ita-latn-10mb-after-ppt-Dp-100mb-ckpt500_seed3407/checkpoint-30000/trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9773719730051607,
  "eval_steps": 3000,
  "global_step": 30000,
  "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
    },
    {
      "entropy": 5.596977090835571,
      "epoch": 0.5959706232631997,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004972115556947518,
      "loss": 5.4779,
      "mean_token_accuracy": 0.1541017025709152,
      "num_tokens": 13190777.0,
      "step": 6005
    },
    {
      "entropy": 5.6757081031799315,
      "epoch": 0.5964668519253672,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004972059920474904,
      "loss": 5.4613,
      "mean_token_accuracy": 0.14994437322020532,
      "num_tokens": 13202179.0,
      "step": 6010
    },
    {
      "entropy": 5.580601167678833,
      "epoch": 0.5969630805875348,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004972004228899991,
      "loss": 5.4181,
      "mean_token_accuracy": 0.1509955532848835,
      "num_tokens": 13212764.0,
      "step": 6015
    },
    {
      "entropy": 5.560248231887817,
      "epoch": 0.5974593092497023,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004971948482224162,
      "loss": 5.4524,
      "mean_token_accuracy": 0.1504641056060791,
      "num_tokens": 13223343.0,
      "step": 6020
    },
    {
      "entropy": 5.629840755462647,
      "epoch": 0.5979555379118698,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004971892680448798,
      "loss": 5.4253,
      "mean_token_accuracy": 0.15569519847631455,
      "num_tokens": 13233839.0,
      "step": 6025
    },
    {
      "entropy": 5.661754989624024,
      "epoch": 0.5984517665740373,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004971836823575283,
      "loss": 5.5301,
      "mean_token_accuracy": 0.14222781658172606,
      "num_tokens": 13243854.0,
      "step": 6030
    },
    {
      "entropy": 5.647662878036499,
      "epoch": 0.5989479952362048,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004971780911605003,
      "loss": 5.4551,
      "mean_token_accuracy": 0.15329526215791703,
      "num_tokens": 13256073.0,
      "step": 6035
    },
    {
      "entropy": 5.599632644653321,
      "epoch": 0.5994442238983724,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004971724944539343,
      "loss": 5.497,
      "mean_token_accuracy": 0.14713055565953254,
      "num_tokens": 13267568.0,
      "step": 6040
    },
    {
      "entropy": 5.649599838256836,
      "epoch": 0.5999404525605398,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004971668922379692,
      "loss": 5.4618,
      "mean_token_accuracy": 0.1436742812395096,
      "num_tokens": 13278829.0,
      "step": 6045
    },
    {
      "entropy": 5.61177659034729,
      "epoch": 0.6004366812227074,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000497161284512744,
      "loss": 5.4336,
      "mean_token_accuracy": 0.15138799101114273,
      "num_tokens": 13289156.0,
      "step": 6050
    },
    {
      "entropy": 5.493603658676148,
      "epoch": 0.6009329098848749,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004971556712783976,
      "loss": 5.3112,
      "mean_token_accuracy": 0.1571349173784256,
      "num_tokens": 13299817.0,
      "step": 6055
    },
    {
      "entropy": 5.506628179550171,
      "epoch": 0.6014291385470425,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004971500525350692,
      "loss": 5.3852,
      "mean_token_accuracy": 0.15314638912677764,
      "num_tokens": 13310290.0,
      "step": 6060
    },
    {
      "entropy": 5.533604288101197,
      "epoch": 0.60192536720921,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004971444282828983,
      "loss": 5.4803,
      "mean_token_accuracy": 0.1499824970960617,
      "num_tokens": 13321747.0,
      "step": 6065
    },
    {
      "entropy": 5.6678197383880615,
      "epoch": 0.6024215958713776,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004971387985220242,
      "loss": 5.4208,
      "mean_token_accuracy": 0.1545249879360199,
      "num_tokens": 13333249.0,
      "step": 6070
    },
    {
      "entropy": 5.5247804641723635,
      "epoch": 0.602917824533545,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004971331632525866,
      "loss": 5.4065,
      "mean_token_accuracy": 0.15247410610318185,
      "num_tokens": 13345336.0,
      "step": 6075
    },
    {
      "entropy": 5.669757747650147,
      "epoch": 0.6034140531957126,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000497127522474725,
      "loss": 5.5524,
      "mean_token_accuracy": 0.14289941713213922,
      "num_tokens": 13355797.0,
      "step": 6080
    },
    {
      "entropy": 5.6765429973602295,
      "epoch": 0.6039102818578801,
      "grad_norm": 0.875,
      "learning_rate": 0.0004971218761885797,
      "loss": 5.4941,
      "mean_token_accuracy": 0.14894456416368484,
      "num_tokens": 13367058.0,
      "step": 6085
    },
    {
      "entropy": 5.602103137969971,
      "epoch": 0.6044065105200477,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004971162243942904,
      "loss": 5.4538,
      "mean_token_accuracy": 0.14952910095453262,
      "num_tokens": 13378579.0,
      "step": 6090
    },
    {
      "entropy": 5.651467227935791,
      "epoch": 0.6049027391822152,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004971105670919975,
      "loss": 5.4143,
      "mean_token_accuracy": 0.15067921876907348,
      "num_tokens": 13389240.0,
      "step": 6095
    },
    {
      "entropy": 5.645436096191406,
      "epoch": 0.6053989678443827,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000497104904281841,
      "loss": 5.5716,
      "mean_token_accuracy": 0.14636854827404022,
      "num_tokens": 13401895.0,
      "step": 6100
    },
    {
      "entropy": 5.68249282836914,
      "epoch": 0.6058951965065502,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004970992359639616,
      "loss": 5.5284,
      "mean_token_accuracy": 0.1433717429637909,
      "num_tokens": 13413334.0,
      "step": 6105
    },
    {
      "entropy": 5.65785961151123,
      "epoch": 0.6063914251687177,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004970935621384996,
      "loss": 5.4466,
      "mean_token_accuracy": 0.15460411906242372,
      "num_tokens": 13423371.0,
      "step": 6110
    },
    {
      "entropy": 5.650516653060913,
      "epoch": 0.6068876538308853,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004970878828055959,
      "loss": 5.4274,
      "mean_token_accuracy": 0.15732789784669876,
      "num_tokens": 13434782.0,
      "step": 6115
    },
    {
      "entropy": 5.564152765274048,
      "epoch": 0.6073838824930528,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004970821979653912,
      "loss": 5.414,
      "mean_token_accuracy": 0.14711159765720366,
      "num_tokens": 13444493.0,
      "step": 6120
    },
    {
      "entropy": 5.58135290145874,
      "epoch": 0.6078801111552203,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004970765076180265,
      "loss": 5.4037,
      "mean_token_accuracy": 0.1521766349673271,
      "num_tokens": 13455518.0,
      "step": 6125
    },
    {
      "entropy": 5.663119697570801,
      "epoch": 0.6083763398173878,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004970708117636431,
      "loss": 5.4881,
      "mean_token_accuracy": 0.1530603289604187,
      "num_tokens": 13467631.0,
      "step": 6130
    },
    {
      "entropy": 5.553421258926392,
      "epoch": 0.6088725684795554,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000497065110402382,
      "loss": 5.4247,
      "mean_token_accuracy": 0.14682705998420714,
      "num_tokens": 13478217.0,
      "step": 6135
    },
    {
      "entropy": 5.6026238918304445,
      "epoch": 0.6093687971417229,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004970594035343846,
      "loss": 5.405,
      "mean_token_accuracy": 0.15311249345541,
      "num_tokens": 13489169.0,
      "step": 6140
    },
    {
      "entropy": 5.589141130447388,
      "epoch": 0.6098650258038905,
      "grad_norm": 1.0,
      "learning_rate": 0.0004970536911597926,
      "loss": 5.4172,
      "mean_token_accuracy": 0.14844012260437012,
      "num_tokens": 13499273.0,
      "step": 6145
    },
    {
      "entropy": 5.6402270793914795,
      "epoch": 0.6103612544660579,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004970479732787476,
      "loss": 5.461,
      "mean_token_accuracy": 0.15369903594255446,
      "num_tokens": 13509664.0,
      "step": 6150
    },
    {
      "entropy": 5.549899244308472,
      "epoch": 0.6108574831282255,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004970422498913911,
      "loss": 5.3488,
      "mean_token_accuracy": 0.14968640953302384,
      "num_tokens": 13520288.0,
      "step": 6155
    },
    {
      "entropy": 5.564370441436767,
      "epoch": 0.611353711790393,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004970365209978655,
      "loss": 5.4345,
      "mean_token_accuracy": 0.15311274081468582,
      "num_tokens": 13531167.0,
      "step": 6160
    },
    {
      "entropy": 5.590855407714844,
      "epoch": 0.6118499404525606,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004970307865983124,
      "loss": 5.4787,
      "mean_token_accuracy": 0.15183627977967262,
      "num_tokens": 13544109.0,
      "step": 6165
    },
    {
      "entropy": 5.6111372947692875,
      "epoch": 0.6123461691147281,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004970250466928744,
      "loss": 5.4087,
      "mean_token_accuracy": 0.15262484773993493,
      "num_tokens": 13554935.0,
      "step": 6170
    },
    {
      "entropy": 5.625780916213989,
      "epoch": 0.6128423977768956,
      "grad_norm": 0.875,
      "learning_rate": 0.0004970193012816936,
      "loss": 5.4213,
      "mean_token_accuracy": 0.15170982629060745,
      "num_tokens": 13565959.0,
      "step": 6175
    },
    {
      "entropy": 5.550536394119263,
      "epoch": 0.6133386264390631,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004970135503649125,
      "loss": 5.4496,
      "mean_token_accuracy": 0.14747072756290436,
      "num_tokens": 13577883.0,
      "step": 6180
    },
    {
      "entropy": 5.58063268661499,
      "epoch": 0.6138348551012306,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004970077939426738,
      "loss": 5.4037,
      "mean_token_accuracy": 0.15324754267930984,
      "num_tokens": 13588597.0,
      "step": 6185
    },
    {
      "entropy": 5.587204360961914,
      "epoch": 0.6143310837633982,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004970020320151202,
      "loss": 5.3461,
      "mean_token_accuracy": 0.15477967262268066,
      "num_tokens": 13598772.0,
      "step": 6190
    },
    {
      "entropy": 5.640543794631958,
      "epoch": 0.6148273124255657,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004969962645823946,
      "loss": 5.4372,
      "mean_token_accuracy": 0.1521898865699768,
      "num_tokens": 13609696.0,
      "step": 6195
    },
    {
      "entropy": 5.557079124450683,
      "epoch": 0.6153235410877332,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00049699049164464,
      "loss": 5.4172,
      "mean_token_accuracy": 0.15375182554125785,
      "num_tokens": 13619389.0,
      "step": 6200
    },
    {
      "entropy": 5.561429595947265,
      "epoch": 0.6158197697499007,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004969847132019995,
      "loss": 5.394,
      "mean_token_accuracy": 0.1538977637887001,
      "num_tokens": 13632133.0,
      "step": 6205
    },
    {
      "entropy": 5.571953344345093,
      "epoch": 0.6163159984120683,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004969789292546164,
      "loss": 5.3465,
      "mean_token_accuracy": 0.1636103093624115,
      "num_tokens": 13642715.0,
      "step": 6210
    },
    {
      "entropy": 5.5559193134307865,
      "epoch": 0.6168122270742358,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004969731398026343,
      "loss": 5.4485,
      "mean_token_accuracy": 0.15080342590808868,
      "num_tokens": 13653223.0,
      "step": 6215
    },
    {
      "entropy": 5.517760848999023,
      "epoch": 0.6173084557364034,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004969673448461966,
      "loss": 5.4323,
      "mean_token_accuracy": 0.1504549577832222,
      "num_tokens": 13664206.0,
      "step": 6220
    },
    {
      "entropy": 5.590613555908203,
      "epoch": 0.6178046843985708,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000496961544385447,
      "loss": 5.4383,
      "mean_token_accuracy": 0.1557757943868637,
      "num_tokens": 13673974.0,
      "step": 6225
    },
    {
      "entropy": 5.583416986465454,
      "epoch": 0.6183009130607384,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004969557384205295,
      "loss": 5.369,
      "mean_token_accuracy": 0.1549408406019211,
      "num_tokens": 13685189.0,
      "step": 6230
    },
    {
      "entropy": 5.61024580001831,
      "epoch": 0.6187971417229059,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004969499269515878,
      "loss": 5.3969,
      "mean_token_accuracy": 0.15812541991472245,
      "num_tokens": 13696181.0,
      "step": 6235
    },
    {
      "entropy": 5.652813959121704,
      "epoch": 0.6192933703850735,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004969441099787663,
      "loss": 5.5008,
      "mean_token_accuracy": 0.14699973165988922,
      "num_tokens": 13706719.0,
      "step": 6240
    },
    {
      "entropy": 5.647890329360962,
      "epoch": 0.619789599047241,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000496938287502209,
      "loss": 5.452,
      "mean_token_accuracy": 0.15290555059909822,
      "num_tokens": 13716653.0,
      "step": 6245
    },
    {
      "entropy": 5.525570392608643,
      "epoch": 0.6202858277094085,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004969324595220605,
      "loss": 5.3681,
      "mean_token_accuracy": 0.15710528790950776,
      "num_tokens": 13728220.0,
      "step": 6250
    },
    {
      "entropy": 5.50266752243042,
      "epoch": 0.620782056371576,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004969266260384652,
      "loss": 5.332,
      "mean_token_accuracy": 0.15505106896162033,
      "num_tokens": 13738103.0,
      "step": 6255
    },
    {
      "entropy": 5.5926953792572025,
      "epoch": 0.6212782850337435,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004969207870515678,
      "loss": 5.422,
      "mean_token_accuracy": 0.15330670326948165,
      "num_tokens": 13749845.0,
      "step": 6260
    },
    {
      "entropy": 5.616781616210938,
      "epoch": 0.6217745136959111,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000496914942561513,
      "loss": 5.4408,
      "mean_token_accuracy": 0.15307224541902542,
      "num_tokens": 13759983.0,
      "step": 6265
    },
    {
      "entropy": 5.643848037719726,
      "epoch": 0.6222707423580786,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004969090925684459,
      "loss": 5.5296,
      "mean_token_accuracy": 0.1525065764784813,
      "num_tokens": 13770045.0,
      "step": 6270
    },
    {
      "entropy": 5.51793851852417,
      "epoch": 0.6227669710202461,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004969032370725114,
      "loss": 5.3821,
      "mean_token_accuracy": 0.15532010644674302,
      "num_tokens": 13779823.0,
      "step": 6275
    },
    {
      "entropy": 5.583812856674195,
      "epoch": 0.6232631996824136,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004968973760738548,
      "loss": 5.3897,
      "mean_token_accuracy": 0.1501995801925659,
      "num_tokens": 13791165.0,
      "step": 6280
    },
    {
      "entropy": 5.661803197860718,
      "epoch": 0.6237594283445812,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004968915095726216,
      "loss": 5.4912,
      "mean_token_accuracy": 0.14553115367889405,
      "num_tokens": 13802872.0,
      "step": 6285
    },
    {
      "entropy": 5.591525983810425,
      "epoch": 0.6242556570067487,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004968856375689568,
      "loss": 5.4198,
      "mean_token_accuracy": 0.1514755442738533,
      "num_tokens": 13813893.0,
      "step": 6290
    },
    {
      "entropy": 5.59356369972229,
      "epoch": 0.6247518856689163,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004968797600630065,
      "loss": 5.4917,
      "mean_token_accuracy": 0.15136266350746155,
      "num_tokens": 13823613.0,
      "step": 6295
    },
    {
      "entropy": 5.6368231773376465,
      "epoch": 0.6252481143310837,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004968738770549162,
      "loss": 5.4971,
      "mean_token_accuracy": 0.1554287150502205,
      "num_tokens": 13835413.0,
      "step": 6300
    },
    {
      "entropy": 5.605167531967163,
      "epoch": 0.6257443429932513,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004968679885448318,
      "loss": 5.4137,
      "mean_token_accuracy": 0.14931111931800842,
      "num_tokens": 13845806.0,
      "step": 6305
    },
    {
      "entropy": 5.620719861984253,
      "epoch": 0.6262405716554188,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004968620945328995,
      "loss": 5.5048,
      "mean_token_accuracy": 0.15276079922914504,
      "num_tokens": 13856540.0,
      "step": 6310
    },
    {
      "entropy": 5.61713719367981,
      "epoch": 0.6267368003175864,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004968561950192652,
      "loss": 5.3625,
      "mean_token_accuracy": 0.15804028809070586,
      "num_tokens": 13867609.0,
      "step": 6315
    },
    {
      "entropy": 5.62848858833313,
      "epoch": 0.6272330289797539,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004968502900040753,
      "loss": 5.4364,
      "mean_token_accuracy": 0.14521164894104005,
      "num_tokens": 13878299.0,
      "step": 6320
    },
    {
      "entropy": 5.604143857955933,
      "epoch": 0.6277292576419214,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004968443794874764,
      "loss": 5.4484,
      "mean_token_accuracy": 0.15260792821645736,
      "num_tokens": 13889625.0,
      "step": 6325
    },
    {
      "entropy": 5.5830577373504635,
      "epoch": 0.6282254863040889,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004968384634696149,
      "loss": 5.4168,
      "mean_token_accuracy": 0.14463182538747787,
      "num_tokens": 13899680.0,
      "step": 6330
    },
    {
      "entropy": 5.550440120697021,
      "epoch": 0.6287217149662565,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004968325419506375,
      "loss": 5.391,
      "mean_token_accuracy": 0.15744505524635316,
      "num_tokens": 13910883.0,
      "step": 6335
    },
    {
      "entropy": 5.549885511398315,
      "epoch": 0.629217943628424,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004968266149306909,
      "loss": 5.4029,
      "mean_token_accuracy": 0.15215905159711837,
      "num_tokens": 13921442.0,
      "step": 6340
    },
    {
      "entropy": 5.583293914794922,
      "epoch": 0.6297141722905915,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004968206824099224,
      "loss": 5.4343,
      "mean_token_accuracy": 0.15862448066473006,
      "num_tokens": 13930309.0,
      "step": 6345
    },
    {
      "entropy": 5.611448574066162,
      "epoch": 0.630210400952759,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004968147443884789,
      "loss": 5.4261,
      "mean_token_accuracy": 0.15601219236850739,
      "num_tokens": 13941346.0,
      "step": 6350
    },
    {
      "entropy": 5.586319446563721,
      "epoch": 0.6307066296149265,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004968088008665078,
      "loss": 5.396,
      "mean_token_accuracy": 0.1572890430688858,
      "num_tokens": 13951933.0,
      "step": 6355
    },
    {
      "entropy": 5.540216064453125,
      "epoch": 0.6312028582770941,
      "grad_norm": 0.875,
      "learning_rate": 0.0004968028518441563,
      "loss": 5.4616,
      "mean_token_accuracy": 0.1499076098203659,
      "num_tokens": 13963112.0,
      "step": 6360
    },
    {
      "entropy": 5.652147722244263,
      "epoch": 0.6316990869392616,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004967968973215721,
      "loss": 5.4874,
      "mean_token_accuracy": 0.15574448704719543,
      "num_tokens": 13974218.0,
      "step": 6365
    },
    {
      "entropy": 5.657738161087036,
      "epoch": 0.6321953156014292,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004967909372989027,
      "loss": 5.4141,
      "mean_token_accuracy": 0.15234902650117874,
      "num_tokens": 13984841.0,
      "step": 6370
    },
    {
      "entropy": 5.5751903533935545,
      "epoch": 0.6326915442635966,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004967849717762959,
      "loss": 5.4299,
      "mean_token_accuracy": 0.15411732643842696,
      "num_tokens": 13997504.0,
      "step": 6375
    },
    {
      "entropy": 5.528285264968872,
      "epoch": 0.6331877729257642,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004967790007538999,
      "loss": 5.3672,
      "mean_token_accuracy": 0.15952090322971343,
      "num_tokens": 14009244.0,
      "step": 6380
    },
    {
      "entropy": 5.540134716033935,
      "epoch": 0.6336840015879317,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004967730242318624,
      "loss": 5.4008,
      "mean_token_accuracy": 0.15605138689279557,
      "num_tokens": 14021190.0,
      "step": 6385
    },
    {
      "entropy": 5.561956739425659,
      "epoch": 0.6341802302500993,
      "grad_norm": 0.890625,
      "learning_rate": 0.000496767042210332,
      "loss": 5.3639,
      "mean_token_accuracy": 0.156063561886549,
      "num_tokens": 14032191.0,
      "step": 6390
    },
    {
      "entropy": 5.500299978256225,
      "epoch": 0.6346764589122668,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004967610546894566,
      "loss": 5.3296,
      "mean_token_accuracy": 0.16037983745336531,
      "num_tokens": 14042817.0,
      "step": 6395
    },
    {
      "entropy": 5.554322195053101,
      "epoch": 0.6351726875744343,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000496755061669385,
      "loss": 5.3512,
      "mean_token_accuracy": 0.15881706178188323,
      "num_tokens": 14052725.0,
      "step": 6400
    },
    {
      "entropy": 5.610175895690918,
      "epoch": 0.6356689162366018,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004967490631502656,
      "loss": 5.4268,
      "mean_token_accuracy": 0.14564218297600745,
      "num_tokens": 14063621.0,
      "step": 6405
    },
    {
      "entropy": 5.705971002578735,
      "epoch": 0.6361651448987694,
      "grad_norm": 0.875,
      "learning_rate": 0.0004967430591322474,
      "loss": 5.5563,
      "mean_token_accuracy": 0.147991943359375,
      "num_tokens": 14074350.0,
      "step": 6410
    },
    {
      "entropy": 5.67575511932373,
      "epoch": 0.6366613735609369,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004967370496154791,
      "loss": 5.486,
      "mean_token_accuracy": 0.14895145893096923,
      "num_tokens": 14084600.0,
      "step": 6415
    },
    {
      "entropy": 5.625611925125122,
      "epoch": 0.6371576022231044,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004967310346001098,
      "loss": 5.4084,
      "mean_token_accuracy": 0.15440958440303804,
      "num_tokens": 14094739.0,
      "step": 6420
    },
    {
      "entropy": 5.702681303024292,
      "epoch": 0.6376538308852719,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004967250140862885,
      "loss": 5.5624,
      "mean_token_accuracy": 0.14849390238523483,
      "num_tokens": 14106020.0,
      "step": 6425
    },
    {
      "entropy": 5.679549694061279,
      "epoch": 0.6381500595474394,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004967189880741647,
      "loss": 5.5236,
      "mean_token_accuracy": 0.14355285838246346,
      "num_tokens": 14117699.0,
      "step": 6430
    },
    {
      "entropy": 5.57834038734436,
      "epoch": 0.638646288209607,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004967129565638878,
      "loss": 5.4054,
      "mean_token_accuracy": 0.15328127294778823,
      "num_tokens": 14128901.0,
      "step": 6435
    },
    {
      "entropy": 5.588972377777099,
      "epoch": 0.6391425168717745,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004967069195556073,
      "loss": 5.4547,
      "mean_token_accuracy": 0.14977715387940407,
      "num_tokens": 14140832.0,
      "step": 6440
    },
    {
      "entropy": 5.637591886520386,
      "epoch": 0.6396387455339421,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000496700877049473,
      "loss": 5.4916,
      "mean_token_accuracy": 0.15187741667032242,
      "num_tokens": 14151936.0,
      "step": 6445
    },
    {
      "entropy": 5.651413059234619,
      "epoch": 0.6401349741961095,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004966948290456345,
      "loss": 5.4796,
      "mean_token_accuracy": 0.15306032747030257,
      "num_tokens": 14160945.0,
      "step": 6450
    },
    {
      "entropy": 5.560344457626343,
      "epoch": 0.6406312028582771,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004966887755442421,
      "loss": 5.3642,
      "mean_token_accuracy": 0.15178609192371367,
      "num_tokens": 14172052.0,
      "step": 6455
    },
    {
      "entropy": 5.631370782852173,
      "epoch": 0.6411274315204446,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004966827165454456,
      "loss": 5.4448,
      "mean_token_accuracy": 0.1570268116891384,
      "num_tokens": 14181788.0,
      "step": 6460
    },
    {
      "entropy": 5.668099117279053,
      "epoch": 0.6416236601826122,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004966766520493956,
      "loss": 5.5322,
      "mean_token_accuracy": 0.14923309087753295,
      "num_tokens": 14193668.0,
      "step": 6465
    },
    {
      "entropy": 5.659458875656128,
      "epoch": 0.6421198888447797,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004966705820562422,
      "loss": 5.4092,
      "mean_token_accuracy": 0.15575137436389924,
      "num_tokens": 14203665.0,
      "step": 6470
    },
    {
      "entropy": 5.593989849090576,
      "epoch": 0.6426161175069472,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004966645065661361,
      "loss": 5.4438,
      "mean_token_accuracy": 0.1480250433087349,
      "num_tokens": 14214213.0,
      "step": 6475
    },
    {
      "entropy": 5.6232795238494875,
      "epoch": 0.6431123461691147,
      "grad_norm": 1.0,
      "learning_rate": 0.0004966584255792279,
      "loss": 5.5451,
      "mean_token_accuracy": 0.1423339731991291,
      "num_tokens": 14225754.0,
      "step": 6480
    },
    {
      "entropy": 5.638065958023072,
      "epoch": 0.6436085748312823,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004966523390956683,
      "loss": 5.4231,
      "mean_token_accuracy": 0.15263875126838683,
      "num_tokens": 14236508.0,
      "step": 6485
    },
    {
      "entropy": 5.632704401016236,
      "epoch": 0.6441048034934498,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004966462471156083,
      "loss": 5.4648,
      "mean_token_accuracy": 0.1550041601061821,
      "num_tokens": 14247719.0,
      "step": 6490
    },
    {
      "entropy": 5.61082239151001,
      "epoch": 0.6446010321556173,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004966401496391989,
      "loss": 5.5119,
      "mean_token_accuracy": 0.1417395904660225,
      "num_tokens": 14258667.0,
      "step": 6495
    },
    {
      "entropy": 5.609544277191162,
      "epoch": 0.6450972608177848,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004966340466665915,
      "loss": 5.5156,
      "mean_token_accuracy": 0.14988971799612044,
      "num_tokens": 14269666.0,
      "step": 6500
    },
    {
      "entropy": 5.559063959121704,
      "epoch": 0.6455934894799523,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004966279381979373,
      "loss": 5.4094,
      "mean_token_accuracy": 0.15216174125671386,
      "num_tokens": 14280862.0,
      "step": 6505
    },
    {
      "entropy": 5.688417387008667,
      "epoch": 0.6460897181421199,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004966218242333879,
      "loss": 5.5598,
      "mean_token_accuracy": 0.13700098246335984,
      "num_tokens": 14292505.0,
      "step": 6510
    },
    {
      "entropy": 5.637891864776611,
      "epoch": 0.6465859468042874,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004966157047730945,
      "loss": 5.3477,
      "mean_token_accuracy": 0.15833587795495987,
      "num_tokens": 14302547.0,
      "step": 6515
    },
    {
      "entropy": 5.537370729446411,
      "epoch": 0.647082175466455,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004966095798172095,
      "loss": 5.4881,
      "mean_token_accuracy": 0.1493873566389084,
      "num_tokens": 14313585.0,
      "step": 6520
    },
    {
      "entropy": 5.529198169708252,
      "epoch": 0.6475784041286224,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004966034493658843,
      "loss": 5.3499,
      "mean_token_accuracy": 0.15040139853954315,
      "num_tokens": 14323665.0,
      "step": 6525
    },
    {
      "entropy": 5.586475276947022,
      "epoch": 0.64807463279079,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0004965973134192711,
      "loss": 5.3843,
      "mean_token_accuracy": 0.1553342416882515,
      "num_tokens": 14334257.0,
      "step": 6530
    },
    {
      "entropy": 5.642743921279907,
      "epoch": 0.6485708614529575,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000496591171977522,
      "loss": 5.4145,
      "mean_token_accuracy": 0.1533318355679512,
      "num_tokens": 14345350.0,
      "step": 6535
    },
    {
      "entropy": 5.611017465591431,
      "epoch": 0.6490670901151251,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004965850250407893,
      "loss": 5.4279,
      "mean_token_accuracy": 0.152386437356472,
      "num_tokens": 14355593.0,
      "step": 6540
    },
    {
      "entropy": 5.564835166931152,
      "epoch": 0.6495633187772926,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004965788726092256,
      "loss": 5.3091,
      "mean_token_accuracy": 0.15809425115585327,
      "num_tokens": 14366405.0,
      "step": 6545
    },
    {
      "entropy": 5.564842176437378,
      "epoch": 0.6500595474394602,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004965727146829832,
      "loss": 5.4131,
      "mean_token_accuracy": 0.1491552025079727,
      "num_tokens": 14377228.0,
      "step": 6550
    },
    {
      "entropy": 5.696504354476929,
      "epoch": 0.6505557761016276,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004965665512622148,
      "loss": 5.5438,
      "mean_token_accuracy": 0.1498422294855118,
      "num_tokens": 14387113.0,
      "step": 6555
    },
    {
      "entropy": 5.623318910598755,
      "epoch": 0.6510520047637952,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004965603823470735,
      "loss": 5.41,
      "mean_token_accuracy": 0.15291822105646133,
      "num_tokens": 14397980.0,
      "step": 6560
    },
    {
      "entropy": 5.563955879211425,
      "epoch": 0.6515482334259627,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004965542079377122,
      "loss": 5.3477,
      "mean_token_accuracy": 0.15694964528083802,
      "num_tokens": 14409063.0,
      "step": 6565
    },
    {
      "entropy": 5.52393217086792,
      "epoch": 0.6520444620881302,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004965480280342839,
      "loss": 5.325,
      "mean_token_accuracy": 0.1627918154001236,
      "num_tokens": 14421056.0,
      "step": 6570
    },
    {
      "entropy": 5.563065719604492,
      "epoch": 0.6525406907502977,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004965418426369419,
      "loss": 5.3247,
      "mean_token_accuracy": 0.16055609956383704,
      "num_tokens": 14430778.0,
      "step": 6575
    },
    {
      "entropy": 5.547006320953369,
      "epoch": 0.6530369194124652,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004965356517458395,
      "loss": 5.3391,
      "mean_token_accuracy": 0.15274855643510818,
      "num_tokens": 14441544.0,
      "step": 6580
    },
    {
      "entropy": 5.579457855224609,
      "epoch": 0.6535331480746328,
      "grad_norm": 0.875,
      "learning_rate": 0.0004965294553611303,
      "loss": 5.4925,
      "mean_token_accuracy": 0.15214899480342864,
      "num_tokens": 14453062.0,
      "step": 6585
    },
    {
      "entropy": 5.602321434020996,
      "epoch": 0.6540293767368003,
      "grad_norm": 0.953125,
      "learning_rate": 0.000496523253482968,
      "loss": 5.4204,
      "mean_token_accuracy": 0.15419695898890495,
      "num_tokens": 14464402.0,
      "step": 6590
    },
    {
      "entropy": 5.635910081863403,
      "epoch": 0.6545256053989679,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004965170461115065,
      "loss": 5.5059,
      "mean_token_accuracy": 0.15288595706224442,
      "num_tokens": 14476595.0,
      "step": 6595
    },
    {
      "entropy": 5.700248050689697,
      "epoch": 0.6550218340611353,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004965108332468994,
      "loss": 5.4883,
      "mean_token_accuracy": 0.14744484722614287,
      "num_tokens": 14487298.0,
      "step": 6600
    },
    {
      "entropy": 5.587152910232544,
      "epoch": 0.6555180627233029,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004965046148893011,
      "loss": 5.3789,
      "mean_token_accuracy": 0.16049299538135528,
      "num_tokens": 14499708.0,
      "step": 6605
    },
    {
      "entropy": 5.576786851882934,
      "epoch": 0.6560142913854704,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004964983910388654,
      "loss": 5.4411,
      "mean_token_accuracy": 0.15392221957445146,
      "num_tokens": 14511773.0,
      "step": 6610
    },
    {
      "entropy": 5.547587585449219,
      "epoch": 0.656510520047638,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004964921616957471,
      "loss": 5.3444,
      "mean_token_accuracy": 0.15164763033390044,
      "num_tokens": 14523024.0,
      "step": 6615
    },
    {
      "entropy": 5.617991828918457,
      "epoch": 0.6570067487098055,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004964859268601005,
      "loss": 5.4126,
      "mean_token_accuracy": 0.154441487044096,
      "num_tokens": 14532915.0,
      "step": 6620
    },
    {
      "entropy": 5.598261404037475,
      "epoch": 0.657502977371973,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004964796865320801,
      "loss": 5.4043,
      "mean_token_accuracy": 0.15950134098529817,
      "num_tokens": 14543627.0,
      "step": 6625
    },
    {
      "entropy": 5.525602531433106,
      "epoch": 0.6579992060341405,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004964734407118406,
      "loss": 5.3893,
      "mean_token_accuracy": 0.15673066079616546,
      "num_tokens": 14555181.0,
      "step": 6630
    },
    {
      "entropy": 5.539927768707275,
      "epoch": 0.6584954346963081,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004964671893995372,
      "loss": 5.3771,
      "mean_token_accuracy": 0.1498602196574211,
      "num_tokens": 14565542.0,
      "step": 6635
    },
    {
      "entropy": 5.630816411972046,
      "epoch": 0.6589916633584756,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004964609325953245,
      "loss": 5.4384,
      "mean_token_accuracy": 0.1532747134566307,
      "num_tokens": 14575395.0,
      "step": 6640
    },
    {
      "entropy": 5.629387712478637,
      "epoch": 0.6594878920206431,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004964546702993581,
      "loss": 5.4587,
      "mean_token_accuracy": 0.15217862278223038,
      "num_tokens": 14587206.0,
      "step": 6645
    },
    {
      "entropy": 5.650357103347778,
      "epoch": 0.6599841206828106,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000496448402511793,
      "loss": 5.4758,
      "mean_token_accuracy": 0.15155555158853531,
      "num_tokens": 14599048.0,
      "step": 6650
    },
    {
      "entropy": 5.654848718643189,
      "epoch": 0.6604803493449781,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004964421292327847,
      "loss": 5.4365,
      "mean_token_accuracy": 0.15383079051971435,
      "num_tokens": 14609633.0,
      "step": 6655
    },
    {
      "entropy": 5.5945721626281735,
      "epoch": 0.6609765780071457,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004964358504624887,
      "loss": 5.346,
      "mean_token_accuracy": 0.15625559240579606,
      "num_tokens": 14620373.0,
      "step": 6660
    },
    {
      "entropy": 5.537472248077393,
      "epoch": 0.6614728066693132,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004964295662010608,
      "loss": 5.2967,
      "mean_token_accuracy": 0.16170741766691207,
      "num_tokens": 14630463.0,
      "step": 6665
    },
    {
      "entropy": 5.740504264831543,
      "epoch": 0.6619690353314808,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004964232764486568,
      "loss": 5.5402,
      "mean_token_accuracy": 0.1473265089094639,
      "num_tokens": 14641523.0,
      "step": 6670
    },
    {
      "entropy": 5.589998054504394,
      "epoch": 0.6624652639936482,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004964169812054326,
      "loss": 5.3418,
      "mean_token_accuracy": 0.15613186061382295,
      "num_tokens": 14651812.0,
      "step": 6675
    },
    {
      "entropy": 5.514553642272949,
      "epoch": 0.6629614926558158,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004964106804715444,
      "loss": 5.3556,
      "mean_token_accuracy": 0.15830594450235366,
      "num_tokens": 14661982.0,
      "step": 6680
    },
    {
      "entropy": 5.569979190826416,
      "epoch": 0.6634577213179833,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004964043742471486,
      "loss": 5.3686,
      "mean_token_accuracy": 0.162179996073246,
      "num_tokens": 14671981.0,
      "step": 6685
    },
    {
      "entropy": 5.622475957870483,
      "epoch": 0.6639539499801509,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004963980625324012,
      "loss": 5.3914,
      "mean_token_accuracy": 0.1518598973751068,
      "num_tokens": 14682478.0,
      "step": 6690
    },
    {
      "entropy": 5.633850479125977,
      "epoch": 0.6644501786423184,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000496391745327459,
      "loss": 5.479,
      "mean_token_accuracy": 0.14283154159784317,
      "num_tokens": 14695197.0,
      "step": 6695
    },
    {
      "entropy": 5.586344242095947,
      "epoch": 0.664946407304486,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004963854226324785,
      "loss": 5.3718,
      "mean_token_accuracy": 0.15447912067174913,
      "num_tokens": 14705702.0,
      "step": 6700
    },
    {
      "entropy": 5.602133846282959,
      "epoch": 0.6654426359666534,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004963790944476167,
      "loss": 5.393,
      "mean_token_accuracy": 0.1590353712439537,
      "num_tokens": 14715370.0,
      "step": 6705
    },
    {
      "entropy": 5.636506462097168,
      "epoch": 0.665938864628821,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004963727607730303,
      "loss": 5.4095,
      "mean_token_accuracy": 0.15574731081724166,
      "num_tokens": 14725481.0,
      "step": 6710
    },
    {
      "entropy": 5.5663489818573,
      "epoch": 0.6664350932909885,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004963664216088764,
      "loss": 5.3308,
      "mean_token_accuracy": 0.15330153852701187,
      "num_tokens": 14735250.0,
      "step": 6715
    },
    {
      "entropy": 5.550839567184449,
      "epoch": 0.666931321953156,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004963600769553124,
      "loss": 5.3423,
      "mean_token_accuracy": 0.16309174597263337,
      "num_tokens": 14744741.0,
      "step": 6720
    },
    {
      "entropy": 5.648164129257202,
      "epoch": 0.6674275506153236,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004963537268124953,
      "loss": 5.3806,
      "mean_token_accuracy": 0.15598916113376618,
      "num_tokens": 14755865.0,
      "step": 6725
    },
    {
      "entropy": 5.5299581527709964,
      "epoch": 0.667923779277491,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004963473711805829,
      "loss": 5.3308,
      "mean_token_accuracy": 0.1584343433380127,
      "num_tokens": 14766154.0,
      "step": 6730
    },
    {
      "entropy": 5.5833440780639645,
      "epoch": 0.6684200079396586,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004963410100597326,
      "loss": 5.388,
      "mean_token_accuracy": 0.1521584615111351,
      "num_tokens": 14776940.0,
      "step": 6735
    },
    {
      "entropy": 5.560282754898071,
      "epoch": 0.6689162366018261,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004963346434501021,
      "loss": 5.4813,
      "mean_token_accuracy": 0.15100081786513328,
      "num_tokens": 14788367.0,
      "step": 6740
    },
    {
      "entropy": 5.56590051651001,
      "epoch": 0.6694124652639937,
      "grad_norm": 0.875,
      "learning_rate": 0.0004963282713518495,
      "loss": 5.3901,
      "mean_token_accuracy": 0.15709468275308608,
      "num_tokens": 14798751.0,
      "step": 6745
    },
    {
      "entropy": 5.539422416687012,
      "epoch": 0.6699086939261611,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004963218937651326,
      "loss": 5.323,
      "mean_token_accuracy": 0.15921037793159484,
      "num_tokens": 14808748.0,
      "step": 6750
    },
    {
      "entropy": 5.621778106689453,
      "epoch": 0.6704049225883287,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004963155106901098,
      "loss": 5.4996,
      "mean_token_accuracy": 0.1455796867609024,
      "num_tokens": 14820524.0,
      "step": 6755
    },
    {
      "entropy": 5.7212916851043705,
      "epoch": 0.6709011512504962,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004963091221269391,
      "loss": 5.5437,
      "mean_token_accuracy": 0.14160653054714203,
      "num_tokens": 14832485.0,
      "step": 6760
    },
    {
      "entropy": 5.581074953079224,
      "epoch": 0.6713973799126638,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000496302728075779,
      "loss": 5.3353,
      "mean_token_accuracy": 0.1552274703979492,
      "num_tokens": 14843951.0,
      "step": 6765
    },
    {
      "entropy": 5.529450416564941,
      "epoch": 0.6718936085748313,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004962963285367882,
      "loss": 5.4265,
      "mean_token_accuracy": 0.1550913542509079,
      "num_tokens": 14855502.0,
      "step": 6770
    },
    {
      "entropy": 5.558300542831421,
      "epoch": 0.6723898372369989,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004962899235101253,
      "loss": 5.338,
      "mean_token_accuracy": 0.15669150203466414,
      "num_tokens": 14866305.0,
      "step": 6775
    },
    {
      "entropy": 5.592230749130249,
      "epoch": 0.6728860658991663,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004962835129959491,
      "loss": 5.3719,
      "mean_token_accuracy": 0.15086598098278045,
      "num_tokens": 14877222.0,
      "step": 6780
    },
    {
      "entropy": 5.55699896812439,
      "epoch": 0.6733822945613339,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004962770969944186,
      "loss": 5.433,
      "mean_token_accuracy": 0.15394371300935744,
      "num_tokens": 14888099.0,
      "step": 6785
    },
    {
      "entropy": 5.572944831848145,
      "epoch": 0.6738785232235014,
      "grad_norm": 0.9375,
      "learning_rate": 0.000496270675505693,
      "loss": 5.3696,
      "mean_token_accuracy": 0.15748813152313232,
      "num_tokens": 14899042.0,
      "step": 6790
    },
    {
      "entropy": 5.584994602203369,
      "epoch": 0.6743747518856689,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004962642485299314,
      "loss": 5.4143,
      "mean_token_accuracy": 0.15487085282802582,
      "num_tokens": 14908837.0,
      "step": 6795
    },
    {
      "entropy": 5.543886661529541,
      "epoch": 0.6748709805478365,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004962578160672931,
      "loss": 5.3594,
      "mean_token_accuracy": 0.16115720719099044,
      "num_tokens": 14920773.0,
      "step": 6800
    },
    {
      "entropy": 5.5483417987823485,
      "epoch": 0.6753672092100039,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004962513781179379,
      "loss": 5.3534,
      "mean_token_accuracy": 0.15471526086330414,
      "num_tokens": 14930970.0,
      "step": 6805
    },
    {
      "entropy": 5.565474987030029,
      "epoch": 0.6758634378721715,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004962449346820251,
      "loss": 5.4262,
      "mean_token_accuracy": 0.1476544000208378,
      "num_tokens": 14942378.0,
      "step": 6810
    },
    {
      "entropy": 5.558213567733764,
      "epoch": 0.676359666534339,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004962384857597148,
      "loss": 5.4566,
      "mean_token_accuracy": 0.15085810124874116,
      "num_tokens": 14953020.0,
      "step": 6815
    },
    {
      "entropy": 5.653578233718872,
      "epoch": 0.6768558951965066,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004962320313511668,
      "loss": 5.5269,
      "mean_token_accuracy": 0.15361799597740172,
      "num_tokens": 14962340.0,
      "step": 6820
    },
    {
      "entropy": 5.625601720809937,
      "epoch": 0.677352123858674,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004962255714565409,
      "loss": 5.4141,
      "mean_token_accuracy": 0.15450359284877777,
      "num_tokens": 14972792.0,
      "step": 6825
    },
    {
      "entropy": 5.571131658554077,
      "epoch": 0.6778483525208416,
      "grad_norm": 0.875,
      "learning_rate": 0.0004962191060759978,
      "loss": 5.3912,
      "mean_token_accuracy": 0.15612896382808686,
      "num_tokens": 14983698.0,
      "step": 6830
    },
    {
      "entropy": 5.613586330413819,
      "epoch": 0.6783445811830091,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004962126352096973,
      "loss": 5.3524,
      "mean_token_accuracy": 0.15863122791051865,
      "num_tokens": 14994583.0,
      "step": 6835
    },
    {
      "entropy": 5.51322569847107,
      "epoch": 0.6788408098451767,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004962061588578004,
      "loss": 5.2983,
      "mean_token_accuracy": 0.15476125478744507,
      "num_tokens": 15005325.0,
      "step": 6840
    },
    {
      "entropy": 5.588961601257324,
      "epoch": 0.6793370385073442,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004961996770204673,
      "loss": 5.3704,
      "mean_token_accuracy": 0.1552986204624176,
      "num_tokens": 15015155.0,
      "step": 6845
    },
    {
      "entropy": 5.578842353820801,
      "epoch": 0.6798332671695118,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004961931896978588,
      "loss": 5.4308,
      "mean_token_accuracy": 0.15067857205867768,
      "num_tokens": 15026529.0,
      "step": 6850
    },
    {
      "entropy": 5.50896635055542,
      "epoch": 0.6803294958316792,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004961866968901359,
      "loss": 5.3618,
      "mean_token_accuracy": 0.15841467380523683,
      "num_tokens": 15037632.0,
      "step": 6855
    },
    {
      "entropy": 5.562885856628418,
      "epoch": 0.6808257244938468,
      "grad_norm": 0.875,
      "learning_rate": 0.0004961801985974595,
      "loss": 5.4317,
      "mean_token_accuracy": 0.15281036049127578,
      "num_tokens": 15050397.0,
      "step": 6860
    },
    {
      "entropy": 5.555966424942016,
      "epoch": 0.6813219531560143,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004961736948199907,
      "loss": 5.3491,
      "mean_token_accuracy": 0.15506883710622787,
      "num_tokens": 15061505.0,
      "step": 6865
    },
    {
      "entropy": 5.660563802719116,
      "epoch": 0.6818181818181818,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004961671855578909,
      "loss": 5.5388,
      "mean_token_accuracy": 0.1459961101412773,
      "num_tokens": 15073752.0,
      "step": 6870
    },
    {
      "entropy": 5.6258409976959225,
      "epoch": 0.6823144104803494,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004961606708113216,
      "loss": 5.4266,
      "mean_token_accuracy": 0.15296203792095184,
      "num_tokens": 15084730.0,
      "step": 6875
    },
    {
      "entropy": 5.583535385131836,
      "epoch": 0.6828106391425168,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004961541505804441,
      "loss": 5.3401,
      "mean_token_accuracy": 0.15618596374988555,
      "num_tokens": 15095811.0,
      "step": 6880
    },
    {
      "entropy": 5.579980182647705,
      "epoch": 0.6833068678046844,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004961476248654202,
      "loss": 5.3299,
      "mean_token_accuracy": 0.16105467826128006,
      "num_tokens": 15106102.0,
      "step": 6885
    },
    {
      "entropy": 5.59256911277771,
      "epoch": 0.6838030964668519,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004961410936664118,
      "loss": 5.4246,
      "mean_token_accuracy": 0.15121853053569795,
      "num_tokens": 15115907.0,
      "step": 6890
    },
    {
      "entropy": 5.580052757263184,
      "epoch": 0.6842993251290195,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004961345569835807,
      "loss": 5.3971,
      "mean_token_accuracy": 0.1558905601501465,
      "num_tokens": 15126788.0,
      "step": 6895
    },
    {
      "entropy": 5.584158658981323,
      "epoch": 0.684795553791187,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004961280148170892,
      "loss": 5.3621,
      "mean_token_accuracy": 0.15638109892606736,
      "num_tokens": 15137814.0,
      "step": 6900
    },
    {
      "entropy": 5.631082773208618,
      "epoch": 0.6852917824533545,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004961214671670994,
      "loss": 5.5891,
      "mean_token_accuracy": 0.1433118350803852,
      "num_tokens": 15148597.0,
      "step": 6905
    },
    {
      "entropy": 5.630214738845825,
      "epoch": 0.685788011115522,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004961149140337737,
      "loss": 5.3957,
      "mean_token_accuracy": 0.1523701250553131,
      "num_tokens": 15158985.0,
      "step": 6910
    },
    {
      "entropy": 5.566757869720459,
      "epoch": 0.6862842397776896,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004961083554172745,
      "loss": 5.3784,
      "mean_token_accuracy": 0.15582308694720268,
      "num_tokens": 15170242.0,
      "step": 6915
    },
    {
      "entropy": 5.5762779712677,
      "epoch": 0.6867804684398571,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004961017913177646,
      "loss": 5.3514,
      "mean_token_accuracy": 0.1617621198296547,
      "num_tokens": 15180353.0,
      "step": 6920
    },
    {
      "entropy": 5.5126060962677,
      "epoch": 0.6872766971020247,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004960952217354066,
      "loss": 5.3043,
      "mean_token_accuracy": 0.16224115192890168,
      "num_tokens": 15191955.0,
      "step": 6925
    },
    {
      "entropy": 5.535926866531372,
      "epoch": 0.6877729257641921,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004960886466703636,
      "loss": 5.3753,
      "mean_token_accuracy": 0.15272082388401031,
      "num_tokens": 15202088.0,
      "step": 6930
    },
    {
      "entropy": 5.515703868865967,
      "epoch": 0.6882691544263597,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004960820661227987,
      "loss": 5.3163,
      "mean_token_accuracy": 0.15685615092515945,
      "num_tokens": 15212291.0,
      "step": 6935
    },
    {
      "entropy": 5.555116033554077,
      "epoch": 0.6887653830885272,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004960754800928748,
      "loss": 5.3691,
      "mean_token_accuracy": 0.16065500527620316,
      "num_tokens": 15223315.0,
      "step": 6940
    },
    {
      "entropy": 5.656789302825928,
      "epoch": 0.6892616117506947,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004960688885807553,
      "loss": 5.4829,
      "mean_token_accuracy": 0.15075226426124572,
      "num_tokens": 15233416.0,
      "step": 6945
    },
    {
      "entropy": 5.625695610046387,
      "epoch": 0.6897578404128623,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004960622915866038,
      "loss": 5.3164,
      "mean_token_accuracy": 0.16170334964990615,
      "num_tokens": 15242951.0,
      "step": 6950
    },
    {
      "entropy": 5.542146015167236,
      "epoch": 0.6902540690750297,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004960556891105839,
      "loss": 5.3995,
      "mean_token_accuracy": 0.15278018414974212,
      "num_tokens": 15253145.0,
      "step": 6955
    },
    {
      "entropy": 5.546169090270996,
      "epoch": 0.6907502977371973,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004960490811528591,
      "loss": 5.3452,
      "mean_token_accuracy": 0.16333898156881332,
      "num_tokens": 15263712.0,
      "step": 6960
    },
    {
      "entropy": 5.671264123916626,
      "epoch": 0.6912465263993648,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004960424677135935,
      "loss": 5.4732,
      "mean_token_accuracy": 0.15577918738126756,
      "num_tokens": 15273609.0,
      "step": 6965
    },
    {
      "entropy": 5.5488487720489506,
      "epoch": 0.6917427550615324,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004960358487929511,
      "loss": 5.3341,
      "mean_token_accuracy": 0.16462758779525757,
      "num_tokens": 15283934.0,
      "step": 6970
    },
    {
      "entropy": 5.548047065734863,
      "epoch": 0.6922389837236999,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004960292243910958,
      "loss": 5.3984,
      "mean_token_accuracy": 0.15148687809705735,
      "num_tokens": 15294394.0,
      "step": 6975
    },
    {
      "entropy": 5.582794809341431,
      "epoch": 0.6927352123858674,
      "grad_norm": 0.921875,
      "learning_rate": 0.000496022594508192,
      "loss": 5.3622,
      "mean_token_accuracy": 0.16121286451816558,
      "num_tokens": 15304622.0,
      "step": 6980
    },
    {
      "entropy": 5.567217922210693,
      "epoch": 0.6932314410480349,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004960159591444041,
      "loss": 5.2711,
      "mean_token_accuracy": 0.1600771963596344,
      "num_tokens": 15314094.0,
      "step": 6985
    },
    {
      "entropy": 5.557595539093017,
      "epoch": 0.6937276697102025,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004960093182998968,
      "loss": 5.3916,
      "mean_token_accuracy": 0.16023263037204744,
      "num_tokens": 15324313.0,
      "step": 6990
    },
    {
      "entropy": 5.487686729431152,
      "epoch": 0.69422389837237,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004960026719748346,
      "loss": 5.2706,
      "mean_token_accuracy": 0.17502587735652925,
      "num_tokens": 15335195.0,
      "step": 6995
    },
    {
      "entropy": 5.690118312835693,
      "epoch": 0.6947201270345376,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004959960201693822,
      "loss": 5.5042,
      "mean_token_accuracy": 0.1464548960328102,
      "num_tokens": 15345798.0,
      "step": 7000
    },
    {
      "entropy": 5.568120956420898,
      "epoch": 0.695216355696705,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004959893628837048,
      "loss": 5.3216,
      "mean_token_accuracy": 0.1585929274559021,
      "num_tokens": 15356340.0,
      "step": 7005
    },
    {
      "entropy": 5.606752204895019,
      "epoch": 0.6957125843588726,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004959827001179675,
      "loss": 5.4273,
      "mean_token_accuracy": 0.1500048279762268,
      "num_tokens": 15367795.0,
      "step": 7010
    },
    {
      "entropy": 5.634905433654785,
      "epoch": 0.6962088130210401,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004959760318723353,
      "loss": 5.5592,
      "mean_token_accuracy": 0.14500176906585693,
      "num_tokens": 15379214.0,
      "step": 7015
    },
    {
      "entropy": 5.589799880981445,
      "epoch": 0.6967050416832076,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004959693581469736,
      "loss": 5.3946,
      "mean_token_accuracy": 0.16276417523622513,
      "num_tokens": 15390242.0,
      "step": 7020
    },
    {
      "entropy": 5.543370485305786,
      "epoch": 0.6972012703453752,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004959626789420481,
      "loss": 5.3472,
      "mean_token_accuracy": 0.16189260482788087,
      "num_tokens": 15402600.0,
      "step": 7025
    },
    {
      "entropy": 5.556691026687622,
      "epoch": 0.6976974990075426,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004959559942577243,
      "loss": 5.3659,
      "mean_token_accuracy": 0.15843235999345778,
      "num_tokens": 15412883.0,
      "step": 7030
    },
    {
      "entropy": 5.643964672088623,
      "epoch": 0.6981937276697102,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004959493040941678,
      "loss": 5.4198,
      "mean_token_accuracy": 0.15234392434358596,
      "num_tokens": 15425919.0,
      "step": 7035
    },
    {
      "entropy": 5.552553415298462,
      "epoch": 0.6986899563318777,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004959426084515448,
      "loss": 5.3401,
      "mean_token_accuracy": 0.1577442169189453,
      "num_tokens": 15437604.0,
      "step": 7040
    },
    {
      "entropy": 5.54770336151123,
      "epoch": 0.6991861849940453,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004959359073300212,
      "loss": 5.4063,
      "mean_token_accuracy": 0.15242261290550232,
      "num_tokens": 15449023.0,
      "step": 7045
    },
    {
      "entropy": 5.640282535552979,
      "epoch": 0.6996824136562128,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004959292007297631,
      "loss": 5.4535,
      "mean_token_accuracy": 0.15502861142158508,
      "num_tokens": 15459567.0,
      "step": 7050
    },
    {
      "entropy": 5.591946411132812,
      "epoch": 0.7001786423183803,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000495922488650937,
      "loss": 5.3815,
      "mean_token_accuracy": 0.15652793496847153,
      "num_tokens": 15470950.0,
      "step": 7055
    },
    {
      "entropy": 5.53835244178772,
      "epoch": 0.7006748709805478,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004959157710937092,
      "loss": 5.3718,
      "mean_token_accuracy": 0.15299536138772965,
      "num_tokens": 15481258.0,
      "step": 7060
    },
    {
      "entropy": 5.589594173431396,
      "epoch": 0.7011710996427154,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004959090480582462,
      "loss": 5.3002,
      "mean_token_accuracy": 0.1572806268930435,
      "num_tokens": 15491672.0,
      "step": 7065
    },
    {
      "entropy": 5.533781099319458,
      "epoch": 0.7016673283048829,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004959023195447148,
      "loss": 5.3717,
      "mean_token_accuracy": 0.15430044680833815,
      "num_tokens": 15502342.0,
      "step": 7070
    },
    {
      "entropy": 5.548447132110596,
      "epoch": 0.7021635569670505,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000495895585553282,
      "loss": 5.4055,
      "mean_token_accuracy": 0.1500248782336712,
      "num_tokens": 15513430.0,
      "step": 7075
    },
    {
      "entropy": 5.773319435119629,
      "epoch": 0.7026597856292179,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004958888460841147,
      "loss": 5.5038,
      "mean_token_accuracy": 0.15315764024853706,
      "num_tokens": 15524889.0,
      "step": 7080
    },
    {
      "entropy": 5.56788067817688,
      "epoch": 0.7031560142913855,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004958821011373798,
      "loss": 5.3231,
      "mean_token_accuracy": 0.15642202347517015,
      "num_tokens": 15534998.0,
      "step": 7085
    },
    {
      "entropy": 5.514524364471436,
      "epoch": 0.703652242953553,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004958753507132449,
      "loss": 5.3353,
      "mean_token_accuracy": 0.15758486092090607,
      "num_tokens": 15544752.0,
      "step": 7090
    },
    {
      "entropy": 5.589644813537598,
      "epoch": 0.7041484716157205,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004958685948118772,
      "loss": 5.3486,
      "mean_token_accuracy": 0.1569026455283165,
      "num_tokens": 15555490.0,
      "step": 7095
    },
    {
      "entropy": 5.5551917552948,
      "epoch": 0.7046447002778881,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004958618334334443,
      "loss": 5.2885,
      "mean_token_accuracy": 0.1576731339097023,
      "num_tokens": 15565752.0,
      "step": 7100
    },
    {
      "entropy": 5.55247278213501,
      "epoch": 0.7051409289400555,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004958550665781138,
      "loss": 5.4262,
      "mean_token_accuracy": 0.15204641222953796,
      "num_tokens": 15576368.0,
      "step": 7105
    },
    {
      "entropy": 5.648264694213867,
      "epoch": 0.7056371576022231,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004958482942460536,
      "loss": 5.4218,
      "mean_token_accuracy": 0.15578450560569762,
      "num_tokens": 15587972.0,
      "step": 7110
    },
    {
      "entropy": 5.701238584518433,
      "epoch": 0.7061333862643906,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004958415164374316,
      "loss": 5.4626,
      "mean_token_accuracy": 0.15444232970476152,
      "num_tokens": 15599951.0,
      "step": 7115
    },
    {
      "entropy": 5.559233951568603,
      "epoch": 0.7066296149265582,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004958347331524158,
      "loss": 5.4389,
      "mean_token_accuracy": 0.15119174122810364,
      "num_tokens": 15610183.0,
      "step": 7120
    },
    {
      "entropy": 5.567929315567016,
      "epoch": 0.7071258435887257,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004958279443911745,
      "loss": 5.4394,
      "mean_token_accuracy": 0.15136580318212509,
      "num_tokens": 15621074.0,
      "step": 7125
    },
    {
      "entropy": 5.691843461990357,
      "epoch": 0.7076220722508932,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000495821150153876,
      "loss": 5.449,
      "mean_token_accuracy": 0.15179289430379866,
      "num_tokens": 15632752.0,
      "step": 7130
    },
    {
      "entropy": 5.610294342041016,
      "epoch": 0.7081183009130607,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000495814350440689,
      "loss": 5.4259,
      "mean_token_accuracy": 0.1501299522817135,
      "num_tokens": 15644703.0,
      "step": 7135
    },
    {
      "entropy": 5.569166755676269,
      "epoch": 0.7086145295752283,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004958075452517816,
      "loss": 5.3879,
      "mean_token_accuracy": 0.1546367108821869,
      "num_tokens": 15655278.0,
      "step": 7140
    },
    {
      "entropy": 5.601422882080078,
      "epoch": 0.7091107582373958,
      "grad_norm": 0.875,
      "learning_rate": 0.0004958007345873232,
      "loss": 5.4085,
      "mean_token_accuracy": 0.15654650926589966,
      "num_tokens": 15665998.0,
      "step": 7145
    },
    {
      "entropy": 5.567332220077515,
      "epoch": 0.7096069868995634,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004957939184474821,
      "loss": 5.2947,
      "mean_token_accuracy": 0.15694765597581864,
      "num_tokens": 15676343.0,
      "step": 7150
    },
    {
      "entropy": 5.5898384094238285,
      "epoch": 0.7101032155617308,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004957870968324277,
      "loss": 5.3913,
      "mean_token_accuracy": 0.1532277412712574,
      "num_tokens": 15686520.0,
      "step": 7155
    },
    {
      "entropy": 5.6117500305175785,
      "epoch": 0.7105994442238984,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004957802697423291,
      "loss": 5.5012,
      "mean_token_accuracy": 0.151394285261631,
      "num_tokens": 15698302.0,
      "step": 7160
    },
    {
      "entropy": 5.52632508277893,
      "epoch": 0.7110956728860659,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004957734371773554,
      "loss": 5.2562,
      "mean_token_accuracy": 0.15840691775083543,
      "num_tokens": 15708462.0,
      "step": 7165
    },
    {
      "entropy": 5.590388298034668,
      "epoch": 0.7115919015482334,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004957665991376763,
      "loss": 5.378,
      "mean_token_accuracy": 0.16070500314235686,
      "num_tokens": 15719234.0,
      "step": 7170
    },
    {
      "entropy": 5.53080644607544,
      "epoch": 0.712088130210401,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004957597556234613,
      "loss": 5.4013,
      "mean_token_accuracy": 0.15340089201927185,
      "num_tokens": 15732245.0,
      "step": 7175
    },
    {
      "entropy": 5.595386743545532,
      "epoch": 0.7125843588725684,
      "grad_norm": 1.0,
      "learning_rate": 0.0004957529066348799,
      "loss": 5.3819,
      "mean_token_accuracy": 0.16044221967458724,
      "num_tokens": 15742403.0,
      "step": 7180
    },
    {
      "entropy": 5.509410572052002,
      "epoch": 0.713080587534736,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004957460521721022,
      "loss": 5.2863,
      "mean_token_accuracy": 0.1583533227443695,
      "num_tokens": 15752100.0,
      "step": 7185
    },
    {
      "entropy": 5.549450254440307,
      "epoch": 0.7135768161969035,
      "grad_norm": 0.921875,
      "learning_rate": 0.000495739192235298,
      "loss": 5.3787,
      "mean_token_accuracy": 0.154872128367424,
      "num_tokens": 15762315.0,
      "step": 7190
    },
    {
      "entropy": 5.552863550186157,
      "epoch": 0.7140730448590711,
      "grad_norm": 1.0,
      "learning_rate": 0.0004957323268246376,
      "loss": 5.3982,
      "mean_token_accuracy": 0.15654573291540147,
      "num_tokens": 15773841.0,
      "step": 7195
    },
    {
      "entropy": 5.593350791931153,
      "epoch": 0.7145692735212386,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004957254559402911,
      "loss": 5.3425,
      "mean_token_accuracy": 0.15729348957538605,
      "num_tokens": 15785511.0,
      "step": 7200
    },
    {
      "entropy": 5.536798667907715,
      "epoch": 0.7150655021834061,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004957185795824288,
      "loss": 5.3716,
      "mean_token_accuracy": 0.1508410856127739,
      "num_tokens": 15795394.0,
      "step": 7205
    },
    {
      "entropy": 5.525018167495728,
      "epoch": 0.7155617308455736,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004957116977512214,
      "loss": 5.3375,
      "mean_token_accuracy": 0.15722720101475715,
      "num_tokens": 15805894.0,
      "step": 7210
    },
    {
      "entropy": 5.599946451187134,
      "epoch": 0.7160579595077412,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004957048104468395,
      "loss": 5.395,
      "mean_token_accuracy": 0.14500413239002227,
      "num_tokens": 15816595.0,
      "step": 7215
    },
    {
      "entropy": 5.561121034622192,
      "epoch": 0.7165541881699087,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004956979176694539,
      "loss": 5.3367,
      "mean_token_accuracy": 0.16101332306861876,
      "num_tokens": 15827638.0,
      "step": 7220
    },
    {
      "entropy": 5.531207656860351,
      "epoch": 0.7170504168320763,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004956910194192355,
      "loss": 5.4097,
      "mean_token_accuracy": 0.1532890185713768,
      "num_tokens": 15838437.0,
      "step": 7225
    },
    {
      "entropy": 5.690220689773559,
      "epoch": 0.7175466454942437,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004956841156963553,
      "loss": 5.5203,
      "mean_token_accuracy": 0.154303640127182,
      "num_tokens": 15850110.0,
      "step": 7230
    },
    {
      "entropy": 5.58256688117981,
      "epoch": 0.7180428741564113,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004956772065009846,
      "loss": 5.3749,
      "mean_token_accuracy": 0.15135821774601937,
      "num_tokens": 15861786.0,
      "step": 7235
    },
    {
      "entropy": 5.567192411422729,
      "epoch": 0.7185391028185788,
      "grad_norm": 1.0,
      "learning_rate": 0.0004956702918332948,
      "loss": 5.3524,
      "mean_token_accuracy": 0.152687868475914,
      "num_tokens": 15871574.0,
      "step": 7240
    },
    {
      "entropy": 5.578045654296875,
      "epoch": 0.7190353314807463,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004956633716934572,
      "loss": 5.3429,
      "mean_token_accuracy": 0.1614610031247139,
      "num_tokens": 15881944.0,
      "step": 7245
    },
    {
      "entropy": 5.522058582305908,
      "epoch": 0.7195315601429139,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004956564460816433,
      "loss": 5.3467,
      "mean_token_accuracy": 0.1552556797862053,
      "num_tokens": 15892511.0,
      "step": 7250
    },
    {
      "entropy": 5.567608165740967,
      "epoch": 0.7200277888050813,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004956495149980252,
      "loss": 5.3759,
      "mean_token_accuracy": 0.15121624171733855,
      "num_tokens": 15903387.0,
      "step": 7255
    },
    {
      "entropy": 5.598010063171387,
      "epoch": 0.7205240174672489,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004956425784427745,
      "loss": 5.4062,
      "mean_token_accuracy": 0.15305554121732712,
      "num_tokens": 15915117.0,
      "step": 7260
    },
    {
      "entropy": 5.5196246147155765,
      "epoch": 0.7210202461294164,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004956356364160634,
      "loss": 5.3573,
      "mean_token_accuracy": 0.15544371008872987,
      "num_tokens": 15925163.0,
      "step": 7265
    },
    {
      "entropy": 5.449028968811035,
      "epoch": 0.721516474791584,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004956286889180639,
      "loss": 5.2431,
      "mean_token_accuracy": 0.16879490315914153,
      "num_tokens": 15935021.0,
      "step": 7270
    },
    {
      "entropy": 5.512518739700317,
      "epoch": 0.7220127034537515,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004956217359489483,
      "loss": 5.3681,
      "mean_token_accuracy": 0.15262435823678971,
      "num_tokens": 15945416.0,
      "step": 7275
    },
    {
      "entropy": 5.648502159118652,
      "epoch": 0.722508932115919,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000495614777508889,
      "loss": 5.4761,
      "mean_token_accuracy": 0.14991702884435654,
      "num_tokens": 15955388.0,
      "step": 7280
    },
    {
      "entropy": 5.64871301651001,
      "epoch": 0.7230051607780865,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004956078135980586,
      "loss": 5.4853,
      "mean_token_accuracy": 0.15207213908433914,
      "num_tokens": 15966721.0,
      "step": 7285
    },
    {
      "entropy": 5.565513181686401,
      "epoch": 0.7235013894402541,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004956008442166299,
      "loss": 5.4075,
      "mean_token_accuracy": 0.15500583946704866,
      "num_tokens": 15976388.0,
      "step": 7290
    },
    {
      "entropy": 5.521198511123657,
      "epoch": 0.7239976181024216,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004955938693647756,
      "loss": 5.3026,
      "mean_token_accuracy": 0.15797873884439467,
      "num_tokens": 15986632.0,
      "step": 7295
    },
    {
      "entropy": 5.5546427249908445,
      "epoch": 0.7244938467645892,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004955868890426686,
      "loss": 5.3512,
      "mean_token_accuracy": 0.15412055999040603,
      "num_tokens": 15996626.0,
      "step": 7300
    },
    {
      "entropy": 5.508546781539917,
      "epoch": 0.7249900754267566,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004955799032504821,
      "loss": 5.3186,
      "mean_token_accuracy": 0.1552100211381912,
      "num_tokens": 16007055.0,
      "step": 7305
    },
    {
      "entropy": 5.623987674713135,
      "epoch": 0.7254863040889242,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004955729119883893,
      "loss": 5.4046,
      "mean_token_accuracy": 0.15193189680576324,
      "num_tokens": 16019087.0,
      "step": 7310
    },
    {
      "entropy": 5.587416172027588,
      "epoch": 0.7259825327510917,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004955659152565636,
      "loss": 5.3931,
      "mean_token_accuracy": 0.1515917584300041,
      "num_tokens": 16029980.0,
      "step": 7315
    },
    {
      "entropy": 5.5827936172485355,
      "epoch": 0.7264787614132592,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004955589130551784,
      "loss": 5.3693,
      "mean_token_accuracy": 0.15805643126368524,
      "num_tokens": 16040932.0,
      "step": 7320
    },
    {
      "entropy": 5.461764478683472,
      "epoch": 0.7269749900754268,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004955519053844076,
      "loss": 5.3339,
      "mean_token_accuracy": 0.1601037234067917,
      "num_tokens": 16051809.0,
      "step": 7325
    },
    {
      "entropy": 5.5657754898071286,
      "epoch": 0.7274712187375942,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004955448922444246,
      "loss": 5.3872,
      "mean_token_accuracy": 0.15353999137878419,
      "num_tokens": 16062957.0,
      "step": 7330
    },
    {
      "entropy": 5.607474231719971,
      "epoch": 0.7279674473997618,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004955378736354036,
      "loss": 5.4259,
      "mean_token_accuracy": 0.15297536104917525,
      "num_tokens": 16073592.0,
      "step": 7335
    },
    {
      "entropy": 5.541664505004883,
      "epoch": 0.7284636760619293,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004955308495575185,
      "loss": 5.368,
      "mean_token_accuracy": 0.15536474511027337,
      "num_tokens": 16085054.0,
      "step": 7340
    },
    {
      "entropy": 5.595587205886841,
      "epoch": 0.7289599047240969,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004955238200109436,
      "loss": 5.4466,
      "mean_token_accuracy": 0.15401146709918975,
      "num_tokens": 16096936.0,
      "step": 7345
    },
    {
      "entropy": 5.619851350784302,
      "epoch": 0.7294561333862644,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004955167849958531,
      "loss": 5.468,
      "mean_token_accuracy": 0.15150101929903032,
      "num_tokens": 16107673.0,
      "step": 7350
    },
    {
      "entropy": 5.576428174972534,
      "epoch": 0.729952362048432,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004955097445124215,
      "loss": 5.3278,
      "mean_token_accuracy": 0.16196224093437195,
      "num_tokens": 16117941.0,
      "step": 7355
    },
    {
      "entropy": 5.561256551742554,
      "epoch": 0.7304485907105994,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004955026985608235,
      "loss": 5.295,
      "mean_token_accuracy": 0.16231960356235503,
      "num_tokens": 16128737.0,
      "step": 7360
    },
    {
      "entropy": 5.662528657913208,
      "epoch": 0.730944819372767,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004954956471412336,
      "loss": 5.4283,
      "mean_token_accuracy": 0.15473217517137527,
      "num_tokens": 16139599.0,
      "step": 7365
    },
    {
      "entropy": 5.456266450881958,
      "epoch": 0.7314410480349345,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004954885902538269,
      "loss": 5.2789,
      "mean_token_accuracy": 0.1517025738954544,
      "num_tokens": 16150796.0,
      "step": 7370
    },
    {
      "entropy": 5.545576524734497,
