.0,
      "step": 890
    },
    {
      "entropy": 6.773042964935303,
      "epoch": 0.0888249305279873,
      "grad_norm": 1.046875,
      "learning_rate": 0.000447,
      "loss": 6.4253,
      "mean_token_accuracy": 0.10482999682426453,
      "num_tokens": 1980212.0,
      "step": 895
    },
    {
      "entropy": 6.5782490253448485,
      "epoch": 0.08932115919015482,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.3457,
      "mean_token_accuracy": 0.1083353117108345,
      "num_tokens": 1990333.0,
      "step": 900
    },
    {
      "entropy": 6.723943090438842,
      "epoch": 0.08981738785232235,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.4208,
      "mean_token_accuracy": 0.11527454107999802,
      "num_tokens": 2001486.0,
      "step": 905
    },
    {
      "entropy": 6.647228240966797,
      "epoch": 0.09031361651448988,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.4152,
      "mean_token_accuracy": 0.10908102542161942,
      "num_tokens": 2012535.0,
      "step": 910
    },
    {
      "entropy": 6.65789384841919,
      "epoch": 0.0908098451766574,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.3697,
      "mean_token_accuracy": 0.11192298531532288,
      "num_tokens": 2024116.0,
      "step": 915
    },
    {
      "entropy": 6.552388858795166,
      "epoch": 0.09130607383882493,
      "grad_norm": 0.82421875,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.4129,
      "mean_token_accuracy": 0.10573532730340958,
      "num_tokens": 2036341.0,
      "step": 920
    },
    {
      "entropy": 6.772451591491699,
      "epoch": 0.09180230250099246,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000462,
      "loss": 6.3831,
      "mean_token_accuracy": 0.11066723242402077,
      "num_tokens": 2046768.0,
      "step": 925
    },
    {
      "entropy": 6.601136875152588,
      "epoch": 0.09229853116315999,
      "grad_norm": 1.0,
      "learning_rate": 0.0004645,
      "loss": 6.3447,
      "mean_token_accuracy": 0.10959907621145248,
      "num_tokens": 2057315.0,
      "step": 930
    },
    {
      "entropy": 6.595295667648315,
      "epoch": 0.0927947598253275,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000467,
      "loss": 6.3258,
      "mean_token_accuracy": 0.1111043132841587,
      "num_tokens": 2068267.0,
      "step": 935
    },
    {
      "entropy": 6.6356977939605715,
      "epoch": 0.09329098848749504,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004695,
      "loss": 6.4018,
      "mean_token_accuracy": 0.10635230094194412,
      "num_tokens": 2078271.0,
      "step": 940
    },
    {
      "entropy": 6.570490837097168,
      "epoch": 0.09378721714966257,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000472,
      "loss": 6.3186,
      "mean_token_accuracy": 0.10399494022130966,
      "num_tokens": 2088381.0,
      "step": 945
    },
    {
      "entropy": 6.635789442062378,
      "epoch": 0.0942834458118301,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004745,
      "loss": 6.2807,
      "mean_token_accuracy": 0.10803479552268982,
      "num_tokens": 2098899.0,
      "step": 950
    },
    {
      "entropy": 6.580207061767578,
      "epoch": 0.09477967447399761,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000477,
      "loss": 6.4469,
      "mean_token_accuracy": 0.1090017631649971,
      "num_tokens": 2110057.0,
      "step": 955
    },
    {
      "entropy": 6.590322017669678,
      "epoch": 0.09527590313616514,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004795,
      "loss": 6.3534,
      "mean_token_accuracy": 0.10890609472990036,
      "num_tokens": 2121758.0,
      "step": 960
    },
    {
      "entropy": 6.626723718643189,
      "epoch": 0.09577213179833267,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000482,
      "loss": 6.3606,
      "mean_token_accuracy": 0.11137763261795045,
      "num_tokens": 2131840.0,
      "step": 965
    },
    {
      "entropy": 6.4795605659484865,
      "epoch": 0.0962683604605002,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004845,
      "loss": 6.3243,
      "mean_token_accuracy": 0.11041747704148293,
      "num_tokens": 2142084.0,
      "step": 970
    },
    {
      "entropy": 6.671319103240966,
      "epoch": 0.09676458912266772,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000487,
      "loss": 6.3937,
      "mean_token_accuracy": 0.10798183381557465,
      "num_tokens": 2152785.0,
      "step": 975
    },
    {
      "entropy": 6.661399173736572,
      "epoch": 0.09726081778483525,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004895,
      "loss": 6.3784,
      "mean_token_accuracy": 0.109221463650465,
      "num_tokens": 2163774.0,
      "step": 980
    },
    {
      "entropy": 6.505852460861206,
      "epoch": 0.09775704644700278,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000492,
      "loss": 6.2916,
      "mean_token_accuracy": 0.1088893674314022,
      "num_tokens": 2175378.0,
      "step": 985
    },
    {
      "entropy": 6.755297660827637,
      "epoch": 0.0982532751091703,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004945,
      "loss": 6.5117,
      "mean_token_accuracy": 0.1058343231678009,
      "num_tokens": 2186510.0,
      "step": 990
    },
    {
      "entropy": 6.522902965545654,
      "epoch": 0.09874950377133783,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000497,
      "loss": 6.2971,
      "mean_token_accuracy": 0.11370693743228913,
      "num_tokens": 2197482.0,
      "step": 995
    },
    {
      "entropy": 6.5511187553405765,
      "epoch": 0.09924573243350536,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004995,
      "loss": 6.3049,
      "mean_token_accuracy": 0.10995083898305893,
      "num_tokens": 2208329.0,
      "step": 1000
    },
    {
      "entropy": 6.5022056102752686,
      "epoch": 0.09974196109567289,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999999982145551,
      "loss": 6.2196,
      "mean_token_accuracy": 0.10750824585556984,
      "num_tokens": 2218632.0,
      "step": 1005
    },
    {
      "entropy": 6.522266244888305,
      "epoch": 0.1002381897578404,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999999909611855,
      "loss": 6.3288,
      "mean_token_accuracy": 0.10453014001250267,
      "num_tokens": 2231335.0,
      "step": 1010
    },
    {
      "entropy": 6.613006448745727,
      "epoch": 0.10073441842000794,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999999781283008,
      "loss": 6.3162,
      "mean_token_accuracy": 0.11068271622061729,
      "num_tokens": 2241704.0,
      "step": 1015
    },
    {
      "entropy": 6.541505002975464,
      "epoch": 0.10123064708217547,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999999597159017,
      "loss": 6.3683,
      "mean_token_accuracy": 0.10505040511488914,
      "num_tokens": 2252310.0,
      "step": 1020
    },
    {
      "entropy": 6.5655745506286625,
      "epoch": 0.101726875744343,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999999357239882,
      "loss": 6.289,
      "mean_token_accuracy": 0.10687798410654067,
      "num_tokens": 2262270.0,
      "step": 1025
    },
    {
      "entropy": 6.599134349822998,
      "epoch": 0.10222310440651051,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999999061525612,
      "loss": 6.4258,
      "mean_token_accuracy": 0.10728801861405372,
      "num_tokens": 2273986.0,
      "step": 1030
    },
    {
      "entropy": 6.644435930252075,
      "epoch": 0.10271933306867805,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999998710016214,
      "loss": 6.3152,
      "mean_token_accuracy": 0.10951637625694274,
      "num_tokens": 2283430.0,
      "step": 1035
    },
    {
      "entropy": 6.572604465484619,
      "epoch": 0.10321556173084558,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999998302711696,
      "loss": 6.3347,
      "mean_token_accuracy": 0.10473796129226684,
      "num_tokens": 2294201.0,
      "step": 1040
    },
    {
      "entropy": 6.451688957214356,
      "epoch": 0.1037117903930131,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999997839612069,
      "loss": 6.2316,
      "mean_token_accuracy": 0.10967583358287811,
      "num_tokens": 2303876.0,
      "step": 1045
    },
    {
      "entropy": 6.53422532081604,
      "epoch": 0.10420801905518062,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999997320717342,
      "loss": 6.2748,
      "mean_token_accuracy": 0.11531499549746513,
      "num_tokens": 2313713.0,
      "step": 1050
    },
    {
      "entropy": 6.462176990509033,
      "epoch": 0.10470424771734815,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999996746027531,
      "loss": 6.281,
      "mean_token_accuracy": 0.1057841993868351,
      "num_tokens": 2324862.0,
      "step": 1055
    },
    {
      "entropy": 6.63836817741394,
      "epoch": 0.10520047637951568,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499999611554265,
      "loss": 6.342,
      "mean_token_accuracy": 0.11243951395153999,
      "num_tokens": 2334974.0,
      "step": 1060
    },
    {
      "entropy": 6.527207708358764,
      "epoch": 0.10569670504168321,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999995429262712,
      "loss": 6.3814,
      "mean_token_accuracy": 0.11008421406149864,
      "num_tokens": 2345945.0,
      "step": 1065
    },
    {
      "entropy": 6.573855066299439,
      "epoch": 0.10619293370385073,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999994687187737,
      "loss": 6.3162,
      "mean_token_accuracy": 0.10868648067116737,
      "num_tokens": 2356749.0,
      "step": 1070
    },
    {
      "entropy": 6.601923894882202,
      "epoch": 0.10668916236601826,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999993889317742,
      "loss": 6.3894,
      "mean_token_accuracy": 0.10912014171481133,
      "num_tokens": 2368181.0,
      "step": 1075
    },
    {
      "entropy": 6.519433784484863,
      "epoch": 0.10718539102818579,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999993035652746,
      "loss": 6.2988,
      "mean_token_accuracy": 0.10856023505330085,
      "num_tokens": 2377737.0,
      "step": 1080
    },
    {
      "entropy": 6.498434209823609,
      "epoch": 0.10768161969035331,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999992126192771,
      "loss": 6.2253,
      "mean_token_accuracy": 0.11343928501009941,
      "num_tokens": 2388944.0,
      "step": 1085
    },
    {
      "entropy": 6.506166744232178,
      "epoch": 0.10817784835252084,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499999116093784,
      "loss": 6.3021,
      "mean_token_accuracy": 0.11269065588712693,
      "num_tokens": 2400758.0,
      "step": 1090
    },
    {
      "entropy": 6.475096273422241,
      "epoch": 0.10867407701468837,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999990139887976,
      "loss": 6.2988,
      "mean_token_accuracy": 0.11071598455309868,
      "num_tokens": 2411916.0,
      "step": 1095
    },
    {
      "entropy": 6.6090761661529545,
      "epoch": 0.1091703056768559,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999989063043206,
      "loss": 6.3162,
      "mean_token_accuracy": 0.10899331420660019,
      "num_tokens": 2423310.0,
      "step": 1100
    },
    {
      "entropy": 6.4430852890014645,
      "epoch": 0.10966653433902342,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999987930403555,
      "loss": 6.255,
      "mean_token_accuracy": 0.11132511496543884,
      "num_tokens": 2432100.0,
      "step": 1105
    },
    {
      "entropy": 6.497279357910156,
      "epoch": 0.11016276300119095,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999986741969051,
      "loss": 6.3283,
      "mean_token_accuracy": 0.10885677859187126,
      "num_tokens": 2443563.0,
      "step": 1110
    },
    {
      "entropy": 6.568647575378418,
      "epoch": 0.11065899166335848,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999985497739724,
      "loss": 6.3328,
      "mean_token_accuracy": 0.10904523879289627,
      "num_tokens": 2455435.0,
      "step": 1115
    },
    {
      "entropy": 6.526462554931641,
      "epoch": 0.11115522032552601,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999984197715606,
      "loss": 6.2693,
      "mean_token_accuracy": 0.11513903066515922,
      "num_tokens": 2466173.0,
      "step": 1120
    },
    {
      "entropy": 6.406380701065063,
      "epoch": 0.11165144898769352,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999982841896728,
      "loss": 6.2217,
      "mean_token_accuracy": 0.11128725931048393,
      "num_tokens": 2476349.0,
      "step": 1125
    },
    {
      "entropy": 6.5459887981414795,
      "epoch": 0.11214767764986106,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999981430283123,
      "loss": 6.3545,
      "mean_token_accuracy": 0.11260213553905488,
      "num_tokens": 2487864.0,
      "step": 1130
    },
    {
      "entropy": 6.459234380722046,
      "epoch": 0.11264390631202859,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999979962874827,
      "loss": 6.2778,
      "mean_token_accuracy": 0.11224315464496612,
      "num_tokens": 2499188.0,
      "step": 1135
    },
    {
      "entropy": 6.541913795471191,
      "epoch": 0.11314013497419612,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999978439671877,
      "loss": 6.2679,
      "mean_token_accuracy": 0.11238265410065651,
      "num_tokens": 2509813.0,
      "step": 1140
    },
    {
      "entropy": 6.452944564819336,
      "epoch": 0.11363636363636363,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999976860674309,
      "loss": 6.2954,
      "mean_token_accuracy": 0.1079881377518177,
      "num_tokens": 2520813.0,
      "step": 1145
    },
    {
      "entropy": 6.451686000823974,
      "epoch": 0.11413259229853116,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999975225882164,
      "loss": 6.2857,
      "mean_token_accuracy": 0.11005399227142335,
      "num_tokens": 2531802.0,
      "step": 1150
    },
    {
      "entropy": 6.419873952865601,
      "epoch": 0.1146288209606987,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000499997353529548,
      "loss": 6.0989,
      "mean_token_accuracy": 0.12534342482686042,
      "num_tokens": 2541935.0,
      "step": 1155
    },
    {
      "entropy": 6.41290922164917,
      "epoch": 0.11512504962286621,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999971788914303,
      "loss": 6.209,
      "mean_token_accuracy": 0.11427329108119011,
      "num_tokens": 2553224.0,
      "step": 1160
    },
    {
      "entropy": 6.4005992889404295,
      "epoch": 0.11562127828503374,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999969986738672,
      "loss": 6.2408,
      "mean_token_accuracy": 0.1071538008749485,
      "num_tokens": 2565355.0,
      "step": 1165
    },
    {
      "entropy": 6.517459058761597,
      "epoch": 0.11611750694720127,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999968128768634,
      "loss": 6.3245,
      "mean_token_accuracy": 0.1080057941377163,
      "num_tokens": 2575962.0,
      "step": 1170
    },
    {
      "entropy": 6.512309551239014,
      "epoch": 0.1166137356093688,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999966215004236,
      "loss": 6.3356,
      "mean_token_accuracy": 0.11455044671893119,
      "num_tokens": 2588136.0,
      "step": 1175
    },
    {
      "entropy": 6.473892450332642,
      "epoch": 0.11710996427153632,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999964245445522,
      "loss": 6.2264,
      "mean_token_accuracy": 0.11278848052024841,
      "num_tokens": 2597785.0,
      "step": 1180
    },
    {
      "entropy": 6.45322551727295,
      "epoch": 0.11760619293370385,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999962220092545,
      "loss": 6.2625,
      "mean_token_accuracy": 0.12006586417555809,
      "num_tokens": 2609738.0,
      "step": 1185
    },
    {
      "entropy": 6.329366683959961,
      "epoch": 0.11810242159587138,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999960138945351,
      "loss": 6.1518,
      "mean_token_accuracy": 0.12184964567422867,
      "num_tokens": 2620179.0,
      "step": 1190
    },
    {
      "entropy": 6.402722692489624,
      "epoch": 0.11859865025803891,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999958002003996,
      "loss": 6.1977,
      "mean_token_accuracy": 0.1172001026570797,
      "num_tokens": 2631550.0,
      "step": 1195
    },
    {
      "entropy": 6.43868670463562,
      "epoch": 0.11909487892020643,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004999955809268529,
      "loss": 6.2334,
      "mean_token_accuracy": 0.1126227967441082,
      "num_tokens": 2642467.0,
      "step": 1200
    },
    {
      "entropy": 6.404077625274658,
      "epoch": 0.11959110758237396,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999953560739008,
      "loss": 6.2134,
      "mean_token_accuracy": 0.11190939992666245,
      "num_tokens": 2652132.0,
      "step": 1205
    },
    {
      "entropy": 6.447581481933594,
      "epoch": 0.12008733624454149,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004999951256415485,
      "loss": 6.2707,
      "mean_token_accuracy": 0.11127389594912529,
      "num_tokens": 2663336.0,
      "step": 1210
    },
    {
      "entropy": 6.5261087894439695,
      "epoch": 0.12058356490670902,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499994889629802,
      "loss": 6.3448,
      "mean_token_accuracy": 0.10134708434343338,
      "num_tokens": 2674201.0,
      "step": 1215
    },
    {
      "entropy": 6.487839126586914,
      "epoch": 0.12107979356887653,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499994648038667,
      "loss": 6.2708,
      "mean_token_accuracy": 0.11672815084457397,
      "num_tokens": 2685073.0,
      "step": 1220
    },
    {
      "entropy": 6.450255823135376,
      "epoch": 0.12157602223104406,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999944008681496,
      "loss": 6.1999,
      "mean_token_accuracy": 0.11382194608449936,
      "num_tokens": 2695024.0,
      "step": 1225
    },
    {
      "entropy": 6.476427698135376,
      "epoch": 0.1220722508932116,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999941481182559,
      "loss": 6.2898,
      "mean_token_accuracy": 0.10765945836901665,
      "num_tokens": 2706232.0,
      "step": 1230
    },
    {
      "entropy": 6.459441423416138,
      "epoch": 0.12256847955537911,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999938897889921,
      "loss": 6.2763,
      "mean_token_accuracy": 0.11456048339605332,
      "num_tokens": 2715874.0,
      "step": 1235
    },
    {
      "entropy": 6.38568229675293,
      "epoch": 0.12306470821754664,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999936258803647,
      "loss": 6.2288,
      "mean_token_accuracy": 0.11625665873289108,
      "num_tokens": 2727409.0,
      "step": 1240
    },
    {
      "entropy": 6.370026016235352,
      "epoch": 0.12356093687971417,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999933563923802,
      "loss": 6.1212,
      "mean_token_accuracy": 0.11923045217990876,
      "num_tokens": 2737259.0,
      "step": 1245
    },
    {
      "entropy": 6.439346504211426,
      "epoch": 0.1240571655418817,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999930813250451,
      "loss": 6.2589,
      "mean_token_accuracy": 0.11177127435803413,
      "num_tokens": 2748416.0,
      "step": 1250
    },
    {
      "entropy": 6.4149510860443115,
      "epoch": 0.12455339420404922,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999928006783667,
      "loss": 6.234,
      "mean_token_accuracy": 0.11361540555953979,
      "num_tokens": 2760470.0,
      "step": 1255
    },
    {
      "entropy": 6.43356351852417,
      "epoch": 0.12504962286621676,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999925144523515,
      "loss": 6.2615,
      "mean_token_accuracy": 0.11003044992685318,
      "num_tokens": 2770222.0,
      "step": 1260
    },
    {
      "entropy": 6.444881439208984,
      "epoch": 0.12554585152838427,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999922226470066,
      "loss": 6.1815,
      "mean_token_accuracy": 0.1138396754860878,
      "num_tokens": 2780212.0,
      "step": 1265
    },
    {
      "entropy": 6.312497234344482,
      "epoch": 0.1260420801905518,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999919252623397,
      "loss": 6.1511,
      "mean_token_accuracy": 0.11787962689995765,
      "num_tokens": 2790239.0,
      "step": 1270
    },
    {
      "entropy": 6.418433570861817,
      "epoch": 0.12653830885271933,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999916222983577,
      "loss": 6.1859,
      "mean_token_accuracy": 0.11903090998530388,
      "num_tokens": 2801902.0,
      "step": 1275
    },
    {
      "entropy": 6.350197839736938,
      "epoch": 0.12703453751488686,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999913137550682,
      "loss": 6.1833,
      "mean_token_accuracy": 0.11472797095775604,
      "num_tokens": 2813268.0,
      "step": 1280
    },
    {
      "entropy": 6.413725900650024,
      "epoch": 0.1275307661770544,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004999909996324791,
      "loss": 6.2532,
      "mean_token_accuracy": 0.11241401582956315,
      "num_tokens": 2824506.0,
      "step": 1285
    },
    {
      "entropy": 6.40827522277832,
      "epoch": 0.12802699483922192,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499990679930598,
      "loss": 6.1575,
      "mean_token_accuracy": 0.11616934686899186,
      "num_tokens": 2835144.0,
      "step": 1290
    },
    {
      "entropy": 6.333975553512573,
      "epoch": 0.12852322350138945,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999903546494327,
      "loss": 6.1793,
      "mean_token_accuracy": 0.11346072256565094,
      "num_tokens": 2845382.0,
      "step": 1295
    },
    {
      "entropy": 6.406157112121582,
      "epoch": 0.12901945216355698,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999900237889916,
      "loss": 6.1977,
      "mean_token_accuracy": 0.11914140731096268,
      "num_tokens": 2854697.0,
      "step": 1300
    },
    {
      "entropy": 6.399594116210937,
      "epoch": 0.12951568082572448,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999896873492827,
      "loss": 6.1988,
      "mean_token_accuracy": 0.11011647656559945,
      "num_tokens": 2866216.0,
      "step": 1305
    },
    {
      "entropy": 6.361563110351563,
      "epoch": 0.130011909487892,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999893453303143,
      "loss": 6.1669,
      "mean_token_accuracy": 0.11769109070301056,
      "num_tokens": 2877734.0,
      "step": 1310
    },
    {
      "entropy": 6.433618021011353,
      "epoch": 0.13050813815005954,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999889977320949,
      "loss": 6.2305,
      "mean_token_accuracy": 0.11750640273094178,
      "num_tokens": 2889445.0,
      "step": 1315
    },
    {
      "entropy": 6.24367208480835,
      "epoch": 0.13100436681222707,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999886445546332,
      "loss": 6.1673,
      "mean_token_accuracy": 0.1189102478325367,
      "num_tokens": 2900485.0,
      "step": 1320
    },
    {
      "entropy": 6.488619422912597,
      "epoch": 0.1315005954743946,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499988285797938,
      "loss": 6.1406,
      "mean_token_accuracy": 0.12021579965949059,
      "num_tokens": 2909795.0,
      "step": 1325
    },
    {
      "entropy": 6.218992137908936,
      "epoch": 0.13199682413656214,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499987921462018,
      "loss": 6.1739,
      "mean_token_accuracy": 0.12471199110150337,
      "num_tokens": 2920153.0,
      "step": 1330
    },
    {
      "entropy": 6.459899282455444,
      "epoch": 0.13249305279872967,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999875515468825,
      "loss": 6.162,
      "mean_token_accuracy": 0.11742881909012795,
      "num_tokens": 2930457.0,
      "step": 1335
    },
    {
      "entropy": 6.303973865509033,
      "epoch": 0.13298928146089717,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999871760525403,
      "loss": 6.2265,
      "mean_token_accuracy": 0.11418904587626458,
      "num_tokens": 2941305.0,
      "step": 1340
    },
    {
      "entropy": 6.43122968673706,
      "epoch": 0.1334855101230647,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999867949790011,
      "loss": 6.1693,
      "mean_token_accuracy": 0.11784612089395523,
      "num_tokens": 2951798.0,
      "step": 1345
    },
    {
      "entropy": 6.290275287628174,
      "epoch": 0.13398173878523223,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999864083262743,
      "loss": 6.067,
      "mean_token_accuracy": 0.12247526496648789,
      "num_tokens": 2962486.0,
      "step": 1350
    },
    {
      "entropy": 6.344256925582886,
      "epoch": 0.13447796744739976,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999860160943691,
      "loss": 6.1817,
      "mean_token_accuracy": 0.11753400415182114,
      "num_tokens": 2974317.0,
      "step": 1355
    },
    {
      "entropy": 6.320338010787964,
      "epoch": 0.1349741961095673,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999856182832958,
      "loss": 6.0902,
      "mean_token_accuracy": 0.12005901709198952,
      "num_tokens": 2985440.0,
      "step": 1360
    },
    {
      "entropy": 6.327242660522461,
      "epoch": 0.13547042477173482,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999852148930637,
      "loss": 6.1637,
      "mean_token_accuracy": 0.11592641919851303,
      "num_tokens": 2997044.0,
      "step": 1365
    },
    {
      "entropy": 6.361653184890747,
      "epoch": 0.13596665343390235,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999848059236832,
      "loss": 6.1792,
      "mean_token_accuracy": 0.11417617872357369,
      "num_tokens": 3005965.0,
      "step": 1370
    },
    {
      "entropy": 6.298781776428223,
      "epoch": 0.13646288209606988,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999843913751643,
      "loss": 6.0483,
      "mean_token_accuracy": 0.11773452535271645,
      "num_tokens": 3015031.0,
      "step": 1375
    },
    {
      "entropy": 6.287301540374756,
      "epoch": 0.13695911075823738,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999839712475174,
      "loss": 6.1151,
      "mean_token_accuracy": 0.1147402547299862,
      "num_tokens": 3026275.0,
      "step": 1380
    },
    {
      "entropy": 6.302942562103271,
      "epoch": 0.13745533942040492,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999835455407526,
      "loss": 6.1485,
      "mean_token_accuracy": 0.11648812666535377,
      "num_tokens": 3036778.0,
      "step": 1385
    },
    {
      "entropy": 6.41059980392456,
      "epoch": 0.13795156808257245,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999831142548808,
      "loss": 6.2536,
      "mean_token_accuracy": 0.11463667005300522,
      "num_tokens": 3048149.0,
      "step": 1390
    },
    {
      "entropy": 6.343495893478393,
      "epoch": 0.13844779674473998,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999826773899124,
      "loss": 6.1206,
      "mean_token_accuracy": 0.11659936383366584,
      "num_tokens": 3059149.0,
      "step": 1395
    },
    {
      "entropy": 6.2959178447723385,
      "epoch": 0.1389440254069075,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004999822349458586,
      "loss": 6.1424,
      "mean_token_accuracy": 0.12055872678756714,
      "num_tokens": 3071613.0,
      "step": 1400
    },
    {
      "entropy": 6.433216142654419,
      "epoch": 0.13944025406907504,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999817869227301,
      "loss": 6.2831,
      "mean_token_accuracy": 0.11233281642198563,
      "num_tokens": 3084742.0,
      "step": 1405
    },
    {
      "entropy": 6.355742263793945,
      "epoch": 0.13993648273124257,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999813333205381,
      "loss": 6.1927,
      "mean_token_accuracy": 0.11989191100001335,
      "num_tokens": 3095356.0,
      "step": 1410
    },
    {
      "entropy": 6.345205307006836,
      "epoch": 0.14043271139341007,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999808741392937,
      "loss": 6.1828,
      "mean_token_accuracy": 0.11414567157626151,
      "num_tokens": 3107717.0,
      "step": 1415
    },
    {
      "entropy": 6.309127426147461,
      "epoch": 0.1409289400555776,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999804093790086,
      "loss": 6.1671,
      "mean_token_accuracy": 0.11339711248874665,
      "num_tokens": 3119205.0,
      "step": 1420
    },
    {
      "entropy": 6.414076328277588,
      "epoch": 0.14142516871774513,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000499979939039694,
      "loss": 6.2164,
      "mean_token_accuracy": 0.11629738509655,
      "num_tokens": 3130564.0,
      "step": 1425
    },
    {
      "entropy": 6.257117033004761,
      "epoch": 0.14192139737991266,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004999794631213618,
      "loss": 6.1496,
      "mean_token_accuracy": 0.11587404683232308,
      "num_tokens": 3141883.0,
      "step": 1430
    },
    {
      "entropy": 6.4088245868682865,
      "epoch": 0.1424176260420802,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999789816240236,
      "loss": 6.1524,
      "mean_token_accuracy": 0.11169286966323852,
      "num_tokens": 3152817.0,
      "step": 1435
    },
    {
      "entropy": 6.226463317871094,
      "epoch": 0.14291385470424772,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999784945476915,
      "loss": 6.0639,
      "mean_token_accuracy": 0.12211388051509857,
      "num_tokens": 3163546.0,
      "step": 1440
    },
    {
      "entropy": 6.368065643310547,
      "epoch": 0.14341008336641525,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999780018923775,
      "loss": 6.1194,
      "mean_token_accuracy": 0.11407855600118637,
      "num_tokens": 3173899.0,
      "step": 1445
    },
    {
      "entropy": 6.161925745010376,
      "epoch": 0.14390631202858278,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999775036580939,
      "loss": 6.0337,
      "mean_token_accuracy": 0.11928667798638344,
      "num_tokens": 3184433.0,
      "step": 1450
    },
    {
      "entropy": 6.412479734420776,
      "epoch": 0.1444025406907503,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999769998448529,
      "loss": 6.1516,
      "mean_token_accuracy": 0.11415581554174423,
      "num_tokens": 3196077.0,
      "step": 1455
    },
    {
      "entropy": 6.141682243347168,
      "epoch": 0.14489876935291782,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999764904526672,
      "loss": 6.0621,
      "mean_token_accuracy": 0.12485457435250283,
      "num_tokens": 3206645.0,
      "step": 1460
    },
    {
      "entropy": 6.4062056064605715,
      "epoch": 0.14539499801508535,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999759754815492,
      "loss": 6.185,
      "mean_token_accuracy": 0.12362410277128219,
      "num_tokens": 3216786.0,
      "step": 1465
    },
    {
      "entropy": 6.270987462997437,
      "epoch": 0.14589122667725288,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999754549315118,
      "loss": 6.1125,
      "mean_token_accuracy": 0.1173293948173523,
      "num_tokens": 3227384.0,
      "step": 1470
    },
    {
      "entropy": 6.215911531448365,
      "epoch": 0.1463874553394204,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499974928802568,
      "loss": 5.9691,
      "mean_token_accuracy": 0.12435413226485252,
      "num_tokens": 3238800.0,
      "step": 1475
    },
    {
      "entropy": 6.228160572052002,
      "epoch": 0.14688368400158794,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999743970947306,
      "loss": 6.1209,
      "mean_token_accuracy": 0.11474767550826073,
      "num_tokens": 3248627.0,
      "step": 1480
    },
    {
      "entropy": 6.310603427886963,
      "epoch": 0.14737991266375547,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499973859808013,
      "loss": 6.1474,
      "mean_token_accuracy": 0.11879593282938003,
      "num_tokens": 3258989.0,
      "step": 1485
    },
    {
      "entropy": 6.224552726745605,
      "epoch": 0.14787614132592297,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999733169424284,
      "loss": 6.1196,
      "mean_token_accuracy": 0.12131216377019882,
      "num_tokens": 3269474.0,
      "step": 1490
    },
    {
      "entropy": 6.298264455795288,
      "epoch": 0.1483723699880905,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999727684979904,
      "loss": 6.1048,
      "mean_token_accuracy": 0.11965034380555153,
      "num_tokens": 3280453.0,
      "step": 1495
    },
    {
      "entropy": 6.228455209732056,
      "epoch": 0.14886859865025803,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999722144747125,
      "loss": 6.0331,
      "mean_token_accuracy": 0.11981930211186409,
      "num_tokens": 3290234.0,
      "step": 1500
    },
    {
      "entropy": 6.207645463943481,
      "epoch": 0.14936482731242556,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999716548726085,
      "loss": 6.0146,
      "mean_token_accuracy": 0.12176985815167427,
      "num_tokens": 3300943.0,
      "step": 1505
    },
    {
      "entropy": 6.184590911865234,
      "epoch": 0.1498610559745931,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999710896916921,
      "loss": 5.9902,
      "mean_token_accuracy": 0.12336606532335281,
      "num_tokens": 3311296.0,
      "step": 1510
    },
    {
      "entropy": 6.256126356124878,
      "epoch": 0.15035728463676062,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999705189319774,
      "loss": 6.1279,
      "mean_token_accuracy": 0.11304511949419975,
      "num_tokens": 3322310.0,
      "step": 1515
    },
    {
      "entropy": 6.240859413146973,
      "epoch": 0.15085351329892815,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999699425934788,
      "loss": 6.1363,
      "mean_token_accuracy": 0.11859385669231415,
      "num_tokens": 3334176.0,
      "step": 1520
    },
    {
      "entropy": 6.342157602310181,
      "epoch": 0.15134974196109569,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999693606762103,
      "loss": 6.1278,
      "mean_token_accuracy": 0.11730345636606217,
      "num_tokens": 3344839.0,
      "step": 1525
    },
    {
      "entropy": 6.24994010925293,
      "epoch": 0.1518459706232632,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999687731801865,
      "loss": 6.1041,
      "mean_token_accuracy": 0.11314293071627617,
      "num_tokens": 3357046.0,
      "step": 1530
    },
    {
      "entropy": 6.32585072517395,
      "epoch": 0.15234219928543072,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999681801054217,
      "loss": 6.1754,
      "mean_token_accuracy": 0.11708802953362465,
      "num_tokens": 3367783.0,
      "step": 1535
    },
    {
      "entropy": 6.213843870162964,
      "epoch": 0.15283842794759825,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499967581451931,
      "loss": 6.0432,
      "mean_token_accuracy": 0.11833713948726654,