      "epoch": 0.7319372766971021,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004954815278987781,
      "loss": 5.3452,
      "mean_token_accuracy": 0.15946361273527146,
      "num_tokens": 16161236.0,
      "step": 7375
    },
    {
      "entropy": 5.6335042953491214,
      "epoch": 0.7324335053592695,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004954744600762627,
      "loss": 5.4484,
      "mean_token_accuracy": 0.15341491848230362,
      "num_tokens": 16172210.0,
      "step": 7380
    },
    {
      "entropy": 5.5065947532653805,
      "epoch": 0.7329297340214371,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004954673867864555,
      "loss": 5.3199,
      "mean_token_accuracy": 0.16296881139278413,
      "num_tokens": 16182950.0,
      "step": 7385
    },
    {
      "entropy": 5.499658250808716,
      "epoch": 0.7334259626836046,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004954603080295323,
      "loss": 5.2341,
      "mean_token_accuracy": 0.16565425992012023,
      "num_tokens": 16193971.0,
      "step": 7390
    },
    {
      "entropy": 5.560158967971802,
      "epoch": 0.7339221913457721,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004954532238056686,
      "loss": 5.3644,
      "mean_token_accuracy": 0.15560306906700133,
      "num_tokens": 16205774.0,
      "step": 7395
    },
    {
      "entropy": 5.539269638061524,
      "epoch": 0.7344184200079397,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004954461341150399,
      "loss": 5.3567,
      "mean_token_accuracy": 0.15304027497768402,
      "num_tokens": 16216626.0,
      "step": 7400
    },
    {
      "entropy": 5.529664468765259,
      "epoch": 0.7349146486701071,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004954390389578222,
      "loss": 5.4158,
      "mean_token_accuracy": 0.1536064475774765,
      "num_tokens": 16227748.0,
      "step": 7405
    },
    {
      "entropy": 5.623319101333618,
      "epoch": 0.7354108773322747,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004954319383341912,
      "loss": 5.4101,
      "mean_token_accuracy": 0.1580138735473156,
      "num_tokens": 16240390.0,
      "step": 7410
    },
    {
      "entropy": 5.578856325149536,
      "epoch": 0.7359071059944422,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004954248322443232,
      "loss": 5.3393,
      "mean_token_accuracy": 0.1636115148663521,
      "num_tokens": 16250563.0,
      "step": 7415
    },
    {
      "entropy": 5.562770986557007,
      "epoch": 0.7364033346566098,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004954177206883943,
      "loss": 5.3993,
      "mean_token_accuracy": 0.14981402605772018,
      "num_tokens": 16260722.0,
      "step": 7420
    },
    {
      "entropy": 5.681869029998779,
      "epoch": 0.7368995633187773,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004954106036665809,
      "loss": 5.4131,
      "mean_token_accuracy": 0.15602878108620644,
      "num_tokens": 16270258.0,
      "step": 7425
    },
    {
      "entropy": 5.556429052352906,
      "epoch": 0.7373957919809448,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004954034811790594,
      "loss": 5.2887,
      "mean_token_accuracy": 0.15933127403259278,
      "num_tokens": 16279504.0,
      "step": 7430
    },
    {
      "entropy": 5.492929124832154,
      "epoch": 0.7378920206431123,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004953963532260067,
      "loss": 5.3013,
      "mean_token_accuracy": 0.16148138046264648,
      "num_tokens": 16291327.0,
      "step": 7435
    },
    {
      "entropy": 5.554673480987549,
      "epoch": 0.7383882493052799,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004953892198075991,
      "loss": 5.4413,
      "mean_token_accuracy": 0.1550499364733696,
      "num_tokens": 16302778.0,
      "step": 7440
    },
    {
      "entropy": 5.6280824661254885,
      "epoch": 0.7388844779674474,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004953820809240138,
      "loss": 5.383,
      "mean_token_accuracy": 0.14624368250370026,
      "num_tokens": 16314373.0,
      "step": 7445
    },
    {
      "entropy": 5.553422307968139,
      "epoch": 0.739380706629615,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004953749365754279,
      "loss": 5.3488,
      "mean_token_accuracy": 0.16128434240818024,
      "num_tokens": 16324518.0,
      "step": 7450
    },
    {
      "entropy": 5.535141515731811,
      "epoch": 0.7398769352917824,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004953677867620184,
      "loss": 5.2936,
      "mean_token_accuracy": 0.16068151593208313,
      "num_tokens": 16335657.0,
      "step": 7455
    },
    {
      "entropy": 5.587289524078369,
      "epoch": 0.74037316395395,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004953606314839626,
      "loss": 5.4918,
      "mean_token_accuracy": 0.1496763989329338,
      "num_tokens": 16346497.0,
      "step": 7460
    },
    {
      "entropy": 5.520878314971924,
      "epoch": 0.7408693926161175,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000495353470741438,
      "loss": 5.3644,
      "mean_token_accuracy": 0.15484014749526978,
      "num_tokens": 16358175.0,
      "step": 7465
    },
    {
      "entropy": 5.629048538208008,
      "epoch": 0.741365621278285,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000495346304534622,
      "loss": 5.4611,
      "mean_token_accuracy": 0.15078660398721694,
      "num_tokens": 16369803.0,
      "step": 7470
    },
    {
      "entropy": 5.515393447875977,
      "epoch": 0.7418618499404526,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004953391328636926,
      "loss": 5.1748,
      "mean_token_accuracy": 0.16214750409126283,
      "num_tokens": 16380535.0,
      "step": 7475
    },
    {
      "entropy": 5.490961980819702,
      "epoch": 0.74235807860262,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004953319557288275,
      "loss": 5.3495,
      "mean_token_accuracy": 0.15547017753124237,
      "num_tokens": 16390518.0,
      "step": 7480
    },
    {
      "entropy": 5.531258535385132,
      "epoch": 0.7428543072647876,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004953247731302046,
      "loss": 5.3727,
      "mean_token_accuracy": 0.1484750598669052,
      "num_tokens": 16400627.0,
      "step": 7485
    },
    {
      "entropy": 5.573671674728393,
      "epoch": 0.7433505359269551,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004953175850680022,
      "loss": 5.436,
      "mean_token_accuracy": 0.1594882406294346,
      "num_tokens": 16411136.0,
      "step": 7490
    },
    {
      "entropy": 5.659179639816284,
      "epoch": 0.7438467645891227,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004953103915423983,
      "loss": 5.3683,
      "mean_token_accuracy": 0.15788947343826293,
      "num_tokens": 16422830.0,
      "step": 7495
    },
    {
      "entropy": 5.513751602172851,
      "epoch": 0.7443429932512902,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004953031925535715,
      "loss": 5.3206,
      "mean_token_accuracy": 0.162714684009552,
      "num_tokens": 16434158.0,
      "step": 7500
    },
    {
      "entropy": 5.512455320358276,
      "epoch": 0.7448392219134577,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004952959881017003,
      "loss": 5.3389,
      "mean_token_accuracy": 0.1615119531750679,
      "num_tokens": 16444562.0,
      "step": 7505
    },
    {
      "entropy": 5.5862266540527346,
      "epoch": 0.7453354505756252,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004952887781869631,
      "loss": 5.3818,
      "mean_token_accuracy": 0.15387180745601653,
      "num_tokens": 16456353.0,
      "step": 7510
    },
    {
      "entropy": 5.5833018779754635,
      "epoch": 0.7458316792377928,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004952815628095391,
      "loss": 5.3976,
      "mean_token_accuracy": 0.14866131767630578,
      "num_tokens": 16465756.0,
      "step": 7515
    },
    {
      "entropy": 5.537196159362793,
      "epoch": 0.7463279078999603,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004952743419696069,
      "loss": 5.3311,
      "mean_token_accuracy": 0.1572333201766014,
      "num_tokens": 16477562.0,
      "step": 7520
    },
    {
      "entropy": 5.603846645355224,
      "epoch": 0.7468241365621279,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004952671156673456,
      "loss": 5.3832,
      "mean_token_accuracy": 0.152971251308918,
      "num_tokens": 16488134.0,
      "step": 7525
    },
    {
      "entropy": 5.591313695907592,
      "epoch": 0.7473203652242953,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004952598839029345,
      "loss": 5.3663,
      "mean_token_accuracy": 0.15624374002218247,
      "num_tokens": 16499146.0,
      "step": 7530
    },
    {
      "entropy": 5.573712968826294,
      "epoch": 0.7478165938864629,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004952526466765529,
      "loss": 5.4145,
      "mean_token_accuracy": 0.15605186074972152,
      "num_tokens": 16510334.0,
      "step": 7535
    },
    {
      "entropy": 5.56377158164978,
      "epoch": 0.7483128225486304,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004952454039883804,
      "loss": 5.2854,
      "mean_token_accuracy": 0.15974623411893846,
      "num_tokens": 16520574.0,
      "step": 7540
    },
    {
      "entropy": 5.597139978408814,
      "epoch": 0.7488090512107979,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004952381558385963,
      "loss": 5.4305,
      "mean_token_accuracy": 0.15135585218667985,
      "num_tokens": 16532246.0,
      "step": 7545
    },
    {
      "entropy": 5.704423427581787,
      "epoch": 0.7493052798729655,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004952309022273805,
      "loss": 5.4,
      "mean_token_accuracy": 0.15351941138505937,
      "num_tokens": 16541756.0,
      "step": 7550
    },
    {
      "entropy": 5.586373615264892,
      "epoch": 0.7498015085351329,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004952236431549129,
      "loss": 5.4054,
      "mean_token_accuracy": 0.1476966083049774,
      "num_tokens": 16551061.0,
      "step": 7555
    },
    {
      "entropy": 5.5230710983276365,
      "epoch": 0.7502977371973005,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004952163786213735,
      "loss": 5.352,
      "mean_token_accuracy": 0.15240754783153534,
      "num_tokens": 16561155.0,
      "step": 7560
    },
    {
      "entropy": 5.549474906921387,
      "epoch": 0.750793965859468,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004952091086269425,
      "loss": 5.3117,
      "mean_token_accuracy": 0.14989598616957664,
      "num_tokens": 16571843.0,
      "step": 7565
    },
    {
      "entropy": 5.5611807346344,
      "epoch": 0.7512901945216356,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004952018331718001,
      "loss": 5.3909,
      "mean_token_accuracy": 0.15381356999278067,
      "num_tokens": 16583395.0,
      "step": 7570
    },
    {
      "entropy": 5.633284044265747,
      "epoch": 0.7517864231838031,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004951945522561268,
      "loss": 5.4386,
      "mean_token_accuracy": 0.15344821512699128,
      "num_tokens": 16594545.0,
      "step": 7575
    },
    {
      "entropy": 5.657517290115356,
      "epoch": 0.7522826518459707,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004951872658801029,
      "loss": 5.4965,
      "mean_token_accuracy": 0.15352494940161704,
      "num_tokens": 16605901.0,
      "step": 7580
    },
    {
      "entropy": 5.569665765762329,
      "epoch": 0.7527788805081381,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004951799740439093,
      "loss": 5.4758,
      "mean_token_accuracy": 0.14774393886327744,
      "num_tokens": 16616783.0,
      "step": 7585
    },
    {
      "entropy": 5.565461921691894,
      "epoch": 0.7532751091703057,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000495172676747727,
      "loss": 5.3861,
      "mean_token_accuracy": 0.15814861059188842,
      "num_tokens": 16628655.0,
      "step": 7590
    },
    {
      "entropy": 5.602010869979859,
      "epoch": 0.7537713378324732,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004951653739917366,
      "loss": 5.3932,
      "mean_token_accuracy": 0.1518102467060089,
      "num_tokens": 16639403.0,
      "step": 7595
    },
    {
      "entropy": 5.6291468143463135,
      "epoch": 0.7542675664946408,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004951580657761192,
      "loss": 5.4034,
      "mean_token_accuracy": 0.15876073837280275,
      "num_tokens": 16649937.0,
      "step": 7600
    },
    {
      "entropy": 5.473993968963623,
      "epoch": 0.7547637951568082,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004951507521010563,
      "loss": 5.2965,
      "mean_token_accuracy": 0.15884831845760344,
      "num_tokens": 16661466.0,
      "step": 7605
    },
    {
      "entropy": 5.507606792449951,
      "epoch": 0.7552600238189758,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004951434329667293,
      "loss": 5.3134,
      "mean_token_accuracy": 0.15895138829946517,
      "num_tokens": 16673362.0,
      "step": 7610
    },
    {
      "entropy": 5.548601722717285,
      "epoch": 0.7557562524811433,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004951361083733192,
      "loss": 5.3543,
      "mean_token_accuracy": 0.1521439790725708,
      "num_tokens": 16685555.0,
      "step": 7615
    },
    {
      "entropy": 5.573924922943116,
      "epoch": 0.7562524811433108,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004951287783210082,
      "loss": 5.336,
      "mean_token_accuracy": 0.15216623693704606,
      "num_tokens": 16696006.0,
      "step": 7620
    },
    {
      "entropy": 5.557129573822022,
      "epoch": 0.7567487098054784,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004951214428099778,
      "loss": 5.335,
      "mean_token_accuracy": 0.15660312324762343,
      "num_tokens": 16706784.0,
      "step": 7625
    },
    {
      "entropy": 5.508156061172485,
      "epoch": 0.7572449384676458,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004951141018404099,
      "loss": 5.3166,
      "mean_token_accuracy": 0.1631629690527916,
      "num_tokens": 16718492.0,
      "step": 7630
    },
    {
      "entropy": 5.532926273345947,
      "epoch": 0.7577411671298134,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004951067554124865,
      "loss": 5.3799,
      "mean_token_accuracy": 0.15535615980625153,
      "num_tokens": 16729711.0,
      "step": 7635
    },
    {
      "entropy": 5.538987731933593,
      "epoch": 0.7582373957919809,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00049509940352639,
      "loss": 5.3595,
      "mean_token_accuracy": 0.15555189549922943,
      "num_tokens": 16739913.0,
      "step": 7640
    },
    {
      "entropy": 5.638753747940063,
      "epoch": 0.7587336244541485,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004950920461823024,
      "loss": 5.4033,
      "mean_token_accuracy": 0.1522645890712738,
      "num_tokens": 16752273.0,
      "step": 7645
    },
    {
      "entropy": 5.5679864406585695,
      "epoch": 0.759229853116316,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004950846833804064,
      "loss": 5.34,
      "mean_token_accuracy": 0.15633462369441986,
      "num_tokens": 16762050.0,
      "step": 7650
    },
    {
      "entropy": 5.5330182075500485,
      "epoch": 0.7597260817784836,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004950773151208845,
      "loss": 5.4032,
      "mean_token_accuracy": 0.15630503445863725,
      "num_tokens": 16772376.0,
      "step": 7655
    },
    {
      "entropy": 5.546814680099487,
      "epoch": 0.760222310440651,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004950699414039194,
      "loss": 5.3648,
      "mean_token_accuracy": 0.15277177542448045,
      "num_tokens": 16783307.0,
      "step": 7660
    },
    {
      "entropy": 5.574085474014282,
      "epoch": 0.7607185391028186,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000495062562229694,
      "loss": 5.2402,
      "mean_token_accuracy": 0.16051952093839644,
      "num_tokens": 16793582.0,
      "step": 7665
    },
    {
      "entropy": 5.505533981323242,
      "epoch": 0.7612147677649861,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004950551775983912,
      "loss": 5.3639,
      "mean_token_accuracy": 0.1557060956954956,
      "num_tokens": 16805160.0,
      "step": 7670
    },
    {
      "entropy": 5.5394867897033695,
      "epoch": 0.7617109964271537,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004950477875101942,
      "loss": 5.4194,
      "mean_token_accuracy": 0.15286368802189826,
      "num_tokens": 16816108.0,
      "step": 7675
    },
    {
      "entropy": 5.521533107757568,
      "epoch": 0.7622072250893211,
      "grad_norm": 0.875,
      "learning_rate": 0.0004950403919652863,
      "loss": 5.2943,
      "mean_token_accuracy": 0.15909659415483474,
      "num_tokens": 16828126.0,
      "step": 7680
    },
    {
      "entropy": 5.584299373626709,
      "epoch": 0.7627034537514887,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004950329909638507,
      "loss": 5.3624,
      "mean_token_accuracy": 0.155840203166008,
      "num_tokens": 16838186.0,
      "step": 7685
    },
    {
      "entropy": 5.565781497955323,
      "epoch": 0.7631996824136562,
      "grad_norm": 0.80078125,
      "learning_rate": 0.000495025584506071,
      "loss": 5.3038,
      "mean_token_accuracy": 0.15668020844459535,
      "num_tokens": 16848588.0,
      "step": 7690
    },
    {
      "entropy": 5.584897136688232,
      "epoch": 0.7636959110758237,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004950181725921311,
      "loss": 5.4198,
      "mean_token_accuracy": 0.1540161445736885,
      "num_tokens": 16860269.0,
      "step": 7695
    },
    {
      "entropy": 5.563757610321045,
      "epoch": 0.7641921397379913,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004950107552222145,
      "loss": 5.4135,
      "mean_token_accuracy": 0.1542084261775017,
      "num_tokens": 16872221.0,
      "step": 7700
    },
    {
      "entropy": 5.539776229858399,
      "epoch": 0.7646883684001587,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004950033323965052,
      "loss": 5.2995,
      "mean_token_accuracy": 0.1555795654654503,
      "num_tokens": 16882588.0,
      "step": 7705
    },
    {
      "entropy": 5.605736160278321,
      "epoch": 0.7651845970623263,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004949959041151875,
      "loss": 5.3689,
      "mean_token_accuracy": 0.15667574852705002,
      "num_tokens": 16894278.0,
      "step": 7710
    },
    {
      "entropy": 5.567713689804077,
      "epoch": 0.7656808257244938,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004949884703784453,
      "loss": 5.3413,
      "mean_token_accuracy": 0.15670668408274652,
      "num_tokens": 16905448.0,
      "step": 7715
    },
    {
      "entropy": 5.53463191986084,
      "epoch": 0.7661770543866614,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000494981031186463,
      "loss": 5.3692,
      "mean_token_accuracy": 0.15420209094882012,
      "num_tokens": 16914960.0,
      "step": 7720
    },
    {
      "entropy": 5.57317419052124,
      "epoch": 0.7666732830488289,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004949735865394253,
      "loss": 5.4368,
      "mean_token_accuracy": 0.15323926880955696,
      "num_tokens": 16926839.0,
      "step": 7725
    },
    {
      "entropy": 5.56858344078064,
      "epoch": 0.7671695117109965,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004949661364375167,
      "loss": 5.312,
      "mean_token_accuracy": 0.16317027509212495,
      "num_tokens": 16938553.0,
      "step": 7730
    },
    {
      "entropy": 5.59964017868042,
      "epoch": 0.7676657403731639,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004949586808809217,
      "loss": 5.3896,
      "mean_token_accuracy": 0.15333618819713593,
      "num_tokens": 16947987.0,
      "step": 7735
    },
    {
      "entropy": 5.489641618728638,
      "epoch": 0.7681619690353315,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004949512198698256,
      "loss": 5.3169,
      "mean_token_accuracy": 0.15825526118278505,
      "num_tokens": 16959359.0,
      "step": 7740
    },
    {
      "entropy": 5.5458088397979735,
      "epoch": 0.768658197697499,
      "grad_norm": 0.875,
      "learning_rate": 0.000494943753404413,
      "loss": 5.3621,
      "mean_token_accuracy": 0.15992436558008194,
      "num_tokens": 16970731.0,
      "step": 7745
    },
    {
      "entropy": 5.507170724868774,
      "epoch": 0.7691544263596666,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004949362814848693,
      "loss": 5.3168,
      "mean_token_accuracy": 0.16060893684625627,
      "num_tokens": 16981868.0,
      "step": 7750
    },
    {
      "entropy": 5.616957569122315,
      "epoch": 0.769650655021834,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004949288041113799,
      "loss": 5.3748,
      "mean_token_accuracy": 0.1563388839364052,
      "num_tokens": 16994639.0,
      "step": 7755
    },
    {
      "entropy": 5.518722724914551,
      "epoch": 0.7701468836840016,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004949213212841299,
      "loss": 5.3671,
      "mean_token_accuracy": 0.15757046043872833,
      "num_tokens": 17005067.0,
      "step": 7760
    },
    {
      "entropy": 5.533456134796142,
      "epoch": 0.7706431123461691,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000494913833003305,
      "loss": 5.2862,
      "mean_token_accuracy": 0.16234548538923263,
      "num_tokens": 17015102.0,
      "step": 7765
    },
    {
      "entropy": 5.556436967849732,
      "epoch": 0.7711393410083366,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000494906339269091,
      "loss": 5.3135,
      "mean_token_accuracy": 0.1586420491337776,
      "num_tokens": 17026290.0,
      "step": 7770
    },
    {
      "entropy": 5.547585296630859,
      "epoch": 0.7716355696705042,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004948988400816736,
      "loss": 5.2758,
      "mean_token_accuracy": 0.1543065443634987,
      "num_tokens": 17035445.0,
      "step": 7775
    },
    {
      "entropy": 5.577733421325684,
      "epoch": 0.7721317983326716,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004948913354412388,
      "loss": 5.3868,
      "mean_token_accuracy": 0.15804281383752822,
      "num_tokens": 17046515.0,
      "step": 7780
    },
    {
      "entropy": 5.586125135421753,
      "epoch": 0.7726280269948392,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004948838253479727,
      "loss": 5.3576,
      "mean_token_accuracy": 0.15887074917554855,
      "num_tokens": 17056988.0,
      "step": 7785
    },
    {
      "entropy": 5.618466567993164,
      "epoch": 0.7731242556570067,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004948763098020616,
      "loss": 5.4159,
      "mean_token_accuracy": 0.14824309647083284,
      "num_tokens": 17068304.0,
      "step": 7790
    },
    {
      "entropy": 5.5217780590057375,
      "epoch": 0.7736204843191743,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004948687888036918,
      "loss": 5.328,
      "mean_token_accuracy": 0.1570253327488899,
      "num_tokens": 17078405.0,
      "step": 7795
    },
    {
      "entropy": 5.502141380310059,
      "epoch": 0.7741167129813418,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004948612623530497,
      "loss": 5.2224,
      "mean_token_accuracy": 0.159689000248909,
      "num_tokens": 17087839.0,
      "step": 7800
    },
    {
      "entropy": 5.461894512176514,
      "epoch": 0.7746129416435094,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004948537304503221,
      "loss": 5.3239,
      "mean_token_accuracy": 0.15929722636938096,
      "num_tokens": 17099386.0,
      "step": 7805
    },
    {
      "entropy": 5.602202606201172,
      "epoch": 0.7751091703056768,
      "grad_norm": 0.875,
      "learning_rate": 0.0004948461930956959,
      "loss": 5.399,
      "mean_token_accuracy": 0.1521478608250618,
      "num_tokens": 17111754.0,
      "step": 7810
    },
    {
      "entropy": 5.561999034881592,
      "epoch": 0.7756053989678444,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004948386502893577,
      "loss": 5.2793,
      "mean_token_accuracy": 0.162553308904171,
      "num_tokens": 17122882.0,
      "step": 7815
    },
    {
      "entropy": 5.533847236633301,
      "epoch": 0.7761016276300119,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004948311020314946,
      "loss": 5.3924,
      "mean_token_accuracy": 0.1520935371518135,
      "num_tokens": 17133986.0,
      "step": 7820
    },
    {
      "entropy": 5.56985559463501,
      "epoch": 0.7765978562921795,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004948235483222941,
      "loss": 5.3505,
      "mean_token_accuracy": 0.16134029030799865,
      "num_tokens": 17144784.0,
      "step": 7825
    },
    {
      "entropy": 5.60482759475708,
      "epoch": 0.777094084954347,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004948159891619431,
      "loss": 5.4663,
      "mean_token_accuracy": 0.14836366921663285,
      "num_tokens": 17156328.0,
      "step": 7830
    },
    {
      "entropy": 5.627248239517212,
      "epoch": 0.7775903136165145,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004948084245506293,
      "loss": 5.2866,
      "mean_token_accuracy": 0.15721032172441482,
      "num_tokens": 17168062.0,
      "step": 7835
    },
    {
      "entropy": 5.542794609069825,
      "epoch": 0.778086542278682,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00049480085448854,
      "loss": 5.3151,
      "mean_token_accuracy": 0.15671407729387282,
      "num_tokens": 17178872.0,
      "step": 7840
    },
    {
      "entropy": 5.547116708755493,
      "epoch": 0.7785827709408495,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004947932789758635,
      "loss": 5.3648,
      "mean_token_accuracy": 0.15643517822027206,
      "num_tokens": 17189445.0,
      "step": 7845
    },
    {
      "entropy": 5.535507345199585,
      "epoch": 0.7790789996030171,
      "grad_norm": 1.015625,
      "learning_rate": 0.000494785698012787,
      "loss": 5.3167,
      "mean_token_accuracy": 0.15999527126550675,
      "num_tokens": 17199454.0,
      "step": 7850
    },
    {
      "entropy": 5.541446352005005,
      "epoch": 0.7795752282651845,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004947781115994989,
      "loss": 5.3837,
      "mean_token_accuracy": 0.15075388327240943,
      "num_tokens": 17210729.0,
      "step": 7855
    },
    {
      "entropy": 5.557825946807862,
      "epoch": 0.7800714569273521,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004947705197361872,
      "loss": 5.4357,
      "mean_token_accuracy": 0.15301606208086013,
      "num_tokens": 17222961.0,
      "step": 7860
    },
    {
      "entropy": 5.54502305984497,
      "epoch": 0.7805676855895196,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004947629224230401,
      "loss": 5.3163,
      "mean_token_accuracy": 0.1613306075334549,
      "num_tokens": 17234026.0,
      "step": 7865
    },
    {
      "entropy": 5.628870868682862,
      "epoch": 0.7810639142516872,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004947553196602461,
      "loss": 5.4426,
      "mean_token_accuracy": 0.15443769246339797,
      "num_tokens": 17244643.0,
      "step": 7870
    },
    {
      "entropy": 5.638962697982788,
      "epoch": 0.7815601429138547,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004947477114479939,
      "loss": 5.3642,
      "mean_token_accuracy": 0.16114916205406188,
      "num_tokens": 17255382.0,
      "step": 7875
    },
    {
      "entropy": 5.512335777282715,
      "epoch": 0.7820563715760223,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004947400977864718,
      "loss": 5.3355,
      "mean_token_accuracy": 0.15462656915187836,
      "num_tokens": 17265860.0,
      "step": 7880
    },
    {
      "entropy": 5.566488218307495,
      "epoch": 0.7825526002381897,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004947324786758688,
      "loss": 5.4347,
      "mean_token_accuracy": 0.15429544374346732,
      "num_tokens": 17277570.0,
      "step": 7885
    },
    {
      "entropy": 5.661877918243408,
      "epoch": 0.7830488289003573,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004947248541163738,
      "loss": 5.4724,
      "mean_token_accuracy": 0.14807592257857322,
      "num_tokens": 17288684.0,
      "step": 7890
    },
    {
      "entropy": 5.573632764816284,
      "epoch": 0.7835450575625248,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004947172241081759,
      "loss": 5.3097,
      "mean_token_accuracy": 0.15547684729099273,
      "num_tokens": 17299541.0,
      "step": 7895
    },
    {
      "entropy": 5.56752610206604,
      "epoch": 0.7840412862246924,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004947095886514643,
      "loss": 5.2732,
      "mean_token_accuracy": 0.16467752754688264,
      "num_tokens": 17310243.0,
      "step": 7900
    },
    {
      "entropy": 5.575685262680054,
      "epoch": 0.7845375148868599,
      "grad_norm": 1.0,
      "learning_rate": 0.0004947019477464283,
      "loss": 5.3846,
      "mean_token_accuracy": 0.15891839414834977,
      "num_tokens": 17321167.0,
      "step": 7905
    },
    {
      "entropy": 5.497011375427246,
      "epoch": 0.7850337435490274,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004946943013932574,
      "loss": 5.3484,
      "mean_token_accuracy": 0.1585032068192959,
      "num_tokens": 17331076.0,
      "step": 7910
    },
    {
      "entropy": 5.605291795730591,
      "epoch": 0.7855299722111949,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004946866495921413,
      "loss": 5.3346,
      "mean_token_accuracy": 0.15110956281423568,
      "num_tokens": 17341337.0,
      "step": 7915
    },
    {
      "entropy": 5.592031002044678,
      "epoch": 0.7860262008733624,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004946789923432695,
      "loss": 5.3579,
      "mean_token_accuracy": 0.1544201672077179,
      "num_tokens": 17353055.0,
      "step": 7920
    },
    {
      "entropy": 5.4604369640350345,
      "epoch": 0.78652242953553,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004946713296468322,
      "loss": 5.2597,
      "mean_token_accuracy": 0.1598164975643158,
      "num_tokens": 17364442.0,
      "step": 7925
    },
    {
      "entropy": 5.5707447052001955,
      "epoch": 0.7870186581976975,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004946636615030192,
      "loss": 5.3883,
      "mean_token_accuracy": 0.15719339847564698,
      "num_tokens": 17376237.0,
      "step": 7930
    },
    {
      "entropy": 5.5942220211029055,
      "epoch": 0.787514886859865,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004946559879120208,
      "loss": 5.31,
      "mean_token_accuracy": 0.1578843578696251,
      "num_tokens": 17386188.0,
      "step": 7935
    },
    {
      "entropy": 5.489248991012573,
      "epoch": 0.7880111155220325,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004946483088740273,
      "loss": 5.3107,
      "mean_token_accuracy": 0.16017146706581115,
      "num_tokens": 17397058.0,
      "step": 7940
    },
    {
      "entropy": 5.601967000961304,
      "epoch": 0.7885073441842001,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004946406243892289,
      "loss": 5.3578,
      "mean_token_accuracy": 0.15778404772281646,
      "num_tokens": 17408191.0,
      "step": 7945
    },
    {
      "entropy": 5.567347288131714,
      "epoch": 0.7890035728463676,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004946329344578163,
      "loss": 5.322,
      "mean_token_accuracy": 0.15752597898244858,
      "num_tokens": 17418472.0,
      "step": 7950
    },
    {
      "entropy": 5.563692188262939,
      "epoch": 0.7894998015085352,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004946252390799803,
      "loss": 5.2642,
      "mean_token_accuracy": 0.15932658165693284,
      "num_tokens": 17429556.0,
      "step": 7955
    },
    {
      "entropy": 5.586987924575806,
      "epoch": 0.7899960301707026,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004946175382559116,
      "loss": 5.4373,
      "mean_token_accuracy": 0.14806602373719216,
      "num_tokens": 17440950.0,
      "step": 7960
    },
    {
      "entropy": 5.503143644332885,
      "epoch": 0.7904922588328702,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000494609831985801,
      "loss": 5.3143,
      "mean_token_accuracy": 0.16252274811267853,
      "num_tokens": 17453724.0,
      "step": 7965
    },
    {
      "entropy": 5.572233009338379,
      "epoch": 0.7909884874950377,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00049460212026984,
      "loss": 5.3442,
      "mean_token_accuracy": 0.1571431815624237,
      "num_tokens": 17464623.0,
      "step": 7970
    },
    {
      "entropy": 5.528927183151245,
      "epoch": 0.7914847161572053,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004945944031082195,
      "loss": 5.3417,
      "mean_token_accuracy": 0.15848769694566728,
      "num_tokens": 17474919.0,
      "step": 7975
    },
    {
      "entropy": 5.589403867721558,
      "epoch": 0.7919809448193728,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000494586680501131,
      "loss": 5.3415,
      "mean_token_accuracy": 0.1526467278599739,
      "num_tokens": 17484747.0,
      "step": 7980
    },
    {
      "entropy": 5.541376209259033,
      "epoch": 0.7924771734815403,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000494578952448766,
      "loss": 5.2792,
      "mean_token_accuracy": 0.15907120555639268,
      "num_tokens": 17496299.0,
      "step": 7985
    },
    {
      "entropy": 5.604167509078979,
      "epoch": 0.7929734021437078,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004945712189513161,
      "loss": 5.441,
      "mean_token_accuracy": 0.14787812530994415,
      "num_tokens": 17507448.0,
      "step": 7990
    },
    {
      "entropy": 5.628121995925904,
      "epoch": 0.7934696308058753,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004945634800089733,
      "loss": 5.4042,
      "mean_token_accuracy": 0.15441112518310546,
      "num_tokens": 17520112.0,
      "step": 7995
    },
    {
      "entropy": 5.613991737365723,
      "epoch": 0.7939658594680429,
      "grad_norm": 0.875,
      "learning_rate": 0.0004945557356219291,
      "loss": 5.4044,
      "mean_token_accuracy": 0.14858510717749596,
      "num_tokens": 17531368.0,
      "step": 8000
    },
    {
      "entropy": 5.536077356338501,
      "epoch": 0.7944620881302104,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004945479857903758,
      "loss": 5.3431,
      "mean_token_accuracy": 0.15966367572546006,
      "num_tokens": 17542082.0,
      "step": 8005
    },
    {
      "entropy": 5.62625732421875,
      "epoch": 0.7949583167923779,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004945402305145056,
      "loss": 5.38,
      "mean_token_accuracy": 0.15365598797798158,
      "num_tokens": 17555518.0,
      "step": 8010
    },
    {
      "entropy": 5.577501440048218,
      "epoch": 0.7954545454545454,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004945324697945107,
      "loss": 5.3423,
      "mean_token_accuracy": 0.1595192164182663,
      "num_tokens": 17566110.0,
      "step": 8015
    },
    {
      "entropy": 5.473816680908203,
      "epoch": 0.795950774116713,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004945247036305836,
      "loss": 5.2663,
      "mean_token_accuracy": 0.1575251117348671,
      "num_tokens": 17575942.0,
      "step": 8020
    },
    {
      "entropy": 5.535625362396241,
      "epoch": 0.7964470027788805,
      "grad_norm": 1.0,
      "learning_rate": 0.0004945169320229169,
      "loss": 5.3364,
      "mean_token_accuracy": 0.162310454249382,
      "num_tokens": 17586001.0,
      "step": 8025
    },
    {
      "entropy": 5.641378355026245,
      "epoch": 0.7969432314410481,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004945091549717033,
      "loss": 5.4955,
      "mean_token_accuracy": 0.14408404529094695,
      "num_tokens": 17596774.0,
      "step": 8030
    },
    {
      "entropy": 5.5768969535827635,
      "epoch": 0.7974394601032155,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004945013724771356,
      "loss": 5.344,
      "mean_token_accuracy": 0.15899086594581605,
      "num_tokens": 17606557.0,
      "step": 8035
    },
    {
      "entropy": 5.46403317451477,
      "epoch": 0.7979356887653831,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004944935845394069,
      "loss": 5.2091,
      "mean_token_accuracy": 0.16562122851610184,
      "num_tokens": 17616465.0,
      "step": 8040
    },
    {
      "entropy": 5.579554080963135,
      "epoch": 0.7984319174275506,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004944857911587103,
      "loss": 5.3864,
      "mean_token_accuracy": 0.15155401676893235,
      "num_tokens": 17626907.0,
      "step": 8045
    },
    {
      "entropy": 5.6786962985992435,
      "epoch": 0.7989281460897182,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004944779923352389,
      "loss": 5.3975,
      "mean_token_accuracy": 0.15766018629074097,
      "num_tokens": 17638660.0,
      "step": 8050
    },
    {
      "entropy": 5.517003345489502,
      "epoch": 0.7994243747518857,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004944701880691864,
      "loss": 5.3138,
      "mean_token_accuracy": 0.15925874710083007,
      "num_tokens": 17650135.0,
      "step": 8055
    },
    {
      "entropy": 5.460254287719726,
      "epoch": 0.7999206034140532,
      "grad_norm": 0.90625,
      "learning_rate": 0.000494462378360746,
      "loss": 5.3311,
      "mean_token_accuracy": 0.1636511966586113,
      "num_tokens": 17660987.0,
      "step": 8060
    },
    {
      "entropy": 5.543812704086304,
      "epoch": 0.8004168320762207,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004944545632101117,
      "loss": 5.2574,
      "mean_token_accuracy": 0.15987827330827714,
      "num_tokens": 17672520.0,
      "step": 8065
    },
    {
      "entropy": 5.575038909912109,
      "epoch": 0.8009130607383882,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004944467426174769,
      "loss": 5.3339,
      "mean_token_accuracy": 0.1598665416240692,
      "num_tokens": 17683688.0,
      "step": 8070
    },
    {
      "entropy": 5.50664210319519,
      "epoch": 0.8014092894005558,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000494438916583036,
      "loss": 5.2699,
      "mean_token_accuracy": 0.16346991956233978,
      "num_tokens": 17695213.0,
      "step": 8075
    },
    {
      "entropy": 5.575667428970337,
      "epoch": 0.8019055180627233,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004944310851069826,
      "loss": 5.3486,
      "mean_token_accuracy": 0.15848584920167924,
      "num_tokens": 17705545.0,
      "step": 8080
    },
    {
      "entropy": 5.558717441558838,
      "epoch": 0.8024017467248908,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004944232481895112,
      "loss": 5.3209,
      "mean_token_accuracy": 0.1664901480078697,
      "num_tokens": 17715855.0,
      "step": 8085
    },
    {
      "entropy": 5.614819765090942,
      "epoch": 0.8028979753870583,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000494415405830816,
      "loss": 5.4158,
      "mean_token_accuracy": 0.15779888331890107,
      "num_tokens": 17727764.0,
      "step": 8090
    },
    {
      "entropy": 5.56912522315979,
      "epoch": 0.8033942040492259,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004944075580310916,
      "loss": 5.398,
      "mean_token_accuracy": 0.15286846309900284,
      "num_tokens": 17738464.0,
      "step": 8095
    },
    {
      "entropy": 5.577541494369507,
      "epoch": 0.8038904327113934,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004943997047905325,
      "loss": 5.3703,
      "mean_token_accuracy": 0.1655590355396271,
      "num_tokens": 17747502.0,
      "step": 8100
    },
    {
      "entropy": 5.551443243026734,
      "epoch": 0.804386661373561,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004943918461093335,
      "loss": 5.367,
      "mean_token_accuracy": 0.1538964331150055,
      "num_tokens": 17759520.0,
      "step": 8105
    },
    {
      "entropy": 5.5665810108184814,
      "epoch": 0.8048828900357284,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004943839819876894,
      "loss": 5.3329,
      "mean_token_accuracy": 0.1593322053551674,
      "num_tokens": 17770631.0,
      "step": 8110
    },
    {
      "entropy": 5.602841758728028,
      "epoch": 0.805379118697896,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004943761124257954,
      "loss": 5.2946,
      "mean_token_accuracy": 0.16228071600198746,
      "num_tokens": 17781320.0,
      "step": 8115
    },
    {
      "entropy": 5.590414524078369,
      "epoch": 0.8058753473600635,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004943682374238464,
      "loss": 5.3649,
      "mean_token_accuracy": 0.15804662331938743,
      "num_tokens": 17792627.0,
      "step": 8120
    },
    {
      "entropy": 5.540787553787231,
      "epoch": 0.8063715760222311,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000494360356982038,
      "loss": 5.406,
      "mean_token_accuracy": 0.15668713301420212,
      "num_tokens": 17803286.0,
      "step": 8125
    },
    {
      "entropy": 5.480075263977051,
      "epoch": 0.8068678046843986,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004943524711005653,
      "loss": 5.2015,
      "mean_token_accuracy": 0.16303706616163255,
      "num_tokens": 17813631.0,
      "step": 8130
    },
    {
      "entropy": 5.504738998413086,
      "epoch": 0.8073640333465661,
      "grad_norm": 0.953125,
      "learning_rate": 0.000494344579779624,
      "loss": 5.3063,
      "mean_token_accuracy": 0.16077257096767425,
      "num_tokens": 17823467.0,
      "step": 8135
    },
    {
      "entropy": 5.5465771675109865,
      "epoch": 0.8078602620087336,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004943366830194098,
      "loss": 5.307,
      "mean_token_accuracy": 0.16680953353643418,
      "num_tokens": 17833600.0,
      "step": 8140
    },
    {
      "entropy": 5.558188152313233,
      "epoch": 0.8083564906709012,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004943287808201183,
      "loss": 5.2986,
      "mean_token_accuracy": 0.15690502524375916,
      "num_tokens": 17843112.0,
      "step": 8145
    },
    {
      "entropy": 5.561684846878052,
      "epoch": 0.8088527193330687,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004943208731819459,
      "loss": 5.3072,
      "mean_token_accuracy": 0.15987997353076935,
      "num_tokens": 17853678.0,
      "step": 8150
    },
    {
      "entropy": 5.5255622386932375,
      "epoch": 0.8093489479952362,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004943129601050883,
      "loss": 5.2885,
      "mean_token_accuracy": 0.1580919548869133,
      "num_tokens": 17863537.0,
      "step": 8155
    },
    {
      "entropy": 5.550281381607055,
      "epoch": 0.8098451766574037,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000494305041589742,
      "loss": 5.3723,
      "mean_token_accuracy": 0.15576187670230865,
      "num_tokens": 17873511.0,
      "step": 8160
    },
    {
      "entropy": 5.533975887298584,
      "epoch": 0.8103414053195712,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004942971176361031,
      "loss": 5.2837,
      "mean_token_accuracy": 0.16323758363723756,
      "num_tokens": 17884546.0,
      "step": 8165
    },
    {
      "entropy": 5.526020526885986,
      "epoch": 0.8108376339817388,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004942891882443682,
      "loss": 5.2443,
      "mean_token_accuracy": 0.16286058723926544,
      "num_tokens": 17895331.0,
      "step": 8170
    },
    {
      "entropy": 5.571474981307984,
      "epoch": 0.8113338626439063,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000494281253414734,
      "loss": 5.4073,
      "mean_token_accuracy": 0.15418355911970139,
      "num_tokens": 17906465.0,
      "step": 8175
    },
    {
      "entropy": 5.532446432113647,
      "epoch": 0.8118300913060739,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004942733131473973,
      "loss": 5.3441,
      "mean_token_accuracy": 0.15361143350601197,
      "num_tokens": 17916936.0,
      "step": 8180
    },
    {
      "entropy": 5.652284479141235,
      "epoch": 0.8123263199682413,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004942653674425548,
      "loss": 5.3573,
      "mean_token_accuracy": 0.1522192656993866,
      "num_tokens": 17927522.0,
      "step": 8185
    },
    {
      "entropy": 5.616598796844483,
      "epoch": 0.8128225486304089,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004942574163004038,
      "loss": 5.4177,
      "mean_token_accuracy": 0.15336811244487764,
      "num_tokens": 17938216.0,
      "step": 8190
    },
    {
      "entropy": 5.556334447860718,
      "epoch": 0.8133187772925764,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004942494597211412,
      "loss": 5.3068,
      "mean_token_accuracy": 0.15798144042491913,
      "num_tokens": 17949333.0,
      "step": 8195
    },
    {
      "entropy": 5.6230260848999025,
      "epoch": 0.813815005954744,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004942414977049646,
      "loss": 5.3813,
      "mean_token_accuracy": 0.14451824128627777,
      "num_tokens": 17960178.0,
      "step": 8200
    },
    {
      "entropy": 5.552642488479615,
      "epoch": 0.8143112346169115,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000494233530252071,
      "loss": 5.3371,
      "mean_token_accuracy": 0.15979109555482865,
      "num_tokens": 17971180.0,
      "step": 8205
    },
    {
      "entropy": 5.567732620239258,
      "epoch": 0.814807463279079,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004942255573626585,
      "loss": 5.3655,
      "mean_token_accuracy": 0.1580875739455223,
      "num_tokens": 17982572.0,
      "step": 8210
    },
    {
      "entropy": 5.600094747543335,
      "epoch": 0.8153036919412465,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004942175790369244,
      "loss": 5.2915,
      "mean_token_accuracy": 0.16242016702890397,
      "num_tokens": 17993236.0,
      "step": 8215
    },
    {
      "entropy": 5.568907976150513,
      "epoch": 0.8157999206034141,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004942095952750669,
      "loss": 5.4123,
      "mean_token_accuracy": 0.15650524497032164,
      "num_tokens": 18004215.0,
      "step": 8220
    },
    {
      "entropy": 5.543283700942993,
      "epoch": 0.8162961492655816,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004942016060772837,
      "loss": 5.3112,
      "mean_token_accuracy": 0.16373128592967987,
      "num_tokens": 18013880.0,
      "step": 8225
    },
    {
      "entropy": 5.578078413009644,
      "epoch": 0.8167923779277491,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004941936114437731,
      "loss": 5.3558,
      "mean_token_accuracy": 0.1561330273747444,
      "num_tokens": 18024928.0,
      "step": 8230
    },
    {
      "entropy": 5.560678339004516,
      "epoch": 0.8172886065899166,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004941856113747331,
      "loss": 5.3184,
      "mean_token_accuracy": 0.15780884623527527,
      "num_tokens": 18035857.0,
      "step": 8235
    },
    {
      "entropy": 5.629302167892456,
      "epoch": 0.8177848352520841,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004941776058703623,
      "loss": 5.4277,
      "mean_token_accuracy": 0.15591926872730255,
      "num_tokens": 18046776.0,
      "step": 8240
    },
    {
      "entropy": 5.582234573364258,
      "epoch": 0.8182810639142517,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004941695949308593,
      "loss": 5.3828,
      "mean_token_accuracy": 0.16022476106882094,
      "num_tokens": 18057167.0,
      "step": 8245
    },
    {
      "entropy": 5.578600931167602,
      "epoch": 0.8187772925764192,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004941615785564224,
      "loss": 5.3976,
      "mean_token_accuracy": 0.15019726008176804,
      "num_tokens": 18066953.0,
      "step": 8250
    },
    {
      "entropy": 5.55739197731018,
      "epoch": 0.8192735212385868,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004941535567472506,
      "loss": 5.2314,
      "mean_token_accuracy": 0.16082544922828673,
      "num_tokens": 18077977.0,
      "step": 8255
    },
    {
      "entropy": 5.53443341255188,
      "epoch": 0.8197697499007542,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000494145529503543,
      "loss": 5.3292,
      "mean_token_accuracy": 0.15666717290878296,
      "num_tokens": 18087581.0,
      "step": 8260
    },
    {
      "entropy": 5.475851774215698,
      "epoch": 0.8202659785629218,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004941374968254985,
      "loss": 5.2356,
      "mean_token_accuracy": 0.1597230166196823,
      "num_tokens": 18098193.0,
      "step": 8265
    },
    {
      "entropy": 5.548586225509643,
      "epoch": 0.8207622072250893,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004941294587133161,
      "loss": 5.3556,
      "mean_token_accuracy": 0.15848664790391923,
      "num_tokens": 18110830.0,
      "step": 8270
    },
    {
      "entropy": 5.597237873077392,
      "epoch": 0.8212584358872569,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004941214151671954,
      "loss": 5.3835,
      "mean_token_accuracy": 0.15455053448677064,
      "num_tokens": 18122073.0,
      "step": 8275
    },
    {
      "entropy": 5.521269798278809,
      "epoch": 0.8217546645494244,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004941133661873357,
      "loss": 5.3386,
      "mean_token_accuracy": 0.15720982998609542,
      "num_tokens": 18132777.0,
      "step": 8280
    },
    {
      "entropy": 5.478314447402954,
      "epoch": 0.822250893211592,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004941053117739367,
      "loss": 5.2657,
      "mean_token_accuracy": 0.15625037103891373,
      "num_tokens": 18143232.0,
      "step": 8285
    },
    {
      "entropy": 5.552276754379273,
      "epoch": 0.8227471218737594,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004940972519271982,
      "loss": 5.4458,
      "mean_token_accuracy": 0.15443039685487747,
      "num_tokens": 18155225.0,
      "step": 8290
    },
    {
      "entropy": 5.566086196899414,
      "epoch": 0.823243350535927,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004940891866473199,
      "loss": 5.3689,
      "mean_token_accuracy": 0.15550159364938737,
      "num_tokens": 18165945.0,
      "step": 8295
    },
    {
      "entropy": 5.621346950531006,
      "epoch": 0.8237395791980945,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004940811159345017,
      "loss": 5.4121,
      "mean_token_accuracy": 0.1536681219935417,
      "num_tokens": 18177271.0,
      "step": 8300
    },
    {
      "entropy": 5.600881147384643,
      "epoch": 0.824235807860262,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004940730397889441,
      "loss": 5.3436,
      "mean_token_accuracy": 0.15792988687753678,
      "num_tokens": 18189732.0,
      "step": 8305
    },
    {
      "entropy": 5.562526035308838,
      "epoch": 0.8247320365224295,
      "grad_norm": 0.875,
      "learning_rate": 0.0004940649582108472,
      "loss": 5.4431,
      "mean_token_accuracy": 0.15375415980815887,
      "num_tokens": 18200731.0,
      "step": 8310
    },
    {
      "entropy": 5.58730411529541,
      "epoch": 0.825228265184597,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004940568712004113,
      "loss": 5.3553,
      "mean_token_accuracy": 0.15794347524642943,
      "num_tokens": 18210873.0,
      "step": 8315
    },
    {
      "entropy": 5.621802616119385,
      "epoch": 0.8257244938467646,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004940487787578369,
      "loss": 5.3582,
      "mean_token_accuracy": 0.15885585248470308,
      "num_tokens": 18222072.0,
      "step": 8320
    },
    {
      "entropy": 5.510449028015136,
      "epoch": 0.8262207225089321,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000494040680883325,
      "loss": 5.2685,
      "mean_token_accuracy": 0.15838711857795715,
      "num_tokens": 18232522.0,
      "step": 8325
    },
    {
      "entropy": 5.585622549057007,
      "epoch": 0.8267169511710997,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004940325775770762,
      "loss": 5.3562,
      "mean_token_accuracy": 0.152874194085598,
      "num_tokens": 18243587.0,
      "step": 8330
    },
    {
      "entropy": 5.499403047561645,
      "epoch": 0.8272131798332671,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004940244688392913,
      "loss": 5.2735,
      "mean_token_accuracy": 0.16457149386405945,
      "num_tokens": 18254965.0,
      "step": 8335
    },
    {
      "entropy": 5.526314306259155,
      "epoch": 0.8277094084954347,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004940163546701716,
      "loss": 5.304,
      "mean_token_accuracy": 0.1621924877166748,
      "num_tokens": 18265892.0,
      "step": 8340
    },
    {
      "entropy": 5.533067941665649,
      "epoch": 0.8282056371576022,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004940082350699183,
      "loss": 5.3054,
      "mean_token_accuracy": 0.1597759246826172,
      "num_tokens": 18276713.0,
      "step": 8345
    },
    {
      "entropy": 5.6015809059143065,
      "epoch": 0.8287018658197698,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004940001100387325,
      "loss": 5.3487,
      "mean_token_accuracy": 0.15081969648599625,
      "num_tokens": 18288846.0,
      "step": 8350
    },
    {
      "entropy": 5.594089126586914,
      "epoch": 0.8291980944819373,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004939919795768161,
      "loss": 5.3696,
      "mean_token_accuracy": 0.16012121886014938,
      "num_tokens": 18298750.0,
      "step": 8355
    },
    {
      "entropy": 5.524968576431275,
      "epoch": 0.8296943231441049,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004939838436843704,
      "loss": 5.3483,
      "mean_token_accuracy": 0.15921227037906646,
      "num_tokens": 18310065.0,
      "step": 8360
    },
    {
      "entropy": 5.545454597473144,
      "epoch": 0.8301905518062723,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004939757023615973,
      "loss": 5.2898,
      "mean_token_accuracy": 0.15805892199277877,
      "num_tokens": 18319498.0,
      "step": 8365
    },
    {
      "entropy": 5.4899983406066895,
      "epoch": 0.8306867804684399,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004939675556086986,
      "loss": 5.2651,
      "mean_token_accuracy": 0.16524300426244737,
      "num_tokens": 18329718.0,
      "step": 8370
    },
    {
      "entropy": 5.522244691848755,
      "epoch": 0.8311830091306074,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004939594034258763,
      "loss": 5.2852,
      "mean_token_accuracy": 0.16079603731632233,
      "num_tokens": 18341628.0,
      "step": 8375
    },
    {
      "entropy": 5.5233237743377686,
      "epoch": 0.8316792377927749,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004939512458133326,
      "loss": 5.3232,
      "mean_token_accuracy": 0.1642431765794754,
      "num_tokens": 18352869.0,
      "step": 8380
    },
    {
      "entropy": 5.534032392501831,
      "epoch": 0.8321754664549424,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004939430827712698,
      "loss": 5.3238,
      "mean_token_accuracy": 0.15860947966575623,
      "num_tokens": 18363664.0,
      "step": 8385
    },
    {
      "entropy": 5.594780349731446,
      "epoch": 0.8326716951171099,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004939349142998905,
      "loss": 5.3113,
      "mean_token_accuracy": 0.15810291171073915,
      "num_tokens": 18374325.0,
      "step": 8390
    },
    {
      "entropy": 5.5217423915863035,
      "epoch": 0.8331679237792775,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004939267403993968,
      "loss": 5.2966,
      "mean_token_accuracy": 0.1612369954586029,
      "num_tokens": 18384547.0,
      "step": 8395
    },
    {
      "entropy": 5.551669120788574,
      "epoch": 0.833664152441445,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000493918561069992,
      "loss": 5.2874,
      "mean_token_accuracy": 0.16169873923063277,
      "num_tokens": 18395141.0,
      "step": 8400
    },
    {
      "entropy": 5.522643756866455,
      "epoch": 0.8341603811036126,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004939103763118784,
      "loss": 5.2917,
      "mean_token_accuracy": 0.15762884318828582,
      "num_tokens": 18406326.0,
      "step": 8405
    },
    {
      "entropy": 5.537695217132568,
      "epoch": 0.83465660976578,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004939021861252593,
      "loss": 5.3415,
      "mean_token_accuracy": 0.16229671239852905,
      "num_tokens": 18417521.0,
      "step": 8410
    },
    {
      "entropy": 5.561509704589843,
      "epoch": 0.8351528384279476,
      "grad_norm": 1.0,
      "learning_rate": 0.0004938939905103376,
      "loss": 5.277,
      "mean_token_accuracy": 0.15979011803865434,
      "num_tokens": 18426813.0,
      "step": 8415
    },
    {
      "entropy": 5.5146078109741214,
      "epoch": 0.8356490670901151,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004938857894673166,
      "loss": 5.2444,
      "mean_token_accuracy": 0.1613193228840828,
      "num_tokens": 18438227.0,
      "step": 8420
    },
    {
      "entropy": 5.525485038757324,
      "epoch": 0.8361452957522827,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004938775829963997,
      "loss": 5.3727,
      "mean_token_accuracy": 0.15905213952064515,
      "num_tokens": 18448915.0,
      "step": 8425
    },
    {
      "entropy": 5.57030463218689,
      "epoch": 0.8366415244144502,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004938693710977904,
      "loss": 5.2672,
      "mean_token_accuracy": 0.16031362116336823,
      "num_tokens": 18459330.0,
      "step": 8430
    },
    {
      "entropy": 5.532431697845459,
      "epoch": 0.8371377530766178,
      "grad_norm": 0.875,
      "learning_rate": 0.0004938611537716922,
      "loss": 5.2386,
      "mean_token_accuracy": 0.16233667135238647,
      "num_tokens": 18469813.0,
      "step": 8435
    },
    {
      "entropy": 5.490002584457398,
      "epoch": 0.8376339817387852,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004938529310183091,
      "loss": 5.2947,
      "mean_token_accuracy": 0.156914222240448,
      "num_tokens": 18479783.0,
      "step": 8440
    },
    {
      "entropy": 5.510782670974732,
      "epoch": 0.8381302104009528,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000493844702837845,
      "loss": 5.3036,
      "mean_token_accuracy": 0.15965810865163804,
      "num_tokens": 18489640.0,
      "step": 8445
    },
    {
      "entropy": 5.560747814178467,
      "epoch": 0.8386264390631203,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004938364692305036,
      "loss": 5.3664,
      "mean_token_accuracy": 0.1613190710544586,
      "num_tokens": 18500744.0,
      "step": 8450
    },
    {
      "entropy": 5.55366849899292,
      "epoch": 0.8391226677252878,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004938282301964896,
      "loss": 5.2796,
      "mean_token_accuracy": 0.16003650277853013,
      "num_tokens": 18513286.0,
      "step": 8455
    },
    {
      "entropy": 5.599881029129028,
      "epoch": 0.8396188963874553,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004938199857360068,
      "loss": 5.2563,
      "mean_token_accuracy": 0.16317568719387054,
      "num_tokens": 18522934.0,
      "step": 8460
    },
    {
      "entropy": 5.523234224319458,
      "epoch": 0.8401151250496228,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00049381173584926,
      "loss": 5.3165,
      "mean_token_accuracy": 0.16004646569490433,
      "num_tokens": 18534828.0,
      "step": 8465
    },
    {
      "entropy": 5.494680166244507,
      "epoch": 0.8406113537117904,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004938034805364535,
      "loss": 5.2605,
      "mean_token_accuracy": 0.16069284826517105,
      "num_tokens": 18544733.0,
      "step": 8470
    },
    {
      "entropy": 5.531710147857666,
      "epoch": 0.8411075823739579,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004937952197977923,
      "loss": 5.3305,
      "mean_token_accuracy": 0.15478749722242355,
      "num_tokens": 18554583.0,
      "step": 8475
    },
    {
      "entropy": 5.6344075202941895,
      "epoch": 0.8416038110361255,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000493786953633481,
      "loss": 5.3668,
      "mean_token_accuracy": 0.15707846879959106,
      "num_tokens": 18566503.0,
      "step": 8480
    },
    {
      "entropy": 5.5227796077728275,
      "epoch": 0.842100039698293,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004937786820437248,
      "loss": 5.3151,
      "mean_token_accuracy": 0.1600247323513031,
      "num_tokens": 18578161.0,
      "step": 8485
    },
    {
      "entropy": 5.525348567962647,
      "epoch": 0.8425962683604605,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004937704050287286,
      "loss": 5.3631,
      "mean_token_accuracy": 0.16227232664823532,
      "num_tokens": 18588212.0,
      "step": 8490
    },
    {
      "entropy": 5.668597173690796,
      "epoch": 0.843092497022628,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004937621225886978,
      "loss": 5.3781,
      "mean_token_accuracy": 0.15517351925373077,
      "num_tokens": 18601293.0,
      "step": 8495
    },
    {
      "entropy": 5.592793798446655,
      "epoch": 0.8435887256847956,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004937538347238378,
      "loss": 5.4242,
      "mean_token_accuracy": 0.15348978191614152,
      "num_tokens": 18613316.0,
      "step": 8500
    },
    {
      "entropy": 5.537801313400268,
      "epoch": 0.8440849543469631,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000493745541434354,
      "loss": 5.3559,
      "mean_token_accuracy": 0.1587686210870743,
      "num_tokens": 18624461.0,
      "step": 8505
    },
    {
      "entropy": 5.5556048393249515,
      "epoch": 0.8445811830091307,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004937372427204522,
      "loss": 5.2578,
      "mean_token_accuracy": 0.16012921929359436,
      "num_tokens": 18635165.0,
      "step": 8510
    },
    {
      "entropy": 5.566375350952148,
      "epoch": 0.8450774116712981,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004937289385823382,
      "loss": 5.3067,
      "mean_token_accuracy": 0.16270599886775017,
      "num_tokens": 18646649.0,
      "step": 8515
    },
    {
      "entropy": 5.5332996368408205,
      "epoch": 0.8455736403334657,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004937206290202176,
      "loss": 5.2953,
      "mean_token_accuracy": 0.15720799714326858,
      "num_tokens": 18658527.0,
      "step": 8520
    },
    {
      "entropy": 5.535870552062988,
      "epoch": 0.8460698689956332,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004937123140342969,
      "loss": 5.2992,
      "mean_token_accuracy": 0.15728411376476287,
      "num_tokens": 18669463.0,
      "step": 8525
    },
    {
      "entropy": 5.582309579849243,
      "epoch": 0.8465660976578007,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000493703993624782,
      "loss": 5.3422,
      "mean_token_accuracy": 0.1569891206920147,
      "num_tokens": 18680806.0,
      "step": 8530
    },
    {
      "entropy": 5.553014659881592,
      "epoch": 0.8470623263199683,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004936956677918794,
      "loss": 5.325,
      "mean_token_accuracy": 0.15743035674095154,
      "num_tokens": 18692124.0,
      "step": 8535
    },
    {
      "entropy": 5.593314981460571,
      "epoch": 0.8475585549821357,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004936873365357955,
      "loss": 5.3591,
      "mean_token_accuracy": 0.15825066715478897,
      "num_tokens": 18702017.0,
      "step": 8540
    },
    {
      "entropy": 5.466948652267456,
      "epoch": 0.8480547836443033,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004936789998567368,
      "loss": 5.2504,
      "mean_token_accuracy": 0.1651918262243271,
      "num_tokens": 18713205.0,
      "step": 8545
    },
    {
      "entropy": 5.537358236312866,
      "epoch": 0.8485510123064708,
      "grad_norm": 0.875,
      "learning_rate": 0.0004936706577549101,
      "loss": 5.2623,
      "mean_token_accuracy": 0.1572915345430374,
      "num_tokens": 18724165.0,
      "step": 8550
    },
    {
      "entropy": 5.533035326004028,
      "epoch": 0.8490472409686384,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004936623102305223,
      "loss": 5.3234,
      "mean_token_accuracy": 0.1630731612443924,
      "num_tokens": 18735155.0,
      "step": 8555
    },
    {
      "entropy": 5.5569404602050785,
      "epoch": 0.8495434696308058,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004936539572837804,
      "loss": 5.4208,
      "mean_token_accuracy": 0.15740315914154052,
      "num_tokens": 18746228.0,
      "step": 8560
    },
    {
      "entropy": 5.556674766540527,
      "epoch": 0.8500396982929734,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004936455989148916,
      "loss": 5.3713,
      "mean_token_accuracy": 0.15661710053682326,
      "num_tokens": 18756638.0,
      "step": 8565
    },
    {
      "entropy": 5.604758405685425,
      "epoch": 0.8505359269551409,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004936372351240629,
      "loss": 5.3341,
      "mean_token_accuracy": 0.16325204819440842,
      "num_tokens": 18768651.0,
      "step": 8570
    },
    {
      "entropy": 5.6614268779754635,
      "epoch": 0.8510321556173085,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000493628865911502,
      "loss": 5.4427,
      "mean_token_accuracy": 0.15434404462575912,
      "num_tokens": 18780399.0,
      "step": 8575
    },
    {
      "entropy": 5.519382762908935,
      "epoch": 0.851528384279476,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004936204912774163,
      "loss": 5.3044,
      "mean_token_accuracy": 0.1619669407606125,
      "num_tokens": 18791927.0,
      "step": 8580
    },
    {
      "entropy": 5.490979290008545,
      "epoch": 0.8520246129416436,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004936121112220135,
      "loss": 5.2797,
      "mean_token_accuracy": 0.15910061895847322,
      "num_tokens": 18802779.0,
      "step": 8585
    },
    {
      "entropy": 5.4997974872589115,
      "epoch": 0.852520841603811,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004936037257455015,
      "loss": 5.2769,
      "mean_token_accuracy": 0.15892841964960097,
      "num_tokens": 18813123.0,
      "step": 8590
    },
    {
      "entropy": 5.527001857757568,
      "epoch": 0.8530170702659786,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000493595334848088,
      "loss": 5.2799,
      "mean_token_accuracy": 0.16557230204343795,
      "num_tokens": 18823699.0,
      "step": 8595
    },
    {
      "entropy": 5.614219713211059,
      "epoch": 0.8535132989281461,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004935869385299813,
      "loss": 5.4374,
      "mean_token_accuracy": 0.15188810378313064,
      "num_tokens": 18835450.0,
      "step": 8600
    },
    {
      "entropy": 5.550320196151733,
      "epoch": 0.8540095275903136,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004935785367913896,
      "loss": 5.2802,
      "mean_token_accuracy": 0.15528021305799483,
      "num_tokens": 18845919.0,
      "step": 8605
    },
    {
      "entropy": 5.5714820384979244,
      "epoch": 0.8545057562524812,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004935701296325212,
      "loss": 5.4018,
      "mean_token_accuracy": 0.15196267366409302,
      "num_tokens": 18858376.0,
      "step": 8610
    },
    {
      "entropy": 5.567236423492432,
      "epoch": 0.8550019849146486,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004935617170535845,
      "loss": 5.357,
      "mean_token_accuracy": 0.16111219227313994,
      "num_tokens": 18870456.0,
      "step": 8615
    },
    {
      "entropy": 5.539654159545899,
      "epoch": 0.8554982135768162,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004935532990547882,
      "loss": 5.3669,
      "mean_token_accuracy": 0.15522739589214324,
      "num_tokens": 18882174.0,
      "step": 8620
    },
    {
      "entropy": 5.640656852722168,
      "epoch": 0.8559944422389837,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004935448756363409,
      "loss": 5.4789,
      "mean_token_accuracy": 0.15537660121917723,
      "num_tokens": 18892982.0,
      "step": 8625
    },
    {
      "entropy": 5.610037469863892,
      "epoch": 0.8564906709011513,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004935364467984518,
      "loss": 5.2944,
      "mean_token_accuracy": 0.16325953602790833,
      "num_tokens": 18903557.0,
      "step": 8630
    },
    {
      "entropy": 5.615718507766724,
      "epoch": 0.8569868995633187,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004935280125413295,
      "loss": 5.4494,
      "mean_token_accuracy": 0.15439070388674736,
      "num_tokens": 18914912.0,
      "step": 8635
    },
    {
      "entropy": 5.599884176254273,
      "epoch": 0.8574831282254863,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004935195728651836,
      "loss": 5.3884,
      "mean_token_accuracy": 0.15769529342651367,
      "num_tokens": 18927773.0,
      "step": 8640
    },
    {
      "entropy": 5.562195873260498,
      "epoch": 0.8579793568876538,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000493511127770223,
      "loss": 5.3292,
      "mean_token_accuracy": 0.15460922122001647,
      "num_tokens": 18938313.0,
      "step": 8645
    },
    {
      "entropy": 5.537086153030396,
      "epoch": 0.8584755855498214,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004935026772566574,
      "loss": 5.2694,
      "mean_token_accuracy": 0.16621735841035842,
      "num_tokens": 18948600.0,
      "step": 8650
    },
    {
      "entropy": 5.6071208953857425,
      "epoch": 0.8589718142119889,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004934942213246961,
      "loss": 5.366,
      "mean_token_accuracy": 0.16027262061834335,
      "num_tokens": 18958927.0,
      "step": 8655
    },
    {
      "entropy": 5.541692781448364,
      "epoch": 0.8594680428741565,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000493485759974549,
      "loss": 5.3137,
      "mean_token_accuracy": 0.15310209542512893,
      "num_tokens": 18968625.0,
      "step": 8660
    },
    {
      "entropy": 5.471400356292724,
      "epoch": 0.8599642715363239,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004934772932064259,
      "loss": 5.2326,
      "mean_token_accuracy": 0.1595799967646599,
      "num_tokens": 18979772.0,
      "step": 8665
    },
    {
      "entropy": 5.564545679092407,
      "epoch": 0.8604605001984915,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004934688210205367,
      "loss": 5.3956,
      "mean_token_accuracy": 0.15682338774204255,
      "num_tokens": 18991522.0,
      "step": 8670
    },
    {
      "entropy": 5.575645589828492,
      "epoch": 0.860956728860659,
      "grad_norm": 0.875,
      "learning_rate": 0.0004934603434170913,
      "loss": 5.3541,
      "mean_token_accuracy": 0.1573867157101631,
      "num_tokens": 19002479.0,
      "step": 8675
    },
    {
      "entropy": 5.603612327575684,
      "epoch": 0.8614529575228265,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004934518603963003,
      "loss": 5.4179,
      "mean_token_accuracy": 0.15240964889526368,
      "num_tokens": 19013254.0,
      "step": 8680
    },
    {
      "entropy": 5.524136972427368,
      "epoch": 0.861949186184994,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004934433719583739,
      "loss": 5.2423,
      "mean_token_accuracy": 0.16270106732845308,
      "num_tokens": 19023943.0,
      "step": 8685
    },
    {
      "entropy": 5.619406223297119,
      "epoch": 0.8624454148471615,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004934348781035226,
      "loss": 5.3791,
      "mean_token_accuracy": 0.15710915327072145,
      "num_tokens": 19034672.0,
      "step": 8690
    },
    {
      "entropy": 5.626100206375122,
      "epoch": 0.8629416435093291,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000493426378831957,
      "loss": 5.3601,
      "mean_token_accuracy": 0.16364834159612657,
      "num_tokens": 19045758.0,
      "step": 8695
    },
    {
      "entropy": 5.546971464157105,
      "epoch": 0.8634378721714966,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004934178741438877,
      "loss": 5.3529,
      "mean_token_accuracy": 0.16413892656564713,
      "num_tokens": 19057526.0,
      "step": 8700
    },
    {
      "entropy": 5.49228720664978,
      "epoch": 0.8639341008336642,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000493409364039526,
      "loss": 5.2166,
      "mean_token_accuracy": 0.16375850439071654,
      "num_tokens": 19069657.0,
      "step": 8705
    },
    {
      "entropy": 5.535357713699341,
      "epoch": 0.8644303294958317,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004934008485190825,
      "loss": 5.2839,
      "mean_token_accuracy": 0.16336794346570968,
      "num_tokens": 19080699.0,
      "step": 8710
    },
    {
      "entropy": 5.536669921875,
      "epoch": 0.8649265581579992,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004933923275827687,
      "loss": 5.3222,
      "mean_token_accuracy": 0.15725933462381364,
      "num_tokens": 19093041.0,
      "step": 8715
    },
    {
      "entropy": 5.627826738357544,
      "epoch": 0.8654227868201667,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004933838012307957,
      "loss": 5.3765,
      "mean_token_accuracy": 0.1537148669362068,
      "num_tokens": 19104249.0,
      "step": 8720
    },
    {
      "entropy": 5.579020833969116,
      "epoch": 0.8659190154823343,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000493375269463375,
      "loss": 5.3417,
      "mean_token_accuracy": 0.1575908273458481,
      "num_tokens": 19114434.0,
      "step": 8725
    },
    {
      "entropy": 5.537295150756836,
      "epoch": 0.8664152441445018,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004933667322807182,
      "loss": 5.3558,
      "mean_token_accuracy": 0.15704239457845687,
      "num_tokens": 19124814.0,
      "step": 8730
    },
    {
      "entropy": 5.565311050415039,
      "epoch": 0.8669114728066694,
      "grad_norm": 0.9375,
      "learning_rate": 0.000493358189683037,
      "loss": 5.372,
      "mean_token_accuracy": 0.15405240654945374,
      "num_tokens": 19135205.0,
      "step": 8735
    },
    {
      "entropy": 5.5298460006713865,
      "epoch": 0.8674077014688368,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004933496416705432,
      "loss": 5.2593,
      "mean_token_accuracy": 0.1580391377210617,
      "num_tokens": 19147117.0,
      "step": 8740
    },
    {
      "entropy": 5.5313962459564205,
      "epoch": 0.8679039301310044,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004933410882434488,
      "loss": 5.3397,
      "mean_token_accuracy": 0.162195847928524,
      "num_tokens": 19158131.0,
      "step": 8745
    },
    {
      "entropy": 5.506332731246948,
      "epoch": 0.8684001587931719,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004933325294019659,