      "num_tokens": 3378529.0,
      "step": 1540
    },
    {
      "entropy": 6.292020654678344,
      "epoch": 0.15333465660976578,
      "grad_norm": 1.078125,
      "learning_rate": 0.000499966977219729,
      "loss": 6.2356,
      "mean_token_accuracy": 0.11526345536112785,
      "num_tokens": 3389318.0,
      "step": 1545
    },
    {
      "entropy": 6.392855215072632,
      "epoch": 0.1538308852719333,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999663674088306,
      "loss": 6.2284,
      "mean_token_accuracy": 0.11469097509980201,
      "num_tokens": 3400603.0,
      "step": 1550
    },
    {
      "entropy": 6.286278486251831,
      "epoch": 0.15432711393410084,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999657520192511,
      "loss": 6.1222,
      "mean_token_accuracy": 0.11655379831790924,
      "num_tokens": 3410316.0,
      "step": 1555
    },
    {
      "entropy": 6.2597517490386965,
      "epoch": 0.15482334259626837,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999651310510058,
      "loss": 6.0634,
      "mean_token_accuracy": 0.12244669795036316,
      "num_tokens": 3422155.0,
      "step": 1560
    },
    {
      "entropy": 6.282853603363037,
      "epoch": 0.15531957125843587,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999645045041099,
      "loss": 6.0028,
      "mean_token_accuracy": 0.1191568560898304,
      "num_tokens": 3432892.0,
      "step": 1565
    },
    {
      "entropy": 6.221517276763916,
      "epoch": 0.1558157999206034,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999638723785791,
      "loss": 6.1304,
      "mean_token_accuracy": 0.12089115679264069,
      "num_tokens": 3443967.0,
      "step": 1570
    },
    {
      "entropy": 6.202426433563232,
      "epoch": 0.15631202858277093,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499963234674429,
      "loss": 6.0213,
      "mean_token_accuracy": 0.12375078648328781,
      "num_tokens": 3454354.0,
      "step": 1575
    },
    {
      "entropy": 6.131307458877563,
      "epoch": 0.15680825724493846,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999625913916754,
      "loss": 6.0405,
      "mean_token_accuracy": 0.11991182640194893,
      "num_tokens": 3465232.0,
      "step": 1580
    },
    {
      "entropy": 6.1873551368713375,
      "epoch": 0.157304485907106,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999619425303343,
      "loss": 6.0184,
      "mean_token_accuracy": 0.12445951774716377,
      "num_tokens": 3475443.0,
      "step": 1585
    },
    {
      "entropy": 6.2718860626220705,
      "epoch": 0.15780071456927353,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999612880904219,
      "loss": 6.0401,
      "mean_token_accuracy": 0.12001754567027093,
      "num_tokens": 3486643.0,
      "step": 1590
    },
    {
      "entropy": 6.218065166473389,
      "epoch": 0.15829694323144106,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999606280719542,
      "loss": 6.1047,
      "mean_token_accuracy": 0.1197525791823864,
      "num_tokens": 3497557.0,
      "step": 1595
    },
    {
      "entropy": 6.363813638687134,
      "epoch": 0.1587931718936086,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999599624749475,
      "loss": 6.1079,
      "mean_token_accuracy": 0.1160752110183239,
      "num_tokens": 3509790.0,
      "step": 1600
    },
    {
      "entropy": 6.141760301589966,
      "epoch": 0.1592894005557761,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999592912994187,
      "loss": 6.0233,
      "mean_token_accuracy": 0.11927468180656434,
      "num_tokens": 3521577.0,
      "step": 1605
    },
    {
      "entropy": 6.263496160507202,
      "epoch": 0.15978562921794362,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999586145453842,
      "loss": 6.0717,
      "mean_token_accuracy": 0.11626604050397873,
      "num_tokens": 3532126.0,
      "step": 1610
    },
    {
      "entropy": 6.084614086151123,
      "epoch": 0.16028185788011115,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999579322128608,
      "loss": 5.9055,
      "mean_token_accuracy": 0.12044439241290092,
      "num_tokens": 3542560.0,
      "step": 1615
    },
    {
      "entropy": 6.201308965682983,
      "epoch": 0.16077808654227868,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999572443018654,
      "loss": 6.065,
      "mean_token_accuracy": 0.11997904106974602,
      "num_tokens": 3553863.0,
      "step": 1620
    },
    {
      "entropy": 6.1602541446685795,
      "epoch": 0.1612743152044462,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999565508124151,
      "loss": 6.0874,
      "mean_token_accuracy": 0.12250801846385002,
      "num_tokens": 3564953.0,
      "step": 1625
    },
    {
      "entropy": 6.266576147079467,
      "epoch": 0.16177054386661374,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000499955851744527,
      "loss": 6.0789,
      "mean_token_accuracy": 0.12218151539564133,
      "num_tokens": 3576247.0,
      "step": 1630
    },
    {
      "entropy": 6.306499862670899,
      "epoch": 0.16226677252878127,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999551470982185,
      "loss": 6.1101,
      "mean_token_accuracy": 0.11723152995109558,
      "num_tokens": 3587600.0,
      "step": 1635
    },
    {
      "entropy": 6.263530349731445,
      "epoch": 0.1627630011909488,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999544368735073,
      "loss": 6.0821,
      "mean_token_accuracy": 0.12516205981373787,
      "num_tokens": 3597727.0,
      "step": 1640
    },
    {
      "entropy": 6.230346632003784,
      "epoch": 0.1632592298531163,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999537210704106,
      "loss": 6.0679,
      "mean_token_accuracy": 0.12371558174490929,
      "num_tokens": 3608066.0,
      "step": 1645
    },
    {
      "entropy": 6.272209596633911,
      "epoch": 0.16375545851528384,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999529996889463,
      "loss": 6.1215,
      "mean_token_accuracy": 0.11630801856517792,
      "num_tokens": 3618263.0,
      "step": 1650
    },
    {
      "entropy": 6.252897644042969,
      "epoch": 0.16425168717745137,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999522727291325,
      "loss": 6.07,
      "mean_token_accuracy": 0.12269503250718117,
      "num_tokens": 3629465.0,
      "step": 1655
    },
    {
      "entropy": 6.205037641525268,
      "epoch": 0.1647479158396189,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499951540190987,
      "loss": 6.0946,
      "mean_token_accuracy": 0.11775854378938674,
      "num_tokens": 3639657.0,
      "step": 1660
    },
    {
      "entropy": 6.27462944984436,
      "epoch": 0.16524414450178643,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999508020745281,
      "loss": 6.0516,
      "mean_token_accuracy": 0.11680529490113259,
      "num_tokens": 3652374.0,
      "step": 1665
    },
    {
      "entropy": 6.219149589538574,
      "epoch": 0.16574037316395396,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999500583797739,
      "loss": 6.0272,
      "mean_token_accuracy": 0.12123104184865952,
      "num_tokens": 3664083.0,
      "step": 1670
    },
    {
      "entropy": 6.1698380470275875,
      "epoch": 0.1662366018261215,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999493091067431,
      "loss": 6.0679,
      "mean_token_accuracy": 0.12645655497908592,
      "num_tokens": 3675314.0,
      "step": 1675
    },
    {
      "entropy": 6.184545946121216,
      "epoch": 0.166732830488289,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004999485542554541,
      "loss": 6.0337,
      "mean_token_accuracy": 0.11284921467304229,
      "num_tokens": 3686019.0,
      "step": 1680
    },
    {
      "entropy": 6.288911151885986,
      "epoch": 0.16722905915045652,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999477938259256,
      "loss": 6.1241,
      "mean_token_accuracy": 0.1167833961546421,
      "num_tokens": 3698170.0,
      "step": 1685
    },
    {
      "entropy": 6.210192346572876,
      "epoch": 0.16772528781262405,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999470278181766,
      "loss": 6.1146,
      "mean_token_accuracy": 0.11569419503211975,
      "num_tokens": 3709243.0,
      "step": 1690
    },
    {
      "entropy": 6.2869891166687015,
      "epoch": 0.16822151647479158,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999462562322261,
      "loss": 6.0797,
      "mean_token_accuracy": 0.12025846838951111,
      "num_tokens": 3720196.0,
      "step": 1695
    },
    {
      "entropy": 6.290680551528931,
      "epoch": 0.1687177451369591,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499945479068093,
      "loss": 6.1355,
      "mean_token_accuracy": 0.12103915438055993,
      "num_tokens": 3732714.0,
      "step": 1700
    },
    {
      "entropy": 6.222851324081421,
      "epoch": 0.16921397379912664,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004999446963257969,
      "loss": 6.0964,
      "mean_token_accuracy": 0.11665565371513367,
      "num_tokens": 3746474.0,
      "step": 1705
    },
    {
      "entropy": 6.142470836639404,
      "epoch": 0.16971020246129417,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999439080053571,
      "loss": 5.9737,
      "mean_token_accuracy": 0.12647735849022865,
      "num_tokens": 3757528.0,
      "step": 1710
    },
    {
      "entropy": 6.312270736694336,
      "epoch": 0.1702064311234617,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999431141067928,
      "loss": 6.1028,
      "mean_token_accuracy": 0.1200647495687008,
      "num_tokens": 3768360.0,
      "step": 1715
    },
    {
      "entropy": 6.174591398239135,
      "epoch": 0.1707026597856292,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999423146301242,
      "loss": 5.9725,
      "mean_token_accuracy": 0.1192103810608387,
      "num_tokens": 3779809.0,
      "step": 1720
    },
    {
      "entropy": 6.11200361251831,
      "epoch": 0.17119888844779674,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999415095753709,
      "loss": 5.9783,
      "mean_token_accuracy": 0.12067282646894455,
      "num_tokens": 3791455.0,
      "step": 1725
    },
    {
      "entropy": 6.143737411499023,
      "epoch": 0.17169511710996427,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999406989425527,
      "loss": 5.9996,
      "mean_token_accuracy": 0.12016435712575912,
      "num_tokens": 3803209.0,
      "step": 1730
    },
    {
      "entropy": 6.139561700820923,
      "epoch": 0.1721913457721318,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999398827316901,
      "loss": 6.0474,
      "mean_token_accuracy": 0.12039401978254319,
      "num_tokens": 3813553.0,
      "step": 1735
    },
    {
      "entropy": 6.202684926986694,
      "epoch": 0.17268757443429933,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499939060942803,
      "loss": 6.0878,
      "mean_token_accuracy": 0.11598005071282387,
      "num_tokens": 3824856.0,
      "step": 1740
    },
    {
      "entropy": 6.242949295043945,
      "epoch": 0.17318380309646686,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999382335759119,
      "loss": 6.1542,
      "mean_token_accuracy": 0.11370768472552299,
      "num_tokens": 3835489.0,
      "step": 1745
    },
    {
      "entropy": 6.265321969985962,
      "epoch": 0.1736800317586344,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999374006310372,
      "loss": 6.0015,
      "mean_token_accuracy": 0.12164871916174888,
      "num_tokens": 3844838.0,
      "step": 1750
    },
    {
      "entropy": 6.10846905708313,
      "epoch": 0.1741762604208019,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999365621081999,
      "loss": 5.9625,
      "mean_token_accuracy": 0.11757483780384063,
      "num_tokens": 3855288.0,
      "step": 1755
    },
    {
      "entropy": 6.1842241287231445,
      "epoch": 0.17467248908296942,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999357180074204,
      "loss": 6.0165,
      "mean_token_accuracy": 0.12564777955412865,
      "num_tokens": 3865585.0,
      "step": 1760
    },
    {
      "entropy": 6.159081268310547,
      "epoch": 0.17516871774513695,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999348683287199,
      "loss": 5.9979,
      "mean_token_accuracy": 0.11737576574087143,
      "num_tokens": 3877694.0,
      "step": 1765
    },
    {
      "entropy": 6.146845006942749,
      "epoch": 0.17566494640730448,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999340130721192,
      "loss": 5.9587,
      "mean_token_accuracy": 0.13080541640520096,
      "num_tokens": 3888802.0,
      "step": 1770
    },
    {
      "entropy": 6.100156831741333,
      "epoch": 0.17616117506947201,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999331522376396,
      "loss": 5.9332,
      "mean_token_accuracy": 0.12021833211183548,
      "num_tokens": 3900294.0,
      "step": 1775
    },
    {
      "entropy": 6.149476766586304,
      "epoch": 0.17665740373163955,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999322858253027,
      "loss": 5.9754,
      "mean_token_accuracy": 0.12123704552650452,
      "num_tokens": 3910865.0,
      "step": 1780
    },
    {
      "entropy": 6.165459203720093,
      "epoch": 0.17715363239380708,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999314138351297,
      "loss": 6.0263,
      "mean_token_accuracy": 0.12229800149798394,
      "num_tokens": 3921960.0,
      "step": 1785
    },
    {
      "entropy": 6.191252756118774,
      "epoch": 0.1776498610559746,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999305362671424,
      "loss": 6.0387,
      "mean_token_accuracy": 0.11913010105490685,
      "num_tokens": 3934085.0,
      "step": 1790
    },
    {
      "entropy": 6.127382612228393,
      "epoch": 0.1781460897181421,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999296531213624,
      "loss": 5.8928,
      "mean_token_accuracy": 0.13389062657952308,
      "num_tokens": 3944532.0,
      "step": 1795
    },
    {
      "entropy": 6.108739185333252,
      "epoch": 0.17864231838030964,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999287643978117,
      "loss": 5.9284,
      "mean_token_accuracy": 0.1266961745917797,
      "num_tokens": 3954995.0,
      "step": 1800
    },
    {
      "entropy": 6.063549137115478,
      "epoch": 0.17913854704247717,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999278700965124,
      "loss": 5.99,
      "mean_token_accuracy": 0.1248068444430828,
      "num_tokens": 3966621.0,
      "step": 1805
    },
    {
      "entropy": 6.155789041519165,
      "epoch": 0.1796347757046447,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999269702174866,
      "loss": 6.0518,
      "mean_token_accuracy": 0.12221262827515603,
      "num_tokens": 3977568.0,
      "step": 1810
    },
    {
      "entropy": 6.161475467681885,
      "epoch": 0.18013100436681223,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999260647607565,
      "loss": 6.0683,
      "mean_token_accuracy": 0.11906974241137505,
      "num_tokens": 3988571.0,
      "step": 1815
    },
    {
      "entropy": 6.180405378341675,
      "epoch": 0.18062723302897976,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999251537263448,
      "loss": 5.8891,
      "mean_token_accuracy": 0.12807116732001306,
      "num_tokens": 3998380.0,
      "step": 1820
    },
    {
      "entropy": 6.073153924942017,
      "epoch": 0.1811234616911473,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999242371142739,
      "loss": 6.0713,
      "mean_token_accuracy": 0.12228276878595352,
      "num_tokens": 4010017.0,
      "step": 1825
    },
    {
      "entropy": 6.301033782958984,
      "epoch": 0.1816196903533148,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999233149245665,
      "loss": 6.0383,
      "mean_token_accuracy": 0.12104798778891564,
      "num_tokens": 4020922.0,
      "step": 1830
    },
    {
      "entropy": 6.202951669692993,
      "epoch": 0.18211591901548232,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999223871572456,
      "loss": 6.0035,
      "mean_token_accuracy": 0.12094608992338181,
      "num_tokens": 4030927.0,
      "step": 1835
    },
    {
      "entropy": 6.1489091396331785,
      "epoch": 0.18261214767764986,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999214538123342,
      "loss": 6.0733,
      "mean_token_accuracy": 0.11611957773566246,
      "num_tokens": 4041375.0,
      "step": 1840
    },
    {
      "entropy": 6.178863096237182,
      "epoch": 0.18310837633981739,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999205148898554,
      "loss": 6.0746,
      "mean_token_accuracy": 0.11968084275722504,
      "num_tokens": 4053213.0,
      "step": 1845
    },
    {
      "entropy": 6.172112369537354,
      "epoch": 0.18360460500198492,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999195703898325,
      "loss": 5.9206,
      "mean_token_accuracy": 0.12607544884085656,
      "num_tokens": 4064536.0,
      "step": 1850
    },
    {
      "entropy": 6.156340265274048,
      "epoch": 0.18410083366415245,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999186203122889,
      "loss": 5.9853,
      "mean_token_accuracy": 0.12510076612234117,
      "num_tokens": 4075282.0,
      "step": 1855
    },
    {
      "entropy": 6.041736364364624,
      "epoch": 0.18459706232631998,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999176646572482,
      "loss": 5.9996,
      "mean_token_accuracy": 0.12671509012579918,
      "num_tokens": 4085501.0,
      "step": 1860
    },
    {
      "entropy": 6.163187837600708,
      "epoch": 0.1850932909884875,
      "grad_norm": 1.0,
      "learning_rate": 0.000499916703424734,
      "loss": 5.9386,
      "mean_token_accuracy": 0.11762271374464035,
      "num_tokens": 4095697.0,
      "step": 1865
    },
    {
      "entropy": 6.143773317337036,
      "epoch": 0.185589519650655,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999157366147702,
      "loss": 6.0013,
      "mean_token_accuracy": 0.12127534747123718,
      "num_tokens": 4106459.0,
      "step": 1870
    },
    {
      "entropy": 6.127069616317749,
      "epoch": 0.18608574831282254,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999147642273809,
      "loss": 5.9328,
      "mean_token_accuracy": 0.12429732158780098,
      "num_tokens": 4118320.0,
      "step": 1875
    },
    {
      "entropy": 6.075209379196167,
      "epoch": 0.18658197697499007,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00049991378626259,
      "loss": 5.9491,
      "mean_token_accuracy": 0.13264545872807504,
      "num_tokens": 4131249.0,
      "step": 1880
    },
    {
      "entropy": 6.168709850311279,
      "epoch": 0.1870782056371576,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999128027204219,
      "loss": 6.0205,
      "mean_token_accuracy": 0.1249746061861515,
      "num_tokens": 4141480.0,
      "step": 1885
    },
    {
      "entropy": 6.181358861923218,
      "epoch": 0.18757443429932513,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499911813600901,
      "loss": 6.0712,
      "mean_token_accuracy": 0.12171259969472885,
      "num_tokens": 4152344.0,
      "step": 1890
    },
    {
      "entropy": 6.18400239944458,
      "epoch": 0.18807066296149266,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999108189040516,
      "loss": 6.0532,
      "mean_token_accuracy": 0.1188795156776905,
      "num_tokens": 4164772.0,
      "step": 1895
    },
    {
      "entropy": 6.120943212509156,
      "epoch": 0.1885668916236602,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999098186298986,
      "loss": 5.9959,
      "mean_token_accuracy": 0.11978129372000694,
      "num_tokens": 4175511.0,
      "step": 1900
    },
    {
      "entropy": 6.083007574081421,
      "epoch": 0.1890631202858277,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999088127784667,
      "loss": 5.9597,
      "mean_token_accuracy": 0.1269135594367981,
      "num_tokens": 4187445.0,
      "step": 1905
    },
    {
      "entropy": 6.157655334472656,
      "epoch": 0.18955934894799523,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000499907801349781,
      "loss": 5.9288,
      "mean_token_accuracy": 0.1249204121530056,
      "num_tokens": 4198124.0,
      "step": 1910
    },
    {
      "entropy": 6.117712736129761,
      "epoch": 0.19005557761016276,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999067843438664,
      "loss": 5.9642,
      "mean_token_accuracy": 0.12320106327533722,
      "num_tokens": 4208954.0,
      "step": 1915
    },
    {
      "entropy": 6.1655356884002686,
      "epoch": 0.1905518062723303,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999057617607482,
      "loss": 6.036,
      "mean_token_accuracy": 0.12222218587994575,
      "num_tokens": 4218896.0,
      "step": 1920
    },
    {
      "entropy": 6.073115491867066,
      "epoch": 0.19104803493449782,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999047336004517,
      "loss": 5.9545,
      "mean_token_accuracy": 0.11575927585363388,
      "num_tokens": 4228851.0,
      "step": 1925
    },
    {
      "entropy": 6.062970113754273,
      "epoch": 0.19154426359666535,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999036998630024,
      "loss": 5.9034,
      "mean_token_accuracy": 0.12176035344600677,
      "num_tokens": 4240377.0,
      "step": 1930
    },
    {
      "entropy": 6.1012085437774655,
      "epoch": 0.19204049225883288,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499902660548426,
      "loss": 5.9005,
      "mean_token_accuracy": 0.12233036682009697,
      "num_tokens": 4251341.0,
      "step": 1935
    },
    {
      "entropy": 6.09600863456726,
      "epoch": 0.1925367209210004,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999016156567482,
      "loss": 5.9999,
      "mean_token_accuracy": 0.12348600029945374,
      "num_tokens": 4261206.0,
      "step": 1940
    },
    {
      "entropy": 6.171257972717285,
      "epoch": 0.1930329495831679,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499900565187995,
      "loss": 6.0095,
      "mean_token_accuracy": 0.11739006116986275,
      "num_tokens": 4272662.0,
      "step": 1945
    },
    {
      "entropy": 6.155332422256469,
      "epoch": 0.19352917824533544,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998995091421924,
      "loss": 5.9894,
      "mean_token_accuracy": 0.12250523716211319,
      "num_tokens": 4283302.0,
      "step": 1950
    },
    {
      "entropy": 6.102492046356201,
      "epoch": 0.19402540690750297,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004998984475193665,
      "loss": 5.9331,
      "mean_token_accuracy": 0.1246795579791069,
      "num_tokens": 4296087.0,
      "step": 1955
    },
    {
      "entropy": 6.071946620941162,
      "epoch": 0.1945216355696705,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998973803195439,
      "loss": 5.9506,
      "mean_token_accuracy": 0.12343922182917595,
      "num_tokens": 4305918.0,
      "step": 1960
    },
    {
      "entropy": 6.098349571228027,
      "epoch": 0.19501786423183803,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998963075427508,
      "loss": 6.0056,
      "mean_token_accuracy": 0.12570720240473748,
      "num_tokens": 4316507.0,
      "step": 1965
    },
    {
      "entropy": 6.172418403625488,
      "epoch": 0.19551409289400556,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004998952291890138,
      "loss": 5.9377,
      "mean_token_accuracy": 0.12697042599320413,
      "num_tokens": 4329091.0,
      "step": 1970
    },
    {
      "entropy": 6.116201543807984,
      "epoch": 0.1960103215561731,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998941452583599,
      "loss": 6.0005,
      "mean_token_accuracy": 0.12526401430368422,
      "num_tokens": 4339266.0,
      "step": 1975
    },
    {
      "entropy": 6.060456848144531,
      "epoch": 0.1965065502183406,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998930557508156,
      "loss": 5.9315,
      "mean_token_accuracy": 0.12996648699045182,
      "num_tokens": 4349300.0,
      "step": 1980
    },
    {
      "entropy": 6.031670236587525,
      "epoch": 0.19700277888050813,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004998919606664082,
      "loss": 5.8523,
      "mean_token_accuracy": 0.1329902485013008,
      "num_tokens": 4360314.0,
      "step": 1985
    },
    {
      "entropy": 6.06371750831604,
      "epoch": 0.19749900754267566,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998908600051648,
      "loss": 5.9022,
      "mean_token_accuracy": 0.12996097579598426,
      "num_tokens": 4370077.0,
      "step": 1990
    },
    {
      "entropy": 6.041788816452026,
      "epoch": 0.1979952362048432,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998897537671126,
      "loss": 5.9163,
      "mean_token_accuracy": 0.12319361120462417,
      "num_tokens": 4381000.0,
      "step": 1995
    },
    {
      "entropy": 6.162951469421387,
      "epoch": 0.19849146486701072,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998886419522792,
      "loss": 6.1233,
      "mean_token_accuracy": 0.11464894115924835,
      "num_tokens": 4393485.0,
      "step": 2000
    },
    {
      "entropy": 6.207508325576782,
      "epoch": 0.19898769352917825,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004998875245606919,
      "loss": 5.9466,
      "mean_token_accuracy": 0.1263269901275635,
      "num_tokens": 4404354.0,
      "step": 2005
    },
    {
      "entropy": 6.115656423568725,
      "epoch": 0.19948392219134578,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998864015923788,
      "loss": 6.0848,
      "mean_token_accuracy": 0.12513597384095193,
      "num_tokens": 4414818.0,
      "step": 2010
    },
    {
      "entropy": 6.1054645538330075,
      "epoch": 0.1999801508535133,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998852730473674,
      "loss": 5.9732,
      "mean_token_accuracy": 0.1186623029410839,
      "num_tokens": 4426271.0,
      "step": 2015
    },
    {
      "entropy": 6.113159704208374,
      "epoch": 0.2004763795156808,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998841389256859,
      "loss": 6.0201,
      "mean_token_accuracy": 0.12535962387919425,
      "num_tokens": 4436981.0,
      "step": 2020
    },
    {
      "entropy": 6.088190221786499,
      "epoch": 0.20097260817784834,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998829992273622,
      "loss": 5.8457,
      "mean_token_accuracy": 0.12926107719540597,
      "num_tokens": 4446494.0,
      "step": 2025
    },
    {
      "entropy": 6.0542316913604735,
      "epoch": 0.20146883684001587,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998818539524248,
      "loss": 5.941,
      "mean_token_accuracy": 0.12792145758867263,
      "num_tokens": 4458535.0,
      "step": 2030
    },
    {
      "entropy": 6.076647806167602,
      "epoch": 0.2019650655021834,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499880703100902,
      "loss": 5.9109,
      "mean_token_accuracy": 0.12890672460198402,
      "num_tokens": 4469025.0,
      "step": 2035
    },
    {
      "entropy": 6.119675874710083,
      "epoch": 0.20246129416435094,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998795466728223,
      "loss": 5.9311,
      "mean_token_accuracy": 0.12614311054348945,
      "num_tokens": 4478431.0,
      "step": 2040
    },
    {
      "entropy": 6.100471878051758,
      "epoch": 0.20295752282651847,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998783846682144,
      "loss": 6.0565,
      "mean_token_accuracy": 0.12169205471873283,
      "num_tokens": 4490215.0,
      "step": 2045
    },
    {
      "entropy": 6.08721661567688,
      "epoch": 0.203453751488686,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004998772170871072,
      "loss": 5.9323,
      "mean_token_accuracy": 0.12300658226013184,
      "num_tokens": 4502895.0,
      "step": 2050
    },
    {
      "entropy": 6.049704933166504,
      "epoch": 0.20394998015085353,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998760439295295,
      "loss": 5.9527,
      "mean_token_accuracy": 0.12245631441473961,
      "num_tokens": 4514468.0,
      "step": 2055
    },
    {
      "entropy": 6.120728397369385,
      "epoch": 0.20444620881302103,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998748651955105,
      "loss": 5.9697,
      "mean_token_accuracy": 0.12270477414131165,
      "num_tokens": 4525787.0,
      "step": 2060
    },
    {
      "entropy": 6.1617701053619385,
      "epoch": 0.20494243747518856,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004998736808850795,
      "loss": 6.0537,
      "mean_token_accuracy": 0.1189173199236393,
      "num_tokens": 4537536.0,
      "step": 2065
    },
    {
      "entropy": 6.137720203399658,
      "epoch": 0.2054386661373561,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998724909982656,
      "loss": 5.9957,
      "mean_token_accuracy": 0.1230063110589981,
      "num_tokens": 4549619.0,
      "step": 2070
    },
    {
      "entropy": 6.054072999954224,
      "epoch": 0.20593489479952362,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004998712955350986,
      "loss": 5.9127,
      "mean_token_accuracy": 0.12620335072278976,
      "num_tokens": 4560248.0,
      "step": 2075
    },
    {
      "entropy": 6.133942365646362,
      "epoch": 0.20643112346169115,
      "grad_norm": 0.875,
      "learning_rate": 0.0004998700944956081,
      "loss": 6.0258,
      "mean_token_accuracy": 0.12564925700426102,
      "num_tokens": 4571888.0,
      "step": 2080
    },
    {
      "entropy": 6.092451715469361,
      "epoch": 0.20692735212385868,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004998688878798237,
      "loss": 5.8719,
      "mean_token_accuracy": 0.13250108286738396,
      "num_tokens": 4583234.0,
      "step": 2085
    },
    {
      "entropy": 6.129697036743164,
      "epoch": 0.2074235807860262,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998676756877753,
      "loss": 6.0849,
      "mean_token_accuracy": 0.1196140132844448,
      "num_tokens": 4595252.0,
      "step": 2090
    },
    {
      "entropy": 6.055380916595459,
      "epoch": 0.20791980944819372,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998664579194934,
      "loss": 5.8636,
      "mean_token_accuracy": 0.12891709432005882,
      "num_tokens": 4607091.0,
      "step": 2095
    },
    {
      "entropy": 6.060617876052857,
      "epoch": 0.20841603811036125,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998652345750077,
      "loss": 5.8251,
      "mean_token_accuracy": 0.13188225999474526,
      "num_tokens": 4618003.0,
      "step": 2100
    },
    {
      "entropy": 6.004556941986084,
      "epoch": 0.20891226677252878,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004998640056543487,
      "loss": 5.8382,
      "mean_token_accuracy": 0.12877558246254922,
      "num_tokens": 4628429.0,
      "step": 2105
    },
    {
      "entropy": 6.042101526260376,
      "epoch": 0.2094084954346963,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998627711575469,
      "loss": 5.9003,
      "mean_token_accuracy": 0.1301546387374401,
      "num_tokens": 4639414.0,
      "step": 2110
    },
    {
      "entropy": 6.125416374206543,
      "epoch": 0.20990472409686384,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000499861531084633,
      "loss": 5.9868,
      "mean_token_accuracy": 0.12510028034448623,
      "num_tokens": 4652594.0,
      "step": 2115
    },
    {
      "entropy": 5.983136796951294,
      "epoch": 0.21040095275903137,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998602854356376,
      "loss": 5.8353,
      "mean_token_accuracy": 0.13099646791815758,
      "num_tokens": 4662981.0,
      "step": 2120
    },
    {
      "entropy": 6.04475212097168,
      "epoch": 0.2108971814211989,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998590342105917,
      "loss": 5.929,
      "mean_token_accuracy": 0.12667660117149354,
      "num_tokens": 4672469.0,
      "step": 2125
    },
    {
      "entropy": 6.043806028366089,
      "epoch": 0.21139341008336643,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998577774095262,
      "loss": 5.9052,
      "mean_token_accuracy": 0.12564729601144792,
      "num_tokens": 4683640.0,
      "step": 2130
    },
    {
      "entropy": 6.0359203815460205,
      "epoch": 0.21188963874553393,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998565150324724,
      "loss": 5.8664,
      "mean_token_accuracy": 0.12986178547143937,
      "num_tokens": 4694032.0,
      "step": 2135
    },
    {
      "entropy": 5.933635091781616,
      "epoch": 0.21238586740770146,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998552470794616,
      "loss": 5.7421,
      "mean_token_accuracy": 0.13141876757144927,
      "num_tokens": 4704253.0,
      "step": 2140
    },
    {
      "entropy": 6.008824586868286,
      "epoch": 0.212882096069869,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998539735505251,
      "loss": 5.8993,
      "mean_token_accuracy": 0.1256441943347454,
      "num_tokens": 4715327.0,
      "step": 2145
    },
    {
      "entropy": 6.166449022293091,
      "epoch": 0.21337832473203652,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998526944456946,
      "loss": 5.9113,
      "mean_token_accuracy": 0.12448364496231079,
      "num_tokens": 4726450.0,
      "step": 2150
    },
    {
      "entropy": 6.0194117546081545,
      "epoch": 0.21387455339420405,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998514097650018,
      "loss": 5.9441,
      "mean_token_accuracy": 0.12404742538928985,
      "num_tokens": 4736630.0,
      "step": 2155
    },
    {
      "entropy": 6.021891641616821,
      "epoch": 0.21437078205637158,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998501195084786,
      "loss": 5.8261,
      "mean_token_accuracy": 0.13041977211833,
      "num_tokens": 4746592.0,
      "step": 2160
    },
    {
      "entropy": 6.075963544845581,
      "epoch": 0.21486701071853911,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998488236761569,
      "loss": 5.9943,
      "mean_token_accuracy": 0.12953520566225052,
      "num_tokens": 4757824.0,
      "step": 2165
    },
    {
      "entropy": 6.078512859344483,
      "epoch": 0.21536323938070662,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998475222680688,
      "loss": 5.9418,
      "mean_token_accuracy": 0.12912883386015891,
      "num_tokens": 4768453.0,
      "step": 2170
    },
    {
      "entropy": 5.984212827682495,
      "epoch": 0.21585946804287415,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998462152842468,
      "loss": 5.9138,
      "mean_token_accuracy": 0.1260770782828331,
      "num_tokens": 4779664.0,
      "step": 2175
    },
    {
      "entropy": 6.1202507495880125,
      "epoch": 0.21635569670504168,
      "grad_norm": 0.859375,
      "learning_rate": 0.000499844902724723,
      "loss": 5.9297,
      "mean_token_accuracy": 0.1291131466627121,
      "num_tokens": 4791012.0,
      "step": 2180
    },
    {
      "entropy": 6.0246453285217285,
      "epoch": 0.2168519253672092,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998435845895302,