      "loss": 5.2187,
      "mean_token_accuracy": 0.1621017947793007,
      "num_tokens": 19168820.0,
      "step": 8750
    },
    {
      "entropy": 5.590275287628174,
      "epoch": 0.8688963874553394,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004933239651463068,
      "loss": 5.3073,
      "mean_token_accuracy": 0.16791141927242278,
      "num_tokens": 19178845.0,
      "step": 8755
    },
    {
      "entropy": 5.573048782348633,
      "epoch": 0.869392616117507,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004933153954766837,
      "loss": 5.2843,
      "mean_token_accuracy": 0.1613463819026947,
      "num_tokens": 19190830.0,
      "step": 8760
    },
    {
      "entropy": 5.404928827285767,
      "epoch": 0.8698888447796744,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004933068203933092,
      "loss": 5.144,
      "mean_token_accuracy": 0.1659684345126152,
      "num_tokens": 19202056.0,
      "step": 8765
    },
    {
      "entropy": 5.511019325256347,
      "epoch": 0.870385073441842,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000493298239896396,
      "loss": 5.3581,
      "mean_token_accuracy": 0.15750947892665862,
      "num_tokens": 19213497.0,
      "step": 8770
    },
    {
      "entropy": 5.633516693115235,
      "epoch": 0.8708813021040095,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004932896539861568,
      "loss": 5.3812,
      "mean_token_accuracy": 0.15700077563524245,
      "num_tokens": 19226103.0,
      "step": 8775
    },
    {
      "entropy": 5.575097703933716,
      "epoch": 0.8713775307661771,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004932810626628047,
      "loss": 5.4008,
      "mean_token_accuracy": 0.15823982506990433,
      "num_tokens": 19237297.0,
      "step": 8780
    },
    {
      "entropy": 5.556234884262085,
      "epoch": 0.8718737594283446,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004932724659265526,
      "loss": 5.3055,
      "mean_token_accuracy": 0.15963849276304246,
      "num_tokens": 19247537.0,
      "step": 8785
    },
    {
      "entropy": 5.48999514579773,
      "epoch": 0.8723699880905121,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004932638637776134,
      "loss": 5.2464,
      "mean_token_accuracy": 0.15756665170192719,
      "num_tokens": 19260465.0,
      "step": 8790
    },
    {
      "entropy": 5.549872970581054,
      "epoch": 0.8728662167526796,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000493255256216201,
      "loss": 5.3144,
      "mean_token_accuracy": 0.15558967739343643,
      "num_tokens": 19270534.0,
      "step": 8795
    },
    {
      "entropy": 5.5368547439575195,
      "epoch": 0.8733624454148472,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004932466432425284,
      "loss": 5.328,
      "mean_token_accuracy": 0.1593007780611515,
      "num_tokens": 19281551.0,
      "step": 8800
    },
    {
      "entropy": 5.53004264831543,
      "epoch": 0.8738586740770147,
      "grad_norm": 1.0,
      "learning_rate": 0.0004932380248568094,
      "loss": 5.2722,
      "mean_token_accuracy": 0.1617216259241104,
      "num_tokens": 19293300.0,
      "step": 8805
    },
    {
      "entropy": 5.522213697433472,
      "epoch": 0.8743549027391823,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004932294010592575,
      "loss": 5.2088,
      "mean_token_accuracy": 0.1621657654643059,
      "num_tokens": 19302701.0,
      "step": 8810
    },
    {
      "entropy": 5.561089992523193,
      "epoch": 0.8748511314013497,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004932207718500868,
      "loss": 5.3038,
      "mean_token_accuracy": 0.1600021854043007,
      "num_tokens": 19313024.0,
      "step": 8815
    },
    {
      "entropy": 5.554006576538086,
      "epoch": 0.8753473600635173,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004932121372295112,
      "loss": 5.3208,
      "mean_token_accuracy": 0.15621675699949264,
      "num_tokens": 19324014.0,
      "step": 8820
    },
    {
      "entropy": 5.5339195251464846,
      "epoch": 0.8758435887256848,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004932034971977448,
      "loss": 5.3327,
      "mean_token_accuracy": 0.15784318298101424,
      "num_tokens": 19334226.0,
      "step": 8825
    },
    {
      "entropy": 5.585725498199463,
      "epoch": 0.8763398173878523,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004931948517550019,
      "loss": 5.3665,
      "mean_token_accuracy": 0.15267175287008286,
      "num_tokens": 19344363.0,
      "step": 8830
    },
    {
      "entropy": 5.573243904113769,
      "epoch": 0.8768360460500199,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004931862009014968,
      "loss": 5.3827,
      "mean_token_accuracy": 0.1558567300438881,
      "num_tokens": 19354915.0,
      "step": 8835
    },
    {
      "entropy": 5.623068523406983,
      "epoch": 0.8773322747121873,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000493177544637444,
      "loss": 5.3872,
      "mean_token_accuracy": 0.15675173848867416,
      "num_tokens": 19365570.0,
      "step": 8840
    },
    {
      "entropy": 5.614332580566407,
      "epoch": 0.8778285033743549,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004931688829630582,
      "loss": 5.3009,
      "mean_token_accuracy": 0.16717157363891602,
      "num_tokens": 19375731.0,
      "step": 8845
    },
    {
      "entropy": 5.511403656005859,
      "epoch": 0.8783247320365224,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004931602158785544,
      "loss": 5.2702,
      "mean_token_accuracy": 0.16010398268699647,
      "num_tokens": 19386248.0,
      "step": 8850
    },
    {
      "entropy": 5.497075843811035,
      "epoch": 0.87882096069869,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004931515433841472,
      "loss": 5.2717,
      "mean_token_accuracy": 0.16560742855072022,
      "num_tokens": 19397593.0,
      "step": 8855
    },
    {
      "entropy": 5.558141374588013,
      "epoch": 0.8793171893608575,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004931428654800518,
      "loss": 5.3527,
      "mean_token_accuracy": 0.15407872051000596,
      "num_tokens": 19409701.0,
      "step": 8860
    },
    {
      "entropy": 5.589240264892578,
      "epoch": 0.879813418023025,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004931341821664834,
      "loss": 5.3276,
      "mean_token_accuracy": 0.15600220263004302,
      "num_tokens": 19420280.0,
      "step": 8865
    },
    {
      "entropy": 5.584590053558349,
      "epoch": 0.8803096466851925,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004931254934436573,
      "loss": 5.314,
      "mean_token_accuracy": 0.15992545783519746,
      "num_tokens": 19432208.0,
      "step": 8870
    },
    {
      "entropy": 5.6468058109283445,
      "epoch": 0.8808058753473601,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004931167993117891,
      "loss": 5.4488,
      "mean_token_accuracy": 0.15389225333929063,
      "num_tokens": 19443333.0,
      "step": 8875
    },
    {
      "entropy": 5.497191619873047,
      "epoch": 0.8813021040095276,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004931080997710942,
      "loss": 5.1827,
      "mean_token_accuracy": 0.16432267278432847,
      "num_tokens": 19452867.0,
      "step": 8880
    },
    {
      "entropy": 5.483661651611328,
      "epoch": 0.8817983326716952,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004930993948217884,
      "loss": 5.2241,
      "mean_token_accuracy": 0.1687967821955681,
      "num_tokens": 19463021.0,
      "step": 8885
    },
    {
      "entropy": 5.5666601181030275,
      "epoch": 0.8822945613338626,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004930906844640876,
      "loss": 5.3581,
      "mean_token_accuracy": 0.1564721181988716,
      "num_tokens": 19474092.0,
      "step": 8890
    },
    {
      "entropy": 5.524877643585205,
      "epoch": 0.8827907899960302,
      "grad_norm": 0.875,
      "learning_rate": 0.0004930819686982077,
      "loss": 5.3319,
      "mean_token_accuracy": 0.1533910483121872,
      "num_tokens": 19486035.0,
      "step": 8895
    },
    {
      "entropy": 5.497438240051269,
      "epoch": 0.8832870186581977,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000493073247524365,
      "loss": 5.2632,
      "mean_token_accuracy": 0.16072648912668228,
      "num_tokens": 19497484.0,
      "step": 8900
    },
    {
      "entropy": 5.568736743927002,
      "epoch": 0.8837832473203652,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004930645209427756,
      "loss": 5.3694,
      "mean_token_accuracy": 0.15094278901815414,
      "num_tokens": 19508463.0,
      "step": 8905
    },
    {
      "entropy": 5.570272445678711,
      "epoch": 0.8842794759825328,
      "grad_norm": 0.84375,
      "learning_rate": 0.000493055788953656,
      "loss": 5.3055,
      "mean_token_accuracy": 0.16139017641544343,
      "num_tokens": 19520260.0,
      "step": 8910
    },
    {
      "entropy": 5.536235666275024,
      "epoch": 0.8847757046447002,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004930470515572227,
      "loss": 5.2565,
      "mean_token_accuracy": 0.16148921102285385,
      "num_tokens": 19531621.0,
      "step": 8915
    },
    {
      "entropy": 5.542524147033691,
      "epoch": 0.8852719333068678,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004930383087536923,
      "loss": 5.2437,
      "mean_token_accuracy": 0.15747498124837875,
      "num_tokens": 19543271.0,
      "step": 8920
    },
    {
      "entropy": 5.6115631580352785,
      "epoch": 0.8857681619690353,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004930295605432818,
      "loss": 5.4049,
      "mean_token_accuracy": 0.16077397018671036,
      "num_tokens": 19555178.0,
      "step": 8925
    },
    {
      "entropy": 5.574784278869629,
      "epoch": 0.8862643906312029,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004930208069262079,
      "loss": 5.3257,
      "mean_token_accuracy": 0.16025338172912598,
      "num_tokens": 19565153.0,
      "step": 8930
    },
    {
      "entropy": 5.488508558273315,
      "epoch": 0.8867606192933704,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004930120479026878,
      "loss": 5.2733,
      "mean_token_accuracy": 0.16610471904277802,
      "num_tokens": 19575608.0,
      "step": 8935
    },
    {
      "entropy": 5.554893922805786,
      "epoch": 0.8872568479555379,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004930032834729388,
      "loss": 5.2273,
      "mean_token_accuracy": 0.16715442389249802,
      "num_tokens": 19587556.0,
      "step": 8940
    },
    {
      "entropy": 5.515234899520874,
      "epoch": 0.8877530766177054,
      "grad_norm": 0.890625,
      "learning_rate": 0.000492994513637178,
      "loss": 5.2688,
      "mean_token_accuracy": 0.16076616793870926,
      "num_tokens": 19597341.0,
      "step": 8945
    },
    {
      "entropy": 5.507076930999756,
      "epoch": 0.888249305279873,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004929857383956229,
      "loss": 5.2866,
      "mean_token_accuracy": 0.1587548226118088,
      "num_tokens": 19608431.0,
      "step": 8950
    },
    {
      "entropy": 5.525849533081055,
      "epoch": 0.8887455339420405,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004929769577484914,
      "loss": 5.2795,
      "mean_token_accuracy": 0.16742042303085328,
      "num_tokens": 19619645.0,
      "step": 8955
    },
    {
      "entropy": 5.509834575653076,
      "epoch": 0.8892417626042081,
      "grad_norm": 0.984375,
      "learning_rate": 0.000492968171696001,
      "loss": 5.2425,
      "mean_token_accuracy": 0.16633469015359878,
      "num_tokens": 19630412.0,
      "step": 8960
    },
    {
      "entropy": 5.550708866119384,
      "epoch": 0.8897379912663755,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004929593802383696,
      "loss": 5.2967,
      "mean_token_accuracy": 0.16355745047330855,
      "num_tokens": 19640055.0,
      "step": 8965
    },
    {
      "entropy": 5.504737377166748,
      "epoch": 0.8902342199285431,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004929505833758152,
      "loss": 5.3128,
      "mean_token_accuracy": 0.15629351809620856,
      "num_tokens": 19650814.0,
      "step": 8970
    },
    {
      "entropy": 5.560285472869873,
      "epoch": 0.8907304485907106,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004929417811085561,
      "loss": 5.3086,
      "mean_token_accuracy": 0.15954534113407134,
      "num_tokens": 19663669.0,
      "step": 8975
    },
    {
      "entropy": 5.634855079650879,
      "epoch": 0.8912266772528781,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004929329734368103,
      "loss": 5.3851,
      "mean_token_accuracy": 0.1564209520816803,
      "num_tokens": 19675851.0,
      "step": 8980
    },
    {
      "entropy": 5.549490594863892,
      "epoch": 0.8917229059150457,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004929241603607964,
      "loss": 5.2718,
      "mean_token_accuracy": 0.1691298305988312,
      "num_tokens": 19686613.0,
      "step": 8985
    },
    {
      "entropy": 5.466474437713623,
      "epoch": 0.8922191345772131,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000492915341880733,
      "loss": 5.1992,
      "mean_token_accuracy": 0.16453103274106978,
      "num_tokens": 19696419.0,
      "step": 8990
    },
    {
      "entropy": 5.518602275848389,
      "epoch": 0.8927153632393807,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004929065179968387,
      "loss": 5.2944,
      "mean_token_accuracy": 0.15724054127931594,
      "num_tokens": 19707269.0,
      "step": 8995
    },
    {
      "entropy": 5.605218172073364,
      "epoch": 0.8932115919015482,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004928976887093321,
      "loss": 5.341,
      "mean_token_accuracy": 0.15894262343645096,
      "num_tokens": 19716099.0,
      "step": 9000
    },
    {
      "epoch": 0.8932115919015482,
      "eval_entropy": 5.3559456596637345,
      "eval_loss": 5.37523889541626,
      "eval_mean_token_accuracy": 0.16191325234025752,
      "eval_num_tokens": 19716099.0,
      "eval_runtime": 20.6605,
      "eval_samples_per_second": 1629.629,
      "eval_steps_per_second": 203.722,
      "step": 9000
    },
    {
      "entropy": 5.54390869140625,
      "epoch": 0.8937078205637158,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004928888540184325,
      "loss": 5.3313,
      "mean_token_accuracy": 0.15764046013355254,
      "num_tokens": 19726632.0,
      "step": 9005
    },
    {
      "entropy": 5.51641879081726,
      "epoch": 0.8942040492258833,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004928800139243587,
      "loss": 5.2468,
      "mean_token_accuracy": 0.16316547393798828,
      "num_tokens": 19737929.0,
      "step": 9010
    },
    {
      "entropy": 5.5370947360992435,
      "epoch": 0.8947002778880508,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004928711684273302,
      "loss": 5.2577,
      "mean_token_accuracy": 0.16472524553537368,
      "num_tokens": 19748191.0,
      "step": 9015
    },
    {
      "entropy": 5.542944002151489,
      "epoch": 0.8951965065502183,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004928623175275661,
      "loss": 5.3148,
      "mean_token_accuracy": 0.16453443765640258,
      "num_tokens": 19758892.0,
      "step": 9020
    },
    {
      "entropy": 5.499523258209228,
      "epoch": 0.8956927352123859,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004928534612252862,
      "loss": 5.1886,
      "mean_token_accuracy": 0.16720210164785385,
      "num_tokens": 19769120.0,
      "step": 9025
    },
    {
      "entropy": 5.499724817276001,
      "epoch": 0.8961889638745534,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004928445995207095,
      "loss": 5.2481,
      "mean_token_accuracy": 0.16502394229173661,
      "num_tokens": 19779484.0,
      "step": 9030
    },
    {
      "entropy": 5.5149274349212645,
      "epoch": 0.896685192536721,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004928357324140564,
      "loss": 5.2963,
      "mean_token_accuracy": 0.16460447013378143,
      "num_tokens": 19790438.0,
      "step": 9035
    },
    {
      "entropy": 5.583444356918335,
      "epoch": 0.8971814211988884,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004928268599055465,
      "loss": 5.37,
      "mean_token_accuracy": 0.15877770036458969,
      "num_tokens": 19801646.0,
      "step": 9040
    },
    {
      "entropy": 5.560933637619018,
      "epoch": 0.897677649861056,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004928179819953998,
      "loss": 5.2606,
      "mean_token_accuracy": 0.16284436732530594,
      "num_tokens": 19812887.0,
      "step": 9045
    },
    {
      "entropy": 5.540252304077148,
      "epoch": 0.8981738785232235,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004928090986838366,
      "loss": 5.3149,
      "mean_token_accuracy": 0.1579166457056999,
      "num_tokens": 19822938.0,
      "step": 9050
    },
    {
      "entropy": 5.507092380523682,
      "epoch": 0.898670107185391,
      "grad_norm": 1.078125,
      "learning_rate": 0.000492800209971077,
      "loss": 5.271,
      "mean_token_accuracy": 0.16255158483982085,
      "num_tokens": 19833988.0,
      "step": 9055
    },
    {
      "entropy": 5.596437597274781,
      "epoch": 0.8991663358475586,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004927913158573416,
      "loss": 5.3345,
      "mean_token_accuracy": 0.15781496465206146,
      "num_tokens": 19845112.0,
      "step": 9060
    },
    {
      "entropy": 5.525338649749756,
      "epoch": 0.899662564509726,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004927824163428508,
      "loss": 5.2947,
      "mean_token_accuracy": 0.1584398239850998,
      "num_tokens": 19856304.0,
      "step": 9065
    },
    {
      "entropy": 5.5639386653900145,
      "epoch": 0.9001587931718936,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004927735114278254,
      "loss": 5.2946,
      "mean_token_accuracy": 0.16316746026277543,
      "num_tokens": 19866776.0,
      "step": 9070
    },
    {
      "entropy": 5.566666889190674,
      "epoch": 0.9006550218340611,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004927646011124861,
      "loss": 5.2598,
      "mean_token_accuracy": 0.1567187190055847,
      "num_tokens": 19876392.0,
      "step": 9075
    },
    {
      "entropy": 5.559143972396851,
      "epoch": 0.9011512504962287,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004927556853970539,
      "loss": 5.2991,
      "mean_token_accuracy": 0.15610761046409607,
      "num_tokens": 19887362.0,
      "step": 9080
    },
    {
      "entropy": 5.508514881134033,
      "epoch": 0.9016474791583962,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004927467642817499,
      "loss": 5.277,
      "mean_token_accuracy": 0.16156834959983826,
      "num_tokens": 19898989.0,
      "step": 9085
    },
    {
      "entropy": 5.551664209365844,
      "epoch": 0.9021437078205637,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004927378377667955,
      "loss": 5.2741,
      "mean_token_accuracy": 0.15657708197832107,
      "num_tokens": 19910112.0,
      "step": 9090
    },
    {
      "entropy": 5.606841802597046,
      "epoch": 0.9026399364827312,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000492728905852412,
      "loss": 5.3956,
      "mean_token_accuracy": 0.15744143724441528,
      "num_tokens": 19922601.0,
      "step": 9095
    },
    {
      "entropy": 5.5513715744018555,
      "epoch": 0.9031361651448988,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004927199685388205,
      "loss": 5.303,
      "mean_token_accuracy": 0.1631772145628929,
      "num_tokens": 19934153.0,
      "step": 9100
    },
    {
      "entropy": 5.535596609115601,
      "epoch": 0.9036323938070663,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000492711025826243,
      "loss": 5.2764,
      "mean_token_accuracy": 0.16301968693733215,
      "num_tokens": 19944572.0,
      "step": 9105
    },
    {
      "entropy": 5.546830320358277,
      "epoch": 0.9041286224692339,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004927020777149014,
      "loss": 5.3893,
      "mean_token_accuracy": 0.15813228785991668,
      "num_tokens": 19956431.0,
      "step": 9110
    },
    {
      "entropy": 5.589463758468628,
      "epoch": 0.9046248511314013,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004926931242050171,
      "loss": 5.3052,
      "mean_token_accuracy": 0.15729486644268037,
      "num_tokens": 19968118.0,
      "step": 9115
    },
    {
      "entropy": 5.540275049209595,
      "epoch": 0.9051210797935689,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004926841652968125,
      "loss": 5.3725,
      "mean_token_accuracy": 0.15925574228167533,
      "num_tokens": 19979545.0,
      "step": 9120
    },
    {
      "entropy": 5.543299961090088,
      "epoch": 0.9056173084557364,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004926752009905099,
      "loss": 5.3314,
      "mean_token_accuracy": 0.15359800308942795,
      "num_tokens": 19989832.0,
      "step": 9125
    },
    {
      "entropy": 5.542871141433716,
      "epoch": 0.9061135371179039,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004926662312863311,
      "loss": 5.3397,
      "mean_token_accuracy": 0.15784236341714858,
      "num_tokens": 20000935.0,
      "step": 9130
    },
    {
      "entropy": 5.546960735321045,
      "epoch": 0.9066097657800715,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004926572561844988,
      "loss": 5.3435,
      "mean_token_accuracy": 0.15512987971305847,
      "num_tokens": 20011485.0,
      "step": 9135
    },
    {
      "entropy": 5.513018083572388,
      "epoch": 0.9071059944422389,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004926482756852356,
      "loss": 5.1953,
      "mean_token_accuracy": 0.16653136759996415,
      "num_tokens": 20023043.0,
      "step": 9140
    },
    {
      "entropy": 5.572578573226929,
      "epoch": 0.9076022231044065,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004926392897887642,
      "loss": 5.2971,
      "mean_token_accuracy": 0.16815751641988755,
      "num_tokens": 20032437.0,
      "step": 9145
    },
    {
      "entropy": 5.553290939331054,
      "epoch": 0.908098451766574,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004926302984953073,
      "loss": 5.3356,
      "mean_token_accuracy": 0.15907517820596695,
      "num_tokens": 20044662.0,
      "step": 9150
    },
    {
      "entropy": 5.527995777130127,
      "epoch": 0.9085946804287416,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004926213018050881,
      "loss": 5.3433,
      "mean_token_accuracy": 0.1597750887274742,
      "num_tokens": 20055924.0,
      "step": 9155
    },
    {
      "entropy": 5.503418397903443,
      "epoch": 0.9090909090909091,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004926122997183295,
      "loss": 5.2934,
      "mean_token_accuracy": 0.15803872793912888,
      "num_tokens": 20067909.0,
      "step": 9160
    },
    {
      "entropy": 5.570617198944092,
      "epoch": 0.9095871377530766,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004926032922352546,
      "loss": 5.2679,
      "mean_token_accuracy": 0.16587397903203965,
      "num_tokens": 20079333.0,
      "step": 9165
    },
    {
      "entropy": 5.548792457580566,
      "epoch": 0.9100833664152441,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004925942793560872,
      "loss": 5.2517,
      "mean_token_accuracy": 0.1625090569257736,
      "num_tokens": 20089741.0,
      "step": 9170
    },
    {
      "entropy": 5.4624285221099855,
      "epoch": 0.9105795950774117,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004925852610810504,
      "loss": 5.2682,
      "mean_token_accuracy": 0.16062215268611907,
      "num_tokens": 20100267.0,
      "step": 9175
    },
    {
      "entropy": 5.504045677185059,
      "epoch": 0.9110758237395792,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000492576237410368,
      "loss": 5.2625,
      "mean_token_accuracy": 0.16514603048563004,
      "num_tokens": 20111964.0,
      "step": 9180
    },
    {
      "entropy": 5.567818880081177,
      "epoch": 0.9115720524017468,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004925672083442638,
      "loss": 5.3108,
      "mean_token_accuracy": 0.1528840556740761,
      "num_tokens": 20123132.0,
      "step": 9185
    },
    {
      "entropy": 5.469148397445679,
      "epoch": 0.9120682810639142,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004925581738829616,
      "loss": 5.2147,
      "mean_token_accuracy": 0.16623493880033494,
      "num_tokens": 20134050.0,
      "step": 9190
    },
    {
      "entropy": 5.490215921401978,
      "epoch": 0.9125645097260818,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004925491340266855,
      "loss": 5.2723,
      "mean_token_accuracy": 0.16605161279439926,
      "num_tokens": 20143814.0,
      "step": 9195
    },
    {
      "entropy": 5.5671141147613525,
      "epoch": 0.9130607383882493,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004925400887756596,
      "loss": 5.3543,
      "mean_token_accuracy": 0.1668687030673027,
      "num_tokens": 20157647.0,
      "step": 9200
    },
    {
      "entropy": 5.553544712066651,
      "epoch": 0.9135569670504168,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004925310381301083,
      "loss": 5.3219,
      "mean_token_accuracy": 0.16231947541236877,
      "num_tokens": 20168366.0,
      "step": 9205
    },
    {
      "entropy": 5.518465995788574,
      "epoch": 0.9140531957125844,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000492521982090256,
      "loss": 5.3712,
      "mean_token_accuracy": 0.1649198353290558,
      "num_tokens": 20179362.0,
      "step": 9210
    },
    {
      "entropy": 5.496027278900146,
      "epoch": 0.9145494243747518,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004925129206563273,
      "loss": 5.2507,
      "mean_token_accuracy": 0.16224647611379622,
      "num_tokens": 20190320.0,
      "step": 9215
    },
    {
      "entropy": 5.514825487136841,
      "epoch": 0.9150456530369194,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004925038538285467,
      "loss": 5.2776,
      "mean_token_accuracy": 0.1612439125776291,
      "num_tokens": 20201114.0,
      "step": 9220
    },
    {
      "entropy": 5.568852090835572,
      "epoch": 0.9155418816990869,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004924947816071394,
      "loss": 5.3015,
      "mean_token_accuracy": 0.16260259449481965,
      "num_tokens": 20213351.0,
      "step": 9225
    },
    {
      "entropy": 5.4828455448150635,
      "epoch": 0.9160381103612545,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00049248570399233,
      "loss": 5.1966,
      "mean_token_accuracy": 0.167659492790699,
      "num_tokens": 20223376.0,
      "step": 9230
    },
    {
      "entropy": 5.5454470157623295,
      "epoch": 0.916534339023422,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000492476620984344,
      "loss": 5.3408,
      "mean_token_accuracy": 0.15844542682170867,
      "num_tokens": 20234248.0,
      "step": 9235
    },
    {
      "entropy": 5.493923091888428,
      "epoch": 0.9170305676855895,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004924675325834063,
      "loss": 5.2665,
      "mean_token_accuracy": 0.16199109256267546,
      "num_tokens": 20245823.0,
      "step": 9240
    },
    {
      "entropy": 5.567879724502563,
      "epoch": 0.917526796347757,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004924584387897424,
      "loss": 5.3182,
      "mean_token_accuracy": 0.15727784782648085,
      "num_tokens": 20257003.0,
      "step": 9245
    },
    {
      "entropy": 5.5870460033416744,
      "epoch": 0.9180230250099246,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004924493396035777,
      "loss": 5.4568,
      "mean_token_accuracy": 0.15009055137634278,
      "num_tokens": 20268500.0,
      "step": 9250
    },
    {
      "entropy": 5.580854177474976,
      "epoch": 0.9185192536720921,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000492440235025138,
      "loss": 5.3211,
      "mean_token_accuracy": 0.16316964626312255,
      "num_tokens": 20279301.0,
      "step": 9255
    },
    {
      "entropy": 5.668870878219605,
      "epoch": 0.9190154823342597,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004924311250546491,
      "loss": 5.4184,
      "mean_token_accuracy": 0.15555869787931442,
      "num_tokens": 20291263.0,
      "step": 9260
    },
    {
      "entropy": 5.546143102645874,
      "epoch": 0.9195117109964271,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004924220096923367,
      "loss": 5.2915,
      "mean_token_accuracy": 0.16107711046934128,
      "num_tokens": 20301554.0,
      "step": 9265
    },
    {
      "entropy": 5.597596216201782,
      "epoch": 0.9200079396585947,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004924128889384271,
      "loss": 5.3434,
      "mean_token_accuracy": 0.1540423795580864,
      "num_tokens": 20314299.0,
      "step": 9270
    },
    {
      "entropy": 5.5163649559021,
      "epoch": 0.9205041683207622,
      "grad_norm": 0.875,
      "learning_rate": 0.0004924037627931463,
      "loss": 5.3096,
      "mean_token_accuracy": 0.1569603279232979,
      "num_tokens": 20324595.0,
      "step": 9275
    },
    {
      "entropy": 5.533094978332519,
      "epoch": 0.9210003969829297,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004923946312567206,
      "loss": 5.2775,
      "mean_token_accuracy": 0.15990834385156633,
      "num_tokens": 20335744.0,
      "step": 9280
    },
    {
      "entropy": 5.477016162872315,
      "epoch": 0.9214966256450973,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004923854943293765,
      "loss": 5.1769,
      "mean_token_accuracy": 0.16062400490045547,
      "num_tokens": 20346072.0,
      "step": 9285
    },
    {
      "entropy": 5.51767783164978,
      "epoch": 0.9219928543072647,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004923763520113406,
      "loss": 5.2671,
      "mean_token_accuracy": 0.17054644227027893,
      "num_tokens": 20357747.0,
      "step": 9290
    },
    {
      "entropy": 5.589597606658936,
      "epoch": 0.9224890829694323,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004923672043028395,
      "loss": 5.2477,
      "mean_token_accuracy": 0.1620402678847313,
      "num_tokens": 20368838.0,
      "step": 9295
    },
    {
      "entropy": 5.542013216018677,
      "epoch": 0.9229853116315998,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004923580512041001,
      "loss": 5.3211,
      "mean_token_accuracy": 0.15841785520315171,
      "num_tokens": 20380844.0,
      "step": 9300
    },
    {
      "entropy": 5.5596929550170895,
      "epoch": 0.9234815402937674,
      "grad_norm": 0.875,
      "learning_rate": 0.0004923488927153495,
      "loss": 5.3283,
      "mean_token_accuracy": 0.15808232724666596,
      "num_tokens": 20391739.0,
      "step": 9305
    },
    {
      "entropy": 5.5594805717468265,
      "epoch": 0.9239777689559349,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004923397288368146,
      "loss": 5.2179,
      "mean_token_accuracy": 0.17128596156835557,
      "num_tokens": 20402571.0,
      "step": 9310
    },
    {
      "entropy": 5.515147495269775,
      "epoch": 0.9244739976181024,
      "grad_norm": 1.0,
      "learning_rate": 0.0004923305595687229,
      "loss": 5.2554,
      "mean_token_accuracy": 0.16113622933626176,
      "num_tokens": 20414766.0,
      "step": 9315
    },
    {
      "entropy": 5.481299829483032,
      "epoch": 0.9249702262802699,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004923213849113015,
      "loss": 5.2268,
      "mean_token_accuracy": 0.16429373770952224,
      "num_tokens": 20425508.0,
      "step": 9320
    },
    {
      "entropy": 5.472104215621949,
      "epoch": 0.9254664549424375,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004923122048647781,
      "loss": 5.2209,
      "mean_token_accuracy": 0.1635184809565544,
      "num_tokens": 20436128.0,
      "step": 9325
    },
    {
      "entropy": 5.50439944267273,
      "epoch": 0.925962683604605,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004923030194293802,
      "loss": 5.3407,
      "mean_token_accuracy": 0.15802000164985658,
      "num_tokens": 20447212.0,
      "step": 9330
    },
    {
      "entropy": 5.541323184967041,
      "epoch": 0.9264589122667726,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004922938286053358,
      "loss": 5.2536,
      "mean_token_accuracy": 0.16571350246667862,
      "num_tokens": 20457770.0,
      "step": 9335
    },
    {
      "entropy": 5.534483146667481,
      "epoch": 0.92695514092894,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004922846323928726,
      "loss": 5.3223,
      "mean_token_accuracy": 0.15829780399799348,
      "num_tokens": 20468172.0,
      "step": 9340
    },
    {
      "entropy": 5.518729972839355,
      "epoch": 0.9274513695911076,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004922754307922188,
      "loss": 5.2452,
      "mean_token_accuracy": 0.16664718389511107,
      "num_tokens": 20479429.0,
      "step": 9345
    },
    {
      "entropy": 5.532600736618042,
      "epoch": 0.9279475982532751,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004922662238036024,
      "loss": 5.2169,
      "mean_token_accuracy": 0.16647566258907318,
      "num_tokens": 20490373.0,
      "step": 9350
    },
    {
      "entropy": 5.591423892974854,
      "epoch": 0.9284438269154426,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004922570114272518,
      "loss": 5.3467,
      "mean_token_accuracy": 0.16103961020708085,
      "num_tokens": 20500002.0,
      "step": 9355
    },
    {
      "entropy": 5.58456506729126,
      "epoch": 0.9289400555776102,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004922477936633955,
      "loss": 5.3858,
      "mean_token_accuracy": 0.1511903539299965,
      "num_tokens": 20512990.0,
      "step": 9360
    },
    {
      "entropy": 5.487053728103637,
      "epoch": 0.9294362842397776,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004922385705122621,
      "loss": 5.2877,
      "mean_token_accuracy": 0.16572093963623047,
      "num_tokens": 20522979.0,
      "step": 9365
    },
    {
      "entropy": 5.482340526580811,
      "epoch": 0.9299325129019452,
      "grad_norm": 0.875,
      "learning_rate": 0.0004922293419740802,
      "loss": 5.1863,
      "mean_token_accuracy": 0.16835395693778993,
      "num_tokens": 20534620.0,
      "step": 9370
    },
    {
      "entropy": 5.542438554763794,
      "epoch": 0.9304287415641127,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004922201080490787,
      "loss": 5.2794,
      "mean_token_accuracy": 0.16248394697904586,
      "num_tokens": 20545526.0,
      "step": 9375
    },
    {
      "entropy": 5.605936765670776,
      "epoch": 0.9309249702262803,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004922108687374867,
      "loss": 5.4447,
      "mean_token_accuracy": 0.15707781314849853,
      "num_tokens": 20556665.0,
      "step": 9380
    },
    {
      "entropy": 5.571640491485596,
      "epoch": 0.9314211988884478,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000492201624039533,
      "loss": 5.2937,
      "mean_token_accuracy": 0.1559583067893982,
      "num_tokens": 20566054.0,
      "step": 9385
    },
    {
      "entropy": 5.631577730178833,
      "epoch": 0.9319174275506154,
      "grad_norm": 0.875,
      "learning_rate": 0.0004921923739554472,
      "loss": 5.355,
      "mean_token_accuracy": 0.15491781681776046,
      "num_tokens": 20577922.0,
      "step": 9390
    },
    {
      "entropy": 5.471510410308838,
      "epoch": 0.9324136562127828,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004921831184854585,
      "loss": 5.1735,
      "mean_token_accuracy": 0.16721322238445283,
      "num_tokens": 20588473.0,
      "step": 9395
    },
    {
      "entropy": 5.51868200302124,
      "epoch": 0.9329098848749504,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004921738576297965,
      "loss": 5.3654,
      "mean_token_accuracy": 0.15697364658117294,
      "num_tokens": 20599371.0,
      "step": 9400
    },
    {
      "entropy": 5.505065202713013,
      "epoch": 0.9334061135371179,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004921645913886907,
      "loss": 5.2431,
      "mean_token_accuracy": 0.15694219917058944,
      "num_tokens": 20610066.0,
      "step": 9405
    },
    {
      "entropy": 5.560163259506226,
      "epoch": 0.9339023421992855,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004921553197623711,
      "loss": 5.3424,
      "mean_token_accuracy": 0.1578836128115654,
      "num_tokens": 20621804.0,
      "step": 9410
    },
    {
      "entropy": 5.563719654083252,
      "epoch": 0.934398570861453,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004921460427510674,
      "loss": 5.2359,
      "mean_token_accuracy": 0.16877928823232652,
      "num_tokens": 20632706.0,
      "step": 9415
    },
    {
      "entropy": 5.520000648498535,
      "epoch": 0.9348947995236205,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004921367603550098,
      "loss": 5.3352,
      "mean_token_accuracy": 0.15229426324367523,
      "num_tokens": 20643303.0,
      "step": 9420
    },
    {
      "entropy": 5.530603313446045,
      "epoch": 0.935391028185788,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004921274725744283,
      "loss": 5.2895,
      "mean_token_accuracy": 0.1678344264626503,
      "num_tokens": 20653681.0,
      "step": 9425
    },
    {
      "entropy": 5.5783344268798825,
      "epoch": 0.9358872568479555,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004921181794095536,
      "loss": 5.3101,
      "mean_token_accuracy": 0.1571393668651581,
      "num_tokens": 20665599.0,
      "step": 9430
    },
    {
      "entropy": 5.48060564994812,
      "epoch": 0.9363834855101231,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004921088808606157,
      "loss": 5.1938,
      "mean_token_accuracy": 0.1647046461701393,
      "num_tokens": 20675726.0,
      "step": 9435
    },
    {
      "entropy": 5.522446250915527,
      "epoch": 0.9368797141722905,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004920995769278454,
      "loss": 5.3764,
      "mean_token_accuracy": 0.1545850783586502,
      "num_tokens": 20688083.0,
      "step": 9440
    },
    {
      "entropy": 5.562725925445557,
      "epoch": 0.9373759428344581,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004920902676114734,
      "loss": 5.2386,
      "mean_token_accuracy": 0.16600545197725297,
      "num_tokens": 20697642.0,
      "step": 9445
    },
    {
      "entropy": 5.575511169433594,
      "epoch": 0.9378721714966256,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004920809529117306,
      "loss": 5.3388,
      "mean_token_accuracy": 0.15686559230089187,
      "num_tokens": 20709172.0,
      "step": 9450
    },
    {
      "entropy": 5.600897645950317,
      "epoch": 0.9383684001587932,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004920716328288479,
      "loss": 5.3896,
      "mean_token_accuracy": 0.1551354184746742,
      "num_tokens": 20720563.0,
      "step": 9455
    },
    {
      "entropy": 5.54432897567749,
      "epoch": 0.9388646288209607,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004920623073630566,
      "loss": 5.2686,
      "mean_token_accuracy": 0.1644684612751007,
      "num_tokens": 20730943.0,
      "step": 9460
    },
    {
      "entropy": 5.539048767089843,
      "epoch": 0.9393608574831283,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004920529765145876,
      "loss": 5.2408,
      "mean_token_accuracy": 0.16011967062950133,
      "num_tokens": 20740903.0,
      "step": 9465
    },
    {
      "entropy": 5.531922006607056,
      "epoch": 0.9398570861452957,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004920436402836726,
      "loss": 5.3323,
      "mean_token_accuracy": 0.16579257398843766,
      "num_tokens": 20750580.0,
      "step": 9470
    },
    {
      "entropy": 5.530846405029297,
      "epoch": 0.9403533148074633,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000492034298670543,
      "loss": 5.2539,
      "mean_token_accuracy": 0.16114427000284196,
      "num_tokens": 20761357.0,
      "step": 9475
    },
    {
      "entropy": 5.625053882598877,
      "epoch": 0.9408495434696308,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004920249516754304,
      "loss": 5.4096,
      "mean_token_accuracy": 0.1506986804306507,
      "num_tokens": 20773273.0,
      "step": 9480
    },
    {
      "entropy": 5.5660483837127686,
      "epoch": 0.9413457721317984,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004920155992985667,
      "loss": 5.2401,
      "mean_token_accuracy": 0.16067681461572647,
      "num_tokens": 20784290.0,
      "step": 9485
    },
    {
      "entropy": 5.531092119216919,
      "epoch": 0.9418420007939659,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004920062415401837,
      "loss": 5.2741,
      "mean_token_accuracy": 0.16394226998090744,
      "num_tokens": 20794764.0,
      "step": 9490
    },
    {
      "entropy": 5.564286994934082,
      "epoch": 0.9423382294561334,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004919968784005136,
      "loss": 5.3155,
      "mean_token_accuracy": 0.15846064537763596,
      "num_tokens": 20804863.0,
      "step": 9495
    },
    {
      "entropy": 5.6152464866638185,
      "epoch": 0.9428344581183009,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004919875098797886,
      "loss": 5.4012,
      "mean_token_accuracy": 0.15524280518293382,
      "num_tokens": 20816445.0,
      "step": 9500
    },
    {
      "entropy": 5.557435178756714,
      "epoch": 0.9433306867804684,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004919781359782407,
      "loss": 5.2244,
      "mean_token_accuracy": 0.16715539544820784,
      "num_tokens": 20827141.0,
      "step": 9505
    },
    {
      "entropy": 5.526000022888184,
      "epoch": 0.943826915442636,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004919687566961027,
      "loss": 5.2707,
      "mean_token_accuracy": 0.15752107948064803,
      "num_tokens": 20838291.0,
      "step": 9510
    },
    {
      "entropy": 5.579127216339112,
      "epoch": 0.9443231441048034,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000491959372033607,
      "loss": 5.2776,
      "mean_token_accuracy": 0.16056058555841446,
      "num_tokens": 20848839.0,
      "step": 9515
    },
    {
      "entropy": 5.5651250839233395,
      "epoch": 0.944819372766971,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004919499819909865,
      "loss": 5.2558,
      "mean_token_accuracy": 0.16248184889554979,
      "num_tokens": 20859961.0,
      "step": 9520
    },
    {
      "entropy": 5.481574821472168,
      "epoch": 0.9453156014291385,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004919405865684738,
      "loss": 5.2034,
      "mean_token_accuracy": 0.16611860692501068,
      "num_tokens": 20869720.0,
      "step": 9525
    },
    {
      "entropy": 5.507126235961914,
      "epoch": 0.9458118300913061,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000491931185766302,
      "loss": 5.3435,
      "mean_token_accuracy": 0.16365325152873994,
      "num_tokens": 20880513.0,
      "step": 9530
    },
    {
      "entropy": 5.566145992279052,
      "epoch": 0.9463080587534736,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004919217795847044,
      "loss": 5.2656,
      "mean_token_accuracy": 0.16495689749717712,
      "num_tokens": 20892456.0,
      "step": 9535
    },
    {
      "entropy": 5.534870147705078,
      "epoch": 0.9468042874156412,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004919123680239139,
      "loss": 5.2722,
      "mean_token_accuracy": 0.16199540197849274,
      "num_tokens": 20904500.0,
      "step": 9540
    },
    {
      "entropy": 5.5694770336151125,
      "epoch": 0.9473005160778086,
      "grad_norm": 0.875,
      "learning_rate": 0.0004919029510841641,
      "loss": 5.2541,
      "mean_token_accuracy": 0.16894175112247467,
      "num_tokens": 20915133.0,
      "step": 9545
    },
    {
      "entropy": 5.453264045715332,
      "epoch": 0.9477967447399762,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004918935287656887,
      "loss": 5.245,
      "mean_token_accuracy": 0.16683667302131652,
      "num_tokens": 20924779.0,
      "step": 9550
    },
    {
      "entropy": 5.459113740921021,
      "epoch": 0.9482929734021437,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000491884101068721,
      "loss": 5.1905,
      "mean_token_accuracy": 0.1652270406484604,
      "num_tokens": 20935387.0,
      "step": 9555
    },
    {
      "entropy": 5.619988679885864,
      "epoch": 0.9487892020643113,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004918746679934949,
      "loss": 5.3682,
      "mean_token_accuracy": 0.1594923883676529,
      "num_tokens": 20945432.0,
      "step": 9560
    },
    {
      "entropy": 5.542451524734497,
      "epoch": 0.9492854307264788,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004918652295402444,
      "loss": 5.2483,
      "mean_token_accuracy": 0.1673049211502075,
      "num_tokens": 20956777.0,
      "step": 9565
    },
    {
      "entropy": 5.500725793838501,
      "epoch": 0.9497816593886463,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004918557857092036,
      "loss": 5.2856,
      "mean_token_accuracy": 0.15981982946395873,
      "num_tokens": 20966057.0,
      "step": 9570
    },
    {
      "entropy": 5.551764345169067,
      "epoch": 0.9502778880508138,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004918463365006065,
      "loss": 5.3556,
      "mean_token_accuracy": 0.15958352386951447,
      "num_tokens": 20976838.0,
      "step": 9575
    },
    {
      "entropy": 5.602881717681885,
      "epoch": 0.9507741167129813,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004918368819146875,
      "loss": 5.2996,
      "mean_token_accuracy": 0.1609395667910576,
      "num_tokens": 20988553.0,
      "step": 9580
    },
    {
      "entropy": 5.53433780670166,
      "epoch": 0.9512703453751489,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004918274219516811,
      "loss": 5.2791,
      "mean_token_accuracy": 0.1604821041226387,
      "num_tokens": 20998983.0,
      "step": 9585
    },
    {
      "entropy": 5.625037336349488,
      "epoch": 0.9517665740373163,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004918179566118218,
      "loss": 5.3883,
      "mean_token_accuracy": 0.1573731333017349,
      "num_tokens": 21010008.0,
      "step": 9590
    },
    {
      "entropy": 5.598756742477417,
      "epoch": 0.9522628026994839,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004918084858953445,
      "loss": 5.3599,
      "mean_token_accuracy": 0.16161780506372453,
      "num_tokens": 21022826.0,
      "step": 9595
    },
    {
      "entropy": 5.555862092971802,
      "epoch": 0.9527590313616514,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004917990098024839,
      "loss": 5.2834,
      "mean_token_accuracy": 0.15864650160074234,
      "num_tokens": 21033457.0,
      "step": 9600
    },
    {
      "entropy": 5.505966281890869,
      "epoch": 0.953255260023819,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004917895283334749,
      "loss": 5.3013,
      "mean_token_accuracy": 0.16056045591831208,
      "num_tokens": 21044640.0,
      "step": 9605
    },
    {
      "entropy": 5.567455577850342,
      "epoch": 0.9537514886859865,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000491780041488553,
      "loss": 5.2756,
      "mean_token_accuracy": 0.1714605748653412,
      "num_tokens": 21055728.0,
      "step": 9610
    },
    {
      "entropy": 5.550451135635376,
      "epoch": 0.9542477173481541,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004917705492679529,
      "loss": 5.2707,
      "mean_token_accuracy": 0.15992705672979354,
      "num_tokens": 21066057.0,
      "step": 9615
    },
    {
      "entropy": 5.578377771377563,
      "epoch": 0.9547439460103215,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004917610516719104,
      "loss": 5.2892,
      "mean_token_accuracy": 0.15747786164283753,
      "num_tokens": 21075717.0,
      "step": 9620
    },
    {
      "entropy": 5.647128391265869,
      "epoch": 0.9552401746724891,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004917515487006609,
      "loss": 5.3876,
      "mean_token_accuracy": 0.15149791836738585,
      "num_tokens": 21087120.0,
      "step": 9625
    },
    {
      "entropy": 5.595139646530152,
      "epoch": 0.9557364033346566,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00049174204035444,
      "loss": 5.4064,
      "mean_token_accuracy": 0.15330189242959022,
      "num_tokens": 21100110.0,
      "step": 9630
    },
    {
      "entropy": 5.641125917434692,
      "epoch": 0.9562326319968242,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004917325266334836,
      "loss": 5.3082,
      "mean_token_accuracy": 0.1670064240694046,
      "num_tokens": 21110411.0,
      "step": 9635
    },
    {
      "entropy": 5.508846139907837,
      "epoch": 0.9567288606589917,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004917230075380276,
      "loss": 5.2215,
      "mean_token_accuracy": 0.16506332606077195,
      "num_tokens": 21120303.0,
      "step": 9640
    },
    {
      "entropy": 5.52659010887146,
      "epoch": 0.9572250893211592,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004917134830683079,
      "loss": 5.3249,
      "mean_token_accuracy": 0.15908250510692595,
      "num_tokens": 21132312.0,
      "step": 9645
    },
    {
      "entropy": 5.587954521179199,
      "epoch": 0.9577213179833267,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004917039532245608,
      "loss": 5.2954,
      "mean_token_accuracy": 0.16296826899051667,
      "num_tokens": 21142472.0,
      "step": 9650
    },
    {
      "entropy": 5.547664308547974,
      "epoch": 0.9582175466454942,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004916944180070226,
      "loss": 5.3104,
      "mean_token_accuracy": 0.1562706708908081,
      "num_tokens": 21153306.0,
      "step": 9655
    },
    {
      "entropy": 5.620442676544189,
      "epoch": 0.9587137753076618,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004916848774159299,
      "loss": 5.3544,
      "mean_token_accuracy": 0.15180012881755828,
      "num_tokens": 21165606.0,
      "step": 9660
    },
    {
      "entropy": 5.582696485519409,
      "epoch": 0.9592100039698293,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000491675331451519,
      "loss": 5.2715,
      "mean_token_accuracy": 0.16641474664211273,
      "num_tokens": 21176734.0,
      "step": 9665
    },
    {
      "entropy": 5.496678686141967,
      "epoch": 0.9597062326319968,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004916657801140269,
      "loss": 5.2196,
      "mean_token_accuracy": 0.16818380653858184,
      "num_tokens": 21187109.0,
      "step": 9670
    },
    {
      "entropy": 5.493848991394043,
      "epoch": 0.9602024612941643,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004916562234036903,
      "loss": 5.2714,
      "mean_token_accuracy": 0.16576831340789794,
      "num_tokens": 21197281.0,
      "step": 9675
    },
    {
      "entropy": 5.538375568389893,
      "epoch": 0.9606986899563319,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004916466613207462,
      "loss": 5.3115,
      "mean_token_accuracy": 0.15971529334783555,
      "num_tokens": 21208767.0,
      "step": 9680
    },
    {
      "entropy": 5.5474122047424315,
      "epoch": 0.9611949186184994,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004916370938654317,
      "loss": 5.2969,
      "mean_token_accuracy": 0.16287417113780975,
      "num_tokens": 21219328.0,
      "step": 9685
    },
    {
      "entropy": 5.602007246017456,
      "epoch": 0.961691147280667,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004916275210379841,
      "loss": 5.4015,
      "mean_token_accuracy": 0.15935112088918685,
      "num_tokens": 21231476.0,
      "step": 9690
    },
    {
      "entropy": 5.639980030059815,
      "epoch": 0.9621873759428344,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004916179428386408,
      "loss": 5.359,
      "mean_token_accuracy": 0.16527327299118041,
      "num_tokens": 21242445.0,
      "step": 9695
    },
    {
      "entropy": 5.519500875473023,
      "epoch": 0.962683604605002,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004916083592676392,
      "loss": 5.1606,
      "mean_token_accuracy": 0.16533677130937577,
      "num_tokens": 21253179.0,
      "step": 9700
    },
    {
      "entropy": 5.5240099906921385,
      "epoch": 0.9631798332671695,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004915987703252171,
      "loss": 5.2015,
      "mean_token_accuracy": 0.1710731193423271,
      "num_tokens": 21263460.0,
      "step": 9705
    },
    {
      "entropy": 5.451681900024414,
      "epoch": 0.9636760619293371,
      "grad_norm": 0.875,
      "learning_rate": 0.0004915891760116123,
      "loss": 5.2632,
      "mean_token_accuracy": 0.1635735422372818,
      "num_tokens": 21274819.0,
      "step": 9710
    },
    {
      "entropy": 5.517275714874268,
      "epoch": 0.9641722905915046,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004915795763270626,
      "loss": 5.2254,
      "mean_token_accuracy": 0.16332022994756698,
      "num_tokens": 21286215.0,
      "step": 9715
    },
    {
      "entropy": 5.557509994506836,
      "epoch": 0.9646685192536721,
      "grad_norm": 0.953125,
      "learning_rate": 0.000491569971271806,
      "loss": 5.2211,
      "mean_token_accuracy": 0.16662717461586,
      "num_tokens": 21297033.0,
      "step": 9720
    },
    {
      "entropy": 5.490494346618652,
      "epoch": 0.9651647479158396,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004915603608460809,
      "loss": 5.2213,
      "mean_token_accuracy": 0.16428214311599731,
      "num_tokens": 21308019.0,
      "step": 9725
    },
    {
      "entropy": 5.536657381057739,
      "epoch": 0.9656609765780071,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004915507450501254,
      "loss": 5.3577,
      "mean_token_accuracy": 0.15414009913802146,
      "num_tokens": 21319313.0,
      "step": 9730
    },
    {
      "entropy": 5.569180727005005,
      "epoch": 0.9661572052401747,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000491541123884178,
      "loss": 5.2682,
      "mean_token_accuracy": 0.16304835081100463,
      "num_tokens": 21329950.0,
      "step": 9735
    },
    {
      "entropy": 5.537885951995849,
      "epoch": 0.9666534339023422,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004915314973484775,
      "loss": 5.2301,
      "mean_token_accuracy": 0.16611171811819075,
      "num_tokens": 21339061.0,
      "step": 9740
    },
    {
      "entropy": 5.530014896392823,
      "epoch": 0.9671496625645097,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004915218654432624,
      "loss": 5.3287,
      "mean_token_accuracy": 0.1598948985338211,
      "num_tokens": 21350114.0,
      "step": 9745
    },
    {
      "entropy": 5.555548477172851,
      "epoch": 0.9676458912266772,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004915122281687716,
      "loss": 5.2516,
      "mean_token_accuracy": 0.16767426282167436,
      "num_tokens": 21361219.0,
      "step": 9750
    },
    {
      "entropy": 5.485579776763916,
      "epoch": 0.9681421198888448,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004915025855252441,
      "loss": 5.2458,
      "mean_token_accuracy": 0.16266645789146422,
      "num_tokens": 21372722.0,
      "step": 9755
    },
    {
      "entropy": 5.5408666133880615,
      "epoch": 0.9686383485510123,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004914929375129188,
      "loss": 5.2766,
      "mean_token_accuracy": 0.16189596205949783,
      "num_tokens": 21383350.0,
      "step": 9760
    },
    {
      "entropy": 5.569598627090454,
      "epoch": 0.9691345772131799,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004914832841320353,
      "loss": 5.3227,
      "mean_token_accuracy": 0.1622638449072838,
      "num_tokens": 21394656.0,
      "step": 9765
    },
    {
      "entropy": 5.544922971725464,
      "epoch": 0.9696308058753473,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000491473625382833,
      "loss": 5.3138,
      "mean_token_accuracy": 0.15711605846881865,
      "num_tokens": 21406163.0,
      "step": 9770
    },
    {
      "entropy": 5.459386777877808,
      "epoch": 0.9701270345375149,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000491463961265551,
      "loss": 5.2086,
      "mean_token_accuracy": 0.16764862686395646,
      "num_tokens": 21418703.0,
      "step": 9775
    },
    {
      "entropy": 5.505564880371094,
      "epoch": 0.9706232631996824,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004914542917804292,
      "loss": 5.2119,
      "mean_token_accuracy": 0.17257481962442398,
      "num_tokens": 21428511.0,
      "step": 9780
    },
    {
      "entropy": 5.557694387435913,
      "epoch": 0.97111949186185,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004914446169277075,
      "loss": 5.2502,
      "mean_token_accuracy": 0.16785378605127335,
      "num_tokens": 21440350.0,
      "step": 9785
    },
    {
      "entropy": 5.60865159034729,
      "epoch": 0.9716157205240175,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004914349367076258,
      "loss": 5.346,
      "mean_token_accuracy": 0.1607807070016861,
      "num_tokens": 21451262.0,
      "step": 9790
    },
    {
      "entropy": 5.567680978775025,
      "epoch": 0.972111949186185,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004914252511204238,
      "loss": 5.3135,
      "mean_token_accuracy": 0.1643596589565277,
      "num_tokens": 21462587.0,
      "step": 9795
    },
    {
      "entropy": 5.5603399753570555,
      "epoch": 0.9726081778483525,
      "grad_norm": 1.0,
      "learning_rate": 0.0004914155601663422,
      "loss": 5.2482,
      "mean_token_accuracy": 0.16613210290670394,
      "num_tokens": 21473452.0,
      "step": 9800
    },
    {
      "entropy": 5.5556717872619625,
      "epoch": 0.97310440651052,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004914058638456208,
      "loss": 5.2258,
      "mean_token_accuracy": 0.16170699894428253,
      "num_tokens": 21484011.0,
      "step": 9805
    },
    {
      "entropy": 5.466827297210694,
      "epoch": 0.9736006351726876,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004913961621585003,
      "loss": 5.2195,
      "mean_token_accuracy": 0.1644126608967781,
      "num_tokens": 21495959.0,
      "step": 9810
    },
    {
      "entropy": 5.481923246383667,
      "epoch": 0.974096863834855,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004913864551052214,
      "loss": 5.2589,
      "mean_token_accuracy": 0.1722146451473236,
      "num_tokens": 21507496.0,
      "step": 9815
    },
    {
      "entropy": 5.621778869628907,
      "epoch": 0.9745930924970226,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004913767426860246,
      "loss": 5.3438,
      "mean_token_accuracy": 0.1552169665694237,
      "num_tokens": 21517532.0,
      "step": 9820
    },
    {
      "entropy": 5.49606409072876,
      "epoch": 0.9750893211591901,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004913670249011508,
      "loss": 5.1665,
      "mean_token_accuracy": 0.16770999878644943,
      "num_tokens": 21527969.0,
      "step": 9825
    },
    {
      "entropy": 5.5104217529296875,
      "epoch": 0.9755855498213577,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004913573017508411,
      "loss": 5.2711,
      "mean_token_accuracy": 0.15946028381586075,
      "num_tokens": 21538243.0,
      "step": 9830
    },
    {
      "entropy": 5.563217544555664,
      "epoch": 0.9760817784835252,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004913475732353365,
      "loss": 5.2803,
      "mean_token_accuracy": 0.16343535631895065,
      "num_tokens": 21548702.0,
      "step": 9835
    },
    {
      "entropy": 5.486120510101318,
      "epoch": 0.9765780071456928,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004913378393548784,
      "loss": 5.2372,
      "mean_token_accuracy": 0.16483518332242966,
      "num_tokens": 21558583.0,
      "step": 9840
    },
    {
      "entropy": 5.504124307632447,
      "epoch": 0.9770742358078602,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004913281001097078,
      "loss": 5.2229,
      "mean_token_accuracy": 0.16806641072034836,
      "num_tokens": 21569845.0,
      "step": 9845
    },
    {
      "entropy": 5.567736673355102,
      "epoch": 0.9775704644700278,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004913183555000667,
      "loss": 5.3681,
      "mean_token_accuracy": 0.15614749044179915,
      "num_tokens": 21580540.0,
      "step": 9850
    },
    {
      "entropy": 5.568388223648071,
      "epoch": 0.9780666931321953,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004913086055261963,
      "loss": 5.2303,
      "mean_token_accuracy": 0.1666613608598709,
      "num_tokens": 21590982.0,
      "step": 9855
    },
    {
      "entropy": 5.524022960662842,
      "epoch": 0.9785629217943629,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004912988501883387,
      "loss": 5.2565,
      "mean_token_accuracy": 0.16447547525167466,
      "num_tokens": 21602817.0,
      "step": 9860
    },
    {
      "entropy": 5.5490241050720215,
      "epoch": 0.9790591504565304,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004912890894867357,
      "loss": 5.3001,
      "mean_token_accuracy": 0.16047051697969436,
      "num_tokens": 21613422.0,
      "step": 9865
    },
    {
      "entropy": 5.536137580871582,
      "epoch": 0.9795553791186979,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004912793234216293,
      "loss": 5.2961,
      "mean_token_accuracy": 0.1544037289917469,
      "num_tokens": 21624241.0,
      "step": 9870
    },
    {
      "entropy": 5.485258722305298,
      "epoch": 0.9800516077808654,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004912695519932618,
      "loss": 5.2854,
      "mean_token_accuracy": 0.164899580180645,
      "num_tokens": 21636243.0,
      "step": 9875
    },
    {
      "entropy": 5.5323365211486815,
      "epoch": 0.9805478364430329,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004912597752018753,
      "loss": 5.3106,
      "mean_token_accuracy": 0.15859518200159073,
      "num_tokens": 21647918.0,
      "step": 9880
    },
    {
      "entropy": 5.5906257152557375,
      "epoch": 0.9810440651052005,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004912499930477126,
      "loss": 5.2838,
      "mean_token_accuracy": 0.16075822114944457,
      "num_tokens": 21659236.0,
      "step": 9885
    },
    {
      "entropy": 5.5201832294464115,
      "epoch": 0.981540293767368,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004912402055310158,
      "loss": 5.2809,
      "mean_token_accuracy": 0.1637316793203354,
      "num_tokens": 21669647.0,
      "step": 9890
    },
    {
      "entropy": 5.509677886962891,
      "epoch": 0.9820365224295355,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004912304126520279,
      "loss": 5.3217,
      "mean_token_accuracy": 0.15920768976211547,
      "num_tokens": 21680369.0,
      "step": 9895
    },
    {
      "entropy": 5.593974637985229,
      "epoch": 0.982532751091703,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004912206144109917,
      "loss": 5.3321,
      "mean_token_accuracy": 0.15918789356946944,
      "num_tokens": 21690942.0,
      "step": 9900
    },
    {
      "entropy": 5.552346706390381,
      "epoch": 0.9830289797538706,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004912108108081503,
      "loss": 5.2555,
      "mean_token_accuracy": 0.16583765149116517,
      "num_tokens": 21701945.0,
      "step": 9905
    },
    {
      "entropy": 5.577497863769532,
      "epoch": 0.9835252084160381,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004912010018437465,
      "loss": 5.284,
      "mean_token_accuracy": 0.1665597915649414,
      "num_tokens": 21712246.0,
      "step": 9910
    },
    {
      "entropy": 5.6315020561218265,
      "epoch": 0.9840214370782057,
      "grad_norm": 0.890625,
      "learning_rate": 0.000491191187518024,
      "loss": 5.2855,
      "mean_token_accuracy": 0.16163774877786635,
      "num_tokens": 21721660.0,
      "step": 9915
    },
    {
      "entropy": 5.506907224655151,
      "epoch": 0.9845176657403731,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004911813678312257,
      "loss": 5.17,
      "mean_token_accuracy": 0.1689017653465271,
      "num_tokens": 21732157.0,
      "step": 9920
    },
    {
      "entropy": 5.512174034118653,
      "epoch": 0.9850138944025407,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004911715427835953,
      "loss": 5.3077,
      "mean_token_accuracy": 0.15705085918307304,
      "num_tokens": 21741727.0,
      "step": 9925
    },
    {
      "entropy": 5.540558862686157,
      "epoch": 0.9855101230647082,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004911617123753765,
      "loss": 5.2454,
      "mean_token_accuracy": 0.1678730845451355,
      "num_tokens": 21752596.0,
      "step": 9930
    },
    {
      "entropy": 5.4900116443634035,
      "epoch": 0.9860063517268758,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004911518766068131,
      "loss": 5.1613,
      "mean_token_accuracy": 0.16437255591154099,
      "num_tokens": 21763978.0,
      "step": 9935
    },
    {
      "entropy": 5.518296909332276,
      "epoch": 0.9865025803890433,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004911420354781487,
      "loss": 5.3046,
      "mean_token_accuracy": 0.15439167469739914,
      "num_tokens": 21774961.0,
      "step": 9940
    },
    {
      "entropy": 5.527509880065918,
      "epoch": 0.9869988090512108,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004911321889896278,
      "loss": 5.2118,
      "mean_token_accuracy": 0.17294728606939316,
      "num_tokens": 21785387.0,
      "step": 9945
    },
    {
      "entropy": 5.514861297607422,
      "epoch": 0.9874950377133783,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004911223371414943,
      "loss": 5.2672,
      "mean_token_accuracy": 0.16925038099288942,
      "num_tokens": 21795982.0,
      "step": 9950
    },
    {
      "entropy": 5.586288785934448,
      "epoch": 0.9879912663755459,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004911124799339924,
      "loss": 5.3018,
      "mean_token_accuracy": 0.15709125995635986,
      "num_tokens": 21806768.0,
      "step": 9955
    },
    {
      "entropy": 5.5430042266845705,
      "epoch": 0.9884874950377134,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004911026173673667,
      "loss": 5.2655,
      "mean_token_accuracy": 0.16700110733509063,
      "num_tokens": 21818563.0,
      "step": 9960
    },
    {
      "entropy": 5.57335844039917,
      "epoch": 0.9889837236998809,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004910927494418619,
      "loss": 5.3358,
      "mean_token_accuracy": 0.16360723674297334,
      "num_tokens": 21830253.0,
      "step": 9965
    },
    {
      "entropy": 5.5798474788665775,
      "epoch": 0.9894799523620484,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004910828761577224,
      "loss": 5.3094,
      "mean_token_accuracy": 0.1650498926639557,
      "num_tokens": 21842353.0,
      "step": 9970
    },
    {
      "entropy": 5.530132532119751,
      "epoch": 0.9899761810242159,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004910729975151933,
      "loss": 5.2442,
      "mean_token_accuracy": 0.1677899479866028,
      "num_tokens": 21852066.0,
      "step": 9975
    },
    {
      "entropy": 5.504552602767944,
      "epoch": 0.9904724096863835,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004910631135145194,
      "loss": 5.2664,
      "mean_token_accuracy": 0.16598305702209473,
      "num_tokens": 21862754.0,
      "step": 9980
    },
    {
      "entropy": 5.441079378128052,
      "epoch": 0.990968638348551,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004910532241559458,
      "loss": 5.1758,
      "mean_token_accuracy": 0.17741164416074753,
      "num_tokens": 21873531.0,
      "step": 9985
    },
    {
      "entropy": 5.444345712661743,
      "epoch": 0.9914648670107186,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004910433294397179,
      "loss": 5.1451,
      "mean_token_accuracy": 0.16984826773405076,
      "num_tokens": 21884339.0,
      "step": 9990
    },
    {
      "entropy": 5.520063018798828,
      "epoch": 0.991961095672886,
      "grad_norm": 0.921875,
      "learning_rate": 0.000491033429366081,
      "loss": 5.2445,
      "mean_token_accuracy": 0.16473416090011597,
      "num_tokens": 21895489.0,
      "step": 9995
    },
    {
      "entropy": 5.527721643447876,
      "epoch": 0.9924573243350536,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004910235239352804,
      "loss": 5.2228,
      "mean_token_accuracy": 0.1649106651544571,
      "num_tokens": 21906832.0,
      "step": 10000
    },
    {
      "entropy": 5.516564750671387,
      "epoch": 0.9929535529972211,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000491013613147562,
      "loss": 5.2757,
      "mean_token_accuracy": 0.16007844358682632,
      "num_tokens": 21916979.0,
      "step": 10005
    },
    {
      "entropy": 5.515347290039062,
      "epoch": 0.9934497816593887,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004910036970031715,
      "loss": 5.1807,
      "mean_token_accuracy": 0.1732892245054245,
      "num_tokens": 21927352.0,
      "step": 10010
    },
    {
      "entropy": 5.514059352874756,
      "epoch": 0.9939460103215562,
      "grad_norm": 0.875,
      "learning_rate": 0.0004909937755023547,
      "loss": 5.3066,
      "mean_token_accuracy": 0.1569930851459503,
      "num_tokens": 21937765.0,
      "step": 10015
    },
    {
      "entropy": 5.46819486618042,
      "epoch": 0.9944422389837237,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004909838486453577,
      "loss": 5.1887,
      "mean_token_accuracy": 0.16520598977804185,
      "num_tokens": 21948659.0,
      "step": 10020
    },
    {
      "entropy": 5.517417621612549,
      "epoch": 0.9949384676458912,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004909739164324267,
      "loss": 5.3063,
      "mean_token_accuracy": 0.15833762884140015,
      "num_tokens": 21959356.0,
      "step": 10025
    },
    {
      "entropy": 5.654903221130371,
      "epoch": 0.9954346963080588,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004909639788638079,
      "loss": 5.4189,
      "mean_token_accuracy": 0.15256367325782777,
      "num_tokens": 21971321.0,
      "step": 10030
    },
    {
      "entropy": 5.534298658370972,
      "epoch": 0.9959309249702263,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004909540359397477,
      "loss": 5.2701,
      "mean_token_accuracy": 0.16025426238775253,
      "num_tokens": 21981252.0,
      "step": 10035
    },
    {
      "entropy": 5.559725046157837,
      "epoch": 0.9964271536323938,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004909440876604929,
      "loss": 5.2782,
      "mean_token_accuracy": 0.15571915805339814,
      "num_tokens": 21991820.0,
      "step": 10040
    },
    {
      "entropy": 5.531323003768921,
      "epoch": 0.9969233822945613,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00049093413402629,
      "loss": 5.2442,
      "mean_token_accuracy": 0.16417673975229263,
      "num_tokens": 22003619.0,
      "step": 10045
    },
    {
      "entropy": 5.542947816848755,
      "epoch": 0.9974196109567288,
      "grad_norm": 1.0,
      "learning_rate": 0.0004909241750373858,
      "loss": 5.2973,
      "mean_token_accuracy": 0.15834089517593383,
      "num_tokens": 22016116.0,
      "step": 10050
    },
    {
      "entropy": 5.580583333969116,
      "epoch": 0.9979158396188964,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004909142106940275,
      "loss": 5.2283,
      "mean_token_accuracy": 0.1644928440451622,
      "num_tokens": 22026115.0,
      "step": 10055
    },
    {
      "entropy": 5.565815401077271,
      "epoch": 0.9984120682810639,
      "grad_norm": 0.9375,
      "learning_rate": 0.000490904240996462,
      "loss": 5.2981,
      "mean_token_accuracy": 0.16727686673402786,
      "num_tokens": 22038889.0,
      "step": 10060
    },
    {
      "entropy": 5.524577999114991,
      "epoch": 0.9989082969432315,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004908942659449364,
      "loss": 5.2504,
      "mean_token_accuracy": 0.1659238651394844,
      "num_tokens": 22050773.0,
      "step": 10065
    },
    {
      "entropy": 5.491831970214844,
      "epoch": 0.9994045256053989,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004908842855396982,
      "loss": 5.2598,
      "mean_token_accuracy": 0.15760456025600433,
      "num_tokens": 22062869.0,
      "step": 10070
    },
    {
      "entropy": 5.530044889450073,
      "epoch": 0.9999007542675665,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004908742997809951,
      "loss": 5.2714,
      "mean_token_accuracy": 0.16301410943269729,
      "num_tokens": 22073898.0,
      "step": 10075
    },
    {
      "entropy": 5.574612617492676,
      "epoch": 1.0003969829297341,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004908643086690744,
      "loss": 5.1844,
      "mean_token_accuracy": 0.16193026006221772,
      "num_tokens": 22084383.0,
      "step": 10080
    },
    {
      "entropy": 5.51653642654419,
      "epoch": 1.0008932115919016,
      "grad_norm": 0.921875,
      "learning_rate": 0.000490854312204184,
      "loss": 5.2201,
      "mean_token_accuracy": 0.16356946527957916,
      "num_tokens": 22094990.0,
      "step": 10085
    },
    {
      "entropy": 5.531592988967896,
      "epoch": 1.001389440254069,
      "grad_norm": 0.875,
      "learning_rate": 0.0004908443103865719,
      "loss": 5.2525,
      "mean_token_accuracy": 0.16223589330911636,
      "num_tokens": 22105259.0,
      "step": 10090
    },
    {
      "entropy": 5.529381132125854,
      "epoch": 1.0018856689162365,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004908343032164859,
      "loss": 5.218,
      "mean_token_accuracy": 0.16315143704414367,
      "num_tokens": 22116348.0,
      "step": 10095
    },
    {
      "entropy": 5.544572830200195,
      "epoch": 1.0023818975784042,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004908242906941743,
      "loss": 5.2082,
      "mean_token_accuracy": 0.16781669706106186,
      "num_tokens": 22127311.0,
      "step": 10100
    },
    {
      "entropy": 5.584514999389649,
      "epoch": 1.0028781262405717,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004908142728198853,
      "loss": 5.2308,
      "mean_token_accuracy": 0.16152353584766388,
      "num_tokens": 22137929.0,
      "step": 10105
    },
    {
      "entropy": 5.626436710357666,
      "epoch": 1.0033743549027392,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004908042495938673,
      "loss": 5.3409,
      "mean_token_accuracy": 0.16043704748153687,
      "num_tokens": 22148202.0,
      "step": 10110
    },
    {
      "entropy": 5.509261560440064,
      "epoch": 1.0038705835649067,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004907942210163692,
      "loss": 5.139,
      "mean_token_accuracy": 0.16677588373422622,
      "num_tokens": 22158955.0,
      "step": 10115
    },
    {
      "entropy": 5.517710399627686,