      "loss": 5.804,
      "mean_token_accuracy": 0.1243359960615635,
      "num_tokens": 4800578.0,
      "step": 2185
    },
    {
      "entropy": 5.987508344650268,
      "epoch": 0.21734815402937674,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499842260878701,
      "loss": 5.8239,
      "mean_token_accuracy": 0.12554812505841256,
      "num_tokens": 4810940.0,
      "step": 2190
    },
    {
      "entropy": 6.046344041824341,
      "epoch": 0.21784438269154427,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004998409315922683,
      "loss": 5.9594,
      "mean_token_accuracy": 0.12463009878993034,
      "num_tokens": 4822696.0,
      "step": 2195
    },
    {
      "entropy": 6.06123194694519,
      "epoch": 0.2183406113537118,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998395967302649,
      "loss": 5.8424,
      "mean_token_accuracy": 0.1316361606121063,
      "num_tokens": 4833669.0,
      "step": 2200
    },
    {
      "entropy": 6.1397562503814695,
      "epoch": 0.21883684001587933,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499838256292724,
      "loss": 6.0156,
      "mean_token_accuracy": 0.12383595556020736,
      "num_tokens": 4844021.0,
      "step": 2205
    },
    {
      "entropy": 5.95759687423706,
      "epoch": 0.21933306867804683,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998369102796787,
      "loss": 5.8932,
      "mean_token_accuracy": 0.12637994140386583,
      "num_tokens": 4855336.0,
      "step": 2210
    },
    {
      "entropy": 6.048965978622436,
      "epoch": 0.21982929734021436,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004998355586911627,
      "loss": 5.9059,
      "mean_token_accuracy": 0.12790409997105598,
      "num_tokens": 4867713.0,
      "step": 2215
    },
    {
      "entropy": 5.9969117641448975,
      "epoch": 0.2203255260023819,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004998342015272093,
      "loss": 5.8694,
      "mean_token_accuracy": 0.12453557699918746,
      "num_tokens": 4879134.0,
      "step": 2220
    },
    {
      "entropy": 6.086967134475708,
      "epoch": 0.22082175466454942,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004998328387878521,
      "loss": 6.0274,
      "mean_token_accuracy": 0.12505436688661575,
      "num_tokens": 4891755.0,
      "step": 2225
    },
    {
      "entropy": 6.152056741714477,
      "epoch": 0.22131798332671695,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998314704731249,
      "loss": 6.0155,
      "mean_token_accuracy": 0.12360429838299751,
      "num_tokens": 4903264.0,
      "step": 2230
    },
    {
      "entropy": 6.021547603607178,
      "epoch": 0.22181421198888449,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004998300965830618,
      "loss": 5.8482,
      "mean_token_accuracy": 0.1265365608036518,
      "num_tokens": 4912533.0,
      "step": 2235
    },
    {
      "entropy": 5.9325371265411375,
      "epoch": 0.22231044065105202,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004998287171176968,
      "loss": 5.8439,
      "mean_token_accuracy": 0.13453373461961746,
      "num_tokens": 4923288.0,
      "step": 2240
    },
    {
      "entropy": 6.013333702087403,
      "epoch": 0.22280666931321952,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998273320770641,
      "loss": 5.861,
      "mean_token_accuracy": 0.1291363902390003,
      "num_tokens": 4934759.0,
      "step": 2245
    },
    {
      "entropy": 6.0565742492675785,
      "epoch": 0.22330289797538705,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499825941461198,
      "loss": 5.9108,
      "mean_token_accuracy": 0.12829016521573067,
      "num_tokens": 4947024.0,
      "step": 2250
    },
    {
      "entropy": 5.990304183959961,
      "epoch": 0.22379912663755458,
      "grad_norm": 0.875,
      "learning_rate": 0.0004998245452701329,
      "loss": 5.8611,
      "mean_token_accuracy": 0.12755537852644921,
      "num_tokens": 4959281.0,
      "step": 2255
    },
    {
      "entropy": 6.069461011886597,
      "epoch": 0.2242953552997221,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998231435039036,
      "loss": 5.9047,
      "mean_token_accuracy": 0.12395669743418694,
      "num_tokens": 4970466.0,
      "step": 2260
    },
    {
      "entropy": 6.021377325057983,
      "epoch": 0.22479158396188964,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998217361625449,
      "loss": 5.8562,
      "mean_token_accuracy": 0.12923893854022026,
      "num_tokens": 4981135.0,
      "step": 2265
    },
    {
      "entropy": 6.048352479934692,
      "epoch": 0.22528781262405717,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998203232460915,
      "loss": 5.9167,
      "mean_token_accuracy": 0.1324473612010479,
      "num_tokens": 4991292.0,
      "step": 2270
    },
    {
      "entropy": 5.9737649917602536,
      "epoch": 0.2257840412862247,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998189047545785,
      "loss": 5.8825,
      "mean_token_accuracy": 0.13191651031374932,
      "num_tokens": 5002806.0,
      "step": 2275
    },
    {
      "entropy": 5.993149375915527,
      "epoch": 0.22628026994839223,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998174806880411,
      "loss": 5.8179,
      "mean_token_accuracy": 0.13567918241024018,
      "num_tokens": 5013474.0,
      "step": 2280
    },
    {
      "entropy": 6.056148862838745,
      "epoch": 0.22677649861055973,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998160510465146,
      "loss": 5.8642,
      "mean_token_accuracy": 0.13143307566642762,
      "num_tokens": 5025151.0,
      "step": 2285
    },
    {
      "entropy": 5.92156720161438,
      "epoch": 0.22727272727272727,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004998146158300345,
      "loss": 5.8638,
      "mean_token_accuracy": 0.13245996236801147,
      "num_tokens": 5037166.0,
      "step": 2290
    },
    {
      "entropy": 6.050130987167359,
      "epoch": 0.2277689559348948,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998131750386364,
      "loss": 5.9076,
      "mean_token_accuracy": 0.1297108441591263,
      "num_tokens": 5047452.0,
      "step": 2295
    },
    {
      "entropy": 6.028859281539917,
      "epoch": 0.22826518459706233,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998117286723559,
      "loss": 5.9272,
      "mean_token_accuracy": 0.11998689249157905,
      "num_tokens": 5058501.0,
      "step": 2300
    },
    {
      "entropy": 6.028175163269043,
      "epoch": 0.22876141325922986,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499810276731229,
      "loss": 5.9102,
      "mean_token_accuracy": 0.12801963984966278,
      "num_tokens": 5069716.0,
      "step": 2305
    },
    {
      "entropy": 6.123485803604126,
      "epoch": 0.2292576419213974,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998088192152918,
      "loss": 5.8893,
      "mean_token_accuracy": 0.1252819001674652,
      "num_tokens": 5078509.0,
      "step": 2310
    },
    {
      "entropy": 6.000077247619629,
      "epoch": 0.22975387058356492,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998073561245801,
      "loss": 5.9347,
      "mean_token_accuracy": 0.12345504015684128,
      "num_tokens": 5090355.0,
      "step": 2315
    },
    {
      "entropy": 6.051438713073731,
      "epoch": 0.23025009924573242,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004998058874591305,
      "loss": 5.8242,
      "mean_token_accuracy": 0.1337508350610733,
      "num_tokens": 5100582.0,
      "step": 2320
    },
    {
      "entropy": 6.000838756561279,
      "epoch": 0.23074632790789995,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004998044132189793,
      "loss": 5.8887,
      "mean_token_accuracy": 0.12203488871455193,
      "num_tokens": 5112176.0,
      "step": 2325
    },
    {
      "entropy": 6.021178150177002,
      "epoch": 0.23124255657006748,
      "grad_norm": 1.28125,
      "learning_rate": 0.000499802933404163,
      "loss": 5.8335,
      "mean_token_accuracy": 0.1279439650475979,
      "num_tokens": 5122840.0,
      "step": 2330
    },
    {
      "entropy": 6.0098320007324215,
      "epoch": 0.231738785232235,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998014480147183,
      "loss": 5.8047,
      "mean_token_accuracy": 0.13865086659789086,
      "num_tokens": 5132926.0,
      "step": 2335
    },
    {
      "entropy": 5.934864521026611,
      "epoch": 0.23223501389440254,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997999570506823,
      "loss": 5.7806,
      "mean_token_accuracy": 0.13042209595441817,
      "num_tokens": 5143394.0,
      "step": 2340
    },
    {
      "entropy": 5.982978010177613,
      "epoch": 0.23273124255657007,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004997984605120917,
      "loss": 5.8147,
      "mean_token_accuracy": 0.12575569450855256,
      "num_tokens": 5155710.0,
      "step": 2345
    },
    {
      "entropy": 5.841717767715454,
      "epoch": 0.2332274712187376,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997969583989835,
      "loss": 5.751,
      "mean_token_accuracy": 0.12713968753814697,
      "num_tokens": 5167065.0,
      "step": 2350
    },
    {
      "entropy": 5.989433431625367,
      "epoch": 0.23372369988090513,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997954507113952,
      "loss": 5.8495,
      "mean_token_accuracy": 0.12923550233244896,
      "num_tokens": 5177535.0,
      "step": 2355
    },
    {
      "entropy": 6.037681818008423,
      "epoch": 0.23421992854307264,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004997939374493642,
      "loss": 5.7806,
      "mean_token_accuracy": 0.13312982767820358,
      "num_tokens": 5188324.0,
      "step": 2360
    },
    {
      "entropy": 5.901477909088134,
      "epoch": 0.23471615720524017,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499792418612928,
      "loss": 5.86,
      "mean_token_accuracy": 0.1304488480091095,
      "num_tokens": 5199223.0,
      "step": 2365
    },
    {
      "entropy": 5.945672512054443,
      "epoch": 0.2352123858674077,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997908942021241,
      "loss": 5.8299,
      "mean_token_accuracy": 0.13325254768133163,
      "num_tokens": 5210080.0,
      "step": 2370
    },
    {
      "entropy": 6.099682092666626,
      "epoch": 0.23570861452957523,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997893642169904,
      "loss": 5.9558,
      "mean_token_accuracy": 0.12958926782011987,
      "num_tokens": 5220867.0,
      "step": 2375
    },
    {
      "entropy": 6.014858865737915,
      "epoch": 0.23620484319174276,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004997878286575648,
      "loss": 5.7952,
      "mean_token_accuracy": 0.12957458645105363,
      "num_tokens": 5231451.0,
      "step": 2380
    },
    {
      "entropy": 5.938134288787841,
      "epoch": 0.2367010718539103,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997862875238855,
      "loss": 5.9034,
      "mean_token_accuracy": 0.12620984017848969,
      "num_tokens": 5244643.0,
      "step": 2385
    },
    {
      "entropy": 6.035072565078735,
      "epoch": 0.23719730051607782,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004997847408159907,
      "loss": 5.9145,
      "mean_token_accuracy": 0.12331874147057534,
      "num_tokens": 5256279.0,
      "step": 2390
    },
    {
      "entropy": 6.0572350978851315,
      "epoch": 0.23769352917824532,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997831885339186,
      "loss": 5.8955,
      "mean_token_accuracy": 0.13024576678872107,
      "num_tokens": 5268498.0,
      "step": 2395
    },
    {
      "entropy": 5.939180183410644,
      "epoch": 0.23818975784041285,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997816306777077,
      "loss": 5.8567,
      "mean_token_accuracy": 0.13194234371185304,
      "num_tokens": 5280869.0,
      "step": 2400
    },
    {
      "entropy": 6.102878284454346,
      "epoch": 0.23868598650258038,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997800672473969,
      "loss": 5.9558,
      "mean_token_accuracy": 0.12405509725213051,
      "num_tokens": 5293082.0,
      "step": 2405
    },
    {
      "entropy": 5.935460233688355,
      "epoch": 0.2391822151647479,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004997784982430247,
      "loss": 5.7407,
      "mean_token_accuracy": 0.14041346311569214,
      "num_tokens": 5304039.0,
      "step": 2410
    },
    {
      "entropy": 5.937913465499878,
      "epoch": 0.23967844382691544,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997769236646301,
      "loss": 5.8253,
      "mean_token_accuracy": 0.13146729916334152,
      "num_tokens": 5315016.0,
      "step": 2415
    },
    {
      "entropy": 6.090373134613037,
      "epoch": 0.24017467248908297,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997753435122522,
      "loss": 5.9481,
      "mean_token_accuracy": 0.12890507653355598,
      "num_tokens": 5326090.0,
      "step": 2420
    },
    {
      "entropy": 5.9364526748657225,
      "epoch": 0.2406709011512505,
      "grad_norm": 1.015625,
      "learning_rate": 0.00049977375778593,
      "loss": 5.7381,
      "mean_token_accuracy": 0.13036757260560988,
      "num_tokens": 5336423.0,
      "step": 2425
    },
    {
      "entropy": 5.988410186767578,
      "epoch": 0.24116712981341804,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997721664857032,
      "loss": 5.9064,
      "mean_token_accuracy": 0.13083979934453965,
      "num_tokens": 5346982.0,
      "step": 2430
    },
    {
      "entropy": 5.9514707088470455,
      "epoch": 0.24166335847558554,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997705696116108,
      "loss": 5.8041,
      "mean_token_accuracy": 0.13514370396733283,
      "num_tokens": 5359210.0,
      "step": 2435
    },
    {
      "entropy": 5.958845949172973,
      "epoch": 0.24215958713775307,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997689671636926,
      "loss": 5.8358,
      "mean_token_accuracy": 0.12565323561429978,
      "num_tokens": 5369891.0,
      "step": 2440
    },
    {
      "entropy": 5.962273597717285,
      "epoch": 0.2426558157999206,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004997673591419884,
      "loss": 5.8116,
      "mean_token_accuracy": 0.1305241070687771,
      "num_tokens": 5380850.0,
      "step": 2445
    },
    {
      "entropy": 5.970908737182617,
      "epoch": 0.24315204446208813,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004997657455465381,
      "loss": 5.8724,
      "mean_token_accuracy": 0.13438860848546028,
      "num_tokens": 5391903.0,
      "step": 2450
    },
    {
      "entropy": 6.027713441848755,
      "epoch": 0.24364827312425566,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997641263773815,
      "loss": 5.9209,
      "mean_token_accuracy": 0.1283269263803959,
      "num_tokens": 5402517.0,
      "step": 2455
    },
    {
      "entropy": 5.9370378971099855,
      "epoch": 0.2441445017864232,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499762501634559,
      "loss": 5.8357,
      "mean_token_accuracy": 0.12951843738555907,
      "num_tokens": 5414110.0,
      "step": 2460
    },
    {
      "entropy": 6.02907428741455,
      "epoch": 0.24464073044859072,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997608713181108,
      "loss": 5.8403,
      "mean_token_accuracy": 0.13202603831887244,
      "num_tokens": 5424466.0,
      "step": 2465
    },
    {
      "entropy": 5.915480756759644,
      "epoch": 0.24513695911075822,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997592354280772,
      "loss": 5.8526,
      "mean_token_accuracy": 0.1366075359284878,
      "num_tokens": 5434461.0,
      "step": 2470
    },
    {
      "entropy": 5.940920686721801,
      "epoch": 0.24563318777292575,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499757593964499,
      "loss": 5.8263,
      "mean_token_accuracy": 0.12434638366103172,
      "num_tokens": 5445921.0,
      "step": 2475
    },
    {
      "entropy": 5.955485916137695,
      "epoch": 0.24612941643509328,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997559469274166,
      "loss": 5.8127,
      "mean_token_accuracy": 0.12727661803364754,
      "num_tokens": 5456283.0,
      "step": 2480
    },
    {
      "entropy": 5.970708322525025,
      "epoch": 0.24662564509726081,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004997542943168712,
      "loss": 5.8501,
      "mean_token_accuracy": 0.1340114578604698,
      "num_tokens": 5467616.0,
      "step": 2485
    },
    {
      "entropy": 5.974401187896729,
      "epoch": 0.24712187375942835,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997526361329035,
      "loss": 5.8048,
      "mean_token_accuracy": 0.13320422694087028,
      "num_tokens": 5478050.0,
      "step": 2490
    },
    {
      "entropy": 6.017988967895508,
      "epoch": 0.24761810242159588,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004997509723755547,
      "loss": 5.8205,
      "mean_token_accuracy": 0.13056738600134848,
      "num_tokens": 5487185.0,
      "step": 2495
    },
    {
      "entropy": 5.945784282684326,
      "epoch": 0.2481143310837634,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997493030448662,
      "loss": 5.8411,
      "mean_token_accuracy": 0.1261320762336254,
      "num_tokens": 5498802.0,
      "step": 2500
    },
    {
      "entropy": 5.9388504981994625,
      "epoch": 0.24861055974593094,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997476281408793,
      "loss": 5.855,
      "mean_token_accuracy": 0.12759978100657463,
      "num_tokens": 5510285.0,
      "step": 2505
    },
    {
      "entropy": 6.00165810585022,
      "epoch": 0.24910678840809844,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997459476636354,
      "loss": 5.7335,
      "mean_token_accuracy": 0.13755400031805037,
      "num_tokens": 5520028.0,
      "step": 2510
    },
    {
      "entropy": 5.905316925048828,
      "epoch": 0.24960301707026597,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004997442616131763,
      "loss": 5.85,
      "mean_token_accuracy": 0.12924634143710137,
      "num_tokens": 5529271.0,
      "step": 2515
    },
    {
      "entropy": 5.97735447883606,
      "epoch": 0.25009924573243353,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004997425699895438,
      "loss": 5.7753,
      "mean_token_accuracy": 0.12766662314534188,
      "num_tokens": 5539396.0,
      "step": 2520
    },
    {
      "entropy": 6.006867456436157,
      "epoch": 0.25059547439460106,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004997408727927798,
      "loss": 5.9196,
      "mean_token_accuracy": 0.13126150593161584,
      "num_tokens": 5551674.0,
      "step": 2525
    },
    {
      "entropy": 5.974583911895752,
      "epoch": 0.25109170305676853,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997391700229264,
      "loss": 5.8413,
      "mean_token_accuracy": 0.1326450914144516,
      "num_tokens": 5562749.0,
      "step": 2530
    },
    {
      "entropy": 5.960433483123779,
      "epoch": 0.25158793171893606,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499737461680026,
      "loss": 5.854,
      "mean_token_accuracy": 0.134577427059412,
      "num_tokens": 5573478.0,
      "step": 2535
    },
    {
      "entropy": 5.884474802017212,
      "epoch": 0.2520841603811036,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997357477641207,
      "loss": 5.8349,
      "mean_token_accuracy": 0.1368376225233078,
      "num_tokens": 5584546.0,
      "step": 2540
    },
    {
      "entropy": 5.904690790176391,
      "epoch": 0.2525803890432711,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997340282752532,
      "loss": 5.7177,
      "mean_token_accuracy": 0.13633103296160698,
      "num_tokens": 5594004.0,
      "step": 2545
    },
    {
      "entropy": 5.9720470905303955,
      "epoch": 0.25307661770543866,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499732303213466,
      "loss": 5.7943,
      "mean_token_accuracy": 0.125521769374609,
      "num_tokens": 5606095.0,
      "step": 2550
    },
    {
      "entropy": 5.999422073364258,
      "epoch": 0.2535728463676062,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004997305725788019,
      "loss": 5.813,
      "mean_token_accuracy": 0.12604894265532493,
      "num_tokens": 5615596.0,
      "step": 2555
    },
    {
      "entropy": 5.867432165145874,
      "epoch": 0.2540690750297737,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499728836371304,
      "loss": 5.771,
      "mean_token_accuracy": 0.13205990493297576,
      "num_tokens": 5626753.0,
      "step": 2560
    },
    {
      "entropy": 5.866491079330444,
      "epoch": 0.25456530369194125,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997270945910151,
      "loss": 5.7828,
      "mean_token_accuracy": 0.12868185043334962,
      "num_tokens": 5636135.0,
      "step": 2565
    },
    {
      "entropy": 6.046658754348755,
      "epoch": 0.2550615323541088,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997253472379785,
      "loss": 5.8504,
      "mean_token_accuracy": 0.1282991424202919,
      "num_tokens": 5647074.0,
      "step": 2570
    },
    {
      "entropy": 5.9704625606536865,
      "epoch": 0.2555577610162763,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004997235943122375,
      "loss": 5.8703,
      "mean_token_accuracy": 0.12940976694226264,
      "num_tokens": 5658600.0,
      "step": 2575
    },
    {
      "entropy": 5.924697065353394,
      "epoch": 0.25605398967844384,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997218358138356,
      "loss": 5.7866,
      "mean_token_accuracy": 0.13074240386486052,
      "num_tokens": 5668453.0,
      "step": 2580
    },
    {
      "entropy": 5.984611177444458,
      "epoch": 0.25655021834061137,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004997200717428165,
      "loss": 5.8295,
      "mean_token_accuracy": 0.12661149725317955,
      "num_tokens": 5678855.0,
      "step": 2585
    },
    {
      "entropy": 5.916502237319946,
      "epoch": 0.2570464470027789,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997183020992237,
      "loss": 5.8903,
      "mean_token_accuracy": 0.12172176912426949,
      "num_tokens": 5690063.0,
      "step": 2590
    },
    {
      "entropy": 5.936478853225708,
      "epoch": 0.25754267566494643,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997165268831014,
      "loss": 5.7896,
      "mean_token_accuracy": 0.13551302403211593,
      "num_tokens": 5701487.0,
      "step": 2595
    },
    {
      "entropy": 5.903074455261231,
      "epoch": 0.25803890432711396,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997147460944933,
      "loss": 5.7725,
      "mean_token_accuracy": 0.1349751755595207,
      "num_tokens": 5713034.0,
      "step": 2600
    },
    {
      "entropy": 6.014874839782715,
      "epoch": 0.25853513298928144,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004997129597334438,
      "loss": 5.8224,
      "mean_token_accuracy": 0.13061220347881317,
      "num_tokens": 5725469.0,
      "step": 2605
    },
    {
      "entropy": 5.862464809417725,
      "epoch": 0.25903136165144897,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004997111677999971,
      "loss": 5.812,
      "mean_token_accuracy": 0.12882769852876663,
      "num_tokens": 5737099.0,
      "step": 2610
    },
    {
      "entropy": 5.97240686416626,
      "epoch": 0.2595275903136165,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004997093702941976,
      "loss": 5.8214,
      "mean_token_accuracy": 0.12686606049537658,
      "num_tokens": 5748756.0,
      "step": 2615
    },
    {
      "entropy": 5.980477237701416,
      "epoch": 0.260023818975784,
      "grad_norm": 1.046875,
      "learning_rate": 0.00049970756721609,
      "loss": 5.8099,
      "mean_token_accuracy": 0.12726252228021623,
      "num_tokens": 5759598.0,
      "step": 2620
    },
    {
      "entropy": 5.953548765182495,
      "epoch": 0.26052004763795156,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000499705758565719,
      "loss": 5.854,
      "mean_token_accuracy": 0.12930647283792496,
      "num_tokens": 5770397.0,
      "step": 2625
    },
    {
      "entropy": 6.005121231079102,
      "epoch": 0.2610162763001191,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997039443431292,
      "loss": 5.9263,
      "mean_token_accuracy": 0.12437670230865479,
      "num_tokens": 5780935.0,
      "step": 2630
    },
    {
      "entropy": 6.0060749530792235,
      "epoch": 0.2615125049622866,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004997021245483659,
      "loss": 5.7808,
      "mean_token_accuracy": 0.13181432709097862,
      "num_tokens": 5791557.0,
      "step": 2635
    },
    {
      "entropy": 5.898467540740967,
      "epoch": 0.26200873362445415,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499700299181474,
      "loss": 5.8199,
      "mean_token_accuracy": 0.1286296509206295,
      "num_tokens": 5802233.0,
      "step": 2640
    },
    {
      "entropy": 5.9045013904571535,
      "epoch": 0.2625049622866217,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004996984682424989,
      "loss": 5.7429,
      "mean_token_accuracy": 0.13722301945090293,
      "num_tokens": 5813424.0,
      "step": 2645
    },
    {
      "entropy": 5.888933420181274,
      "epoch": 0.2630011909487892,
      "grad_norm": 0.7890625,
      "learning_rate": 0.000499696631731486,
      "loss": 5.7238,
      "mean_token_accuracy": 0.13307972252368927,
      "num_tokens": 5825652.0,
      "step": 2650
    },
    {
      "entropy": 5.953473615646362,
      "epoch": 0.26349741961095674,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996947896484809,
      "loss": 5.8126,
      "mean_token_accuracy": 0.13535162359476088,
      "num_tokens": 5836603.0,
      "step": 2655
    },
    {
      "entropy": 5.948557996749878,
      "epoch": 0.26399364827312427,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499692941993529,
      "loss": 5.8636,
      "mean_token_accuracy": 0.1336210548877716,
      "num_tokens": 5849365.0,
      "step": 2660
    },
    {
      "entropy": 5.975414085388183,
      "epoch": 0.2644898769352918,
      "grad_norm": 0.875,
      "learning_rate": 0.0004996910887666765,
      "loss": 5.7797,
      "mean_token_accuracy": 0.1360326051712036,
      "num_tokens": 5860305.0,
      "step": 2665
    },
    {
      "entropy": 5.86191725730896,
      "epoch": 0.26498610559745933,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000499689229967969,
      "loss": 5.7249,
      "mean_token_accuracy": 0.1364116244018078,
      "num_tokens": 5872272.0,
      "step": 2670
    },
    {
      "entropy": 5.931841802597046,
      "epoch": 0.26548233425962686,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996873655974529,
      "loss": 5.7536,
      "mean_token_accuracy": 0.13195567727088928,
      "num_tokens": 5884107.0,
      "step": 2675
    },
    {
      "entropy": 5.839572381973267,
      "epoch": 0.26597856292179434,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996854956551743,
      "loss": 5.7433,
      "mean_token_accuracy": 0.13461884558200837,
      "num_tokens": 5894060.0,
      "step": 2680
    },
    {
      "entropy": 5.867028617858887,
      "epoch": 0.26647479158396187,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996836201411794,
      "loss": 5.7661,
      "mean_token_accuracy": 0.13390844464302062,
      "num_tokens": 5904804.0,
      "step": 2685
    },
    {
      "entropy": 6.046043062210083,
      "epoch": 0.2669710202461294,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499681739055515,
      "loss": 5.8361,
      "mean_token_accuracy": 0.12845558002591134,
      "num_tokens": 5916496.0,
      "step": 2690
    },
    {
      "entropy": 5.9300178527832035,
      "epoch": 0.26746724890829693,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996798523982276,
      "loss": 5.7818,
      "mean_token_accuracy": 0.1285025343298912,
      "num_tokens": 5926920.0,
      "step": 2695
    },
    {
      "entropy": 5.965683031082153,
      "epoch": 0.26796347757046446,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499677960169364,
      "loss": 5.8865,
      "mean_token_accuracy": 0.12686679884791374,
      "num_tokens": 5940098.0,
      "step": 2700
    },
    {
      "entropy": 5.910781335830689,
      "epoch": 0.268459706232632,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996760623689711,
      "loss": 5.7374,
      "mean_token_accuracy": 0.13134501054883002,
      "num_tokens": 5949687.0,
      "step": 2705
    },
    {
      "entropy": 5.929884958267212,
      "epoch": 0.2689559348947995,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000499674158997096,
      "loss": 5.8442,
      "mean_token_accuracy": 0.12607498317956925,
      "num_tokens": 5960777.0,
      "step": 2710
    },
    {
      "entropy": 5.978472423553467,
      "epoch": 0.26945216355696705,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004996722500537859,
      "loss": 5.7503,
      "mean_token_accuracy": 0.13114108070731162,
      "num_tokens": 5972298.0,
      "step": 2715
    },
    {
      "entropy": 5.902798414230347,
      "epoch": 0.2699483922191346,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004996703355390881,
      "loss": 5.8696,
      "mean_token_accuracy": 0.12787102088332175,
      "num_tokens": 5985613.0,
      "step": 2720
    },
    {
      "entropy": 5.998373651504517,
      "epoch": 0.2704446208813021,
      "grad_norm": 0.9375,
      "learning_rate": 0.00049966841545305,
      "loss": 5.9368,
      "mean_token_accuracy": 0.1289162516593933,
      "num_tokens": 5997530.0,
      "step": 2725
    },
    {
      "entropy": 5.979059410095215,
      "epoch": 0.27094084954346964,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004996664897957195,
      "loss": 5.7906,
      "mean_token_accuracy": 0.12967565059661865,
      "num_tokens": 6008123.0,
      "step": 2730
    },
    {
      "entropy": 5.898594331741333,
      "epoch": 0.2714370782056372,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004996645585671441,
      "loss": 5.8159,
      "mean_token_accuracy": 0.13189680576324464,
      "num_tokens": 6020578.0,
      "step": 2735
    },
    {
      "entropy": 5.885920572280884,
      "epoch": 0.2719333068678047,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004996626217673717,
      "loss": 5.7329,
      "mean_token_accuracy": 0.13771061301231385,
      "num_tokens": 6031625.0,
      "step": 2740
    },
    {
      "entropy": 5.859991264343262,
      "epoch": 0.27242953552997223,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004996606793964504,
      "loss": 5.771,
      "mean_token_accuracy": 0.13899061530828477,
      "num_tokens": 6041320.0,
      "step": 2745
    },
    {
      "entropy": 5.95008225440979,
      "epoch": 0.27292576419213976,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004996587314544284,
      "loss": 5.793,
      "mean_token_accuracy": 0.13212788924574853,
      "num_tokens": 6052627.0,
      "step": 2750
    },
    {
      "entropy": 5.905590677261353,
      "epoch": 0.27342199285430724,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004996567779413539,
      "loss": 5.7442,
      "mean_token_accuracy": 0.13712375462055207,
      "num_tokens": 6063468.0,
      "step": 2755
    },
    {
      "entropy": 5.918176126480103,
      "epoch": 0.27391822151647477,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004996548188572755,
      "loss": 5.7254,
      "mean_token_accuracy": 0.13772219493985177,
      "num_tokens": 6074208.0,
      "step": 2760
    },
    {
      "entropy": 5.93715763092041,
      "epoch": 0.2744144501786423,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004996528542022416,
      "loss": 5.7761,
      "mean_token_accuracy": 0.1317790240049362,
      "num_tokens": 6085068.0,
      "step": 2765
    },
    {
      "entropy": 5.8894490718841555,
      "epoch": 0.27491067884080983,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000499650883976301,
      "loss": 5.7723,
      "mean_token_accuracy": 0.13286536261439325,
      "num_tokens": 6096445.0,
      "step": 2770
    },
    {
      "entropy": 5.8424560070037845,
      "epoch": 0.27540690750297736,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004996489081795026,
      "loss": 5.6885,
      "mean_token_accuracy": 0.137538892775774,
      "num_tokens": 6106527.0,
      "step": 2775
    },
    {
      "entropy": 5.817946577072144,
      "epoch": 0.2759031361651449,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996469268118953,
      "loss": 5.7127,
      "mean_token_accuracy": 0.14144881442189217,
      "num_tokens": 6115923.0,
      "step": 2780
    },
    {
      "entropy": 5.838368940353393,
      "epoch": 0.2763993648273124,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004996449398735283,
      "loss": 5.6099,
      "mean_token_accuracy": 0.14129377156496048,
      "num_tokens": 6125861.0,
      "step": 2785
    },
    {
      "entropy": 5.945985269546509,
      "epoch": 0.27689559348947995,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004996429473644509,
      "loss": 5.8072,
      "mean_token_accuracy": 0.13425228521227836,
      "num_tokens": 6137189.0,
      "step": 2790
    },
    {
      "entropy": 5.8859100341796875,
      "epoch": 0.2773918221516475,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004996409492847124,
      "loss": 5.7305,
      "mean_token_accuracy": 0.13389328271150588,
      "num_tokens": 6147920.0,
      "step": 2795
    },
    {
      "entropy": 5.892881202697754,
      "epoch": 0.277888050813815,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004996389456343624,
      "loss": 5.7287,
      "mean_token_accuracy": 0.14077215641736984,
      "num_tokens": 6159371.0,
      "step": 2800
    },
    {
      "entropy": 6.015835094451904,
      "epoch": 0.27838427947598254,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004996369364134507,
      "loss": 5.9235,
      "mean_token_accuracy": 0.12242046073079109,
      "num_tokens": 6172679.0,
      "step": 2805
    },
    {
      "entropy": 5.894385576248169,
      "epoch": 0.2788805081381501,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000499634921622027,
      "loss": 5.7718,
      "mean_token_accuracy": 0.12801498249173165,
      "num_tokens": 6184244.0,
      "step": 2810
    },
    {
      "entropy": 5.8594602108001705,
      "epoch": 0.2793767368003176,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004996329012601413,
      "loss": 5.6501,
      "mean_token_accuracy": 0.1308717854321003,
      "num_tokens": 6194826.0,
      "step": 2815
    },
    {
      "entropy": 5.934268379211426,
      "epoch": 0.27987296546248513,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004996308753278437,
      "loss": 5.7955,
      "mean_token_accuracy": 0.13045783266425132,
      "num_tokens": 6206808.0,
      "step": 2820
    },
    {
      "entropy": 5.903773975372315,
      "epoch": 0.28036919412465267,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996288438251843,
      "loss": 5.8036,
      "mean_token_accuracy": 0.1288069613277912,