      "epoch": 1.0043668122270741,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004907841870876392,
      "loss": 5.1703,
      "mean_token_accuracy": 0.16501835733652115,
      "num_tokens": 22169569.0,
      "step": 10120
    },
    {
      "entropy": 5.489158821105957,
      "epoch": 1.0048630408892418,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004907741478079263,
      "loss": 5.1477,
      "mean_token_accuracy": 0.16425950676202775,
      "num_tokens": 22180171.0,
      "step": 10125
    },
    {
      "entropy": 5.505987215042114,
      "epoch": 1.0053592695514093,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004907641031774796,
      "loss": 5.2712,
      "mean_token_accuracy": 0.16013553142547607,
      "num_tokens": 22190531.0,
      "step": 10130
    },
    {
      "entropy": 5.506381320953369,
      "epoch": 1.0058554982135768,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000490754053196548,
      "loss": 5.1612,
      "mean_token_accuracy": 0.16621319204568863,
      "num_tokens": 22201495.0,
      "step": 10135
    },
    {
      "entropy": 5.585003519058228,
      "epoch": 1.0063517268757443,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004907439978653809,
      "loss": 5.2439,
      "mean_token_accuracy": 0.15559897422790528,
      "num_tokens": 22212941.0,
      "step": 10140
    },
    {
      "entropy": 5.596948575973511,
      "epoch": 1.006847955537912,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004907339371842275,
      "loss": 5.2728,
      "mean_token_accuracy": 0.1628827027976513,
      "num_tokens": 22225212.0,
      "step": 10145
    },
    {
      "entropy": 5.503411531448364,
      "epoch": 1.0073441842000794,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004907238711533373,
      "loss": 5.2133,
      "mean_token_accuracy": 0.1630281016230583,
      "num_tokens": 22236370.0,
      "step": 10150
    },
    {
      "entropy": 5.5146815299987795,
      "epoch": 1.007840412862247,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004907137997729599,
      "loss": 5.1852,
      "mean_token_accuracy": 0.16718839406967162,
      "num_tokens": 22247641.0,
      "step": 10155
    },
    {
      "entropy": 5.473132514953614,
      "epoch": 1.0083366415244144,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000490703723043345,
      "loss": 5.2471,
      "mean_token_accuracy": 0.166613307595253,
      "num_tokens": 22258599.0,
      "step": 10160
    },
    {
      "entropy": 5.531865882873535,
      "epoch": 1.008832870186582,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004906936409647428,
      "loss": 5.1721,
      "mean_token_accuracy": 0.1649941012263298,
      "num_tokens": 22270913.0,
      "step": 10165
    },
    {
      "entropy": 5.511922311782837,
      "epoch": 1.0093290988487496,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000490683553537403,
      "loss": 5.0885,
      "mean_token_accuracy": 0.16907574087381363,
      "num_tokens": 22281053.0,
      "step": 10170
    },
    {
      "entropy": 5.4852986335754395,
      "epoch": 1.009825327510917,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004906734607615758,
      "loss": 5.0978,
      "mean_token_accuracy": 0.1731666073203087,
      "num_tokens": 22292570.0,
      "step": 10175
    },
    {
      "entropy": 5.496685314178467,
      "epoch": 1.0103215561730845,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004906633626375114,
      "loss": 5.1819,
      "mean_token_accuracy": 0.16467416435480117,
      "num_tokens": 22303548.0,
      "step": 10180
    },
    {
      "entropy": 5.523727798461914,
      "epoch": 1.010817784835252,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004906532591654604,
      "loss": 5.2105,
      "mean_token_accuracy": 0.16738889664411544,
      "num_tokens": 22314133.0,
      "step": 10185
    },
    {
      "entropy": 5.545677757263183,
      "epoch": 1.0113140134974197,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004906431503456733,
      "loss": 5.1641,
      "mean_token_accuracy": 0.1664816200733185,
      "num_tokens": 22326431.0,
      "step": 10190
    },
    {
      "entropy": 5.537136840820312,
      "epoch": 1.0118102421595871,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004906330361784007,
      "loss": 5.218,
      "mean_token_accuracy": 0.16480584293603898,
      "num_tokens": 22338507.0,
      "step": 10195
    },
    {
      "entropy": 5.5465387344360355,
      "epoch": 1.0123064708217546,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004906229166638936,
      "loss": 5.1878,
      "mean_token_accuracy": 0.16493188440799714,
      "num_tokens": 22348863.0,
      "step": 10200
    },
    {
      "entropy": 5.474705839157105,
      "epoch": 1.012802699483922,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004906127918024026,
      "loss": 5.139,
      "mean_token_accuracy": 0.1639999732375145,
      "num_tokens": 22359000.0,
      "step": 10205
    },
    {
      "entropy": 5.56318154335022,
      "epoch": 1.0132989281460898,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004906026615941791,
      "loss": 5.2609,
      "mean_token_accuracy": 0.16299256980419158,
      "num_tokens": 22369539.0,
      "step": 10210
    },
    {
      "entropy": 5.552541208267212,
      "epoch": 1.0137951568082573,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004905925260394741,
      "loss": 5.2348,
      "mean_token_accuracy": 0.163515704870224,
      "num_tokens": 22380700.0,
      "step": 10215
    },
    {
      "entropy": 5.525241994857788,
      "epoch": 1.0142913854704247,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000490582385138539,
      "loss": 5.211,
      "mean_token_accuracy": 0.1586480900645256,
      "num_tokens": 22392151.0,
      "step": 10220
    },
    {
      "entropy": 5.4952363014221195,
      "epoch": 1.0147876141325922,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004905722388916254,
      "loss": 5.188,
      "mean_token_accuracy": 0.17093088775873183,
      "num_tokens": 22401735.0,
      "step": 10225
    },
    {
      "entropy": 5.524148988723755,
      "epoch": 1.01528384279476,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004905620872989846,
      "loss": 5.2472,
      "mean_token_accuracy": 0.1665743976831436,
      "num_tokens": 22412549.0,
      "step": 10230
    },
    {
      "entropy": 5.612970018386841,
      "epoch": 1.0157800714569274,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004905519303608687,
      "loss": 5.2171,
      "mean_token_accuracy": 0.16105247288942337,
      "num_tokens": 22424728.0,
      "step": 10235
    },
    {
      "entropy": 5.472014379501343,
      "epoch": 1.0162763001190949,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004905417680775294,
      "loss": 5.1082,
      "mean_token_accuracy": 0.16330609768629073,
      "num_tokens": 22434197.0,
      "step": 10240
    },
    {
      "entropy": 5.578816938400268,
      "epoch": 1.0167725287812623,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004905316004492186,
      "loss": 5.2974,
      "mean_token_accuracy": 0.15572961866855622,
      "num_tokens": 22444671.0,
      "step": 10245
    },
    {
      "entropy": 5.495514726638794,
      "epoch": 1.01726875744343,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004905214274761885,
      "loss": 5.1684,
      "mean_token_accuracy": 0.1653987407684326,
      "num_tokens": 22456636.0,
      "step": 10250
    },
    {
      "entropy": 5.4824018478393555,
      "epoch": 1.0177649861055975,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004905112491586915,
      "loss": 5.1076,
      "mean_token_accuracy": 0.17306455075740815,
      "num_tokens": 22467011.0,
      "step": 10255
    },
    {
      "entropy": 5.50503191947937,
      "epoch": 1.018261214767765,
      "grad_norm": 0.96875,
      "learning_rate": 0.00049050106549698,
      "loss": 5.2013,
      "mean_token_accuracy": 0.16318316012620926,
      "num_tokens": 22477630.0,
      "step": 10260
    },
    {
      "entropy": 5.486586093902588,
      "epoch": 1.0187574434299325,
      "grad_norm": 1.0,
      "learning_rate": 0.0004904908764913063,
      "loss": 5.1528,
      "mean_token_accuracy": 0.17257197499275206,
      "num_tokens": 22488132.0,
      "step": 10265
    },
    {
      "entropy": 5.468431901931763,
      "epoch": 1.0192536720921,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004904806821419233,
      "loss": 5.2124,
      "mean_token_accuracy": 0.16477933526039124,
      "num_tokens": 22499430.0,
      "step": 10270
    },
    {
      "entropy": 5.486281061172486,
      "epoch": 1.0197499007542676,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004904704824490835,
      "loss": 5.1387,
      "mean_token_accuracy": 0.16794027537107467,
      "num_tokens": 22509728.0,
      "step": 10275
    },
    {
      "entropy": 5.530235671997071,
      "epoch": 1.020246129416435,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004904602774130402,
      "loss": 5.2485,
      "mean_token_accuracy": 0.15583620667457582,
      "num_tokens": 22520955.0,
      "step": 10280
    },
    {
      "entropy": 5.578283500671387,
      "epoch": 1.0207423580786026,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004904500670340463,
      "loss": 5.2563,
      "mean_token_accuracy": 0.15621759369969368,
      "num_tokens": 22530440.0,
      "step": 10285
    },
    {
      "entropy": 5.51908540725708,
      "epoch": 1.02123858674077,
      "grad_norm": 0.859375,
      "learning_rate": 0.000490439851312355,
      "loss": 5.2293,
      "mean_token_accuracy": 0.17005994617938996,
      "num_tokens": 22541539.0,
      "step": 10290
    },
    {
      "entropy": 5.478838062286377,
      "epoch": 1.0217348154029378,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004904296302482196,
      "loss": 5.2226,
      "mean_token_accuracy": 0.164718359708786,
      "num_tokens": 22555717.0,
      "step": 10295
    },
    {
      "entropy": 5.59019660949707,
      "epoch": 1.0222310440651052,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004904194038418935,
      "loss": 5.3002,
      "mean_token_accuracy": 0.15941984802484513,
      "num_tokens": 22566938.0,
      "step": 10300
    },
    {
      "entropy": 5.628157997131348,
      "epoch": 1.0227272727272727,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004904091720936305,
      "loss": 5.1681,
      "mean_token_accuracy": 0.16496185809373856,
      "num_tokens": 22578725.0,
      "step": 10305
    },
    {
      "entropy": 5.531756830215454,
      "epoch": 1.0232235013894402,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004903989350036841,
      "loss": 5.0983,
      "mean_token_accuracy": 0.17594655752182006,
      "num_tokens": 22588476.0,
      "step": 10310
    },
    {
      "entropy": 5.53554630279541,
      "epoch": 1.0237197300516079,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004903886925723083,
      "loss": 5.3216,
      "mean_token_accuracy": 0.1539576232433319,
      "num_tokens": 22600294.0,
      "step": 10315
    },
    {
      "entropy": 5.569748830795288,
      "epoch": 1.0242159587137754,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000490378444799757,
      "loss": 5.198,
      "mean_token_accuracy": 0.16748596355319023,
      "num_tokens": 22610578.0,
      "step": 10320
    },
    {
      "entropy": 5.591857004165649,
      "epoch": 1.0247121873759428,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004903681916862844,
      "loss": 5.2055,
      "mean_token_accuracy": 0.16512291729450226,
      "num_tokens": 22620988.0,
      "step": 10325
    },
    {
      "entropy": 5.520949745178223,
      "epoch": 1.0252084160381103,
      "grad_norm": 0.875,
      "learning_rate": 0.0004903579332321447,
      "loss": 5.1952,
      "mean_token_accuracy": 0.1655699610710144,
      "num_tokens": 22632198.0,
      "step": 10330
    },
    {
      "entropy": 5.524732303619385,
      "epoch": 1.0257046447002778,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004903476694375924,
      "loss": 5.201,
      "mean_token_accuracy": 0.17436613887548447,
      "num_tokens": 22642543.0,
      "step": 10335
    },
    {
      "entropy": 5.542492485046386,
      "epoch": 1.0262008733624455,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004903374003028819,
      "loss": 5.2051,
      "mean_token_accuracy": 0.16694827228784562,
      "num_tokens": 22655453.0,
      "step": 10340
    },
    {
      "entropy": 5.478806352615356,
      "epoch": 1.026697102024613,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004903271258282678,
      "loss": 5.1442,
      "mean_token_accuracy": 0.17118590772151948,
      "num_tokens": 22667342.0,
      "step": 10345
    },
    {
      "entropy": 5.581794404983521,
      "epoch": 1.0271933306867804,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004903168460140049,
      "loss": 5.2131,
      "mean_token_accuracy": 0.17006220370531083,
      "num_tokens": 22676890.0,
      "step": 10350
    },
    {
      "entropy": 5.539871406555176,
      "epoch": 1.027689559348948,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004903065608603484,
      "loss": 5.2111,
      "mean_token_accuracy": 0.16357226967811583,
      "num_tokens": 22687809.0,
      "step": 10355
    },
    {
      "entropy": 5.488600587844848,
      "epoch": 1.0281857880111156,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000490296270367553,
      "loss": 5.224,
      "mean_token_accuracy": 0.16250108629465104,
      "num_tokens": 22697938.0,
      "step": 10360
    },
    {
      "entropy": 5.61170334815979,
      "epoch": 1.028682016673283,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004902859745358741,
      "loss": 5.2199,
      "mean_token_accuracy": 0.17221857160329818,
      "num_tokens": 22708303.0,
      "step": 10365
    },
    {
      "entropy": 5.50864953994751,
      "epoch": 1.0291782453354505,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004902756733655668,
      "loss": 5.1535,
      "mean_token_accuracy": 0.1687371775507927,
      "num_tokens": 22717987.0,
      "step": 10370
    },
    {
      "entropy": 5.512455415725708,
      "epoch": 1.029674473997618,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004902653668568868,
      "loss": 5.2313,
      "mean_token_accuracy": 0.16371186822652817,
      "num_tokens": 22729672.0,
      "step": 10375
    },
    {
      "entropy": 5.576651382446289,
      "epoch": 1.0301707026597857,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004902550550100896,
      "loss": 5.1842,
      "mean_token_accuracy": 0.17124547809362411,
      "num_tokens": 22741296.0,
      "step": 10380
    },
    {
      "entropy": 5.529744672775268,
      "epoch": 1.0306669313219532,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004902447378254307,
      "loss": 5.1767,
      "mean_token_accuracy": 0.1712263137102127,
      "num_tokens": 22752337.0,
      "step": 10385
    },
    {
      "entropy": 5.553467845916748,
      "epoch": 1.0311631599841207,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004902344153031663,
      "loss": 5.2157,
      "mean_token_accuracy": 0.16358236074447632,
      "num_tokens": 22762681.0,
      "step": 10390
    },
    {
      "entropy": 5.600700044631958,
      "epoch": 1.0316593886462881,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000490224087443552,
      "loss": 5.2636,
      "mean_token_accuracy": 0.15973918735980988,
      "num_tokens": 22774275.0,
      "step": 10395
    },
    {
      "entropy": 5.502658748626709,
      "epoch": 1.0321556173084558,
      "grad_norm": 1.0,
      "learning_rate": 0.0004902137542468442,
      "loss": 5.1395,
      "mean_token_accuracy": 0.17113078981637955,
      "num_tokens": 22785944.0,
      "step": 10400
    },
    {
      "entropy": 5.451526737213134,
      "epoch": 1.0326518459706233,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004902034157132991,
      "loss": 5.0633,
      "mean_token_accuracy": 0.16619019359350204,
      "num_tokens": 22796543.0,
      "step": 10405
    },
    {
      "entropy": 5.454790830612183,
      "epoch": 1.0331480746327908,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004901930718431729,
      "loss": 5.1876,
      "mean_token_accuracy": 0.16527335494756698,
      "num_tokens": 22807957.0,
      "step": 10410
    },
    {
      "entropy": 5.563516902923584,
      "epoch": 1.0336443032949583,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004901827226367223,
      "loss": 5.2828,
      "mean_token_accuracy": 0.16433306634426117,
      "num_tokens": 22818903.0,
      "step": 10415
    },
    {
      "entropy": 5.579446649551391,
      "epoch": 1.0341405319571257,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004901723680942038,
      "loss": 5.3108,
      "mean_token_accuracy": 0.16274386048316955,
      "num_tokens": 22829286.0,
      "step": 10420
    },
    {
      "entropy": 5.599664211273193,
      "epoch": 1.0346367606192934,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004901620082158742,
      "loss": 5.2202,
      "mean_token_accuracy": 0.16135814189910888,
      "num_tokens": 22839860.0,
      "step": 10425
    },
    {
      "entropy": 5.625802993774414,
      "epoch": 1.035132989281461,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004901516430019905,
      "loss": 5.2771,
      "mean_token_accuracy": 0.15363107174634932,
      "num_tokens": 22852238.0,
      "step": 10430
    },
    {
      "entropy": 5.5464558601379395,
      "epoch": 1.0356292179436284,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004901412724528097,
      "loss": 5.2357,
      "mean_token_accuracy": 0.1671655386686325,
      "num_tokens": 22863013.0,
      "step": 10435
    },
    {
      "entropy": 5.477605962753296,
      "epoch": 1.0361254466057959,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004901308965685889,
      "loss": 5.0898,
      "mean_token_accuracy": 0.166864375770092,
      "num_tokens": 22874233.0,
      "step": 10440
    },
    {
      "entropy": 5.533360528945923,
      "epoch": 1.0366216752679636,
      "grad_norm": 0.875,
      "learning_rate": 0.0004901205153495853,
      "loss": 5.273,
      "mean_token_accuracy": 0.16261522918939592,
      "num_tokens": 22885313.0,
      "step": 10445
    },
    {
      "entropy": 5.524601840972901,
      "epoch": 1.037117903930131,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004901101287960566,
      "loss": 5.1543,
      "mean_token_accuracy": 0.16824844479560852,
      "num_tokens": 22895365.0,
      "step": 10450
    },
    {
      "entropy": 5.499677896499634,
      "epoch": 1.0376141325922985,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004900997369082602,
      "loss": 5.1551,
      "mean_token_accuracy": 0.16431809812784196,
      "num_tokens": 22906392.0,
      "step": 10455
    },
    {
      "entropy": 5.454514503479004,
      "epoch": 1.038110361254466,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004900893396864539,
      "loss": 5.1099,
      "mean_token_accuracy": 0.17248469442129136,
      "num_tokens": 22916760.0,
      "step": 10460
    },
    {
      "entropy": 5.541943502426148,
      "epoch": 1.0386065899166337,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004900789371308953,
      "loss": 5.1727,
      "mean_token_accuracy": 0.16550545692443847,
      "num_tokens": 22928070.0,
      "step": 10465
    },
    {
      "entropy": 5.448971796035766,
      "epoch": 1.0391028185788012,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004900685292418425,
      "loss": 5.1535,
      "mean_token_accuracy": 0.1698997437953949,
      "num_tokens": 22938815.0,
      "step": 10470
    },
    {
      "entropy": 5.519144487380982,
      "epoch": 1.0395990472409686,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004900581160195537,
      "loss": 5.228,
      "mean_token_accuracy": 0.15901804566383362,
      "num_tokens": 22949806.0,
      "step": 10475
    },
    {
      "entropy": 5.454152297973633,
      "epoch": 1.040095275903136,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000490047697464287,
      "loss": 5.0665,
      "mean_token_accuracy": 0.16806019246578216,
      "num_tokens": 22960880.0,
      "step": 10480
    },
    {
      "entropy": 5.5343646049499515,
      "epoch": 1.0405915045653038,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004900372735763009,
      "loss": 5.2213,
      "mean_token_accuracy": 0.16154549568891524,
      "num_tokens": 22972828.0,
      "step": 10485
    },
    {
      "entropy": 5.451268148422241,
      "epoch": 1.0410877332274713,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004900268443558536,
      "loss": 5.1264,
      "mean_token_accuracy": 0.1684920996427536,
      "num_tokens": 22981869.0,
      "step": 10490
    },
    {
      "entropy": 5.549597644805909,
      "epoch": 1.0415839618896388,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000490016409803204,
      "loss": 5.2251,
      "mean_token_accuracy": 0.15966662913560867,
      "num_tokens": 22993111.0,
      "step": 10495
    },
    {
      "entropy": 5.527101325988769,
      "epoch": 1.0420801905518062,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004900059699186108,
      "loss": 5.208,
      "mean_token_accuracy": 0.1624673679471016,
      "num_tokens": 23003115.0,
      "step": 10500
    },
    {
      "entropy": 5.56898045539856,
      "epoch": 1.0425764192139737,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004899955247023328,
      "loss": 5.2478,
      "mean_token_accuracy": 0.16275709122419357,
      "num_tokens": 23012587.0,
      "step": 10505
    },
    {
      "entropy": 5.5545085906982425,
      "epoch": 1.0430726478761414,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000489985074154629,
      "loss": 5.2612,
      "mean_token_accuracy": 0.16268475502729415,
      "num_tokens": 23024455.0,
      "step": 10510
    },
    {
      "entropy": 5.506721019744873,
      "epoch": 1.0435688765383089,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004899746182757587,
      "loss": 5.1938,
      "mean_token_accuracy": 0.16569384187459946,
      "num_tokens": 23034430.0,
      "step": 10515
    },
    {
      "entropy": 5.547843074798584,
      "epoch": 1.0440651052004764,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004899641570659811,
      "loss": 5.2456,
      "mean_token_accuracy": 0.16971434727311135,
      "num_tokens": 23044811.0,
      "step": 10520
    },
    {
      "entropy": 5.573384952545166,
      "epoch": 1.0445613338626438,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004899536905255555,
      "loss": 5.2911,
      "mean_token_accuracy": 0.1627610817551613,
      "num_tokens": 23056179.0,
      "step": 10525
    },
    {
      "entropy": 5.552228784561157,
      "epoch": 1.0450575625248115,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004899432186547416,
      "loss": 5.2078,
      "mean_token_accuracy": 0.1670810952782631,
      "num_tokens": 23068331.0,
      "step": 10530
    },
    {
      "entropy": 5.5211748600006105,
      "epoch": 1.045553791186979,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004899327414537991,
      "loss": 5.1866,
      "mean_token_accuracy": 0.16346872746944427,
      "num_tokens": 23079213.0,
      "step": 10535
    },
    {
      "entropy": 5.63276047706604,
      "epoch": 1.0460500198491465,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004899222589229877,
      "loss": 5.2932,
      "mean_token_accuracy": 0.1583160325884819,
      "num_tokens": 23089860.0,
      "step": 10540
    },
    {
      "entropy": 5.4699718952178955,
      "epoch": 1.046546248511314,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004899117710625674,
      "loss": 5.1626,
      "mean_token_accuracy": 0.17572366297245026,
      "num_tokens": 23099431.0,
      "step": 10545
    },
    {
      "entropy": 5.507458305358886,
      "epoch": 1.0470424771734816,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004899012778727983,
      "loss": 5.228,
      "mean_token_accuracy": 0.1616830661892891,
      "num_tokens": 23110947.0,
      "step": 10550
    },
    {
      "entropy": 5.591244840621949,
      "epoch": 1.0475387058356491,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004898907793539406,
      "loss": 5.3145,
      "mean_token_accuracy": 0.16271693259477615,
      "num_tokens": 23122782.0,
      "step": 10555
    },
    {
      "entropy": 5.6101878643035885,
      "epoch": 1.0480349344978166,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004898802755062546,
      "loss": 5.2243,
      "mean_token_accuracy": 0.164697103202343,
      "num_tokens": 23134254.0,
      "step": 10560
    },
    {
      "entropy": 5.476492500305175,
      "epoch": 1.048531163159984,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004898697663300007,
      "loss": 5.0673,
      "mean_token_accuracy": 0.17472380995750428,
      "num_tokens": 23143851.0,
      "step": 10565
    },
    {
      "entropy": 5.486402606964111,
      "epoch": 1.0490273918221515,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004898592518254398,
      "loss": 5.1824,
      "mean_token_accuracy": 0.16846754848957063,
      "num_tokens": 23155966.0,
      "step": 10570
    },
    {
      "entropy": 5.50768814086914,
      "epoch": 1.0495236204843192,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004898487319928322,
      "loss": 5.1788,
      "mean_token_accuracy": 0.16355437636375428,
      "num_tokens": 23165896.0,
      "step": 10575
    },
    {
      "entropy": 5.493472671508789,
      "epoch": 1.0500198491464867,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004898382068324392,
      "loss": 5.1764,
      "mean_token_accuracy": 0.16505965292453767,
      "num_tokens": 23177821.0,
      "step": 10580
    },
    {
      "entropy": 5.52282304763794,
      "epoch": 1.0505160778086542,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004898276763445215,
      "loss": 5.2375,
      "mean_token_accuracy": 0.16412696093320847,
      "num_tokens": 23188916.0,
      "step": 10585
    },
    {
      "entropy": 5.607707548141479,
      "epoch": 1.0510123064708217,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004898171405293404,
      "loss": 5.2358,
      "mean_token_accuracy": 0.16866102665662766,
      "num_tokens": 23200490.0,
      "step": 10590
    },
    {
      "entropy": 5.484441757202148,
      "epoch": 1.0515085351329894,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004898065993871571,
      "loss": 5.1784,
      "mean_token_accuracy": 0.1647914916276932,
      "num_tokens": 23212991.0,
      "step": 10595
    },
    {
      "entropy": 5.551609706878662,
      "epoch": 1.0520047637951568,
      "grad_norm": 0.9375,
      "learning_rate": 0.000489796052918233,
      "loss": 5.2828,
      "mean_token_accuracy": 0.15930335968732834,
      "num_tokens": 23223361.0,
      "step": 10600
    },
    {
      "entropy": 5.553236389160157,
      "epoch": 1.0525009924573243,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004897855011228295,
      "loss": 5.1785,
      "mean_token_accuracy": 0.16796258538961412,
      "num_tokens": 23233765.0,
      "step": 10605
    },
    {
      "entropy": 5.492867755889892,
      "epoch": 1.0529972211194918,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004897749440012085,
      "loss": 5.1303,
      "mean_token_accuracy": 0.17219723612070084,
      "num_tokens": 23246562.0,
      "step": 10610
    },
    {
      "entropy": 5.469007396697998,
      "epoch": 1.0534934497816595,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004897643815536316,
      "loss": 5.1474,
      "mean_token_accuracy": 0.1698348864912987,
      "num_tokens": 23256648.0,
      "step": 10615
    },
    {
      "entropy": 5.5521628856658936,
      "epoch": 1.053989678443827,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000489753813780361,
      "loss": 5.2127,
      "mean_token_accuracy": 0.16471905261278152,
      "num_tokens": 23267428.0,
      "step": 10620
    },
    {
      "entropy": 5.521107912063599,
      "epoch": 1.0544859071059944,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004897432406816585,
      "loss": 5.1463,
      "mean_token_accuracy": 0.16491916477680207,
      "num_tokens": 23278440.0,
      "step": 10625
    },
    {
      "entropy": 5.505084609985351,
      "epoch": 1.054982135768162,
      "grad_norm": 0.875,
      "learning_rate": 0.0004897326622577863,
      "loss": 5.1963,
      "mean_token_accuracy": 0.16055031418800353,
      "num_tokens": 23290571.0,
      "step": 10630
    },
    {
      "entropy": 5.5806097984313965,
      "epoch": 1.0554783644303294,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004897220785090067,
      "loss": 5.2633,
      "mean_token_accuracy": 0.16173502206802368,
      "num_tokens": 23301634.0,
      "step": 10635
    },
    {
      "entropy": 5.478116703033447,
      "epoch": 1.055974593092497,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004897114894355824,
      "loss": 5.1385,
      "mean_token_accuracy": 0.1717913791537285,
      "num_tokens": 23313125.0,
      "step": 10640
    },
    {
      "entropy": 5.479493474960327,
      "epoch": 1.0564708217546646,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004897008950377758,
      "loss": 5.1635,
      "mean_token_accuracy": 0.16100295931100844,
      "num_tokens": 23323610.0,
      "step": 10645
    },
    {
      "entropy": 5.451137685775757,
      "epoch": 1.056967050416832,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004896902953158497,
      "loss": 5.1394,
      "mean_token_accuracy": 0.1681753695011139,
      "num_tokens": 23333652.0,
      "step": 10650
    },
    {
      "entropy": 5.476093864440918,
      "epoch": 1.0574632790789995,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004896796902700667,
      "loss": 5.1827,
      "mean_token_accuracy": 0.16579088419675828,
      "num_tokens": 23345954.0,
      "step": 10655
    },
    {
      "entropy": 5.472770929336548,
      "epoch": 1.0579595077411672,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004896690799006901,
      "loss": 5.1541,
      "mean_token_accuracy": 0.1691794514656067,
      "num_tokens": 23357052.0,
      "step": 10660
    },
    {
      "entropy": 5.495822286605835,
      "epoch": 1.0584557364033347,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004896584642079828,
      "loss": 5.1812,
      "mean_token_accuracy": 0.17066397219896318,
      "num_tokens": 23368837.0,
      "step": 10665
    },
    {
      "entropy": 5.489329147338867,
      "epoch": 1.0589519650655022,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004896478431922083,
      "loss": 5.1548,
      "mean_token_accuracy": 0.16938245445489883,
      "num_tokens": 23378608.0,
      "step": 10670
    },
    {
      "entropy": 5.544188547134399,
      "epoch": 1.0594481937276696,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004896372168536297,
      "loss": 5.2215,
      "mean_token_accuracy": 0.16721830517053604,
      "num_tokens": 23390094.0,
      "step": 10675
    },
    {
      "entropy": 5.45720853805542,
      "epoch": 1.0599444223898373,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004896265851925107,
      "loss": 5.0931,
      "mean_token_accuracy": 0.1752452790737152,
      "num_tokens": 23399991.0,
      "step": 10680
    },
    {
      "entropy": 5.539272975921631,
      "epoch": 1.0604406510520048,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004896159482091148,
      "loss": 5.2897,
      "mean_token_accuracy": 0.15691623836755753,
      "num_tokens": 23410705.0,
      "step": 10685
    },
    {
      "entropy": 5.461822128295898,
      "epoch": 1.0609368797141723,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004896053059037058,
      "loss": 5.0726,
      "mean_token_accuracy": 0.18398592472076417,
      "num_tokens": 23421939.0,
      "step": 10690
    },
    {
      "entropy": 5.515606451034546,
      "epoch": 1.0614331083763398,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004895946582765477,
      "loss": 5.2103,
      "mean_token_accuracy": 0.16736398935317992,
      "num_tokens": 23433252.0,
      "step": 10695
    },
    {
      "entropy": 5.613071775436401,
      "epoch": 1.0619293370385074,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004895840053279045,
      "loss": 5.2772,
      "mean_token_accuracy": 0.15813700258731841,
      "num_tokens": 23445582.0,
      "step": 10700
    },
    {
      "entropy": 5.571057605743408,
      "epoch": 1.062425565700675,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004895733470580404,
      "loss": 5.2339,
      "mean_token_accuracy": 0.1600451797246933,
      "num_tokens": 23456360.0,
      "step": 10705
    },
    {
      "entropy": 5.524703931808472,
      "epoch": 1.0629217943628424,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004895626834672196,
      "loss": 5.1241,
      "mean_token_accuracy": 0.16391288936138154,
      "num_tokens": 23466576.0,
      "step": 10710
    },
    {
      "entropy": 5.421665954589844,
      "epoch": 1.0634180230250099,
      "grad_norm": 1.0,
      "learning_rate": 0.0004895520145557066,
      "loss": 5.0801,
      "mean_token_accuracy": 0.17319126725196837,
      "num_tokens": 23476200.0,
      "step": 10715
    },
    {
      "entropy": 5.533837461471558,
      "epoch": 1.0639142516871773,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004895413403237659,
      "loss": 5.2805,
      "mean_token_accuracy": 0.16643560528755189,
      "num_tokens": 23487776.0,
      "step": 10720
    },
    {
      "entropy": 5.519957256317139,
      "epoch": 1.064410480349345,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004895306607716624,
      "loss": 5.1327,
      "mean_token_accuracy": 0.1732277974486351,
      "num_tokens": 23497517.0,
      "step": 10725
    },
    {
      "entropy": 5.5163506984710695,
      "epoch": 1.0649067090115125,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004895199758996607,
      "loss": 5.1638,
      "mean_token_accuracy": 0.17083370983600615,
      "num_tokens": 23507436.0,
      "step": 10730
    },
    {
      "entropy": 5.479664707183838,
      "epoch": 1.06540293767368,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004895092857080258,
      "loss": 5.1094,
      "mean_token_accuracy": 0.1597520738840103,
      "num_tokens": 23517170.0,
      "step": 10735
    },
    {
      "entropy": 5.479441976547241,
      "epoch": 1.0658991663358475,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004894985901970229,
      "loss": 5.1174,
      "mean_token_accuracy": 0.173440021276474,
      "num_tokens": 23527329.0,
      "step": 10740
    },
    {
      "entropy": 5.561740398406982,
      "epoch": 1.0663953949980152,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004894878893669171,
      "loss": 5.2226,
      "mean_token_accuracy": 0.16073554754257202,
      "num_tokens": 23538341.0,
      "step": 10745
    },
    {
      "entropy": 5.505781459808349,
      "epoch": 1.0668916236601826,
      "grad_norm": 0.875,
      "learning_rate": 0.0004894771832179739,
      "loss": 5.1448,
      "mean_token_accuracy": 0.16851148307323455,
      "num_tokens": 23549522.0,
      "step": 10750
    },
    {
      "entropy": 5.569004535675049,
      "epoch": 1.0673878523223501,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004894664717504587,
      "loss": 5.186,
      "mean_token_accuracy": 0.1654226675629616,
      "num_tokens": 23559274.0,
      "step": 10755
    },
    {
      "entropy": 5.525196743011475,
      "epoch": 1.0678840809845176,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004894557549646371,
      "loss": 5.1539,
      "mean_token_accuracy": 0.17346994131803511,
      "num_tokens": 23570852.0,
      "step": 10760
    },
    {
      "entropy": 5.4818190574646,
      "epoch": 1.0683803096466853,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004894450328607749,
      "loss": 5.1488,
      "mean_token_accuracy": 0.16973407864570617,
      "num_tokens": 23582918.0,
      "step": 10765
    },
    {
      "entropy": 5.473199987411499,
      "epoch": 1.0688765383088528,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004894343054391379,
      "loss": 5.1838,
      "mean_token_accuracy": 0.1745926633477211,
      "num_tokens": 23594177.0,
      "step": 10770
    },
    {
      "entropy": 5.504260063171387,
      "epoch": 1.0693727669710202,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004894235726999924,
      "loss": 5.1418,
      "mean_token_accuracy": 0.17183480858802797,
      "num_tokens": 23605901.0,
      "step": 10775
    },
    {
      "entropy": 5.4763895034790036,
      "epoch": 1.0698689956331877,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004894128346436042,
      "loss": 5.1963,
      "mean_token_accuracy": 0.1675628513097763,
      "num_tokens": 23617354.0,
      "step": 10780
    },
    {
      "entropy": 5.522255086898804,
      "epoch": 1.0703652242953554,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004894020912702397,
      "loss": 5.221,
      "mean_token_accuracy": 0.1643326163291931,
      "num_tokens": 23627603.0,
      "step": 10785
    },
    {
      "entropy": 5.440590858459473,
      "epoch": 1.0708614529575229,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004893913425801653,
      "loss": 5.0436,
      "mean_token_accuracy": 0.17594035416841508,
      "num_tokens": 23637280.0,
      "step": 10790
    },
    {
      "entropy": 5.517892932891845,
      "epoch": 1.0713576816196904,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004893805885736477,
      "loss": 5.2475,
      "mean_token_accuracy": 0.16413678079843522,
      "num_tokens": 23649326.0,
      "step": 10795
    },
    {
      "entropy": 5.568311405181885,
      "epoch": 1.0718539102818578,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004893698292509534,
      "loss": 5.2488,
      "mean_token_accuracy": 0.1645419031381607,
      "num_tokens": 23659561.0,
      "step": 10800
    },
    {
      "entropy": 5.518482828140259,
      "epoch": 1.0723501389440253,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004893590646123492,
      "loss": 5.1611,
      "mean_token_accuracy": 0.1690828889608383,
      "num_tokens": 23673206.0,
      "step": 10805
    },
    {
      "entropy": 5.491669321060181,
      "epoch": 1.072846367606193,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004893482946581022,
      "loss": 5.1949,
      "mean_token_accuracy": 0.16448067724704743,
      "num_tokens": 23683920.0,
      "step": 10810
    },
    {
      "entropy": 5.461504507064819,
      "epoch": 1.0733425962683605,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004893375193884793,
      "loss": 5.1147,
      "mean_token_accuracy": 0.17293360084295273,
      "num_tokens": 23693990.0,
      "step": 10815
    },
    {
      "entropy": 5.5457727909088135,
      "epoch": 1.073838824930528,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004893267388037478,
      "loss": 5.2244,
      "mean_token_accuracy": 0.16552564948797227,
      "num_tokens": 23704870.0,
      "step": 10820
    },
    {
      "entropy": 5.45700101852417,
      "epoch": 1.0743350535926954,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000489315952904175,
      "loss": 5.2099,
      "mean_token_accuracy": 0.16702677607536315,
      "num_tokens": 23715159.0,
      "step": 10825
    },
    {
      "entropy": 5.489004421234131,
      "epoch": 1.0748312822548631,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004893051616900283,
      "loss": 5.138,
      "mean_token_accuracy": 0.16661283373832703,
      "num_tokens": 23725252.0,
      "step": 10830
    },
    {
      "entropy": 5.484555721282959,
      "epoch": 1.0753275109170306,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004892943651615756,
      "loss": 5.1542,
      "mean_token_accuracy": 0.17478515952825546,
      "num_tokens": 23736472.0,
      "step": 10835
    },
    {
      "entropy": 5.500439023971557,
      "epoch": 1.075823739579198,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004892835633190843,
      "loss": 5.0729,
      "mean_token_accuracy": 0.16555219292640685,
      "num_tokens": 23747565.0,
      "step": 10840
    },
    {
      "entropy": 5.488166046142578,
      "epoch": 1.0763199682413656,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004892727561628223,
      "loss": 5.1693,
      "mean_token_accuracy": 0.16721499860286712,
      "num_tokens": 23758201.0,
      "step": 10845
    },
    {
      "entropy": 5.543577861785889,
      "epoch": 1.0768161969035333,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004892619436930577,
      "loss": 5.2183,
      "mean_token_accuracy": 0.1680361270904541,
      "num_tokens": 23769087.0,
      "step": 10850
    },
    {
      "entropy": 5.541718244552612,
      "epoch": 1.0773124255657007,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004892511259100585,
      "loss": 5.2133,
      "mean_token_accuracy": 0.16787203550338745,
      "num_tokens": 23780598.0,
      "step": 10855
    },
    {
      "entropy": 5.514758491516114,
      "epoch": 1.0778086542278682,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004892403028140932,
      "loss": 5.1403,
      "mean_token_accuracy": 0.16158319115638733,
      "num_tokens": 23791449.0,
      "step": 10860
    },
    {
      "entropy": 5.4929993629455565,
      "epoch": 1.0783048828900357,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048922947440543,
      "loss": 5.1332,
      "mean_token_accuracy": 0.17096055895090104,
      "num_tokens": 23803071.0,
      "step": 10865
    },
    {
      "entropy": 5.533701801300049,
      "epoch": 1.0788011115522032,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004892186406843374,
      "loss": 5.2057,
      "mean_token_accuracy": 0.16061035692691802,
      "num_tokens": 23813372.0,
      "step": 10870
    },
    {
      "entropy": 5.491838121414185,
      "epoch": 1.0792973402143708,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004892078016510841,
      "loss": 5.1706,
      "mean_token_accuracy": 0.16986196339130402,
      "num_tokens": 23824979.0,
      "step": 10875
    },
    {
      "entropy": 5.502412128448486,
      "epoch": 1.0797935688765383,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000489196957305939,
      "loss": 5.1622,
      "mean_token_accuracy": 0.1653856858611107,
      "num_tokens": 23836335.0,
      "step": 10880
    },
    {
      "entropy": 5.4904509544372555,
      "epoch": 1.0802897975387058,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004891861076491709,
      "loss": 5.1567,
      "mean_token_accuracy": 0.17029475420713425,
      "num_tokens": 23848147.0,
      "step": 10885
    },
    {
      "entropy": 5.451874876022339,
      "epoch": 1.0807860262008733,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004891752526810488,
      "loss": 5.1069,
      "mean_token_accuracy": 0.17364071756601335,
      "num_tokens": 23859502.0,
      "step": 10890
    },
    {
      "entropy": 5.522591018676758,
      "epoch": 1.081282254863041,
      "grad_norm": 0.84375,
      "learning_rate": 0.000489164392401842,
      "loss": 5.2518,
      "mean_token_accuracy": 0.16657178103923798,
      "num_tokens": 23870433.0,
      "step": 10895
    },
    {
      "entropy": 5.476272773742676,
      "epoch": 1.0817784835252084,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004891535268118196,
      "loss": 5.0635,
      "mean_token_accuracy": 0.1754612922668457,
      "num_tokens": 23879581.0,
      "step": 10900
    },
    {
      "entropy": 5.424292087554932,
      "epoch": 1.082274712187376,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004891426559112514,
      "loss": 5.1412,
      "mean_token_accuracy": 0.17278343439102173,
      "num_tokens": 23890889.0,
      "step": 10905
    },
    {
      "entropy": 5.4880101680755615,
      "epoch": 1.0827709408495434,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004891317797004066,
      "loss": 5.1348,
      "mean_token_accuracy": 0.17125614285469054,
      "num_tokens": 23901577.0,
      "step": 10910
    },
    {
      "entropy": 5.470410585403442,
      "epoch": 1.083267169511711,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004891208981795552,
      "loss": 5.2133,
      "mean_token_accuracy": 0.16936561316251755,
      "num_tokens": 23912698.0,
      "step": 10915
    },
    {
      "entropy": 5.417942237854004,
      "epoch": 1.0837633981738786,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004891100113489669,
      "loss": 5.0855,
      "mean_token_accuracy": 0.1729748159646988,
      "num_tokens": 23923056.0,
      "step": 10920
    },
    {
      "entropy": 5.475066804885865,
      "epoch": 1.084259626836046,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004890991192089116,
      "loss": 5.2276,
      "mean_token_accuracy": 0.16701621562242508,
      "num_tokens": 23933998.0,
      "step": 10925
    },
    {
      "entropy": 5.541318655014038,
      "epoch": 1.0847558554982135,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004890882217596595,
      "loss": 5.1268,
      "mean_token_accuracy": 0.1713723510503769,
      "num_tokens": 23943980.0,
      "step": 10930
    },
    {
      "entropy": 5.499484539031982,
      "epoch": 1.085252084160381,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004890773190014809,
      "loss": 5.193,
      "mean_token_accuracy": 0.16638527512550355,
      "num_tokens": 23955646.0,
      "step": 10935
    },
    {
      "entropy": 5.456666898727417,
      "epoch": 1.0857483128225487,
      "grad_norm": 0.9375,
      "learning_rate": 0.000489066410934646,
      "loss": 5.1657,
      "mean_token_accuracy": 0.1730557292699814,
      "num_tokens": 23966381.0,
      "step": 10940
    },
    {
      "entropy": 5.5801537990570065,
      "epoch": 1.0862445414847162,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004890554975594255,
      "loss": 5.2391,
      "mean_token_accuracy": 0.15644838511943818,
      "num_tokens": 23978669.0,
      "step": 10945
    },
    {
      "entropy": 5.626119947433471,
      "epoch": 1.0867407701468836,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004890445788760898,
      "loss": 5.2926,
      "mean_token_accuracy": 0.16053656488656998,
      "num_tokens": 23989467.0,
      "step": 10950
    },
    {
      "entropy": 5.492971706390381,
      "epoch": 1.0872369988090511,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004890336548849098,
      "loss": 5.1128,
      "mean_token_accuracy": 0.17263050973415375,
      "num_tokens": 24000797.0,
      "step": 10955
    },
    {
      "entropy": 5.582439327239991,
      "epoch": 1.0877332274712188,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004890227255861564,
      "loss": 5.2621,
      "mean_token_accuracy": 0.1681663379073143,
      "num_tokens": 24013182.0,
      "step": 10960
    },
    {
      "entropy": 5.5321564197540285,
      "epoch": 1.0882294561333863,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004890117909801005,
      "loss": 5.1396,
      "mean_token_accuracy": 0.17512231767177583,
      "num_tokens": 24024005.0,
      "step": 10965
    },
    {
      "entropy": 5.502088308334351,
      "epoch": 1.0887256847955538,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004890008510670134,
      "loss": 5.1377,
      "mean_token_accuracy": 0.1663498252630234,
      "num_tokens": 24034524.0,
      "step": 10970
    },
    {
      "entropy": 5.533322477340699,
      "epoch": 1.0892219134577212,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004889899058471663,
      "loss": 5.2013,
      "mean_token_accuracy": 0.16822793036699296,
      "num_tokens": 24045944.0,
      "step": 10975
    },
    {
      "entropy": 5.575636863708496,
      "epoch": 1.089718142119889,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004889789553208306,
      "loss": 5.268,
      "mean_token_accuracy": 0.1660327509045601,
      "num_tokens": 24058612.0,
      "step": 10980
    },
    {
      "entropy": 5.520413494110107,
      "epoch": 1.0902143707820564,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000488967999488278,
      "loss": 5.1379,
      "mean_token_accuracy": 0.17078853994607926,
      "num_tokens": 24068912.0,
      "step": 10985
    },
    {
      "entropy": 5.419206237792968,
      "epoch": 1.0907105994442239,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048895703834978,
      "loss": 5.1451,
      "mean_token_accuracy": 0.1665818512439728,
      "num_tokens": 24079879.0,
      "step": 10990
    },
    {
      "entropy": 5.503745603561401,
      "epoch": 1.0912068281063914,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004889460719056085,
      "loss": 5.1674,
      "mean_token_accuracy": 0.16840779334306716,
      "num_tokens": 24091268.0,
      "step": 10995
    },
    {
      "entropy": 5.518901920318603,
      "epoch": 1.091703056768559,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004889351001560354,
      "loss": 5.1914,
      "mean_token_accuracy": 0.16596200168132783,
      "num_tokens": 24101428.0,
      "step": 11000
    },
    {
      "entropy": 5.464192247390747,
      "epoch": 1.0921992854307265,
      "grad_norm": 0.875,
      "learning_rate": 0.0004889241231013329,
      "loss": 5.1585,
      "mean_token_accuracy": 0.1704880937933922,
      "num_tokens": 24112920.0,
      "step": 11005
    },
    {
      "entropy": 5.458786821365356,
      "epoch": 1.092695514092894,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000488913140741773,
      "loss": 5.0893,
      "mean_token_accuracy": 0.16703478693962098,
      "num_tokens": 24125731.0,
      "step": 11010
    },
    {
      "entropy": 5.479241228103637,
      "epoch": 1.0931917427550615,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004889021530776282,
      "loss": 5.1051,
      "mean_token_accuracy": 0.16768449544906616,
      "num_tokens": 24136606.0,
      "step": 11015
    },
    {
      "entropy": 5.550204944610596,
      "epoch": 1.093687971417229,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004888911601091709,
      "loss": 5.125,
      "mean_token_accuracy": 0.1734997034072876,
      "num_tokens": 24146783.0,
      "step": 11020
    },
    {
      "entropy": 5.499466609954834,
      "epoch": 1.0941842000793967,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004888801618366737,
      "loss": 5.226,
      "mean_token_accuracy": 0.1696036234498024,
      "num_tokens": 24157412.0,
      "step": 11025
    },
    {
      "entropy": 5.516845607757569,
      "epoch": 1.0946804287415641,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004888691582604095,
      "loss": 5.209,
      "mean_token_accuracy": 0.16557887345552444,
      "num_tokens": 24168077.0,
      "step": 11030
    },
    {
      "entropy": 5.553576278686523,
      "epoch": 1.0951766574037316,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000488858149380651,
      "loss": 5.2585,
      "mean_token_accuracy": 0.16936648190021514,
      "num_tokens": 24178913.0,
      "step": 11035
    },
    {
      "entropy": 5.5928089141845705,
      "epoch": 1.095672886065899,
      "grad_norm": 0.96875,
      "learning_rate": 0.000488847135197671,
      "loss": 5.3005,
      "mean_token_accuracy": 0.158344267308712,
      "num_tokens": 24190356.0,
      "step": 11040
    },
    {
      "entropy": 5.57515230178833,
      "epoch": 1.0961691147280668,
      "grad_norm": 1.046875,
      "learning_rate": 0.000488836115711743,
      "loss": 5.2473,
      "mean_token_accuracy": 0.161076557636261,
      "num_tokens": 24199772.0,
      "step": 11045
    },
    {
      "entropy": 5.540526580810547,
      "epoch": 1.0966653433902342,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004888250909231401,
      "loss": 5.1557,
      "mean_token_accuracy": 0.16764594614505768,
      "num_tokens": 24211807.0,
      "step": 11050
    },
    {
      "entropy": 5.476969623565674,
      "epoch": 1.0971615720524017,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004888140608321357,
      "loss": 5.1485,
      "mean_token_accuracy": 0.17423913180828093,
      "num_tokens": 24223156.0,
      "step": 11055
    },
    {
      "entropy": 5.567568063735962,
      "epoch": 1.0976578007145692,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004888030254390033,
      "loss": 5.2115,
      "mean_token_accuracy": 0.16370397061109543,
      "num_tokens": 24234352.0,
      "step": 11060
    },
    {
      "entropy": 5.516723585128784,
      "epoch": 1.098154029376737,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004887919847440166,
      "loss": 5.1298,
      "mean_token_accuracy": 0.1740872636437416,
      "num_tokens": 24244137.0,
      "step": 11065
    },
    {
      "entropy": 5.463329601287842,
      "epoch": 1.0986502580389044,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004887809387474494,
      "loss": 5.1369,
      "mean_token_accuracy": 0.16834257394075394,
      "num_tokens": 24253250.0,
      "step": 11070
    },
    {
      "entropy": 5.537700366973877,
      "epoch": 1.0991464867010718,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004887698874495755,
      "loss": 5.252,
      "mean_token_accuracy": 0.16806362569332123,
      "num_tokens": 24264551.0,
      "step": 11075
    },
    {
      "entropy": 5.508148956298828,
      "epoch": 1.0996427153632393,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000488758830850669,
      "loss": 5.0704,
      "mean_token_accuracy": 0.17682563811540603,
      "num_tokens": 24275522.0,
      "step": 11080
    },
    {
      "entropy": 5.475867557525635,
      "epoch": 1.100138944025407,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004887477689510042,
      "loss": 5.2243,
      "mean_token_accuracy": 0.16861492395401,
      "num_tokens": 24286333.0,
      "step": 11085
    },
    {
      "entropy": 5.475257730484008,
      "epoch": 1.1006351726875745,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004887367017508553,
      "loss": 5.2385,
      "mean_token_accuracy": 0.16946501284837723,
      "num_tokens": 24297627.0,
      "step": 11090
    },
    {
      "entropy": 5.495471620559693,
      "epoch": 1.101131401349742,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004887256292504967,
      "loss": 5.1766,
      "mean_token_accuracy": 0.16979029923677444,
      "num_tokens": 24308246.0,
      "step": 11095
    },
    {
      "entropy": 5.541315174102783,
      "epoch": 1.1016276300119094,
      "grad_norm": 1.078125,
      "learning_rate": 0.000488714551450203,
      "loss": 5.2241,
      "mean_token_accuracy": 0.162785604596138,
      "num_tokens": 24317718.0,
      "step": 11100
    },
    {
      "entropy": 5.501359224319458,
      "epoch": 1.102123858674077,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000488703468350249,
      "loss": 5.1548,
      "mean_token_accuracy": 0.16744553297758102,
      "num_tokens": 24329242.0,
      "step": 11105
    },
    {
      "entropy": 5.616970062255859,
      "epoch": 1.1026200873362446,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004886923799509095,
      "loss": 5.3133,
      "mean_token_accuracy": 0.16358482390642165,
      "num_tokens": 24340146.0,
      "step": 11110
    },
    {
      "entropy": 5.564360427856445,
      "epoch": 1.103116315998412,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004886812862524595,
      "loss": 5.2128,
      "mean_token_accuracy": 0.16251914203166962,
      "num_tokens": 24349370.0,
      "step": 11115
    },
    {
      "entropy": 5.573534393310547,
      "epoch": 1.1036125446605796,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004886701872551739,
      "loss": 5.1599,
      "mean_token_accuracy": 0.1653120204806328,
      "num_tokens": 24359770.0,
      "step": 11120
    },
    {
      "entropy": 5.5279176235198975,
      "epoch": 1.104108773322747,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004886590829593281,
      "loss": 5.2082,
      "mean_token_accuracy": 0.1692720517516136,
      "num_tokens": 24370974.0,
      "step": 11125
    },
    {
      "entropy": 5.529518032073975,
      "epoch": 1.1046050019849147,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004886479733651974,
      "loss": 5.205,
      "mean_token_accuracy": 0.167129485309124,
      "num_tokens": 24381771.0,
      "step": 11130
    },
    {
      "entropy": 5.481357479095459,
      "epoch": 1.1051012306470822,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004886368584730575,
      "loss": 5.172,
      "mean_token_accuracy": 0.16281089633703233,
      "num_tokens": 24393067.0,
      "step": 11135
    },
    {
      "entropy": 5.505965518951416,
      "epoch": 1.1055974593092497,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004886257382831837,
      "loss": 5.2394,
      "mean_token_accuracy": 0.1625291511416435,
      "num_tokens": 24402866.0,
      "step": 11140
    },
    {
      "entropy": 5.577435398101807,
      "epoch": 1.1060936879714172,
      "grad_norm": 0.953125,
      "learning_rate": 0.000488614612795852,
      "loss": 5.1683,
      "mean_token_accuracy": 0.169407357275486,
      "num_tokens": 24412956.0,
      "step": 11145
    },
    {
      "entropy": 5.573137950897217,
      "epoch": 1.1065899166335849,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004886034820113382,
      "loss": 5.2377,
      "mean_token_accuracy": 0.16365133225917816,
      "num_tokens": 24424650.0,
      "step": 11150
    },
    {
      "entropy": 5.555062007904053,
      "epoch": 1.1070861452957523,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004885923459299182,
      "loss": 5.2526,
      "mean_token_accuracy": 0.16688985675573348,
      "num_tokens": 24436541.0,
      "step": 11155
    },
    {
      "entropy": 5.571127605438233,
      "epoch": 1.1075823739579198,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004885812045518685,
      "loss": 5.2387,
      "mean_token_accuracy": 0.16248765736818313,
      "num_tokens": 24447588.0,
      "step": 11160
    },
    {
      "entropy": 5.570699787139892,
      "epoch": 1.1080786026200873,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000488570057877465,
      "loss": 5.1487,
      "mean_token_accuracy": 0.1717093899846077,
      "num_tokens": 24458854.0,
      "step": 11165
    },
    {
      "entropy": 5.5340282917022705,
      "epoch": 1.108574831282255,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004885589059069842,
      "loss": 5.2161,
      "mean_token_accuracy": 0.16735346764326095,
      "num_tokens": 24469427.0,
      "step": 11170
    },
    {
      "entropy": 5.455790281295776,
      "epoch": 1.1090710599444225,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004885477486407028,
      "loss": 5.2115,
      "mean_token_accuracy": 0.16059117913246154,
      "num_tokens": 24479352.0,
      "step": 11175
    },
    {
      "entropy": 5.601591444015503,
      "epoch": 1.10956728860659,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004885365860788974,
      "loss": 5.2,
      "mean_token_accuracy": 0.16488547176122664,
      "num_tokens": 24489580.0,
      "step": 11180
    },
    {
      "entropy": 5.519624948501587,
      "epoch": 1.1100635172687574,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004885254182218448,
      "loss": 5.1626,
      "mean_token_accuracy": 0.1684727668762207,
      "num_tokens": 24501141.0,
      "step": 11185
    },
    {
      "entropy": 5.479819393157959,
      "epoch": 1.1105597459309249,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004885142450698219,
      "loss": 5.1642,
      "mean_token_accuracy": 0.16967890560626983,
      "num_tokens": 24512079.0,
      "step": 11190
    },
    {
      "entropy": 5.552234888076782,
      "epoch": 1.1110559745930926,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004885030666231057,
      "loss": 5.2164,
      "mean_token_accuracy": 0.1667739197611809,
      "num_tokens": 24521839.0,
      "step": 11195
    },
    {
      "entropy": 5.551833391189575,
      "epoch": 1.11155220325526,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004884918828819738,
      "loss": 5.259,
      "mean_token_accuracy": 0.16346664428710939,
      "num_tokens": 24532704.0,
      "step": 11200
    },
    {
      "entropy": 5.524718379974365,
      "epoch": 1.1120484319174275,
      "grad_norm": 0.90625,
      "learning_rate": 0.000488480693846703,
      "loss": 5.1619,
      "mean_token_accuracy": 0.1717764914035797,
      "num_tokens": 24543790.0,
      "step": 11205
    },
    {
      "entropy": 5.499316644668579,
      "epoch": 1.112544660579595,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004884694995175711,
      "loss": 5.2149,
      "mean_token_accuracy": 0.16575611680746077,
      "num_tokens": 24554765.0,
      "step": 11210
    },
    {
      "entropy": 5.64555983543396,
      "epoch": 1.1130408892417627,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004884582998948555,
      "loss": 5.2889,
      "mean_token_accuracy": 0.15843822062015533,
      "num_tokens": 24565888.0,
      "step": 11215
    },
    {
      "entropy": 5.487444019317627,
      "epoch": 1.1135371179039302,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004884470949788342,
      "loss": 5.1487,
      "mean_token_accuracy": 0.1711251765489578,
      "num_tokens": 24575482.0,
      "step": 11220
    },
    {
      "entropy": 5.562432670593262,
      "epoch": 1.1140333465660976,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004884358847697847,
      "loss": 5.2724,
      "mean_token_accuracy": 0.15471834540367127,
      "num_tokens": 24588249.0,
      "step": 11225
    },
    {
      "entropy": 5.59770097732544,
      "epoch": 1.1145295752282651,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004884246692679852,
      "loss": 5.2105,
      "mean_token_accuracy": 0.1702192947268486,
      "num_tokens": 24599546.0,
      "step": 11230
    },
    {
      "entropy": 5.517202663421631,
      "epoch": 1.1150258038904326,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004884134484737139,
      "loss": 5.1551,
      "mean_token_accuracy": 0.16802441328763962,
      "num_tokens": 24611777.0,
      "step": 11235
    },
    {
      "entropy": 5.498545980453491,
      "epoch": 1.1155220325526003,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004884022223872489,
      "loss": 5.1362,
      "mean_token_accuracy": 0.16747527122497557,
      "num_tokens": 24623675.0,
      "step": 11240
    },
    {
      "entropy": 5.52301025390625,
      "epoch": 1.1160182612147678,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004883909910088686,
      "loss": 5.1652,
      "mean_token_accuracy": 0.16839252710342406,
      "num_tokens": 24633929.0,
      "step": 11245
    },
    {
      "entropy": 5.485877895355225,
      "epoch": 1.1165144898769352,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004883797543388515,
      "loss": 5.127,
      "mean_token_accuracy": 0.1698381707072258,
      "num_tokens": 24642800.0,
      "step": 11250
    },
    {
      "entropy": 5.456842374801636,
      "epoch": 1.1170107185391027,
      "grad_norm": 1.0,
      "learning_rate": 0.0004883685123774764,
      "loss": 5.1215,
      "mean_token_accuracy": 0.17325924783945085,
      "num_tokens": 24653780.0,
      "step": 11255
    },
    {
      "entropy": 5.472145652770996,
      "epoch": 1.1175069472012704,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004883572651250218,
      "loss": 5.1606,
      "mean_token_accuracy": 0.16740617752075196,
      "num_tokens": 24663912.0,
      "step": 11260
    },
    {
      "entropy": 5.536586141586303,
      "epoch": 1.118003175863438,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004883460125817669,
      "loss": 5.2486,
      "mean_token_accuracy": 0.1640252113342285,
      "num_tokens": 24675222.0,
      "step": 11265
    },
    {
      "entropy": 5.563428974151611,
      "epoch": 1.1184994045256054,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004883347547479905,
      "loss": 5.253,
      "mean_token_accuracy": 0.16681550741195678,
      "num_tokens": 24686697.0,
      "step": 11270
    },
    {
      "entropy": 5.571768045425415,
      "epoch": 1.1189956331877728,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000488323491623972,
      "loss": 5.1623,
      "mean_token_accuracy": 0.17624045610427858,
      "num_tokens": 24697067.0,
      "step": 11275
    },
    {
      "entropy": 5.548267364501953,
      "epoch": 1.1194918618499405,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004883122232099905,
      "loss": 5.1754,
      "mean_token_accuracy": 0.16944562941789626,
      "num_tokens": 24707989.0,
      "step": 11280
    },
    {
      "entropy": 5.431782197952271,
      "epoch": 1.119988090512108,
      "grad_norm": 0.875,
      "learning_rate": 0.0004883009495063256,
      "loss": 5.1139,
      "mean_token_accuracy": 0.17899062782526015,
      "num_tokens": 24718826.0,
      "step": 11285
    },
    {
      "entropy": 5.541382598876953,
      "epoch": 1.1204843191742755,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004882896705132566,
      "loss": 5.2302,
      "mean_token_accuracy": 0.1619640827178955,
      "num_tokens": 24729913.0,
      "step": 11290
    },
    {
      "entropy": 5.454885005950928,
      "epoch": 1.120980547836443,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048827838623106355,
      "loss": 5.0974,
      "mean_token_accuracy": 0.1714236170053482,
      "num_tokens": 24741621.0,
      "step": 11295
    },
    {
      "entropy": 5.5092963695526125,
      "epoch": 1.1214767764986107,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000488267096660026,
      "loss": 5.147,
      "mean_token_accuracy": 0.16761314421892165,
      "num_tokens": 24753427.0,
      "step": 11300
    },
    {
      "entropy": 5.549616384506225,
      "epoch": 1.1219730051607781,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048825580180042406,
      "loss": 5.1816,
      "mean_token_accuracy": 0.17004798501729965,
      "num_tokens": 24763580.0,
      "step": 11305
    },
    {
      "entropy": 5.481188058853149,
      "epoch": 1.1224692338229456,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004882445016525377,
      "loss": 5.1681,
      "mean_token_accuracy": 0.16722124814987183,
      "num_tokens": 24773955.0,
      "step": 11310
    },
    {
      "entropy": 5.511306381225586,
      "epoch": 1.122965462485113,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048823319621664723,
      "loss": 5.0865,
      "mean_token_accuracy": 0.1783554509282112,
      "num_tokens": 24784039.0,
      "step": 11315
    },
    {
      "entropy": 5.583564329147339,
      "epoch": 1.1234616911472806,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048822188549303296,
      "loss": 5.2022,
      "mean_token_accuracy": 0.16035133898258208,
      "num_tokens": 24794630.0,
      "step": 11320
    },
    {
      "entropy": 5.557933139801025,
      "epoch": 1.1239579198094483,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004882105694819754,
      "loss": 5.2514,
      "mean_token_accuracy": 0.1608917698264122,
      "num_tokens": 24805625.0,
      "step": 11325
    },
    {
      "entropy": 5.555304145812988,
      "epoch": 1.1244541484716157,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004881992481837551,
      "loss": 5.1837,
      "mean_token_accuracy": 0.175361330807209,
      "num_tokens": 24817356.0,
      "step": 11330
    },
    {
      "entropy": 5.503820419311523,
      "epoch": 1.1249503771337832,
      "grad_norm": 0.875,
      "learning_rate": 0.0004881879215986529,
      "loss": 5.1497,
      "mean_token_accuracy": 0.16104721277952194,
      "num_tokens": 24828349.0,
      "step": 11335
    },
    {
      "entropy": 5.518275547027588,
      "epoch": 1.1254466057959507,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048817658972694967,
      "loss": 5.1913,
      "mean_token_accuracy": 0.1651931494474411,
      "num_tokens": 24840138.0,
      "step": 11340
    },
    {
      "entropy": 5.532198524475097,
      "epoch": 1.1259428344581184,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004881652525689263,
      "loss": 5.1866,
      "mean_token_accuracy": 0.16392576992511748,
      "num_tokens": 24851570.0,
      "step": 11345
    },
    {
      "entropy": 5.406888103485107,
      "epoch": 1.1264390631202859,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000488153910124864,
      "loss": 5.0813,
      "mean_token_accuracy": 0.17003836184740068,
      "num_tokens": 24861396.0,
      "step": 11350
    },
    {
      "entropy": 5.527017307281494,
      "epoch": 1.1269352917824533,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000488142562395044,
      "loss": 5.2067,
      "mean_token_accuracy": 0.16171378791332244,
      "num_tokens": 24873519.0,
      "step": 11355
    },
    {
      "entropy": 5.5367402076721195,
      "epoch": 1.1274315204446208,
      "grad_norm": 0.7890625,
      "learning_rate": 0.00048813120937974774,
      "loss": 5.2196,
      "mean_token_accuracy": 0.16119163632392883,
      "num_tokens": 24886856.0,
      "step": 11360
    },
    {
      "entropy": 5.489326763153076,
      "epoch": 1.1279277491067885,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048811985107925684,
      "loss": 5.1744,
      "mean_token_accuracy": 0.16307065784931182,
      "num_tokens": 24897292.0,
      "step": 11365
    },
    {
      "entropy": 5.536104440689087,
      "epoch": 1.128423977768956,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048810848749385286,
      "loss": 5.1807,
      "mean_token_accuracy": 0.16587122082710265,
      "num_tokens": 24907757.0,
      "step": 11370
    },
    {
      "entropy": 5.453400373458862,
      "epoch": 1.1289202064311235,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048809711862381753,
      "loss": 5.1591,
      "mean_token_accuracy": 0.1707037642598152,
      "num_tokens": 24918071.0,
      "step": 11375
    },
    {
      "entropy": 5.48599762916565,
      "epoch": 1.129416435093291,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048808574446943274,
      "loss": 5.1332,
      "mean_token_accuracy": 0.17266378104686736,
      "num_tokens": 24927967.0,
      "step": 11380
    },
    {
      "entropy": 5.5505107879638675,
      "epoch": 1.1299126637554586,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004880743650309808,
      "loss": 5.207,
      "mean_token_accuracy": 0.17195654958486556,
      "num_tokens": 24939170.0,
      "step": 11385
    },
    {
      "entropy": 5.512000846862793,
      "epoch": 1.130408892417626,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00048806298030874373,
      "loss": 5.1922,
      "mean_token_accuracy": 0.1708056300878525,
      "num_tokens": 24950937.0,
      "step": 11390
    },
    {
      "entropy": 5.500658226013184,
      "epoch": 1.1309051210797936,
      "grad_norm": 0.81640625,
      "learning_rate": 0.00048805159030300383,
      "loss": 5.1834,
      "mean_token_accuracy": 0.16870294362306595,
      "num_tokens": 24962813.0,
      "step": 11395
    },
    {
      "entropy": 5.497263860702515,
      "epoch": 1.131401349741961,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048804019501404357,
      "loss": 5.0778,
      "mean_token_accuracy": 0.17271201312541962,
      "num_tokens": 24974192.0,
      "step": 11400
    },
    {
      "entropy": 5.461179494857788,
      "epoch": 1.1318975784041285,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048802879444214553,
      "loss": 5.0843,
      "mean_token_accuracy": 0.17591992020606995,
      "num_tokens": 24985353.0,
      "step": 11405
    },
    {
      "entropy": 5.498987674713135,
      "epoch": 1.1323938070662962,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00048801738858759243,
      "loss": 5.1904,
      "mean_token_accuracy": 0.17054355442523955,
      "num_tokens": 24997259.0,
      "step": 11410
    },
    {
      "entropy": 5.507398080825806,
      "epoch": 1.1328900357284637,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004880059774506671,
      "loss": 5.1777,
      "mean_token_accuracy": 0.16675892919301988,
      "num_tokens": 25008222.0,
      "step": 11415
    },
    {
      "entropy": 5.512610197067261,
      "epoch": 1.1333862643906312,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004879945610316525,
      "loss": 5.2189,
      "mean_token_accuracy": 0.16371117830276488,
      "num_tokens": 25020267.0,
      "step": 11420
    },
    {
      "entropy": 5.516874551773071,
      "epoch": 1.1338824930527986,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048798313933083175,
      "loss": 5.1886,
      "mean_token_accuracy": 0.1679120123386383,
      "num_tokens": 25031212.0,
      "step": 11425
    },
    {
      "entropy": 5.514549446105957,
      "epoch": 1.1343787217149663,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004879717123484882,
      "loss": 5.1548,
      "mean_token_accuracy": 0.16509698778390886,
      "num_tokens": 25042204.0,
      "step": 11430
    },
    {
      "entropy": 5.542244005203247,
      "epoch": 1.1348749503771338,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000487960280084905,
      "loss": 5.1812,
      "mean_token_accuracy": 0.1701316550374031,
      "num_tokens": 25052010.0,
      "step": 11435
    },
    {
      "entropy": 5.551565837860108,
      "epoch": 1.1353711790393013,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004879488425403657,
      "loss": 5.2296,
      "mean_token_accuracy": 0.161693225055933,
      "num_tokens": 25062209.0,
      "step": 11440
    },
    {
      "entropy": 5.504443120956421,
      "epoch": 1.1358674077014688,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048793739971515407,
      "loss": 5.1864,
      "mean_token_accuracy": 0.16541998088359833,
      "num_tokens": 25072758.0,
      "step": 11445
    },
    {
      "entropy": 5.549315118789673,
      "epoch": 1.1363636363636362,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048792595160955374,
      "loss": 5.2664,
      "mean_token_accuracy": 0.16343289166688918,
      "num_tokens": 25084318.0,
      "step": 11450
    },
    {
      "entropy": 5.550451278686523,
      "epoch": 1.136859865025804,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048791449822384856,
      "loss": 5.154,
      "mean_token_accuracy": 0.17242386043071747,
      "num_tokens": 25095715.0,
      "step": 11455
    },
    {
      "entropy": 5.570097970962524,
      "epoch": 1.1373560936879714,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004879030395583226,
      "loss": 5.1702,
      "mean_token_accuracy": 0.16954847127199174,
      "num_tokens": 25106163.0,
      "step": 11460
    },
    {
      "entropy": 5.579524564743042,
      "epoch": 1.137852322350139,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004878915756132601,
      "loss": 5.2476,
      "mean_token_accuracy": 0.16438795924186705,
      "num_tokens": 25118745.0,
      "step": 11465
    },
    {
      "entropy": 5.528561401367187,
      "epoch": 1.1383485510123066,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048788010638894523,
      "loss": 5.2274,
      "mean_token_accuracy": 0.15801485627889633,
      "num_tokens": 25130066.0,
      "step": 11470
    },
    {
      "entropy": 5.497963285446167,
      "epoch": 1.138844779674474,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004878686318856624,
      "loss": 5.0723,
      "mean_token_accuracy": 0.1734581857919693,
      "num_tokens": 25139895.0,
      "step": 11475
    },
    {
      "entropy": 5.63120584487915,
      "epoch": 1.1393410083366415,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048785715210369635,
      "loss": 5.3008,
      "mean_token_accuracy": 0.15998752117156984,
      "num_tokens": 25150644.0,
      "step": 11480
    },
    {
      "entropy": 5.549851989746093,
      "epoch": 1.139837236998809,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004878456670433314,
      "loss": 5.2144,
      "mean_token_accuracy": 0.16592725813388826,
      "num_tokens": 25163038.0,
      "step": 11485
    },
    {
      "entropy": 5.490525531768799,
      "epoch": 1.1403334656609765,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004878341767048527,