      "num_tokens": 6217303.0,
      "step": 2825
    },
    {
      "entropy": 5.918454885482788,
      "epoch": 0.28086542278682014,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996268067522137,
      "loss": 5.7907,
      "mean_token_accuracy": 0.13076610937714578,
      "num_tokens": 6227965.0,
      "step": 2830
    },
    {
      "entropy": 6.045854377746582,
      "epoch": 0.28136165144898767,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996247641089823,
      "loss": 5.8534,
      "mean_token_accuracy": 0.12538014501333236,
      "num_tokens": 6239011.0,
      "step": 2835
    },
    {
      "entropy": 5.820587921142578,
      "epoch": 0.2818578801111552,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996227158955409,
      "loss": 5.7089,
      "mean_token_accuracy": 0.13381239399313927,
      "num_tokens": 6249084.0,
      "step": 2840
    },
    {
      "entropy": 5.854635334014892,
      "epoch": 0.28235410877332273,
      "grad_norm": 0.96875,
      "learning_rate": 0.00049962066211194,
      "loss": 5.7811,
      "mean_token_accuracy": 0.12858823388814927,
      "num_tokens": 6260706.0,
      "step": 2845
    },
    {
      "entropy": 5.910700273513794,
      "epoch": 0.28285033743549026,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996186027582308,
      "loss": 5.719,
      "mean_token_accuracy": 0.13755066841840743,
      "num_tokens": 6272848.0,
      "step": 2850
    },
    {
      "entropy": 5.831938123703003,
      "epoch": 0.2833465660976578,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004996165378344642,
      "loss": 5.7198,
      "mean_token_accuracy": 0.1345575749874115,
      "num_tokens": 6283820.0,
      "step": 2855
    },
    {
      "entropy": 5.887689590454102,
      "epoch": 0.2838427947598253,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004996144673406914,
      "loss": 5.8847,
      "mean_token_accuracy": 0.1203585371375084,
      "num_tokens": 6294193.0,
      "step": 2860
    },
    {
      "entropy": 5.846189975738525,
      "epoch": 0.28433902342199285,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996123912769639,
      "loss": 5.7111,
      "mean_token_accuracy": 0.14153243526816367,
      "num_tokens": 6304407.0,
      "step": 2865
    },
    {
      "entropy": 5.857138013839721,
      "epoch": 0.2848352520841604,
      "grad_norm": 0.875,
      "learning_rate": 0.0004996103096433332,
      "loss": 5.7866,
      "mean_token_accuracy": 0.13395510241389275,
      "num_tokens": 6316311.0,
      "step": 2870
    },
    {
      "entropy": 5.944645738601684,
      "epoch": 0.2853314807463279,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004996082224398506,
      "loss": 5.7471,
      "mean_token_accuracy": 0.13548968210816384,
      "num_tokens": 6327248.0,
      "step": 2875
    },
    {
      "entropy": 5.888983535766601,
      "epoch": 0.28582770940849545,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004996061296665682,
      "loss": 5.7629,
      "mean_token_accuracy": 0.13373192697763442,
      "num_tokens": 6338494.0,
      "step": 2880
    },
    {
      "entropy": 5.86532621383667,
      "epoch": 0.286323938070663,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996040313235377,
      "loss": 5.6606,
      "mean_token_accuracy": 0.1401577278971672,
      "num_tokens": 6349038.0,
      "step": 2885
    },
    {
      "entropy": 5.888543939590454,
      "epoch": 0.2868201667328305,
      "grad_norm": 0.875,
      "learning_rate": 0.0004996019274108112,
      "loss": 5.7737,
      "mean_token_accuracy": 0.12937842831015586,
      "num_tokens": 6360180.0,
      "step": 2890
    },
    {
      "entropy": 5.789392042160034,
      "epoch": 0.28731639539499804,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499599817928441,
      "loss": 5.7128,
      "mean_token_accuracy": 0.13721143081784248,
      "num_tokens": 6370931.0,
      "step": 2895
    },
    {
      "entropy": 5.796161222457886,
      "epoch": 0.28781262405716557,
      "grad_norm": 0.875,
      "learning_rate": 0.0004995977028764792,
      "loss": 5.7034,
      "mean_token_accuracy": 0.14205339923501015,
      "num_tokens": 6381810.0,
      "step": 2900
    },
    {
      "entropy": 5.8964122295379635,
      "epoch": 0.28830885271933304,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004995955822549783,
      "loss": 5.7075,
      "mean_token_accuracy": 0.1379415936768055,
      "num_tokens": 6392268.0,
      "step": 2905
    },
    {
      "entropy": 5.881337690353393,
      "epoch": 0.2888050813815006,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004995934560639909,
      "loss": 5.7543,
      "mean_token_accuracy": 0.132367368042469,
      "num_tokens": 6403650.0,
      "step": 2910
    },
    {
      "entropy": 5.904406881332397,
      "epoch": 0.2893013100436681,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004995913243035699,
      "loss": 5.7468,
      "mean_token_accuracy": 0.13186144679784775,
      "num_tokens": 6414672.0,
      "step": 2915
    },
    {
      "entropy": 5.880802202224731,
      "epoch": 0.28979753870583563,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995891869737678,
      "loss": 5.8088,
      "mean_token_accuracy": 0.1321220226585865,
      "num_tokens": 6424916.0,
      "step": 2920
    },
    {
      "entropy": 5.9205621719360355,
      "epoch": 0.29029376736800316,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995870440746379,
      "loss": 5.7555,
      "mean_token_accuracy": 0.13907312899827956,
      "num_tokens": 6435950.0,
      "step": 2925
    },
    {
      "entropy": 5.813934850692749,
      "epoch": 0.2907899960301707,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004995848956062333,
      "loss": 5.7223,
      "mean_token_accuracy": 0.1334433563053608,
      "num_tokens": 6446336.0,
      "step": 2930
    },
    {
      "entropy": 5.862525606155396,
      "epoch": 0.2912862246923382,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004995827415686071,
      "loss": 5.7939,
      "mean_token_accuracy": 0.1313887841999531,
      "num_tokens": 6457655.0,
      "step": 2935
    },
    {
      "entropy": 5.933152437210083,
      "epoch": 0.29178245335450576,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499580581961813,
      "loss": 5.6738,
      "mean_token_accuracy": 0.1359466515481472,
      "num_tokens": 6469627.0,
      "step": 2940
    },
    {
      "entropy": 5.85411057472229,
      "epoch": 0.2922786820166733,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004995784167859043,
      "loss": 5.7708,
      "mean_token_accuracy": 0.1325305663049221,
      "num_tokens": 6480351.0,
      "step": 2945
    },
    {
      "entropy": 5.930560064315796,
      "epoch": 0.2927749106788408,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004995762460409349,
      "loss": 5.7976,
      "mean_token_accuracy": 0.1347188226878643,
      "num_tokens": 6491651.0,
      "step": 2950
    },
    {
      "entropy": 5.843895006179809,
      "epoch": 0.29327113934100835,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004995740697269584,
      "loss": 5.6555,
      "mean_token_accuracy": 0.13501327559351922,
      "num_tokens": 6502531.0,
      "step": 2955
    },
    {
      "entropy": 5.776656246185302,
      "epoch": 0.2937673680031759,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004995718878440289,
      "loss": 5.6508,
      "mean_token_accuracy": 0.13918386697769164,
      "num_tokens": 6511938.0,
      "step": 2960
    },
    {
      "entropy": 5.909338426589966,
      "epoch": 0.2942635966653434,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004995697003922006,
      "loss": 5.8071,
      "mean_token_accuracy": 0.1281385228037834,
      "num_tokens": 6522775.0,
      "step": 2965
    },
    {
      "entropy": 5.889328289031982,
      "epoch": 0.29475982532751094,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004995675073715274,
      "loss": 5.6384,
      "mean_token_accuracy": 0.1326303191483021,
      "num_tokens": 6534660.0,
      "step": 2970
    },
    {
      "entropy": 5.867802667617798,
      "epoch": 0.29525605398967847,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004995653087820641,
      "loss": 5.7662,
      "mean_token_accuracy": 0.1363363467156887,
      "num_tokens": 6545673.0,
      "step": 2975
    },
    {
      "entropy": 5.872935438156128,
      "epoch": 0.29575228265184594,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499563104623865,
      "loss": 5.7364,
      "mean_token_accuracy": 0.13667785823345185,
      "num_tokens": 6556912.0,
      "step": 2980
    },
    {
      "entropy": 5.885501098632813,
      "epoch": 0.2962485113140135,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004995608948969849,
      "loss": 5.8271,
      "mean_token_accuracy": 0.13097959756851196,
      "num_tokens": 6567532.0,
      "step": 2985
    },
    {
      "entropy": 5.90928168296814,
      "epoch": 0.296744739976181,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004995586796014784,
      "loss": 5.796,
      "mean_token_accuracy": 0.13155989721417427,
      "num_tokens": 6577830.0,
      "step": 2990
    },
    {
      "entropy": 5.823900842666626,
      "epoch": 0.29724096863834853,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004995564587374005,
      "loss": 5.6926,
      "mean_token_accuracy": 0.13435450717806816,
      "num_tokens": 6587254.0,
      "step": 2995
    },
    {
      "entropy": 5.902713966369629,
      "epoch": 0.29773719730051607,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004995542323048063,
      "loss": 5.8848,
      "mean_token_accuracy": 0.12699431478977202,
      "num_tokens": 6598294.0,
      "step": 3000
    },
    {
      "epoch": 0.29773719730051607,
      "eval_entropy": 5.728751434966033,
      "eval_loss": 5.813499450683594,
      "eval_mean_token_accuracy": 0.1359337329577737,
      "eval_num_tokens": 6598294.0,
      "eval_runtime": 20.8245,
      "eval_samples_per_second": 1616.798,
      "eval_steps_per_second": 202.118,
      "step": 3000
    },
    {
      "entropy": 5.949578332901001,
      "epoch": 0.2982334259626836,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499552000303751,
      "loss": 5.7558,
      "mean_token_accuracy": 0.13457091972231866,
      "num_tokens": 6608772.0,
      "step": 3005
    },
    {
      "entropy": 5.853357172012329,
      "epoch": 0.2987296546248511,
      "grad_norm": 0.90625,
      "learning_rate": 0.00049954976273429,
      "loss": 5.7668,
      "mean_token_accuracy": 0.13631939515471458,
      "num_tokens": 6619973.0,
      "step": 3010
    },
    {
      "entropy": 5.872531938552856,
      "epoch": 0.29922588328701866,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004995475195964787,
      "loss": 5.782,
      "mean_token_accuracy": 0.13367341533303262,
      "num_tokens": 6631172.0,
      "step": 3015
    },
    {
      "entropy": 5.848301219940185,
      "epoch": 0.2997221119491862,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004995452708903727,
      "loss": 5.7763,
      "mean_token_accuracy": 0.13015636652708054,
      "num_tokens": 6641343.0,
      "step": 3020
    },
    {
      "entropy": 5.876894283294678,
      "epoch": 0.3002183406113537,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499543016616028,
      "loss": 5.7196,
      "mean_token_accuracy": 0.1363053061068058,
      "num_tokens": 6652892.0,
      "step": 3025
    },
    {
      "entropy": 5.898422145843506,
      "epoch": 0.30071456927352125,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004995407567735002,
      "loss": 5.7331,
      "mean_token_accuracy": 0.13731808662414552,
      "num_tokens": 6664177.0,
      "step": 3030
    },
    {
      "entropy": 5.860565662384033,
      "epoch": 0.3012107979356888,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995384913628456,
      "loss": 5.6982,
      "mean_token_accuracy": 0.13397607430815697,
      "num_tokens": 6674402.0,
      "step": 3035
    },
    {
      "entropy": 5.862356185913086,
      "epoch": 0.3017070265978563,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004995362203841202,
      "loss": 5.8452,
      "mean_token_accuracy": 0.12625229880213737,
      "num_tokens": 6684886.0,
      "step": 3040
    },
    {
      "entropy": 5.876009559631347,
      "epoch": 0.30220325526002384,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995339438373804,
      "loss": 5.6665,
      "mean_token_accuracy": 0.1372012011706829,
      "num_tokens": 6694774.0,
      "step": 3045
    },
    {
      "entropy": 5.883480262756348,
      "epoch": 0.30269948392219137,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004995316617226825,
      "loss": 5.7381,
      "mean_token_accuracy": 0.1354772590100765,
      "num_tokens": 6705346.0,
      "step": 3050
    },
    {
      "entropy": 5.802752208709717,
      "epoch": 0.30319571258435885,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004995293740400833,
      "loss": 5.6781,
      "mean_token_accuracy": 0.14116692245006562,
      "num_tokens": 6715579.0,
      "step": 3055
    },
    {
      "entropy": 5.831467723846435,
      "epoch": 0.3036919412465264,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004995270807896395,
      "loss": 5.6444,
      "mean_token_accuracy": 0.14167544841766358,
      "num_tokens": 6726382.0,
      "step": 3060
    },
    {
      "entropy": 5.8336011409759525,
      "epoch": 0.3041881699086939,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004995247819714079,
      "loss": 5.7137,
      "mean_token_accuracy": 0.13432926014065744,
      "num_tokens": 6737245.0,
      "step": 3065
    },
    {
      "entropy": 5.889817237854004,
      "epoch": 0.30468439857086144,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995224775854454,
      "loss": 5.9247,
      "mean_token_accuracy": 0.12740089893341064,
      "num_tokens": 6749113.0,
      "step": 3070
    },
    {
      "entropy": 6.012298154830932,
      "epoch": 0.30518062723302897,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004995201676318093,
      "loss": 5.7805,
      "mean_token_accuracy": 0.13419591411948203,
      "num_tokens": 6761987.0,
      "step": 3075
    },
    {
      "entropy": 5.858676910400391,
      "epoch": 0.3056768558951965,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004995178521105568,
      "loss": 5.7065,
      "mean_token_accuracy": 0.13019177466630935,
      "num_tokens": 6771279.0,
      "step": 3080
    },
    {
      "entropy": 5.801815176010132,
      "epoch": 0.30617308455736403,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004995155310217454,
      "loss": 5.6501,
      "mean_token_accuracy": 0.13885203823447229,
      "num_tokens": 6781038.0,
      "step": 3085
    },
    {
      "entropy": 5.84817328453064,
      "epoch": 0.30666931321953156,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004995132043654327,
      "loss": 5.7956,
      "mean_token_accuracy": 0.13456297814846038,
      "num_tokens": 6792686.0,
      "step": 3090
    },
    {
      "entropy": 5.865736818313598,
      "epoch": 0.3071655418816991,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004995108721416763,
      "loss": 5.7511,
      "mean_token_accuracy": 0.1393015667796135,
      "num_tokens": 6804414.0,
      "step": 3095
    },
    {
      "entropy": 5.889206123352051,
      "epoch": 0.3076617705438666,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004995085343505339,
      "loss": 5.7185,
      "mean_token_accuracy": 0.13796254768967628,
      "num_tokens": 6816180.0,
      "step": 3100
    },
    {
      "entropy": 5.851957178115844,
      "epoch": 0.30815799920603415,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004995061909920638,
      "loss": 5.7194,
      "mean_token_accuracy": 0.13511088341474534,
      "num_tokens": 6827114.0,
      "step": 3105
    },
    {
      "entropy": 5.898450088500977,
      "epoch": 0.3086542278682017,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995038420663238,
      "loss": 5.7754,
      "mean_token_accuracy": 0.13403737246990205,
      "num_tokens": 6836953.0,
      "step": 3110
    },
    {
      "entropy": 5.8968480110168455,
      "epoch": 0.3091504565303692,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995014875733724,
      "loss": 5.8208,
      "mean_token_accuracy": 0.12903424575924874,
      "num_tokens": 6847963.0,
      "step": 3115
    },
    {
      "entropy": 5.90019793510437,
      "epoch": 0.30964668519253674,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004994991275132678,
      "loss": 5.8363,
      "mean_token_accuracy": 0.13068509995937347,
      "num_tokens": 6858264.0,
      "step": 3120
    },
    {
      "entropy": 5.890040588378906,
      "epoch": 0.31014291385470427,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004994967618860685,
      "loss": 5.725,
      "mean_token_accuracy": 0.13812246397137642,
      "num_tokens": 6868861.0,
      "step": 3125
    },
    {
      "entropy": 5.879378128051758,
      "epoch": 0.31063914251687175,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004994943906918333,
      "loss": 5.7576,
      "mean_token_accuracy": 0.1395530015230179,
      "num_tokens": 6879797.0,
      "step": 3130
    },
    {
      "entropy": 5.851549291610718,
      "epoch": 0.3111353711790393,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004994920139306211,
      "loss": 5.6931,
      "mean_token_accuracy": 0.1380985639989376,
      "num_tokens": 6891323.0,
      "step": 3135
    },
    {
      "entropy": 5.807861804962158,
      "epoch": 0.3116315998412068,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004994896316024906,
      "loss": 5.7844,
      "mean_token_accuracy": 0.13645399808883668,
      "num_tokens": 6903345.0,
      "step": 3140
    },
    {
      "entropy": 5.850348901748657,
      "epoch": 0.31212782850337434,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499487243707501,
      "loss": 5.7002,
      "mean_token_accuracy": 0.1344092607498169,
      "num_tokens": 6914342.0,
      "step": 3145
    },
    {
      "entropy": 5.878314256668091,
      "epoch": 0.31262405716554187,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004994848502457114,
      "loss": 5.7757,
      "mean_token_accuracy": 0.13398355096578599,
      "num_tokens": 6926691.0,
      "step": 3150
    },
    {
      "entropy": 5.797974109649658,
      "epoch": 0.3131202858277094,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004994824512171812,
      "loss": 5.6434,
      "mean_token_accuracy": 0.1319235309958458,
      "num_tokens": 6937841.0,
      "step": 3155
    },
    {
      "entropy": 5.810340785980225,
      "epoch": 0.31361651448987693,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994800466219701,
      "loss": 5.7249,
      "mean_token_accuracy": 0.13448656126856803,
      "num_tokens": 6948601.0,
      "step": 3160
    },
    {
      "entropy": 5.866660022735596,
      "epoch": 0.31411274315204446,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004994776364601374,
      "loss": 5.663,
      "mean_token_accuracy": 0.13846807479858397,
      "num_tokens": 6959775.0,
      "step": 3165
    },
    {
      "entropy": 5.769833517074585,
      "epoch": 0.314608971814212,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004994752207317431,
      "loss": 5.6916,
      "mean_token_accuracy": 0.13938113674521446,
      "num_tokens": 6969202.0,
      "step": 3170
    },
    {
      "entropy": 5.793140506744384,
      "epoch": 0.3151052004763795,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499472799436847,
      "loss": 5.6714,
      "mean_token_accuracy": 0.1376807488501072,
      "num_tokens": 6978633.0,
      "step": 3175
    },
    {
      "entropy": 5.9285777568817135,
      "epoch": 0.31560142913854705,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004994703725755092,
      "loss": 5.731,
      "mean_token_accuracy": 0.13278935626149177,
      "num_tokens": 6989638.0,
      "step": 3180
    },
    {
      "entropy": 5.860269212722779,
      "epoch": 0.3160976578007146,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994679401477899,
      "loss": 5.7104,
      "mean_token_accuracy": 0.13575214371085167,
      "num_tokens": 7001606.0,
      "step": 3185
    },
    {
      "entropy": 5.792490386962891,
      "epoch": 0.3165938864628821,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994655021537492,
      "loss": 5.6791,
      "mean_token_accuracy": 0.13449148833751678,
      "num_tokens": 7012915.0,
      "step": 3190
    },
    {
      "entropy": 5.800523090362549,
      "epoch": 0.31709011512504964,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004994630585934478,
      "loss": 5.7451,
      "mean_token_accuracy": 0.13046567216515542,
      "num_tokens": 7023191.0,
      "step": 3195
    },
    {
      "entropy": 5.797707414627075,
      "epoch": 0.3175863437872172,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004994606094669463,
      "loss": 5.6272,
      "mean_token_accuracy": 0.14198058173060418,
      "num_tokens": 7032506.0,
      "step": 3200
    },
    {
      "entropy": 5.8543891429901125,
      "epoch": 0.3180825724493847,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004994581547743053,
      "loss": 5.7768,
      "mean_token_accuracy": 0.12855609208345414,
      "num_tokens": 7046098.0,
      "step": 3205
    },
    {
      "entropy": 5.933334302902222,
      "epoch": 0.3185788011115522,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004994556945155857,
      "loss": 5.7842,
      "mean_token_accuracy": 0.13946380093693733,
      "num_tokens": 7057434.0,
      "step": 3210
    },
    {
      "entropy": 5.781634664535522,
      "epoch": 0.3190750297737197,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004994532286908486,
      "loss": 5.7094,
      "mean_token_accuracy": 0.1315760038793087,
      "num_tokens": 7067674.0,
      "step": 3215
    },
    {
      "entropy": 5.888899612426758,
      "epoch": 0.31957125843588724,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000499450757300155,
      "loss": 5.7672,
      "mean_token_accuracy": 0.1315600000321865,
      "num_tokens": 7077898.0,
      "step": 3220
    },
    {
      "entropy": 5.922313070297241,
      "epoch": 0.32006748709805477,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004994482803435664,
      "loss": 5.7324,
      "mean_token_accuracy": 0.13344367668032647,
      "num_tokens": 7088566.0,
      "step": 3225
    },
    {
      "entropy": 5.8241376876831055,
      "epoch": 0.3205637157602223,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499445797821144,
      "loss": 5.7409,
      "mean_token_accuracy": 0.13400137275457383,
      "num_tokens": 7099533.0,
      "step": 3230
    },
    {
      "entropy": 5.858956956863404,
      "epoch": 0.32105994442238983,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004994433097329494,
      "loss": 5.7326,
      "mean_token_accuracy": 0.13736375123262407,
      "num_tokens": 7112254.0,
      "step": 3235
    },
    {
      "entropy": 5.934496688842773,
      "epoch": 0.32155617308455736,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004994408160790444,
      "loss": 5.7663,
      "mean_token_accuracy": 0.1309788353741169,
      "num_tokens": 7124704.0,
      "step": 3240
    },
    {
      "entropy": 5.988117361068726,
      "epoch": 0.3220524017467249,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004994383168594907,
      "loss": 5.931,
      "mean_token_accuracy": 0.1246418759226799,
      "num_tokens": 7137741.0,
      "step": 3245
    },
    {
      "entropy": 5.875287818908691,
      "epoch": 0.3225486304088924,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004994358120743506,
      "loss": 5.7035,
      "mean_token_accuracy": 0.1314437985420227,
      "num_tokens": 7147888.0,
      "step": 3250
    },
    {
      "entropy": 5.739679145812988,
      "epoch": 0.32304485907105995,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004994333017236858,
      "loss": 5.6589,
      "mean_token_accuracy": 0.1450750708580017,
      "num_tokens": 7158950.0,
      "step": 3255
    },
    {
      "entropy": 5.861381769180298,
      "epoch": 0.3235410877332275,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004994307858075587,
      "loss": 5.8056,
      "mean_token_accuracy": 0.13047620952129363,
      "num_tokens": 7170156.0,
      "step": 3260
    },
    {
      "entropy": 5.854081630706787,
      "epoch": 0.324037316395395,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994282643260319,
      "loss": 5.6497,
      "mean_token_accuracy": 0.13866678774356841,
      "num_tokens": 7180576.0,
      "step": 3265
    },
    {
      "entropy": 5.776857566833496,
      "epoch": 0.32453354505756254,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004994257372791676,
      "loss": 5.7051,
      "mean_token_accuracy": 0.1358933262526989,
      "num_tokens": 7192380.0,
      "step": 3270
    },
    {
      "entropy": 5.895743036270142,
      "epoch": 0.3250297737197301,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004994232046670286,
      "loss": 5.7705,
      "mean_token_accuracy": 0.13492169752717018,
      "num_tokens": 7203898.0,
      "step": 3275
    },
    {
      "entropy": 5.9295378684997555,
      "epoch": 0.3255260023818976,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004994206664896778,
      "loss": 5.7211,
      "mean_token_accuracy": 0.13575714454054832,
      "num_tokens": 7215241.0,
      "step": 3280
    },
    {
      "entropy": 5.788207483291626,
      "epoch": 0.3260222310440651,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004994181227471779,
      "loss": 5.6898,
      "mean_token_accuracy": 0.13606705367565156,
      "num_tokens": 7226039.0,
      "step": 3285
    },
    {
      "entropy": 5.8000298023223875,
      "epoch": 0.3265184597062326,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004994155734395923,
      "loss": 5.7174,
      "mean_token_accuracy": 0.13560251593589784,
      "num_tokens": 7237148.0,
      "step": 3290
    },
    {
      "entropy": 5.844682121276856,
      "epoch": 0.32701468836840014,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499413018566984,
      "loss": 5.7074,
      "mean_token_accuracy": 0.13221757486462593,
      "num_tokens": 7249284.0,
      "step": 3295
    },
    {
      "entropy": 5.780368804931641,
      "epoch": 0.32751091703056767,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004994104581294163,
      "loss": 5.7178,
      "mean_token_accuracy": 0.13415811136364936,
      "num_tokens": 7259967.0,
      "step": 3300
    },
    {
      "entropy": 5.746554756164551,
      "epoch": 0.3280071456927352,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004994078921269529,
      "loss": 5.6759,
      "mean_token_accuracy": 0.1414274677634239,
      "num_tokens": 7271145.0,
      "step": 3305
    },
    {
      "entropy": 5.904040813446045,
      "epoch": 0.32850337435490273,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004994053205596573,
      "loss": 5.6909,
      "mean_token_accuracy": 0.13865874335169792,
      "num_tokens": 7281872.0,
      "step": 3310
    },
    {
      "entropy": 5.795816040039062,
      "epoch": 0.32899960301707026,
      "grad_norm": 0.875,
      "learning_rate": 0.0004994027434275934,
      "loss": 5.7103,
      "mean_token_accuracy": 0.13829881325364113,
      "num_tokens": 7293330.0,
      "step": 3315
    },
    {
      "entropy": 5.911486101150513,
      "epoch": 0.3294958316792378,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004994001607308249,
      "loss": 5.7653,
      "mean_token_accuracy": 0.1293144330382347,
      "num_tokens": 7304187.0,
      "step": 3320
    },
    {
      "entropy": 5.881396007537842,
      "epoch": 0.3299920603414053,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000499397572469416,
      "loss": 5.667,
      "mean_token_accuracy": 0.1366066187620163,
      "num_tokens": 7314756.0,
      "step": 3325
    },
    {
      "entropy": 5.746714353561401,
      "epoch": 0.33048828900357285,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004993949786434308,
      "loss": 5.7122,
      "mean_token_accuracy": 0.1379957266151905,
      "num_tokens": 7325711.0,
      "step": 3330
    },
    {
      "entropy": 5.778922271728516,
      "epoch": 0.3309845176657404,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004993923792529337,
      "loss": 5.6361,
      "mean_token_accuracy": 0.14142280891537667,
      "num_tokens": 7335639.0,
      "step": 3335
    },
    {
      "entropy": 5.88486647605896,
      "epoch": 0.3314807463279079,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004993897742979892,
      "loss": 5.7531,
      "mean_token_accuracy": 0.13650422394275666,
      "num_tokens": 7346621.0,
      "step": 3340
    },
    {
      "entropy": 5.86642370223999,
      "epoch": 0.33197697499007545,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004993871637786616,
      "loss": 5.7269,
      "mean_token_accuracy": 0.12956238090991973,
      "num_tokens": 7358950.0,
      "step": 3345
    },
    {
      "entropy": 5.761933898925781,
      "epoch": 0.332473203652243,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499384547695016,
      "loss": 5.5806,
      "mean_token_accuracy": 0.14225346446037293,
      "num_tokens": 7369629.0,
      "step": 3350
    },
    {
      "entropy": 5.821819877624511,
      "epoch": 0.3329694323144105,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499381926047117,
      "loss": 5.6669,
      "mean_token_accuracy": 0.14004866182804107,
      "num_tokens": 7379853.0,
      "step": 3355
    },
    {
      "entropy": 5.812605714797973,
      "epoch": 0.333465660976578,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00049937929883503,
      "loss": 5.7162,
      "mean_token_accuracy": 0.13662909269332885,
      "num_tokens": 7391323.0,
      "step": 3360
    },
    {
      "entropy": 5.839717102050781,
      "epoch": 0.3339618896387455,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004993766660588196,
      "loss": 5.6707,
      "mean_token_accuracy": 0.1323630690574646,
      "num_tokens": 7402201.0,
      "step": 3365
    },
    {
      "entropy": 5.884400796890259,
      "epoch": 0.33445811830091304,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004993740277185515,
      "loss": 5.9218,
      "mean_token_accuracy": 0.12637710124254226,
      "num_tokens": 7413783.0,
      "step": 3370
    },
    {
      "entropy": 5.816495943069458,
      "epoch": 0.3349543469630806,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499371383814291,
      "loss": 5.6916,
      "mean_token_accuracy": 0.1357066087424755,
      "num_tokens": 7425909.0,
      "step": 3375
    },
    {
      "entropy": 5.770580053329468,
      "epoch": 0.3354505756252481,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004993687343461036,
      "loss": 5.5844,
      "mean_token_accuracy": 0.14517004415392876,
      "num_tokens": 7436346.0,
      "step": 3380
    },
    {
      "entropy": 5.808633470535279,
      "epoch": 0.33594680428741563,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004993660793140551,
      "loss": 5.6471,
      "mean_token_accuracy": 0.14266916885972022,
      "num_tokens": 7446596.0,
      "step": 3385
    },
    {
      "entropy": 5.769165945053101,
      "epoch": 0.33644303294958317,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993634187182113,
      "loss": 5.7441,
      "mean_token_accuracy": 0.13104177489876748,
      "num_tokens": 7457697.0,
      "step": 3390
    },
    {
      "entropy": 5.9399810314178465,
      "epoch": 0.3369392616117507,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004993607525586381,
      "loss": 5.8457,
      "mean_token_accuracy": 0.13390295282006265,
      "num_tokens": 7468541.0,
      "step": 3395
    },
    {
      "entropy": 5.8909464359283445,
      "epoch": 0.3374354902739182,
      "grad_norm": 0.875,
      "learning_rate": 0.0004993580808354018,
      "loss": 5.726,
      "mean_token_accuracy": 0.1409744441509247,
      "num_tokens": 7479781.0,
      "step": 3400
    },
    {
      "entropy": 5.798458576202393,
      "epoch": 0.33793171893608576,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004993554035485685,
      "loss": 5.6354,
      "mean_token_accuracy": 0.13420383185148238,
      "num_tokens": 7490722.0,
      "step": 3405
    },
    {
      "entropy": 5.815723514556884,
      "epoch": 0.3384279475982533,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993527206982047,
      "loss": 5.7456,
      "mean_token_accuracy": 0.1327146239578724,
      "num_tokens": 7502293.0,
      "step": 3410
    },
    {
      "entropy": 5.8476334571838375,
      "epoch": 0.3389241762604208,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004993500322843767,
      "loss": 5.6813,
      "mean_token_accuracy": 0.14118870049715043,
      "num_tokens": 7513846.0,
      "step": 3415
    },
    {
      "entropy": 5.772971153259277,
      "epoch": 0.33942040492258835,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004993473383071514,
      "loss": 5.6888,
      "mean_token_accuracy": 0.13772222474217416,
      "num_tokens": 7524062.0,
      "step": 3420
    },
    {
      "entropy": 5.8234320163726805,
      "epoch": 0.3399166335847559,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004993446387665956,
      "loss": 5.7599,
      "mean_token_accuracy": 0.13640008345246316,
      "num_tokens": 7535533.0,
      "step": 3425
    },
    {
      "entropy": 5.834895324707031,
      "epoch": 0.3404128622469234,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004993419336627761,
      "loss": 5.6559,
      "mean_token_accuracy": 0.13569505959749223,
      "num_tokens": 7545951.0,
      "step": 3430
    },
    {
      "entropy": 5.771151876449585,
      "epoch": 0.3409090909090909,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004993392229957602,
      "loss": 5.596,
      "mean_token_accuracy": 0.13585764318704605,
      "num_tokens": 7557320.0,
      "step": 3435
    },
    {
      "entropy": 5.7666747093200685,
      "epoch": 0.3414053195712584,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004993365067656148,
      "loss": 5.708,
      "mean_token_accuracy": 0.13677700236439705,
      "num_tokens": 7568078.0,
      "step": 3440
    },
    {
      "entropy": 5.792825222015381,
      "epoch": 0.34190154823342594,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004993337849724076,
      "loss": 5.6956,
      "mean_token_accuracy": 0.13778189420700074,
      "num_tokens": 7578940.0,
      "step": 3445
    },
    {
      "entropy": 5.818231916427612,
      "epoch": 0.3423977768955935,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004993310576162059,
      "loss": 5.7024,
      "mean_token_accuracy": 0.1383884370326996,
      "num_tokens": 7589864.0,
      "step": 3450
    },
    {
      "entropy": 5.852412557601928,
      "epoch": 0.342894005557761,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004993283246970772,
      "loss": 5.6502,
      "mean_token_accuracy": 0.13423066288232804,
      "num_tokens": 7599795.0,
      "step": 3455
    },
    {
      "entropy": 5.793929576873779,
      "epoch": 0.34339023421992854,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004993255862150896,