      "loss": 5.1225,
      "mean_token_accuracy": 0.17211485654115677,
      "num_tokens": 25174154.0,
      "step": 11490
    },
    {
      "entropy": 5.5640708923339846,
      "epoch": 1.1408296943231442,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048782268108854494,
      "loss": 5.2429,
      "mean_token_accuracy": 0.16499737352132798,
      "num_tokens": 25185369.0,
      "step": 11495
    },
    {
      "entropy": 5.511232233047485,
      "epoch": 1.1413259229853117,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004878111801946933,
      "loss": 5.2328,
      "mean_token_accuracy": 0.1616926297545433,
      "num_tokens": 25197875.0,
      "step": 11500
    },
    {
      "entropy": 5.4827024936676025,
      "epoch": 1.1418221516474791,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048779967402358296,
      "loss": 5.1164,
      "mean_token_accuracy": 0.1780473992228508,
      "num_tokens": 25208362.0,
      "step": 11505
    },
    {
      "entropy": 5.589849996566772,
      "epoch": 1.1423183803096466,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004877881625754993,
      "loss": 5.3077,
      "mean_token_accuracy": 0.16069869101047515,
      "num_tokens": 25220441.0,
      "step": 11510
    },
    {
      "entropy": 5.489728593826294,
      "epoch": 1.1428146089718143,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048777664585072765,
      "loss": 5.0851,
      "mean_token_accuracy": 0.16956905275583267,
      "num_tokens": 25231134.0,
      "step": 11515
    },
    {
      "entropy": 5.618770551681519,
      "epoch": 1.1433108376339818,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004877651238495537,
      "loss": 5.3481,
      "mean_token_accuracy": 0.15938334316015243,
      "num_tokens": 25242327.0,
      "step": 11520
    },
    {
      "entropy": 5.551085948944092,
      "epoch": 1.1438070662961493,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048775359657226316,
      "loss": 5.1846,
      "mean_token_accuracy": 0.16684814542531967,
      "num_tokens": 25253571.0,
      "step": 11525
    },
    {
      "entropy": 5.470210313796997,
      "epoch": 1.1443032949583167,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004877420640191418,
      "loss": 5.1238,
      "mean_token_accuracy": 0.17145076692104338,
      "num_tokens": 25264573.0,
      "step": 11530
    },
    {
      "entropy": 5.529512643814087,
      "epoch": 1.1447995236204842,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004877305261904758,
      "loss": 5.2137,
      "mean_token_accuracy": 0.1657499223947525,
      "num_tokens": 25275197.0,
      "step": 11535
    },
    {
      "entropy": 5.516482925415039,
      "epoch": 1.145295752282652,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048771898308655107,
      "loss": 5.097,
      "mean_token_accuracy": 0.17488718926906585,
      "num_tokens": 25286564.0,
      "step": 11540
    },
    {
      "entropy": 5.564264726638794,
      "epoch": 1.1457919809448194,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004877074347076539,
      "loss": 5.1505,
      "mean_token_accuracy": 0.17080543786287308,
      "num_tokens": 25297874.0,
      "step": 11545
    },
    {
      "entropy": 5.543221855163575,
      "epoch": 1.1462882096069869,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048769588105407075,
      "loss": 5.2173,
      "mean_token_accuracy": 0.16473061442375184,
      "num_tokens": 25308047.0,
      "step": 11550
    },
    {
      "entropy": 5.579469966888428,
      "epoch": 1.1467844382691545,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048768432212608807,
      "loss": 5.2563,
      "mean_token_accuracy": 0.1589917406439781,
      "num_tokens": 25318527.0,
      "step": 11555
    },
    {
      "entropy": 5.531277179718018,
      "epoch": 1.147280666931322,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004876727579239925,
      "loss": 5.1707,
      "mean_token_accuracy": 0.17014151066541672,
      "num_tokens": 25328242.0,
      "step": 11560
    },
    {
      "entropy": 5.5312587261199955,
      "epoch": 1.1477768955934895,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004876611884480708,
      "loss": 5.1572,
      "mean_token_accuracy": 0.16783188432455062,
      "num_tokens": 25338214.0,
      "step": 11565
    },
    {
      "entropy": 5.524855613708496,
      "epoch": 1.148273124255657,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048764961369860977,
      "loss": 5.2066,
      "mean_token_accuracy": 0.16303918212652208,
      "num_tokens": 25349762.0,
      "step": 11570
    },
    {
      "entropy": 5.506272411346435,
      "epoch": 1.1487693529178244,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048763803367589663,
      "loss": 5.1949,
      "mean_token_accuracy": 0.16914894729852675,
      "num_tokens": 25361241.0,
      "step": 11575
    },
    {
      "entropy": 5.553636074066162,
      "epoch": 1.1492655815799921,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048762644838021843,
      "loss": 5.2317,
      "mean_token_accuracy": 0.16020854711532592,
      "num_tokens": 25373846.0,
      "step": 11580
    },
    {
      "entropy": 5.527219963073731,
      "epoch": 1.1497618102421596,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048761485781186243,
      "loss": 5.1486,
      "mean_token_accuracy": 0.16828452944755554,
      "num_tokens": 25385331.0,
      "step": 11585
    },
    {
      "entropy": 5.5011152744293215,
      "epoch": 1.150258038904327,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004876032619711162,
      "loss": 5.1743,
      "mean_token_accuracy": 0.17516097724437713,
      "num_tokens": 25395140.0,
      "step": 11590
    },
    {
      "entropy": 5.5760979652404785,
      "epoch": 1.1507542675664946,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048759166085826714,
      "loss": 5.2234,
      "mean_token_accuracy": 0.15750466883182526,
      "num_tokens": 25405455.0,
      "step": 11595
    },
    {
      "entropy": 5.5412132263183596,
      "epoch": 1.1512504962286623,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000487580054473603,
      "loss": 5.2322,
      "mean_token_accuracy": 0.16742897033691406,
      "num_tokens": 25417000.0,
      "step": 11600
    },
    {
      "entropy": 5.440884304046631,
      "epoch": 1.1517467248908297,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048756844281741156,
      "loss": 5.1338,
      "mean_token_accuracy": 0.17535168528556824,
      "num_tokens": 25427808.0,
      "step": 11605
    },
    {
      "entropy": 5.516225719451905,
      "epoch": 1.1522429535529972,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048755682588998085,
      "loss": 5.1628,
      "mean_token_accuracy": 0.1629212662577629,
      "num_tokens": 25438302.0,
      "step": 11610
    },
    {
      "entropy": 5.4552528858184814,
      "epoch": 1.1527391822151647,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004875452036915987,
      "loss": 5.0679,
      "mean_token_accuracy": 0.16725915521383286,
      "num_tokens": 25449335.0,
      "step": 11615
    },
    {
      "entropy": 5.596983766555786,
      "epoch": 1.1532354108773322,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004875335762225536,
      "loss": 5.2561,
      "mean_token_accuracy": 0.16031309962272644,
      "num_tokens": 25461173.0,
      "step": 11620
    },
    {
      "entropy": 5.561198377609253,
      "epoch": 1.1537316395394999,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048752194348313383,
      "loss": 5.1535,
      "mean_token_accuracy": 0.1659163236618042,
      "num_tokens": 25472538.0,
      "step": 11625
    },
    {
      "entropy": 5.49793119430542,
      "epoch": 1.1542278682016673,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004875103054736278,
      "loss": 5.1902,
      "mean_token_accuracy": 0.170088492333889,
      "num_tokens": 25483386.0,
      "step": 11630
    },
    {
      "entropy": 5.547458934783935,
      "epoch": 1.1547240968638348,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004874986621943241,
      "loss": 5.2026,
      "mean_token_accuracy": 0.17032596319913865,
      "num_tokens": 25495505.0,
      "step": 11635
    },
    {
      "entropy": 5.521256303787231,
      "epoch": 1.1552203255260023,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048748701364551147,
      "loss": 5.1882,
      "mean_token_accuracy": 0.16937849819660186,
      "num_tokens": 25506144.0,
      "step": 11640
    },
    {
      "entropy": 5.524990606307983,
      "epoch": 1.15571655418817,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004874753598274787,
      "loss": 5.1835,
      "mean_token_accuracy": 0.1634795621037483,
      "num_tokens": 25516185.0,
      "step": 11645
    },
    {
      "entropy": 5.6168482303619385,
      "epoch": 1.1562127828503375,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004874637007405149,
      "loss": 5.266,
      "mean_token_accuracy": 0.1643349274992943,
      "num_tokens": 25526657.0,
      "step": 11650
    },
    {
      "entropy": 5.4739500999450685,
      "epoch": 1.156709011512505,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048745203638490917,
      "loss": 5.1242,
      "mean_token_accuracy": 0.17248900085687638,
      "num_tokens": 25537101.0,
      "step": 11655
    },
    {
      "entropy": 5.509799146652222,
      "epoch": 1.1572052401746724,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048744036676095067,
      "loss": 5.2268,
      "mean_token_accuracy": 0.15992130041122438,
      "num_tokens": 25547672.0,
      "step": 11660
    },
    {
      "entropy": 5.59275450706482,
      "epoch": 1.15770146883684,
      "grad_norm": 1.0,
      "learning_rate": 0.0004874286918689289,
      "loss": 5.2681,
      "mean_token_accuracy": 0.1656245231628418,
      "num_tokens": 25558408.0,
      "step": 11665
    },
    {
      "entropy": 5.589129781723022,
      "epoch": 1.1581976974990076,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048741701170913326,
      "loss": 5.1705,
      "mean_token_accuracy": 0.1653664618730545,
      "num_tokens": 25569823.0,
      "step": 11670
    },
    {
      "entropy": 5.51324872970581,
      "epoch": 1.158693926161175,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048740532628185353,
      "loss": 5.168,
      "mean_token_accuracy": 0.17340276390314102,
      "num_tokens": 25579705.0,
      "step": 11675
    },
    {
      "entropy": 5.521526765823364,
      "epoch": 1.1591901548233425,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004873936355873793,
      "loss": 5.2393,
      "mean_token_accuracy": 0.1646953508257866,
      "num_tokens": 25590573.0,
      "step": 11680
    },
    {
      "entropy": 5.509283542633057,
      "epoch": 1.1596863834855102,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004873819396260007,
      "loss": 5.1496,
      "mean_token_accuracy": 0.17044238299131392,
      "num_tokens": 25601418.0,
      "step": 11685
    },
    {
      "entropy": 5.5421528816223145,
      "epoch": 1.1601826121476777,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048737023839800754,
      "loss": 5.2243,
      "mean_token_accuracy": 0.164617358148098,
      "num_tokens": 25612409.0,
      "step": 11690
    },
    {
      "entropy": 5.496366214752197,
      "epoch": 1.1606788408098452,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004873585319036901,
      "loss": 5.1403,
      "mean_token_accuracy": 0.17199216336011885,
      "num_tokens": 25623547.0,
      "step": 11695
    },
    {
      "entropy": 5.570702028274536,
      "epoch": 1.1611750694720127,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004873468201433387,
      "loss": 5.319,
      "mean_token_accuracy": 0.160432893037796,
      "num_tokens": 25636537.0,
      "step": 11700
    },
    {
      "entropy": 5.542683935165405,
      "epoch": 1.1616712981341801,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004873351031172437,
      "loss": 5.1433,
      "mean_token_accuracy": 0.1731248676776886,
      "num_tokens": 25646734.0,
      "step": 11705
    },
    {
      "entropy": 5.560605478286743,
      "epoch": 1.1621675267963478,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048732338082569564,
      "loss": 5.2509,
      "mean_token_accuracy": 0.1691226616501808,
      "num_tokens": 25658058.0,
      "step": 11710
    },
    {
      "entropy": 5.492860174179077,
      "epoch": 1.1626637554585153,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004873116532689853,
      "loss": 5.088,
      "mean_token_accuracy": 0.17136604934930802,
      "num_tokens": 25669630.0,
      "step": 11715
    },
    {
      "entropy": 5.450148391723633,
      "epoch": 1.1631599841206828,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004872999204474035,
      "loss": 5.1037,
      "mean_token_accuracy": 0.1707699716091156,
      "num_tokens": 25681928.0,
      "step": 11720
    },
    {
      "entropy": 5.440466070175171,
      "epoch": 1.1636562127828503,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000487288182361241,
      "loss": 5.1108,
      "mean_token_accuracy": 0.1661658838391304,
      "num_tokens": 25690653.0,
      "step": 11725
    },
    {
      "entropy": 5.538868236541748,
      "epoch": 1.164152441445018,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004872764390107891,
      "loss": 5.2163,
      "mean_token_accuracy": 0.16166536659002304,
      "num_tokens": 25701201.0,
      "step": 11730
    },
    {
      "entropy": 5.457179927825928,
      "epoch": 1.1646486701071854,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048726469039633895,
      "loss": 5.133,
      "mean_token_accuracy": 0.17010793089866638,
      "num_tokens": 25710927.0,
      "step": 11735
    },
    {
      "entropy": 5.515346240997315,
      "epoch": 1.165144898769353,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048725293651818183,
      "loss": 5.1783,
      "mean_token_accuracy": 0.16830242574214935,
      "num_tokens": 25722136.0,
      "step": 11740
    },
    {
      "entropy": 5.513708066940308,
      "epoch": 1.1656411274315204,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004872411773766093,
      "loss": 5.0631,
      "mean_token_accuracy": 0.17500446289777755,
      "num_tokens": 25733386.0,
      "step": 11745
    },
    {
      "entropy": 5.467298269271851,
      "epoch": 1.1661373560936878,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048722941297191284,
      "loss": 5.1494,
      "mean_token_accuracy": 0.16997703164815903,
      "num_tokens": 25742801.0,
      "step": 11750
    },
    {
      "entropy": 5.5761877536773685,
      "epoch": 1.1666335847558555,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004872176433043843,
      "loss": 5.2554,
      "mean_token_accuracy": 0.16377057135105133,
      "num_tokens": 25755594.0,
      "step": 11755
    },
    {
      "entropy": 5.559096097946167,
      "epoch": 1.167129813418023,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004872058683743155,
      "loss": 5.187,
      "mean_token_accuracy": 0.1699764460325241,
      "num_tokens": 25766271.0,
      "step": 11760
    },
    {
      "entropy": 5.56765661239624,
      "epoch": 1.1676260420801905,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048719408818199836,
      "loss": 5.2355,
      "mean_token_accuracy": 0.1647241234779358,
      "num_tokens": 25777338.0,
      "step": 11765
    },
    {
      "entropy": 5.510936880111695,
      "epoch": 1.1681222707423582,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048718230272772517,
      "loss": 5.1173,
      "mean_token_accuracy": 0.16859878301620485,
      "num_tokens": 25787968.0,
      "step": 11770
    },
    {
      "entropy": 5.519161939620972,
      "epoch": 1.1686184994045257,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048717051201178804,
      "loss": 5.1988,
      "mean_token_accuracy": 0.17095001190900802,
      "num_tokens": 25798274.0,
      "step": 11775
    },
    {
      "entropy": 5.480053472518921,
      "epoch": 1.1691147280666931,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004871587160344794,
      "loss": 5.158,
      "mean_token_accuracy": 0.16201619505882264,
      "num_tokens": 25809140.0,
      "step": 11780
    },
    {
      "entropy": 5.546340751647949,
      "epoch": 1.1696109567288606,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004871469147960917,
      "loss": 5.1983,
      "mean_token_accuracy": 0.1667727842926979,
      "num_tokens": 25821502.0,
      "step": 11785
    },
    {
      "entropy": 5.529561853408813,
      "epoch": 1.170107185391028,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004871351082969177,
      "loss": 5.0859,
      "mean_token_accuracy": 0.17633658647537231,
      "num_tokens": 25833231.0,
      "step": 11790
    },
    {
      "entropy": 5.568858480453491,
      "epoch": 1.1706034140531958,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004871232965372501,
      "loss": 5.2379,
      "mean_token_accuracy": 0.16485683172941207,
      "num_tokens": 25844867.0,
      "step": 11795
    },
    {
      "entropy": 5.52788896560669,
      "epoch": 1.1710996427153633,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004871114795173819,
      "loss": 5.2151,
      "mean_token_accuracy": 0.17525176405906678,
      "num_tokens": 25855959.0,
      "step": 11800
    },
    {
      "entropy": 5.576307678222657,
      "epoch": 1.1715958713775307,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004870996572376061,
      "loss": 5.2943,
      "mean_token_accuracy": 0.16473801285028458,
      "num_tokens": 25866926.0,
      "step": 11805
    },
    {
      "entropy": 5.598384284973145,
      "epoch": 1.1720921000396982,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004870878296982157,
      "loss": 5.228,
      "mean_token_accuracy": 0.16459791511297225,
      "num_tokens": 25877699.0,
      "step": 11810
    },
    {
      "entropy": 5.592456245422364,
      "epoch": 1.172588328701866,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004870759968995042,
      "loss": 5.174,
      "mean_token_accuracy": 0.17203024327754973,
      "num_tokens": 25888640.0,
      "step": 11815
    },
    {
      "entropy": 5.517426633834839,
      "epoch": 1.1730845573640334,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048706415884176493,
      "loss": 5.173,
      "mean_token_accuracy": 0.1669470727443695,
      "num_tokens": 25900651.0,
      "step": 11820
    },
    {
      "entropy": 5.508384704589844,
      "epoch": 1.1735807860262009,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004870523155252915,
      "loss": 5.1593,
      "mean_token_accuracy": 0.16976964324712754,
      "num_tokens": 25911775.0,
      "step": 11825
    },
    {
      "entropy": 5.508563232421875,
      "epoch": 1.1740770146883683,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004870404669503776,
      "loss": 5.1163,
      "mean_token_accuracy": 0.16939260065555573,
      "num_tokens": 25922126.0,
      "step": 11830
    },
    {
      "entropy": 5.565032005310059,
      "epoch": 1.1745732433505358,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000487028613117317,
      "loss": 5.1804,
      "mean_token_accuracy": 0.1685253396630287,
      "num_tokens": 25932945.0,
      "step": 11835
    },
    {
      "entropy": 5.596076774597168,
      "epoch": 1.1750694720127035,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004870167540264037,
      "loss": 5.2499,
      "mean_token_accuracy": 0.15927001237869262,
      "num_tokens": 25944277.0,
      "step": 11840
    },
    {
      "entropy": 5.453707075119018,
      "epoch": 1.175565700674871,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004870048896779317,
      "loss": 5.1124,
      "mean_token_accuracy": 0.17226413041353225,
      "num_tokens": 25956838.0,
      "step": 11845
    },
    {
      "entropy": 5.478300619125366,
      "epoch": 1.1760619293370385,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048699302007219533,
      "loss": 5.1597,
      "mean_token_accuracy": 0.16248856782913207,
      "num_tokens": 25966639.0,
      "step": 11850
    },
    {
      "entropy": 5.521308898925781,
      "epoch": 1.1765581579992062,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004869811452094889,
      "loss": 5.2009,
      "mean_token_accuracy": 0.173517669737339,
      "num_tokens": 25978209.0,
      "step": 11855
    },
    {
      "entropy": 5.498318338394165,
      "epoch": 1.1770543866613736,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048696926509010673,
      "loss": 5.1334,
      "mean_token_accuracy": 0.1738154858350754,
      "num_tokens": 25989306.0,
      "step": 11860
    },
    {
      "entropy": 5.537467432022095,
      "epoch": 1.177550615323541,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004869573797143437,
      "loss": 5.2327,
      "mean_token_accuracy": 0.16106203198432922,
      "num_tokens": 26000235.0,
      "step": 11865
    },
    {
      "entropy": 5.511281061172485,
      "epoch": 1.1780468439857086,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004869454890824942,
      "loss": 5.1268,
      "mean_token_accuracy": 0.17023174166679383,
      "num_tokens": 26010616.0,
      "step": 11870
    },
    {
      "entropy": 5.511196327209473,
      "epoch": 1.178543072647876,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00048693359319485337,
      "loss": 5.2079,
      "mean_token_accuracy": 0.16644639074802398,
      "num_tokens": 26021303.0,
      "step": 11875
    },
    {
      "entropy": 5.527012968063355,
      "epoch": 1.1790393013100438,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004869216920517161,
      "loss": 5.1298,
      "mean_token_accuracy": 0.16764458566904067,
      "num_tokens": 26032386.0,
      "step": 11880
    },
    {
      "entropy": 5.454592275619507,
      "epoch": 1.1795355299722112,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048690978565337753,
      "loss": 5.0962,
      "mean_token_accuracy": 0.1689310885965824,
      "num_tokens": 26042889.0,
      "step": 11885
    },
    {
      "entropy": 5.479773139953613,
      "epoch": 1.1800317586343787,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048689787400013287,
      "loss": 5.2047,
      "mean_token_accuracy": 0.16476578414440154,
      "num_tokens": 26054299.0,
      "step": 11890
    },
    {
      "entropy": 5.61276683807373,
      "epoch": 1.1805279872965462,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004868859570922775,
      "loss": 5.2453,
      "mean_token_accuracy": 0.16201778948307038,
      "num_tokens": 26064587.0,
      "step": 11895
    },
    {
      "entropy": 5.580943870544433,
      "epoch": 1.1810242159587139,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004868740349301071,
      "loss": 5.1712,
      "mean_token_accuracy": 0.17701984643936158,
      "num_tokens": 26076495.0,
      "step": 11900
    },
    {
      "entropy": 5.543245077133179,
      "epoch": 1.1815204446208813,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000486862107513917,
      "loss": 5.1831,
      "mean_token_accuracy": 0.16747597455978394,
      "num_tokens": 26086496.0,
      "step": 11905
    },
    {
      "entropy": 5.488832187652588,
      "epoch": 1.1820166732830488,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004868501748440033,
      "loss": 5.157,
      "mean_token_accuracy": 0.16652124971151352,
      "num_tokens": 26097263.0,
      "step": 11910
    },
    {
      "entropy": 5.536653661727906,
      "epoch": 1.1825129019452163,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004868382369206617,
      "loss": 5.161,
      "mean_token_accuracy": 0.168246990442276,
      "num_tokens": 26108264.0,
      "step": 11915
    },
    {
      "entropy": 5.532262706756592,
      "epoch": 1.1830091306073838,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004868262937441883,
      "loss": 5.2233,
      "mean_token_accuracy": 0.17646037936210632,
      "num_tokens": 26119730.0,
      "step": 11920
    },
    {
      "entropy": 5.55259370803833,
      "epoch": 1.1835053592695515,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048681434531487925,
      "loss": 5.1289,
      "mean_token_accuracy": 0.16772469878196716,
      "num_tokens": 26131055.0,
      "step": 11925
    },
    {
      "entropy": 5.4555566787719725,
      "epoch": 1.184001587931719,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048680239163303085,
      "loss": 5.121,
      "mean_token_accuracy": 0.17310939133167266,
      "num_tokens": 26141577.0,
      "step": 11930
    },
    {
      "entropy": 5.5206818103790285,
      "epoch": 1.1844978165938864,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004867904326989396,
      "loss": 5.1716,
      "mean_token_accuracy": 0.16741517931222916,
      "num_tokens": 26152733.0,
      "step": 11935
    },
    {
      "entropy": 5.569200706481934,
      "epoch": 1.184994045256054,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004867784685129019,
      "loss": 5.2112,
      "mean_token_accuracy": 0.16788169592618943,
      "num_tokens": 26164763.0,
      "step": 11940
    },
    {
      "entropy": 5.486721515655518,
      "epoch": 1.1854902739182216,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048676649907521455,
      "loss": 5.0908,
      "mean_token_accuracy": 0.17138195186853408,
      "num_tokens": 26174391.0,
      "step": 11945
    },
    {
      "entropy": 5.417409515380859,
      "epoch": 1.185986502580389,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048675452438617434,
      "loss": 5.0737,
      "mean_token_accuracy": 0.17694325000047684,
      "num_tokens": 26184772.0,
      "step": 11950
    },
    {
      "entropy": 5.583131980895996,
      "epoch": 1.1864827312425565,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004867425444460782,
      "loss": 5.2208,
      "mean_token_accuracy": 0.15865879356861115,
      "num_tokens": 26195411.0,
      "step": 11955
    },
    {
      "entropy": 5.488093328475952,
      "epoch": 1.186978959904724,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004867305592552233,
      "loss": 5.1463,
      "mean_token_accuracy": 0.16958065927028657,
      "num_tokens": 26206719.0,
      "step": 11960
    },
    {
      "entropy": 5.47322154045105,
      "epoch": 1.1874751885668917,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004867185688139067,
      "loss": 5.0898,
      "mean_token_accuracy": 0.17300083339214326,
      "num_tokens": 26217469.0,
      "step": 11965
    },
    {
      "entropy": 5.454719018936157,
      "epoch": 1.1879714172290592,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004867065731224258,
      "loss": 5.1056,
      "mean_token_accuracy": 0.16581848561763762,
      "num_tokens": 26228556.0,
      "step": 11970
    },
    {
      "entropy": 5.533361625671387,
      "epoch": 1.1884676458912267,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004866945721810781,
      "loss": 5.2115,
      "mean_token_accuracy": 0.16659659743309022,
      "num_tokens": 26239891.0,
      "step": 11975
    },
    {
      "entropy": 5.521927070617676,
      "epoch": 1.1889638745533941,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004866825659901612,
      "loss": 5.1733,
      "mean_token_accuracy": 0.1729780450463295,
      "num_tokens": 26251793.0,
      "step": 11980
    },
    {
      "entropy": 5.426382732391358,
      "epoch": 1.1894601032155618,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048667055454997276,
      "loss": 5.0653,
      "mean_token_accuracy": 0.17323132008314132,
      "num_tokens": 26262367.0,
      "step": 11985
    },
    {
      "entropy": 5.483523797988892,
      "epoch": 1.1899563318777293,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048665853786081063,
      "loss": 5.1624,
      "mean_token_accuracy": 0.167740099132061,
      "num_tokens": 26272681.0,
      "step": 11990
    },
    {
      "entropy": 5.540319919586182,
      "epoch": 1.1904525605398968,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048664651592297294,
      "loss": 5.2232,
      "mean_token_accuracy": 0.16316691040992737,
      "num_tokens": 26281194.0,
      "step": 11995
    },
    {
      "entropy": 5.543879842758178,
      "epoch": 1.1909487892020643,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048663448873675766,
      "loss": 5.1236,
      "mean_token_accuracy": 0.1654999375343323,
      "num_tokens": 26292682.0,
      "step": 12000
    },
    {
      "epoch": 1.1909487892020643,
      "eval_entropy": 5.388492889997936,
      "eval_loss": 5.299468517303467,
      "eval_mean_token_accuracy": 0.1675439624137509,
      "eval_num_tokens": 26292682.0,
      "eval_runtime": 20.8063,
      "eval_samples_per_second": 1618.211,
      "eval_steps_per_second": 202.294,
      "step": 12000
    },
    {
      "entropy": 5.559010601043701,
      "epoch": 1.1914450178642317,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004866224563024631,
      "loss": 5.2223,
      "mean_token_accuracy": 0.16385785341262818,
      "num_tokens": 26305500.0,
      "step": 12005
    },
    {
      "entropy": 5.515285015106201,
      "epoch": 1.1919412465263994,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004866104186203877,
      "loss": 5.1131,
      "mean_token_accuracy": 0.1688894212245941,
      "num_tokens": 26317207.0,
      "step": 12010
    },
    {
      "entropy": 5.517972230911255,
      "epoch": 1.192437475188567,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048659837569082977,
      "loss": 5.1217,
      "mean_token_accuracy": 0.1721375048160553,
      "num_tokens": 26328728.0,
      "step": 12015
    },
    {
      "entropy": 5.580038928985596,
      "epoch": 1.1929337038507344,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004865863275140882,
      "loss": 5.1815,
      "mean_token_accuracy": 0.1676922097802162,
      "num_tokens": 26338622.0,
      "step": 12020
    },
    {
      "entropy": 5.538651657104492,
      "epoch": 1.1934299325129019,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004865742740904615,
      "loss": 5.1573,
      "mean_token_accuracy": 0.1691615954041481,
      "num_tokens": 26348807.0,
      "step": 12025
    },
    {
      "entropy": 5.446991109848023,
      "epoch": 1.1939261611750696,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004865622154202488,
      "loss": 5.1225,
      "mean_token_accuracy": 0.17199401259422303,
      "num_tokens": 26358481.0,
      "step": 12030
    },
    {
      "entropy": 5.521640825271606,
      "epoch": 1.194422389837237,
      "grad_norm": 0.875,
      "learning_rate": 0.00048655015150374904,
      "loss": 5.1968,
      "mean_token_accuracy": 0.16276587247848512,
      "num_tokens": 26370519.0,
      "step": 12035
    },
    {
      "entropy": 5.566204071044922,
      "epoch": 1.1949186184994045,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004865380823412613,
      "loss": 5.196,
      "mean_token_accuracy": 0.17306286990642547,
      "num_tokens": 26380988.0,
      "step": 12040
    },
    {
      "entropy": 5.499672985076904,
      "epoch": 1.195414847161572,
      "grad_norm": 0.96875,
      "learning_rate": 0.000486526007933085,
      "loss": 5.0886,
      "mean_token_accuracy": 0.17434173077344894,
      "num_tokens": 26390308.0,
      "step": 12045
    },
    {
      "entropy": 5.5383542537689205,
      "epoch": 1.1959110758237395,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048651392827951944,
      "loss": 5.2133,
      "mean_token_accuracy": 0.1591360315680504,
      "num_tokens": 26401145.0,
      "step": 12050
    },
    {
      "entropy": 5.5039387226104735,
      "epoch": 1.1964073044859072,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048650184338086436,
      "loss": 5.1461,
      "mean_token_accuracy": 0.16698650270700455,
      "num_tokens": 26411930.0,
      "step": 12055
    },
    {
      "entropy": 5.519377422332764,
      "epoch": 1.1969035331480746,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048648975323741916,
      "loss": 5.0848,
      "mean_token_accuracy": 0.16916512101888656,
      "num_tokens": 26422320.0,
      "step": 12060
    },
    {
      "entropy": 5.638296699523925,
      "epoch": 1.197399761810242,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048647765784948395,
      "loss": 5.2167,
      "mean_token_accuracy": 0.16180272549390792,
      "num_tokens": 26433760.0,
      "step": 12065
    },
    {
      "entropy": 5.5516104221344,
      "epoch": 1.1978959904724098,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004864655572173584,
      "loss": 5.131,
      "mean_token_accuracy": 0.1768320232629776,
      "num_tokens": 26444256.0,
      "step": 12070
    },
    {
      "entropy": 5.507214641571045,
      "epoch": 1.1983922191345773,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048645345134134274,
      "loss": 5.2217,
      "mean_token_accuracy": 0.16871612817049025,
      "num_tokens": 26456433.0,
      "step": 12075
    },
    {
      "entropy": 5.551804685592652,
      "epoch": 1.1988884477967447,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048644134022173707,
      "loss": 5.2021,
      "mean_token_accuracy": 0.16770168393850327,
      "num_tokens": 26466909.0,
      "step": 12080
    },
    {
      "entropy": 5.5663196563720705,
      "epoch": 1.1993846764589122,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004864292238588419,
      "loss": 5.1945,
      "mean_token_accuracy": 0.1715540662407875,
      "num_tokens": 26478785.0,
      "step": 12085
    },
    {
      "entropy": 5.50177788734436,
      "epoch": 1.1998809051210797,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048641710225295746,
      "loss": 5.2012,
      "mean_token_accuracy": 0.16975893527269365,
      "num_tokens": 26489596.0,
      "step": 12090
    },
    {
      "entropy": 5.4807556629180905,
      "epoch": 1.2003771337832474,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048640497540438455,
      "loss": 5.1388,
      "mean_token_accuracy": 0.17157377004623414,
      "num_tokens": 26500376.0,
      "step": 12095
    },
    {
      "entropy": 5.503072261810303,
      "epoch": 1.2008733624454149,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004863928433134237,
      "loss": 5.0882,
      "mean_token_accuracy": 0.16729052662849425,
      "num_tokens": 26510339.0,
      "step": 12100
    },
    {
      "entropy": 5.528472137451172,
      "epoch": 1.2013695911075823,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048638070598037585,
      "loss": 5.1794,
      "mean_token_accuracy": 0.16958650946617126,
      "num_tokens": 26521359.0,
      "step": 12105
    },
    {
      "entropy": 5.578713893890381,
      "epoch": 1.2018658197697498,
      "grad_norm": 1.03125,
      "learning_rate": 0.000486368563405542,
      "loss": 5.2437,
      "mean_token_accuracy": 0.16707834005355834,
      "num_tokens": 26532716.0,
      "step": 12110
    },
    {
      "entropy": 5.496238422393799,
      "epoch": 1.2023620484319175,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004863564155892232,
      "loss": 5.0969,
      "mean_token_accuracy": 0.1705687537789345,
      "num_tokens": 26542624.0,
      "step": 12115
    },
    {
      "entropy": 5.566644954681396,
      "epoch": 1.202858277094085,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004863442625317208,
      "loss": 5.2012,
      "mean_token_accuracy": 0.17068732231855394,
      "num_tokens": 26552907.0,
      "step": 12120
    },
    {
      "entropy": 5.484821224212647,
      "epoch": 1.2033545057562525,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048633210423333607,
      "loss": 5.1571,
      "mean_token_accuracy": 0.1678165689110756,
      "num_tokens": 26564871.0,
      "step": 12125
    },
    {
      "entropy": 5.482084131240844,
      "epoch": 1.20385073441842,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004863199406943705,
      "loss": 5.1392,
      "mean_token_accuracy": 0.17085685431957245,
      "num_tokens": 26576254.0,
      "step": 12130
    },
    {
      "entropy": 5.509066438674926,
      "epoch": 1.2043469630805874,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048630777191512584,
      "loss": 5.1693,
      "mean_token_accuracy": 0.17062569707632064,
      "num_tokens": 26587518.0,
      "step": 12135
    },
    {
      "entropy": 5.483522033691406,
      "epoch": 1.2048431917427551,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048629559789590366,
      "loss": 5.0852,
      "mean_token_accuracy": 0.1703111633658409,
      "num_tokens": 26597485.0,
      "step": 12140
    },
    {
      "entropy": 5.469754028320312,
      "epoch": 1.2053394204049226,
      "grad_norm": 0.984375,
      "learning_rate": 0.000486283418637006,
      "loss": 5.1367,
      "mean_token_accuracy": 0.1717156320810318,
      "num_tokens": 26607696.0,
      "step": 12145
    },
    {
      "entropy": 5.552721881866455,
      "epoch": 1.20583564906709,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004862712341387348,
      "loss": 5.1612,
      "mean_token_accuracy": 0.1707509994506836,
      "num_tokens": 26618974.0,
      "step": 12150
    },
    {
      "entropy": 5.576195907592774,
      "epoch": 1.2063318777292578,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004862590444013923,
      "loss": 5.2022,
      "mean_token_accuracy": 0.1651910662651062,
      "num_tokens": 26629949.0,
      "step": 12155
    },
    {
      "entropy": 5.54221248626709,
      "epoch": 1.2068281063914252,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004862468494252806,
      "loss": 5.1396,
      "mean_token_accuracy": 0.16881280839443208,
      "num_tokens": 26641460.0,
      "step": 12160
    },
    {
      "entropy": 5.530866956710815,
      "epoch": 1.2073243350535927,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004862346492107024,
      "loss": 5.2338,
      "mean_token_accuracy": 0.16301158666610718,
      "num_tokens": 26652553.0,
      "step": 12165
    },
    {
      "entropy": 5.480940151214599,
      "epoch": 1.2078205637157602,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048622244375796,
      "loss": 5.1511,
      "mean_token_accuracy": 0.16281547099351884,
      "num_tokens": 26662479.0,
      "step": 12170
    },
    {
      "entropy": 5.560932159423828,
      "epoch": 1.2083167923779277,
      "grad_norm": 0.875,
      "learning_rate": 0.0004862102330673561,
      "loss": 5.2016,
      "mean_token_accuracy": 0.16211413890123366,
      "num_tokens": 26673326.0,
      "step": 12175
    },
    {
      "entropy": 5.589660501480102,
      "epoch": 1.2088130210400954,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048619801713919357,
      "loss": 5.258,
      "mean_token_accuracy": 0.16120460033416747,
      "num_tokens": 26684189.0,
      "step": 12180
    },
    {
      "entropy": 5.592612171173096,
      "epoch": 1.2093092497022628,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004861857959737753,
      "loss": 5.2566,
      "mean_token_accuracy": 0.1649405136704445,
      "num_tokens": 26695399.0,
      "step": 12185
    },
    {
      "entropy": 5.570272493362427,
      "epoch": 1.2098054783644303,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004861735695714044,
      "loss": 5.1903,
      "mean_token_accuracy": 0.16377503871917726,
      "num_tokens": 26705993.0,
      "step": 12190
    },
    {
      "entropy": 5.542041778564453,
      "epoch": 1.2103017070265978,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048616133793238394,
      "loss": 5.1604,
      "mean_token_accuracy": 0.16853890866041182,
      "num_tokens": 26716120.0,
      "step": 12195
    },
    {
      "entropy": 5.574769496917725,
      "epoch": 1.2107979356887655,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048614910105701737,
      "loss": 5.2868,
      "mean_token_accuracy": 0.15985732674598693,
      "num_tokens": 26727736.0,
      "step": 12200
    },
    {
      "entropy": 5.582148265838623,
      "epoch": 1.211294164350933,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004861368589456081,
      "loss": 5.1344,
      "mean_token_accuracy": 0.17238374352455138,
      "num_tokens": 26737825.0,
      "step": 12205
    },
    {
      "entropy": 5.6176838874816895,
      "epoch": 1.2117903930131004,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004861246115984596,
      "loss": 5.2256,
      "mean_token_accuracy": 0.16376420259475707,
      "num_tokens": 26748035.0,
      "step": 12210
    },
    {
      "entropy": 5.570216703414917,
      "epoch": 1.212286621675268,
      "grad_norm": 0.8125,
      "learning_rate": 0.00048611235901587575,
      "loss": 5.1437,
      "mean_token_accuracy": 0.17248266339302062,
      "num_tokens": 26759899.0,
      "step": 12215
    },
    {
      "entropy": 5.483235931396484,
      "epoch": 1.2127828503374354,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004861001011981602,
      "loss": 5.0948,
      "mean_token_accuracy": 0.174762000143528,
      "num_tokens": 26771562.0,
      "step": 12220
    },
    {
      "entropy": 5.572576427459717,
      "epoch": 1.213279078999603,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004860878381456171,
      "loss": 5.2247,
      "mean_token_accuracy": 0.1695061057806015,
      "num_tokens": 26783002.0,
      "step": 12225
    },
    {
      "entropy": 5.534043502807617,
      "epoch": 1.2137753076617706,
      "grad_norm": 1.0,
      "learning_rate": 0.00048607556985855047,
      "loss": 5.2025,
      "mean_token_accuracy": 0.16991137713193893,
      "num_tokens": 26794136.0,
      "step": 12230
    },
    {
      "entropy": 5.45085597038269,
      "epoch": 1.214271536323938,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048606329633726447,
      "loss": 5.1786,
      "mean_token_accuracy": 0.16698249131441117,
      "num_tokens": 26804776.0,
      "step": 12235
    },
    {
      "entropy": 5.497370624542237,
      "epoch": 1.2147677649861055,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004860510175820636,
      "loss": 5.106,
      "mean_token_accuracy": 0.17415401041507722,
      "num_tokens": 26816696.0,
      "step": 12240
    },
    {
      "entropy": 5.462891387939453,
      "epoch": 1.2152639936482732,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004860387335932522,
      "loss": 5.1486,
      "mean_token_accuracy": 0.1721884787082672,
      "num_tokens": 26827789.0,
      "step": 12245
    },
    {
      "entropy": 5.540329074859619,
      "epoch": 1.2157602223104407,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004860264443711351,
      "loss": 5.1082,
      "mean_token_accuracy": 0.17041248679161072,
      "num_tokens": 26838763.0,
      "step": 12250
    },
    {
      "entropy": 5.461474275588989,
      "epoch": 1.2162564509726081,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004860141499160167,
      "loss": 5.1394,
      "mean_token_accuracy": 0.16416517943143843,
      "num_tokens": 26849304.0,
      "step": 12255
    },
    {
      "entropy": 5.502264976501465,
      "epoch": 1.2167526796347756,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004860018502282022,
      "loss": 5.2206,
      "mean_token_accuracy": 0.1691025137901306,
      "num_tokens": 26858777.0,
      "step": 12260
    },
    {
      "entropy": 5.557392120361328,
      "epoch": 1.2172489082969433,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048598954530799644,
      "loss": 5.1737,
      "mean_token_accuracy": 0.1635236620903015,
      "num_tokens": 26870286.0,
      "step": 12265
    },
    {
      "entropy": 5.545029783248902,
      "epoch": 1.2177451369591108,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048597723515570465,
      "loss": 5.1466,
      "mean_token_accuracy": 0.17426633685827256,
      "num_tokens": 26880485.0,
      "step": 12270
    },
    {
      "entropy": 5.437982892990112,
      "epoch": 1.2182413656212783,
      "grad_norm": 0.8125,
      "learning_rate": 0.000485964919771632,
      "loss": 5.0658,
      "mean_token_accuracy": 0.17580669820308686,
      "num_tokens": 26893556.0,
      "step": 12275
    },
    {
      "entropy": 5.494607257843017,
      "epoch": 1.2187375942834457,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048595259915608396,
      "loss": 5.1416,
      "mean_token_accuracy": 0.16553050130605698,
      "num_tokens": 26905699.0,
      "step": 12280
    },
    {
      "entropy": 5.532801342010498,
      "epoch": 1.2192338229456134,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048594027330936594,
      "loss": 5.1605,
      "mean_token_accuracy": 0.16534959375858307,
      "num_tokens": 26917188.0,
      "step": 12285
    },
    {
      "entropy": 5.546386432647705,
      "epoch": 1.219730051607781,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048592794223178374,
      "loss": 5.1587,
      "mean_token_accuracy": 0.16714199930429458,
      "num_tokens": 26927905.0,
      "step": 12290
    },
    {
      "entropy": 5.456134510040283,
      "epoch": 1.2202262802699484,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048591560592364314,
      "loss": 5.1271,
      "mean_token_accuracy": 0.17412807792425156,
      "num_tokens": 26938665.0,
      "step": 12295
    },
    {
      "entropy": 5.508511304855347,
      "epoch": 1.2207225089321159,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004859032643852499,
      "loss": 5.2274,
      "mean_token_accuracy": 0.16469652503728865,
      "num_tokens": 26950516.0,
      "step": 12300
    },
    {
      "entropy": 5.593306541442871,
      "epoch": 1.2212187375942833,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004858909176169102,
      "loss": 5.2159,
      "mean_token_accuracy": 0.17017593383789062,
      "num_tokens": 26960018.0,
      "step": 12305
    },
    {
      "entropy": 5.524492979049683,
      "epoch": 1.221714966256451,
      "grad_norm": 1.0,
      "learning_rate": 0.00048587856561893014,
      "loss": 5.1814,
      "mean_token_accuracy": 0.16484884470701217,
      "num_tokens": 26969825.0,
      "step": 12310
    },
    {
      "entropy": 5.563662004470825,
      "epoch": 1.2222111949186185,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004858662083916161,
      "loss": 5.2684,
      "mean_token_accuracy": 0.16246035248041152,
      "num_tokens": 26979873.0,
      "step": 12315
    },
    {
      "entropy": 5.585719299316406,
      "epoch": 1.222707423580786,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004858538459352744,
      "loss": 5.25,
      "mean_token_accuracy": 0.1645125925540924,
      "num_tokens": 26990577.0,
      "step": 12320
    },
    {
      "entropy": 5.519025754928589,
      "epoch": 1.2232036522429535,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004858414782502117,
      "loss": 5.0913,
      "mean_token_accuracy": 0.1726545885205269,
      "num_tokens": 27001598.0,
      "step": 12325
    },
    {
      "entropy": 5.451758193969726,
      "epoch": 1.2236998809051212,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048582910533673474,
      "loss": 5.0994,
      "mean_token_accuracy": 0.17842347621917726,
      "num_tokens": 27012535.0,
      "step": 12330
    },
    {
      "entropy": 5.509342813491822,
      "epoch": 1.2241961095672886,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004858167271951502,
      "loss": 5.1768,
      "mean_token_accuracy": 0.16546874791383742,
      "num_tokens": 27021867.0,
      "step": 12335
    },
    {
      "entropy": 5.454013967514038,
      "epoch": 1.2246923382294561,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004858043438257651,
      "loss": 5.1572,
      "mean_token_accuracy": 0.1679985836148262,
      "num_tokens": 27032540.0,
      "step": 12340
    },
    {
      "entropy": 5.428150987625122,
      "epoch": 1.2251885668916236,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004857919552288865,
      "loss": 5.1154,
      "mean_token_accuracy": 0.1721821054816246,
      "num_tokens": 27044008.0,
      "step": 12345
    },
    {
      "entropy": 5.546235132217407,
      "epoch": 1.225684795553791,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004857795614048217,
      "loss": 5.1418,
      "mean_token_accuracy": 0.17591281235218048,
      "num_tokens": 27053646.0,
      "step": 12350
    },
    {
      "entropy": 5.612762546539306,
      "epoch": 1.2261810242159588,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048576716235387786,
      "loss": 5.2688,
      "mean_token_accuracy": 0.16229404360055924,
      "num_tokens": 27065834.0,
      "step": 12355
    },
    {
      "entropy": 5.516036462783814,
      "epoch": 1.2266772528781262,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048575475807636267,
      "loss": 5.2008,
      "mean_token_accuracy": 0.1633292019367218,
      "num_tokens": 27077544.0,
      "step": 12360
    },
    {
      "entropy": 5.559359359741211,
      "epoch": 1.2271734815402937,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048574234857258356,
      "loss": 5.2486,
      "mean_token_accuracy": 0.1663970723748207,
      "num_tokens": 27088294.0,
      "step": 12365
    },
    {
      "entropy": 5.554310035705567,
      "epoch": 1.2276697102024614,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048572993384284833,
      "loss": 5.1804,
      "mean_token_accuracy": 0.16529619842767715,
      "num_tokens": 27098848.0,
      "step": 12370
    },
    {
      "entropy": 5.4490556716918945,
      "epoch": 1.2281659388646289,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004857175138874648,
      "loss": 5.1498,
      "mean_token_accuracy": 0.17082599699497222,
      "num_tokens": 27112319.0,
      "step": 12375
    },
    {
      "entropy": 5.548217105865478,
      "epoch": 1.2286621675267964,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048570508870674097,
      "loss": 5.2044,
      "mean_token_accuracy": 0.17085884213447572,
      "num_tokens": 27121840.0,
      "step": 12380
    },
    {
      "entropy": 5.558666038513183,
      "epoch": 1.2291583961889638,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048569265830098493,
      "loss": 5.2095,
      "mean_token_accuracy": 0.16960825473070146,
      "num_tokens": 27133619.0,
      "step": 12385
    },
    {
      "entropy": 5.523226165771485,
      "epoch": 1.2296546248511313,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000485680222670505,
      "loss": 5.1867,
      "mean_token_accuracy": 0.1673790395259857,
      "num_tokens": 27144161.0,
      "step": 12390
    },
    {
      "entropy": 5.499545097351074,
      "epoch": 1.230150853513299,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004856677818156096,
      "loss": 5.1088,
      "mean_token_accuracy": 0.17127839028835296,
      "num_tokens": 27154906.0,
      "step": 12395
    },
    {
      "entropy": 5.50167932510376,
      "epoch": 1.2306470821754665,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048565533573660707,
      "loss": 5.1376,
      "mean_token_accuracy": 0.1709315612912178,
      "num_tokens": 27165746.0,
      "step": 12400
    },
    {
      "entropy": 5.4959306716918945,
      "epoch": 1.231143310837634,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004856428844338062,
      "loss": 5.1221,
      "mean_token_accuracy": 0.1723973944783211,
      "num_tokens": 27176108.0,
      "step": 12405
    },
    {
      "entropy": 5.496760559082031,
      "epoch": 1.2316395394998014,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004856304279075156,
      "loss": 5.2017,
      "mean_token_accuracy": 0.1688024580478668,
      "num_tokens": 27186721.0,
      "step": 12410
    },
    {
      "entropy": 5.500027847290039,
      "epoch": 1.2321357681619691,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004856179661580444,
      "loss": 5.0749,
      "mean_token_accuracy": 0.17624875605106355,
      "num_tokens": 27196120.0,
      "step": 12415
    },
    {
      "entropy": 5.55590295791626,
      "epoch": 1.2326319968241366,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004856054991857013,
      "loss": 5.2036,
      "mean_token_accuracy": 0.1612488016486168,
      "num_tokens": 27207431.0,
      "step": 12420
    },
    {
      "entropy": 5.4861139297485355,
      "epoch": 1.233128225486304,
      "grad_norm": 0.79296875,
      "learning_rate": 0.00048559302699079576,
      "loss": 5.0917,
      "mean_token_accuracy": 0.1792071908712387,
      "num_tokens": 27218489.0,
      "step": 12425
    },
    {
      "entropy": 5.582187223434448,
      "epoch": 1.2336244541484715,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004855805495736369,
      "loss": 5.2163,
      "mean_token_accuracy": 0.1659310668706894,
      "num_tokens": 27230849.0,
      "step": 12430
    },
    {
      "entropy": 5.551375389099121,
      "epoch": 1.234120682810639,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004855680669345342,
      "loss": 5.1717,
      "mean_token_accuracy": 0.16809328496456147,
      "num_tokens": 27242838.0,
      "step": 12435
    },
    {
      "entropy": 5.552640104293824,
      "epoch": 1.2346169114728067,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00048555557907379714,
      "loss": 5.1923,
      "mean_token_accuracy": 0.17570284008979797,
      "num_tokens": 27253945.0,
      "step": 12440
    },
    {
      "entropy": 5.518435430526734,
      "epoch": 1.2351131401349742,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048554308599173544,
      "loss": 5.2638,
      "mean_token_accuracy": 0.16034649312496185,
      "num_tokens": 27263999.0,
      "step": 12445
    },
    {
      "entropy": 5.5560914993286135,
      "epoch": 1.2356093687971417,
      "grad_norm": 1.0,
      "learning_rate": 0.0004855305876886589,
      "loss": 5.0993,
      "mean_token_accuracy": 0.175088270008564,
      "num_tokens": 27274677.0,
      "step": 12450
    },
    {
      "entropy": 5.53433256149292,
      "epoch": 1.2361055974593094,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00048551808416487743,
      "loss": 5.1836,
      "mean_token_accuracy": 0.16513915061950685,
      "num_tokens": 27286512.0,
      "step": 12455
    },
    {
      "entropy": 5.4974696159362795,
      "epoch": 1.2366018261214768,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004855055754207011,
      "loss": 5.1836,
      "mean_token_accuracy": 0.17457702606916428,
      "num_tokens": 27297620.0,
      "step": 12460
    },
    {
      "entropy": 5.564794731140137,
      "epoch": 1.2370980547836443,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004854930614564401,
      "loss": 5.1921,
      "mean_token_accuracy": 0.16229198276996612,
      "num_tokens": 27309963.0,
      "step": 12465
    },
    {
      "entropy": 5.628265762329102,
      "epoch": 1.2375942834458118,
      "grad_norm": 1.0,
      "learning_rate": 0.0004854805422724047,
      "loss": 5.2779,
      "mean_token_accuracy": 0.16460468918085097,
      "num_tokens": 27320740.0,
      "step": 12470
    },
    {
      "entropy": 5.462673139572144,
      "epoch": 1.2380905121079793,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004854680178689055,
      "loss": 5.0968,
      "mean_token_accuracy": 0.17474846988916398,
      "num_tokens": 27331543.0,
      "step": 12475
    },
    {
      "entropy": 5.442281913757324,
      "epoch": 1.238586740770147,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048545548824625287,
      "loss": 5.0829,
      "mean_token_accuracy": 0.17312445342540742,
      "num_tokens": 27343222.0,
      "step": 12480
    },
    {
      "entropy": 5.532667112350464,
      "epoch": 1.2390829694323144,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004854429534047577,
      "loss": 5.1463,
      "mean_token_accuracy": 0.16827509850263594,
      "num_tokens": 27354694.0,
      "step": 12485
    },
    {
      "entropy": 5.481135368347168,
      "epoch": 1.239579198094482,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004854304133447307,
      "loss": 5.1338,
      "mean_token_accuracy": 0.1677487835288048,
      "num_tokens": 27365462.0,
      "step": 12490
    },
    {
      "entropy": 5.442475605010986,
      "epoch": 1.2400754267566494,
      "grad_norm": 0.875,
      "learning_rate": 0.00048541786806648293,
      "loss": 5.0657,
      "mean_token_accuracy": 0.1727943554520607,
      "num_tokens": 27377548.0,
      "step": 12495
    },
    {
      "entropy": 5.555824899673462,
      "epoch": 1.240571655418817,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004854053175703254,
      "loss": 5.1794,
      "mean_token_accuracy": 0.16493744403123856,
      "num_tokens": 27388221.0,
      "step": 12500
    },
    {
      "entropy": 5.5570111751556395,
      "epoch": 1.2410678840809846,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004853927618565694,
      "loss": 5.1765,
      "mean_token_accuracy": 0.16697217226028443,
      "num_tokens": 27400497.0,
      "step": 12505
    },
    {
      "entropy": 5.483710765838623,
      "epoch": 1.241564112743152,
      "grad_norm": 1.0,
      "learning_rate": 0.00048538020092552626,
      "loss": 5.0649,
      "mean_token_accuracy": 0.17475847899913788,
      "num_tokens": 27410700.0,
      "step": 12510
    },
    {
      "entropy": 5.513055229187012,
      "epoch": 1.2420603414053195,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048536763477750747,
      "loss": 5.2065,
      "mean_token_accuracy": 0.17092445492744446,
      "num_tokens": 27421286.0,
      "step": 12515
    },
    {
      "entropy": 5.579060459136963,
      "epoch": 1.242556570067487,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004853550634128246,
      "loss": 5.2226,
      "mean_token_accuracy": 0.1669715538620949,
      "num_tokens": 27433922.0,
      "step": 12520
    },
    {
      "entropy": 5.494808101654053,
      "epoch": 1.2430527987296547,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004853424868317895,
      "loss": 5.1517,
      "mean_token_accuracy": 0.16734260618686675,
      "num_tokens": 27444684.0,
      "step": 12525
    },
    {
      "entropy": 5.490879917144776,
      "epoch": 1.2435490273918222,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004853299050347139,
      "loss": 5.1393,
      "mean_token_accuracy": 0.16786016672849655,
      "num_tokens": 27455494.0,
      "step": 12530
    },
    {
      "entropy": 5.602056550979614,
      "epoch": 1.2440452560539896,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048531731802190994,
      "loss": 5.1743,
      "mean_token_accuracy": 0.167215932905674,
      "num_tokens": 27464927.0,
      "step": 12535
    },
    {
      "entropy": 5.576707363128662,
      "epoch": 1.244541484716157,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004853047257936897,
      "loss": 5.182,
      "mean_token_accuracy": 0.16496121436357497,
      "num_tokens": 27476228.0,
      "step": 12540
    },
    {
      "entropy": 5.550760507583618,
      "epoch": 1.2450377133783248,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004852921283503653,
      "loss": 5.1847,
      "mean_token_accuracy": 0.17001437842845918,
      "num_tokens": 27488835.0,
      "step": 12545
    },
    {
      "entropy": 5.517337465286255,
      "epoch": 1.2455339420404923,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004852795256922494,
      "loss": 5.1586,
      "mean_token_accuracy": 0.16673565804958343,
      "num_tokens": 27501481.0,
      "step": 12550
    },
    {
      "entropy": 5.536225080490112,
      "epoch": 1.2460301707026598,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048526691781965425,
      "loss": 5.1774,
      "mean_token_accuracy": 0.16835180521011353,
      "num_tokens": 27512234.0,
      "step": 12555
    },
    {
      "entropy": 5.605270719528198,
      "epoch": 1.2465263993648272,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048525430473289274,
      "loss": 5.2305,
      "mean_token_accuracy": 0.16284125298261642,
      "num_tokens": 27523756.0,
      "step": 12560
    },
    {
      "entropy": 5.5720970153808596,
      "epoch": 1.247022628026995,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048524168643227747,
      "loss": 5.2201,
      "mean_token_accuracy": 0.1646725729107857,
      "num_tokens": 27534579.0,
      "step": 12565
    },
    {
      "entropy": 5.514043283462525,
      "epoch": 1.2475188566891624,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048522906291812145,
      "loss": 5.2538,
      "mean_token_accuracy": 0.16246421188116072,
      "num_tokens": 27546420.0,
      "step": 12570
    },
    {
      "entropy": 5.549548816680908,
      "epoch": 1.2480150853513299,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048521643419073763,
      "loss": 5.119,
      "mean_token_accuracy": 0.16686424612998962,
      "num_tokens": 27557701.0,
      "step": 12575
    },
    {
      "entropy": 5.480019235610962,
      "epoch": 1.2485113140134974,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004852038002504392,
      "loss": 5.1447,
      "mean_token_accuracy": 0.17416032105684282,
      "num_tokens": 27568592.0,
      "step": 12580
    },
    {
      "entropy": 5.536455869674683,
      "epoch": 1.249007542675665,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048519116109753955,
      "loss": 5.177,
      "mean_token_accuracy": 0.1732865184545517,
      "num_tokens": 27579286.0,
      "step": 12585
    },
    {
      "entropy": 5.517602252960205,
      "epoch": 1.2495037713378325,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048517851673235196,
      "loss": 5.1178,
      "mean_token_accuracy": 0.16958713978528978,
      "num_tokens": 27590031.0,
      "step": 12590
    },
    {
      "entropy": 5.517959785461426,
      "epoch": 1.25,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004851658671551901,
      "loss": 5.1522,
      "mean_token_accuracy": 0.16750041544437408,
      "num_tokens": 27600999.0,
      "step": 12595
    },
    {
      "entropy": 5.492237997055054,
      "epoch": 1.2504962286621675,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048515321236636754,
      "loss": 5.1472,
      "mean_token_accuracy": 0.16664251834154128,
      "num_tokens": 27613394.0,
      "step": 12600
    },
    {
      "entropy": 5.505588483810425,
      "epoch": 1.250992457324335,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004851405523661982,
      "loss": 5.1262,
      "mean_token_accuracy": 0.17925361543893814,
      "num_tokens": 27624819.0,
      "step": 12605
    },
    {
      "entropy": 5.46622142791748,
      "epoch": 1.2514886859865026,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048512788715499597,
      "loss": 5.1584,
      "mean_token_accuracy": 0.17537716329097747,
      "num_tokens": 27635993.0,
      "step": 12610
    },
    {
      "entropy": 5.516826629638672,
      "epoch": 1.2519849146486701,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004851152167330749,
      "loss": 5.2176,
      "mean_token_accuracy": 0.16609541624784468,
      "num_tokens": 27647298.0,
      "step": 12615
    },
    {
      "entropy": 5.586860036849975,
      "epoch": 1.2524811433108376,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048510254110074923,
      "loss": 5.1887,
      "mean_token_accuracy": 0.17025675028562545,
      "num_tokens": 27659381.0,
      "step": 12620
    },
    {
      "entropy": 5.547328615188599,
      "epoch": 1.2529773719730053,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004850898602583333,
      "loss": 5.175,
      "mean_token_accuracy": 0.16522342711687088,
      "num_tokens": 27671122.0,
      "step": 12625
    },
    {
      "entropy": 5.532905769348145,
      "epoch": 1.2534736006351728,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048507717420614153,
      "loss": 5.1663,
      "mean_token_accuracy": 0.16918721348047255,
      "num_tokens": 27683258.0,
      "step": 12630
    },
    {
      "entropy": 5.427686595916748,
      "epoch": 1.2539698292973402,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004850644829444884,
      "loss": 5.031,
      "mean_token_accuracy": 0.1729172334074974,
      "num_tokens": 27693116.0,
      "step": 12635
    },
    {
      "entropy": 5.469030237197876,
      "epoch": 1.2544660579595077,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048505178647368887,
      "loss": 5.1263,
      "mean_token_accuracy": 0.1712686911225319,
      "num_tokens": 27704859.0,
      "step": 12640
    },
    {
      "entropy": 5.478932428359985,
      "epoch": 1.2549622866216752,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004850390847940576,
      "loss": 5.0782,
      "mean_token_accuracy": 0.1733901709318161,
      "num_tokens": 27715854.0,
      "step": 12645
    },
    {
      "entropy": 5.507164812088012,
      "epoch": 1.2554585152838427,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004850263779059097,
      "loss": 5.1127,
      "mean_token_accuracy": 0.17149351239204408,
      "num_tokens": 27726398.0,
      "step": 12650
    },
    {
      "entropy": 5.538310480117798,
      "epoch": 1.2559547439460104,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004850136658095601,
      "loss": 5.1627,
      "mean_token_accuracy": 0.17338285744190216,
      "num_tokens": 27737337.0,
      "step": 12655
    },
    {
      "entropy": 5.516957759857178,
      "epoch": 1.2564509726081778,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048500094850532417,
      "loss": 5.1781,
      "mean_token_accuracy": 0.17100569158792495,
      "num_tokens": 27748593.0,
      "step": 12660
    },
    {
      "entropy": 5.572302961349488,
      "epoch": 1.2569472012703453,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004849882259935172,
      "loss": 5.1399,
      "mean_token_accuracy": 0.1682426154613495,
      "num_tokens": 27760908.0,
      "step": 12665
    },
    {
      "entropy": 5.55339789390564,
      "epoch": 1.257443429932513,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004849754982744548,
      "loss": 5.2207,
      "mean_token_accuracy": 0.16648263186216355,
      "num_tokens": 27772104.0,
      "step": 12670
    },
    {
      "entropy": 5.467471742630005,
      "epoch": 1.2579396585946805,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004849627653484524,
      "loss": 5.1654,
      "mean_token_accuracy": 0.17432896196842193,
      "num_tokens": 27781654.0,
      "step": 12675
    },
    {
      "entropy": 5.527649021148681,
      "epoch": 1.258435887256848,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004849500272158259,
      "loss": 5.1492,
      "mean_token_accuracy": 0.17107137888669968,
      "num_tokens": 27793026.0,
      "step": 12680
    },
    {
      "entropy": 5.543030500411987,
      "epoch": 1.2589321159190154,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004849372838768911,
      "loss": 5.1253,
      "mean_token_accuracy": 0.17380276918411255,
      "num_tokens": 27803447.0,
      "step": 12685
    },
    {
      "entropy": 5.6290388107299805,
      "epoch": 1.259428344581183,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000484924535331964,
      "loss": 5.1359,
      "mean_token_accuracy": 0.1681838095188141,
      "num_tokens": 27814632.0,
      "step": 12690
    },
    {
      "entropy": 5.545263957977295,
      "epoch": 1.2599245732433506,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048491178158136086,
      "loss": 5.2237,
      "mean_token_accuracy": 0.1642076313495636,
      "num_tokens": 27826582.0,
      "step": 12695
    },
    {
      "entropy": 5.516428422927857,
      "epoch": 1.260420801905518,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004848990226253978,
      "loss": 5.0708,
      "mean_token_accuracy": 0.1688187152147293,
      "num_tokens": 27838065.0,
      "step": 12700
    },
    {
      "entropy": 5.484880828857422,
      "epoch": 1.2609170305676856,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048488625846439135,
      "loss": 5.1378,
      "mean_token_accuracy": 0.17219123542308806,
      "num_tokens": 27848182.0,
      "step": 12705
    },
    {
      "entropy": 5.576004219055176,
      "epoch": 1.261413259229853,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004848734890986578,
      "loss": 5.1405,
      "mean_token_accuracy": 0.1705386146903038,
      "num_tokens": 27858524.0,
      "step": 12710
    },
    {
      "entropy": 5.535070991516113,
      "epoch": 1.2619094878920207,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004848607145285141,
      "loss": 5.0627,
      "mean_token_accuracy": 0.17471428215503693,
      "num_tokens": 27868588.0,
      "step": 12715
    },
    {
      "entropy": 5.494311952590943,
      "epoch": 1.2624057165541882,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004848479347542768,
      "loss": 5.0815,
      "mean_token_accuracy": 0.1724834606051445,
      "num_tokens": 27878829.0,
      "step": 12720
    },
    {
      "entropy": 5.500310945510864,
      "epoch": 1.2629019452163557,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000484835149776263,
      "loss": 5.1559,
      "mean_token_accuracy": 0.17567149102687835,
      "num_tokens": 27889872.0,
      "step": 12725
    },
    {
      "entropy": 5.510600996017456,
      "epoch": 1.2633981738785232,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048482235959478953,
      "loss": 5.2206,
      "mean_token_accuracy": 0.16926396936178206,
      "num_tokens": 27900567.0,
      "step": 12730
    },
    {
      "entropy": 5.555635976791382,
      "epoch": 1.2638944025406906,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004848095642101737,
      "loss": 5.1013,
      "mean_token_accuracy": 0.17302893251180648,
      "num_tokens": 27911594.0,
      "step": 12735
    },
    {
      "entropy": 5.600049161911011,
      "epoch": 1.2643906312028583,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048479676362273277,
      "loss": 5.1975,
      "mean_token_accuracy": 0.16255032867193223,
      "num_tokens": 27923694.0,
      "step": 12740
    },
    {
      "entropy": 5.554832839965821,
      "epoch": 1.2648868598650258,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048478395783278425,
      "loss": 5.2844,
      "mean_token_accuracy": 0.15380905121564864,
      "num_tokens": 27934168.0,
      "step": 12745
    },
    {
      "entropy": 5.585643911361695,
      "epoch": 1.2653830885271933,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004847711468406455,
      "loss": 5.171,
      "mean_token_accuracy": 0.16681780964136123,
      "num_tokens": 27944401.0,
      "step": 12750
    },
    {
      "entropy": 5.479246711730957,
      "epoch": 1.265879317189361,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048475833064663443,
      "loss": 5.0627,
      "mean_token_accuracy": 0.17939431369304656,
      "num_tokens": 27956219.0,
      "step": 12755
    },
    {
      "entropy": 5.5214111328125,
      "epoch": 1.2663755458515285,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004847455092510687,
      "loss": 5.1664,
      "mean_token_accuracy": 0.17302629798650743,
      "num_tokens": 27967457.0,
      "step": 12760
    },
    {
      "entropy": 5.510754060745239,
      "epoch": 1.266871774513696,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004847326826542663,
      "loss": 5.1747,
      "mean_token_accuracy": 0.16621131598949432,
      "num_tokens": 27978287.0,
      "step": 12765
    },
    {
      "entropy": 5.55013256072998,
      "epoch": 1.2673680031758634,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048471985085654537,
      "loss": 5.1987,
      "mean_token_accuracy": 0.17439829111099242,
      "num_tokens": 27989703.0,
      "step": 12770
    },
    {
      "entropy": 5.536842489242554,
      "epoch": 1.2678642318380309,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048470701385822394,
      "loss": 5.1628,
      "mean_token_accuracy": 0.1690773770213127,
      "num_tokens": 28002071.0,
      "step": 12775
    },
    {
      "entropy": 5.555324363708496,
      "epoch": 1.2683604605001986,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048469417165962047,
      "loss": 5.2169,
      "mean_token_accuracy": 0.16876873821020127,
      "num_tokens": 28014517.0,
      "step": 12780
    },
    {
      "entropy": 5.541184186935425,
      "epoch": 1.268856689162366,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004846813242610535,
      "loss": 5.1695,
      "mean_token_accuracy": 0.17397904694080352,
      "num_tokens": 28025135.0,
      "step": 12785
    },
    {
      "entropy": 5.485084056854248,
      "epoch": 1.2693529178245335,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004846684716628414,
      "loss": 5.1031,
      "mean_token_accuracy": 0.1720394656062126,
      "num_tokens": 28036105.0,
      "step": 12790
    },
    {
      "entropy": 5.490474700927734,
      "epoch": 1.269849146486701,
      "grad_norm": 0.96875,
      "learning_rate": 0.000484655613865303,
      "loss": 5.1457,
      "mean_token_accuracy": 0.17245976030826568,
      "num_tokens": 28048825.0,
      "step": 12795
    },
    {
      "entropy": 5.5143452167510985,
      "epoch": 1.2703453751488687,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048464275086875717,
      "loss": 5.0982,
      "mean_token_accuracy": 0.17062480598688126,
      "num_tokens": 28060408.0,
      "step": 12800
    },
    {
      "entropy": 5.554420852661133,
      "epoch": 1.2708416038110362,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004846298826735229,
      "loss": 5.18,
      "mean_token_accuracy": 0.1619243785738945,
      "num_tokens": 28072179.0,
      "step": 12805
    },
    {
      "entropy": 5.509398078918457,
      "epoch": 1.2713378324732036,
      "grad_norm": 1.125,
      "learning_rate": 0.0004846170092799192,
      "loss": 5.1451,
      "mean_token_accuracy": 0.16915974915027618,
      "num_tokens": 28083825.0,
      "step": 12810
    },
    {
      "entropy": 5.6818922996521,
      "epoch": 1.2718340611353711,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048460413068826536,
      "loss": 5.3056,
      "mean_token_accuracy": 0.16124261915683746,
      "num_tokens": 28095136.0,
      "step": 12815
    },
    {
      "entropy": 5.570952320098877,
      "epoch": 1.2723302897975386,
      "grad_norm": 1.0,
      "learning_rate": 0.0004845912468988807,
      "loss": 5.1261,
      "mean_token_accuracy": 0.16855156868696214,
      "num_tokens": 28104763.0,
      "step": 12820
    },
    {
      "entropy": 5.528598594665527,
      "epoch": 1.2728265184597063,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004845783579120849,
      "loss": 5.1798,
      "mean_token_accuracy": 0.16893855780363082,
      "num_tokens": 28115771.0,
      "step": 12825
    },
    {
      "entropy": 5.492273235321045,
      "epoch": 1.2733227471218738,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004845654637281973,
      "loss": 5.1078,
      "mean_token_accuracy": 0.16819423139095308,
      "num_tokens": 28128039.0,
      "step": 12830
    },
    {
      "entropy": 5.486709690093994,
      "epoch": 1.2738189757840412,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048455256434753774,
      "loss": 5.1206,
      "mean_token_accuracy": 0.1762232467532158,
      "num_tokens": 28139163.0,
      "step": 12835
    },
    {
      "entropy": 5.484812641143799,
      "epoch": 1.274315204446209,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004845396597704262,
      "loss": 5.1585,
      "mean_token_accuracy": 0.16834336072206496,
      "num_tokens": 28150625.0,
      "step": 12840
    },
    {
      "entropy": 5.569616365432739,
      "epoch": 1.2748114331083764,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048452674999718257,
      "loss": 5.2309,
      "mean_token_accuracy": 0.16711384654045106,
      "num_tokens": 28163072.0,
      "step": 12845
    },
    {
      "entropy": 5.623217678070068,
      "epoch": 1.2753076617705439,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048451383502812705,
      "loss": 5.2518,
      "mean_token_accuracy": 0.16317146867513657,
      "num_tokens": 28174192.0,
      "step": 12850
    },
    {
      "entropy": 5.48516263961792,
      "epoch": 1.2758038904327114,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048450091486357987,