      "loss": 5.6382,
      "mean_token_accuracy": 0.1384445235133171,
      "num_tokens": 7610056.0,
      "step": 3460
    },
    {
      "entropy": 5.756991195678711,
      "epoch": 0.34388646288209607,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004993228421703108,
      "loss": 5.6375,
      "mean_token_accuracy": 0.13536739200353623,
      "num_tokens": 7620218.0,
      "step": 3465
    },
    {
      "entropy": 5.7849561214447025,
      "epoch": 0.3443826915442636,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004993200925628088,
      "loss": 5.7457,
      "mean_token_accuracy": 0.13547196015715599,
      "num_tokens": 7632781.0,
      "step": 3470
    },
    {
      "entropy": 5.846343946456909,
      "epoch": 0.3448789202064311,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000499317337392652,
      "loss": 5.7067,
      "mean_token_accuracy": 0.13641866371035577,
      "num_tokens": 7643845.0,
      "step": 3475
    },
    {
      "entropy": 5.814862680435181,
      "epoch": 0.34537514886859866,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004993145766599085,
      "loss": 5.7323,
      "mean_token_accuracy": 0.13265105634927749,
      "num_tokens": 7654583.0,
      "step": 3480
    },
    {
      "entropy": 5.779863452911377,
      "epoch": 0.3458713775307662,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004993118103646469,
      "loss": 5.581,
      "mean_token_accuracy": 0.14222516641020774,
      "num_tokens": 7665714.0,
      "step": 3485
    },
    {
      "entropy": 5.807524061203003,
      "epoch": 0.3463676061929337,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004993090385069357,
      "loss": 5.6426,
      "mean_token_accuracy": 0.1403677485883236,
      "num_tokens": 7676011.0,
      "step": 3490
    },
    {
      "entropy": 5.743056392669677,
      "epoch": 0.34686383485510125,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004993062610868437,
      "loss": 5.6994,
      "mean_token_accuracy": 0.13592455089092254,
      "num_tokens": 7686927.0,
      "step": 3495
    },
    {
      "entropy": 5.7684183597564695,
      "epoch": 0.3473600635172688,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004993034781044398,
      "loss": 5.6349,
      "mean_token_accuracy": 0.13586694225668908,
      "num_tokens": 7698419.0,
      "step": 3500
    },
    {
      "entropy": 5.849251174926758,
      "epoch": 0.3478562921794363,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499300689559793,
      "loss": 5.627,
      "mean_token_accuracy": 0.13978302255272865,
      "num_tokens": 7709695.0,
      "step": 3505
    },
    {
      "entropy": 5.78063006401062,
      "epoch": 0.3483525208416038,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004992978954529724,
      "loss": 5.6018,
      "mean_token_accuracy": 0.14354961216449738,
      "num_tokens": 7721309.0,
      "step": 3510
    },
    {
      "entropy": 5.775903558731079,
      "epoch": 0.3488487495037713,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004992950957840473,
      "loss": 5.6792,
      "mean_token_accuracy": 0.13921934366226196,
      "num_tokens": 7732074.0,
      "step": 3515
    },
    {
      "entropy": 5.859705686569214,
      "epoch": 0.34934497816593885,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499292290553087,
      "loss": 5.7478,
      "mean_token_accuracy": 0.13691921010613442,
      "num_tokens": 7742869.0,
      "step": 3520
    },
    {
      "entropy": 5.745819091796875,
      "epoch": 0.3498412068281064,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004992894797601614,
      "loss": 5.6456,
      "mean_token_accuracy": 0.13721488118171693,
      "num_tokens": 7754895.0,
      "step": 3525
    },
    {
      "entropy": 5.811453056335449,
      "epoch": 0.3503374354902739,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00049928666340534,
      "loss": 5.613,
      "mean_token_accuracy": 0.1413191042840481,
      "num_tokens": 7765262.0,
      "step": 3530
    },
    {
      "entropy": 5.722553396224976,
      "epoch": 0.35083366415244144,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004992838414886925,
      "loss": 5.6921,
      "mean_token_accuracy": 0.13405056223273276,
      "num_tokens": 7777348.0,
      "step": 3535
    },
    {
      "entropy": 5.832651519775391,
      "epoch": 0.35132989281460897,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004992810140102891,
      "loss": 5.6802,
      "mean_token_accuracy": 0.136273655295372,
      "num_tokens": 7788161.0,
      "step": 3540
    },
    {
      "entropy": 5.782558441162109,
      "epoch": 0.3518261214767765,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004992781809701999,
      "loss": 5.6726,
      "mean_token_accuracy": 0.1378345899283886,
      "num_tokens": 7798629.0,
      "step": 3545
    },
    {
      "entropy": 5.735908031463623,
      "epoch": 0.35232235013894403,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004992753423684951,
      "loss": 5.6783,
      "mean_token_accuracy": 0.137686438113451,
      "num_tokens": 7810148.0,
      "step": 3550
    },
    {
      "entropy": 5.815752363204956,
      "epoch": 0.35281857880111156,
      "grad_norm": 1.0,
      "learning_rate": 0.0004992724982052449,
      "loss": 5.6681,
      "mean_token_accuracy": 0.13577194958925248,
      "num_tokens": 7821667.0,
      "step": 3555
    },
    {
      "entropy": 5.76047043800354,
      "epoch": 0.3533148074632791,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00049926964848052,
      "loss": 5.653,
      "mean_token_accuracy": 0.14100263491272927,
      "num_tokens": 7832184.0,
      "step": 3560
    },
    {
      "entropy": 5.8020179271698,
      "epoch": 0.3538110361254466,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004992667931943913,
      "loss": 5.6656,
      "mean_token_accuracy": 0.13793855458498,
      "num_tokens": 7842983.0,
      "step": 3565
    },
    {
      "entropy": 5.791328716278076,
      "epoch": 0.35430726478761415,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004992639323469293,
      "loss": 5.6032,
      "mean_token_accuracy": 0.14528444707393645,
      "num_tokens": 7854783.0,
      "step": 3570
    },
    {
      "entropy": 5.734807443618775,
      "epoch": 0.3548034934497817,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499261065938205,
      "loss": 5.6097,
      "mean_token_accuracy": 0.14882165640592576,
      "num_tokens": 7864684.0,
      "step": 3575
    },
    {
      "entropy": 5.812001514434814,
      "epoch": 0.3552997221119492,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004992581939682895,
      "loss": 5.774,
      "mean_token_accuracy": 0.1278185188770294,
      "num_tokens": 7876846.0,
      "step": 3580
    },
    {
      "entropy": 5.900391387939453,
      "epoch": 0.3557959507741167,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499255316437254,
      "loss": 5.64,
      "mean_token_accuracy": 0.13711518049240112,
      "num_tokens": 7886826.0,
      "step": 3585
    },
    {
      "entropy": 5.703480863571167,
      "epoch": 0.3562921794362842,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004992524333451699,
      "loss": 5.7114,
      "mean_token_accuracy": 0.13370427638292312,
      "num_tokens": 7898941.0,
      "step": 3590
    },
    {
      "entropy": 5.773089456558227,
      "epoch": 0.35678840809845175,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004992495446921087,
      "loss": 5.616,
      "mean_token_accuracy": 0.1364353835582733,
      "num_tokens": 7909916.0,
      "step": 3595
    },
    {
      "entropy": 5.7855347156524655,
      "epoch": 0.3572846367606193,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004992466504781419,
      "loss": 5.6485,
      "mean_token_accuracy": 0.13997598960995675,
      "num_tokens": 7922137.0,
      "step": 3600
    },
    {
      "entropy": 5.682146310806274,
      "epoch": 0.3577808654227868,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004992437507033415,
      "loss": 5.5599,
      "mean_token_accuracy": 0.14679991602897643,
      "num_tokens": 7932940.0,
      "step": 3605
    },
    {
      "entropy": 5.837021684646606,
      "epoch": 0.35827709408495434,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004992408453677792,
      "loss": 5.7332,
      "mean_token_accuracy": 0.13032081127166747,
      "num_tokens": 7943735.0,
      "step": 3610
    },
    {
      "entropy": 5.760317420959472,
      "epoch": 0.35877332274712187,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004992379344715273,
      "loss": 5.6161,
      "mean_token_accuracy": 0.13913318663835525,
      "num_tokens": 7955323.0,
      "step": 3615
    },
    {
      "entropy": 5.6868006706237795,
      "epoch": 0.3592695514092894,
      "grad_norm": 1.0,
      "learning_rate": 0.0004992350180146577,
      "loss": 5.5981,
      "mean_token_accuracy": 0.14091636314988137,
      "num_tokens": 7966073.0,
      "step": 3620
    },
    {
      "entropy": 5.820367622375488,
      "epoch": 0.35976578007145693,
      "grad_norm": 1.0,
      "learning_rate": 0.0004992320959972428,
      "loss": 5.7517,
      "mean_token_accuracy": 0.13469664603471757,
      "num_tokens": 7977027.0,
      "step": 3625
    },
    {
      "entropy": 5.770542049407959,
      "epoch": 0.36026200873362446,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004992291684193551,
      "loss": 5.6092,
      "mean_token_accuracy": 0.14524339586496354,
      "num_tokens": 7987368.0,
      "step": 3630
    },
    {
      "entropy": 5.802702474594116,
      "epoch": 0.360758237395792,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004992262352810672,
      "loss": 5.6917,
      "mean_token_accuracy": 0.13706899285316468,
      "num_tokens": 8000060.0,
      "step": 3635
    },
    {
      "entropy": 5.845965242385864,
      "epoch": 0.3612544660579595,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004992232965824518,
      "loss": 5.7226,
      "mean_token_accuracy": 0.13262399807572364,
      "num_tokens": 8011388.0,
      "step": 3640
    },
    {
      "entropy": 5.746702432632446,
      "epoch": 0.36175069472012705,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499220352323582,
      "loss": 5.605,
      "mean_token_accuracy": 0.14428586438298224,
      "num_tokens": 8021682.0,
      "step": 3645
    },
    {
      "entropy": 5.750078248977661,
      "epoch": 0.3622469233822946,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992174025045304,
      "loss": 5.6234,
      "mean_token_accuracy": 0.13938501179218293,
      "num_tokens": 8032799.0,
      "step": 3650
    },
    {
      "entropy": 5.704130172729492,
      "epoch": 0.3627431520444621,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992144471253704,
      "loss": 5.5055,
      "mean_token_accuracy": 0.14247506186366082,
      "num_tokens": 8042742.0,
      "step": 3655
    },
    {
      "entropy": 5.786503791809082,
      "epoch": 0.3632393807066296,
      "grad_norm": 0.875,
      "learning_rate": 0.0004992114861861753,
      "loss": 5.6827,
      "mean_token_accuracy": 0.13944674730300904,
      "num_tokens": 8054068.0,
      "step": 3660
    },
    {
      "entropy": 5.790756273269653,
      "epoch": 0.3637356093687971,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004992085196870184,
      "loss": 5.6695,
      "mean_token_accuracy": 0.13868657797574996,
      "num_tokens": 8065286.0,
      "step": 3665
    },
    {
      "entropy": 5.766940307617188,
      "epoch": 0.36423183803096465,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004992055476279734,
      "loss": 5.7167,
      "mean_token_accuracy": 0.12711743265390396,
      "num_tokens": 8077122.0,
      "step": 3670
    },
    {
      "entropy": 5.846836996078491,
      "epoch": 0.3647280666931322,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992025700091139,
      "loss": 5.66,
      "mean_token_accuracy": 0.1416809819638729,
      "num_tokens": 8087036.0,
      "step": 3675
    },
    {
      "entropy": 5.781601667404175,
      "epoch": 0.3652242953552997,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004991995868305138,
      "loss": 5.6973,
      "mean_token_accuracy": 0.1388478934764862,
      "num_tokens": 8098517.0,
      "step": 3680
    },
    {
      "entropy": 5.680510234832764,
      "epoch": 0.36572052401746724,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004991965980922471,
      "loss": 5.5717,
      "mean_token_accuracy": 0.14402113407850264,
      "num_tokens": 8107475.0,
      "step": 3685
    },
    {
      "entropy": 5.778713607788086,
      "epoch": 0.36621675267963477,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004991936037943878,
      "loss": 5.6709,
      "mean_token_accuracy": 0.13664408326148986,
      "num_tokens": 8118529.0,
      "step": 3690
    },
    {
      "entropy": 5.82429313659668,
      "epoch": 0.3667129813418023,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004991906039370104,
      "loss": 5.5781,
      "mean_token_accuracy": 0.14057729691267012,
      "num_tokens": 8128951.0,
      "step": 3695
    },
    {
      "entropy": 5.6867591381073,
      "epoch": 0.36720921000396983,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004991875985201889,
      "loss": 5.5914,
      "mean_token_accuracy": 0.14267491921782494,
      "num_tokens": 8139385.0,
      "step": 3700
    },
    {
      "entropy": 5.760690116882325,
      "epoch": 0.36770543866613736,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499184587543998,
      "loss": 5.6417,
      "mean_token_accuracy": 0.14163167104125024,
      "num_tokens": 8149376.0,
      "step": 3705
    },
    {
      "entropy": 5.830694532394409,
      "epoch": 0.3682016673283049,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004991815710085126,
      "loss": 5.6757,
      "mean_token_accuracy": 0.13829845562577248,
      "num_tokens": 8160687.0,
      "step": 3710
    },
    {
      "entropy": 5.783746433258057,
      "epoch": 0.3686978959904724,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004991785489138072,
      "loss": 5.673,
      "mean_token_accuracy": 0.13856422975659372,
      "num_tokens": 8171762.0,
      "step": 3715
    },
    {
      "entropy": 5.668252658843994,
      "epoch": 0.36919412465263995,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004991755212599569,
      "loss": 5.4415,
      "mean_token_accuracy": 0.14850418344140054,
      "num_tokens": 8181643.0,
      "step": 3720
    },
    {
      "entropy": 5.653321266174316,
      "epoch": 0.3696903533148075,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004991724880470366,
      "loss": 5.6961,
      "mean_token_accuracy": 0.14297981560230255,
      "num_tokens": 8192692.0,
      "step": 3725
    },
    {
      "entropy": 5.8245847702026365,
      "epoch": 0.370186581976975,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004991694492751217,
      "loss": 5.6274,
      "mean_token_accuracy": 0.1364683873951435,
      "num_tokens": 8202941.0,
      "step": 3730
    },
    {
      "entropy": 5.747600936889649,
      "epoch": 0.3706828106391425,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004991664049442876,
      "loss": 5.6263,
      "mean_token_accuracy": 0.14124569520354271,
      "num_tokens": 8214206.0,
      "step": 3735
    },
    {
      "entropy": 5.694499111175537,
      "epoch": 0.37117903930131,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004991633550546097,
      "loss": 5.4604,
      "mean_token_accuracy": 0.1512777641415596,
      "num_tokens": 8223672.0,
      "step": 3740
    },
    {
      "entropy": 5.720300483703613,
      "epoch": 0.37167526796347755,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004991602996061635,
      "loss": 5.6137,
      "mean_token_accuracy": 0.14432198852300643,
      "num_tokens": 8234189.0,
      "step": 3745
    },
    {
      "entropy": 5.8249866485595705,
      "epoch": 0.3721714966256451,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004991572385990249,
      "loss": 5.6509,
      "mean_token_accuracy": 0.14039039388298988,
      "num_tokens": 8245382.0,
      "step": 3750
    },
    {
      "entropy": 5.78410153388977,
      "epoch": 0.3726677252878126,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004991541720332698,
      "loss": 5.6006,
      "mean_token_accuracy": 0.14047230705618857,
      "num_tokens": 8254530.0,
      "step": 3755
    },
    {
      "entropy": 5.685094213485717,
      "epoch": 0.37316395394998014,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004991510999089743,
      "loss": 5.6452,
      "mean_token_accuracy": 0.1451885238289833,
      "num_tokens": 8264030.0,
      "step": 3760
    },
    {
      "entropy": 5.694255018234253,
      "epoch": 0.3736601826121477,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004991480222262145,
      "loss": 5.6106,
      "mean_token_accuracy": 0.14497761949896812,
      "num_tokens": 8273537.0,
      "step": 3765
    },
    {
      "entropy": 5.816177749633789,
      "epoch": 0.3741564112743152,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004991449389850668,
      "loss": 5.6572,
      "mean_token_accuracy": 0.13783644288778304,
      "num_tokens": 8286123.0,
      "step": 3770
    },
    {
      "entropy": 5.77542405128479,
      "epoch": 0.37465263993648273,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004991418501856076,
      "loss": 5.6482,
      "mean_token_accuracy": 0.13695365637540818,
      "num_tokens": 8298115.0,
      "step": 3775
    },
    {
      "entropy": 5.780150699615478,
      "epoch": 0.37514886859865026,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004991387558279134,
      "loss": 5.6735,
      "mean_token_accuracy": 0.13566352427005768,
      "num_tokens": 8309376.0,
      "step": 3780
    },
    {
      "entropy": 5.761322498321533,
      "epoch": 0.3756450972608178,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004991356559120612,
      "loss": 5.6279,
      "mean_token_accuracy": 0.13963163048028945,
      "num_tokens": 8319272.0,
      "step": 3785
    },
    {
      "entropy": 5.761824226379394,
      "epoch": 0.3761413259229853,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004991325504381275,
      "loss": 5.629,
      "mean_token_accuracy": 0.13797441720962525,
      "num_tokens": 8330042.0,
      "step": 3790
    },
    {
      "entropy": 5.743603467941284,
      "epoch": 0.37663755458515286,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004991294394061896,
      "loss": 5.596,
      "mean_token_accuracy": 0.14118768870830536,
      "num_tokens": 8340394.0,
      "step": 3795
    },
    {
      "entropy": 5.70108060836792,
      "epoch": 0.3771337832473204,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004991263228163246,
      "loss": 5.6309,
      "mean_token_accuracy": 0.1392921596765518,
      "num_tokens": 8351103.0,
      "step": 3800
    },
    {
      "entropy": 5.783044767379761,
      "epoch": 0.3776300119094879,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004991232006686096,
      "loss": 5.6576,
      "mean_token_accuracy": 0.13801991567015648,
      "num_tokens": 8363215.0,
      "step": 3805
    },
    {
      "entropy": 5.8357092380523685,
      "epoch": 0.3781262405716554,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004991200729631223,
      "loss": 5.7,
      "mean_token_accuracy": 0.13135057240724562,
      "num_tokens": 8374602.0,
      "step": 3810
    },
    {
      "entropy": 5.7686113834381105,
      "epoch": 0.3786224692338229,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004991169396999401,
      "loss": 5.6115,
      "mean_token_accuracy": 0.13531650006771087,
      "num_tokens": 8386470.0,
      "step": 3815
    },
    {
      "entropy": 5.760716915130615,
      "epoch": 0.37911869789599045,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004991138008791408,
      "loss": 5.6257,
      "mean_token_accuracy": 0.14137942716479301,
      "num_tokens": 8396284.0,
      "step": 3820
    },
    {
      "entropy": 5.784114170074463,
      "epoch": 0.379614926558158,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004991106565008021,
      "loss": 5.6669,
      "mean_token_accuracy": 0.13694209456443787,
      "num_tokens": 8407856.0,
      "step": 3825
    },
    {
      "entropy": 5.8008862972259525,
      "epoch": 0.3801111552203255,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004991075065650019,
      "loss": 5.5639,
      "mean_token_accuracy": 0.14441588819026946,
      "num_tokens": 8419560.0,
      "step": 3830
    },
    {
      "entropy": 5.788234567642212,
      "epoch": 0.38060738388249304,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004991043510718186,
      "loss": 5.5603,
      "mean_token_accuracy": 0.14501854926347732,
      "num_tokens": 8430594.0,
      "step": 3835
    },
    {
      "entropy": 5.681872177124023,
      "epoch": 0.3811036125446606,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004991011900213303,
      "loss": 5.5325,
      "mean_token_accuracy": 0.14425235539674758,
      "num_tokens": 8440549.0,
      "step": 3840
    },
    {
      "entropy": 5.670712041854858,
      "epoch": 0.3815998412068281,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004990980234136152,
      "loss": 5.579,
      "mean_token_accuracy": 0.1425841838121414,
      "num_tokens": 8451553.0,
      "step": 3845
    },
    {
      "entropy": 5.710864925384522,
      "epoch": 0.38209606986899564,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004990948512487522,
      "loss": 5.5829,
      "mean_token_accuracy": 0.14589926972985268,
      "num_tokens": 8462899.0,
      "step": 3850
    },
    {
      "entropy": 5.752317094802857,
      "epoch": 0.38259229853116317,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004990916735268198,
      "loss": 5.6404,
      "mean_token_accuracy": 0.1422861024737358,
      "num_tokens": 8474566.0,
      "step": 3855
    },
    {
      "entropy": 5.808441305160523,
      "epoch": 0.3830885271933307,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004990884902478966,
      "loss": 5.645,
      "mean_token_accuracy": 0.14045381546020508,
      "num_tokens": 8484786.0,
      "step": 3860
    },
    {
      "entropy": 5.6865410804748535,
      "epoch": 0.3835847558554982,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004990853014120619,
      "loss": 5.571,
      "mean_token_accuracy": 0.14521799385547637,
      "num_tokens": 8496416.0,
      "step": 3865
    },
    {
      "entropy": 5.733003807067871,
      "epoch": 0.38408098451766576,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004990821070193944,
      "loss": 5.6045,
      "mean_token_accuracy": 0.14117611795663834,
      "num_tokens": 8507287.0,
      "step": 3870
    },
    {
      "entropy": 5.785660648345948,
      "epoch": 0.3845772131798333,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004990789070699736,
      "loss": 5.6507,
      "mean_token_accuracy": 0.134336905926466,
      "num_tokens": 8517922.0,
      "step": 3875
    },
    {
      "entropy": 5.702974510192871,
      "epoch": 0.3850734418420008,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004990757015638787,
      "loss": 5.5823,
      "mean_token_accuracy": 0.1453675426542759,
      "num_tokens": 8529532.0,
      "step": 3880
    },
    {
      "entropy": 5.749368524551391,
      "epoch": 0.3855696705041683,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004990724905011894,
      "loss": 5.5415,
      "mean_token_accuracy": 0.14694293439388276,
      "num_tokens": 8540207.0,
      "step": 3885
    },
    {
      "entropy": 5.764535760879516,
      "epoch": 0.3860658991663358,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499069273881985,
      "loss": 5.6017,
      "mean_token_accuracy": 0.1401916190981865,
      "num_tokens": 8551634.0,
      "step": 3890
    },
    {
      "entropy": 5.759525489807129,
      "epoch": 0.38656212782850335,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004990660517063455,
      "loss": 5.65,
      "mean_token_accuracy": 0.13647492751479148,
      "num_tokens": 8563256.0,
      "step": 3895
    },
    {
      "entropy": 5.7427960395812985,
      "epoch": 0.3870583564906709,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004990628239743508,
      "loss": 5.6617,
      "mean_token_accuracy": 0.1428539350628853,
      "num_tokens": 8573833.0,
      "step": 3900
    },
    {
      "entropy": 5.713337182998657,
      "epoch": 0.3875545851528384,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004990595906860807,
      "loss": 5.5842,
      "mean_token_accuracy": 0.1448860026896,
      "num_tokens": 8584283.0,
      "step": 3905
    },
    {
      "entropy": 5.746373891830444,
      "epoch": 0.38805081381500595,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004990563518416157,
      "loss": 5.6359,
      "mean_token_accuracy": 0.140063364058733,
      "num_tokens": 8595344.0,
      "step": 3910
    },
    {
      "entropy": 5.745514059066773,
      "epoch": 0.3885470424771735,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004990531074410358,
      "loss": 5.5535,
      "mean_token_accuracy": 0.14560891613364219,
      "num_tokens": 8605625.0,
      "step": 3915
    },
    {
      "entropy": 5.709172630310059,
      "epoch": 0.389043271139341,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004990498574844218,
      "loss": 5.5433,
      "mean_token_accuracy": 0.14344219714403153,
      "num_tokens": 8616259.0,
      "step": 3920
    },
    {
      "entropy": 5.773303079605102,
      "epoch": 0.38953949980150854,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004990466019718542,
      "loss": 5.5579,
      "mean_token_accuracy": 0.14597627222537995,
      "num_tokens": 8627179.0,
      "step": 3925
    },
    {
      "entropy": 5.7424726486206055,
      "epoch": 0.39003572846367607,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004990433409034134,
      "loss": 5.6674,
      "mean_token_accuracy": 0.1455268532037735,
      "num_tokens": 8637448.0,
      "step": 3930
    },
    {
      "entropy": 5.678447723388672,
      "epoch": 0.3905319571258436,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990400742791806,
      "loss": 5.5955,
      "mean_token_accuracy": 0.1387963816523552,
      "num_tokens": 8649168.0,
      "step": 3935
    },
    {
      "entropy": 5.773329401016236,
      "epoch": 0.39102818578801113,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004990368020992368,
      "loss": 5.6941,
      "mean_token_accuracy": 0.1367001138627529,
      "num_tokens": 8660653.0,
      "step": 3940
    },
    {
      "entropy": 5.7819396495819095,
      "epoch": 0.39152441445017866,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499033524363663,
      "loss": 5.6639,
      "mean_token_accuracy": 0.14397113472223283,
      "num_tokens": 8672000.0,
      "step": 3945
    },
    {
      "entropy": 5.703604078292846,
      "epoch": 0.3920206431123462,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004990302410725405,
      "loss": 5.5842,
      "mean_token_accuracy": 0.14821155071258546,
      "num_tokens": 8682455.0,
      "step": 3950
    },
    {
      "entropy": 5.719656085968017,
      "epoch": 0.3925168717745137,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004990269522259509,
      "loss": 5.5607,
      "mean_token_accuracy": 0.14346430450677872,
      "num_tokens": 8692975.0,
      "step": 3955
    },
    {
      "entropy": 5.700722599029541,
      "epoch": 0.3930131004366812,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004990236578239755,
      "loss": 5.5429,
      "mean_token_accuracy": 0.14142076373100282,
      "num_tokens": 8703569.0,
      "step": 3960
    },
    {
      "entropy": 5.761720275878906,
      "epoch": 0.3935093290988487,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004990203578666962,
      "loss": 5.6559,
      "mean_token_accuracy": 0.13742631003260614,
      "num_tokens": 8713734.0,
      "step": 3965
    },
    {
      "entropy": 5.743472719192505,
      "epoch": 0.39400555776101626,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004990170523541948,
      "loss": 5.5357,
      "mean_token_accuracy": 0.14768768101930618,
      "num_tokens": 8723852.0,
      "step": 3970
    },
    {
      "entropy": 5.710507488250732,
      "epoch": 0.3945017864231838,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004990137412865531,
      "loss": 5.6887,
      "mean_token_accuracy": 0.14348620921373367,
      "num_tokens": 8734300.0,
      "step": 3975
    },
    {
      "entropy": 5.805154275894165,
      "epoch": 0.3949980150853513,
      "grad_norm": 1.125,
      "learning_rate": 0.0004990104246638533,
      "loss": 5.7054,
      "mean_token_accuracy": 0.13508821055293083,
      "num_tokens": 8745022.0,
      "step": 3980
    },
    {
      "entropy": 5.774425458908081,
      "epoch": 0.39549424374751885,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004990071024861778,
      "loss": 5.5833,
      "mean_token_accuracy": 0.13405360728502275,
      "num_tokens": 8754870.0,
      "step": 3985
    },
    {
      "entropy": 5.83691554069519,
      "epoch": 0.3959904724096864,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004990037747536087,
      "loss": 5.6401,
      "mean_token_accuracy": 0.14253394007682801,
      "num_tokens": 8765536.0,
      "step": 3990
    },
    {
      "entropy": 5.679643011093139,
      "epoch": 0.3964867010718539,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004990004414662288,
      "loss": 5.6109,
      "mean_token_accuracy": 0.14047741740942002,
      "num_tokens": 8776898.0,
      "step": 3995
    },
    {
      "entropy": 5.740188264846802,
      "epoch": 0.39698292973402144,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004989971026241205,
      "loss": 5.6714,
      "mean_token_accuracy": 0.13848020657896995,
      "num_tokens": 8788176.0,
      "step": 4000
    },
    {
      "entropy": 5.700374555587769,
      "epoch": 0.39747915839618897,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004989937582273668,
      "loss": 5.5856,
      "mean_token_accuracy": 0.14035871252417564,
      "num_tokens": 8798939.0,
      "step": 4005
    },
    {
      "entropy": 5.715577459335327,
      "epoch": 0.3979753870583565,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004989904082760506,
      "loss": 5.6308,
      "mean_token_accuracy": 0.14493223130702973,
      "num_tokens": 8809092.0,
      "step": 4010
    },
    {
      "entropy": 5.731601619720459,
      "epoch": 0.39847161572052403,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000498987052770255,
      "loss": 5.549,
      "mean_token_accuracy": 0.14861732870340347,
      "num_tokens": 8819352.0,
      "step": 4015
    },
    {
      "entropy": 5.744062423706055,
      "epoch": 0.39896784438269156,
      "grad_norm": 0.9375,
      "learning_rate": 0.000498983691710063,
      "loss": 5.6741,
      "mean_token_accuracy": 0.1401342362165451,
      "num_tokens": 8829878.0,
      "step": 4020
    },
    {
      "entropy": 5.737670612335205,
      "epoch": 0.3994640730448591,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000498980325095558,
      "loss": 5.5346,
      "mean_token_accuracy": 0.1425821863114834,
      "num_tokens": 8840364.0,
      "step": 4025
    },
    {
      "entropy": 5.703105068206787,
      "epoch": 0.3999603017070266,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004989769529268238,
      "loss": 5.6547,
      "mean_token_accuracy": 0.1329083226621151,
      "num_tokens": 8851340.0,
      "step": 4030
    },
    {
      "entropy": 5.688699102401733,
      "epoch": 0.4004565303691941,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004989735752039436,
      "loss": 5.6232,
      "mean_token_accuracy": 0.13763951659202575,
      "num_tokens": 8862571.0,
      "step": 4035
    },
    {
      "entropy": 5.747498846054077,
      "epoch": 0.4009527590313616,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004989701919270015,
      "loss": 5.6114,
      "mean_token_accuracy": 0.14419316872954369,
      "num_tokens": 8873927.0,
      "step": 4040
    },
    {
      "entropy": 5.835363340377808,
      "epoch": 0.40144898769352916,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004989668030960812,
      "loss": 5.8179,
      "mean_token_accuracy": 0.13200713470578193,
      "num_tokens": 8886542.0,
      "step": 4045
    },
    {
      "entropy": 5.795601892471313,
      "epoch": 0.4019452163556967,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004989634087112667,
      "loss": 5.5541,
      "mean_token_accuracy": 0.1411711633205414,
      "num_tokens": 8896114.0,
      "step": 4050
    },
    {
      "entropy": 5.712681913375855,
      "epoch": 0.4024414450178642,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004989600087726425,
      "loss": 5.6337,
      "mean_token_accuracy": 0.135611891746521,
      "num_tokens": 8906303.0,
      "step": 4055
    },
    {
      "entropy": 5.780783605575562,
      "epoch": 0.40293767368003175,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004989566032802924,
      "loss": 5.666,
      "mean_token_accuracy": 0.13578887283802032,
      "num_tokens": 8916961.0,
      "step": 4060
    },
    {
      "entropy": 5.8231781959533695,
      "epoch": 0.4034339023421993,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004989531922343013,
      "loss": 5.7521,
      "mean_token_accuracy": 0.13754109293222427,
      "num_tokens": 8927700.0,
      "step": 4065
    },
    {
      "entropy": 5.754679441452026,
      "epoch": 0.4039301310043668,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004989497756347534,
      "loss": 5.5834,
      "mean_token_accuracy": 0.1437943920493126,
      "num_tokens": 8939655.0,
      "step": 4070
    },
    {
      "entropy": 5.698803377151489,
      "epoch": 0.40442635966653434,
      "grad_norm": 0.875,
      "learning_rate": 0.0004989463534817339,
      "loss": 5.525,
      "mean_token_accuracy": 0.14996594488620757,
      "num_tokens": 8951238.0,
      "step": 4075
    },
    {
      "entropy": 5.7337535381317135,
      "epoch": 0.40492258832870187,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004989429257753273,
      "loss": 5.5954,
      "mean_token_accuracy": 0.1474266193807125,
      "num_tokens": 8962019.0,
      "step": 4080
    },
    {
      "entropy": 5.738568735122681,
      "epoch": 0.4054188169908694,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004989394925156187,
      "loss": 5.595,
      "mean_token_accuracy": 0.1431464821100235,
      "num_tokens": 8973861.0,
      "step": 4085
    },
    {
      "entropy": 5.689034509658813,
      "epoch": 0.40591504565303693,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004989360537026932,
      "loss": 5.5913,
      "mean_token_accuracy": 0.14298243969678878,
      "num_tokens": 8985981.0,
      "step": 4090
    },
    {
      "entropy": 5.71806526184082,
      "epoch": 0.40641127431520446,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000498932609336636,
      "loss": 5.4788,
      "mean_token_accuracy": 0.14892211481928824,
      "num_tokens": 8995318.0,
      "step": 4095
    },
    {
      "entropy": 5.727901840209961,
      "epoch": 0.406907502977372,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004989291594175327,