      "loss": 5.1015,
      "mean_token_accuracy": 0.17857202589511872,
      "num_tokens": 28184477.0,
      "step": 12855
    },
    {
      "entropy": 5.472021150588989,
      "epoch": 1.2763001190948788,
      "grad_norm": 0.875,
      "learning_rate": 0.00048448798950386147,
      "loss": 5.0274,
      "mean_token_accuracy": 0.17695969194173813,
      "num_tokens": 28194552.0,
      "step": 12860
    },
    {
      "entropy": 5.510885238647461,
      "epoch": 1.2767963477570463,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048447505894929233,
      "loss": 5.1973,
      "mean_token_accuracy": 0.1698785126209259,
      "num_tokens": 28205712.0,
      "step": 12865
    },
    {
      "entropy": 5.549971055984497,
      "epoch": 1.277292576419214,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048446212320019306,
      "loss": 5.1305,
      "mean_token_accuracy": 0.17509013861417771,
      "num_tokens": 28217246.0,
      "step": 12870
    },
    {
      "entropy": 5.52444543838501,
      "epoch": 1.2777888050813815,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004844491822568846,
      "loss": 5.1486,
      "mean_token_accuracy": 0.16876955032348634,
      "num_tokens": 28227933.0,
      "step": 12875
    },
    {
      "entropy": 5.517069387435913,
      "epoch": 1.278285033743549,
      "grad_norm": 0.84375,
      "learning_rate": 0.00048443623611968764,
      "loss": 5.1709,
      "mean_token_accuracy": 0.17192982733249665,
      "num_tokens": 28239692.0,
      "step": 12880
    },
    {
      "entropy": 5.503536605834961,
      "epoch": 1.2787812624057167,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048442328478892334,
      "loss": 5.1422,
      "mean_token_accuracy": 0.1741974472999573,
      "num_tokens": 28249815.0,
      "step": 12885
    },
    {
      "entropy": 5.57446665763855,
      "epoch": 1.2792774910678841,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004844103282649128,
      "loss": 5.171,
      "mean_token_accuracy": 0.16653064787387847,
      "num_tokens": 28260385.0,
      "step": 12890
    },
    {
      "entropy": 5.502856588363647,
      "epoch": 1.2797737197300516,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004843973665479774,
      "loss": 5.1807,
      "mean_token_accuracy": 0.16114374548196791,
      "num_tokens": 28271858.0,
      "step": 12895
    },
    {
      "entropy": 5.567032337188721,
      "epoch": 1.280269948392219,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048438439963843853,
      "loss": 5.1788,
      "mean_token_accuracy": 0.1681448847055435,
      "num_tokens": 28283988.0,
      "step": 12900
    },
    {
      "entropy": 5.5415338516235355,
      "epoch": 1.2807661770543866,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048437142753661775,
      "loss": 5.1976,
      "mean_token_accuracy": 0.17178643643856048,
      "num_tokens": 28293281.0,
      "step": 12905
    },
    {
      "entropy": 5.614867639541626,
      "epoch": 1.2812624057165543,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004843584502428367,
      "loss": 5.1541,
      "mean_token_accuracy": 0.16519233137369155,
      "num_tokens": 28304468.0,
      "step": 12910
    },
    {
      "entropy": 5.446540594100952,
      "epoch": 1.2817586343787217,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004843454677574173,
      "loss": 5.1069,
      "mean_token_accuracy": 0.1761067196726799,
      "num_tokens": 28315864.0,
      "step": 12915
    },
    {
      "entropy": 5.501544523239136,
      "epoch": 1.2822548630408892,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048433248008068123,
      "loss": 5.1761,
      "mean_token_accuracy": 0.1747245654463768,
      "num_tokens": 28327212.0,
      "step": 12920
    },
    {
      "entropy": 5.572963809967041,
      "epoch": 1.282751091703057,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048431948721295085,
      "loss": 5.1713,
      "mean_token_accuracy": 0.17358099520206452,
      "num_tokens": 28339256.0,
      "step": 12925
    },
    {
      "entropy": 5.55533127784729,
      "epoch": 1.2832473203652244,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004843064891545482,
      "loss": 5.1795,
      "mean_token_accuracy": 0.17052376717329026,
      "num_tokens": 28350196.0,
      "step": 12930
    },
    {
      "entropy": 5.530222606658936,
      "epoch": 1.2837435490273919,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004842934859057956,
      "loss": 5.1561,
      "mean_token_accuracy": 0.16836240887641907,
      "num_tokens": 28361241.0,
      "step": 12935
    },
    {
      "entropy": 5.589635801315308,
      "epoch": 1.2842397776895593,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004842804774670156,
      "loss": 5.2759,
      "mean_token_accuracy": 0.16310614496469497,
      "num_tokens": 28373369.0,
      "step": 12940
    },
    {
      "entropy": 5.547092342376709,
      "epoch": 1.2847360063517268,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048426746383853066,
      "loss": 5.1043,
      "mean_token_accuracy": 0.17407662719488143,
      "num_tokens": 28384970.0,
      "step": 12945
    },
    {
      "entropy": 5.540127944946289,
      "epoch": 1.2852322350138943,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048425444502066354,
      "loss": 5.1157,
      "mean_token_accuracy": 0.17156493216753005,
      "num_tokens": 28396469.0,
      "step": 12950
    },
    {
      "entropy": 5.512262153625488,
      "epoch": 1.285728463676062,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048424142101373706,
      "loss": 5.1805,
      "mean_token_accuracy": 0.169954714179039,
      "num_tokens": 28406975.0,
      "step": 12955
    },
    {
      "entropy": 5.466426658630371,
      "epoch": 1.2862246923382294,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048422839181807426,
      "loss": 5.0496,
      "mean_token_accuracy": 0.18006895929574968,
      "num_tokens": 28418285.0,
      "step": 12960
    },
    {
      "entropy": 5.495448303222656,
      "epoch": 1.286720921000397,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004842153574339982,
      "loss": 5.1984,
      "mean_token_accuracy": 0.16441502571105956,
      "num_tokens": 28429769.0,
      "step": 12965
    },
    {
      "entropy": 5.499081182479858,
      "epoch": 1.2872171496625646,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048420231786183216,
      "loss": 5.1408,
      "mean_token_accuracy": 0.17683082073926926,
      "num_tokens": 28441188.0,
      "step": 12970
    },
    {
      "entropy": 5.597442388534546,
      "epoch": 1.287713378324732,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004841892731018994,
      "loss": 5.2034,
      "mean_token_accuracy": 0.16595323234796525,
      "num_tokens": 28453504.0,
      "step": 12975
    },
    {
      "entropy": 5.568818998336792,
      "epoch": 1.2882096069868996,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004841762231545233,
      "loss": 5.186,
      "mean_token_accuracy": 0.1743595689535141,
      "num_tokens": 28465930.0,
      "step": 12980
    },
    {
      "entropy": 5.509972286224365,
      "epoch": 1.288705835649067,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048416316802002776,
      "loss": 5.1023,
      "mean_token_accuracy": 0.17614503800868989,
      "num_tokens": 28477803.0,
      "step": 12985
    },
    {
      "entropy": 5.570358085632324,
      "epoch": 1.2892020643112345,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048415010769873625,
      "loss": 5.2103,
      "mean_token_accuracy": 0.1656086727976799,
      "num_tokens": 28488556.0,
      "step": 12990
    },
    {
      "entropy": 5.561723375320435,
      "epoch": 1.2896982929734022,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004841370421909728,
      "loss": 5.1122,
      "mean_token_accuracy": 0.17091791033744813,
      "num_tokens": 28499275.0,
      "step": 12995
    },
    {
      "entropy": 5.56761794090271,
      "epoch": 1.2901945216355697,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048412397149706133,
      "loss": 5.149,
      "mean_token_accuracy": 0.17340620160102843,
      "num_tokens": 28511327.0,
      "step": 13000
    },
    {
      "entropy": 5.521489238739013,
      "epoch": 1.2906907502977372,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000484110895617326,
      "loss": 5.1802,
      "mean_token_accuracy": 0.17062077671289444,
      "num_tokens": 28523937.0,
      "step": 13005
    },
    {
      "entropy": 5.437697410583496,
      "epoch": 1.2911869789599046,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048409781455209104,
      "loss": 5.0882,
      "mean_token_accuracy": 0.17024388015270234,
      "num_tokens": 28534028.0,
      "step": 13010
    },
    {
      "entropy": 5.588116025924682,
      "epoch": 1.2916832076220723,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004840847283016809,
      "loss": 5.1797,
      "mean_token_accuracy": 0.15891446620225907,
      "num_tokens": 28545480.0,
      "step": 13015
    },
    {
      "entropy": 5.5265740871429445,
      "epoch": 1.2921794362842398,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048407163686642,
      "loss": 5.1218,
      "mean_token_accuracy": 0.17879820466041565,
      "num_tokens": 28555401.0,
      "step": 13020
    },
    {
      "entropy": 5.5234174728393555,
      "epoch": 1.2926756649464073,
      "grad_norm": 1.078125,
      "learning_rate": 0.000484058540246633,
      "loss": 5.1953,
      "mean_token_accuracy": 0.17124780863523484,
      "num_tokens": 28565497.0,
      "step": 13025
    },
    {
      "entropy": 5.5675750255584715,
      "epoch": 1.2931718936085748,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004840454384426447,
      "loss": 5.2786,
      "mean_token_accuracy": 0.1606290251016617,
      "num_tokens": 28576186.0,
      "step": 13030
    },
    {
      "entropy": 5.598295259475708,
      "epoch": 1.2936681222707422,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004840323314547799,
      "loss": 5.1463,
      "mean_token_accuracy": 0.1722848042845726,
      "num_tokens": 28586202.0,
      "step": 13035
    },
    {
      "entropy": 5.519181776046753,
      "epoch": 1.29416435093291,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048401921928336385,
      "loss": 5.1141,
      "mean_token_accuracy": 0.16592354327440262,
      "num_tokens": 28596486.0,
      "step": 13040
    },
    {
      "entropy": 5.55699315071106,
      "epoch": 1.2946605795950774,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004840061019287215,
      "loss": 5.1574,
      "mean_token_accuracy": 0.171697960793972,
      "num_tokens": 28607934.0,
      "step": 13045
    },
    {
      "entropy": 5.614859914779663,
      "epoch": 1.2951568082572449,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048399297939117817,
      "loss": 5.1934,
      "mean_token_accuracy": 0.16277693808078766,
      "num_tokens": 28617673.0,
      "step": 13050
    },
    {
      "entropy": 5.6102077007293705,
      "epoch": 1.2956530369194126,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004839798516710593,
      "loss": 5.1832,
      "mean_token_accuracy": 0.17069528847932816,
      "num_tokens": 28629811.0,
      "step": 13055
    },
    {
      "entropy": 5.596235704421997,
      "epoch": 1.29614926558158,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004839667187686904,
      "loss": 5.2612,
      "mean_token_accuracy": 0.1647556334733963,
      "num_tokens": 28640371.0,
      "step": 13060
    },
    {
      "entropy": 5.548670816421509,
      "epoch": 1.2966454942437475,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004839535806843972,
      "loss": 5.1893,
      "mean_token_accuracy": 0.16967585012316705,
      "num_tokens": 28651904.0,
      "step": 13065
    },
    {
      "entropy": 5.488644027709961,
      "epoch": 1.297141722905915,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004839404374185054,
      "loss": 5.1144,
      "mean_token_accuracy": 0.1673978701233864,
      "num_tokens": 28662241.0,
      "step": 13070
    },
    {
      "entropy": 5.491305589675903,
      "epoch": 1.2976379515680825,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000483927288971341,
      "loss": 5.0856,
      "mean_token_accuracy": 0.17415677905082702,
      "num_tokens": 28672791.0,
      "step": 13075
    },
    {
      "entropy": 5.525943326950073,
      "epoch": 1.2981341802302502,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004839141353432301,
      "loss": 5.2222,
      "mean_token_accuracy": 0.16960235834121704,
      "num_tokens": 28684218.0,
      "step": 13080
    },
    {
      "entropy": 5.547878217697144,
      "epoch": 1.2986304088924177,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004839009765344987,
      "loss": 5.153,
      "mean_token_accuracy": 0.1657433107495308,
      "num_tokens": 28694690.0,
      "step": 13085
    },
    {
      "entropy": 5.523418760299682,
      "epoch": 1.2991266375545851,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048388781254547324,
      "loss": 5.1449,
      "mean_token_accuracy": 0.17216846644878386,
      "num_tokens": 28705890.0,
      "step": 13090
    },
    {
      "entropy": 5.548996448516846,
      "epoch": 1.2996228662167526,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048387464337648026,
      "loss": 5.1767,
      "mean_token_accuracy": 0.17067211419343947,
      "num_tokens": 28716935.0,
      "step": 13095
    },
    {
      "entropy": 5.563096284866333,
      "epoch": 1.3001190948789203,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048386146902784606,
      "loss": 5.1636,
      "mean_token_accuracy": 0.17251261323690414,
      "num_tokens": 28728053.0,
      "step": 13100
    },
    {
      "entropy": 5.48268313407898,
      "epoch": 1.3006153235410878,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048384828949989757,
      "loss": 5.1171,
      "mean_token_accuracy": 0.17329630851745606,
      "num_tokens": 28738940.0,
      "step": 13105
    },
    {
      "entropy": 5.572353172302246,
      "epoch": 1.3011115522032553,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048383510479296153,
      "loss": 5.1976,
      "mean_token_accuracy": 0.17301127016544343,
      "num_tokens": 28750259.0,
      "step": 13110
    },
    {
      "entropy": 5.55026683807373,
      "epoch": 1.3016077808654227,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004838219149073649,
      "loss": 5.1369,
      "mean_token_accuracy": 0.17226935178041458,
      "num_tokens": 28760119.0,
      "step": 13115
    },
    {
      "entropy": 5.564360618591309,
      "epoch": 1.3021040095275902,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004838087198434347,
      "loss": 5.1804,
      "mean_token_accuracy": 0.16659301668405532,
      "num_tokens": 28772022.0,
      "step": 13120
    },
    {
      "entropy": 5.519193410873413,
      "epoch": 1.302600238189758,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004837955196014982,
      "loss": 5.1281,
      "mean_token_accuracy": 0.16583479195833206,
      "num_tokens": 28782935.0,
      "step": 13125
    },
    {
      "entropy": 5.583038854598999,
      "epoch": 1.3030964668519254,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048378231418188275,
      "loss": 5.2522,
      "mean_token_accuracy": 0.17253718376159669,
      "num_tokens": 28793604.0,
      "step": 13130
    },
    {
      "entropy": 5.563717365264893,
      "epoch": 1.3035926955140928,
      "grad_norm": 1.0,
      "learning_rate": 0.00048376910358491576,
      "loss": 5.2046,
      "mean_token_accuracy": 0.16530329287052153,
      "num_tokens": 28803494.0,
      "step": 13135
    },
    {
      "entropy": 5.524064350128174,
      "epoch": 1.3040889241762605,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004837558878109249,
      "loss": 5.1686,
      "mean_token_accuracy": 0.16294043362140656,
      "num_tokens": 28816232.0,
      "step": 13140
    },
    {
      "entropy": 5.543507099151611,
      "epoch": 1.304585152838428,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004837426668602378,
      "loss": 5.1229,
      "mean_token_accuracy": 0.17397501468658447,
      "num_tokens": 28827056.0,
      "step": 13145
    },
    {
      "entropy": 5.488945960998535,
      "epoch": 1.3050813815005955,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004837294407331825,
      "loss": 5.1199,
      "mean_token_accuracy": 0.17302099019289016,
      "num_tokens": 28837217.0,
      "step": 13150
    },
    {
      "entropy": 5.457980966567993,
      "epoch": 1.305577610162763,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004837162094300867,
      "loss": 5.0278,
      "mean_token_accuracy": 0.17647732198238372,
      "num_tokens": 28846838.0,
      "step": 13155
    },
    {
      "entropy": 5.524992418289185,
      "epoch": 1.3060738388249304,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004837029729512787,
      "loss": 5.076,
      "mean_token_accuracy": 0.1775840237736702,
      "num_tokens": 28856269.0,
      "step": 13160
    },
    {
      "entropy": 5.53138484954834,
      "epoch": 1.306570067487098,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048368973129708667,
      "loss": 5.1132,
      "mean_token_accuracy": 0.17624222338199616,
      "num_tokens": 28867142.0,
      "step": 13165
    },
    {
      "entropy": 5.518300867080688,
      "epoch": 1.3070662961492656,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048367648446783894,
      "loss": 5.0767,
      "mean_token_accuracy": 0.17221258133649825,
      "num_tokens": 28877346.0,
      "step": 13170
    },
    {
      "entropy": 5.5300086498260494,
      "epoch": 1.307562524811433,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004836632324638642,
      "loss": 5.0827,
      "mean_token_accuracy": 0.1707124665379524,
      "num_tokens": 28888425.0,
      "step": 13175
    },
    {
      "entropy": 5.5115738868713375,
      "epoch": 1.3080587534736006,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048364997528549077,
      "loss": 5.1278,
      "mean_token_accuracy": 0.16951045542955398,
      "num_tokens": 28899575.0,
      "step": 13180
    },
    {
      "entropy": 5.497907924652099,
      "epoch": 1.3085549821357683,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048363671293304763,
      "loss": 5.1094,
      "mean_token_accuracy": 0.1746928110718727,
      "num_tokens": 28910769.0,
      "step": 13185
    },
    {
      "entropy": 5.532970237731933,
      "epoch": 1.3090512107979357,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004836234454068636,
      "loss": 5.1157,
      "mean_token_accuracy": 0.1782885417342186,
      "num_tokens": 28921224.0,
      "step": 13190
    },
    {
      "entropy": 5.497860527038574,
      "epoch": 1.3095474394601032,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048361017270726764,
      "loss": 5.1157,
      "mean_token_accuracy": 0.17214779704809188,
      "num_tokens": 28932585.0,
      "step": 13195
    },
    {
      "entropy": 5.498167324066162,
      "epoch": 1.3100436681222707,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004835968948345889,
      "loss": 5.1642,
      "mean_token_accuracy": 0.1656687542796135,
      "num_tokens": 28943471.0,
      "step": 13200
    },
    {
      "entropy": 5.475327730178833,
      "epoch": 1.3105398967844382,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048358361178915666,
      "loss": 5.0728,
      "mean_token_accuracy": 0.18596817404031754,
      "num_tokens": 28956336.0,
      "step": 13205
    },
    {
      "entropy": 5.603514766693115,
      "epoch": 1.3110361254466059,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004835703235713003,
      "loss": 5.2125,
      "mean_token_accuracy": 0.1666790872812271,
      "num_tokens": 28966689.0,
      "step": 13210
    },
    {
      "entropy": 5.589122915267945,
      "epoch": 1.3115323541087733,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048355703018134935,
      "loss": 5.3069,
      "mean_token_accuracy": 0.1742893025279045,
      "num_tokens": 28978548.0,
      "step": 13215
    },
    {
      "entropy": 5.587890911102295,
      "epoch": 1.3120285827709408,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004835437316196335,
      "loss": 5.2115,
      "mean_token_accuracy": 0.16657934039831163,
      "num_tokens": 28988899.0,
      "step": 13220
    },
    {
      "entropy": 5.508882236480713,
      "epoch": 1.3125248114331085,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004835304278864824,
      "loss": 5.134,
      "mean_token_accuracy": 0.17632086873054503,
      "num_tokens": 28999666.0,
      "step": 13225
    },
    {
      "entropy": 5.546355867385865,
      "epoch": 1.313021040095276,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004835171189822261,
      "loss": 5.1365,
      "mean_token_accuracy": 0.16771157681941987,
      "num_tokens": 29010185.0,
      "step": 13230
    },
    {
      "entropy": 5.495452642440796,
      "epoch": 1.3135172687574435,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004835038049071945,
      "loss": 5.1506,
      "mean_token_accuracy": 0.16959853321313859,
      "num_tokens": 29020761.0,
      "step": 13235
    },
    {
      "entropy": 5.624324655532837,
      "epoch": 1.314013497419611,
      "grad_norm": 0.875,
      "learning_rate": 0.0004834904856617178,
      "loss": 5.2992,
      "mean_token_accuracy": 0.16602130234241486,
      "num_tokens": 29031844.0,
      "step": 13240
    },
    {
      "entropy": 5.645944356918335,
      "epoch": 1.3145097260817784,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048347716124612643,
      "loss": 5.2515,
      "mean_token_accuracy": 0.1639154449105263,
      "num_tokens": 29043101.0,
      "step": 13245
    },
    {
      "entropy": 5.585401725769043,
      "epoch": 1.3150059547439459,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004834638316607506,
      "loss": 5.1857,
      "mean_token_accuracy": 0.17143767178058625,
      "num_tokens": 29054016.0,
      "step": 13250
    },
    {
      "entropy": 5.46118278503418,
      "epoch": 1.3155021834061136,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048345049690592105,
      "loss": 5.1558,
      "mean_token_accuracy": 0.16566675156354904,
      "num_tokens": 29065127.0,
      "step": 13255
    },
    {
      "entropy": 5.568687295913696,
      "epoch": 1.315998412068281,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048343715698196814,
      "loss": 5.128,
      "mean_token_accuracy": 0.16824142634868622,
      "num_tokens": 29075130.0,
      "step": 13260
    },
    {
      "entropy": 5.602110385894775,
      "epoch": 1.3164946407304485,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004834238118892231,
      "loss": 5.2484,
      "mean_token_accuracy": 0.16015904396772385,
      "num_tokens": 29086405.0,
      "step": 13265
    },
    {
      "entropy": 5.580699920654297,
      "epoch": 1.3169908693926162,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004834104616280165,
      "loss": 5.1951,
      "mean_token_accuracy": 0.16180069744586945,
      "num_tokens": 29096945.0,
      "step": 13270
    },
    {
      "entropy": 5.598345518112183,
      "epoch": 1.3174870980547837,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004833971061986796,
      "loss": 5.2685,
      "mean_token_accuracy": 0.17143092602491378,
      "num_tokens": 29108275.0,
      "step": 13275
    },
    {
      "entropy": 5.546763277053833,
      "epoch": 1.3179833267169512,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048338374560154344,
      "loss": 5.1045,
      "mean_token_accuracy": 0.16976259797811508,
      "num_tokens": 29118832.0,
      "step": 13280
    },
    {
      "entropy": 5.493811464309692,
      "epoch": 1.3184795553791187,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004833703798369395,
      "loss": 5.1323,
      "mean_token_accuracy": 0.17113442569971085,
      "num_tokens": 29130720.0,
      "step": 13285
    },
    {
      "entropy": 5.5654833793640135,
      "epoch": 1.3189757840412861,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004833570089051992,
      "loss": 5.1177,
      "mean_token_accuracy": 0.16917258203029634,
      "num_tokens": 29140441.0,
      "step": 13290
    },
    {
      "entropy": 5.544336223602295,
      "epoch": 1.3194720127034538,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048334363280665393,
      "loss": 5.1756,
      "mean_token_accuracy": 0.16945491433143617,
      "num_tokens": 29151544.0,
      "step": 13295
    },
    {
      "entropy": 5.593157339096069,
      "epoch": 1.3199682413656213,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004833302515416355,
      "loss": 5.2466,
      "mean_token_accuracy": 0.16966851353645324,
      "num_tokens": 29165475.0,
      "step": 13300
    },
    {
      "entropy": 5.5261537551879885,
      "epoch": 1.3204644700277888,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048331686511047584,
      "loss": 5.0912,
      "mean_token_accuracy": 0.1693265989422798,
      "num_tokens": 29177139.0,
      "step": 13305
    },
    {
      "entropy": 5.587190103530884,
      "epoch": 1.3209606986899565,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048330347351350676,
      "loss": 5.1689,
      "mean_token_accuracy": 0.17337473332881928,
      "num_tokens": 29187812.0,
      "step": 13310
    },
    {
      "entropy": 5.503566312789917,
      "epoch": 1.321456927352124,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048329007675106036,
      "loss": 5.0666,
      "mean_token_accuracy": 0.17029308080673217,
      "num_tokens": 29197932.0,
      "step": 13315
    },
    {
      "entropy": 5.472479248046875,
      "epoch": 1.3219531560142914,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00048327667482346895,
      "loss": 5.0908,
      "mean_token_accuracy": 0.17620163708925246,
      "num_tokens": 29209529.0,
      "step": 13320
    },
    {
      "entropy": 5.516754531860352,
      "epoch": 1.322449384676459,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004832632677310647,
      "loss": 5.1357,
      "mean_token_accuracy": 0.17173322290182114,
      "num_tokens": 29219027.0,
      "step": 13325
    },
    {
      "entropy": 5.556044960021973,
      "epoch": 1.3229456133386264,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048324985547418026,
      "loss": 5.1696,
      "mean_token_accuracy": 0.16752573400735854,
      "num_tokens": 29229647.0,
      "step": 13330
    },
    {
      "entropy": 5.575425386428833,
      "epoch": 1.3234418420007938,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004832364380531482,
      "loss": 5.1823,
      "mean_token_accuracy": 0.17596760541200637,
      "num_tokens": 29240263.0,
      "step": 13335
    },
    {
      "entropy": 5.576332330703735,
      "epoch": 1.3239380706629615,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048322301546830103,
      "loss": 5.2026,
      "mean_token_accuracy": 0.16378069072961807,
      "num_tokens": 29251115.0,
      "step": 13340
    },
    {
      "entropy": 5.514184141159058,
      "epoch": 1.324434299325129,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004832095877199718,
      "loss": 5.1748,
      "mean_token_accuracy": 0.16429441869258882,
      "num_tokens": 29261888.0,
      "step": 13345
    },
    {
      "entropy": 5.46731743812561,
      "epoch": 1.3249305279872965,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048319615480849347,
      "loss": 5.0739,
      "mean_token_accuracy": 0.17850470691919326,
      "num_tokens": 29271948.0,
      "step": 13350
    },
    {
      "entropy": 5.542776489257813,
      "epoch": 1.3254267566494642,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048318271673419914,
      "loss": 5.1444,
      "mean_token_accuracy": 0.16894257217645645,
      "num_tokens": 29282449.0,
      "step": 13355
    },
    {
      "entropy": 5.539471769332886,
      "epoch": 1.3259229853116317,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048316927349742207,
      "loss": 5.1662,
      "mean_token_accuracy": 0.1671816036105156,
      "num_tokens": 29294280.0,
      "step": 13360
    },
    {
      "entropy": 5.571553325653076,
      "epoch": 1.3264192139737991,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004831558250984955,
      "loss": 5.1328,
      "mean_token_accuracy": 0.1717141091823578,
      "num_tokens": 29305143.0,
      "step": 13365
    },
    {
      "entropy": 5.535297060012818,
      "epoch": 1.3269154426359666,
      "grad_norm": 1.125,
      "learning_rate": 0.00048314237153775303,
      "loss": 5.0694,
      "mean_token_accuracy": 0.17472027838230134,
      "num_tokens": 29314697.0,
      "step": 13370
    },
    {
      "entropy": 5.450375318527222,
      "epoch": 1.327411671298134,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004831289128155283,
      "loss": 5.0989,
      "mean_token_accuracy": 0.16731479465961457,
      "num_tokens": 29325257.0,
      "step": 13375
    },
    {
      "entropy": 5.537502670288086,
      "epoch": 1.3279078999603018,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048311544893215496,
      "loss": 5.1876,
      "mean_token_accuracy": 0.1664431095123291,
      "num_tokens": 29335449.0,
      "step": 13380
    },
    {
      "entropy": 5.552964925765991,
      "epoch": 1.3284041286224693,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000483101979887967,
      "loss": 5.1819,
      "mean_token_accuracy": 0.16522295773029327,
      "num_tokens": 29346898.0,
      "step": 13385
    },
    {
      "entropy": 5.596976900100708,
      "epoch": 1.3289003572846367,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048308850568329825,
      "loss": 5.1364,
      "mean_token_accuracy": 0.16628507971763612,
      "num_tokens": 29356118.0,
      "step": 13390
    },
    {
      "entropy": 5.535267114639282,
      "epoch": 1.3293965859468042,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048307502631848306,
      "loss": 5.2179,
      "mean_token_accuracy": 0.1626502111554146,
      "num_tokens": 29367597.0,
      "step": 13395
    },
    {
      "entropy": 5.480478763580322,
      "epoch": 1.329892814608972,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048306154179385546,
      "loss": 5.0871,
      "mean_token_accuracy": 0.16878528892993927,
      "num_tokens": 29377829.0,
      "step": 13400
    },
    {
      "entropy": 5.504371070861817,
      "epoch": 1.3303890432711394,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004830480521097501,
      "loss": 5.0596,
      "mean_token_accuracy": 0.1710459351539612,
      "num_tokens": 29387945.0,
      "step": 13405
    },
    {
      "entropy": 5.455507373809814,
      "epoch": 1.3308852719333069,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004830345572665012,
      "loss": 5.0287,
      "mean_token_accuracy": 0.18020383268594742,
      "num_tokens": 29398269.0,
      "step": 13410
    },
    {
      "entropy": 5.593930053710937,
      "epoch": 1.3313815005954743,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048302105726444366,
      "loss": 5.2105,
      "mean_token_accuracy": 0.16092029213905334,
      "num_tokens": 29410946.0,
      "step": 13415
    },
    {
      "entropy": 5.518677616119385,
      "epoch": 1.3318777292576418,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048300755210391205,
      "loss": 5.0567,
      "mean_token_accuracy": 0.1721974492073059,
      "num_tokens": 29421770.0,
      "step": 13420
    },
    {
      "entropy": 5.52910008430481,
      "epoch": 1.3323739579198095,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004829940417852415,
      "loss": 5.2157,
      "mean_token_accuracy": 0.16812992990016937,
      "num_tokens": 29433717.0,
      "step": 13425
    },
    {
      "entropy": 5.470255279541016,
      "epoch": 1.332870186581977,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004829805263087668,
      "loss": 5.0539,
      "mean_token_accuracy": 0.17629618048667908,
      "num_tokens": 29444382.0,
      "step": 13430
    },
    {
      "entropy": 5.564712810516357,
      "epoch": 1.3333664152441445,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048296700567482324,
      "loss": 5.1767,
      "mean_token_accuracy": 0.16681267470121383,
      "num_tokens": 29454680.0,
      "step": 13435
    },
    {
      "entropy": 5.577697801589966,
      "epoch": 1.3338626439063122,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004829534798837461,
      "loss": 5.1812,
      "mean_token_accuracy": 0.17264432609081268,
      "num_tokens": 29465065.0,
      "step": 13440
    },
    {
      "entropy": 5.579254531860352,
      "epoch": 1.3343588725684796,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004829399489358707,
      "loss": 5.1219,
      "mean_token_accuracy": 0.1692115396261215,
      "num_tokens": 29475704.0,
      "step": 13445
    },
    {
      "entropy": 5.503735017776489,
      "epoch": 1.334855101230647,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004829264128315326,
      "loss": 5.1457,
      "mean_token_accuracy": 0.16468473374843598,
      "num_tokens": 29486418.0,
      "step": 13450
    },
    {
      "entropy": 5.58544545173645,
      "epoch": 1.3353513298928146,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004829128715710676,
      "loss": 5.2362,
      "mean_token_accuracy": 0.1592664211988449,
      "num_tokens": 29498939.0,
      "step": 13455
    },
    {
      "entropy": 5.514646100997925,
      "epoch": 1.335847558554982,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004828993251548114,
      "loss": 5.0616,
      "mean_token_accuracy": 0.1790044367313385,
      "num_tokens": 29508525.0,
      "step": 13460
    },
    {
      "entropy": 5.556839656829834,
      "epoch": 1.3363437872171495,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004828857735830999,
      "loss": 5.1682,
      "mean_token_accuracy": 0.17363756895065308,
      "num_tokens": 29519962.0,
      "step": 13465
    },
    {
      "entropy": 5.493147087097168,
      "epoch": 1.3368400158793172,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004828722168562691,
      "loss": 5.1238,
      "mean_token_accuracy": 0.17261976152658462,
      "num_tokens": 29529586.0,
      "step": 13470
    },
    {
      "entropy": 5.520703887939453,
      "epoch": 1.3373362445414847,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004828586549746553,
      "loss": 5.2173,
      "mean_token_accuracy": 0.16360023617744446,
      "num_tokens": 29540670.0,
      "step": 13475
    },
    {
      "entropy": 5.56369047164917,
      "epoch": 1.3378324732036522,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004828450879385947,
      "loss": 5.1736,
      "mean_token_accuracy": 0.16526678800582886,
      "num_tokens": 29551484.0,
      "step": 13480
    },
    {
      "entropy": 5.548616695404053,
      "epoch": 1.3383287018658199,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004828315157484239,
      "loss": 5.1156,
      "mean_token_accuracy": 0.17327941507101058,
      "num_tokens": 29561596.0,
      "step": 13485
    },
    {
      "entropy": 5.56624665260315,
      "epoch": 1.3388249305279873,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004828179384044793,
      "loss": 5.1339,
      "mean_token_accuracy": 0.1695005089044571,
      "num_tokens": 29572837.0,
      "step": 13490
    },
    {
      "entropy": 5.533002471923828,
      "epoch": 1.3393211591901548,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048280435590709756,
      "loss": 5.0882,
      "mean_token_accuracy": 0.17123755514621736,
      "num_tokens": 29584095.0,
      "step": 13495
    },
    {
      "entropy": 5.530926465988159,
      "epoch": 1.3398173878523223,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004827907682566157,
      "loss": 5.1395,
      "mean_token_accuracy": 0.16533828973770143,
      "num_tokens": 29594903.0,
      "step": 13500
    },
    {
      "entropy": 5.530222988128662,
      "epoch": 1.3403136165144898,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004827771754533704,
      "loss": 5.1182,
      "mean_token_accuracy": 0.17224762886762618,
      "num_tokens": 29605199.0,
      "step": 13505
    },
    {
      "entropy": 5.5968035697937015,
      "epoch": 1.3408098451766575,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048276357749769895,
      "loss": 5.2949,
      "mean_token_accuracy": 0.16060142517089843,
      "num_tokens": 29617535.0,
      "step": 13510
    },
    {
      "entropy": 5.529836797714234,
      "epoch": 1.341306073838825,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048274997438993845,
      "loss": 5.088,
      "mean_token_accuracy": 0.171018186211586,
      "num_tokens": 29628227.0,
      "step": 13515
    },
    {
      "entropy": 5.57895073890686,
      "epoch": 1.3418023025009924,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048273636613042626,
      "loss": 5.2037,
      "mean_token_accuracy": 0.1692424848675728,
      "num_tokens": 29639112.0,
      "step": 13520
    },
    {
      "entropy": 5.455804634094238,
      "epoch": 1.3422985311631601,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004827227527194998,
      "loss": 5.0881,
      "mean_token_accuracy": 0.16489313095808028,
      "num_tokens": 29650768.0,
      "step": 13525
    },
    {
      "entropy": 5.565947532653809,
      "epoch": 1.3427947598253276,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004827091341574967,
      "loss": 5.2026,
      "mean_token_accuracy": 0.17328959852457046,
      "num_tokens": 29662388.0,
      "step": 13530
    },
    {
      "entropy": 5.551366996765137,
      "epoch": 1.343290988487495,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004826955104447546,
      "loss": 5.0939,
      "mean_token_accuracy": 0.17867274582386017,
      "num_tokens": 29672231.0,
      "step": 13535
    },
    {
      "entropy": 5.536306858062744,
      "epoch": 1.3437872171496625,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004826818815816114,
      "loss": 5.1445,
      "mean_token_accuracy": 0.17367480844259262,
      "num_tokens": 29683019.0,
      "step": 13540
    },
    {
      "entropy": 5.534232902526855,
      "epoch": 1.34428344581183,
      "grad_norm": 1.0,
      "learning_rate": 0.00048266824756840513,
      "loss": 5.1594,
      "mean_token_accuracy": 0.17224482148885728,
      "num_tokens": 29694233.0,
      "step": 13545
    },
    {
      "entropy": 5.562705039978027,
      "epoch": 1.3447796744739975,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004826546084054737,
      "loss": 5.1977,
      "mean_token_accuracy": 0.16854349821805953,
      "num_tokens": 29704023.0,
      "step": 13550
    },
    {
      "entropy": 5.641705417633057,
      "epoch": 1.3452759031361652,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004826409640931555,
      "loss": 5.2865,
      "mean_token_accuracy": 0.16259726881980896,
      "num_tokens": 29715407.0,
      "step": 13555
    },
    {
      "entropy": 5.48348593711853,
      "epoch": 1.3457721317983327,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048262731463178884,
      "loss": 5.1015,
      "mean_token_accuracy": 0.17188044041395187,
      "num_tokens": 29726223.0,
      "step": 13560
    },
    {
      "entropy": 5.495322227478027,
      "epoch": 1.3462683604605001,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004826136600217121,
      "loss": 5.0278,
      "mean_token_accuracy": 0.17780915945768355,
      "num_tokens": 29735850.0,
      "step": 13565
    },
    {
      "entropy": 5.557669258117675,
      "epoch": 1.3467645891226678,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000482600000263264,
      "loss": 5.1982,
      "mean_token_accuracy": 0.1690421849489212,
      "num_tokens": 29747381.0,
      "step": 13570
    },
    {
      "entropy": 5.56163182258606,
      "epoch": 1.3472608177848353,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004825863353567832,
      "loss": 5.2256,
      "mean_token_accuracy": 0.16919496208429335,
      "num_tokens": 29758440.0,
      "step": 13575
    },
    {
      "entropy": 5.4747802257537845,
      "epoch": 1.3477570464470028,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048257266530260865,
      "loss": 5.0804,
      "mean_token_accuracy": 0.17054817378520964,
      "num_tokens": 29770606.0,
      "step": 13580
    },
    {
      "entropy": 5.508941602706909,
      "epoch": 1.3482532751091703,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004825589901010792,
      "loss": 5.1555,
      "mean_token_accuracy": 0.17294624596834182,
      "num_tokens": 29781469.0,
      "step": 13585
    },
    {
      "entropy": 5.538801288604736,
      "epoch": 1.3487495037713377,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004825453097525341,
      "loss": 5.2181,
      "mean_token_accuracy": 0.16858044266700745,
      "num_tokens": 29792847.0,
      "step": 13590
    },
    {
      "entropy": 5.5267258644104,
      "epoch": 1.3492457324335054,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048253162425731253,
      "loss": 5.131,
      "mean_token_accuracy": 0.17018972039222718,
      "num_tokens": 29802784.0,
      "step": 13595
    },
    {
      "entropy": 5.582158184051513,
      "epoch": 1.349741961095673,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004825179336157539,
      "loss": 5.1481,
      "mean_token_accuracy": 0.17191785126924514,
      "num_tokens": 29812849.0,
      "step": 13600
    },
    {
      "entropy": 5.526828813552856,
      "epoch": 1.3502381897578404,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048250423782819765,
      "loss": 5.1204,
      "mean_token_accuracy": 0.16780119240283967,
      "num_tokens": 29823495.0,
      "step": 13605
    },
    {
      "entropy": 5.529862880706787,
      "epoch": 1.350734418420008,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004824905368949834,
      "loss": 5.1778,
      "mean_token_accuracy": 0.1669362425804138,
      "num_tokens": 29835295.0,
      "step": 13610
    },
    {
      "entropy": 5.510252904891968,
      "epoch": 1.3512306470821756,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004824768308164509,
      "loss": 5.1085,
      "mean_token_accuracy": 0.176770481467247,
      "num_tokens": 29845747.0,
      "step": 13615
    },
    {
      "entropy": 5.538020896911621,
      "epoch": 1.351726875744343,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004824631195929402,
      "loss": 5.171,
      "mean_token_accuracy": 0.17418198883533478,
      "num_tokens": 29857111.0,
      "step": 13620
    },
    {
      "entropy": 5.494154548645019,
      "epoch": 1.3522231044065105,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004824494032247911,
      "loss": 5.1597,
      "mean_token_accuracy": 0.164189974963665,
      "num_tokens": 29866919.0,
      "step": 13625
    },
    {
      "entropy": 5.511776733398437,
      "epoch": 1.352719333068678,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004824356817123439,
      "loss": 5.1129,
      "mean_token_accuracy": 0.16926506757736207,
      "num_tokens": 29878631.0,
      "step": 13630
    },
    {
      "entropy": 5.524524545669555,
      "epoch": 1.3532155617308455,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048242195505593865,
      "loss": 5.1422,
      "mean_token_accuracy": 0.17134637683629989,
      "num_tokens": 29890809.0,
      "step": 13635
    },
    {
      "entropy": 5.562575626373291,
      "epoch": 1.3537117903930131,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048240822325591596,
      "loss": 5.2187,
      "mean_token_accuracy": 0.17203592360019684,
      "num_tokens": 29901789.0,
      "step": 13640
    },
    {
      "entropy": 5.498739385604859,
      "epoch": 1.3542080190551806,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004823944863126162,
      "loss": 5.1246,
      "mean_token_accuracy": 0.16743547171354295,
      "num_tokens": 29912576.0,
      "step": 13645
    },
    {
      "entropy": 5.580034875869751,
      "epoch": 1.354704247717348,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004823807442263801,
      "loss": 5.2168,
      "mean_token_accuracy": 0.1606443256139755,
      "num_tokens": 29924076.0,
      "step": 13650
    },
    {
      "entropy": 5.504463481903076,
      "epoch": 1.3552004763795158,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048236699699754845,
      "loss": 5.1359,
      "mean_token_accuracy": 0.1691555291414261,
      "num_tokens": 29935854.0,
      "step": 13655
    },
    {
      "entropy": 5.536391353607177,
      "epoch": 1.3556967050416833,
      "grad_norm": 0.875,
      "learning_rate": 0.000482353244626462,
      "loss": 5.093,
      "mean_token_accuracy": 0.1683717966079712,
      "num_tokens": 29946389.0,
      "step": 13660
    },
    {
      "entropy": 5.559522819519043,
      "epoch": 1.3561929337038507,
      "grad_norm": 1.0,
      "learning_rate": 0.00048233948711346196,
      "loss": 5.1595,
      "mean_token_accuracy": 0.16595248281955718,
      "num_tokens": 29958170.0,
      "step": 13665
    },
    {
      "entropy": 5.485475683212281,
      "epoch": 1.3566891623660182,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048232572445888945,
      "loss": 5.0464,
      "mean_token_accuracy": 0.1755625918507576,
      "num_tokens": 29969067.0,
      "step": 13670
    },
    {
      "entropy": 5.565349864959717,
      "epoch": 1.3571853910281857,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004823119566630857,
      "loss": 5.1858,
      "mean_token_accuracy": 0.16782769113779067,
      "num_tokens": 29979727.0,
      "step": 13675
    },
    {
      "entropy": 5.573714685440064,
      "epoch": 1.3576816196903534,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048229818372639216,
      "loss": 5.1516,
      "mean_token_accuracy": 0.17022989243268966,
      "num_tokens": 29989397.0,
      "step": 13680
    },
    {
      "entropy": 5.467497873306274,
      "epoch": 1.3581778483525209,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048228440564915026,
      "loss": 5.0803,
      "mean_token_accuracy": 0.1684701457619667,
      "num_tokens": 29999415.0,
      "step": 13685
    },
    {
      "entropy": 5.51286563873291,
      "epoch": 1.3586740770146883,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048227062243170186,
      "loss": 5.1199,
      "mean_token_accuracy": 0.17517461329698564,
      "num_tokens": 30010277.0,
      "step": 13690
    },
    {
      "entropy": 5.466383600234986,
      "epoch": 1.3591703056768558,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004822568340743886,
      "loss": 5.0119,
      "mean_token_accuracy": 0.18072010576725006,
      "num_tokens": 30020684.0,
      "step": 13695
    },
    {
      "entropy": 5.600068140029907,
      "epoch": 1.3596665343390235,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004822430405775525,
      "loss": 5.2303,
      "mean_token_accuracy": 0.16519418954849244,
      "num_tokens": 30030916.0,
      "step": 13700
    },
    {
      "entropy": 5.595063591003418,
      "epoch": 1.360162763001191,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004822292419415355,
      "loss": 5.1497,
      "mean_token_accuracy": 0.16804620772600173,
      "num_tokens": 30041140.0,
      "step": 13705
    },
    {
      "entropy": 5.51181640625,
      "epoch": 1.3606589916633585,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048221543816667984,
      "loss": 5.1126,
      "mean_token_accuracy": 0.18026674091815947,
      "num_tokens": 30052765.0,
      "step": 13710
    },
    {
      "entropy": 5.4777802467346195,
      "epoch": 1.361155220325526,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048220162925332786,
      "loss": 5.1272,
      "mean_token_accuracy": 0.1705651581287384,
      "num_tokens": 30062511.0,
      "step": 13715
    },
    {
      "entropy": 5.622525358200074,
      "epoch": 1.3616514489876934,
      "grad_norm": 1.296875,
      "learning_rate": 0.000482187815201822,
      "loss": 5.2131,
      "mean_token_accuracy": 0.16916419714689254,
      "num_tokens": 30073404.0,
      "step": 13720
    },
    {
      "entropy": 5.506596279144287,
      "epoch": 1.362147677649861,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048217399601250464,
      "loss": 5.0694,
      "mean_token_accuracy": 0.17430136799812318,
      "num_tokens": 30084978.0,
      "step": 13725
    },
    {
      "entropy": 5.538654232025147,
      "epoch": 1.3626439063120286,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004821601716857187,
      "loss": 5.1431,
      "mean_token_accuracy": 0.17120698988437652,
      "num_tokens": 30094498.0,
      "step": 13730
    },
    {
      "entropy": 5.496025657653808,
      "epoch": 1.363140134974196,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048214634222180686,
      "loss": 5.1591,
      "mean_token_accuracy": 0.17425190806388854,
      "num_tokens": 30104545.0,
      "step": 13735
    },
    {
      "entropy": 5.588847208023071,
      "epoch": 1.3636363636363638,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004821325076211121,
      "loss": 5.2407,
      "mean_token_accuracy": 0.17144907712936402,
      "num_tokens": 30114863.0,
      "step": 13740
    },
    {
      "entropy": 5.591871690750122,
      "epoch": 1.3641325922985312,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048211866788397753,
      "loss": 5.1412,
      "mean_token_accuracy": 0.1668848067522049,
      "num_tokens": 30125101.0,
      "step": 13745
    },
    {
      "entropy": 5.5326512336730955,
      "epoch": 1.3646288209606987,
      "grad_norm": 1.125,
      "learning_rate": 0.0004821048230107462,
      "loss": 5.2082,
      "mean_token_accuracy": 0.1738513931632042,
      "num_tokens": 30135081.0,
      "step": 13750
    },
    {
      "entropy": 5.6686928272247314,
      "epoch": 1.3651250496228662,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004820909730017616,
      "loss": 5.2494,
      "mean_token_accuracy": 0.1575665593147278,
      "num_tokens": 30147169.0,
      "step": 13755
    },
    {
      "entropy": 5.563359212875366,
      "epoch": 1.3656212782850337,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048207711785736707,
      "loss": 5.0489,
      "mean_token_accuracy": 0.17081768214702606,
      "num_tokens": 30157523.0,
      "step": 13760
    },
    {
      "entropy": 5.461745977401733,
      "epoch": 1.3661175069472011,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00048206325757790624,
      "loss": 5.1337,
      "mean_token_accuracy": 0.17506172955036164,
      "num_tokens": 30168313.0,
      "step": 13765
    },
    {
      "entropy": 5.461941576004028,
      "epoch": 1.3666137356093688,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048204939216372285,
      "loss": 5.0874,
      "mean_token_accuracy": 0.1732782557606697,
      "num_tokens": 30179173.0,
      "step": 13770
    },
    {
      "entropy": 5.602393531799317,
      "epoch": 1.3671099642715363,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004820355216151607,
      "loss": 5.1646,
      "mean_token_accuracy": 0.16877421587705613,
      "num_tokens": 30191666.0,
      "step": 13775
    },
    {
      "entropy": 5.600204801559448,
      "epoch": 1.3676061929337038,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048202164593256363,
      "loss": 5.1257,
      "mean_token_accuracy": 0.17001240849494934,
      "num_tokens": 30202752.0,
      "step": 13780
    },
    {
      "entropy": 5.517510795593262,
      "epoch": 1.3681024215958715,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048200776511627585,
      "loss": 5.1529,
      "mean_token_accuracy": 0.1731338307261467,
      "num_tokens": 30213127.0,
      "step": 13785
    },
    {
      "entropy": 5.57043423652649,
      "epoch": 1.368598650258039,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004819938791666416,
      "loss": 5.191,
      "mean_token_accuracy": 0.17228126674890518,
      "num_tokens": 30224782.0,
      "step": 13790
    },
    {
      "entropy": 5.538125705718994,
      "epoch": 1.3690948789202064,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048197998808400524,
      "loss": 5.1864,
      "mean_token_accuracy": 0.1719948098063469,
      "num_tokens": 30235129.0,
      "step": 13795
    },
    {
      "entropy": 5.557019567489624,
      "epoch": 1.369591107582374,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004819660918687111,
      "loss": 5.2072,
      "mean_token_accuracy": 0.16350836157798768,
      "num_tokens": 30247304.0,
      "step": 13800
    },
    {
      "entropy": 5.519007587432862,
      "epoch": 1.3700873362445414,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004819521905211039,
      "loss": 5.0757,
      "mean_token_accuracy": 0.17202291637659073,
      "num_tokens": 30257900.0,
      "step": 13805
    },
    {
      "entropy": 5.616189479827881,
      "epoch": 1.370583564906709,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048193828404152826,
      "loss": 5.2009,
      "mean_token_accuracy": 0.16968987584114076,
      "num_tokens": 30269741.0,
      "step": 13810
    },
    {
      "entropy": 5.531930732727051,
      "epoch": 1.3710797935688765,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004819243724303291,
      "loss": 5.2029,
      "mean_token_accuracy": 0.16319651156663895,
      "num_tokens": 30279382.0,
      "step": 13815
    },
    {
      "entropy": 5.553898525238037,
      "epoch": 1.371576022231044,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048191045568785144,
      "loss": 5.175,
      "mean_token_accuracy": 0.1693979024887085,
      "num_tokens": 30289837.0,
      "step": 13820
    },
    {
      "entropy": 5.514196348190308,
      "epoch": 1.3720722508932117,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048189653381444024,
      "loss": 5.1263,
      "mean_token_accuracy": 0.17463000416755675,
      "num_tokens": 30302138.0,
      "step": 13825
    },
    {
      "entropy": 5.561945962905884,
      "epoch": 1.3725684795553792,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004818826068104409,
      "loss": 5.2059,
      "mean_token_accuracy": 0.1698312595486641,
      "num_tokens": 30313623.0,
      "step": 13830
    },
    {
      "entropy": 5.574228954315186,
      "epoch": 1.3730647082175467,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004818686746761987,
      "loss": 5.1538,
      "mean_token_accuracy": 0.1668202981352806,
      "num_tokens": 30324533.0,
      "step": 13835
    },
    {
      "entropy": 5.5970619201660154,
      "epoch": 1.3735609368797141,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004818547374120591,
      "loss": 5.2531,
      "mean_token_accuracy": 0.15952227115631104,
      "num_tokens": 30335819.0,
      "step": 13840
    },
    {
      "entropy": 5.530732536315918,
      "epoch": 1.3740571655418816,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004818407950183678,
      "loss": 5.1269,
      "mean_token_accuracy": 0.17419376373291015,
      "num_tokens": 30345947.0,
      "step": 13845
    },
    {
      "entropy": 5.502305793762207,
      "epoch": 1.374553394204049,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004818268474954704,
      "loss": 5.1146,
      "mean_token_accuracy": 0.17125113159418107,
      "num_tokens": 30356972.0,
      "step": 13850
    },
    {
      "entropy": 5.584651374816895,
      "epoch": 1.3750496228662168,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004818128948437129,
      "loss": 5.1962,
      "mean_token_accuracy": 0.16647759228944778,
      "num_tokens": 30368414.0,
      "step": 13855
    },
    {
      "entropy": 5.595959424972534,
      "epoch": 1.3755458515283843,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004817989370634412,
      "loss": 5.1248,
      "mean_token_accuracy": 0.17206144481897354,
      "num_tokens": 30379141.0,
      "step": 13860
    },
    {
      "entropy": 5.536151313781739,
      "epoch": 1.3760420801905517,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048178497415500153,
      "loss": 5.0588,
      "mean_token_accuracy": 0.18021752834320068,
      "num_tokens": 30388985.0,
      "step": 13865
    },
    {
      "entropy": 5.529819536209106,
      "epoch": 1.3765383088527194,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004817710061187401,
      "loss": 5.1155,
      "mean_token_accuracy": 0.18031678348779678,
      "num_tokens": 30399599.0,
      "step": 13870
    },
    {
      "entropy": 5.560130500793457,
      "epoch": 1.377034537514887,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048175703295500323,
      "loss": 5.2095,
      "mean_token_accuracy": 0.1690106675028801,
      "num_tokens": 30411143.0,
      "step": 13875
    },
    {
      "entropy": 5.634877109527588,
      "epoch": 1.3775307661770544,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048174305466413755,
      "loss": 5.1973,
      "mean_token_accuracy": 0.17103828340768815,
      "num_tokens": 30422339.0,
      "step": 13880
    },
    {
      "entropy": 5.531447649002075,
      "epoch": 1.3780269948392219,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048172907124648947,
      "loss": 5.1205,
      "mean_token_accuracy": 0.17195411920547485,
      "num_tokens": 30432588.0,
      "step": 13885
    },
    {
      "entropy": 5.566704368591308,
      "epoch": 1.3785232235013893,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004817150827024059,
      "loss": 5.1805,
      "mean_token_accuracy": 0.17412401586771012,
      "num_tokens": 30442272.0,
      "step": 13890
    },
    {
      "entropy": 5.504783725738525,
      "epoch": 1.379019452163557,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004817010890322338,
      "loss": 5.0915,
      "mean_token_accuracy": 0.17417070567607879,
      "num_tokens": 30453142.0,
      "step": 13895
    },
    {
      "entropy": 5.577582311630249,
      "epoch": 1.3795156808257245,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048168709023632006,
      "loss": 5.2018,
      "mean_token_accuracy": 0.16694744378328324,
      "num_tokens": 30465136.0,
      "step": 13900
    },
    {
      "entropy": 5.635681009292602,
      "epoch": 1.380011909487892,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048167308631501184,
      "loss": 5.1228,
      "mean_token_accuracy": 0.17227154970169067,
      "num_tokens": 30476511.0,
      "step": 13905
    },
    {
      "entropy": 5.577196025848389,
      "epoch": 1.3805081381500597,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048165907726865643,
      "loss": 5.2315,
      "mean_token_accuracy": 0.1627224341034889,
      "num_tokens": 30486547.0,
      "step": 13910
    },
    {
      "entropy": 5.573952627182007,
      "epoch": 1.3810043668122272,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004816450630976012,
      "loss": 5.1948,
      "mean_token_accuracy": 0.17420513927936554,
      "num_tokens": 30496163.0,
      "step": 13915
    },
    {
      "entropy": 5.555355358123779,
      "epoch": 1.3815005954743946,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004816310438021937,
      "loss": 5.1876,
      "mean_token_accuracy": 0.17177563160657883,
      "num_tokens": 30506979.0,
      "step": 13920
    },
    {
      "entropy": 5.58404278755188,
      "epoch": 1.381996824136562,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048161701938278156,
      "loss": 5.2208,
      "mean_token_accuracy": 0.1606489047408104,
      "num_tokens": 30517492.0,
      "step": 13925
    },
    {
      "entropy": 5.500133323669433,
      "epoch": 1.3824930527987296,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048160298983971256,
      "loss": 5.0776,
      "mean_token_accuracy": 0.1745465651154518,
      "num_tokens": 30528614.0,
      "step": 13930
    },
    {
      "entropy": 5.528719902038574,
      "epoch": 1.382989281460897,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004815889551733346,
      "loss": 5.1855,
      "mean_token_accuracy": 0.16944511532783507,
      "num_tokens": 30539479.0,
      "step": 13935
    },
    {
      "entropy": 5.614548969268799,
      "epoch": 1.3834855101230648,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048157491538399573,
      "loss": 5.2142,
      "mean_token_accuracy": 0.16844443678855897,
      "num_tokens": 30549909.0,
      "step": 13940
    },
    {
      "entropy": 5.598271369934082,
      "epoch": 1.3839817387852322,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048156087047204407,
      "loss": 5.0996,
      "mean_token_accuracy": 0.1766225203871727,
      "num_tokens": 30562265.0,
      "step": 13945
    },
    {
      "entropy": 5.582998132705688,
      "epoch": 1.3844779674473997,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004815468204378279,
      "loss": 5.1352,
      "mean_token_accuracy": 0.16981200724840165,
      "num_tokens": 30572622.0,
      "step": 13950
    },
    {
      "entropy": 5.509753274917602,
      "epoch": 1.3849741961095674,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004815327652816957,
      "loss": 5.1119,
      "mean_token_accuracy": 0.17214690297842025,
      "num_tokens": 30583300.0,
      "step": 13955
    },
    {
      "entropy": 5.514261627197266,
      "epoch": 1.3854704247717349,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004815187050039959,
      "loss": 5.0686,
      "mean_token_accuracy": 0.17511713355779648,
      "num_tokens": 30594460.0,
      "step": 13960
    },
    {
      "entropy": 5.537472295761108,
      "epoch": 1.3859666534339024,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048150463960507724,
      "loss": 5.1024,
      "mean_token_accuracy": 0.1711593374609947,
      "num_tokens": 30604658.0,
      "step": 13965
    },
    {
      "entropy": 5.547265911102295,
      "epoch": 1.3864628820960698,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048149056908528855,
      "loss": 5.1047,
      "mean_token_accuracy": 0.16818841993808747,
      "num_tokens": 30616151.0,
      "step": 13970
    },
    {
      "entropy": 5.531367063522339,
      "epoch": 1.3869591107582373,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004814764934449786,
      "loss": 5.1633,
      "mean_token_accuracy": 0.1750291481614113,
      "num_tokens": 30627560.0,
      "step": 13975
    },
    {
      "entropy": 5.536222124099732,
      "epoch": 1.387455339420405,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048146241268449665,
      "loss": 5.1423,
      "mean_token_accuracy": 0.17360917925834657,
      "num_tokens": 30638490.0,
      "step": 13980
    },
    {
      "entropy": 5.598623752593994,
      "epoch": 1.3879515680825725,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004814483268041917,
      "loss": 5.2527,
      "mean_token_accuracy": 0.1691885322332382,
      "num_tokens": 30649501.0,
      "step": 13985
    },
    {
      "entropy": 5.567948818206787,
      "epoch": 1.38844779674474,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004814342358044131,
      "loss": 5.1279,
      "mean_token_accuracy": 0.17392905205488204,
      "num_tokens": 30660479.0,
      "step": 13990
    },
    {
      "entropy": 5.64867262840271,
      "epoch": 1.3889440254069074,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048142013968551026,
      "loss": 5.3088,
      "mean_token_accuracy": 0.15852625221014022,
      "num_tokens": 30672015.0,
      "step": 13995
    },
    {
      "entropy": 5.467982387542724,
      "epoch": 1.3894402540690751,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048140603844783273,
      "loss": 5.1537,
      "mean_token_accuracy": 0.16933163851499558,
      "num_tokens": 30682612.0,
      "step": 14000
    },
    {
      "entropy": 5.495663833618164,
      "epoch": 1.3899364827312426,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004813919320917303,
      "loss": 5.0957,
      "mean_token_accuracy": 0.1712799847126007,
      "num_tokens": 30693289.0,
      "step": 14005
    },
    {
      "entropy": 5.572775840759277,
      "epoch": 1.39043271139341,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048137782061755265,
      "loss": 5.1474,
      "mean_token_accuracy": 0.16950712203979493,
      "num_tokens": 30705346.0,
      "step": 14010
    },
    {
      "entropy": 5.509809923171997,
      "epoch": 1.3909289400555775,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048136370402564974,
      "loss": 5.1506,
      "mean_token_accuracy": 0.1659905955195427,
      "num_tokens": 30715968.0,
      "step": 14015
    },
    {
      "entropy": 5.519743728637695,
      "epoch": 1.391425168717745,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004813495823163716,
      "loss": 5.0999,
      "mean_token_accuracy": 0.17194507420063018,
      "num_tokens": 30726225.0,
      "step": 14020
    },
    {
      "entropy": 5.507151126861572,
      "epoch": 1.3919213973799127,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004813354554900685,
      "loss": 5.0673,
      "mean_token_accuracy": 0.1738160088658333,
      "num_tokens": 30735603.0,
      "step": 14025
    },
    {
      "entropy": 5.487907886505127,
      "epoch": 1.3924176260420802,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004813213235470907,
      "loss": 5.1307,
      "mean_token_accuracy": 0.16985554099082947,
      "num_tokens": 30747414.0,
      "step": 14030
    },
    {
      "entropy": 5.582440805435181,
      "epoch": 1.3929138547042477,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004813071864877887,
      "loss": 5.1873,
      "mean_token_accuracy": 0.17258260846138002,
      "num_tokens": 30758389.0,
      "step": 14035
    },
    {
      "entropy": 5.588983154296875,
      "epoch": 1.3934100833664154,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000481293044312513,
      "loss": 5.203,
      "mean_token_accuracy": 0.16748234033584594,
      "num_tokens": 30769484.0,
      "step": 14040
    },
    {
      "entropy": 5.490986490249634,
      "epoch": 1.3939063120285828,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048127889702161435,
      "loss": 5.1042,
      "mean_token_accuracy": 0.17436276972293854,
      "num_tokens": 30779775.0,
      "step": 14045
    },
    {
      "entropy": 5.558767890930175,
      "epoch": 1.3944025406907503,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004812647446154435,
      "loss": 5.2127,
      "mean_token_accuracy": 0.1640891522169113,
      "num_tokens": 30791456.0,
      "step": 14050
    },
    {
      "entropy": 5.537540006637573,
      "epoch": 1.3948987693529178,
      "grad_norm": 0.875,
      "learning_rate": 0.0004812505870943515,
      "loss": 5.0344,
      "mean_token_accuracy": 0.17639026194810867,
      "num_tokens": 30801659.0,
      "step": 14055
    },
    {
      "entropy": 5.624769496917724,
      "epoch": 1.3953949980150853,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004812364244586893,
      "loss": 5.2659,
      "mean_token_accuracy": 0.16538924127817153,
      "num_tokens": 30813389.0,
      "step": 14060
    },
    {
      "entropy": 5.477582693099976,
      "epoch": 1.395891226677253,
      "grad_norm": 1.0,
      "learning_rate": 0.0004812222567088082,
      "loss": 5.0251,
      "mean_token_accuracy": 0.1695793554186821,
      "num_tokens": 30823725.0,
      "step": 14065
    },
    {
      "entropy": 5.483001232147217,
      "epoch": 1.3963874553394204,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004812080838450596,
      "loss": 5.0484,
      "mean_token_accuracy": 0.17452310621738434,
      "num_tokens": 30834448.0,
      "step": 14070
    },
    {
      "entropy": 5.534590435028076,
      "epoch": 1.396883684001588,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004811939058677947,
      "loss": 5.0952,
      "mean_token_accuracy": 0.17181549817323685,
      "num_tokens": 30845891.0,
      "step": 14075
    },
    {
      "entropy": 5.519650173187256,
      "epoch": 1.3973799126637554,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004811797227773654,
      "loss": 5.0451,
      "mean_token_accuracy": 0.1788552522659302,
      "num_tokens": 30855170.0,
      "step": 14080
    },
    {
      "entropy": 5.499523878097534,
      "epoch": 1.397876141325923,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048116553457412317,
      "loss": 5.1354,
      "mean_token_accuracy": 0.17040596008300782,
      "num_tokens": 30866527.0,
      "step": 14085
    },
    {
      "entropy": 5.500823593139648,
      "epoch": 1.3983723699880906,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048115134125841993,
      "loss": 5.0865,
      "mean_token_accuracy": 0.17504869252443314,
      "num_tokens": 30876292.0,
      "step": 14090
    },
    {
      "entropy": 5.5390700340271,
      "epoch": 1.398868598650258,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004811371428306076,
      "loss": 5.0694,
      "mean_token_accuracy": 0.1723286345601082,
      "num_tokens": 30887440.0,
      "step": 14095
    },
    {
      "entropy": 5.523568105697632,
      "epoch": 1.3993648273124255,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004811229392910384,
      "loss": 5.1191,
      "mean_token_accuracy": 0.17340546995401382,
      "num_tokens": 30897925.0,
      "step": 14100
    },
    {
      "entropy": 5.484356164932251,
      "epoch": 1.399861055974593,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004811087306400644,
      "loss": 5.1529,
      "mean_token_accuracy": 0.17690719962120055,
      "num_tokens": 30908413.0,
      "step": 14105
    },
    {
      "entropy": 5.580122327804565,
      "epoch": 1.4003572846367607,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048109451687803804,
      "loss": 5.1758,
      "mean_token_accuracy": 0.16955263614654542,
      "num_tokens": 30919149.0,
      "step": 14110
    },
    {
      "entropy": 5.6071288108825685,
      "epoch": 1.4008535132989282,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004810802980053117,
      "loss": 5.2003,
      "mean_token_accuracy": 0.16742929220199584,
      "num_tokens": 30929883.0,
      "step": 14115
    },
    {
      "entropy": 5.519658041000366,
      "epoch": 1.4013497419610956,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048106607402223815,
      "loss": 5.1158,
      "mean_token_accuracy": 0.16678639501333237,
      "num_tokens": 30941358.0,
      "step": 14120
    },
    {
      "entropy": 5.548413944244385,
      "epoch": 1.4018459706232633,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048105184492916995,
      "loss": 5.1051,
      "mean_token_accuracy": 0.17210416346788407,
      "num_tokens": 30953289.0,
      "step": 14125
    },
    {
      "entropy": 5.518923664093018,
      "epoch": 1.4023421992854308,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048103761072645993,
      "loss": 5.0465,
      "mean_token_accuracy": 0.17729029059410095,
      "num_tokens": 30964304.0,
      "step": 14130
    },
    {
      "entropy": 5.471157598495483,
      "epoch": 1.4028384279475983,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048102337141446116,
      "loss": 5.0713,
      "mean_token_accuracy": 0.1787961408495903,
      "num_tokens": 30974840.0,
      "step": 14135
    },
    {
      "entropy": 5.501088714599609,
      "epoch": 1.4033346566097658,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048100912699352666,
      "loss": 5.131,
      "mean_token_accuracy": 0.17864849269390107,
      "num_tokens": 30985805.0,
      "step": 14140
    },
    {
      "entropy": 5.557447481155395,
      "epoch": 1.4038308852719332,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004809948774640098,
      "loss": 5.1365,
      "mean_token_accuracy": 0.1727934792637825,
      "num_tokens": 30998174.0,
      "step": 14145
    },
    {
      "entropy": 5.559416151046753,
      "epoch": 1.4043271139341007,
      "grad_norm": 0.875,
      "learning_rate": 0.0004809806228262639,
      "loss": 5.1466,
      "mean_token_accuracy": 0.17206069678068162,
      "num_tokens": 31010050.0,
      "step": 14150
    },
    {
      "entropy": 5.612583446502685,
      "epoch": 1.4048233425962684,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048096636308064236,
      "loss": 5.2717,
      "mean_token_accuracy": 0.16032503694295883,
      "num_tokens": 31023069.0,
      "step": 14155
    },
    {
      "entropy": 5.556841230392456,
      "epoch": 1.4053195712584359,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004809520982274989,
      "loss": 5.0798,
      "mean_token_accuracy": 0.17819133996963502,
      "num_tokens": 31033512.0,
      "step": 14160
    },
    {
      "entropy": 5.515301084518432,
      "epoch": 1.4058157999206033,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048093782826718706,
      "loss": 5.1136,
      "mean_token_accuracy": 0.16862879693508148,
      "num_tokens": 31045731.0,
      "step": 14165
    },
    {
      "entropy": 5.493713808059693,
      "epoch": 1.406312028582771,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048092355320006086,
      "loss": 5.0293,
      "mean_token_accuracy": 0.1763712137937546,
      "num_tokens": 31055443.0,
      "step": 14170
    },
    {
      "entropy": 5.48050217628479,
      "epoch": 1.4068082572449385,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004809092730264744,
      "loss": 5.1242,
      "mean_token_accuracy": 0.17357977628707885,
      "num_tokens": 31067575.0,
      "step": 14175
    },
    {
      "entropy": 5.529447078704834,
      "epoch": 1.407304485907106,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048089498774678155,
      "loss": 5.2136,
      "mean_token_accuracy": 0.16673895865678787,
      "num_tokens": 31078198.0,
      "step": 14180
    },
    {
      "entropy": 5.519463539123535,
      "epoch": 1.4078007145692735,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048088069736133664,
      "loss": 5.0996,
      "mean_token_accuracy": 0.17052169889211655,
      "num_tokens": 31089961.0,
      "step": 14185
    },
    {
      "entropy": 5.537969446182251,
      "epoch": 1.408296943231441,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004808664018704941,
      "loss": 5.1126,
      "mean_token_accuracy": 0.1709460660815239,
      "num_tokens": 31100534.0,
      "step": 14190
    },
    {
      "entropy": 5.641386461257935,
      "epoch": 1.4087931718936086,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004808521012746085,
      "loss": 5.1461,
      "mean_token_accuracy": 0.17591289877891542,
      "num_tokens": 31112016.0,
      "step": 14195
    },
    {
      "entropy": 5.498534440994263,
      "epoch": 1.4092894005557761,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048083779557403424,
      "loss": 5.1653,
      "mean_token_accuracy": 0.17351087778806687,
      "num_tokens": 31123659.0,
      "step": 14200
    },
    {
      "entropy": 5.54283504486084,
      "epoch": 1.4097856292179436,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004808234847691262,
      "loss": 5.1138,
      "mean_token_accuracy": 0.16809820830821992,
      "num_tokens": 31133872.0,
      "step": 14205
    },
    {
      "entropy": 5.57604603767395,
      "epoch": 1.4102818578801113,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004808091688602393,
      "loss": 5.1701,
      "mean_token_accuracy": 0.16956085115671157,
      "num_tokens": 31145834.0,
      "step": 14210
    },
    {
      "entropy": 5.564661598205566,
      "epoch": 1.4107780865422788,
      "grad_norm": 0.875,
      "learning_rate": 0.00048079484784772854,
      "loss": 5.2306,
      "mean_token_accuracy": 0.16539136320352554,
      "num_tokens": 31157037.0,
      "step": 14215
    },
    {
      "entropy": 5.532263612747192,
      "epoch": 1.4112743152044462,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048078052173194886,
      "loss": 5.1133,
      "mean_token_accuracy": 0.17727053314447402,
      "num_tokens": 31166812.0,
      "step": 14220
    },
    {
      "entropy": 5.548606300354004,
      "epoch": 1.4117705438666137,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004807661905132557,
      "loss": 5.1133,
      "mean_token_accuracy": 0.17000300586223602,