      "loss": 5.5317,
      "mean_token_accuracy": 0.1424754813313484,
      "num_tokens": 9004679.0,
      "step": 4100
    },
    {
      "entropy": 5.833022928237915,
      "epoch": 0.4074037316395395,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004989257039454689,
      "loss": 5.6549,
      "mean_token_accuracy": 0.14180289059877396,
      "num_tokens": 9016422.0,
      "step": 4105
    },
    {
      "entropy": 5.692260265350342,
      "epoch": 0.40789996030170705,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00049892224292053,
      "loss": 5.5695,
      "mean_token_accuracy": 0.1427057534456253,
      "num_tokens": 9027914.0,
      "step": 4110
    },
    {
      "entropy": 5.702508449554443,
      "epoch": 0.40839618896387453,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004989187763428021,
      "loss": 5.5434,
      "mean_token_accuracy": 0.14706938862800598,
      "num_tokens": 9038701.0,
      "step": 4115
    },
    {
      "entropy": 5.695271348953247,
      "epoch": 0.40889241762604206,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000498915304212371,
      "loss": 5.6123,
      "mean_token_accuracy": 0.13733669966459275,
      "num_tokens": 9050620.0,
      "step": 4120
    },
    {
      "entropy": 5.745882606506347,
      "epoch": 0.4093886462882096,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004989118265293228,
      "loss": 5.6348,
      "mean_token_accuracy": 0.143435500562191,
      "num_tokens": 9060900.0,
      "step": 4125
    },
    {
      "entropy": 5.723922824859619,
      "epoch": 0.4098848749503771,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004989083432937439,
      "loss": 5.5599,
      "mean_token_accuracy": 0.1416636064648628,
      "num_tokens": 9072122.0,
      "step": 4130
    },
    {
      "entropy": 5.714425945281983,
      "epoch": 0.41038110361254465,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004989048545057204,
      "loss": 5.5717,
      "mean_token_accuracy": 0.14303382486104965,
      "num_tokens": 9082780.0,
      "step": 4135
    },
    {
      "entropy": 5.707315826416016,
      "epoch": 0.4108773322747122,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004989013601653391,
      "loss": 5.6098,
      "mean_token_accuracy": 0.14234035238623619,
      "num_tokens": 9093518.0,
      "step": 4140
    },
    {
      "entropy": 5.705009508132934,
      "epoch": 0.4113735609368797,
      "grad_norm": 0.875,
      "learning_rate": 0.0004988978602726866,
      "loss": 5.6371,
      "mean_token_accuracy": 0.14491504356265067,
      "num_tokens": 9105035.0,
      "step": 4145
    },
    {
      "entropy": 5.72711033821106,
      "epoch": 0.41186978959904724,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004988943548278494,
      "loss": 5.5522,
      "mean_token_accuracy": 0.14476445242762565,
      "num_tokens": 9117154.0,
      "step": 4150
    },
    {
      "entropy": 5.721778726577758,
      "epoch": 0.4123660182612148,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004988908438309149,
      "loss": 5.6244,
      "mean_token_accuracy": 0.14237917512655257,
      "num_tokens": 9128818.0,
      "step": 4155
    },
    {
      "entropy": 5.778127670288086,
      "epoch": 0.4128622469233823,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004988873272819697,
      "loss": 5.6435,
      "mean_token_accuracy": 0.14399504363536836,
      "num_tokens": 9140188.0,
      "step": 4160
    },
    {
      "entropy": 5.867915105819702,
      "epoch": 0.41335847558554983,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004988838051811014,
      "loss": 5.7094,
      "mean_token_accuracy": 0.1409727469086647,
      "num_tokens": 9150565.0,
      "step": 4165
    },
    {
      "entropy": 5.710624837875367,
      "epoch": 0.41385470424771736,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000498880277528397,
      "loss": 5.5585,
      "mean_token_accuracy": 0.14689609557390212,
      "num_tokens": 9160130.0,
      "step": 4170
    },
    {
      "entropy": 5.71880955696106,
      "epoch": 0.4143509329098849,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004988767443239443,
      "loss": 5.5574,
      "mean_token_accuracy": 0.14629159718751908,
      "num_tokens": 9171740.0,
      "step": 4175
    },
    {
      "entropy": 5.68405818939209,
      "epoch": 0.4148471615720524,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004988732055678307,
      "loss": 5.5073,
      "mean_token_accuracy": 0.15795112699270247,
      "num_tokens": 9182459.0,
      "step": 4180
    },
    {
      "entropy": 5.7594013690948485,
      "epoch": 0.41534339023421996,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000498869661260144,
      "loss": 5.5727,
      "mean_token_accuracy": 0.14156269729137422,
      "num_tokens": 9192284.0,
      "step": 4185
    },
    {
      "entropy": 5.685678911209107,
      "epoch": 0.41583961889638743,
      "grad_norm": 1.015625,
      "learning_rate": 0.000498866111400972,
      "loss": 5.566,
      "mean_token_accuracy": 0.14145209789276122,
      "num_tokens": 9202700.0,
      "step": 4190
    },
    {
      "entropy": 5.729474592208862,
      "epoch": 0.41633584755855496,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004988625559904029,
      "loss": 5.5868,
      "mean_token_accuracy": 0.13673395067453384,
      "num_tokens": 9213013.0,
      "step": 4195
    },
    {
      "entropy": 5.746359014511109,
      "epoch": 0.4168320762207225,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004988589950285248,
      "loss": 5.6554,
      "mean_token_accuracy": 0.13856094107031822,
      "num_tokens": 9224568.0,
      "step": 4200
    },
    {
      "entropy": 5.674574184417724,
      "epoch": 0.41732830488289,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004988554285154259,
      "loss": 5.5154,
      "mean_token_accuracy": 0.14188847690820694,
      "num_tokens": 9233791.0,
      "step": 4205
    },
    {
      "entropy": 5.743446207046508,
      "epoch": 0.41782453354505755,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004988518564511948,
      "loss": 5.5913,
      "mean_token_accuracy": 0.13892836272716522,
      "num_tokens": 9245048.0,
      "step": 4210
    },
    {
      "entropy": 5.789479351043701,
      "epoch": 0.4183207622072251,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004988482788359202,
      "loss": 5.5935,
      "mean_token_accuracy": 0.14784503281116484,
      "num_tokens": 9256508.0,
      "step": 4215
    },
    {
      "entropy": 5.7150181293487545,
      "epoch": 0.4188169908693926,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004988446956696904,
      "loss": 5.6271,
      "mean_token_accuracy": 0.14091657102108002,
      "num_tokens": 9267871.0,
      "step": 4220
    },
    {
      "entropy": 5.779106283187867,
      "epoch": 0.41931321953156014,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004988411069525947,
      "loss": 5.6809,
      "mean_token_accuracy": 0.1356142893433571,
      "num_tokens": 9279287.0,
      "step": 4225
    },
    {
      "entropy": 5.804977607727051,
      "epoch": 0.4198094481937277,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004988375126847216,
      "loss": 5.6935,
      "mean_token_accuracy": 0.13691136837005616,
      "num_tokens": 9291257.0,
      "step": 4230
    },
    {
      "entropy": 5.812367677688599,
      "epoch": 0.4203056768558952,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004988339128661607,
      "loss": 5.6579,
      "mean_token_accuracy": 0.14066813588142396,
      "num_tokens": 9302590.0,
      "step": 4235
    },
    {
      "entropy": 5.618175888061524,
      "epoch": 0.42080190551806274,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000498830307497001,
      "loss": 5.515,
      "mean_token_accuracy": 0.1433468133211136,
      "num_tokens": 9313174.0,
      "step": 4240
    },
    {
      "entropy": 5.695446538925171,
      "epoch": 0.42129813418023027,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000498826696577332,
      "loss": 5.6063,
      "mean_token_accuracy": 0.14157773107290267,
      "num_tokens": 9325034.0,
      "step": 4245
    },
    {
      "entropy": 5.748422145843506,
      "epoch": 0.4217943628423978,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004988230801072432,
      "loss": 5.5873,
      "mean_token_accuracy": 0.14445904046297073,
      "num_tokens": 9336834.0,
      "step": 4250
    },
    {
      "entropy": 5.697594404220581,
      "epoch": 0.4222905915045653,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004988194580868244,
      "loss": 5.5885,
      "mean_token_accuracy": 0.1432207852602005,
      "num_tokens": 9347747.0,
      "step": 4255
    },
    {
      "entropy": 5.68045597076416,
      "epoch": 0.42278682016673286,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004988158305161654,
      "loss": 5.5934,
      "mean_token_accuracy": 0.1437683343887329,
      "num_tokens": 9359186.0,
      "step": 4260
    },
    {
      "entropy": 5.722374105453492,
      "epoch": 0.42328304882890033,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000498812197395356,
      "loss": 5.514,
      "mean_token_accuracy": 0.14643879979848862,
      "num_tokens": 9369485.0,
      "step": 4265
    },
    {
      "entropy": 5.757485628128052,
      "epoch": 0.42377927749106786,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004988085587244863,
      "loss": 5.6179,
      "mean_token_accuracy": 0.13760613054037094,
      "num_tokens": 9382658.0,
      "step": 4270
    },
    {
      "entropy": 5.686766576766968,
      "epoch": 0.4242755061532354,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004988049145036466,
      "loss": 5.5651,
      "mean_token_accuracy": 0.13786769136786461,
      "num_tokens": 9394151.0,
      "step": 4275
    },
    {
      "entropy": 5.803422689437866,
      "epoch": 0.4247717348154029,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004988012647329273,
      "loss": 5.6641,
      "mean_token_accuracy": 0.13631192743778228,
      "num_tokens": 9405956.0,
      "step": 4280
    },
    {
      "entropy": 5.7161619663238525,
      "epoch": 0.42526796347757045,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000498797609412419,
      "loss": 5.5532,
      "mean_token_accuracy": 0.14562675803899766,
      "num_tokens": 9416889.0,
      "step": 4285
    },
    {
      "entropy": 5.713563346862793,
      "epoch": 0.425764192139738,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004987939485422122,
      "loss": 5.6597,
      "mean_token_accuracy": 0.13954902291297913,
      "num_tokens": 9427602.0,
      "step": 4290
    },
    {
      "entropy": 5.756892251968384,
      "epoch": 0.4262604208019055,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004987902821223976,
      "loss": 5.5697,
      "mean_token_accuracy": 0.14587784707546234,
      "num_tokens": 9438557.0,
      "step": 4295
    },
    {
      "entropy": 5.696601390838623,
      "epoch": 0.42675664946407305,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004987866101530663,
      "loss": 5.6446,
      "mean_token_accuracy": 0.14027416184544564,
      "num_tokens": 9449598.0,
      "step": 4300
    },
    {
      "entropy": 5.736512470245361,
      "epoch": 0.4272528781262406,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004987829326343092,
      "loss": 5.555,
      "mean_token_accuracy": 0.14784253388643265,
      "num_tokens": 9459740.0,
      "step": 4305
    },
    {
      "entropy": 5.7235182762146,
      "epoch": 0.4277491067884081,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004987792495662177,
      "loss": 5.5779,
      "mean_token_accuracy": 0.13899921551346778,
      "num_tokens": 9469835.0,
      "step": 4310
    },
    {
      "entropy": 5.7507165431976315,
      "epoch": 0.42824533545057564,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000498775560948883,
      "loss": 5.6549,
      "mean_token_accuracy": 0.14052388891577722,
      "num_tokens": 9480735.0,
      "step": 4315
    },
    {
      "entropy": 5.712874031066894,
      "epoch": 0.42874156411274317,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004987718667823965,
      "loss": 5.538,
      "mean_token_accuracy": 0.14749773293733598,
      "num_tokens": 9491321.0,
      "step": 4320
    },
    {
      "entropy": 5.732460165023804,
      "epoch": 0.4292377927749107,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00049876816706685,
      "loss": 5.6957,
      "mean_token_accuracy": 0.13651810511946677,
      "num_tokens": 9503507.0,
      "step": 4325
    },
    {
      "entropy": 5.802684688568116,
      "epoch": 0.42973402143707823,
      "grad_norm": 1.7734375,
      "learning_rate": 0.000498764461802335,
      "loss": 5.6834,
      "mean_token_accuracy": 0.14579661712050437,
      "num_tokens": 9514404.0,
      "step": 4330
    },
    {
      "entropy": 5.7324159145355225,
      "epoch": 0.43023025009924576,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004987607509889437,
      "loss": 5.5574,
      "mean_token_accuracy": 0.1418476790189743,
      "num_tokens": 9525555.0,
      "step": 4335
    },
    {
      "entropy": 5.744416856765747,
      "epoch": 0.43072647876141323,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004987570346267677,
      "loss": 5.5349,
      "mean_token_accuracy": 0.1448458880186081,
      "num_tokens": 9535084.0,
      "step": 4340
    },
    {
      "entropy": 5.688948249816894,
      "epoch": 0.43122270742358076,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004987533127158996,
      "loss": 5.5599,
      "mean_token_accuracy": 0.14783779084682463,
      "num_tokens": 9546170.0,
      "step": 4345
    },
    {
      "entropy": 5.707106018066407,
      "epoch": 0.4317189360857483,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004987495852564314,
      "loss": 5.5708,
      "mean_token_accuracy": 0.1440203867852688,
      "num_tokens": 9556793.0,
      "step": 4350
    },
    {
      "entropy": 5.728043746948242,
      "epoch": 0.4322151647479158,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004987458522484556,
      "loss": 5.5347,
      "mean_token_accuracy": 0.14598321914672852,
      "num_tokens": 9568114.0,
      "step": 4355
    },
    {
      "entropy": 5.6903989791870115,
      "epoch": 0.43271139341008336,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004987421136920648,
      "loss": 5.5325,
      "mean_token_accuracy": 0.14110969603061677,
      "num_tokens": 9579119.0,
      "step": 4360
    },
    {
      "entropy": 5.637863302230835,
      "epoch": 0.4332076220722509,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004987383695873517,
      "loss": 5.5679,
      "mean_token_accuracy": 0.1449605032801628,
      "num_tokens": 9590259.0,
      "step": 4365
    },
    {
      "entropy": 5.7717413902282715,
      "epoch": 0.4337038507344184,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004987346199344092,
      "loss": 5.6091,
      "mean_token_accuracy": 0.13644771203398703,
      "num_tokens": 9600431.0,
      "step": 4370
    },
    {
      "entropy": 5.666158962249756,
      "epoch": 0.43420007939658595,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004987308647333303,
      "loss": 5.5863,
      "mean_token_accuracy": 0.14474302157759666,
      "num_tokens": 9610286.0,
      "step": 4375
    },
    {
      "entropy": 5.7676201343536375,
      "epoch": 0.4346963080587535,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004987271039842079,
      "loss": 5.6053,
      "mean_token_accuracy": 0.14020158275961875,
      "num_tokens": 9621148.0,
      "step": 4380
    },
    {
      "entropy": 5.6954185962677,
      "epoch": 0.435192536720921,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004987233376871355,
      "loss": 5.489,
      "mean_token_accuracy": 0.14512064382433892,
      "num_tokens": 9631300.0,
      "step": 4385
    },
    {
      "entropy": 5.684174966812134,
      "epoch": 0.43568876538308854,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004987195658422063,
      "loss": 5.5825,
      "mean_token_accuracy": 0.14152609258890153,
      "num_tokens": 9641986.0,
      "step": 4390
    },
    {
      "entropy": 5.750404691696167,
      "epoch": 0.43618499404525607,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004987157884495141,
      "loss": 5.6017,
      "mean_token_accuracy": 0.14466868340969086,
      "num_tokens": 9652962.0,
      "step": 4395
    },
    {
      "entropy": 5.7012653827667235,
      "epoch": 0.4366812227074236,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004987120055091523,
      "loss": 5.6411,
      "mean_token_accuracy": 0.1366007797420025,
      "num_tokens": 9664084.0,
      "step": 4400
    },
    {
      "entropy": 5.6462547302246096,
      "epoch": 0.43717745136959113,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004987082170212149,
      "loss": 5.5308,
      "mean_token_accuracy": 0.13876494839787484,
      "num_tokens": 9675821.0,
      "step": 4405
    },
    {
      "entropy": 5.772474527359009,
      "epoch": 0.43767368003175866,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004987044229857957,
      "loss": 5.6539,
      "mean_token_accuracy": 0.1392872929573059,
      "num_tokens": 9687964.0,
      "step": 4410
    },
    {
      "entropy": 5.742184209823608,
      "epoch": 0.43816990869392614,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004987006234029889,
      "loss": 5.6274,
      "mean_token_accuracy": 0.14196077436208726,
      "num_tokens": 9699653.0,
      "step": 4415
    },
    {
      "entropy": 5.769703722000122,
      "epoch": 0.43866613735609367,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004986968182728885,
      "loss": 5.6332,
      "mean_token_accuracy": 0.14099507704377173,
      "num_tokens": 9709872.0,
      "step": 4420
    },
    {
      "entropy": 5.763581466674805,
      "epoch": 0.4391623660182612,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004986930075955892,
      "loss": 5.5604,
      "mean_token_accuracy": 0.13856668621301652,
      "num_tokens": 9720589.0,
      "step": 4425
    },
    {
      "entropy": 5.648024702072144,
      "epoch": 0.4396585946804287,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004986891913711853,
      "loss": 5.4917,
      "mean_token_accuracy": 0.15007439404726028,
      "num_tokens": 9731862.0,
      "step": 4430
    },
    {
      "entropy": 5.676476955413818,
      "epoch": 0.44015482334259626,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004986853695997715,
      "loss": 5.6447,
      "mean_token_accuracy": 0.1467808410525322,
      "num_tokens": 9743710.0,
      "step": 4435
    },
    {
      "entropy": 5.7441404342651365,
      "epoch": 0.4406510520047638,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004986815422814425,
      "loss": 5.5928,
      "mean_token_accuracy": 0.13124571591615677,
      "num_tokens": 9753855.0,
      "step": 4440
    },
    {
      "entropy": 5.832332277297974,
      "epoch": 0.4411472806669313,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004986777094162932,
      "loss": 5.6,
      "mean_token_accuracy": 0.14974353909492494,
      "num_tokens": 9763883.0,
      "step": 4445
    },
    {
      "entropy": 5.690779829025269,
      "epoch": 0.44164350932909885,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004986738710044187,
      "loss": 5.5678,
      "mean_token_accuracy": 0.1452265501022339,
      "num_tokens": 9774890.0,
      "step": 4450
    },
    {
      "entropy": 5.675231504440307,
      "epoch": 0.4421397379912664,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004986700270459142,
      "loss": 5.5401,
      "mean_token_accuracy": 0.14680951908230783,
      "num_tokens": 9785249.0,
      "step": 4455
    },
    {
      "entropy": 5.715075492858887,
      "epoch": 0.4426359666534339,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004986661775408749,
      "loss": 5.5418,
      "mean_token_accuracy": 0.1484077528119087,
      "num_tokens": 9795723.0,
      "step": 4460
    },
    {
      "entropy": 5.701987457275391,
      "epoch": 0.44313219531560144,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004986623224893965,
      "loss": 5.4956,
      "mean_token_accuracy": 0.14353932589292526,
      "num_tokens": 9807637.0,
      "step": 4465
    },
    {
      "entropy": 5.737668371200561,
      "epoch": 0.44362842397776897,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004986584618915744,
      "loss": 5.617,
      "mean_token_accuracy": 0.14440734907984734,
      "num_tokens": 9820298.0,
      "step": 4470
    },
    {
      "entropy": 5.6621157169342045,
      "epoch": 0.4441246526399365,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004986545957475045,
      "loss": 5.4962,
      "mean_token_accuracy": 0.15061585903167723,
      "num_tokens": 9831251.0,
      "step": 4475
    },
    {
      "entropy": 5.658903312683106,
      "epoch": 0.44462088130210403,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004986507240572825,
      "loss": 5.5639,
      "mean_token_accuracy": 0.14525441080331802,
      "num_tokens": 9842884.0,
      "step": 4480
    },
    {
      "entropy": 5.655770349502563,
      "epoch": 0.44511710996427156,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004986468468210042,
      "loss": 5.467,
      "mean_token_accuracy": 0.14644730389118193,
      "num_tokens": 9854413.0,
      "step": 4485
    },
    {
      "entropy": 5.705828237533569,
      "epoch": 0.44561333862643904,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004986429640387663,
      "loss": 5.6078,
      "mean_token_accuracy": 0.1486542619764805,
      "num_tokens": 9865771.0,
      "step": 4490
    },
    {
      "entropy": 5.735600233078003,
      "epoch": 0.44610956728860657,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004986390757106648,
      "loss": 5.521,
      "mean_token_accuracy": 0.1486384704709053,
      "num_tokens": 9876717.0,
      "step": 4495
    },
    {
      "entropy": 5.674195289611816,
      "epoch": 0.4466057959507741,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004986351818367959,
      "loss": 5.6052,
      "mean_token_accuracy": 0.14445709586143493,
      "num_tokens": 9887616.0,
      "step": 4500
    },
    {
      "entropy": 5.7188373565673825,
      "epoch": 0.44710202461294163,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004986312824172566,
      "loss": 5.5535,
      "mean_token_accuracy": 0.14710604920983314,
      "num_tokens": 9898979.0,
      "step": 4505
    },
    {
      "entropy": 5.7393416404724125,
      "epoch": 0.44759825327510916,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004986273774521433,
      "loss": 5.598,
      "mean_token_accuracy": 0.13431652784347534,
      "num_tokens": 9912158.0,
      "step": 4510
    },
    {
      "entropy": 5.729260540008545,
      "epoch": 0.4480944819372767,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004986234669415529,
      "loss": 5.6112,
      "mean_token_accuracy": 0.13989987969398499,
      "num_tokens": 9923886.0,
      "step": 4515
    },
    {
      "entropy": 5.710967636108398,
      "epoch": 0.4485907105994442,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004986195508855824,
      "loss": 5.568,
      "mean_token_accuracy": 0.14522476345300675,
      "num_tokens": 9934979.0,
      "step": 4520
    },
    {
      "entropy": 5.73377742767334,
      "epoch": 0.44908693926161175,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000498615629284329,
      "loss": 5.6753,
      "mean_token_accuracy": 0.13699942901730539,
      "num_tokens": 9944715.0,
      "step": 4525
    },
    {
      "entropy": 5.720676136016846,
      "epoch": 0.4495831679237793,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004986117021378896,
      "loss": 5.5443,
      "mean_token_accuracy": 0.14494104832410812,
      "num_tokens": 9954736.0,
      "step": 4530
    },
    {
      "entropy": 5.709241533279419,
      "epoch": 0.4500793965859468,
      "grad_norm": 0.8125,
      "learning_rate": 0.000498607769446362,
      "loss": 5.6216,
      "mean_token_accuracy": 0.1421697661280632,
      "num_tokens": 9965952.0,
      "step": 4535
    },
    {
      "entropy": 5.6858762264251705,
      "epoch": 0.45057562524811434,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004986038312098435,
      "loss": 5.5741,
      "mean_token_accuracy": 0.14561808109283447,
      "num_tokens": 9976428.0,
      "step": 4540
    },
    {
      "entropy": 5.689672613143921,
      "epoch": 0.4510718539102819,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004985998874284319,
      "loss": 5.5091,
      "mean_token_accuracy": 0.15442328453063964,
      "num_tokens": 9986100.0,
      "step": 4545
    },
    {
      "entropy": 5.609327793121338,
      "epoch": 0.4515680825724494,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004985959381022248,
      "loss": 5.5176,
      "mean_token_accuracy": 0.14544908851385116,
      "num_tokens": 9996584.0,
      "step": 4550
    },
    {
      "entropy": 5.65582242012024,
      "epoch": 0.45206431123461693,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004985919832313203,
      "loss": 5.5209,
      "mean_token_accuracy": 0.1447821080684662,
      "num_tokens": 10007792.0,
      "step": 4555
    },
    {
      "entropy": 5.783624315261841,
      "epoch": 0.45256053989678446,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004985880228158164,
      "loss": 5.7078,
      "mean_token_accuracy": 0.13523221537470817,
      "num_tokens": 10019615.0,
      "step": 4560
    },
    {
      "entropy": 5.654605865478516,
      "epoch": 0.45305676855895194,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004985840568558112,
      "loss": 5.5268,
      "mean_token_accuracy": 0.14684706926345825,
      "num_tokens": 10030223.0,
      "step": 4565
    },
    {
      "entropy": 5.660999774932861,
      "epoch": 0.45355299722111947,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004985800853514033,
      "loss": 5.5514,
      "mean_token_accuracy": 0.14329646304249763,
      "num_tokens": 10041522.0,
      "step": 4570
    },
    {
      "entropy": 5.739897918701172,
      "epoch": 0.454049225883287,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004985761083026911,
      "loss": 5.4876,
      "mean_token_accuracy": 0.1519765079021454,
      "num_tokens": 10052847.0,
      "step": 4575
    },
    {
      "entropy": 5.634422731399536,
      "epoch": 0.45454545454545453,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004985721257097731,
      "loss": 5.5151,
      "mean_token_accuracy": 0.14605328142642976,
      "num_tokens": 10062259.0,
      "step": 4580
    },
    {
      "entropy": 5.746643114089966,
      "epoch": 0.45504168320762206,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004985681375727482,
      "loss": 5.6664,
      "mean_token_accuracy": 0.13895911127328872,
      "num_tokens": 10073341.0,
      "step": 4585
    },
    {
      "entropy": 5.727483177185059,
      "epoch": 0.4555379118697896,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004985641438917151,
      "loss": 5.5213,
      "mean_token_accuracy": 0.15104319751262665,
      "num_tokens": 10083992.0,
      "step": 4590
    },
    {
      "entropy": 5.7053868770599365,
      "epoch": 0.4560341405319571,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004985601446667731,
      "loss": 5.5157,
      "mean_token_accuracy": 0.14777568504214286,
      "num_tokens": 10094245.0,
      "step": 4595
    },
    {
      "entropy": 5.688941955566406,
      "epoch": 0.45653036919412465,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004985561398980213,
      "loss": 5.5144,
      "mean_token_accuracy": 0.14640356823801995,
      "num_tokens": 10105976.0,
      "step": 4600
    },
    {
      "entropy": 5.641877317428589,
      "epoch": 0.4570265978562922,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004985521295855588,
      "loss": 5.5733,
      "mean_token_accuracy": 0.14555746018886567,
      "num_tokens": 10117373.0,
      "step": 4605
    },
    {
      "entropy": 5.679246425628662,
      "epoch": 0.4575228265184597,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004985481137294853,
      "loss": 5.5183,
      "mean_token_accuracy": 0.13970350474119186,
      "num_tokens": 10128470.0,
      "step": 4610
    },
    {
      "entropy": 5.621714115142822,
      "epoch": 0.45801905518062724,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004985440923299002,
      "loss": 5.5212,
      "mean_token_accuracy": 0.14765125513076782,
      "num_tokens": 10140217.0,
      "step": 4615
    },
    {
      "entropy": 5.667717123031617,
      "epoch": 0.4585152838427948,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004985400653869034,
      "loss": 5.5875,
      "mean_token_accuracy": 0.14465930163860322,
      "num_tokens": 10152308.0,
      "step": 4620
    },
    {
      "entropy": 5.760606670379639,
      "epoch": 0.4590115125049623,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004985360329005945,
      "loss": 5.6278,
      "mean_token_accuracy": 0.14379535466432572,
      "num_tokens": 10164234.0,
      "step": 4625
    },
    {
      "entropy": 5.850100708007813,
      "epoch": 0.45950774116712984,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004985319948710738,
      "loss": 5.6696,
      "mean_token_accuracy": 0.14368383213877678,
      "num_tokens": 10175287.0,
      "step": 4630
    },
    {
      "entropy": 5.714134311676025,
      "epoch": 0.46000396982929737,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004985279512984413,
      "loss": 5.5401,
      "mean_token_accuracy": 0.1474534623324871,
      "num_tokens": 10184658.0,
      "step": 4635
    },
    {
      "entropy": 5.683081340789795,
      "epoch": 0.46050019849146484,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004985239021827972,
      "loss": 5.5407,
      "mean_token_accuracy": 0.14281928688287734,
      "num_tokens": 10194593.0,
      "step": 4640
    },
    {
      "entropy": 5.643809127807617,
      "epoch": 0.46099642715363237,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000498519847524242,
      "loss": 5.4609,
      "mean_token_accuracy": 0.1482139438390732,
      "num_tokens": 10205200.0,
      "step": 4645
    },
    {
      "entropy": 5.627954339981079,
      "epoch": 0.4614926558157999,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004985157873228762,
      "loss": 5.4901,
      "mean_token_accuracy": 0.1481429785490036,
      "num_tokens": 10216848.0,
      "step": 4650
    },
    {
      "entropy": 5.701125526428223,
      "epoch": 0.46198888447796743,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004985117215788005,
      "loss": 5.5382,
      "mean_token_accuracy": 0.1468523286283016,
      "num_tokens": 10228188.0,
      "step": 4655
    },
    {
      "entropy": 5.6984175682067875,
      "epoch": 0.46248511314013496,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004985076502921158,
      "loss": 5.6165,
      "mean_token_accuracy": 0.14205382913351058,
      "num_tokens": 10238253.0,
      "step": 4660
    },
    {
      "entropy": 5.732058429718018,
      "epoch": 0.4629813418023025,
      "grad_norm": 1.015625,
      "learning_rate": 0.000498503573462923,
      "loss": 5.5538,
      "mean_token_accuracy": 0.1446946606040001,
      "num_tokens": 10248169.0,
      "step": 4665
    },
    {
      "entropy": 5.6297468662261965,
      "epoch": 0.46347757046447,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004984994910913232,
      "loss": 5.4295,
      "mean_token_accuracy": 0.1478232428431511,
      "num_tokens": 10258962.0,
      "step": 4670
    },
    {
      "entropy": 5.6839498519897464,
      "epoch": 0.46397379912663755,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004984954031774175,
      "loss": 5.5555,
      "mean_token_accuracy": 0.14447656720876695,
      "num_tokens": 10270303.0,
      "step": 4675
    },
    {
      "entropy": 5.606175422668457,
      "epoch": 0.4644700277888051,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004984913097213075,
      "loss": 5.4721,
      "mean_token_accuracy": 0.14679303541779518,
      "num_tokens": 10281782.0,
      "step": 4680
    },
    {
      "entropy": 5.665231561660766,
      "epoch": 0.4649662564509726,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004984872107230945,
      "loss": 5.5035,
      "mean_token_accuracy": 0.14818502962589264,
      "num_tokens": 10292937.0,
      "step": 4685
    },
    {
      "entropy": 5.675636720657349,
      "epoch": 0.46546248511314015,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004984831061828803,
      "loss": 5.5788,
      "mean_token_accuracy": 0.14178428724408149,
      "num_tokens": 10305443.0,
      "step": 4690
    },
    {
      "entropy": 5.661796236038208,
      "epoch": 0.4659587137753077,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004984789961007667,
      "loss": 5.5021,
      "mean_token_accuracy": 0.14874687641859055,
      "num_tokens": 10316489.0,
      "step": 4695
    },
    {
      "entropy": 5.598919534683228,
      "epoch": 0.4664549424374752,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004984748804768554,
      "loss": 5.5602,
      "mean_token_accuracy": 0.13966515585780143,
      "num_tokens": 10326690.0,
      "step": 4700
    },
    {
      "entropy": 5.724484539031982,
      "epoch": 0.46695117109964274,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004984707593112487,
      "loss": 5.548,
      "mean_token_accuracy": 0.14317874610424042,
      "num_tokens": 10337059.0,
      "step": 4705
    },
    {
      "entropy": 5.647740983963013,
      "epoch": 0.46744739976181027,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004984666326040487,
      "loss": 5.4308,
      "mean_token_accuracy": 0.1530276544392109,
      "num_tokens": 10347035.0,
      "step": 4710
    },
    {
      "entropy": 5.6334373474121096,
      "epoch": 0.46794362842397774,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004984625003553578,
      "loss": 5.5579,
      "mean_token_accuracy": 0.14425093233585357,
      "num_tokens": 10359492.0,
      "step": 4715
    },
    {
      "entropy": 5.715490818023682,
      "epoch": 0.4684398570861453,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004984583625652783,
      "loss": 5.5419,
      "mean_token_accuracy": 0.14544321745634078,
      "num_tokens": 10370128.0,
      "step": 4720
    },
    {
      "entropy": 5.689985609054565,
      "epoch": 0.4689360857483128,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000498454219233913,
      "loss": 5.5389,
      "mean_token_accuracy": 0.14008901715278627,
      "num_tokens": 10380276.0,
      "step": 4725
    },
    {
      "entropy": 5.613158082962036,
      "epoch": 0.46943231441048033,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004984500703613645,
      "loss": 5.5232,
      "mean_token_accuracy": 0.14934480264782907,
      "num_tokens": 10392489.0,
      "step": 4730
    },
    {
      "entropy": 5.769829893112183,
      "epoch": 0.46992854307264786,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004984459159477358,
      "loss": 5.6114,
      "mean_token_accuracy": 0.14083144292235375,
      "num_tokens": 10404259.0,
      "step": 4735
    },
    {