      "num_tokens": 31177162.0,
      "step": 14225
    },
    {
      "entropy": 5.551578426361084,
      "epoch": 1.4122667725287812,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004807518541920045,
      "loss": 5.1521,
      "mean_token_accuracy": 0.17382025569677353,
      "num_tokens": 31187333.0,
      "step": 14230
    },
    {
      "entropy": 5.581780099868775,
      "epoch": 1.4127630011909487,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048073751276855056,
      "loss": 5.1608,
      "mean_token_accuracy": 0.16805825978517533,
      "num_tokens": 31197869.0,
      "step": 14235
    },
    {
      "entropy": 5.555130338668823,
      "epoch": 1.4132592298531164,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004807231662432497,
      "loss": 5.164,
      "mean_token_accuracy": 0.1665928155183792,
      "num_tokens": 31208199.0,
      "step": 14240
    },
    {
      "entropy": 5.557612657546997,
      "epoch": 1.4137554585152838,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004807088146164575,
      "loss": 5.1467,
      "mean_token_accuracy": 0.1648218020796776,
      "num_tokens": 31218903.0,
      "step": 14245
    },
    {
      "entropy": 5.517922306060791,
      "epoch": 1.4142516871774513,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004806944578885301,
      "loss": 5.1555,
      "mean_token_accuracy": 0.1690744251012802,
      "num_tokens": 31231146.0,
      "step": 14250
    },
    {
      "entropy": 5.541270542144775,
      "epoch": 1.414747915839619,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048068009605982333,
      "loss": 5.0739,
      "mean_token_accuracy": 0.1735615149140358,
      "num_tokens": 31240980.0,
      "step": 14255
    },
    {
      "entropy": 5.5217297077178955,
      "epoch": 1.4152441445017865,
      "grad_norm": 1.0,
      "learning_rate": 0.00048066572913069334,
      "loss": 5.1142,
      "mean_token_accuracy": 0.169989150762558,
      "num_tokens": 31252458.0,
      "step": 14260
    },
    {
      "entropy": 5.514530181884766,
      "epoch": 1.415740373163954,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004806513571014965,
      "loss": 5.1039,
      "mean_token_accuracy": 0.17186874896287918,
      "num_tokens": 31262979.0,
      "step": 14265
    },
    {
      "entropy": 5.5504576683044435,
      "epoch": 1.4162366018261214,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004806369799725891,
      "loss": 5.0961,
      "mean_token_accuracy": 0.16948116570711136,
      "num_tokens": 31274811.0,
      "step": 14270
    },
    {
      "entropy": 5.507457304000854,
      "epoch": 1.416732830488289,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004806225977443277,
      "loss": 5.0625,
      "mean_token_accuracy": 0.17748325318098068,
      "num_tokens": 31285573.0,
      "step": 14275
    },
    {
      "entropy": 5.597534418106079,
      "epoch": 1.4172290591504566,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004806082104170689,
      "loss": 5.2375,
      "mean_token_accuracy": 0.16136072874069213,
      "num_tokens": 31296993.0,
      "step": 14280
    },
    {
      "entropy": 5.50703616142273,
      "epoch": 1.417725287812624,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048059381799116967,
      "loss": 5.1192,
      "mean_token_accuracy": 0.1678570181131363,
      "num_tokens": 31307217.0,
      "step": 14285
    },
    {
      "entropy": 5.491989517211914,
      "epoch": 1.4182215164747916,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004805794204669867,
      "loss": 5.1443,
      "mean_token_accuracy": 0.17037644237279892,
      "num_tokens": 31317939.0,
      "step": 14290
    },
    {
      "entropy": 5.602749443054199,
      "epoch": 1.418717745136959,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048056501784487707,
      "loss": 5.1205,
      "mean_token_accuracy": 0.16901961714029312,
      "num_tokens": 31327731.0,
      "step": 14295
    },
    {
      "entropy": 5.719781351089478,
      "epoch": 1.4192139737991267,
      "grad_norm": 1.046875,
      "learning_rate": 0.000480550610125198,
      "loss": 5.2978,
      "mean_token_accuracy": 0.15915951281785964,
      "num_tokens": 31339388.0,
      "step": 14300
    },
    {
      "entropy": 5.490464496612549,
      "epoch": 1.4197102024612942,
      "grad_norm": 1.0,
      "learning_rate": 0.0004805361973083066,
      "loss": 4.9961,
      "mean_token_accuracy": 0.17786178588867188,
      "num_tokens": 31350186.0,
      "step": 14305
    },
    {
      "entropy": 5.531382703781128,
      "epoch": 1.4202064311234617,
      "grad_norm": 1.0,
      "learning_rate": 0.0004805217793945605,
      "loss": 5.143,
      "mean_token_accuracy": 0.1653599351644516,
      "num_tokens": 31361219.0,
      "step": 14310
    },
    {
      "entropy": 5.550638246536255,
      "epoch": 1.4207026597856292,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004805073563843171,
      "loss": 5.1803,
      "mean_token_accuracy": 0.16556326001882554,
      "num_tokens": 31372820.0,
      "step": 14315
    },
    {
      "entropy": 5.536877870559692,
      "epoch": 1.4211988884477966,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048049292827793413,
      "loss": 5.0608,
      "mean_token_accuracy": 0.17393716126680375,
      "num_tokens": 31382779.0,
      "step": 14320
    },
    {
      "entropy": 5.6161552429199215,
      "epoch": 1.4216951171099643,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004804784950757693,
      "loss": 5.2638,
      "mean_token_accuracy": 0.1676454499363899,
      "num_tokens": 31394872.0,
      "step": 14325
    },
    {
      "entropy": 5.56783390045166,
      "epoch": 1.4221913457721318,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048046405677818057,
      "loss": 5.1713,
      "mean_token_accuracy": 0.17211804091930388,
      "num_tokens": 31405767.0,
      "step": 14330
    },
    {
      "entropy": 5.586855125427246,
      "epoch": 1.4226875744342993,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048044961338552595,
      "loss": 5.1709,
      "mean_token_accuracy": 0.1698562055826187,
      "num_tokens": 31417255.0,
      "step": 14335
    },
    {
      "entropy": 5.561547899246216,
      "epoch": 1.423183803096467,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004804351648981636,
      "loss": 5.1226,
      "mean_token_accuracy": 0.17501015365123748,
      "num_tokens": 31428494.0,
      "step": 14340
    },
    {
      "entropy": 5.516804885864258,
      "epoch": 1.4236800317586344,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004804207113164519,
      "loss": 5.0327,
      "mean_token_accuracy": 0.17773139029741286,
      "num_tokens": 31439221.0,
      "step": 14345
    },
    {
      "entropy": 5.537303018569946,
      "epoch": 1.424176260420802,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048040625264074915,
      "loss": 5.1175,
      "mean_token_accuracy": 0.17708641439676284,
      "num_tokens": 31450575.0,
      "step": 14350
    },
    {
      "entropy": 5.452661561965942,
      "epoch": 1.4246724890829694,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048039178887141404,
      "loss": 5.1208,
      "mean_token_accuracy": 0.17472280263900758,
      "num_tokens": 31462168.0,
      "step": 14355
    },
    {
      "entropy": 5.541571760177613,
      "epoch": 1.4251687177451369,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048037732000880503,
      "loss": 5.1417,
      "mean_token_accuracy": 0.17525738179683686,
      "num_tokens": 31472619.0,
      "step": 14360
    },
    {
      "entropy": 5.586092805862426,
      "epoch": 1.4256649464073046,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048036284605328106,
      "loss": 5.2185,
      "mean_token_accuracy": 0.1665370911359787,
      "num_tokens": 31485246.0,
      "step": 14365
    },
    {
      "entropy": 5.548651313781738,
      "epoch": 1.426161175069472,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004803483670052011,
      "loss": 5.1325,
      "mean_token_accuracy": 0.16484898328781128,
      "num_tokens": 31496258.0,
      "step": 14370
    },
    {
      "entropy": 5.525687313079834,
      "epoch": 1.4266574037316395,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000480333882864924,
      "loss": 5.1004,
      "mean_token_accuracy": 0.1732739731669426,
      "num_tokens": 31506671.0,
      "step": 14375
    },
    {
      "entropy": 5.556588077545166,
      "epoch": 1.427153632393807,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004803193936328092,
      "loss": 5.1417,
      "mean_token_accuracy": 0.166272109746933,
      "num_tokens": 31517001.0,
      "step": 14380
    },
    {
      "entropy": 5.544458532333374,
      "epoch": 1.4276498610559747,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004803048993092158,
      "loss": 5.1428,
      "mean_token_accuracy": 0.17396878749132155,
      "num_tokens": 31529649.0,
      "step": 14385
    },
    {
      "entropy": 5.462076568603516,
      "epoch": 1.4281460897181422,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048029039989450317,
      "loss": 5.0617,
      "mean_token_accuracy": 0.17557339519262313,
      "num_tokens": 31540083.0,
      "step": 14390
    },
    {
      "entropy": 5.574583101272583,
      "epoch": 1.4286423183803096,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048027589538903115,
      "loss": 5.1891,
      "mean_token_accuracy": 0.17000586837530135,
      "num_tokens": 31551400.0,
      "step": 14395
    },
    {
      "entropy": 5.558568286895752,
      "epoch": 1.4291385470424771,
      "grad_norm": 1.0,
      "learning_rate": 0.0004802613857931592,
      "loss": 5.1761,
      "mean_token_accuracy": 0.16776325404644013,
      "num_tokens": 31562430.0,
      "step": 14400
    },
    {
      "entropy": 5.561457729339599,
      "epoch": 1.4296347757046446,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004802468711072472,
      "loss": 5.2149,
      "mean_token_accuracy": 0.1718543663620949,
      "num_tokens": 31573173.0,
      "step": 14405
    },
    {
      "entropy": 5.550090551376343,
      "epoch": 1.4301310043668123,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000480232351331655,
      "loss": 5.1495,
      "mean_token_accuracy": 0.17230795919895173,
      "num_tokens": 31584072.0,
      "step": 14410
    },
    {
      "entropy": 5.483279132843018,
      "epoch": 1.4306272330289798,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048021782646674277,
      "loss": 5.0859,
      "mean_token_accuracy": 0.17675964236259462,
      "num_tokens": 31594993.0,
      "step": 14415
    },
    {
      "entropy": 5.484372758865357,
      "epoch": 1.4311234616911472,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048020329651287064,
      "loss": 5.109,
      "mean_token_accuracy": 0.17678418904542922,
      "num_tokens": 31604284.0,
      "step": 14420
    },
    {
      "entropy": 5.505160856246948,
      "epoch": 1.431619690353315,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000480188761470399,
      "loss": 5.193,
      "mean_token_accuracy": 0.16898860186338424,
      "num_tokens": 31616188.0,
      "step": 14425
    },
    {
      "entropy": 5.615882444381714,
      "epoch": 1.4321159190154824,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004801742213396881,
      "loss": 5.1558,
      "mean_token_accuracy": 0.17215863317251207,
      "num_tokens": 31626606.0,
      "step": 14430
    },
    {
      "entropy": 5.499743223190308,
      "epoch": 1.4326121476776499,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048015967612109875,
      "loss": 5.0689,
      "mean_token_accuracy": 0.18403681367635727,
      "num_tokens": 31638050.0,
      "step": 14435
    },
    {
      "entropy": 5.521497535705566,
      "epoch": 1.4331083763398174,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048014512581499136,
      "loss": 5.0896,
      "mean_token_accuracy": 0.17138998955488205,
      "num_tokens": 31648451.0,
      "step": 14440
    },
    {
      "entropy": 5.519690799713135,
      "epoch": 1.4336046050019848,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000480130570421727,
      "loss": 5.1064,
      "mean_token_accuracy": 0.16932046562433242,
      "num_tokens": 31659310.0,
      "step": 14445
    },
    {
      "entropy": 5.504786491394043,
      "epoch": 1.4341008336641523,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004801160099416665,
      "loss": 5.1378,
      "mean_token_accuracy": 0.16748925894498826,
      "num_tokens": 31669594.0,
      "step": 14450
    },
    {
      "entropy": 5.528512287139892,
      "epoch": 1.43459706232632,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004801014443751709,
      "loss": 5.14,
      "mean_token_accuracy": 0.1666400671005249,
      "num_tokens": 31682034.0,
      "step": 14455
    },
    {
      "entropy": 5.589808320999145,
      "epoch": 1.4350932909884875,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004800868737226015,
      "loss": 5.1655,
      "mean_token_accuracy": 0.17436257749795914,
      "num_tokens": 31692390.0,
      "step": 14460
    },
    {
      "entropy": 5.557921600341797,
      "epoch": 1.435589519650655,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048007229798431946,
      "loss": 5.1457,
      "mean_token_accuracy": 0.17328951209783555,
      "num_tokens": 31703690.0,
      "step": 14465
    },
    {
      "entropy": 5.545484733581543,
      "epoch": 1.4360857483128227,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004800577171606864,
      "loss": 5.1409,
      "mean_token_accuracy": 0.17311709821224214,
      "num_tokens": 31714655.0,
      "step": 14470
    },
    {
      "entropy": 5.488937044143677,
      "epoch": 1.4365819769749901,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004800431312520638,
      "loss": 5.0594,
      "mean_token_accuracy": 0.1795497789978981,
      "num_tokens": 31727022.0,
      "step": 14475
    },
    {
      "entropy": 5.548110818862915,
      "epoch": 1.4370782056371576,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004800285402588134,
      "loss": 5.1085,
      "mean_token_accuracy": 0.1772718459367752,
      "num_tokens": 31738071.0,
      "step": 14480
    },
    {
      "entropy": 5.59354567527771,
      "epoch": 1.437574434299325,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004800139441812969,
      "loss": 5.2143,
      "mean_token_accuracy": 0.16117098033428193,
      "num_tokens": 31749342.0,
      "step": 14485
    },
    {
      "entropy": 5.556843280792236,
      "epoch": 1.4380706629614926,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047999934301987644,
      "loss": 5.1927,
      "mean_token_accuracy": 0.16907583326101303,
      "num_tokens": 31761218.0,
      "step": 14490
    },
    {
      "entropy": 5.564563035964966,
      "epoch": 1.4385668916236602,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000479984736774914,
      "loss": 5.2028,
      "mean_token_accuracy": 0.16832948178052903,
      "num_tokens": 31773111.0,
      "step": 14495
    },
    {
      "entropy": 5.603694868087769,
      "epoch": 1.4390631202858277,
      "grad_norm": 1.0,
      "learning_rate": 0.00047997012544677176,
      "loss": 5.153,
      "mean_token_accuracy": 0.17313717901706696,
      "num_tokens": 31782948.0,
      "step": 14500
    },
    {
      "entropy": 5.590463399887085,
      "epoch": 1.4395593489479952,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004799555090358121,
      "loss": 5.1388,
      "mean_token_accuracy": 0.17268241345882415,
      "num_tokens": 31793992.0,
      "step": 14505
    },
    {
      "entropy": 5.566668081283569,
      "epoch": 1.440055577610163,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004799408875423975,
      "loss": 5.153,
      "mean_token_accuracy": 0.16334629505872728,
      "num_tokens": 31804830.0,
      "step": 14510
    },
    {
      "entropy": 5.515792655944824,
      "epoch": 1.4405518062723304,
      "grad_norm": 0.875,
      "learning_rate": 0.0004799262609668904,
      "loss": 5.1384,
      "mean_token_accuracy": 0.1739792615175247,
      "num_tokens": 31817200.0,
      "step": 14515
    },
    {
      "entropy": 5.52047028541565,
      "epoch": 1.4410480349344978,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004799116293096537,
      "loss": 5.1811,
      "mean_token_accuracy": 0.17165776044130326,
      "num_tokens": 31828226.0,
      "step": 14520
    },
    {
      "entropy": 5.5173523902893065,
      "epoch": 1.4415442635966653,
      "grad_norm": 1.0,
      "learning_rate": 0.0004798969925710501,
      "loss": 5.0669,
      "mean_token_accuracy": 0.18043017238378525,
      "num_tokens": 31839635.0,
      "step": 14525
    },
    {
      "entropy": 5.54227991104126,
      "epoch": 1.4420404922588328,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047988235075144254,
      "loss": 5.1097,
      "mean_token_accuracy": 0.17895270884037018,
      "num_tokens": 31850333.0,
      "step": 14530
    },
    {
      "entropy": 5.489082527160645,
      "epoch": 1.4425367209210003,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004798677038511942,
      "loss": 5.092,
      "mean_token_accuracy": 0.16857707649469375,
      "num_tokens": 31861002.0,
      "step": 14535
    },
    {
      "entropy": 5.548662710189819,
      "epoch": 1.443032949583168,
      "grad_norm": 1.0,
      "learning_rate": 0.0004798530518706683,
      "loss": 5.1678,
      "mean_token_accuracy": 0.16767743676900865,
      "num_tokens": 31871713.0,
      "step": 14540
    },
    {
      "entropy": 5.54739818572998,
      "epoch": 1.4435291782453354,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004798383948102281,
      "loss": 5.0995,
      "mean_token_accuracy": 0.16896482259035112,
      "num_tokens": 31882381.0,
      "step": 14545
    },
    {
      "entropy": 5.569896125793457,
      "epoch": 1.444025406907503,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004798237326702371,
      "loss": 5.1257,
      "mean_token_accuracy": 0.17677302062511444,
      "num_tokens": 31892347.0,
      "step": 14550
    },
    {
      "entropy": 5.626822662353516,
      "epoch": 1.4445216355696706,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047980906545105884,
      "loss": 5.2405,
      "mean_token_accuracy": 0.16343855410814284,
      "num_tokens": 31902751.0,
      "step": 14555
    },
    {
      "entropy": 5.539401912689209,
      "epoch": 1.445017864231838,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004797943931530571,
      "loss": 5.1066,
      "mean_token_accuracy": 0.1770730197429657,
      "num_tokens": 31913294.0,
      "step": 14560
    },
    {
      "entropy": 5.541674518585205,
      "epoch": 1.4455140928940056,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004797797157765958,
      "loss": 5.2252,
      "mean_token_accuracy": 0.1736141175031662,
      "num_tokens": 31925678.0,
      "step": 14565
    },
    {
      "entropy": 5.477121734619141,
      "epoch": 1.446010321556173,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004797650333220387,
      "loss": 5.0455,
      "mean_token_accuracy": 0.17725862711668014,
      "num_tokens": 31934578.0,
      "step": 14570
    },
    {
      "entropy": 5.483195114135742,
      "epoch": 1.4465065502183405,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047975034578975006,
      "loss": 5.0982,
      "mean_token_accuracy": 0.17120981961488724,
      "num_tokens": 31945482.0,
      "step": 14575
    },
    {
      "entropy": 5.53853964805603,
      "epoch": 1.4470027788805082,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000479735653180094,
      "loss": 5.2027,
      "mean_token_accuracy": 0.17311585694551468,
      "num_tokens": 31957303.0,
      "step": 14580
    },
    {
      "entropy": 5.544617652893066,
      "epoch": 1.4474990075426757,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004797209554934349,
      "loss": 5.1182,
      "mean_token_accuracy": 0.16760422587394713,
      "num_tokens": 31968601.0,
      "step": 14585
    },
    {
      "entropy": 5.556090831756592,
      "epoch": 1.4479952362048432,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004797062527301373,
      "loss": 5.1688,
      "mean_token_accuracy": 0.16858005821704863,
      "num_tokens": 31978987.0,
      "step": 14590
    },
    {
      "entropy": 5.530074882507324,
      "epoch": 1.4484914648670106,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004796915448905657,
      "loss": 5.1344,
      "mean_token_accuracy": 0.16919378638267518,
      "num_tokens": 31990950.0,
      "step": 14595
    },
    {
      "entropy": 5.601096820831299,
      "epoch": 1.4489876935291783,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047967683197508486,
      "loss": 5.1385,
      "mean_token_accuracy": 0.16988271921873094,
      "num_tokens": 32001795.0,
      "step": 14600
    },
    {
      "entropy": 5.574069786071777,
      "epoch": 1.4494839221913458,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047966211398405966,
      "loss": 5.1625,
      "mean_token_accuracy": 0.16445329189300537,
      "num_tokens": 32011775.0,
      "step": 14605
    },
    {
      "entropy": 5.645728683471679,
      "epoch": 1.4499801508535133,
      "grad_norm": 1.015625,
      "learning_rate": 0.000479647390917855,
      "loss": 5.2472,
      "mean_token_accuracy": 0.15949815660715103,
      "num_tokens": 32022674.0,
      "step": 14610
    },
    {
      "entropy": 5.586968660354614,
      "epoch": 1.4504763795156808,
      "grad_norm": 1.015625,
      "learning_rate": 0.000479632662776836,
      "loss": 5.202,
      "mean_token_accuracy": 0.17336668223142623,
      "num_tokens": 32033383.0,
      "step": 14615
    },
    {
      "entropy": 5.559762620925904,
      "epoch": 1.4509726081778482,
      "grad_norm": 0.953125,
      "learning_rate": 0.000479617929561368,
      "loss": 5.0982,
      "mean_token_accuracy": 0.17610179781913757,
      "num_tokens": 32044066.0,
      "step": 14620
    },
    {
      "entropy": 5.544317197799683,
      "epoch": 1.451468836840016,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004796031912718162,
      "loss": 5.1845,
      "mean_token_accuracy": 0.1644149601459503,
      "num_tokens": 32054650.0,
      "step": 14625
    },
    {
      "entropy": 5.616777038574218,
      "epoch": 1.4519650655021834,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047958844790854606,
      "loss": 5.1798,
      "mean_token_accuracy": 0.16292149126529692,
      "num_tokens": 32065786.0,
      "step": 14630
    },
    {
      "entropy": 5.607652378082276,
      "epoch": 1.4524612941643509,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047957369947192337,
      "loss": 5.1222,
      "mean_token_accuracy": 0.17663830071687697,
      "num_tokens": 32076326.0,
      "step": 14635
    },
    {
      "entropy": 5.505214834213257,
      "epoch": 1.4529575228265186,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004795589459623136,
      "loss": 5.0797,
      "mean_token_accuracy": 0.17234665751457215,
      "num_tokens": 32086915.0,
      "step": 14640
    },
    {
      "entropy": 5.524808263778686,
      "epoch": 1.453453751488686,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047954418738008295,
      "loss": 5.0431,
      "mean_token_accuracy": 0.17361759543418884,
      "num_tokens": 32096614.0,
      "step": 14645
    },
    {
      "entropy": 5.542260694503784,
      "epoch": 1.4539499801508535,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004795294237255971,
      "loss": 5.132,
      "mean_token_accuracy": 0.1735408589243889,
      "num_tokens": 32107721.0,
      "step": 14650
    },
    {
      "entropy": 5.5165914535522464,
      "epoch": 1.454446208813021,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047951465499922234,
      "loss": 5.1471,
      "mean_token_accuracy": 0.16828598380088805,
      "num_tokens": 32116621.0,
      "step": 14655
    },
    {
      "entropy": 5.576553440093994,
      "epoch": 1.4549424374751885,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004794998812013248,
      "loss": 5.1793,
      "mean_token_accuracy": 0.16863167881965638,
      "num_tokens": 32126536.0,
      "step": 14660
    },
    {
      "entropy": 5.586959505081177,
      "epoch": 1.4554386661373562,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047948510233227084,
      "loss": 5.2409,
      "mean_token_accuracy": 0.16116394251585006,
      "num_tokens": 32137059.0,
      "step": 14665
    },
    {
      "entropy": 5.518677139282227,
      "epoch": 1.4559348947995236,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004794703183924269,
      "loss": 5.1785,
      "mean_token_accuracy": 0.17294724136590958,
      "num_tokens": 32148186.0,
      "step": 14670
    },
    {
      "entropy": 5.614913368225098,
      "epoch": 1.4564311234616911,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004794555293821597,
      "loss": 5.2262,
      "mean_token_accuracy": 0.15936787873506547,
      "num_tokens": 32158244.0,
      "step": 14675
    },
    {
      "entropy": 5.5571283340454105,
      "epoch": 1.4569273521238586,
      "grad_norm": 0.953125,
      "learning_rate": 0.000479440735301836,
      "loss": 5.0787,
      "mean_token_accuracy": 0.17744283825159074,
      "num_tokens": 32168404.0,
      "step": 14680
    },
    {
      "entropy": 5.509984922409058,
      "epoch": 1.4574235807860263,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004794259361518226,
      "loss": 5.0635,
      "mean_token_accuracy": 0.17613533586263658,
      "num_tokens": 32178744.0,
      "step": 14685
    },
    {
      "entropy": 5.513152837753296,
      "epoch": 1.4579198094481938,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047941113193248643,
      "loss": 5.1795,
      "mean_token_accuracy": 0.1680013969540596,
      "num_tokens": 32189948.0,
      "step": 14690
    },
    {
      "entropy": 5.577406024932861,
      "epoch": 1.4584160381103612,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004793963226441947,
      "loss": 5.1663,
      "mean_token_accuracy": 0.16641853302717208,
      "num_tokens": 32200979.0,
      "step": 14695
    },
    {
      "entropy": 5.536780452728271,
      "epoch": 1.4589122667725287,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047938150828731455,
      "loss": 5.1467,
      "mean_token_accuracy": 0.17314921766519548,
      "num_tokens": 32211569.0,
      "step": 14700
    },
    {
      "entropy": 5.561519908905029,
      "epoch": 1.4594084954346962,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004793666888622134,
      "loss": 5.1347,
      "mean_token_accuracy": 0.1694324180483818,
      "num_tokens": 32222429.0,
      "step": 14705
    },
    {
      "entropy": 5.571921062469483,
      "epoch": 1.459904724096864,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004793518643692587,
      "loss": 5.1893,
      "mean_token_accuracy": 0.16576670706272126,
      "num_tokens": 32234332.0,
      "step": 14710
    },
    {
      "entropy": 5.620045900344849,
      "epoch": 1.4604009527590314,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004793370348088182,
      "loss": 5.2234,
      "mean_token_accuracy": 0.17236478626728058,
      "num_tokens": 32244798.0,
      "step": 14715
    },
    {
      "entropy": 5.590567874908447,
      "epoch": 1.4608971814211988,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004793222001812595,
      "loss": 5.137,
      "mean_token_accuracy": 0.1752204939723015,
      "num_tokens": 32257746.0,
      "step": 14720
    },
    {
      "entropy": 5.577586555480957,
      "epoch": 1.4613934100833665,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047930736048695045,
      "loss": 5.2356,
      "mean_token_accuracy": 0.16509189307689667,
      "num_tokens": 32270692.0,
      "step": 14725
    },
    {
      "entropy": 5.531466436386109,
      "epoch": 1.461889638745534,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004792925157262592,
      "loss": 5.1171,
      "mean_token_accuracy": 0.17350534796714784,
      "num_tokens": 32281350.0,
      "step": 14730
    },
    {
      "entropy": 5.565432596206665,
      "epoch": 1.4623858674077015,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004792776658995537,
      "loss": 5.153,
      "mean_token_accuracy": 0.1720444455742836,
      "num_tokens": 32293182.0,
      "step": 14735
    },
    {
      "entropy": 5.546879720687866,
      "epoch": 1.462882096069869,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004792628110072023,
      "loss": 5.1268,
      "mean_token_accuracy": 0.16702627539634704,
      "num_tokens": 32304903.0,
      "step": 14740
    },
    {
      "entropy": 5.620081615447998,
      "epoch": 1.4633783247320364,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004792479510495734,
      "loss": 5.3009,
      "mean_token_accuracy": 0.1631435364484787,
      "num_tokens": 32316074.0,
      "step": 14745
    },
    {
      "entropy": 5.567771053314209,
      "epoch": 1.463874553394204,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047923308602703535,
      "loss": 5.1125,
      "mean_token_accuracy": 0.18134698122739792,
      "num_tokens": 32327594.0,
      "step": 14750
    },
    {
      "entropy": 5.53625898361206,
      "epoch": 1.4643707820563716,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047921821593995684,
      "loss": 5.1584,
      "mean_token_accuracy": 0.17243526726961136,
      "num_tokens": 32339224.0,
      "step": 14755
    },
    {
      "entropy": 5.581178379058838,
      "epoch": 1.464867010718539,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004792033407887066,
      "loss": 5.1224,
      "mean_token_accuracy": 0.16923604309558868,
      "num_tokens": 32349451.0,
      "step": 14760
    },
    {
      "entropy": 5.681001234054565,
      "epoch": 1.4653632393807066,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004791884605736536,
      "loss": 5.2149,
      "mean_token_accuracy": 0.16261758506298066,
      "num_tokens": 32361177.0,
      "step": 14765
    },
    {
      "entropy": 5.58430027961731,
      "epoch": 1.4658594680428743,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047917357529516676,
      "loss": 5.1255,
      "mean_token_accuracy": 0.16506303399801253,
      "num_tokens": 32372031.0,
      "step": 14770
    },
    {
      "entropy": 5.512924289703369,
      "epoch": 1.4663556967050417,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004791586849536152,
      "loss": 5.0396,
      "mean_token_accuracy": 0.1780196025967598,
      "num_tokens": 32383936.0,
      "step": 14775
    },
    {
      "entropy": 5.5187695026397705,
      "epoch": 1.4668519253672092,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004791437895493682,
      "loss": 5.1891,
      "mean_token_accuracy": 0.17052170336246492,
      "num_tokens": 32394283.0,
      "step": 14780
    },
    {
      "entropy": 5.529397821426391,
      "epoch": 1.4673481540293767,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047912888908279507,
      "loss": 5.1767,
      "mean_token_accuracy": 0.16659256368875502,
      "num_tokens": 32405549.0,
      "step": 14785
    },
    {
      "entropy": 5.513508605957031,
      "epoch": 1.4678443826915442,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004791139835542653,
      "loss": 5.1124,
      "mean_token_accuracy": 0.17101800739765166,
      "num_tokens": 32414896.0,
      "step": 14790
    },
    {
      "entropy": 5.560007190704345,
      "epoch": 1.4683406113537119,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047909907296414864,
      "loss": 5.1734,
      "mean_token_accuracy": 0.17036933153867723,
      "num_tokens": 32426863.0,
      "step": 14795
    },
    {
      "entropy": 5.5639026165008545,
      "epoch": 1.4688368400158793,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047908415731281475,
      "loss": 5.1499,
      "mean_token_accuracy": 0.1683117762207985,
      "num_tokens": 32437703.0,
      "step": 14800
    },
    {
      "entropy": 5.570926094055176,
      "epoch": 1.4693330686780468,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047906923660063344,
      "loss": 5.1815,
      "mean_token_accuracy": 0.17697264552116393,
      "num_tokens": 32448517.0,
      "step": 14805
    },
    {
      "entropy": 5.527656412124633,
      "epoch": 1.4698292973402145,
      "grad_norm": 1.046875,
      "learning_rate": 0.000479054310827975,
      "loss": 5.1725,
      "mean_token_accuracy": 0.1672141045331955,
      "num_tokens": 32458782.0,
      "step": 14810
    },
    {
      "entropy": 5.514373922348023,
      "epoch": 1.470325526002382,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047903937999520913,
      "loss": 5.0472,
      "mean_token_accuracy": 0.17484400570392608,
      "num_tokens": 32470113.0,
      "step": 14815
    },
    {
      "entropy": 5.54478759765625,
      "epoch": 1.4708217546645495,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004790244441027064,
      "loss": 5.1238,
      "mean_token_accuracy": 0.1723346710205078,
      "num_tokens": 32482455.0,
      "step": 14820
    },
    {
      "entropy": 5.60060396194458,
      "epoch": 1.471317983326717,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004790095031508371,
      "loss": 5.1715,
      "mean_token_accuracy": 0.1745301142334938,
      "num_tokens": 32493536.0,
      "step": 14825
    },
    {
      "entropy": 5.522411918640136,
      "epoch": 1.4718142119888844,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047899455713997164,
      "loss": 5.0806,
      "mean_token_accuracy": 0.17717416882514953,
      "num_tokens": 32503991.0,
      "step": 14830
    },
    {
      "entropy": 5.515948867797851,
      "epoch": 1.4723104406510519,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047897960607048084,
      "loss": 5.1634,
      "mean_token_accuracy": 0.17042644321918488,
      "num_tokens": 32515474.0,
      "step": 14835
    },
    {
      "entropy": 5.509422159194946,
      "epoch": 1.4728066693132196,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004789646499427353,
      "loss": 5.0692,
      "mean_token_accuracy": 0.1759020745754242,
      "num_tokens": 32526751.0,
      "step": 14840
    },
    {
      "entropy": 5.621194648742676,
      "epoch": 1.473302897975387,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004789496887571059,
      "loss": 5.2313,
      "mean_token_accuracy": 0.1665905609726906,
      "num_tokens": 32538191.0,
      "step": 14845
    },
    {
      "entropy": 5.552267169952392,
      "epoch": 1.4737991266375545,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047893472251396373,
      "loss": 5.1644,
      "mean_token_accuracy": 0.17348639965057372,
      "num_tokens": 32548841.0,
      "step": 14850
    },
    {
      "entropy": 5.5161089420318605,
      "epoch": 1.4742953552997222,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047891975121368,
      "loss": 5.1003,
      "mean_token_accuracy": 0.17140769064426423,
      "num_tokens": 32558599.0,
      "step": 14855
    },
    {
      "entropy": 5.4889500617980955,
      "epoch": 1.4747915839618897,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047890477485662563,
      "loss": 5.0768,
      "mean_token_accuracy": 0.1801529198884964,
      "num_tokens": 32568641.0,
      "step": 14860
    },
    {
      "entropy": 5.551357698440552,
      "epoch": 1.4752878126240572,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047888979344317236,
      "loss": 5.1708,
      "mean_token_accuracy": 0.1650012657046318,
      "num_tokens": 32578924.0,
      "step": 14865
    },
    {
      "entropy": 5.601125812530517,
      "epoch": 1.4757840412862246,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047887480697369155,
      "loss": 5.1333,
      "mean_token_accuracy": 0.1767551302909851,
      "num_tokens": 32590233.0,
      "step": 14870
    },
    {
      "entropy": 5.545289659500122,
      "epoch": 1.4762802699483921,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047885981544855475,
      "loss": 5.1438,
      "mean_token_accuracy": 0.1724060893058777,
      "num_tokens": 32600782.0,
      "step": 14875
    },
    {
      "entropy": 5.491605663299561,
      "epoch": 1.4767764986105598,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004788448188681339,
      "loss": 5.1344,
      "mean_token_accuracy": 0.1780883178114891,
      "num_tokens": 32612611.0,
      "step": 14880
    },
    {
      "entropy": 5.5843908309936525,
      "epoch": 1.4772727272727273,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004788298172328007,
      "loss": 5.1912,
      "mean_token_accuracy": 0.17206951677799226,
      "num_tokens": 32623436.0,
      "step": 14885
    },
    {
      "entropy": 5.551280403137207,
      "epoch": 1.4777689559348948,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047881481054292736,
      "loss": 5.0796,
      "mean_token_accuracy": 0.17461806535720825,
      "num_tokens": 32633356.0,
      "step": 14890
    },
    {
      "entropy": 5.517702865600586,
      "epoch": 1.4782651845970625,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047879979879888584,
      "loss": 5.1198,
      "mean_token_accuracy": 0.18242467045783997,
      "num_tokens": 32643927.0,
      "step": 14895
    },
    {
      "entropy": 5.483425855636597,
      "epoch": 1.47876141325923,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047878478200104845,
      "loss": 5.0922,
      "mean_token_accuracy": 0.1793556034564972,
      "num_tokens": 32655086.0,
      "step": 14900
    },
    {
      "entropy": 5.512010192871093,
      "epoch": 1.4792576419213974,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004787697601497876,
      "loss": 5.0409,
      "mean_token_accuracy": 0.17739986032247543,
      "num_tokens": 32665298.0,
      "step": 14905
    },
    {
      "entropy": 5.546717119216919,
      "epoch": 1.479753870583565,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004787547332454758,
      "loss": 5.1035,
      "mean_token_accuracy": 0.167855703830719,
      "num_tokens": 32676717.0,
      "step": 14910
    },
    {
      "entropy": 5.584608888626098,
      "epoch": 1.4802500992457324,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047873970128848567,
      "loss": 5.167,
      "mean_token_accuracy": 0.1665666699409485,
      "num_tokens": 32686886.0,
      "step": 14915
    },
    {
      "entropy": 5.538633918762207,
      "epoch": 1.4807463279078998,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004787246642791899,
      "loss": 5.0916,
      "mean_token_accuracy": 0.17911523431539536,
      "num_tokens": 32698631.0,
      "step": 14920
    },
    {
      "entropy": 5.550447654724121,
      "epoch": 1.4812425565700675,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00047870962221796147,
      "loss": 5.1171,
      "mean_token_accuracy": 0.16417475044727325,
      "num_tokens": 32710672.0,
      "step": 14925
    },
    {
      "entropy": 5.564852905273438,
      "epoch": 1.481738785232235,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004786945751051734,
      "loss": 5.1567,
      "mean_token_accuracy": 0.17314314395189284,
      "num_tokens": 32720858.0,
      "step": 14930
    },
    {
      "entropy": 5.503535413742066,
      "epoch": 1.4822350138944025,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004786795229411988,
      "loss": 5.0882,
      "mean_token_accuracy": 0.1734174072742462,
      "num_tokens": 32731592.0,
      "step": 14935
    },
    {
      "entropy": 5.617137622833252,
      "epoch": 1.4827312425565702,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047866446572641084,
      "loss": 5.1544,
      "mean_token_accuracy": 0.1668742388486862,
      "num_tokens": 32742777.0,
      "step": 14940
    },
    {
      "entropy": 5.569394874572754,
      "epoch": 1.4832274712187377,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004786494034611831,
      "loss": 5.0828,
      "mean_token_accuracy": 0.1707368016242981,
      "num_tokens": 32753035.0,
      "step": 14945
    },
    {
      "entropy": 5.5360575199127195,
      "epoch": 1.4837236998809051,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000478634336145889,
      "loss": 5.1169,
      "mean_token_accuracy": 0.17117978781461715,
      "num_tokens": 32764318.0,
      "step": 14950
    },
    {
      "entropy": 5.5492650985717775,
      "epoch": 1.4842199285430726,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047861926378090197,
      "loss": 5.1355,
      "mean_token_accuracy": 0.1678936168551445,
      "num_tokens": 32775521.0,
      "step": 14955
    },
    {
      "entropy": 5.595921468734741,
      "epoch": 1.48471615720524,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004786041863665961,
      "loss": 5.1455,
      "mean_token_accuracy": 0.17291181832551955,
      "num_tokens": 32787101.0,
      "step": 14960
    },
    {
      "entropy": 5.538467836380005,
      "epoch": 1.4852123858674078,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004785891039033451,
      "loss": 5.1517,
      "mean_token_accuracy": 0.17064756751060486,
      "num_tokens": 32798446.0,
      "step": 14965
    },
    {
      "entropy": 5.555763387680054,
      "epoch": 1.4857086145295753,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000478574016391523,
      "loss": 5.1529,
      "mean_token_accuracy": 0.1614686831831932,
      "num_tokens": 32809636.0,
      "step": 14970
    },
    {
      "entropy": 5.50410566329956,
      "epoch": 1.4862048431917427,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000478558923831504,
      "loss": 5.056,
      "mean_token_accuracy": 0.1740719422698021,
      "num_tokens": 32820132.0,
      "step": 14975
    },
    {
      "entropy": 5.5598358631134035,
      "epoch": 1.4867010718539102,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047854382622366235,
      "loss": 5.1429,
      "mean_token_accuracy": 0.17363158166408538,
      "num_tokens": 32830720.0,
      "step": 14980
    },
    {
      "entropy": 5.510737133026123,
      "epoch": 1.487197300516078,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047852872356837236,
      "loss": 5.09,
      "mean_token_accuracy": 0.17761751115322114,
      "num_tokens": 32841621.0,
      "step": 14985
    },
    {
      "entropy": 5.635169219970703,
      "epoch": 1.4876935291782454,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004785136158660085,
      "loss": 5.2308,
      "mean_token_accuracy": 0.1646098554134369,
      "num_tokens": 32852074.0,
      "step": 14990
    },
    {
      "entropy": 5.641831541061402,
      "epoch": 1.4881897578404129,
      "grad_norm": 1.0,
      "learning_rate": 0.00047849850311694567,
      "loss": 5.1816,
      "mean_token_accuracy": 0.1754940703511238,
      "num_tokens": 32862556.0,
      "step": 14995
    },
    {
      "entropy": 5.536677646636963,
      "epoch": 1.4886859865025803,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047848338532155833,
      "loss": 5.1138,
      "mean_token_accuracy": 0.17840631902217866,
      "num_tokens": 32874407.0,
      "step": 15000
    },
    {
      "epoch": 1.4886859865025803,
      "eval_entropy": 5.356222114757847,
      "eval_loss": 5.242770671844482,
      "eval_mean_token_accuracy": 0.17171108076757363,
      "eval_num_tokens": 32874407.0,
      "eval_runtime": 20.9474,
      "eval_samples_per_second": 1607.312,
      "eval_steps_per_second": 200.932,
      "step": 15000
    },
    {
      "entropy": 5.548680400848388,
      "epoch": 1.4891822151647478,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047846826248022155,
      "loss": 5.1454,
      "mean_token_accuracy": 0.16540502160787582,
      "num_tokens": 32884986.0,
      "step": 15005
    },
    {
      "entropy": 5.618879175186157,
      "epoch": 1.4896784438269155,
      "grad_norm": 0.859375,
      "learning_rate": 0.00047845313459331034,
      "loss": 5.2049,
      "mean_token_accuracy": 0.17036499977111816,
      "num_tokens": 32896016.0,
      "step": 15010
    },
    {
      "entropy": 5.530941724777222,
      "epoch": 1.490174672489083,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047843800166119974,
      "loss": 5.1048,
      "mean_token_accuracy": 0.17776186615228654,
      "num_tokens": 32905648.0,
      "step": 15015
    },
    {
      "entropy": 5.597233629226684,
      "epoch": 1.4906709011512504,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000478422863684265,
      "loss": 5.1645,
      "mean_token_accuracy": 0.17507992535829545,
      "num_tokens": 32917103.0,
      "step": 15020
    },
    {
      "entropy": 5.577707052230835,
      "epoch": 1.4911671298134181,
      "grad_norm": 0.859375,
      "learning_rate": 0.00047840772066288167,
      "loss": 5.0961,
      "mean_token_accuracy": 0.17214539498090745,
      "num_tokens": 32928741.0,
      "step": 15025
    },
    {
      "entropy": 5.533963251113891,
      "epoch": 1.4916633584755856,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004783925725974252,
      "loss": 5.1665,
      "mean_token_accuracy": 0.16858091801404954,
      "num_tokens": 32939292.0,
      "step": 15030
    },
    {
      "entropy": 5.60145378112793,
      "epoch": 1.492159587137753,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004783774194882712,
      "loss": 5.1551,
      "mean_token_accuracy": 0.17159147411584855,
      "num_tokens": 32949940.0,
      "step": 15035
    },
    {
      "entropy": 5.544801235198975,
      "epoch": 1.4926558157999206,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004783622613357954,
      "loss": 5.1068,
      "mean_token_accuracy": 0.17158510982990266,
      "num_tokens": 32961222.0,
      "step": 15040
    },
    {
      "entropy": 5.496397590637207,
      "epoch": 1.493152044462088,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004783470981403737,
      "loss": 5.1061,
      "mean_token_accuracy": 0.17940518260002136,
      "num_tokens": 32972095.0,
      "step": 15045
    },
    {
      "entropy": 5.57369704246521,
      "epoch": 1.4936482731242555,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004783319299023821,
      "loss": 5.145,
      "mean_token_accuracy": 0.17390543967485428,
      "num_tokens": 32981834.0,
      "step": 15050
    },
    {
      "entropy": 5.665181493759155,
      "epoch": 1.4941445017864232,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004783167566221969,
      "loss": 5.2349,
      "mean_token_accuracy": 0.16971024125814438,
      "num_tokens": 32994347.0,
      "step": 15055
    },
    {
      "entropy": 5.635477066040039,
      "epoch": 1.4946407304485907,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047830157830019423,
      "loss": 5.1918,
      "mean_token_accuracy": 0.16387041807174682,
      "num_tokens": 33004780.0,
      "step": 15060
    },
    {
      "entropy": 5.588273239135742,
      "epoch": 1.4951369591107582,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004782863949367505,
      "loss": 5.158,
      "mean_token_accuracy": 0.16703834682703017,
      "num_tokens": 33015697.0,
      "step": 15065
    },
    {
      "entropy": 5.503173780441284,
      "epoch": 1.4956331877729259,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004782712065322421,
      "loss": 5.1271,
      "mean_token_accuracy": 0.16697831302881241,
      "num_tokens": 33026563.0,
      "step": 15070
    },
    {
      "entropy": 5.5939418315887455,
      "epoch": 1.4961294164350933,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004782560130870459,
      "loss": 5.1456,
      "mean_token_accuracy": 0.1666735142469406,
      "num_tokens": 33037290.0,
      "step": 15075
    },
    {
      "entropy": 5.601247882843017,
      "epoch": 1.4966256450972608,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047824081460153846,
      "loss": 5.0831,
      "mean_token_accuracy": 0.17054253816604614,
      "num_tokens": 33047199.0,
      "step": 15080
    },
    {
      "entropy": 5.526590490341187,
      "epoch": 1.4971218737594283,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004782256110760969,
      "loss": 5.1038,
      "mean_token_accuracy": 0.17255692929029465,
      "num_tokens": 33057253.0,
      "step": 15085
    },
    {
      "entropy": 5.5876223087310795,
      "epoch": 1.4976181024215958,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000478210402511098,
      "loss": 5.2528,
      "mean_token_accuracy": 0.16506772637367248,
      "num_tokens": 33068228.0,
      "step": 15090
    },
    {
      "entropy": 5.6244049072265625,
      "epoch": 1.4981143310837635,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004781951889069189,
      "loss": 5.194,
      "mean_token_accuracy": 0.17383694052696227,
      "num_tokens": 33079285.0,
      "step": 15095
    },
    {
      "entropy": 5.561894941329956,
      "epoch": 1.498610559745931,
      "grad_norm": 0.84375,
      "learning_rate": 0.00047817997026393714,
      "loss": 5.0647,
      "mean_token_accuracy": 0.18118080496788025,
      "num_tokens": 33091253.0,
      "step": 15100
    },
    {
      "entropy": 5.483845281600952,
      "epoch": 1.4991067884080984,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047816474658252985,
      "loss": 5.0959,
      "mean_token_accuracy": 0.17563858926296233,
      "num_tokens": 33102492.0,
      "step": 15105
    },
    {
      "entropy": 5.500487232208252,
      "epoch": 1.499603017070266,
      "grad_norm": 1.0,
      "learning_rate": 0.00047814951786307467,
      "loss": 5.095,
      "mean_token_accuracy": 0.1764495015144348,
      "num_tokens": 33115038.0,
      "step": 15110
    },
    {
      "entropy": 5.565530395507812,
      "epoch": 1.5000992457324336,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004781342841059492,
      "loss": 5.0803,
      "mean_token_accuracy": 0.1781868427991867,
      "num_tokens": 33124470.0,
      "step": 15115
    },
    {
      "entropy": 5.5187924861907955,
      "epoch": 1.500595474394601,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004781190453115311,
      "loss": 5.099,
      "mean_token_accuracy": 0.1759433701634407,
      "num_tokens": 33135994.0,
      "step": 15120
    },
    {
      "entropy": 5.536745023727417,
      "epoch": 1.5010917030567685,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047810380148019847,
      "loss": 5.1202,
      "mean_token_accuracy": 0.17155790477991104,
      "num_tokens": 33145783.0,
      "step": 15125
    },
    {
      "entropy": 5.548599195480347,
      "epoch": 1.501587931718936,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047808855261232915,
      "loss": 5.2034,
      "mean_token_accuracy": 0.17392805367708206,
      "num_tokens": 33156404.0,
      "step": 15130
    },
    {
      "entropy": 5.564274835586548,
      "epoch": 1.5020841603811035,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047807329870830134,
      "loss": 5.196,
      "mean_token_accuracy": 0.1735937312245369,
      "num_tokens": 33168226.0,
      "step": 15135
    },
    {
      "entropy": 5.567821931838989,
      "epoch": 1.5025803890432712,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004780580397684933,
      "loss": 5.1268,
      "mean_token_accuracy": 0.16919144988059998,
      "num_tokens": 33178265.0,
      "step": 15140
    },
    {
      "entropy": 5.529192304611206,
      "epoch": 1.5030766177054387,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004780427757932835,
      "loss": 5.0839,
      "mean_token_accuracy": 0.17965191453695298,
      "num_tokens": 33188992.0,
      "step": 15145
    },
    {
      "entropy": 5.527132368087768,
      "epoch": 1.5035728463676064,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004780275067830503,
      "loss": 5.1671,
      "mean_token_accuracy": 0.17157556712627411,
      "num_tokens": 33199424.0,
      "step": 15150
    },
    {
      "entropy": 5.515825700759888,
      "epoch": 1.5040690750297738,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004780122327381725,
      "loss": 5.0797,
      "mean_token_accuracy": 0.18236674666404723,
      "num_tokens": 33210900.0,
      "step": 15155
    },
    {
      "entropy": 5.615907859802246,
      "epoch": 1.5045653036919413,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047799695365902865,
      "loss": 5.166,
      "mean_token_accuracy": 0.16845391392707826,
      "num_tokens": 33220796.0,
      "step": 15160
    },
    {
      "entropy": 5.5539452075958256,
      "epoch": 1.5050615323541088,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004779816695459979,
      "loss": 5.1315,
      "mean_token_accuracy": 0.17035460323095322,
      "num_tokens": 33233454.0,
      "step": 15165
    },
    {
      "entropy": 5.557143783569336,
      "epoch": 1.5055577610162763,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004779663803994591,
      "loss": 5.0641,
      "mean_token_accuracy": 0.17768268138170243,
      "num_tokens": 33243574.0,
      "step": 15170
    },
    {
      "entropy": 5.550082588195801,
      "epoch": 1.5060539896784437,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004779510862197914,
      "loss": 5.1149,
      "mean_token_accuracy": 0.17416314482688905,
      "num_tokens": 33256298.0,
      "step": 15175
    },
    {
      "entropy": 5.597191667556762,
      "epoch": 1.5065502183406112,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004779357870073741,
      "loss": 5.2187,
      "mean_token_accuracy": 0.16762249320745468,
      "num_tokens": 33268089.0,
      "step": 15180
    },
    {
      "entropy": 5.527649068832398,
      "epoch": 1.507046447002779,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047792048276258653,
      "loss": 5.1397,
      "mean_token_accuracy": 0.16818636655807495,
      "num_tokens": 33279955.0,
      "step": 15185
    },
    {
      "entropy": 5.5470537662506105,
      "epoch": 1.5075426756649464,
      "grad_norm": 1.0,
      "learning_rate": 0.0004779051734858083,
      "loss": 5.1183,
      "mean_token_accuracy": 0.17061854004859925,
      "num_tokens": 33291470.0,
      "step": 15190
    },
    {
      "entropy": 5.582837200164795,
      "epoch": 1.508038904327114,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047788985917741894,
      "loss": 5.1537,
      "mean_token_accuracy": 0.17951596677303314,
      "num_tokens": 33304169.0,
      "step": 15195
    },
    {
      "entropy": 5.462422704696655,
      "epoch": 1.5085351329892815,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047787453983779826,
      "loss": 4.9843,
      "mean_token_accuracy": 0.18072272092103958,
      "num_tokens": 33314436.0,
      "step": 15200
    },
    {
      "entropy": 5.515456438064575,
      "epoch": 1.509031361651449,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004778592154673262,
      "loss": 5.2156,
      "mean_token_accuracy": 0.16572293639183044,
      "num_tokens": 33325361.0,
      "step": 15205
    },
    {
      "entropy": 5.464729118347168,
      "epoch": 1.5095275903136165,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047784388606638263,
      "loss": 5.0869,
      "mean_token_accuracy": 0.1688888967037201,
      "num_tokens": 33336644.0,
      "step": 15210
    },
    {
      "entropy": 5.547032070159912,
      "epoch": 1.510023818975784,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004778285516353478,
      "loss": 5.142,
      "mean_token_accuracy": 0.16770277172327042,
      "num_tokens": 33348418.0,
      "step": 15215
    },
    {
      "entropy": 5.548987483978271,
      "epoch": 1.5105200476379514,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047781321217460205,
      "loss": 5.0922,
      "mean_token_accuracy": 0.17391715198755264,
      "num_tokens": 33360615.0,
      "step": 15220
    },
    {
      "entropy": 5.531084823608398,
      "epoch": 1.5110162763001191,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047779786768452554,
      "loss": 5.0947,
      "mean_token_accuracy": 0.1704520285129547,
      "num_tokens": 33372030.0,
      "step": 15225
    },
    {
      "entropy": 5.539376640319825,
      "epoch": 1.5115125049622866,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047778251816549895,
      "loss": 5.1059,
      "mean_token_accuracy": 0.17253375947475433,
      "num_tokens": 33382834.0,
      "step": 15230
    },
    {
      "entropy": 5.528509044647217,
      "epoch": 1.512008733624454,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047776716361790284,
      "loss": 5.1251,
      "mean_token_accuracy": 0.1769322633743286,
      "num_tokens": 33393250.0,
      "step": 15235
    },
    {
      "entropy": 5.5322105884552,
      "epoch": 1.5125049622866218,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047775180404211804,
      "loss": 5.0779,
      "mean_token_accuracy": 0.1776392266154289,
      "num_tokens": 33402778.0,
      "step": 15240
    },
    {
      "entropy": 5.602771425247193,
      "epoch": 1.5130011909487893,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004777364394385254,
      "loss": 5.1553,
      "mean_token_accuracy": 0.16944981515407562,
      "num_tokens": 33413565.0,
      "step": 15245
    },
    {
      "entropy": 5.591983222961426,
      "epoch": 1.5134974196109567,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047772106980750593,
      "loss": 5.0954,
      "mean_token_accuracy": 0.17246437817811966,
      "num_tokens": 33424108.0,
      "step": 15250
    },
    {
      "entropy": 5.521729278564453,
      "epoch": 1.5139936482731242,
      "grad_norm": 1.0,
      "learning_rate": 0.0004777056951494407,
      "loss": 5.1729,
      "mean_token_accuracy": 0.16888258308172227,
      "num_tokens": 33435683.0,
      "step": 15255
    },
    {
      "entropy": 5.49679160118103,
      "epoch": 1.5144898769352917,
      "grad_norm": 1.078125,
      "learning_rate": 0.000477690315464711,
      "loss": 5.1372,
      "mean_token_accuracy": 0.17263571918010712,
      "num_tokens": 33446267.0,
      "step": 15260
    },
    {
      "entropy": 5.511676931381226,
      "epoch": 1.5149861055974592,
      "grad_norm": 0.875,
      "learning_rate": 0.00047767493075369827,
      "loss": 5.0597,
      "mean_token_accuracy": 0.19019062966108322,
      "num_tokens": 33457901.0,
      "step": 15265
    },
    {
      "entropy": 5.61328821182251,
      "epoch": 1.5154823342596269,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000477659541016784,
      "loss": 5.2317,
      "mean_token_accuracy": 0.17092126458883286,
      "num_tokens": 33469215.0,
      "step": 15270
    },
    {
      "entropy": 5.572698926925659,
      "epoch": 1.5159785629217943,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047764414625434983,
      "loss": 5.0951,
      "mean_token_accuracy": 0.17466504722833634,
      "num_tokens": 33480384.0,
      "step": 15275
    },
    {
      "entropy": 5.574563980102539,
      "epoch": 1.516474791583962,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004776287464667774,
      "loss": 5.0944,
      "mean_token_accuracy": 0.17736030817031861,
      "num_tokens": 33491404.0,
      "step": 15280
    },
    {
      "entropy": 5.632802820205688,
      "epoch": 1.5169710202461295,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047761334165444876,
      "loss": 5.2227,
      "mean_token_accuracy": 0.16314250975847244,
      "num_tokens": 33501888.0,
      "step": 15285
    },
    {
      "entropy": 5.571307516098022,
      "epoch": 1.517467248908297,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047759793181774586,
      "loss": 5.1781,
      "mean_token_accuracy": 0.17024213820695877,
      "num_tokens": 33513274.0,
      "step": 15290
    },
    {
      "entropy": 5.545615291595459,
      "epoch": 1.5179634775704645,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047758251695705073,
      "loss": 5.1408,
      "mean_token_accuracy": 0.17191704213619233,
      "num_tokens": 33524188.0,
      "step": 15295
    },
    {
      "entropy": 5.599178743362427,
      "epoch": 1.518459706232632,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047756709707274576,
      "loss": 5.2006,
      "mean_token_accuracy": 0.1723843589425087,
      "num_tokens": 33535207.0,
      "step": 15300
    },
    {
      "entropy": 5.6099165916442875,
      "epoch": 1.5189559348947994,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004775516721652133,
      "loss": 5.143,
      "mean_token_accuracy": 0.1672565758228302,
      "num_tokens": 33546154.0,
      "step": 15305
    },
    {
      "entropy": 5.5444395542144775,
      "epoch": 1.519452163556967,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004775362422348358,
      "loss": 5.073,
      "mean_token_accuracy": 0.17534394264221193,
      "num_tokens": 33555365.0,
      "step": 15310
    },
    {
      "entropy": 5.535499095916748,
      "epoch": 1.5199483922191346,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047752080728199586,
      "loss": 5.1298,
      "mean_token_accuracy": 0.1708448588848114,
      "num_tokens": 33566316.0,
      "step": 15315
    },
    {
      "entropy": 5.550029373168945,
      "epoch": 1.520444620881302,
      "grad_norm": 1.0,
      "learning_rate": 0.0004775053673070764,
      "loss": 5.0615,
      "mean_token_accuracy": 0.18068691492080688,
      "num_tokens": 33576549.0,
      "step": 15320
    },
    {
      "entropy": 5.535349607467651,
      "epoch": 1.5209408495434698,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004774899223104601,
      "loss": 5.1249,
      "mean_token_accuracy": 0.17240564227104188,
      "num_tokens": 33586681.0,
      "step": 15325
    },
    {
      "entropy": 5.644279193878174,
      "epoch": 1.5214370782056372,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004774744722925301,
      "loss": 5.2632,
      "mean_token_accuracy": 0.16375259011983873,
      "num_tokens": 33599201.0,
      "step": 15330
    },
    {
      "entropy": 5.585297775268555,
      "epoch": 1.5219333068678047,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004774590172536695,
      "loss": 5.1178,
      "mean_token_accuracy": 0.17082442194223404,
      "num_tokens": 33610217.0,
      "step": 15335
    },
    {
      "entropy": 5.556169700622559,
      "epoch": 1.5224295355299722,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047744355719426146,
      "loss": 5.1431,
      "mean_token_accuracy": 0.17410140484571457,
      "num_tokens": 33620586.0,
      "step": 15340
    },
    {
      "entropy": 5.5452680587768555,
      "epoch": 1.5229257641921397,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047742809211468953,
      "loss": 5.1149,
      "mean_token_accuracy": 0.17649095505475998,
      "num_tokens": 33630712.0,
      "step": 15345
    },
    {
      "entropy": 5.5179625988006595,
      "epoch": 1.5234219928543071,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000477412622015337,
      "loss": 5.0834,
      "mean_token_accuracy": 0.1743198424577713,
      "num_tokens": 33640367.0,
      "step": 15350
    },
    {
      "entropy": 5.573330020904541,
      "epoch": 1.5239182215164748,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004773971468965876,
      "loss": 5.1792,
      "mean_token_accuracy": 0.16839204132556915,
      "num_tokens": 33651816.0,
      "step": 15355
    },
    {
      "entropy": 5.562081670761108,
      "epoch": 1.5244144501786423,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047738166675882516,
      "loss": 5.0497,
      "mean_token_accuracy": 0.17609239518642425,
      "num_tokens": 33663661.0,
      "step": 15360
    },
    {
      "entropy": 5.471735429763794,
      "epoch": 1.52491067884081,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004773661816024335,
      "loss": 4.9736,
      "mean_token_accuracy": 0.18471864908933638,
      "num_tokens": 33673670.0,
      "step": 15365
    },
    {
      "entropy": 5.5232867240905765,
      "epoch": 1.5254069075029775,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004773506914277965,
      "loss": 5.211,
      "mean_token_accuracy": 0.16723523586988448,
      "num_tokens": 33685434.0,
      "step": 15370
    },
    {
      "entropy": 5.568566513061524,
      "epoch": 1.525903136165145,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047733519623529853,
      "loss": 5.1559,
      "mean_token_accuracy": 0.1741061955690384,
      "num_tokens": 33696751.0,
      "step": 15375
    },
    {
      "entropy": 5.565272235870362,
      "epoch": 1.5263993648273124,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004773196960253236,
      "loss": 5.0989,
      "mean_token_accuracy": 0.17368203550577163,
      "num_tokens": 33708141.0,
      "step": 15380
    },
    {
      "entropy": 5.4547648429870605,
      "epoch": 1.52689559348948,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004773041907982561,
      "loss": 4.9609,
      "mean_token_accuracy": 0.18090272843837737,
      "num_tokens": 33719219.0,
      "step": 15385
    },
    {
      "entropy": 5.539753818511963,
      "epoch": 1.5273918221516474,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047728868055448064,
      "loss": 5.1293,
      "mean_token_accuracy": 0.1711447536945343,
      "num_tokens": 33730493.0,
      "step": 15390
    },
    {
      "entropy": 5.598049592971802,
      "epoch": 1.5278880508138148,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004772731652943818,
      "loss": 5.1692,
      "mean_token_accuracy": 0.16518838554620743,
      "num_tokens": 33740273.0,
      "step": 15395
    },
    {
      "entropy": 5.598770332336426,
      "epoch": 1.5283842794759825,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004772576450183444,
      "loss": 5.138,
      "mean_token_accuracy": 0.17564401179552078,
      "num_tokens": 33752162.0,
      "step": 15400
    },
    {
      "entropy": 5.5537309646606445,
      "epoch": 1.52888050813815,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047724211972675315,
      "loss": 5.0851,
      "mean_token_accuracy": 0.1708499774336815,
      "num_tokens": 33763139.0,
      "step": 15405
    },
    {
      "entropy": 5.5139909267425535,
      "epoch": 1.5293767368003177,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004772265894199931,
      "loss": 5.1407,
      "mean_token_accuracy": 0.1749879702925682,
      "num_tokens": 33774641.0,
      "step": 15410
    },
    {
      "entropy": 5.554297733306885,
      "epoch": 1.5298729654624852,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004772110540984494,
      "loss": 5.1115,
      "mean_token_accuracy": 0.18277724981307983,
      "num_tokens": 33785644.0,
      "step": 15415
    },
    {
      "entropy": 5.662728357315063,
      "epoch": 1.5303691941246527,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004771955137625073,
      "loss": 5.2568,
      "mean_token_accuracy": 0.16644036620855332,
      "num_tokens": 33797880.0,
      "step": 15420
    },
    {
      "entropy": 5.509476852416992,
      "epoch": 1.5308654227868201,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047717996841255224,
      "loss": 5.063,
      "mean_token_accuracy": 0.17345482110977173,
      "num_tokens": 33808782.0,
      "step": 15425
    },
    {
      "entropy": 5.498629474639893,
      "epoch": 1.5313616514489876,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047716441804896955,
      "loss": 5.1058,
      "mean_token_accuracy": 0.1760659098625183,
      "num_tokens": 33818483.0,
      "step": 15430
    },
    {
      "entropy": 5.548745012283325,
      "epoch": 1.531857880111155,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004771488626721449,
      "loss": 5.1022,
      "mean_token_accuracy": 0.16741147190332412,
      "num_tokens": 33828590.0,
      "step": 15435
    },
    {
      "entropy": 5.624418115615844,
      "epoch": 1.5323541087733228,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047713330228246406,
      "loss": 5.1929,
      "mean_token_accuracy": 0.16959738433361055,
      "num_tokens": 33838978.0,
      "step": 15440
    },
    {
      "entropy": 5.543429565429688,
      "epoch": 1.5328503374354903,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004771177368803129,
      "loss": 5.1269,
      "mean_token_accuracy": 0.17334332168102265,
      "num_tokens": 33851244.0,
      "step": 15445
    },
    {
      "entropy": 5.545010232925415,
      "epoch": 1.533346566097658,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047710216646607733,
      "loss": 5.0202,
      "mean_token_accuracy": 0.18064377903938295,
      "num_tokens": 33861666.0,
      "step": 15450
    },
    {
      "entropy": 5.608187913894653,
      "epoch": 1.5338427947598254,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047708659104014363,
      "loss": 5.1025,
      "mean_token_accuracy": 0.17549319565296173,
      "num_tokens": 33872416.0,
      "step": 15455
    },
    {
      "entropy": 5.499054622650147,
      "epoch": 1.534339023421993,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004770710106028978,
      "loss": 5.0874,
      "mean_token_accuracy": 0.1733687475323677,
      "num_tokens": 33881808.0,
      "step": 15460
    },
    {
      "entropy": 5.5644495487213135,
      "epoch": 1.5348352520841604,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047705542515472647,
      "loss": 5.1817,
      "mean_token_accuracy": 0.17119942903518676,
      "num_tokens": 33892925.0,
      "step": 15465
    },
    {
      "entropy": 5.585255765914917,
      "epoch": 1.5353314807463279,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004770398346960159,
      "loss": 5.1149,
      "mean_token_accuracy": 0.16412315219640733,
      "num_tokens": 33903702.0,
      "step": 15470
    },
    {
      "entropy": 5.614564657211304,
      "epoch": 1.5358277094084953,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047702423922715284,
      "loss": 5.1958,
      "mean_token_accuracy": 0.17241449356079103,
      "num_tokens": 33917233.0,
      "step": 15475
    },
    {
      "entropy": 5.588253831863403,
      "epoch": 1.5363239380706628,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004770086387485239,
      "loss": 5.1624,
      "mean_token_accuracy": 0.17151085287332535,
      "num_tokens": 33928362.0,
      "step": 15480
    },
    {
      "entropy": 5.633815765380859,
      "epoch": 1.5368201667328305,
      "grad_norm": 1.015625,
      "learning_rate": 0.000476993033260516,
      "loss": 5.1821,
      "mean_token_accuracy": 0.17045846134424208,
      "num_tokens": 33938624.0,
      "step": 15485
    },
    {
      "entropy": 5.53687162399292,
      "epoch": 1.537316395394998,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004769774227635162,
      "loss": 5.075,
      "mean_token_accuracy": 0.17419005781412125,
      "num_tokens": 33950058.0,
      "step": 15490
    },
    {
      "entropy": 5.571239185333252,
      "epoch": 1.5378126240571657,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004769618072579115,
      "loss": 5.187,
      "mean_token_accuracy": 0.16624055951833724,
      "num_tokens": 33962404.0,
      "step": 15495
    },
    {
      "entropy": 5.58046555519104,
      "epoch": 1.5383088527193332,
      "grad_norm": 0.859375,
      "learning_rate": 0.00047694618674408917,
      "loss": 5.1171,
      "mean_token_accuracy": 0.1723117709159851,
      "num_tokens": 33973095.0,
      "step": 15500
    },
    {
      "entropy": 5.579331111907959,
      "epoch": 1.5388050813815006,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047693056122243655,
      "loss": 5.1391,
      "mean_token_accuracy": 0.17182512134313582,
      "num_tokens": 33984152.0,
      "step": 15505
    },
    {
      "entropy": 5.61326231956482,
      "epoch": 1.539301310043668,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047691493069334124,
      "loss": 5.1652,
      "mean_token_accuracy": 0.1681300088763237,
      "num_tokens": 33994854.0,
      "step": 15510
    },
    {
      "entropy": 5.589783716201782,
      "epoch": 1.5397975387058356,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047689929515719066,
      "loss": 5.1696,
      "mean_token_accuracy": 0.17352217584848403,
      "num_tokens": 34006573.0,
      "step": 15515
    },
    {
      "entropy": 5.5608540058135985,
      "epoch": 1.540293767368003,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047688365461437263,
      "loss": 5.0862,
      "mean_token_accuracy": 0.17178861945867538,
      "num_tokens": 34016537.0,
      "step": 15520
    },
    {
      "entropy": 5.551642656326294,
      "epoch": 1.5407899960301707,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047686800906527493,
      "loss": 5.0983,
      "mean_token_accuracy": 0.18281222134828568,
      "num_tokens": 34026825.0,
      "step": 15525
    },
    {
      "entropy": 5.599949502944947,
      "epoch": 1.5412862246923382,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004768523585102856,
      "loss": 5.1179,
      "mean_token_accuracy": 0.17747576534748077,
      "num_tokens": 34036786.0,
      "step": 15530
    },
    {
      "entropy": 5.584811496734619,
      "epoch": 1.541782453354506,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004768367029497928,
      "loss": 5.1135,
      "mean_token_accuracy": 0.17588718235492706,
      "num_tokens": 34046784.0,
      "step": 15535
    },
    {
      "entropy": 5.587221336364746,
      "epoch": 1.5422786820166734,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047682104238418464,
      "loss": 5.1689,
      "mean_token_accuracy": 0.17410801351070404,
      "num_tokens": 34058250.0,
      "step": 15540
    },
    {
      "entropy": 5.5062182426452635,
      "epoch": 1.5427749106788409,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047680537681384956,
      "loss": 5.093,
      "mean_token_accuracy": 0.17905824929475783,
      "num_tokens": 34069969.0,
      "step": 15545
    },
    {
      "entropy": 5.597629547119141,
      "epoch": 1.5432711393410083,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004767897062391759,
      "loss": 5.1934,
      "mean_token_accuracy": 0.17195533514022826,
      "num_tokens": 34080817.0,
      "step": 15550
    },
    {
      "entropy": 5.561704778671265,
      "epoch": 1.5437673680031758,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004767740306605524,
      "loss": 5.1355,
      "mean_token_accuracy": 0.16943931579589844,
      "num_tokens": 34091138.0,
      "step": 15555
    },
    {
      "entropy": 5.52356858253479,
      "epoch": 1.5442635966653433,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004767583500783677,
      "loss": 5.0681,
      "mean_token_accuracy": 0.1783754661679268,
      "num_tokens": 34102035.0,
      "step": 15560
    },
    {
      "entropy": 5.564588594436645,
      "epoch": 1.5447598253275108,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047674266449301074,
      "loss": 5.1371,
      "mean_token_accuracy": 0.17570264488458634,
      "num_tokens": 34112533.0,
      "step": 15565
    },
    {
      "entropy": 5.5875993251800535,
      "epoch": 1.5452560539896785,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047672697390487034,
      "loss": 5.126,
      "mean_token_accuracy": 0.17441647946834565,
      "num_tokens": 34124196.0,
      "step": 15570
    },
    {
      "entropy": 5.575704431533813,
      "epoch": 1.545752282651846,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004767112783143357,
      "loss": 5.1092,
      "mean_token_accuracy": 0.17847241312265397,
      "num_tokens": 34135635.0,
      "step": 15575
    },
    {
      "entropy": 5.5376934051513675,
      "epoch": 1.5462485113140136,
      "grad_norm": 0.984375,
      "learning_rate": 0.000476695577721796,
      "loss": 5.037,
      "mean_token_accuracy": 0.1835579514503479,
      "num_tokens": 34146202.0,
      "step": 15580
    },
    {
      "entropy": 5.549660778045654,
      "epoch": 1.5467447399761811,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047667987212764065,
      "loss": 5.0853,
      "mean_token_accuracy": 0.17492646872997283,
      "num_tokens": 34157354.0,
      "step": 15585
    },
    {
      "entropy": 5.554770660400391,
      "epoch": 1.5472409686383486,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047666416153225894,
      "loss": 5.1413,
      "mean_token_accuracy": 0.169989937543869,