      "entropy": 5.626959705352784,
      "epoch": 0.4704247717348154,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004984417559931298,
      "loss": 5.474,
      "mean_token_accuracy": 0.14718540757894516,
      "num_tokens": 10415000.0,
      "step": 4740
    },
    {
      "entropy": 5.70870304107666,
      "epoch": 0.4709210003969829,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004984375904976498,
      "loss": 5.5996,
      "mean_token_accuracy": 0.1425722323358059,
      "num_tokens": 10426078.0,
      "step": 4745
    },
    {
      "entropy": 5.705552864074707,
      "epoch": 0.47141722905915046,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004984334194613989,
      "loss": 5.5221,
      "mean_token_accuracy": 0.14522773921489715,
      "num_tokens": 10436838.0,
      "step": 4750
    },
    {
      "entropy": 5.6620279312133786,
      "epoch": 0.471913457721318,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004984292428844809,
      "loss": 5.5104,
      "mean_token_accuracy": 0.1477622404694557,
      "num_tokens": 10448544.0,
      "step": 4755
    },
    {
      "entropy": 5.723327922821045,
      "epoch": 0.4724096863834855,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004984250607669989,
      "loss": 5.5926,
      "mean_token_accuracy": 0.13789047300815582,
      "num_tokens": 10459730.0,
      "step": 4760
    },
    {
      "entropy": 5.760179853439331,
      "epoch": 0.47290591504565305,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004984208731090569,
      "loss": 5.6464,
      "mean_token_accuracy": 0.14240211620926857,
      "num_tokens": 10470257.0,
      "step": 4765
    },
    {
      "entropy": 5.725982236862182,
      "epoch": 0.4734021437078206,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004984166799107587,
      "loss": 5.5736,
      "mean_token_accuracy": 0.1431686021387577,
      "num_tokens": 10480835.0,
      "step": 4770
    },
    {
      "entropy": 5.673721694946289,
      "epoch": 0.4738983723699881,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004984124811722081,
      "loss": 5.6174,
      "mean_token_accuracy": 0.14570664018392562,
      "num_tokens": 10491720.0,
      "step": 4775
    },
    {
      "entropy": 5.677192306518554,
      "epoch": 0.47439460103215564,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004984082768935096,
      "loss": 5.6288,
      "mean_token_accuracy": 0.14744448885321618,
      "num_tokens": 10502313.0,
      "step": 4780
    },
    {
      "entropy": 5.717935991287232,
      "epoch": 0.47489082969432317,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004984040670747672,
      "loss": 5.5059,
      "mean_token_accuracy": 0.14801656156778337,
      "num_tokens": 10513333.0,
      "step": 4785
    },
    {
      "entropy": 5.642219924926758,
      "epoch": 0.47538705835649064,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004983998517160852,
      "loss": 5.4789,
      "mean_token_accuracy": 0.15268410742282867,
      "num_tokens": 10522766.0,
      "step": 4790
    },
    {
      "entropy": 5.641929483413696,
      "epoch": 0.4758832870186582,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004983956308175684,
      "loss": 5.5273,
      "mean_token_accuracy": 0.1454133540391922,
      "num_tokens": 10534362.0,
      "step": 4795
    },
    {
      "entropy": 5.717424297332764,
      "epoch": 0.4763795156808257,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004983914043793212,
      "loss": 5.5404,
      "mean_token_accuracy": 0.14541754499077797,
      "num_tokens": 10545072.0,
      "step": 4800
    },
    {
      "entropy": 5.662140846252441,
      "epoch": 0.47687574434299324,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004983871724014487,
      "loss": 5.4779,
      "mean_token_accuracy": 0.14503563791513444,
      "num_tokens": 10555953.0,
      "step": 4805
    },
    {
      "entropy": 5.675953578948975,
      "epoch": 0.47737197300516077,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004983829348840556,
      "loss": 5.5266,
      "mean_token_accuracy": 0.15017574280500412,
      "num_tokens": 10567757.0,
      "step": 4810
    },
    {
      "entropy": 5.5973015308380125,
      "epoch": 0.4778682016673283,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000498378691827247,
      "loss": 5.4211,
      "mean_token_accuracy": 0.14696255847811698,
      "num_tokens": 10579390.0,
      "step": 4815
    },
    {
      "entropy": 5.703798151016235,
      "epoch": 0.4783644303294958,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004983744432311282,
      "loss": 5.6241,
      "mean_token_accuracy": 0.1413818195462227,
      "num_tokens": 10589624.0,
      "step": 4820
    },
    {
      "entropy": 5.770223808288574,
      "epoch": 0.47886065899166336,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004983701890958046,
      "loss": 5.6033,
      "mean_token_accuracy": 0.13691395223140718,
      "num_tokens": 10601927.0,
      "step": 4825
    },
    {
      "entropy": 5.598169279098511,
      "epoch": 0.4793568876538309,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004983659294213815,
      "loss": 5.4757,
      "mean_token_accuracy": 0.15195037573575973,
      "num_tokens": 10614073.0,
      "step": 4830
    },
    {
      "entropy": 5.658059740066529,
      "epoch": 0.4798531163159984,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004983616642079649,
      "loss": 5.5083,
      "mean_token_accuracy": 0.14958997219800949,
      "num_tokens": 10625535.0,
      "step": 4835
    },
    {
      "entropy": 5.777277183532715,
      "epoch": 0.48034934497816595,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004983573934556601,
      "loss": 5.4825,
      "mean_token_accuracy": 0.15245484858751296,
      "num_tokens": 10637160.0,
      "step": 4840
    },
    {
      "entropy": 5.7017354488372805,
      "epoch": 0.4808455736403335,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004983531171645734,
      "loss": 5.5988,
      "mean_token_accuracy": 0.14110324680805206,
      "num_tokens": 10649061.0,
      "step": 4845
    },
    {
      "entropy": 5.562698173522949,
      "epoch": 0.481341802302501,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004983488353348106,
      "loss": 5.437,
      "mean_token_accuracy": 0.15287891477346421,
      "num_tokens": 10661130.0,
      "step": 4850
    },
    {
      "entropy": 5.650864458084106,
      "epoch": 0.48183803096466854,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004983445479664779,
      "loss": 5.4996,
      "mean_token_accuracy": 0.151998171210289,
      "num_tokens": 10673017.0,
      "step": 4855
    },
    {
      "entropy": 5.735723400115967,
      "epoch": 0.48233425962683607,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004983402550596817,
      "loss": 5.5697,
      "mean_token_accuracy": 0.14529768973588944,
      "num_tokens": 10683268.0,
      "step": 4860
    },
    {
      "entropy": 5.782104110717773,
      "epoch": 0.48283048828900355,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004983359566145285,
      "loss": 5.5835,
      "mean_token_accuracy": 0.14277169853448868,
      "num_tokens": 10694220.0,
      "step": 4865
    },
    {
      "entropy": 5.681487417221069,
      "epoch": 0.4833267169511711,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004983316526311248,
      "loss": 5.4965,
      "mean_token_accuracy": 0.14053841307759285,
      "num_tokens": 10704829.0,
      "step": 4870
    },
    {
      "entropy": 5.690288496017456,
      "epoch": 0.4838229456133386,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004983273431095773,
      "loss": 5.6949,
      "mean_token_accuracy": 0.1336314469575882,
      "num_tokens": 10717160.0,
      "step": 4875
    },
    {
      "entropy": 5.73415470123291,
      "epoch": 0.48431917427550614,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004983230280499929,
      "loss": 5.498,
      "mean_token_accuracy": 0.14990181624889373,
      "num_tokens": 10728638.0,
      "step": 4880
    },
    {
      "entropy": 5.670762395858764,
      "epoch": 0.48481540293767367,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004983187074524787,
      "loss": 5.5515,
      "mean_token_accuracy": 0.14588793516159057,
      "num_tokens": 10740122.0,
      "step": 4885
    },
    {
      "entropy": 5.650857067108154,
      "epoch": 0.4853116315998412,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004983143813171418,
      "loss": 5.5085,
      "mean_token_accuracy": 0.1518013060092926,
      "num_tokens": 10751103.0,
      "step": 4890
    },
    {
      "entropy": 5.625285959243774,
      "epoch": 0.48580786026200873,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004983100496440895,
      "loss": 5.4702,
      "mean_token_accuracy": 0.15078884214162827,
      "num_tokens": 10761937.0,
      "step": 4895
    },
    {
      "entropy": 5.783539056777954,
      "epoch": 0.48630408892417626,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000498305712433429,
      "loss": 5.6993,
      "mean_token_accuracy": 0.13927533775568007,
      "num_tokens": 10773527.0,
      "step": 4900
    },
    {
      "entropy": 5.689421606063843,
      "epoch": 0.4868003175863438,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000498301369685268,
      "loss": 5.5252,
      "mean_token_accuracy": 0.14703643843531608,
      "num_tokens": 10783873.0,
      "step": 4905
    },
    {
      "entropy": 5.613701868057251,
      "epoch": 0.4872965462485113,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004982970213997142,
      "loss": 5.381,
      "mean_token_accuracy": 0.15223184525966643,
      "num_tokens": 10796124.0,
      "step": 4910
    },
    {
      "entropy": 5.588850116729736,
      "epoch": 0.48779277491067885,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004982926675768756,
      "loss": 5.49,
      "mean_token_accuracy": 0.14737663865089418,
      "num_tokens": 10806468.0,
      "step": 4915
    },
    {
      "entropy": 5.650877809524536,
      "epoch": 0.4882890035728464,
      "grad_norm": 0.875,
      "learning_rate": 0.0004982883082168599,
      "loss": 5.5443,
      "mean_token_accuracy": 0.14648173600435258,
      "num_tokens": 10817126.0,
      "step": 4920
    },
    {
      "entropy": 5.713279438018799,
      "epoch": 0.4887852322350139,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004982839433197752,
      "loss": 5.5715,
      "mean_token_accuracy": 0.1365367963910103,
      "num_tokens": 10828985.0,
      "step": 4925
    },
    {
      "entropy": 5.649182891845703,
      "epoch": 0.48928146089718144,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00049827957288573,
      "loss": 5.4752,
      "mean_token_accuracy": 0.14631946086883546,
      "num_tokens": 10840528.0,
      "step": 4930
    },
    {
      "entropy": 5.659715747833252,
      "epoch": 0.48977768955934897,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004982751969148324,
      "loss": 5.4823,
      "mean_token_accuracy": 0.1495818853378296,
      "num_tokens": 10851312.0,
      "step": 4935
    },
    {
      "entropy": 5.6238807201385494,
      "epoch": 0.49027391822151645,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004982708154071912,
      "loss": 5.4437,
      "mean_token_accuracy": 0.14900542199611663,
      "num_tokens": 10861882.0,
      "step": 4940
    },
    {
      "entropy": 5.668314647674561,
      "epoch": 0.490770146883684,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004982664283629147,
      "loss": 5.5941,
      "mean_token_accuracy": 0.13935748636722564,
      "num_tokens": 10873187.0,
      "step": 4945
    },
    {
      "entropy": 5.6875807762146,
      "epoch": 0.4912663755458515,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004982620357821118,
      "loss": 5.4848,
      "mean_token_accuracy": 0.15269453823566437,
      "num_tokens": 10884254.0,
      "step": 4950
    },
    {
      "entropy": 5.59837589263916,
      "epoch": 0.49176260420801904,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004982576376648918,
      "loss": 5.4791,
      "mean_token_accuracy": 0.14832527041435242,
      "num_tokens": 10895770.0,
      "step": 4955
    },
    {
      "entropy": 5.6609045505523685,
      "epoch": 0.49225883287018657,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004982532340113633,
      "loss": 5.4942,
      "mean_token_accuracy": 0.1478537067770958,
      "num_tokens": 10906436.0,
      "step": 4960
    },
    {
      "entropy": 5.698761129379273,
      "epoch": 0.4927550615323541,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004982488248216357,
      "loss": 5.4663,
      "mean_token_accuracy": 0.14812514334917068,
      "num_tokens": 10917655.0,
      "step": 4965
    },
    {
      "entropy": 5.648922395706177,
      "epoch": 0.49325129019452163,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004982444100958182,
      "loss": 5.4702,
      "mean_token_accuracy": 0.15194573104381562,
      "num_tokens": 10928359.0,
      "step": 4970
    },
    {
      "entropy": 5.663105535507202,
      "epoch": 0.49374751885668916,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004982399898340205,
      "loss": 5.637,
      "mean_token_accuracy": 0.13940111473202704,
      "num_tokens": 10940494.0,
      "step": 4975
    },
    {
      "entropy": 5.739015007019043,
      "epoch": 0.4942437475188567,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000498235564036352,
      "loss": 5.6336,
      "mean_token_accuracy": 0.1354987032711506,
      "num_tokens": 10951298.0,
      "step": 4980
    },
    {
      "entropy": 5.670893716812134,
      "epoch": 0.4947399761810242,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004982311327029226,
      "loss": 5.5232,
      "mean_token_accuracy": 0.14640430435538293,
      "num_tokens": 10961680.0,
      "step": 4985
    },
    {
      "entropy": 5.775196647644043,
      "epoch": 0.49523620484319175,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004982266958338421,
      "loss": 5.6519,
      "mean_token_accuracy": 0.14135513976216316,
      "num_tokens": 10974968.0,
      "step": 4990
    },
    {
      "entropy": 5.642757892608643,
      "epoch": 0.4957324335053593,
      "grad_norm": 0.875,
      "learning_rate": 0.0004982222534292205,
      "loss": 5.5701,
      "mean_token_accuracy": 0.1384653314948082,
      "num_tokens": 10986383.0,
      "step": 4995
    },
    {
      "entropy": 5.758541297912598,
      "epoch": 0.4962286621675268,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004982178054891681,
      "loss": 5.6179,
      "mean_token_accuracy": 0.14474987238645554,
      "num_tokens": 10998384.0,
      "step": 5000
    },
    {
      "entropy": 5.681232976913452,
      "epoch": 0.49672489082969434,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004982133520137951,
      "loss": 5.389,
      "mean_token_accuracy": 0.15461313128471374,
      "num_tokens": 11009420.0,
      "step": 5005
    },
    {
      "entropy": 5.6752293586730955,
      "epoch": 0.4972211194918619,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004982088930032119,
      "loss": 5.5462,
      "mean_token_accuracy": 0.14680904522538185,
      "num_tokens": 11019601.0,
      "step": 5010
    },
    {
      "entropy": 5.649302768707275,
      "epoch": 0.4977173481540294,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004982044284575293,
      "loss": 5.4825,
      "mean_token_accuracy": 0.15077524483203888,
      "num_tokens": 11031138.0,
      "step": 5015
    },
    {
      "entropy": 5.6932378768920895,
      "epoch": 0.4982135768161969,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004981999583768576,
      "loss": 5.4972,
      "mean_token_accuracy": 0.14869402647018432,
      "num_tokens": 11041643.0,
      "step": 5020
    },
    {
      "entropy": 5.639556217193603,
      "epoch": 0.4987098054783644,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000498195482761308,
      "loss": 5.5166,
      "mean_token_accuracy": 0.13713015839457512,
      "num_tokens": 11052116.0,
      "step": 5025
    },
    {
      "entropy": 5.763868427276611,
      "epoch": 0.49920603414053194,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004981910016109914,
      "loss": 5.6962,
      "mean_token_accuracy": 0.1334000639617443,
      "num_tokens": 11064796.0,
      "step": 5030
    },
    {
      "entropy": 5.795920133590698,
      "epoch": 0.49970226280269947,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004981865149260189,
      "loss": 5.5422,
      "mean_token_accuracy": 0.14919748455286025,
      "num_tokens": 11075029.0,
      "step": 5035
    },
    {
      "entropy": 5.691865539550781,
      "epoch": 0.5001984914648671,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004981820227065016,
      "loss": 5.5688,
      "mean_token_accuracy": 0.1439378082752228,
      "num_tokens": 11086702.0,
      "step": 5040
    },
    {
      "entropy": 5.636436605453492,
      "epoch": 0.5006947201270345,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004981775249525513,
      "loss": 5.5487,
      "mean_token_accuracy": 0.14065493866801262,
      "num_tokens": 11097417.0,
      "step": 5045
    },
    {
      "entropy": 5.619754123687744,
      "epoch": 0.5011909487892021,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004981730216642791,
      "loss": 5.4312,
      "mean_token_accuracy": 0.15412285625934602,
      "num_tokens": 11108458.0,
      "step": 5050
    },
    {
      "entropy": 5.71159348487854,
      "epoch": 0.5016871774513696,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004981685128417969,
      "loss": 5.606,
      "mean_token_accuracy": 0.14178437143564224,
      "num_tokens": 11121466.0,
      "step": 5055
    },
    {
      "entropy": 5.693905544281006,
      "epoch": 0.5021834061135371,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004981639984852165,
      "loss": 5.5735,
      "mean_token_accuracy": 0.14030421599745752,
      "num_tokens": 11132489.0,
      "step": 5060
    },
    {
      "entropy": 5.643926811218262,
      "epoch": 0.5026796347757047,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004981594785946498,
      "loss": 5.5035,
      "mean_token_accuracy": 0.14792548716068268,
      "num_tokens": 11143087.0,
      "step": 5065
    },
    {
      "entropy": 5.711865997314453,
      "epoch": 0.5031758634378721,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004981549531702089,
      "loss": 5.5463,
      "mean_token_accuracy": 0.13861528635025025,
      "num_tokens": 11154229.0,
      "step": 5070
    },
    {
      "entropy": 5.684705591201782,
      "epoch": 0.5036720921000397,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000498150422212006,
      "loss": 5.5091,
      "mean_token_accuracy": 0.1434071958065033,
      "num_tokens": 11165698.0,
      "step": 5075
    },
    {
      "entropy": 5.68370566368103,
      "epoch": 0.5041683207622072,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004981458857201534,
      "loss": 5.5029,
      "mean_token_accuracy": 0.14430436715483666,
      "num_tokens": 11177025.0,
      "step": 5080
    },
    {
      "entropy": 5.708558464050293,
      "epoch": 0.5046645494243748,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004981413436947639,
      "loss": 5.5667,
      "mean_token_accuracy": 0.14835076332092284,
      "num_tokens": 11188766.0,
      "step": 5085
    },
    {
      "entropy": 5.664322233200073,
      "epoch": 0.5051607780865423,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004981367961359496,
      "loss": 5.6283,
      "mean_token_accuracy": 0.13773046880960466,
      "num_tokens": 11201626.0,
      "step": 5090
    },
    {
      "entropy": 5.6853649616241455,
      "epoch": 0.5056570067487098,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004981322430438239,
      "loss": 5.533,
      "mean_token_accuracy": 0.14464087933301925,
      "num_tokens": 11213975.0,
      "step": 5095
    },
    {
      "entropy": 5.66510968208313,
      "epoch": 0.5061532354108773,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004981276844184991,
      "loss": 5.4853,
      "mean_token_accuracy": 0.1474761888384819,
      "num_tokens": 11223636.0,
      "step": 5100
    },
    {
      "entropy": 5.611086797714234,
      "epoch": 0.5066494640730449,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004981231202600886,
      "loss": 5.4442,
      "mean_token_accuracy": 0.14874403178691864,
      "num_tokens": 11233549.0,
      "step": 5105
    },
    {
      "entropy": 5.596583366394043,
      "epoch": 0.5071456927352124,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004981185505687056,
      "loss": 5.4526,
      "mean_token_accuracy": 0.1505897492170334,
      "num_tokens": 11242915.0,
      "step": 5110
    },
    {
      "entropy": 5.740023183822632,
      "epoch": 0.50764192139738,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004981139753444632,
      "loss": 5.5914,
      "mean_token_accuracy": 0.1402211457490921,
      "num_tokens": 11254370.0,
      "step": 5115
    },
    {
      "entropy": 5.672887802124023,
      "epoch": 0.5081381500595474,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004981093945874752,
      "loss": 5.5167,
      "mean_token_accuracy": 0.1479219064116478,
      "num_tokens": 11264226.0,
      "step": 5120
    },
    {
      "entropy": 5.613948059082031,
      "epoch": 0.508634378721715,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004981048082978547,
      "loss": 5.4124,
      "mean_token_accuracy": 0.15597307235002517,
      "num_tokens": 11275304.0,
      "step": 5125
    },
    {
      "entropy": 5.612861156463623,
      "epoch": 0.5091306073838825,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004981002164757158,
      "loss": 5.4731,
      "mean_token_accuracy": 0.14841337352991105,
      "num_tokens": 11286179.0,
      "step": 5130
    },
    {
      "entropy": 5.707505702972412,
      "epoch": 0.50962683604605,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004980956191211723,
      "loss": 5.5044,
      "mean_token_accuracy": 0.15050283074378967,
      "num_tokens": 11297244.0,
      "step": 5135
    },
    {
      "entropy": 5.691849660873413,
      "epoch": 0.5101230647082176,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004980910162343382,
      "loss": 5.5116,
      "mean_token_accuracy": 0.15071954280138017,
      "num_tokens": 11307251.0,
      "step": 5140
    },
    {
      "entropy": 5.67484941482544,
      "epoch": 0.510619293370385,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004980864078153275,
      "loss": 5.4943,
      "mean_token_accuracy": 0.14468513876199723,
      "num_tokens": 11318441.0,
      "step": 5145
    },
    {
      "entropy": 5.574022626876831,
      "epoch": 0.5111155220325526,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004980817938642548,
      "loss": 5.4672,
      "mean_token_accuracy": 0.1504348635673523,
      "num_tokens": 11328743.0,
      "step": 5150
    },
    {
      "entropy": 5.621407842636108,
      "epoch": 0.5116117506947201,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000498077174381234,
      "loss": 5.4571,
      "mean_token_accuracy": 0.1529311530292034,
      "num_tokens": 11340745.0,
      "step": 5155
    },
    {
      "entropy": 5.703346395492554,
      "epoch": 0.5121079793568877,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004980725493663802,
      "loss": 5.5203,
      "mean_token_accuracy": 0.1512206144630909,
      "num_tokens": 11352060.0,
      "step": 5160
    },
    {
      "entropy": 5.640616035461425,
      "epoch": 0.5126042080190552,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004980679188198078,
      "loss": 5.5279,
      "mean_token_accuracy": 0.14537407457828522,
      "num_tokens": 11363097.0,
      "step": 5165
    },
    {
      "entropy": 5.687601852416992,
      "epoch": 0.5131004366812227,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004980632827416317,
      "loss": 5.5476,
      "mean_token_accuracy": 0.16172647774219512,
      "num_tokens": 11373410.0,
      "step": 5170
    },
    {
      "entropy": 5.705497169494629,
      "epoch": 0.5135966653433902,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004980586411319669,
      "loss": 5.6674,
      "mean_token_accuracy": 0.14129337221384047,
      "num_tokens": 11385615.0,
      "step": 5175
    },
    {
      "entropy": 5.601473951339722,
      "epoch": 0.5140928940055578,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004980539939909283,
      "loss": 5.4292,
      "mean_token_accuracy": 0.14421194940805435,
      "num_tokens": 11396492.0,
      "step": 5180
    },
    {
      "entropy": 5.699062871932983,
      "epoch": 0.5145891226677253,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004980493413186314,
      "loss": 5.5396,
      "mean_token_accuracy": 0.1392931617796421,
      "num_tokens": 11406377.0,
      "step": 5185
    },
    {
      "entropy": 5.685783576965332,
      "epoch": 0.5150853513298929,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004980446831151914,
      "loss": 5.5101,
      "mean_token_accuracy": 0.14487480521202087,
      "num_tokens": 11416862.0,
      "step": 5190
    },
    {
      "entropy": 5.63937292098999,
      "epoch": 0.5155815799920603,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004980400193807239,
      "loss": 5.4585,
      "mean_token_accuracy": 0.15452056229114533,
      "num_tokens": 11428246.0,
      "step": 5195
    },
    {
      "entropy": 5.656797409057617,
      "epoch": 0.5160778086542279,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004980353501153446,
      "loss": 5.4988,
      "mean_token_accuracy": 0.1489669419825077,
      "num_tokens": 11438147.0,
      "step": 5200
    },
    {
      "entropy": 5.71633825302124,
      "epoch": 0.5165740373163954,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004980306753191691,
      "loss": 5.6226,
      "mean_token_accuracy": 0.14095016047358513,
      "num_tokens": 11449154.0,
      "step": 5205
    },
    {
      "entropy": 5.693401622772217,
      "epoch": 0.5170702659785629,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004980259949923135,
      "loss": 5.5201,
      "mean_token_accuracy": 0.14495897442102432,
      "num_tokens": 11459262.0,
      "step": 5210
    },
    {
      "entropy": 5.697604036331176,
      "epoch": 0.5175664946407305,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004980213091348938,
      "loss": 5.5126,
      "mean_token_accuracy": 0.14773540049791337,
      "num_tokens": 11471991.0,
      "step": 5215
    },
    {
      "entropy": 5.627869701385498,
      "epoch": 0.5180627233028979,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004980166177470262,
      "loss": 5.4505,
      "mean_token_accuracy": 0.14557459205389023,
      "num_tokens": 11483176.0,
      "step": 5220
    },
    {
      "entropy": 5.701073884963989,
      "epoch": 0.5185589519650655,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004980119208288271,
      "loss": 5.5483,
      "mean_token_accuracy": 0.14131275564432144,
      "num_tokens": 11494015.0,
      "step": 5225
    },
    {
      "entropy": 5.613106107711792,
      "epoch": 0.519055180627233,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004980072183804128,
      "loss": 5.4654,
      "mean_token_accuracy": 0.1503124602138996,
      "num_tokens": 11506365.0,
      "step": 5230
    },
    {
      "entropy": 5.626119470596313,
      "epoch": 0.5195514092894006,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004980025104019,
      "loss": 5.5456,
      "mean_token_accuracy": 0.15200978443026542,
      "num_tokens": 11517199.0,
      "step": 5235
    },
    {
      "entropy": 5.624682331085205,
      "epoch": 0.520047637951568,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004979977968934056,
      "loss": 5.3987,
      "mean_token_accuracy": 0.15699853301048278,
      "num_tokens": 11527591.0,
      "step": 5240
    },
    {
      "entropy": 5.696583080291748,
      "epoch": 0.5205438666137356,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004979930778550463,
      "loss": 5.5699,
      "mean_token_accuracy": 0.14902450740337372,
      "num_tokens": 11538657.0,
      "step": 5245
    },
    {
      "entropy": 5.607371425628662,
      "epoch": 0.5210400952759031,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004979883532869392,
      "loss": 5.3807,
      "mean_token_accuracy": 0.16424401253461837,
      "num_tokens": 11549513.0,
      "step": 5250
    },
    {
      "entropy": 5.681191349029541,
      "epoch": 0.5215363239380707,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004979836231892013,
      "loss": 5.5036,
      "mean_token_accuracy": 0.1442022740840912,
      "num_tokens": 11560376.0,
      "step": 5255
    },
    {
      "entropy": 5.589213800430298,
      "epoch": 0.5220325526002382,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004979788875619501,
      "loss": 5.4602,
      "mean_token_accuracy": 0.14428113251924515,
      "num_tokens": 11571218.0,
      "step": 5260
    },
    {
      "entropy": 5.65407133102417,
      "epoch": 0.5225287812624058,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000497974146405303,
      "loss": 5.5262,
      "mean_token_accuracy": 0.14889778792858124,
      "num_tokens": 11581981.0,
      "step": 5265
    },
    {
      "entropy": 5.662479019165039,
      "epoch": 0.5230250099245732,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004979693997193776,
      "loss": 5.4796,
      "mean_token_accuracy": 0.14792142063379288,
      "num_tokens": 11592561.0,
      "step": 5270
    },
    {
      "entropy": 5.681849956512451,
      "epoch": 0.5235212385867408,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004979646475042914,
      "loss": 5.4594,
      "mean_token_accuracy": 0.14910868704319,
      "num_tokens": 11601933.0,
      "step": 5275
    },
    {
      "entropy": 5.601380777359009,
      "epoch": 0.5240174672489083,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004979598897601625,
      "loss": 5.3937,
      "mean_token_accuracy": 0.1597262978553772,
      "num_tokens": 11613938.0,
      "step": 5280
    },
    {
      "entropy": 5.693992376327515,
      "epoch": 0.5245136959110758,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004979551264871088,
      "loss": 5.5608,
      "mean_token_accuracy": 0.14049421474337578,
      "num_tokens": 11624890.0,
      "step": 5285
    },
    {
      "entropy": 5.656209802627563,
      "epoch": 0.5250099245732434,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004979503576852482,
      "loss": 5.5634,
      "mean_token_accuracy": 0.14738093614578246,
      "num_tokens": 11635080.0,
      "step": 5290
    },
    {
      "entropy": 5.559004783630371,
      "epoch": 0.5255061532354108,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004979455833546994,
      "loss": 5.4507,
      "mean_token_accuracy": 0.150251816958189,
      "num_tokens": 11645180.0,
      "step": 5295
    },
    {
      "entropy": 5.698576164245606,
      "epoch": 0.5260023818975784,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004979408034955804,
      "loss": 5.5122,
      "mean_token_accuracy": 0.1495388776063919,
      "num_tokens": 11657599.0,
      "step": 5300
    },
    {
      "entropy": 5.709982585906983,
      "epoch": 0.5264986105597459,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004979360181080099,
      "loss": 5.554,
      "mean_token_accuracy": 0.13851185664534568,
      "num_tokens": 11669177.0,
      "step": 5305
    },
    {
      "entropy": 5.611147880554199,
      "epoch": 0.5269948392219135,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004979312271921065,
      "loss": 5.4022,
      "mean_token_accuracy": 0.14745137840509415,
      "num_tokens": 11679756.0,
      "step": 5310
    },
    {
      "entropy": 5.605481243133545,
      "epoch": 0.527491067884081,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000497926430747989,
      "loss": 5.4801,
      "mean_token_accuracy": 0.1458248570561409,
      "num_tokens": 11689805.0,
      "step": 5315
    },
    {
      "entropy": 5.648860311508178,
      "epoch": 0.5279872965462485,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004979216287757764,
      "loss": 5.4855,
      "mean_token_accuracy": 0.15177066773176193,
      "num_tokens": 11701150.0,
      "step": 5320
    },
    {
      "entropy": 5.582261037826538,
      "epoch": 0.528483525208416,
      "grad_norm": 0.90625,
      "learning_rate": 0.000497916821275588,
      "loss": 5.4237,
      "mean_token_accuracy": 0.1469492107629776,
      "num_tokens": 11712256.0,
      "step": 5325
    },
    {
      "entropy": 5.678399562835693,
      "epoch": 0.5289797538705836,
      "grad_norm": 0.875,
      "learning_rate": 0.0004979120082475424,
      "loss": 5.4753,
      "mean_token_accuracy": 0.15113273933529853,
      "num_tokens": 11722547.0,
      "step": 5330
    },
    {
      "entropy": 5.63621883392334,
      "epoch": 0.5294759825327511,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004979071896917595,
      "loss": 5.4934,
      "mean_token_accuracy": 0.14984367564320564,
      "num_tokens": 11732309.0,
      "step": 5335
    },
    {
      "entropy": 5.65453200340271,
      "epoch": 0.5299722111949187,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004979023656083586,
      "loss": 5.5385,
      "mean_token_accuracy": 0.1465301111340523,
      "num_tokens": 11743982.0,
      "step": 5340
    },
    {
      "entropy": 5.62791337966919,
      "epoch": 0.5304684398570861,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004978975359974594,
      "loss": 5.4629,
      "mean_token_accuracy": 0.14960571080446244,
      "num_tokens": 11755506.0,
      "step": 5345
    },
    {
      "entropy": 5.659137630462647,
      "epoch": 0.5309646685192537,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004978927008591816,
      "loss": 5.4984,
      "mean_token_accuracy": 0.14654770717024804,
      "num_tokens": 11767024.0,
      "step": 5350
    },
    {
      "entropy": 5.666571617126465,
      "epoch": 0.5314608971814212,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004978878601936451,
      "loss": 5.4781,
      "mean_token_accuracy": 0.14480792880058288,
      "num_tokens": 11777625.0,
      "step": 5355
    },
    {
      "entropy": 5.589890098571777,
      "epoch": 0.5319571258435887,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004978830140009698,
      "loss": 5.464,
      "mean_token_accuracy": 0.15269244760274886,
      "num_tokens": 11788610.0,
      "step": 5360
    },
    {
      "entropy": 5.667603826522827,
      "epoch": 0.5324533545057563,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004978781622812762,
      "loss": 5.5754,
      "mean_token_accuracy": 0.14783891588449477,
      "num_tokens": 11800828.0,
      "step": 5365
    },
    {
      "entropy": 5.670748043060303,
      "epoch": 0.5329495831679237,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004978733050346844,
      "loss": 5.5546,
      "mean_token_accuracy": 0.15211703032255172,
      "num_tokens": 11811729.0,
      "step": 5370
    },
    {
      "entropy": 5.684667110443115,
      "epoch": 0.5334458118300913,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004978684422613149,
      "loss": 5.5888,
      "mean_token_accuracy": 0.14408276826143265,
      "num_tokens": 11822081.0,
      "step": 5375
    },
    {
      "entropy": 5.630998611450195,
      "epoch": 0.5339420404922588,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004978635739612882,
      "loss": 5.4549,
      "mean_token_accuracy": 0.15072351321578026,
      "num_tokens": 11832984.0,
      "step": 5380
    },
    {
      "entropy": 5.620605039596557,
      "epoch": 0.5344382691544264,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004978587001347252,
      "loss": 5.4317,
      "mean_token_accuracy": 0.14728395640850067,
      "num_tokens": 11842918.0,
      "step": 5385
    },
    {
      "entropy": 5.566420698165894,
      "epoch": 0.5349344978165939,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004978538207817465,
      "loss": 5.4613,
      "mean_token_accuracy": 0.1546679764986038,
      "num_tokens": 11854364.0,
      "step": 5390
    },
    {
      "entropy": 5.636065721511841,
      "epoch": 0.5354307264787614,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004978489359024734,
      "loss": 5.551,
      "mean_token_accuracy": 0.1455624908208847,
      "num_tokens": 11866400.0,
      "step": 5395
    },
    {
      "entropy": 5.71275749206543,
      "epoch": 0.5359269551409289,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004978440454970268,
      "loss": 5.5318,
      "mean_token_accuracy": 0.14682282507419586,
      "num_tokens": 11877094.0,
      "step": 5400
    },
    {
      "entropy": 5.656115961074829,
      "epoch": 0.5364231838030965,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004978391495655281,
      "loss": 5.4814,
      "mean_token_accuracy": 0.15012079179286958,
      "num_tokens": 11887289.0,
      "step": 5405
    },
    {
      "entropy": 5.633537435531617,
      "epoch": 0.536919412465264,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004978342481080987,
      "loss": 5.4713,
      "mean_token_accuracy": 0.14486898928880693,
      "num_tokens": 11897700.0,
      "step": 5410
    },
    {
      "entropy": 5.528839206695556,
      "epoch": 0.5374156411274316,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004978293411248601,
      "loss": 5.3919,
      "mean_token_accuracy": 0.15333716124296187,
      "num_tokens": 11908623.0,
      "step": 5415
    },
    {
      "entropy": 5.619808053970337,
      "epoch": 0.537911869789599,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004978244286159339,
      "loss": 5.4142,
      "mean_token_accuracy": 0.15683804154396058,
      "num_tokens": 11917883.0,
      "step": 5420
    },
    {
      "entropy": 5.594229030609131,
      "epoch": 0.5384080984517666,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004978195105814421,
      "loss": 5.4572,
      "mean_token_accuracy": 0.15008229464292527,
      "num_tokens": 11929145.0,
      "step": 5425
    },
    {
      "entropy": 5.644812965393067,
      "epoch": 0.5389043271139341,
      "grad_norm": 0.875,
      "learning_rate": 0.0004978145870215065,
      "loss": 5.4511,
      "mean_token_accuracy": 0.14874640107154846,
      "num_tokens": 11940073.0,
      "step": 5430
    },
    {
      "entropy": 5.658492660522461,
      "epoch": 0.5394005557761016,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004978096579362493,
      "loss": 5.5777,
      "mean_token_accuracy": 0.14769978895783425,
      "num_tokens": 11953140.0,
      "step": 5435
    },
    {
      "entropy": 5.670812654495239,
      "epoch": 0.5398967844382692,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004978047233257927,
      "loss": 5.4993,
      "mean_token_accuracy": 0.15172669440507888,
      "num_tokens": 11963853.0,
      "step": 5440
    },
    {
      "entropy": 5.62924633026123,
      "epoch": 0.5403930131004366,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004977997831902591,
      "loss": 5.4477,
      "mean_token_accuracy": 0.15352184772491456,
      "num_tokens": 11975469.0,
      "step": 5445
    },
    {
      "entropy": 5.611998891830444,
      "epoch": 0.5408892417626042,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004977948375297709,
      "loss": 5.4473,
      "mean_token_accuracy": 0.15664440840482713,
      "num_tokens": 11985227.0,
      "step": 5450
    },
    {
      "entropy": 5.523879051208496,
      "epoch": 0.5413854704247717,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004977898863444509,
      "loss": 5.4524,
      "mean_token_accuracy": 0.14888169020414352,
      "num_tokens": 11995459.0,
      "step": 5455
    },
    {
      "entropy": 5.634063768386841,
      "epoch": 0.5418816990869393,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004977849296344217,
      "loss": 5.4949,
      "mean_token_accuracy": 0.1459165871143341,
      "num_tokens": 12007705.0,
      "step": 5460
    },
    {
      "entropy": 5.597906303405762,
      "epoch": 0.5423779277491068,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004977799673998063,
      "loss": 5.3994,
      "mean_token_accuracy": 0.1513359174132347,
      "num_tokens": 12019729.0,
      "step": 5465
    },
    {
      "entropy": 5.621879291534424,
      "epoch": 0.5428741564112743,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004977749996407278,
      "loss": 5.4778,
      "mean_token_accuracy": 0.14614856690168382,
      "num_tokens": 12031009.0,
      "step": 5470
    },
    {
      "entropy": 5.717142868041992,
      "epoch": 0.5433703850734418,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004977700263573094,
      "loss": 5.4436,
      "mean_token_accuracy": 0.15253398865461348,
      "num_tokens": 12041396.0,
      "step": 5475
    },
    {
      "entropy": 5.6496118068695065,
      "epoch": 0.5438666137356094,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004977650475496743,
      "loss": 5.5907,
      "mean_token_accuracy": 0.14329303279519082,
      "num_tokens": 12053112.0,
      "step": 5480
    },
    {
      "entropy": 5.657518005371093,
      "epoch": 0.5443628423977769,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004977600632179459,
      "loss": 5.4678,
      "mean_token_accuracy": 0.14574399664998056,
      "num_tokens": 12064823.0,
      "step": 5485
    },
    {
      "entropy": 5.6894450187683105,
      "epoch": 0.5448590710599445,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004977550733622482,
      "loss": 5.5912,
      "mean_token_accuracy": 0.14535179287195205,
      "num_tokens": 12076656.0,
      "step": 5490
    },
    {
      "entropy": 5.602209186553955,
      "epoch": 0.5453552997221119,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004977500779827045,
      "loss": 5.3786,
      "mean_token_accuracy": 0.1520494595170021,
      "num_tokens": 12086324.0,
      "step": 5495
    },
    {
      "entropy": 5.543206691741943,
      "epoch": 0.5458515283842795,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004977450770794389,
      "loss": 5.4208,
      "mean_token_accuracy": 0.14765701517462732,
      "num_tokens": 12098076.0,
      "step": 5500
    },
    {
      "entropy": 5.603697919845581,
      "epoch": 0.546347757046447,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004977400706525755,
      "loss": 5.4906,
      "mean_token_accuracy": 0.1446567229926586,
      "num_tokens": 12111282.0,
      "step": 5505
    },
    {
      "entropy": 5.623009395599365,
      "epoch": 0.5468439857086145,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004977350587022381,
      "loss": 5.4193,
      "mean_token_accuracy": 0.15579552501440047,
      "num_tokens": 12121932.0,
      "step": 5510
    },
    {
      "entropy": 5.595358753204346,
      "epoch": 0.5473402143707821,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004977300412285513,
      "loss": 5.4218,
      "mean_token_accuracy": 0.15426837503910065,
      "num_tokens": 12133040.0,
      "step": 5515
    },
    {
      "entropy": 5.622896909713745,
      "epoch": 0.5478364430329495,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004977250182316394,
      "loss": 5.485,
      "mean_token_accuracy": 0.15093854367733,
      "num_tokens": 12142577.0,
      "step": 5520
    },
    {
      "entropy": 5.6181727886199955,
      "epoch": 0.5483326716951171,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000497719989711627,
      "loss": 5.4525,
      "mean_token_accuracy": 0.15297684520483018,
      "num_tokens": 12153941.0,
      "step": 5525
    },
    {
      "entropy": 5.608196783065796,
      "epoch": 0.5488289003572846,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004977149556686387,
      "loss": 5.5174,
      "mean_token_accuracy": 0.14787946194410323,
      "num_tokens": 12165347.0,
      "step": 5530
    },
    {
      "entropy": 5.545637702941894,
      "epoch": 0.5493251290194522,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004977099161027996,
      "loss": 5.3864,
      "mean_token_accuracy": 0.15185628309845925,
      "num_tokens": 12175705.0,
      "step": 5535
    },
    {
      "entropy": 5.652411890029907,
      "epoch": 0.5498213576816197,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004977048710142342,
      "loss": 5.4364,
      "mean_token_accuracy": 0.1537328541278839,
      "num_tokens": 12186781.0,
      "step": 5540
    },
    {
      "entropy": 5.666394376754761,
      "epoch": 0.5503175863437872,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004976998204030681,
      "loss": 5.4772,
      "mean_token_accuracy": 0.14764580726623536,
      "num_tokens": 12197145.0,
      "step": 5545
    },
    {
      "entropy": 5.549248790740966,
      "epoch": 0.5508138150059547,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004976947642694262,
      "loss": 5.4741,
      "mean_token_accuracy": 0.15212984159588813,
      "num_tokens": 12207637.0,
      "step": 5550
    },
    {
      "entropy": 5.556308317184448,
      "epoch": 0.5513100436681223,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004976897026134341,
      "loss": 5.3988,
      "mean_token_accuracy": 0.15345845073461534,
      "num_tokens": 12218156.0,
      "step": 5555
    },
    {
      "entropy": 5.653855180740356,
      "epoch": 0.5518062723302898,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004976846354352172,
      "loss": 5.4957,
      "mean_token_accuracy": 0.14872472286224364,
      "num_tokens": 12229987.0,
      "step": 5560
    },
    {
      "entropy": 5.643626880645752,
      "epoch": 0.5523025009924574,
      "grad_norm": 0.875,
      "learning_rate": 0.0004976795627349012,
      "loss": 5.4659,
      "mean_token_accuracy": 0.14570514112710953,
      "num_tokens": 12240508.0,
      "step": 5565
    },
    {
      "entropy": 5.559814167022705,
      "epoch": 0.5527987296546248,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004976744845126118,
      "loss": 5.4069,
      "mean_token_accuracy": 0.15187161713838576,
      "num_tokens": 12250317.0,
      "step": 5570
    },
    {
      "entropy": 5.577573823928833,
      "epoch": 0.5532949583167924,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004976694007684751,
      "loss": 5.4525,
      "mean_token_accuracy": 0.15375591665506363,
      "num_tokens": 12262104.0,
      "step": 5575
    },
    {
      "entropy": 5.699751091003418,
      "epoch": 0.5537911869789599,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000497664311502617,
      "loss": 5.4941,
      "mean_token_accuracy": 0.15318164080381394,
      "num_tokens": 12273259.0,
      "step": 5580
    },
    {
      "entropy": 5.682411003112793,
      "epoch": 0.5542874156411274,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004976592167151638,
      "loss": 5.5331,
      "mean_token_accuracy": 0.14391404539346694,
      "num_tokens": 12285504.0,
      "step": 5585
    },
    {
      "entropy": 5.644524049758911,
      "epoch": 0.554783644303295,
      "grad_norm": 1.0,
      "learning_rate": 0.0004976541164062418,
      "loss": 5.4584,
      "mean_token_accuracy": 0.14769638627767562,
      "num_tokens": 12296512.0,
      "step": 5590
    },
    {
      "entropy": 5.512802314758301,
      "epoch": 0.5552798729654624,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004976490105759774,
      "loss": 5.3308,
      "mean_token_accuracy": 0.15871942639350892,
      "num_tokens": 12307948.0,
      "step": 5595
    },
    {
      "entropy": 5.612875652313233,
      "epoch": 0.55577610162763,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004976438992244976,
      "loss": 5.5017,
      "mean_token_accuracy": 0.14463188201189042,
      "num_tokens": 12320020.0,
      "step": 5600
    },
    {
      "entropy": 5.6500156879425045,
      "epoch": 0.5562723302897975,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004976387823519286,
      "loss": 5.3592,
      "mean_token_accuracy": 0.15429917871952056,
      "num_tokens": 12331091.0,
      "step": 5605
    },
    {
      "entropy": 5.6125414848327635,
      "epoch": 0.5567685589519651,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004976336599583975,
      "loss": 5.4673,
      "mean_token_accuracy": 0.14630457013845444,
      "num_tokens": 12342621.0,
      "step": 5610
    },
    {
      "entropy": 5.741844367980957,
      "epoch": 0.5572647876141326,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004976285320440315,
      "loss": 5.7035,
      "mean_token_accuracy": 0.13585652485489846,
      "num_tokens": 12353330.0,
      "step": 5615
    },
    {
      "entropy": 5.693860006332398,
      "epoch": 0.5577610162763001,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004976233986089577,
      "loss": 5.4623,
      "mean_token_accuracy": 0.14677181020379065,
      "num_tokens": 12364418.0,
      "step": 5620
    },
    {
      "entropy": 5.577788209915161,
      "epoch": 0.5582572449384676,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004976182596533031,
      "loss": 5.4472,
      "mean_token_accuracy": 0.15234579965472222,
      "num_tokens": 12377022.0,
      "step": 5625
    },
    {
      "entropy": 5.659402227401733,
      "epoch": 0.5587534736006352,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004976131151771955,
      "loss": 5.5228,
      "mean_token_accuracy": 0.14480556398630143,
      "num_tokens": 12387799.0,
      "step": 5630
    },
    {
      "entropy": 5.5746238231658936,
      "epoch": 0.5592497022628027,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004976079651807623,
      "loss": 5.4466,
      "mean_token_accuracy": 0.14236423075199128,
      "num_tokens": 12397387.0,
      "step": 5635
    },
    {
      "entropy": 5.55435004234314,
      "epoch": 0.5597459309249703,
      "grad_norm": 0.875,
      "learning_rate": 0.0004976028096641313,
      "loss": 5.3789,
      "mean_token_accuracy": 0.15377479195594787,
      "num_tokens": 12407994.0,
      "step": 5640
    },
    {
      "entropy": 5.710486555099488,
      "epoch": 0.5602421595871377,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004975976486274302,
      "loss": 5.5692,
      "mean_token_accuracy": 0.1448943331837654,
      "num_tokens": 12419124.0,
      "step": 5645
    },
    {
      "entropy": 5.650815391540528,
      "epoch": 0.5607383882493053,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004975924820707872,
      "loss": 5.4444,
      "mean_token_accuracy": 0.1529550939798355,
      "num_tokens": 12428699.0,
      "step": 5650
    },
    {
      "entropy": 5.646983575820923,
      "epoch": 0.5612346169114728,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004975873099943302,
      "loss": 5.5266,
      "mean_token_accuracy": 0.14555705338716507,
      "num_tokens": 12439025.0,
      "step": 5655
    },
    {
      "entropy": 5.656699848175049,
      "epoch": 0.5617308455736403,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004975821323981875,
      "loss": 5.4778,
      "mean_token_accuracy": 0.14742110371589662,
      "num_tokens": 12449439.0,
      "step": 5660
    },
    {
      "entropy": 5.654876232147217,
      "epoch": 0.5622270742358079,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004975769492824877,
      "loss": 5.421,
      "mean_token_accuracy": 0.15080877095460893,
      "num_tokens": 12460066.0,
      "step": 5665
    },
    {
      "entropy": 5.6223005771636965,
      "epoch": 0.5627233028979753,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004975717606473589,
      "loss": 5.4397,
      "mean_token_accuracy": 0.15309426784515381,
      "num_tokens": 12471093.0,
      "step": 5670
    },
    {
      "entropy": 5.651199150085449,
      "epoch": 0.5632195315601429,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004975665664929302,
      "loss": 5.4579,
      "mean_token_accuracy": 0.1513369768857956,
      "num_tokens": 12481181.0,
      "step": 5675
    },
    {
      "entropy": 5.674373435974121,
      "epoch": 0.5637157602223104,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004975613668193303,
      "loss": 5.5064,
      "mean_token_accuracy": 0.14478076249361038,
      "num_tokens": 12493068.0,
      "step": 5680
    },
    {
      "entropy": 5.5631945610046385,
      "epoch": 0.564211988884478,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000497556161626688,
      "loss": 5.4497,
      "mean_token_accuracy": 0.1487413913011551,
      "num_tokens": 12504025.0,
      "step": 5685
    },
    {
      "entropy": 5.607772922515869,
      "epoch": 0.5647082175466455,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004975509509151325,
      "loss": 5.4486,
      "mean_token_accuracy": 0.1523358181118965,
      "num_tokens": 12514567.0,
      "step": 5690
    },
    {
      "entropy": 5.702475070953369,
      "epoch": 0.565204446208813,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004975457346847929,
      "loss": 5.4804,
      "mean_token_accuracy": 0.14762943536043166,
      "num_tokens": 12525088.0,
      "step": 5695
    },
    {
      "entropy": 5.618128538131714,
      "epoch": 0.5657006748709805,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004975405129357986,
      "loss": 5.4489,
      "mean_token_accuracy": 0.1503733605146408,
      "num_tokens": 12536002.0,
      "step": 5700
    },
    {
      "entropy": 5.632601690292359,
      "epoch": 0.5661969035331481,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004975352856682792,
      "loss": 5.4669,
      "mean_token_accuracy": 0.15055823028087617,
      "num_tokens": 12547468.0,
      "step": 5705
    },
    {
      "entropy": 5.667331027984619,
      "epoch": 0.5666931321953156,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004975300528823642,
      "loss": 5.4344,
      "mean_token_accuracy": 0.14942139089107515,
      "num_tokens": 12557670.0,
      "step": 5710
    },
    {
      "entropy": 5.5996208667755125,
      "epoch": 0.5671893608574832,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004975248145781834,
      "loss": 5.479,
      "mean_token_accuracy": 0.14795504957437516,
      "num_tokens": 12568425.0,
      "step": 5715
    },
    {
      "entropy": 5.55127854347229,
      "epoch": 0.5676855895196506,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004975195707558667,
      "loss": 5.4463,
      "mean_token_accuracy": 0.14945666939020158,
      "num_tokens": 12579138.0,
      "step": 5720
    },
    {
      "entropy": 5.597843265533447,
      "epoch": 0.5681818181818182,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004975143214155443,
      "loss": 5.4643,
      "mean_token_accuracy": 0.14927031174302102,
      "num_tokens": 12589743.0,
      "step": 5725
    },
    {
      "entropy": 5.586509704589844,
      "epoch": 0.5686780468439857,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004975090665573459,
      "loss": 5.447,
      "mean_token_accuracy": 0.15376916229724885,
      "num_tokens": 12600648.0,
      "step": 5730
    },
    {
      "entropy": 5.580662059783935,
      "epoch": 0.5691742755061532,
      "grad_norm": 0.875,
      "learning_rate": 0.0004975038061814023,
      "loss": 5.452,
      "mean_token_accuracy": 0.1524510756134987,
      "num_tokens": 12611369.0,
      "step": 5735
    },
    {
      "entropy": 5.678642797470093,
      "epoch": 0.5696705041683208,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004974985402878437,
      "loss": 5.4363,
      "mean_token_accuracy": 0.1489766277372837,
      "num_tokens": 12621992.0,
      "step": 5740
    },
    {
      "entropy": 5.610474920272827,
      "epoch": 0.5701667328304882,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004974932688768007,
      "loss": 5.5164,
      "mean_token_accuracy": 0.14607659056782724,
      "num_tokens": 12632463.0,
      "step": 5745
    },
    {
      "entropy": 5.548481416702271,
      "epoch": 0.5706629614926558,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004974879919484042,
      "loss": 5.3712,
      "mean_token_accuracy": 0.15110822319984435,
      "num_tokens": 12642282.0,
      "step": 5750
    },
    {
      "entropy": 5.5748039245605465,
      "epoch": 0.5711591901548233,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004974827095027848,
      "loss": 5.4002,
      "mean_token_accuracy": 0.1541622370481491,
      "num_tokens": 12653495.0,
      "step": 5755
    },
    {
      "entropy": 5.586807537078857,
      "epoch": 0.5716554188169909,
      "grad_norm": 0.875,
      "learning_rate": 0.0004974774215400736,
      "loss": 5.3364,
      "mean_token_accuracy": 0.15577918887138367,
      "num_tokens": 12664322.0,
      "step": 5760
    },
    {
      "entropy": 5.507334613800049,
      "epoch": 0.5721516474791584,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004974721280604019,
      "loss": 5.4354,
      "mean_token_accuracy": 0.15019209086894988,
      "num_tokens": 12673981.0,
      "step": 5765
    },
    {
      "entropy": 5.526956558227539,
      "epoch": 0.572647876141326,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004974668290639007,
      "loss": 5.4469,
      "mean_token_accuracy": 0.1492769457399845,
      "num_tokens": 12684778.0,
      "step": 5770
    },
    {
      "entropy": 5.641401767730713,
      "epoch": 0.5731441048034934,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004974615245507014,
      "loss": 5.4059,
      "mean_token_accuracy": 0.1538984641432762,
      "num_tokens": 12695287.0,
      "step": 5775
    },
    {
      "entropy": 5.5436258792877195,
      "epoch": 0.573640333465661,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004974562145209359,
      "loss": 5.4537,
      "mean_token_accuracy": 0.14565087854862213,
      "num_tokens": 12705773.0,
      "step": 5780
    },
    {
      "entropy": 5.538467264175415,
      "epoch": 0.5741365621278285,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004974508989747354,
      "loss": 5.4019,
      "mean_token_accuracy": 0.1574517771601677,
      "num_tokens": 12716783.0,
      "step": 5785
    },
    {
      "entropy": 5.607895183563232,
      "epoch": 0.5746327907899961,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004974455779122321,
      "loss": 5.4516,
      "mean_token_accuracy": 0.14383325651288031,
      "num_tokens": 12728146.0,
      "step": 5790
    },
    {
      "entropy": 5.656476640701294,
      "epoch": 0.5751290194521635,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004974402513335577,
      "loss": 5.4746,
      "mean_token_accuracy": 0.14959257021546363,
      "num_tokens": 12739683.0,
      "step": 5795
    },
    {
      "entropy": 5.6649665355682375,
      "epoch": 0.5756252481143311,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004974349192388444,
      "loss": 5.4589,
      "mean_token_accuracy": 0.14662942439317703,
      "num_tokens": 12750257.0,
      "step": 5800
    },
    {
      "entropy": 5.688394498825073,
      "epoch": 0.5761214767764986,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004974295816282245,
      "loss": 5.5324,
      "mean_token_accuracy": 0.1480068877339363,
      "num_tokens": 12761055.0,
      "step": 5805
    },
    {
      "entropy": 5.610796308517456,
      "epoch": 0.5766177054386661,
      "grad_norm": 1.0,
      "learning_rate": 0.0004974242385018303,
      "loss": 5.4721,
      "mean_token_accuracy": 0.15184674710035323,
      "num_tokens": 12771014.0,
      "step": 5810
    },
    {
      "entropy": 5.567866468429566,
      "epoch": 0.5771139341008337,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000497418889859794,
      "loss": 5.3818,
      "mean_token_accuracy": 0.15442348420619964,
      "num_tokens": 12782623.0,
      "step": 5815
    },
    {
      "entropy": 5.669680786132813,
      "epoch": 0.5776101627630011,
      "grad_norm": 1.0,
      "learning_rate": 0.0004974135357022488,
      "loss": 5.5093,
      "mean_token_accuracy": 0.14822030067443848,
      "num_tokens": 12793533.0,
      "step": 5820
    },
    {
      "entropy": 5.5356302738189695,
      "epoch": 0.5781063914251687,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004974081760293269,
      "loss": 5.3729,
      "mean_token_accuracy": 0.15088124424219132,
      "num_tokens": 12804389.0,
      "step": 5825
    },
    {
      "entropy": 5.58860149383545,
      "epoch": 0.5786026200873362,
      "grad_norm": 0.875,
      "learning_rate": 0.0004974028108411617,
      "loss": 5.4681,
      "mean_token_accuracy": 0.15093735009431838,
      "num_tokens": 12815672.0,
      "step": 5830
    },
    {
      "entropy": 5.639421892166138,
      "epoch": 0.5790988487495038,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004973974401378859,
      "loss": 5.4668,
      "mean_token_accuracy": 0.14916255176067353,
      "num_tokens": 12826113.0,
      "step": 5835
    },
    {
      "entropy": 5.580629539489746,
      "epoch": 0.5795950774116713,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004973920639196329,
      "loss": 5.4045,
      "mean_token_accuracy": 0.15356406718492507,
      "num_tokens": 12835941.0,
      "step": 5840
    },
    {
      "entropy": 5.596003293991089,
      "epoch": 0.5800913060738389,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004973866821865358,
      "loss": 5.4899,
      "mean_token_accuracy": 0.14188268631696702,
      "num_tokens": 12846186.0,
      "step": 5845
    },
    {
      "entropy": 5.65446286201477,
      "epoch": 0.5805875347360063,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004973812949387283,
      "loss": 5.4459,
      "mean_token_accuracy": 0.15039265230298043,
      "num_tokens": 12856380.0,
      "step": 5850
    },
    {
      "entropy": 5.568442964553833,
      "epoch": 0.5810837633981739,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004973759021763439,
      "loss": 5.4795,
      "mean_token_accuracy": 0.15042850375175476,
      "num_tokens": 12865999.0,
      "step": 5855
    },
    {
      "entropy": 5.599948740005493,
      "epoch": 0.5815799920603414,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004973705038995162,
      "loss": 5.3781,
      "mean_token_accuracy": 0.14677538573741913,
      "num_tokens": 12876388.0,
      "step": 5860
    },
    {
      "entropy": 5.578648567199707,
      "epoch": 0.582076220722509,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004973651001083792,
      "loss": 5.3704,
      "mean_token_accuracy": 0.1494818851351738,
      "num_tokens": 12885862.0,
      "step": 5865
    },
    {
      "entropy": 5.555972909927368,
      "epoch": 0.5825724493846764,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004973596908030669,
      "loss": 5.4721,
      "mean_token_accuracy": 0.14565781652927398,
      "num_tokens": 12897489.0,
      "step": 5870
    },
    {
      "entropy": 5.656794214248658,
      "epoch": 0.583068678046844,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004973542759837135,
      "loss": 5.4925,
      "mean_token_accuracy": 0.14666834399104117,
      "num_tokens": 12907395.0,
      "step": 5875
    },
    {
      "entropy": 5.604556083679199,
      "epoch": 0.5835649067090115,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000497348855650453,
      "loss": 5.4358,
      "mean_token_accuracy": 0.15660476684570312,
      "num_tokens": 12919439.0,
      "step": 5880
    },
    {
      "entropy": 5.640682888031006,
      "epoch": 0.584061135371179,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004973434298034201,
      "loss": 5.5249,
      "mean_token_accuracy": 0.14780937358736992,
      "num_tokens": 12931892.0,
      "step": 5885
    },
    {
      "entropy": 5.666601848602295,
      "epoch": 0.5845573640333466,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004973379984427492,
      "loss": 5.5692,
      "mean_token_accuracy": 0.14671001508831977,
      "num_tokens": 12943189.0,
      "step": 5890
    },
    {
      "entropy": 5.634539604187012,
      "epoch": 0.585053592695514,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000497332561568575,
      "loss": 5.5137,
      "mean_token_accuracy": 0.1456242695450783,
      "num_tokens": 12954662.0,
      "step": 5895
    },
    {
      "entropy": 5.624458932876587,
      "epoch": 0.5855498213576816,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004973271191810325,
      "loss": 5.4486,
      "mean_token_accuracy": 0.14609898179769515,
      "num_tokens": 12964924.0,
      "step": 5900
    },
    {
      "entropy": 5.618928527832031,
      "epoch": 0.5860460500198491,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004973216712802564,
      "loss": 5.5248,
      "mean_token_accuracy": 0.14595869034528733,
      "num_tokens": 12976483.0,
      "step": 5905
    },
    {
      "entropy": 5.583964586257935,
      "epoch": 0.5865422786820167,
      "grad_norm": 0.875,
      "learning_rate": 0.000497316217866382,
      "loss": 5.4224,
      "mean_token_accuracy": 0.15400040447711943,
      "num_tokens": 12988054.0,
      "step": 5910
    },
    {
      "entropy": 5.636979436874389,
      "epoch": 0.5870385073441842,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004973107589395444,
      "loss": 5.4794,
      "mean_token_accuracy": 0.15200079306960107,
      "num_tokens": 13000596.0,
      "step": 5915
    },
    {
      "entropy": 5.478622388839722,
      "epoch": 0.5875347360063518,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004973052944998791,
      "loss": 5.3041,
      "mean_token_accuracy": 0.15711194574832915,
      "num_tokens": 13012360.0,
      "step": 5920
    },
    {
      "entropy": 5.5978460788726805,
      "epoch": 0.5880309646685192,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004972998245475214,
      "loss": 5.4745,
      "mean_token_accuracy": 0.14887128621339799,
      "num_tokens": 13023806.0,
      "step": 5925
    },
    {
      "entropy": 5.599380207061768,
      "epoch": 0.5885271933306868,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000497294349082607,
      "loss": 5.3887,
      "mean_token_accuracy": 0.1511357419192791,
      "num_tokens": 13034049.0,
      "step": 5930
    },
    {
      "entropy": 5.611827325820923,
      "epoch": 0.5890234219928543,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004972888681052719,
      "loss": 5.5051,
      "mean_token_accuracy": 0.14897997081279754,
      "num_tokens": 13045006.0,
      "step": 5935
    },
    {
      "entropy": 5.687719774246216,
      "epoch": 0.5895196506550219,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004972833816156518,
      "loss": 5.4084,
      "mean_token_accuracy": 0.14930811673402786,
      "num_tokens": 13055337.0,
      "step": 5940
    },
    {
      "entropy": 5.543897342681885,
      "epoch": 0.5900158793171894,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004972778896138828,
      "loss": 5.4027,
      "mean_token_accuracy": 0.15560031235218047,
      "num_tokens": 13065267.0,
      "step": 5945
    },
    {
      "entropy": 5.578498983383179,
      "epoch": 0.5905121079793569,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004972723921001011,
      "loss": 5.467,
      "mean_token_accuracy": 0.15096500515937805,
      "num_tokens": 13075173.0,
      "step": 5950
    },
    {
      "entropy": 5.658969116210938,
      "epoch": 0.5910083366415244,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004972668890744431,
      "loss": 5.4562,
      "mean_token_accuracy": 0.14907519966363908,
      "num_tokens": 13086848.0,
      "step": 5955
    },
    {
      "entropy": 5.581852436065674,
      "epoch": 0.5915045653036919,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004972613805370452,
      "loss": 5.4341,
      "mean_token_accuracy": 0.15141052827239038,
      "num_tokens": 13096865.0,
      "step": 5960
    },
    {
      "entropy": 5.562904024124146,
      "epoch": 0.5920007939658595,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004972558664880439,
      "loss": 5.415,
      "mean_token_accuracy": 0.15210264623165132,
      "num_tokens": 13107907.0,
      "step": 5965
    },
    {
      "entropy": 5.540717506408692,
      "epoch": 0.592497022628027,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000497250346927576,
      "loss": 5.3277,
      "mean_token_accuracy": 0.15800193101167678,
      "num_tokens": 13118171.0,
      "step": 5970
    },
    {
      "entropy": 5.570704555511474,
      "epoch": 0.5929932512901945,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004972448218557785,
      "loss": 5.3144,
      "mean_token_accuracy": 0.16167542934417725,
      "num_tokens": 13127042.0,
      "step": 5975
    },
    {
      "entropy": 5.638939285278321,
      "epoch": 0.593489479952362,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004972392912727882,
      "loss": 5.5744,
      "mean_token_accuracy": 0.15045227780938147,
      "num_tokens": 13137026.0,
      "step": 5980
    },
    {
      "entropy": 5.616703224182129,
      "epoch": 0.5939857086145296,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004972337551787425,
      "loss": 5.4344,
      "mean_token_accuracy": 0.15230552107095718,
      "num_tokens": 13146947.0,
      "step": 5985
    },
    {
      "entropy": 5.668242979049682,
      "epoch": 0.5944819372766971,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004972282135737785,
      "loss": 5.4786,
      "mean_token_accuracy": 0.14680141657590867,
      "num_tokens": 13158714.0,
      "step": 5990
    },
    {
      "entropy": 5.639590787887573,
      "epoch": 0.5949781659388647,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004972226664580337,
      "loss": 5.4931,
      "mean_token_accuracy": 0.1462373726069927,
      "num_tokens": 13170156.0,
      "step": 5995
    },
    {
      "entropy": 5.515760707855224,
      "epoch": 0.5954743946010321,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004972171138316456,
      "loss": 5.3974,
      "mean_token_accuracy": 0.1493689924478531,
      "num_tokens": 13179919.0,
      "step": 6000
    },
    {
      "epoch": 0.5954743946010321,
      "eval_entropy": 5.392710745773623,
      "eval_loss": 5.513803482055664,
      "eval_mean_token_accuracy": 0.15261998519223527,
      "eval_num_tokens": 13179919.0,
      "eval_runtime": 20.7006,
      "eval_samples_per_second": 1626.475,
      "eval_steps_per_second": 203.327,
      "step": 6000
    }
  ],
  "logging_steps": 5,
  "max_steps": 100750,
  "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": 3273706621992960.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}