      "num_tokens": 34168563.0,
      "step": 15590
    },
    {
      "entropy": 5.554446125030518,
      "epoch": 1.547737197300516,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004766484459360406,
      "loss": 5.0871,
      "mean_token_accuracy": 0.17454278469085693,
      "num_tokens": 34178604.0,
      "step": 15595
    },
    {
      "entropy": 5.561854362487793,
      "epoch": 1.5482334259626835,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047663272533937537,
      "loss": 5.1132,
      "mean_token_accuracy": 0.17516882568597794,
      "num_tokens": 34190321.0,
      "step": 15600
    },
    {
      "entropy": 5.509166717529297,
      "epoch": 1.548729654624851,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047661699974265294,
      "loss": 5.1359,
      "mean_token_accuracy": 0.17086947113275527,
      "num_tokens": 34201519.0,
      "step": 15605
    },
    {
      "entropy": 5.509905767440796,
      "epoch": 1.5492258832870187,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047660126914626344,
      "loss": 5.075,
      "mean_token_accuracy": 0.1700996831059456,
      "num_tokens": 34212710.0,
      "step": 15610
    },
    {
      "entropy": 5.6021894931793215,
      "epoch": 1.5497221119491862,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004765855335505969,
      "loss": 5.2123,
      "mean_token_accuracy": 0.1705973342061043,
      "num_tokens": 34223187.0,
      "step": 15615
    },
    {
      "entropy": 5.586636590957641,
      "epoch": 1.5502183406113537,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004765697929560434,
      "loss": 5.1522,
      "mean_token_accuracy": 0.17125836312770842,
      "num_tokens": 34234530.0,
      "step": 15620
    },
    {
      "entropy": 5.5975878715515135,
      "epoch": 1.5507145692735214,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047655404736299354,
      "loss": 5.1575,
      "mean_token_accuracy": 0.16789256334304808,
      "num_tokens": 34245003.0,
      "step": 15625
    },
    {
      "entropy": 5.539123106002807,
      "epoch": 1.5512107979356888,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047653829677183745,
      "loss": 5.1192,
      "mean_token_accuracy": 0.17660011649131774,
      "num_tokens": 34255630.0,
      "step": 15630
    },
    {
      "entropy": 5.536657333374023,
      "epoch": 1.5517070265978563,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047652254118296596,
      "loss": 5.1103,
      "mean_token_accuracy": 0.17759755849838257,
      "num_tokens": 34266716.0,
      "step": 15635
    },
    {
      "entropy": 5.584641551971435,
      "epoch": 1.5522032552600238,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004765067805967697,
      "loss": 5.1409,
      "mean_token_accuracy": 0.1703515827655792,
      "num_tokens": 34278905.0,
      "step": 15640
    },
    {
      "entropy": 5.558624410629273,
      "epoch": 1.5526994839221913,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004764910150136395,
      "loss": 5.084,
      "mean_token_accuracy": 0.17589716464281083,
      "num_tokens": 34289688.0,
      "step": 15645
    },
    {
      "entropy": 5.551822566986084,
      "epoch": 1.5531957125843587,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047647524443396625,
      "loss": 5.1233,
      "mean_token_accuracy": 0.17196367383003236,
      "num_tokens": 34300964.0,
      "step": 15650
    },
    {
      "entropy": 5.543437528610229,
      "epoch": 1.5536919412465264,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047645946885814105,
      "loss": 5.1207,
      "mean_token_accuracy": 0.17714701741933822,
      "num_tokens": 34311435.0,
      "step": 15655
    },
    {
      "entropy": 5.5781783103942875,
      "epoch": 1.554188169908694,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004764436882865552,
      "loss": 5.1141,
      "mean_token_accuracy": 0.169697043299675,
      "num_tokens": 34321742.0,
      "step": 15660
    },
    {
      "entropy": 5.6170392513275145,
      "epoch": 1.5546843985708616,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004764279027195999,
      "loss": 5.2035,
      "mean_token_accuracy": 0.16768767833709716,
      "num_tokens": 34333796.0,
      "step": 15665
    },
    {
      "entropy": 5.69798469543457,
      "epoch": 1.555180627233029,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047641211215766673,
      "loss": 5.2398,
      "mean_token_accuracy": 0.16938203871250151,
      "num_tokens": 34344521.0,
      "step": 15670
    },
    {
      "entropy": 5.528249597549438,
      "epoch": 1.5556768558951966,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004763963166011471,
      "loss": 5.1131,
      "mean_token_accuracy": 0.17630386501550674,
      "num_tokens": 34355629.0,
      "step": 15675
    },
    {
      "entropy": 5.522589874267578,
      "epoch": 1.556173084557364,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004763805160504328,
      "loss": 5.1747,
      "mean_token_accuracy": 0.17522377967834474,
      "num_tokens": 34366935.0,
      "step": 15680
    },
    {
      "entropy": 5.526805305480957,
      "epoch": 1.5566693132195315,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004763647105059157,
      "loss": 5.0822,
      "mean_token_accuracy": 0.17030524909496308,
      "num_tokens": 34378355.0,
      "step": 15685
    },
    {
      "entropy": 5.573115396499634,
      "epoch": 1.557165541881699,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004763488999679877,
      "loss": 5.0376,
      "mean_token_accuracy": 0.17525028735399245,
      "num_tokens": 34388811.0,
      "step": 15690
    },
    {
      "entropy": 5.637838459014892,
      "epoch": 1.5576617705438665,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00047633308443704076,
      "loss": 5.1588,
      "mean_token_accuracy": 0.16870612353086473,
      "num_tokens": 34399905.0,
      "step": 15695
    },
    {
      "entropy": 5.515767526626587,
      "epoch": 1.5581579992060341,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047631726391346717,
      "loss": 5.0347,
      "mean_token_accuracy": 0.17637524157762527,
      "num_tokens": 34410338.0,
      "step": 15700
    },
    {
      "entropy": 5.620912837982178,
      "epoch": 1.5586542278682016,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004763014383976593,
      "loss": 5.2098,
      "mean_token_accuracy": 0.1691242277622223,
      "num_tokens": 34421204.0,
      "step": 15705
    },
    {
      "entropy": 5.493432855606079,
      "epoch": 1.5591504565303693,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047628560789000945,
      "loss": 5.0195,
      "mean_token_accuracy": 0.1807053953409195,
      "num_tokens": 34431002.0,
      "step": 15710
    },
    {
      "entropy": 5.49546103477478,
      "epoch": 1.5596466851925368,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047626977239091027,
      "loss": 5.0809,
      "mean_token_accuracy": 0.17558871060609818,
      "num_tokens": 34442446.0,
      "step": 15715
    },
    {
      "entropy": 5.539222002029419,
      "epoch": 1.5601429138547043,
      "grad_norm": 1.0,
      "learning_rate": 0.0004762539319007545,
      "loss": 5.0368,
      "mean_token_accuracy": 0.1773058295249939,
      "num_tokens": 34453590.0,
      "step": 15720
    },
    {
      "entropy": 5.6049315452575685,
      "epoch": 1.5606391425168717,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004762380864199348,
      "loss": 5.1663,
      "mean_token_accuracy": 0.17339849323034287,
      "num_tokens": 34465614.0,
      "step": 15725
    },
    {
      "entropy": 5.625221204757691,
      "epoch": 1.5611353711790392,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004762222359488443,
      "loss": 5.2832,
      "mean_token_accuracy": 0.16179838627576829,
      "num_tokens": 34476616.0,
      "step": 15730
    },
    {
      "entropy": 5.573788499832153,
      "epoch": 1.5616315998412067,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004762063804878758,
      "loss": 5.0783,
      "mean_token_accuracy": 0.17143784761428832,
      "num_tokens": 34486260.0,
      "step": 15735
    },
    {
      "entropy": 5.55121488571167,
      "epoch": 1.5621278285033744,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047619052003742277,
      "loss": 5.1422,
      "mean_token_accuracy": 0.17481964975595474,
      "num_tokens": 34496392.0,
      "step": 15740
    },
    {
      "entropy": 5.573394870758056,
      "epoch": 1.5626240571655419,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004761746545978782,
      "loss": 5.1287,
      "mean_token_accuracy": 0.18143410980701447,
      "num_tokens": 34507218.0,
      "step": 15745
    },
    {
      "entropy": 5.6107978343963625,
      "epoch": 1.5631202858277096,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004761587841696358,
      "loss": 5.174,
      "mean_token_accuracy": 0.17035481035709382,
      "num_tokens": 34518888.0,
      "step": 15750
    },
    {
      "entropy": 5.562266778945923,
      "epoch": 1.563616514489877,
      "grad_norm": 1.015625,
      "learning_rate": 0.000476142908753089,
      "loss": 5.1039,
      "mean_token_accuracy": 0.17415506392717361,
      "num_tokens": 34529865.0,
      "step": 15755
    },
    {
      "entropy": 5.60737566947937,
      "epoch": 1.5641127431520445,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004761270283486314,
      "loss": 5.1699,
      "mean_token_accuracy": 0.17135939747095108,
      "num_tokens": 34542541.0,
      "step": 15760
    },
    {
      "entropy": 5.630389404296875,
      "epoch": 1.564608971814212,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047611114295665697,
      "loss": 5.1788,
      "mean_token_accuracy": 0.17126974016427993,
      "num_tokens": 34553056.0,
      "step": 15765
    },
    {
      "entropy": 5.6379937648773195,
      "epoch": 1.5651052004763795,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047609525257755967,
      "loss": 5.1639,
      "mean_token_accuracy": 0.17363667339086533,
      "num_tokens": 34563344.0,
      "step": 15770
    },
    {
      "entropy": 5.585951137542724,
      "epoch": 1.565601429138547,
      "grad_norm": 1.0,
      "learning_rate": 0.0004760793572117333,
      "loss": 5.0617,
      "mean_token_accuracy": 0.17982448041439056,
      "num_tokens": 34574320.0,
      "step": 15775
    },
    {
      "entropy": 5.520719909667969,
      "epoch": 1.5660976578007144,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004760634568595721,
      "loss": 5.1224,
      "mean_token_accuracy": 0.1677686557173729,
      "num_tokens": 34584728.0,
      "step": 15780
    },
    {
      "entropy": 5.613148403167725,
      "epoch": 1.5665938864628821,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004760475515214704,
      "loss": 5.2267,
      "mean_token_accuracy": 0.17143196910619735,
      "num_tokens": 34596395.0,
      "step": 15785
    },
    {
      "entropy": 5.525284719467163,
      "epoch": 1.5670901151250496,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004760316411978228,
      "loss": 5.117,
      "mean_token_accuracy": 0.17188528031110764,
      "num_tokens": 34607950.0,
      "step": 15790
    },
    {
      "entropy": 5.550534439086914,
      "epoch": 1.5675863437872173,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047601572588902363,
      "loss": 5.1205,
      "mean_token_accuracy": 0.1808302029967308,
      "num_tokens": 34617640.0,
      "step": 15795
    },
    {
      "entropy": 5.5752359390258786,
      "epoch": 1.5680825724493848,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047599980559546757,
      "loss": 5.1724,
      "mean_token_accuracy": 0.16710701137781142,
      "num_tokens": 34629475.0,
      "step": 15800
    },
    {
      "entropy": 5.555687284469604,
      "epoch": 1.5685788011115522,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004759838803175495,
      "loss": 5.1089,
      "mean_token_accuracy": 0.1698293536901474,
      "num_tokens": 34641423.0,
      "step": 15805
    },
    {
      "entropy": 5.66741943359375,
      "epoch": 1.5690750297737197,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047596795005566423,
      "loss": 5.1473,
      "mean_token_accuracy": 0.1724866062402725,
      "num_tokens": 34651894.0,
      "step": 15810
    },
    {
      "entropy": 5.606314182281494,
      "epoch": 1.5695712584358872,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047595201481020685,
      "loss": 5.1645,
      "mean_token_accuracy": 0.16915772408246993,
      "num_tokens": 34664177.0,
      "step": 15815
    },
    {
      "entropy": 5.5669067859649655,
      "epoch": 1.5700674870980547,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047593607458157257,
      "loss": 5.1207,
      "mean_token_accuracy": 0.17064954042434693,
      "num_tokens": 34674850.0,
      "step": 15820
    },
    {
      "entropy": 5.585564422607422,
      "epoch": 1.5705637157602224,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047592012937015647,
      "loss": 5.1613,
      "mean_token_accuracy": 0.17777508199214936,
      "num_tokens": 34685719.0,
      "step": 15825
    },
    {
      "entropy": 5.629309129714966,
      "epoch": 1.5710599444223898,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047590417917635423,
      "loss": 5.1591,
      "mean_token_accuracy": 0.16732092201709747,
      "num_tokens": 34697261.0,
      "step": 15830
    },
    {
      "entropy": 5.606767702102661,
      "epoch": 1.5715561730845575,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004758882240005613,
      "loss": 5.1607,
      "mean_token_accuracy": 0.16877879202365875,
      "num_tokens": 34708196.0,
      "step": 15835
    },
    {
      "entropy": 5.623531723022461,
      "epoch": 1.572052401746725,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047587226384317323,
      "loss": 5.1752,
      "mean_token_accuracy": 0.169980725646019,
      "num_tokens": 34718855.0,
      "step": 15840
    },
    {
      "entropy": 5.564963436126709,
      "epoch": 1.5725486304088925,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047585629870458584,
      "loss": 5.0921,
      "mean_token_accuracy": 0.18022489100694655,
      "num_tokens": 34731137.0,
      "step": 15845
    },
    {
      "entropy": 5.589971685409546,
      "epoch": 1.57304485907106,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004758403285851951,
      "loss": 5.0466,
      "mean_token_accuracy": 0.18069732934236526,
      "num_tokens": 34741628.0,
      "step": 15850
    },
    {
      "entropy": 5.529052305221557,
      "epoch": 1.5735410877332274,
      "grad_norm": 0.80859375,
      "learning_rate": 0.00047582435348539695,
      "loss": 5.1417,
      "mean_token_accuracy": 0.17226557284593583,
      "num_tokens": 34752984.0,
      "step": 15855
    },
    {
      "entropy": 5.575853395462036,
      "epoch": 1.574037316395395,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004758083734055875,
      "loss": 5.2048,
      "mean_token_accuracy": 0.16819843500852585,
      "num_tokens": 34764363.0,
      "step": 15860
    },
    {
      "entropy": 5.587705516815186,
      "epoch": 1.5745335450575624,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004757923883461632,
      "loss": 5.1445,
      "mean_token_accuracy": 0.17044635564088823,
      "num_tokens": 34776289.0,
      "step": 15865
    },
    {
      "entropy": 5.62997145652771,
      "epoch": 1.57502977371973,
      "grad_norm": 1.0,
      "learning_rate": 0.00047577639830752034,
      "loss": 5.1264,
      "mean_token_accuracy": 0.17434465736150742,
      "num_tokens": 34786760.0,
      "step": 15870
    },
    {
      "entropy": 5.5833601474761965,
      "epoch": 1.5755260023818976,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047576040329005536,
      "loss": 5.0992,
      "mean_token_accuracy": 0.1728442579507828,
      "num_tokens": 34797737.0,
      "step": 15875
    },
    {
      "entropy": 5.558037137985229,
      "epoch": 1.5760222310440652,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000475744403294165,
      "loss": 5.1482,
      "mean_token_accuracy": 0.17665326297283174,
      "num_tokens": 34809650.0,
      "step": 15880
    },
    {
      "entropy": 5.6620673656463625,
      "epoch": 1.5765184597062327,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047572839832024606,
      "loss": 5.1687,
      "mean_token_accuracy": 0.1658117175102234,
      "num_tokens": 34820171.0,
      "step": 15885
    },
    {
      "entropy": 5.536234951019287,
      "epoch": 1.5770146883684002,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047571238836869537,
      "loss": 5.0375,
      "mean_token_accuracy": 0.17794111222028733,
      "num_tokens": 34831286.0,
      "step": 15890
    },
    {
      "entropy": 5.495041465759277,
      "epoch": 1.5775109170305677,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004756963734399099,
      "loss": 5.048,
      "mean_token_accuracy": 0.17999424487352372,
      "num_tokens": 34841767.0,
      "step": 15895
    },
    {
      "entropy": 5.561532640457154,
      "epoch": 1.5780071456927351,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047568035353428683,
      "loss": 5.1489,
      "mean_token_accuracy": 0.17454658597707748,
      "num_tokens": 34852644.0,
      "step": 15900
    },
    {
      "entropy": 5.636040258407593,
      "epoch": 1.5785033743549026,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047566432865222343,
      "loss": 5.1654,
      "mean_token_accuracy": 0.17310801446437835,
      "num_tokens": 34863605.0,
      "step": 15905
    },
    {
      "entropy": 5.540699052810669,
      "epoch": 1.5789996030170703,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047564829879411706,
      "loss": 5.1014,
      "mean_token_accuracy": 0.17733003795146943,
      "num_tokens": 34874594.0,
      "step": 15910
    },
    {
      "entropy": 5.581782865524292,
      "epoch": 1.5794958316792378,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004756322639603653,
      "loss": 5.1983,
      "mean_token_accuracy": 0.17092593312263488,
      "num_tokens": 34884627.0,
      "step": 15915
    },
    {
      "entropy": 5.627491903305054,
      "epoch": 1.5799920603414053,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047561622415136565,
      "loss": 5.1566,
      "mean_token_accuracy": 0.17587376534938812,
      "num_tokens": 34895679.0,
      "step": 15920
    },
    {
      "entropy": 5.542557621002198,
      "epoch": 1.580488289003573,
      "grad_norm": 1.0,
      "learning_rate": 0.00047560017936751597,
      "loss": 5.0914,
      "mean_token_accuracy": 0.1790750116109848,
      "num_tokens": 34906811.0,
      "step": 15925
    },
    {
      "entropy": 5.576779127120972,
      "epoch": 1.5809845176657404,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004755841296092141,
      "loss": 5.1271,
      "mean_token_accuracy": 0.16672847419977188,
      "num_tokens": 34918655.0,
      "step": 15930
    },
    {
      "entropy": 5.566650342941284,
      "epoch": 1.581480746327908,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047556807487685797,
      "loss": 5.1247,
      "mean_token_accuracy": 0.17290757447481156,
      "num_tokens": 34931034.0,
      "step": 15935
    },
    {
      "entropy": 5.5927551746368405,
      "epoch": 1.5819769749900754,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004755520151708458,
      "loss": 5.0654,
      "mean_token_accuracy": 0.17453733533620835,
      "num_tokens": 34939707.0,
      "step": 15940
    },
    {
      "entropy": 5.554121732711792,
      "epoch": 1.5824732036522429,
      "grad_norm": 1.0,
      "learning_rate": 0.0004755359504915758,
      "loss": 5.1776,
      "mean_token_accuracy": 0.16669235676527022,
      "num_tokens": 34949237.0,
      "step": 15945
    },
    {
      "entropy": 5.634067249298096,
      "epoch": 1.5829694323144103,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004755198808394464,
      "loss": 5.1951,
      "mean_token_accuracy": 0.17249667793512344,
      "num_tokens": 34960380.0,
      "step": 15950
    },
    {
      "entropy": 5.582531976699829,
      "epoch": 1.583465660976578,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000475503806214856,
      "loss": 5.1662,
      "mean_token_accuracy": 0.16854514181613922,
      "num_tokens": 34971699.0,
      "step": 15955
    },
    {
      "entropy": 5.571568346023559,
      "epoch": 1.5839618896387455,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004754877266182033,
      "loss": 5.1298,
      "mean_token_accuracy": 0.17670858055353164,
      "num_tokens": 34982333.0,
      "step": 15960
    },
    {
      "entropy": 5.5827538013458256,
      "epoch": 1.5844581183009132,
      "grad_norm": 1.015625,
      "learning_rate": 0.000475471642049887,
      "loss": 5.1575,
      "mean_token_accuracy": 0.17220501005649566,
      "num_tokens": 34993277.0,
      "step": 15965
    },
    {
      "entropy": 5.571511173248291,
      "epoch": 1.5849543469630807,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004754555525103058,
      "loss": 5.1032,
      "mean_token_accuracy": 0.17402896881103516,
      "num_tokens": 35004252.0,
      "step": 15970
    },
    {
      "entropy": 5.533213329315186,
      "epoch": 1.5854505756252482,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000475439457999859,
      "loss": 5.0856,
      "mean_token_accuracy": 0.17765643298625947,
      "num_tokens": 35016974.0,
      "step": 15975
    },
    {
      "entropy": 5.605645370483399,
      "epoch": 1.5859468042874156,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047542335851894553,
      "loss": 5.1738,
      "mean_token_accuracy": 0.17135521918535232,
      "num_tokens": 35029017.0,
      "step": 15980
    },
    {
      "entropy": 5.6136940002441404,
      "epoch": 1.586443032949583,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047540725406796454,
      "loss": 5.1889,
      "mean_token_accuracy": 0.17290778160095216,
      "num_tokens": 35040239.0,
      "step": 15985
    },
    {
      "entropy": 5.617549562454224,
      "epoch": 1.5869392616117506,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004753911446473156,
      "loss": 5.1088,
      "mean_token_accuracy": 0.1707977443933487,
      "num_tokens": 35051426.0,
      "step": 15990
    },
    {
      "entropy": 5.506148004531861,
      "epoch": 1.587435490273918,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000475375030257398,
      "loss": 5.0854,
      "mean_token_accuracy": 0.17198949456214904,
      "num_tokens": 35061777.0,
      "step": 15995
    },
    {
      "entropy": 5.557938432693481,
      "epoch": 1.5879317189360858,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004753589108986114,
      "loss": 5.0472,
      "mean_token_accuracy": 0.18720826208591462,
      "num_tokens": 35071822.0,
      "step": 16000
    },
    {
      "entropy": 5.594207620620727,
      "epoch": 1.5884279475982532,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047534278657135565,
      "loss": 5.14,
      "mean_token_accuracy": 0.17440089583396912,
      "num_tokens": 35081900.0,
      "step": 16005
    },
    {
      "entropy": 5.602560949325562,
      "epoch": 1.588924176260421,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004753266572760304,
      "loss": 5.1832,
      "mean_token_accuracy": 0.16803516447544098,
      "num_tokens": 35092866.0,
      "step": 16010
    },
    {
      "entropy": 5.7396979331970215,
      "epoch": 1.5894204049225884,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047531052301303577,
      "loss": 5.2614,
      "mean_token_accuracy": 0.16951098591089248,
      "num_tokens": 35103425.0,
      "step": 16015
    },
    {
      "entropy": 5.522967052459717,
      "epoch": 1.5899166335847559,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004752943837827718,
      "loss": 5.0579,
      "mean_token_accuracy": 0.17411424666643144,
      "num_tokens": 35114620.0,
      "step": 16020
    },
    {
      "entropy": 5.61639175415039,
      "epoch": 1.5904128622469234,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004752782395856387,
      "loss": 5.1524,
      "mean_token_accuracy": 0.1711709275841713,
      "num_tokens": 35125199.0,
      "step": 16025
    },
    {
      "entropy": 5.564209794998169,
      "epoch": 1.5909090909090908,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004752620904220368,
      "loss": 5.1177,
      "mean_token_accuracy": 0.17136414498090743,
      "num_tokens": 35135892.0,
      "step": 16030
    },
    {
      "entropy": 5.573184776306152,
      "epoch": 1.5914053195712583,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004752459362923666,
      "loss": 5.1392,
      "mean_token_accuracy": 0.1697489380836487,
      "num_tokens": 35146186.0,
      "step": 16035
    },
    {
      "entropy": 5.56048436164856,
      "epoch": 1.591901548233426,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004752297771970287,
      "loss": 5.1827,
      "mean_token_accuracy": 0.16847035586833953,
      "num_tokens": 35157261.0,
      "step": 16040
    },
    {
      "entropy": 5.600918340682983,
      "epoch": 1.5923977768955935,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047521361313642373,
      "loss": 5.145,
      "mean_token_accuracy": 0.17414494454860688,
      "num_tokens": 35168508.0,
      "step": 16045
    },
    {
      "entropy": 5.5400285720825195,
      "epoch": 1.5928940055577612,
      "grad_norm": 0.875,
      "learning_rate": 0.0004751974441109526,
      "loss": 5.0965,
      "mean_token_accuracy": 0.16977155655622483,
      "num_tokens": 35179444.0,
      "step": 16050
    },
    {
      "entropy": 5.523671722412109,
      "epoch": 1.5933902342199286,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047518127012101624,
      "loss": 5.0962,
      "mean_token_accuracy": 0.17849835604429246,
      "num_tokens": 35189517.0,
      "step": 16055
    },
    {
      "entropy": 5.562128925323487,
      "epoch": 1.5938864628820961,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004751650911670157,
      "loss": 5.1309,
      "mean_token_accuracy": 0.1715974047780037,
      "num_tokens": 35200191.0,
      "step": 16060
    },
    {
      "entropy": 5.628333711624146,
      "epoch": 1.5943826915442636,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047514890724935236,
      "loss": 5.1551,
      "mean_token_accuracy": 0.1664279356598854,
      "num_tokens": 35211260.0,
      "step": 16065
    },
    {
      "entropy": 5.53777961730957,
      "epoch": 1.594878920206431,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004751327183684272,
      "loss": 5.1336,
      "mean_token_accuracy": 0.1674045130610466,
      "num_tokens": 35222408.0,
      "step": 16070
    },
    {
      "entropy": 5.570901346206665,
      "epoch": 1.5953751488685985,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047511652452464196,
      "loss": 5.1842,
      "mean_token_accuracy": 0.1745487704873085,
      "num_tokens": 35232760.0,
      "step": 16075
    },
    {
      "entropy": 5.587699460983276,
      "epoch": 1.595871377530766,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047510032571839816,
      "loss": 5.1416,
      "mean_token_accuracy": 0.17014646977186204,
      "num_tokens": 35243849.0,
      "step": 16080
    },
    {
      "entropy": 5.6216460227966305,
      "epoch": 1.5963676061929337,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004750841219500975,
      "loss": 5.1533,
      "mean_token_accuracy": 0.1730295717716217,
      "num_tokens": 35254962.0,
      "step": 16085
    },
    {
      "entropy": 5.588691711425781,
      "epoch": 1.5968638348551012,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004750679132201416,
      "loss": 5.218,
      "mean_token_accuracy": 0.16684424132108688,
      "num_tokens": 35265895.0,
      "step": 16090
    },
    {
      "entropy": 5.684676742553711,
      "epoch": 1.597360063517269,
      "grad_norm": 1.0,
      "learning_rate": 0.00047505169952893276,
      "loss": 5.2252,
      "mean_token_accuracy": 0.16927957087755202,
      "num_tokens": 35277543.0,
      "step": 16095
    },
    {
      "entropy": 5.609101343154907,
      "epoch": 1.5978562921794364,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047503548087687267,
      "loss": 5.1114,
      "mean_token_accuracy": 0.17922525554895402,
      "num_tokens": 35289130.0,
      "step": 16100
    },
    {
      "entropy": 5.520952415466309,
      "epoch": 1.5983525208416038,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047501925726436376,
      "loss": 5.066,
      "mean_token_accuracy": 0.17004935294389725,
      "num_tokens": 35298414.0,
      "step": 16105
    },
    {
      "entropy": 5.530217695236206,
      "epoch": 1.5988487495037713,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004750030286918082,
      "loss": 5.1327,
      "mean_token_accuracy": 0.1705824002623558,
      "num_tokens": 35310629.0,
      "step": 16110
    },
    {
      "entropy": 5.637357521057129,
      "epoch": 1.5993449781659388,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004749867951596086,
      "loss": 5.1578,
      "mean_token_accuracy": 0.17160857915878297,
      "num_tokens": 35321810.0,
      "step": 16115
    },
    {
      "entropy": 5.608400344848633,
      "epoch": 1.5998412068281063,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047497055666816736,
      "loss": 5.0862,
      "mean_token_accuracy": 0.177343513071537,
      "num_tokens": 35333517.0,
      "step": 16120
    },
    {
      "entropy": 5.52254843711853,
      "epoch": 1.600337435490274,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004749543132178873,
      "loss": 5.0201,
      "mean_token_accuracy": 0.18037809431552887,
      "num_tokens": 35344170.0,
      "step": 16125
    },
    {
      "entropy": 5.574146795272827,
      "epoch": 1.6008336641524414,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000474938064809171,
      "loss": 5.2097,
      "mean_token_accuracy": 0.17083947807550431,
      "num_tokens": 35354963.0,
      "step": 16130
    },
    {
      "entropy": 5.618002271652221,
      "epoch": 1.6013298928146091,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004749218114424216,
      "loss": 5.1696,
      "mean_token_accuracy": 0.1733885020017624,
      "num_tokens": 35365716.0,
      "step": 16135
    },
    {
      "entropy": 5.5617574691772464,
      "epoch": 1.6018261214767766,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047490555311804206,
      "loss": 5.0998,
      "mean_token_accuracy": 0.17365415245294571,
      "num_tokens": 35376113.0,
      "step": 16140
    },
    {
      "entropy": 5.520939302444458,
      "epoch": 1.602322350138944,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047488928983643565,
      "loss": 5.0645,
      "mean_token_accuracy": 0.17254266440868377,
      "num_tokens": 35386178.0,
      "step": 16145
    },
    {
      "entropy": 5.5097838878631595,
      "epoch": 1.6028185788011116,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047487302159800545,
      "loss": 4.9951,
      "mean_token_accuracy": 0.18097468167543412,
      "num_tokens": 35396874.0,
      "step": 16150
    },
    {
      "entropy": 5.577455711364746,
      "epoch": 1.603314807463279,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000474856748403155,
      "loss": 5.2103,
      "mean_token_accuracy": 0.16899162381887436,
      "num_tokens": 35409405.0,
      "step": 16155
    },
    {
      "entropy": 5.622749853134155,
      "epoch": 1.6038110361254465,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004748404702522879,
      "loss": 5.1522,
      "mean_token_accuracy": 0.17081642150878906,
      "num_tokens": 35421449.0,
      "step": 16160
    },
    {
      "entropy": 5.592343187332153,
      "epoch": 1.604307264787614,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047482418714580774,
      "loss": 5.1967,
      "mean_token_accuracy": 0.16900321394205092,
      "num_tokens": 35434612.0,
      "step": 16165
    },
    {
      "entropy": 5.532162523269653,
      "epoch": 1.6048034934497817,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004748078990841183,
      "loss": 5.0898,
      "mean_token_accuracy": 0.17857665866613387,
      "num_tokens": 35446016.0,
      "step": 16170
    },
    {
      "entropy": 5.631016492843628,
      "epoch": 1.6052997221119492,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004747916060676235,
      "loss": 5.1556,
      "mean_token_accuracy": 0.1717934399843216,
      "num_tokens": 35457405.0,
      "step": 16175
    },
    {
      "entropy": 5.590787601470947,
      "epoch": 1.6057959507741169,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047477530809672747,
      "loss": 5.1935,
      "mean_token_accuracy": 0.16721851825714112,
      "num_tokens": 35468361.0,
      "step": 16180
    },
    {
      "entropy": 5.545227575302124,
      "epoch": 1.6062921794362843,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004747590051718342,
      "loss": 5.0791,
      "mean_token_accuracy": 0.17175506204366683,
      "num_tokens": 35480092.0,
      "step": 16185
    },
    {
      "entropy": 5.576374530792236,
      "epoch": 1.6067884080984518,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047474269729334806,
      "loss": 5.2066,
      "mean_token_accuracy": 0.16445457339286804,
      "num_tokens": 35492131.0,
      "step": 16190
    },
    {
      "entropy": 5.626881551742554,
      "epoch": 1.6072846367606193,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047472638446167346,
      "loss": 5.1595,
      "mean_token_accuracy": 0.17052288353443146,
      "num_tokens": 35503600.0,
      "step": 16195
    },
    {
      "entropy": 5.636068630218506,
      "epoch": 1.6077808654227868,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047471006667721484,
      "loss": 5.2039,
      "mean_token_accuracy": 0.16823625564575195,
      "num_tokens": 35514716.0,
      "step": 16200
    },
    {
      "entropy": 5.587717914581299,
      "epoch": 1.6082770940849542,
      "grad_norm": 0.890625,
      "learning_rate": 0.000474693743940377,
      "loss": 5.1173,
      "mean_token_accuracy": 0.18077134788036348,
      "num_tokens": 35525688.0,
      "step": 16205
    },
    {
      "entropy": 5.583987998962402,
      "epoch": 1.608773322747122,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047467741625156453,
      "loss": 5.1286,
      "mean_token_accuracy": 0.17703331708908082,
      "num_tokens": 35536391.0,
      "step": 16210
    },
    {
      "entropy": 5.534589099884033,
      "epoch": 1.6092695514092894,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047466108361118235,
      "loss": 5.0611,
      "mean_token_accuracy": 0.1796290844678879,
      "num_tokens": 35547016.0,
      "step": 16215
    },
    {
      "entropy": 5.568278503417969,
      "epoch": 1.6097657800714569,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047464474601963564,
      "loss": 5.1397,
      "mean_token_accuracy": 0.16884369999170304,
      "num_tokens": 35558402.0,
      "step": 16220
    },
    {
      "entropy": 5.554005813598633,
      "epoch": 1.6102620087336246,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004746284034773294,
      "loss": 5.1328,
      "mean_token_accuracy": 0.17020624577999116,
      "num_tokens": 35569779.0,
      "step": 16225
    },
    {
      "entropy": 5.613527536392212,
      "epoch": 1.610758237395792,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004746120559846689,
      "loss": 5.1503,
      "mean_token_accuracy": 0.1753134772181511,
      "num_tokens": 35580044.0,
      "step": 16230
    },
    {
      "entropy": 5.6223033428192135,
      "epoch": 1.6112544660579595,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047459570354205954,
      "loss": 5.2295,
      "mean_token_accuracy": 0.17122981548309327,
      "num_tokens": 35591216.0,
      "step": 16235
    },
    {
      "entropy": 5.516415309906006,
      "epoch": 1.611750694720127,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047457934614990684,
      "loss": 5.0467,
      "mean_token_accuracy": 0.17445323169231414,
      "num_tokens": 35602376.0,
      "step": 16240
    },
    {
      "entropy": 5.549879741668701,
      "epoch": 1.6122469233822945,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004745629838086164,
      "loss": 5.033,
      "mean_token_accuracy": 0.18038317561149597,
      "num_tokens": 35612184.0,
      "step": 16245
    },
    {
      "entropy": 5.583459377288818,
      "epoch": 1.612743152044462,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047454661651859394,
      "loss": 5.2002,
      "mean_token_accuracy": 0.17160967141389846,
      "num_tokens": 35624864.0,
      "step": 16250
    },
    {
      "entropy": 5.600981950759888,
      "epoch": 1.6132393807066296,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004745302442802454,
      "loss": 5.1252,
      "mean_token_accuracy": 0.17504289001226425,
      "num_tokens": 35636492.0,
      "step": 16255
    },
    {
      "entropy": 5.528806924819946,
      "epoch": 1.6137356093687971,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004745138670939768,
      "loss": 5.0666,
      "mean_token_accuracy": 0.178764508664608,
      "num_tokens": 35646831.0,
      "step": 16260
    },
    {
      "entropy": 5.598997449874878,
      "epoch": 1.6142318380309648,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004744974849601941,
      "loss": 5.1859,
      "mean_token_accuracy": 0.17179030179977417,
      "num_tokens": 35657703.0,
      "step": 16265
    },
    {
      "entropy": 5.492500257492066,
      "epoch": 1.6147280666931323,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004744810978793038,
      "loss": 5.0024,
      "mean_token_accuracy": 0.18277074545621871,
      "num_tokens": 35668334.0,
      "step": 16270
    },
    {
      "entropy": 5.552626848220825,
      "epoch": 1.6152242953552998,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047446470585171206,
      "loss": 5.1598,
      "mean_token_accuracy": 0.17104385495185853,
      "num_tokens": 35679796.0,
      "step": 16275
    },
    {
      "entropy": 5.499025058746338,
      "epoch": 1.6157205240174672,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047444830887782537,
      "loss": 5.0518,
      "mean_token_accuracy": 0.17568161934614182,
      "num_tokens": 35690825.0,
      "step": 16280
    },
    {
      "entropy": 5.6292157649993895,
      "epoch": 1.6162167526796347,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004744319069580504,
      "loss": 5.1519,
      "mean_token_accuracy": 0.166786290705204,
      "num_tokens": 35701824.0,
      "step": 16285
    },
    {
      "entropy": 5.647288084030151,
      "epoch": 1.6167129813418022,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047441550009279395,
      "loss": 5.0955,
      "mean_token_accuracy": 0.17449122965335845,
      "num_tokens": 35713250.0,
      "step": 16290
    },
    {
      "entropy": 5.572403907775879,
      "epoch": 1.6172092100039697,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004743990882824627,
      "loss": 5.1573,
      "mean_token_accuracy": 0.17184995859861374,
      "num_tokens": 35725049.0,
      "step": 16295
    },
    {
      "entropy": 5.50085301399231,
      "epoch": 1.6177054386661374,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047438267152746376,
      "loss": 5.0977,
      "mean_token_accuracy": 0.1714867725968361,
      "num_tokens": 35734676.0,
      "step": 16300
    },
    {
      "entropy": 5.602110958099365,
      "epoch": 1.6182016673283048,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047436624982820423,
      "loss": 5.1,
      "mean_token_accuracy": 0.17056845128536224,
      "num_tokens": 35743953.0,
      "step": 16305
    },
    {
      "entropy": 5.563694858551026,
      "epoch": 1.6186978959904725,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004743498231850913,
      "loss": 5.0451,
      "mean_token_accuracy": 0.18186824321746825,
      "num_tokens": 35753149.0,
      "step": 16310
    },
    {
      "entropy": 5.528791046142578,
      "epoch": 1.61919412465264,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004743333915985323,
      "loss": 5.1073,
      "mean_token_accuracy": 0.17305390685796737,
      "num_tokens": 35763783.0,
      "step": 16315
    },
    {
      "entropy": 5.533765077590942,
      "epoch": 1.6196903533148075,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047431695506893476,
      "loss": 5.0605,
      "mean_token_accuracy": 0.17982203364372254,
      "num_tokens": 35774663.0,
      "step": 16320
    },
    {
      "entropy": 5.571882152557373,
      "epoch": 1.620186581976975,
      "grad_norm": 1.0,
      "learning_rate": 0.00047430051359670616,
      "loss": 5.2165,
      "mean_token_accuracy": 0.1683632642030716,
      "num_tokens": 35786990.0,
      "step": 16325
    },
    {
      "entropy": 5.652010726928711,
      "epoch": 1.6206828106391424,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004742840671822542,
      "loss": 5.2063,
      "mean_token_accuracy": 0.17310591489076615,
      "num_tokens": 35798818.0,
      "step": 16330
    },
    {
      "entropy": 5.62846417427063,
      "epoch": 1.62117903930131,
      "grad_norm": 1.0,
      "learning_rate": 0.000474267615825987,
      "loss": 5.1607,
      "mean_token_accuracy": 0.16817210763692855,
      "num_tokens": 35808813.0,
      "step": 16335
    },
    {
      "entropy": 5.576836442947387,
      "epoch": 1.6216752679634776,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047425115952831213,
      "loss": 5.1409,
      "mean_token_accuracy": 0.17503860145807265,
      "num_tokens": 35818664.0,
      "step": 16340
    },
    {
      "entropy": 5.490856122970581,
      "epoch": 1.622171496625645,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000474234698289638,
      "loss": 5.0686,
      "mean_token_accuracy": 0.1828345015645027,
      "num_tokens": 35828975.0,
      "step": 16345
    },
    {
      "entropy": 5.50765905380249,
      "epoch": 1.6226677252878128,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004742182321103725,
      "loss": 5.1016,
      "mean_token_accuracy": 0.17467779219150542,
      "num_tokens": 35839437.0,
      "step": 16350
    },
    {
      "entropy": 5.559318494796753,
      "epoch": 1.6231639539499803,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004742017609909242,
      "loss": 5.0534,
      "mean_token_accuracy": 0.17858787178993224,
      "num_tokens": 35850413.0,
      "step": 16355
    },
    {
      "entropy": 5.604243230819702,
      "epoch": 1.6236601826121477,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047418528493170157,
      "loss": 5.1356,
      "mean_token_accuracy": 0.16888745725154877,
      "num_tokens": 35862745.0,
      "step": 16360
    },
    {
      "entropy": 5.622930431365967,
      "epoch": 1.6241564112743152,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004741688039331129,
      "loss": 5.1082,
      "mean_token_accuracy": 0.17218896001577377,
      "num_tokens": 35873934.0,
      "step": 16365
    },
    {
      "entropy": 5.524751281738281,
      "epoch": 1.6246526399364827,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047415231799556725,
      "loss": 5.0428,
      "mean_token_accuracy": 0.17814824730157852,
      "num_tokens": 35884279.0,
      "step": 16370
    },
    {
      "entropy": 5.5249066829681395,
      "epoch": 1.6251488685986502,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047413582711947305,
      "loss": 5.1023,
      "mean_token_accuracy": 0.17733576744794846,
      "num_tokens": 35894144.0,
      "step": 16375
    },
    {
      "entropy": 5.552412033081055,
      "epoch": 1.6256450972608176,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047411933130523954,
      "loss": 5.0946,
      "mean_token_accuracy": 0.1717037722468376,
      "num_tokens": 35903641.0,
      "step": 16380
    },
    {
      "entropy": 5.522723865509033,
      "epoch": 1.6261413259229853,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004741028305532757,
      "loss": 5.0639,
      "mean_token_accuracy": 0.1720563530921936,
      "num_tokens": 35913634.0,
      "step": 16385
    },
    {
      "entropy": 5.563013887405395,
      "epoch": 1.6266375545851528,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004740863248639907,
      "loss": 5.0989,
      "mean_token_accuracy": 0.17695355117321016,
      "num_tokens": 35923106.0,
      "step": 16390
    },
    {
      "entropy": 5.5690265655517575,
      "epoch": 1.6271337832473205,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047406981423779376,
      "loss": 5.096,
      "mean_token_accuracy": 0.17420815527439118,
      "num_tokens": 35933709.0,
      "step": 16395
    },
    {
      "entropy": 5.56211838722229,
      "epoch": 1.627630011909488,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047405329867509447,
      "loss": 5.1235,
      "mean_token_accuracy": 0.16502735912799835,
      "num_tokens": 35944734.0,
      "step": 16400
    },
    {
      "entropy": 5.537131118774414,
      "epoch": 1.6281262405716554,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047403677817630226,
      "loss": 5.0866,
      "mean_token_accuracy": 0.17275551855564117,
      "num_tokens": 35956513.0,
      "step": 16405
    },
    {
      "entropy": 5.52769923210144,
      "epoch": 1.628622469233823,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004740202527418269,
      "loss": 5.1052,
      "mean_token_accuracy": 0.1734310656785965,
      "num_tokens": 35967096.0,
      "step": 16410
    },
    {
      "entropy": 5.57195897102356,
      "epoch": 1.6291186978959904,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000474003722372078,
      "loss": 5.17,
      "mean_token_accuracy": 0.17220955342054367,
      "num_tokens": 35979010.0,
      "step": 16415
    },
    {
      "entropy": 5.510465049743653,
      "epoch": 1.6296149265581579,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047398718706746556,
      "loss": 5.0296,
      "mean_token_accuracy": 0.1757589876651764,
      "num_tokens": 35989253.0,
      "step": 16420
    },
    {
      "entropy": 5.546857833862305,
      "epoch": 1.6301111552203256,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004739706468283998,
      "loss": 5.0466,
      "mean_token_accuracy": 0.17714656442403792,
      "num_tokens": 35999659.0,
      "step": 16425
    },
    {
      "entropy": 5.517600107192993,
      "epoch": 1.630607383882493,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047395410165529075,
      "loss": 5.1023,
      "mean_token_accuracy": 0.17066733688116073,
      "num_tokens": 36010295.0,
      "step": 16430
    },
    {
      "entropy": 5.564552116394043,
      "epoch": 1.6311036125446607,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047393755154854855,
      "loss": 5.1284,
      "mean_token_accuracy": 0.16993907392024993,
      "num_tokens": 36021575.0,
      "step": 16435
    },
    {
      "entropy": 5.5968342304229735,
      "epoch": 1.6315998412068282,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004739209965085839,
      "loss": 5.1246,
      "mean_token_accuracy": 0.16845714300870895,
      "num_tokens": 36032764.0,
      "step": 16440
    },
    {
      "entropy": 5.548697853088379,
      "epoch": 1.6320960698689957,
      "grad_norm": 1.0625,
      "learning_rate": 0.000473904436535807,
      "loss": 5.0882,
      "mean_token_accuracy": 0.18048504441976548,
      "num_tokens": 36041864.0,
      "step": 16445
    },
    {
      "entropy": 5.484388113021851,
      "epoch": 1.6325922985311632,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004738878716306288,
      "loss": 5.0742,
      "mean_token_accuracy": 0.17158215790987014,
      "num_tokens": 36051321.0,
      "step": 16450
    },
    {
      "entropy": 5.529311943054199,
      "epoch": 1.6330885271933306,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047387130179345996,
      "loss": 5.0549,
      "mean_token_accuracy": 0.17032379060983657,
      "num_tokens": 36061448.0,
      "step": 16455
    },
    {
      "entropy": 5.592094659805298,
      "epoch": 1.6335847558554981,
      "grad_norm": 1.0,
      "learning_rate": 0.00047385472702471133,
      "loss": 5.174,
      "mean_token_accuracy": 0.17453288584947585,
      "num_tokens": 36071832.0,
      "step": 16460
    },
    {
      "entropy": 5.5310187339782715,
      "epoch": 1.6340809845176656,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000473838147324794,
      "loss": 5.0915,
      "mean_token_accuracy": 0.17437421083450316,
      "num_tokens": 36082103.0,
      "step": 16465
    },
    {
      "entropy": 5.567050981521606,
      "epoch": 1.6345772131798333,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000473821562694119,
      "loss": 5.1412,
      "mean_token_accuracy": 0.1688063144683838,
      "num_tokens": 36092063.0,
      "step": 16470
    },
    {
      "entropy": 5.562640571594239,
      "epoch": 1.6350734418420008,
      "grad_norm": 1.0,
      "learning_rate": 0.0004738049731330977,
      "loss": 5.12,
      "mean_token_accuracy": 0.16789728552103042,
      "num_tokens": 36102573.0,
      "step": 16475
    },
    {
      "entropy": 5.614202451705933,
      "epoch": 1.6355696705041685,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004737883786421414,
      "loss": 5.0973,
      "mean_token_accuracy": 0.17387198954820632,
      "num_tokens": 36112677.0,
      "step": 16480
    },
    {
      "entropy": 5.576464748382568,
      "epoch": 1.636065899166336,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047377177922166175,
      "loss": 5.1774,
      "mean_token_accuracy": 0.16393067687749863,
      "num_tokens": 36123337.0,
      "step": 16485
    },
    {
      "entropy": 5.6107264995574955,
      "epoch": 1.6365621278285034,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047375517487207024,
      "loss": 5.1445,
      "mean_token_accuracy": 0.1689080849289894,
      "num_tokens": 36134503.0,
      "step": 16490
    },
    {
      "entropy": 5.661848306655884,
      "epoch": 1.6370583564906709,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047373856559377876,
      "loss": 5.2042,
      "mean_token_accuracy": 0.16746170371770858,
      "num_tokens": 36145766.0,
      "step": 16495
    },
    {
      "entropy": 5.580369186401367,
      "epoch": 1.6375545851528384,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004737219513871989,
      "loss": 5.1465,
      "mean_token_accuracy": 0.17452765703201295,
      "num_tokens": 36156584.0,
      "step": 16500
    },
    {
      "entropy": 5.550521469116211,
      "epoch": 1.6380508138150058,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047370533225274305,
      "loss": 5.1263,
      "mean_token_accuracy": 0.1735871210694313,
      "num_tokens": 36166999.0,
      "step": 16505
    },
    {
      "entropy": 5.5824799060821535,
      "epoch": 1.6385470424771735,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004736887081908231,
      "loss": 5.1047,
      "mean_token_accuracy": 0.17466113567352295,
      "num_tokens": 36178724.0,
      "step": 16510
    },
    {
      "entropy": 5.661315774917602,
      "epoch": 1.639043271139341,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004736720792018513,
      "loss": 5.197,
      "mean_token_accuracy": 0.16900236904621124,
      "num_tokens": 36188912.0,
      "step": 16515
    },
    {
      "entropy": 5.539143085479736,
      "epoch": 1.6395394998015085,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004736554452862399,
      "loss": 5.0925,
      "mean_token_accuracy": 0.17414916306734085,
      "num_tokens": 36198525.0,
      "step": 16520
    },
    {
      "entropy": 5.533590269088745,
      "epoch": 1.6400357284636762,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047363880644440173,
      "loss": 5.0592,
      "mean_token_accuracy": 0.17656663358211516,
      "num_tokens": 36209384.0,
      "step": 16525
    },
    {
      "entropy": 5.6189697265625,
      "epoch": 1.6405319571258437,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047362216267674906,
      "loss": 5.1443,
      "mean_token_accuracy": 0.16970728039741517,
      "num_tokens": 36220597.0,
      "step": 16530
    },
    {
      "entropy": 5.628946590423584,
      "epoch": 1.6410281857880111,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004736055139836948,
      "loss": 5.1954,
      "mean_token_accuracy": 0.16823187172412873,
      "num_tokens": 36230317.0,
      "step": 16535
    },
    {
      "entropy": 5.686033439636231,
      "epoch": 1.6415244144501786,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047358886036565176,
      "loss": 5.2847,
      "mean_token_accuracy": 0.16957342475652695,
      "num_tokens": 36241645.0,
      "step": 16540
    },
    {
      "entropy": 5.618369483947754,
      "epoch": 1.642020643112346,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047357220182303285,
      "loss": 5.2052,
      "mean_token_accuracy": 0.16547121107578278,
      "num_tokens": 36253098.0,
      "step": 16545
    },
    {
      "entropy": 5.524405574798584,
      "epoch": 1.6425168717745136,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004735555383562513,
      "loss": 5.0948,
      "mean_token_accuracy": 0.17753568887710572,
      "num_tokens": 36264417.0,
      "step": 16550
    },
    {
      "entropy": 5.652443218231201,
      "epoch": 1.6430131004366813,
      "grad_norm": 0.81640625,
      "learning_rate": 0.00047353886996572014,
      "loss": 5.1984,
      "mean_token_accuracy": 0.16566038578748704,
      "num_tokens": 36276589.0,
      "step": 16555
    },
    {
      "entropy": 5.550014877319336,
      "epoch": 1.6435093290988487,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047352219665185293,
      "loss": 5.1076,
      "mean_token_accuracy": 0.1702713429927826,
      "num_tokens": 36287308.0,
      "step": 16560
    },
    {
      "entropy": 5.5406848907470705,
      "epoch": 1.6440055577610164,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004735055184150629,
      "loss": 5.1256,
      "mean_token_accuracy": 0.18044352829456328,
      "num_tokens": 36298966.0,
      "step": 16565
    },
    {
      "entropy": 5.566959810256958,
      "epoch": 1.644501786423184,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047348883525576385,
      "loss": 5.1174,
      "mean_token_accuracy": 0.16923878341913223,
      "num_tokens": 36310498.0,
      "step": 16570
    },
    {
      "entropy": 5.604967498779297,
      "epoch": 1.6449980150853514,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004734721471743694,
      "loss": 5.1767,
      "mean_token_accuracy": 0.1748863011598587,
      "num_tokens": 36321589.0,
      "step": 16575
    },
    {
      "entropy": 5.548197507858276,
      "epoch": 1.6454942437475188,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004734554541712933,
      "loss": 5.0999,
      "mean_token_accuracy": 0.17060055136680602,
      "num_tokens": 36333041.0,
      "step": 16580
    },
    {
      "entropy": 5.599398422241211,
      "epoch": 1.6459904724096863,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004734387562469496,
      "loss": 5.1766,
      "mean_token_accuracy": 0.1683475062251091,
      "num_tokens": 36344225.0,
      "step": 16585
    },
    {
      "entropy": 5.593447923660278,
      "epoch": 1.6464867010718538,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047342205340175235,
      "loss": 5.054,
      "mean_token_accuracy": 0.1773373767733574,
      "num_tokens": 36355524.0,
      "step": 16590
    },
    {
      "entropy": 5.60854139328003,
      "epoch": 1.6469829297340213,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047340534563611574,
      "loss": 5.134,
      "mean_token_accuracy": 0.17127421796321868,
      "num_tokens": 36366495.0,
      "step": 16595
    },
    {
      "entropy": 5.584811162948609,
      "epoch": 1.647479158396189,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047338863295045414,
      "loss": 5.1276,
      "mean_token_accuracy": 0.17236628979444504,
      "num_tokens": 36378316.0,
      "step": 16600
    },
    {
      "entropy": 5.61781873703003,
      "epoch": 1.6479753870583564,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047337191534518186,
      "loss": 5.2174,
      "mean_token_accuracy": 0.16255452185869218,
      "num_tokens": 36390500.0,
      "step": 16605
    },
    {
      "entropy": 5.609574317932129,
      "epoch": 1.6484716157205241,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004733551928207135,
      "loss": 5.1509,
      "mean_token_accuracy": 0.1722570538520813,
      "num_tokens": 36403831.0,
      "step": 16610
    },
    {
      "entropy": 5.590184497833252,
      "epoch": 1.6489678443826916,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004733384653774639,
      "loss": 5.1146,
      "mean_token_accuracy": 0.17070410698652266,
      "num_tokens": 36414896.0,
      "step": 16615
    },
    {
      "entropy": 5.540698909759522,
      "epoch": 1.649464073044859,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047332173301584764,
      "loss": 5.0591,
      "mean_token_accuracy": 0.17539832144975662,
      "num_tokens": 36426260.0,
      "step": 16620
    },
    {
      "entropy": 5.4968959331512455,
      "epoch": 1.6499603017070266,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047330499573627975,
      "loss": 5.0116,
      "mean_token_accuracy": 0.18272652179002763,
      "num_tokens": 36437125.0,
      "step": 16625
    },
    {
      "entropy": 5.55114426612854,
      "epoch": 1.650456530369194,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047328825353917526,
      "loss": 5.0502,
      "mean_token_accuracy": 0.17725541591644287,
      "num_tokens": 36447507.0,
      "step": 16630
    },
    {
      "entropy": 5.626207447052002,
      "epoch": 1.6509527590313615,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047327150642494944,
      "loss": 5.2273,
      "mean_token_accuracy": 0.17234772443771362,
      "num_tokens": 36459134.0,
      "step": 16635
    },
    {
      "entropy": 5.593615388870239,
      "epoch": 1.6514489876935292,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004732547543940174,
      "loss": 5.1335,
      "mean_token_accuracy": 0.16771827787160873,
      "num_tokens": 36469594.0,
      "step": 16640
    },
    {
      "entropy": 5.572950553894043,
      "epoch": 1.6519452163556967,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004732379974467947,
      "loss": 5.1304,
      "mean_token_accuracy": 0.17616445422172547,
      "num_tokens": 36480477.0,
      "step": 16645
    },
    {
      "entropy": 5.540207862854004,
      "epoch": 1.6524414450178644,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004732212355836968,
      "loss": 5.1281,
      "mean_token_accuracy": 0.17252252846956254,
      "num_tokens": 36491304.0,
      "step": 16650
    },
    {
      "entropy": 5.644248294830322,
      "epoch": 1.6529376736800319,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004732044688051395,
      "loss": 5.2047,
      "mean_token_accuracy": 0.1667452335357666,
      "num_tokens": 36502030.0,
      "step": 16655
    },
    {
      "entropy": 5.590003776550293,
      "epoch": 1.6534339023421993,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047318769711153833,
      "loss": 5.2311,
      "mean_token_accuracy": 0.16826645284891129,
      "num_tokens": 36514405.0,
      "step": 16660
    },
    {
      "entropy": 5.626352453231812,
      "epoch": 1.6539301310043668,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004731709205033094,
      "loss": 5.0856,
      "mean_token_accuracy": 0.1716939926147461,
      "num_tokens": 36524626.0,
      "step": 16665
    },
    {
      "entropy": 5.597578477859497,
      "epoch": 1.6544263596665343,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047315413898086863,
      "loss": 5.1067,
      "mean_token_accuracy": 0.17728606462478638,
      "num_tokens": 36535340.0,
      "step": 16670
    },
    {
      "entropy": 5.599070358276367,
      "epoch": 1.6549225883287018,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004731373525446322,
      "loss": 5.1096,
      "mean_token_accuracy": 0.17382441461086273,
      "num_tokens": 36547189.0,
      "step": 16675
    },
    {
      "entropy": 5.583989477157592,
      "epoch": 1.6554188169908692,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004731205611950164,
      "loss": 5.1797,
      "mean_token_accuracy": 0.17134556770324708,
      "num_tokens": 36558177.0,
      "step": 16680
    },
    {
      "entropy": 5.569241094589233,
      "epoch": 1.655915045653037,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047310376493243754,
      "loss": 4.9962,
      "mean_token_accuracy": 0.1793772280216217,
      "num_tokens": 36569559.0,
      "step": 16685
    },
    {
      "entropy": 5.578179407119751,
      "epoch": 1.6564112743152044,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004730869637573122,
      "loss": 5.0783,
      "mean_token_accuracy": 0.1765996739268303,
      "num_tokens": 36579617.0,
      "step": 16690
    },
    {
      "entropy": 5.52358193397522,
      "epoch": 1.656907502977372,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00047307015767005705,
      "loss": 5.0643,
      "mean_token_accuracy": 0.17602996230125428,
      "num_tokens": 36590427.0,
      "step": 16695
    },
    {
      "entropy": 5.5282731533050535,
      "epoch": 1.6574037316395396,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004730533466710888,
      "loss": 5.0727,
      "mean_token_accuracy": 0.17800467908382417,
      "num_tokens": 36602313.0,
      "step": 16700
    },
    {
      "entropy": 5.629635286331177,
      "epoch": 1.657899960301707,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004730365307608242,
      "loss": 5.1766,
      "mean_token_accuracy": 0.17333924174308776,
      "num_tokens": 36615191.0,
      "step": 16705
    },
    {
      "entropy": 5.557540416717529,
      "epoch": 1.6583961889638745,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047301970993968054,
      "loss": 5.083,
      "mean_token_accuracy": 0.1832567945122719,
      "num_tokens": 36624692.0,
      "step": 16710
    },
    {
      "entropy": 5.5930627346038815,
      "epoch": 1.658892417626042,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047300288420807463,
      "loss": 5.1616,
      "mean_token_accuracy": 0.1690323457121849,
      "num_tokens": 36635961.0,
      "step": 16715
    },
    {
      "entropy": 5.523765182495117,
      "epoch": 1.6593886462882095,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047298605356642387,
      "loss": 5.1141,
      "mean_token_accuracy": 0.1736500769853592,
      "num_tokens": 36646710.0,
      "step": 16720
    },
    {
      "entropy": 5.5658910274505615,
      "epoch": 1.6598848749503772,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047296921801514565,
      "loss": 5.1985,
      "mean_token_accuracy": 0.1703078866004944,
      "num_tokens": 36657168.0,
      "step": 16725
    },
    {
      "entropy": 5.598827171325683,
      "epoch": 1.6603811036125447,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004729523775546574,
      "loss": 5.1816,
      "mean_token_accuracy": 0.17437418103218078,
      "num_tokens": 36668428.0,
      "step": 16730
    },
    {
      "entropy": 5.600645112991333,
      "epoch": 1.6608773322747123,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004729355321853767,
      "loss": 5.1376,
      "mean_token_accuracy": 0.17496313154697418,
      "num_tokens": 36679361.0,
      "step": 16735
    },
    {
      "entropy": 5.655172443389892,
      "epoch": 1.6613735609368798,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004729186819077214,
      "loss": 5.184,
      "mean_token_accuracy": 0.17449210733175277,
      "num_tokens": 36689480.0,
      "step": 16740
    },
    {
      "entropy": 5.650704288482666,
      "epoch": 1.6618697895990473,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004729018267221092,
      "loss": 5.2244,
      "mean_token_accuracy": 0.1695423275232315,
      "num_tokens": 36700420.0,
      "step": 16745
    },
    {
      "entropy": 5.623846864700317,
      "epoch": 1.6623660182612148,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004728849666289581,
      "loss": 5.1755,
      "mean_token_accuracy": 0.16846733093261718,
      "num_tokens": 36710106.0,
      "step": 16750
    },
    {
      "entropy": 5.599697923660278,
      "epoch": 1.6628622469233822,
      "grad_norm": 1.0,
      "learning_rate": 0.00047286810162868625,
      "loss": 5.049,
      "mean_token_accuracy": 0.177377749979496,
      "num_tokens": 36719630.0,
      "step": 16755
    },
    {
      "entropy": 5.516613435745239,
      "epoch": 1.6633584755855497,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004728512317217119,
      "loss": 5.0223,
      "mean_token_accuracy": 0.17715496569871902,
      "num_tokens": 36729379.0,
      "step": 16760
    },
    {
      "entropy": 5.543292427062989,
      "epoch": 1.6638547042477172,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047283435690845336,
      "loss": 5.0718,
      "mean_token_accuracy": 0.17650600373744965,
      "num_tokens": 36740103.0,
      "step": 16765
    },
    {
      "entropy": 5.574830055236816,
      "epoch": 1.664350932909885,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047281747718932906,
      "loss": 5.1332,
      "mean_token_accuracy": 0.17653509229421616,
      "num_tokens": 36750928.0,
      "step": 16770
    },
    {
      "entropy": 5.561835479736328,
      "epoch": 1.6648471615720524,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004728005925647575,
      "loss": 5.1192,
      "mean_token_accuracy": 0.17223663479089737,
      "num_tokens": 36761415.0,
      "step": 16775
    },
    {
      "entropy": 5.522421503067017,
      "epoch": 1.66534339023422,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047278370303515744,
      "loss": 5.1556,
      "mean_token_accuracy": 0.17013748735189438,
      "num_tokens": 36772560.0,
      "step": 16780
    },
    {
      "entropy": 5.595023679733276,
      "epoch": 1.6658396188963875,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004727668086009478,
      "loss": 5.1574,
      "mean_token_accuracy": 0.17707760334014894,
      "num_tokens": 36783597.0,
      "step": 16785
    },
    {
      "entropy": 5.656346273422241,
      "epoch": 1.666335847558555,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047274990926254753,
      "loss": 5.2331,
      "mean_token_accuracy": 0.16883091479539872,
      "num_tokens": 36794219.0,
      "step": 16790
    },
    {
      "entropy": 5.53803071975708,
      "epoch": 1.6668320762207225,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004727330050203755,
      "loss": 5.0657,
      "mean_token_accuracy": 0.1763283535838127,
      "num_tokens": 36804156.0,
      "step": 16795
    },
    {
      "entropy": 5.5270322322845455,
      "epoch": 1.66732830488289,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004727160958748511,
      "loss": 5.0861,
      "mean_token_accuracy": 0.174847774207592,
      "num_tokens": 36814414.0,
      "step": 16800
    },
    {
      "entropy": 5.5711485862731935,
      "epoch": 1.6678245335450574,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047269918182639355,
      "loss": 5.0631,
      "mean_token_accuracy": 0.17212093025445938,
      "num_tokens": 36826173.0,
      "step": 16805
    },
    {
      "entropy": 5.620727968215943,
      "epoch": 1.6683207622072251,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004726822628754222,
      "loss": 5.1485,
      "mean_token_accuracy": 0.17004480808973313,
      "num_tokens": 36837726.0,
      "step": 16810
    },
    {
      "entropy": 5.605885124206543,
      "epoch": 1.6688169908693926,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047266533902235683,
      "loss": 5.1029,
      "mean_token_accuracy": 0.16931190192699433,
      "num_tokens": 36848605.0,
      "step": 16815
    },
    {
      "entropy": 5.592045450210572,
      "epoch": 1.66931321953156,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004726484102676169,
      "loss": 5.1372,
      "mean_token_accuracy": 0.17337658554315566,
      "num_tokens": 36858730.0,
      "step": 16820
    },
    {
      "entropy": 5.577565383911133,
      "epoch": 1.6698094481937278,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004726314766116223,
      "loss": 4.9978,
      "mean_token_accuracy": 0.17982637882232666,
      "num_tokens": 36868239.0,
      "step": 16825
    },
    {
      "entropy": 5.564413547515869,
      "epoch": 1.6703056768558953,
      "grad_norm": 1.21875,
      "learning_rate": 0.000472614538054793,
      "loss": 5.0942,
      "mean_token_accuracy": 0.17396096438169478,
      "num_tokens": 36878077.0,
      "step": 16830
    },
    {
      "entropy": 5.566725587844848,
      "epoch": 1.6708019055180627,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047259759459754893,
      "loss": 5.1315,
      "mean_token_accuracy": 0.1650699034333229,
      "num_tokens": 36888874.0,
      "step": 16835
    },
    {
      "entropy": 5.618921756744385,
      "epoch": 1.6712981341802302,
      "grad_norm": 1.0,
      "learning_rate": 0.0004725806462403103,
      "loss": 5.1395,
      "mean_token_accuracy": 0.17310104072093963,
      "num_tokens": 36898584.0,
      "step": 16840
    },
    {
      "entropy": 5.616058588027954,
      "epoch": 1.6717943628423977,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047256369298349736,
      "loss": 5.1319,
      "mean_token_accuracy": 0.16681381762027742,
      "num_tokens": 36908662.0,
      "step": 16845
    },
    {
      "entropy": 5.566705560684204,
      "epoch": 1.6722905915045652,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047254673482753065,
      "loss": 5.0569,
      "mean_token_accuracy": 0.17420319616794586,
      "num_tokens": 36919482.0,
      "step": 16850
    },
    {
      "entropy": 5.555973768234253,
      "epoch": 1.6727868201667329,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047252977177283045,
      "loss": 5.1094,
      "mean_token_accuracy": 0.16897981464862824,
      "num_tokens": 36930470.0,
      "step": 16855
    },
    {
      "entropy": 5.551867246627808,
      "epoch": 1.6732830488289003,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047251280381981764,
      "loss": 5.0966,
      "mean_token_accuracy": 0.17732679694890977,
      "num_tokens": 36942837.0,
      "step": 16860
    },
    {
      "entropy": 5.646102046966552,
      "epoch": 1.673779277491068,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004724958309689129,
      "loss": 5.1549,
      "mean_token_accuracy": 0.17051972895860673,
      "num_tokens": 36952994.0,
      "step": 16865
    },
    {
      "entropy": 5.597016000747681,
      "epoch": 1.6742755061532355,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047247885322053705,
      "loss": 5.1358,
      "mean_token_accuracy": 0.17145314067602158,
      "num_tokens": 36963973.0,
      "step": 16870
    },
    {
      "entropy": 5.528643465042114,
      "epoch": 1.674771734815403,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047246187057511127,
      "loss": 5.0729,
      "mean_token_accuracy": 0.17897389233112335,
      "num_tokens": 36975897.0,
      "step": 16875
    },
    {
      "entropy": 5.584465074539184,
      "epoch": 1.6752679634775705,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004724448830330565,
      "loss": 5.091,
      "mean_token_accuracy": 0.17511254251003266,
      "num_tokens": 36986481.0,
      "step": 16880
    },
    {
      "entropy": 5.591222810745239,
      "epoch": 1.675764192139738,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047242789059479413,
      "loss": 5.1752,
      "mean_token_accuracy": 0.16985351592302322,
      "num_tokens": 36997863.0,
      "step": 16885
    },
    {
      "entropy": 5.615442180633545,
      "epoch": 1.6762604208019054,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047241089326074547,
      "loss": 5.1377,
      "mean_token_accuracy": 0.16638160198926927,
      "num_tokens": 37009294.0,
      "step": 16890
    },
    {
      "entropy": 5.605202913284302,
      "epoch": 1.6767566494640729,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004723938910313321,
      "loss": 5.1052,
      "mean_token_accuracy": 0.17749632447957991,
      "num_tokens": 37019748.0,
      "step": 16895
    },
    {
      "entropy": 5.625437784194946,
      "epoch": 1.6772528781262406,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047237688390697555,
      "loss": 5.1811,
      "mean_token_accuracy": 0.17610367089509965,
      "num_tokens": 37030903.0,
      "step": 16900
    },
    {
      "entropy": 5.593409442901612,
      "epoch": 1.677749106788408,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004723598718880975,
      "loss": 5.1018,
      "mean_token_accuracy": 0.17399507611989976,
      "num_tokens": 37042551.0,
      "step": 16905
    },
    {
      "entropy": 5.599956846237182,
      "epoch": 1.6782453354505757,
      "grad_norm": 1.0,
      "learning_rate": 0.00047234285497512005,
      "loss": 5.1548,
      "mean_token_accuracy": 0.16995330303907394,
      "num_tokens": 37053151.0,
      "step": 16910
    },
    {
      "entropy": 5.630886030197144,
      "epoch": 1.6787415641127432,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047232583316846494,
      "loss": 5.1726,
      "mean_token_accuracy": 0.16549286991357803,
      "num_tokens": 37064421.0,
      "step": 16915
    },
    {
      "entropy": 5.591755962371826,
      "epoch": 1.6792377927749107,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004723088064685544,
      "loss": 5.1131,
      "mean_token_accuracy": 0.1808774635195732,
      "num_tokens": 37074758.0,
      "step": 16920
    },
    {
      "entropy": 5.601459074020386,
      "epoch": 1.6797340214370782,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047229177487581055,
      "loss": 5.0734,
      "mean_token_accuracy": 0.1723829671740532,
      "num_tokens": 37084898.0,
      "step": 16925
    },
    {
      "entropy": 5.652775764465332,
      "epoch": 1.6802302500992456,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047227473839065587,
      "loss": 5.162,
      "mean_token_accuracy": 0.1681185245513916,
      "num_tokens": 37094366.0,
      "step": 16930
    },
    {
      "entropy": 5.7097423553466795,
      "epoch": 1.6807264787614131,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047225769701351284,
      "loss": 5.2276,
      "mean_token_accuracy": 0.1659492939710617,
      "num_tokens": 37104965.0,
      "step": 16935
    },
    {
      "entropy": 5.562363147735596,
      "epoch": 1.6812227074235808,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004722406507448038,
      "loss": 5.0861,
      "mean_token_accuracy": 0.17326509952545166,
      "num_tokens": 37115622.0,
      "step": 16940
    },
    {
      "entropy": 5.616692590713501,
      "epoch": 1.6817189360857483,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00047222359958495175,
      "loss": 5.1753,
      "mean_token_accuracy": 0.16832805126905442,
      "num_tokens": 37127234.0,
      "step": 16945
    },
    {
      "entropy": 5.611864423751831,
      "epoch": 1.682215164747916,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004722065435343794,
      "loss": 5.0529,
      "mean_token_accuracy": 0.17526630461215972,
      "num_tokens": 37136533.0,
      "step": 16950
    },
    {
      "entropy": 5.630573654174805,
      "epoch": 1.6827113934100835,
      "grad_norm": 1.0,
      "learning_rate": 0.0004721894825935097,
      "loss": 5.1411,
      "mean_token_accuracy": 0.173529514670372,
      "num_tokens": 37147821.0,
      "step": 16955
    },
    {
      "entropy": 5.534932374954224,
      "epoch": 1.683207622072251,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004721724167627657,
      "loss": 5.1487,
      "mean_token_accuracy": 0.1736000433564186,
      "num_tokens": 37158929.0,
      "step": 16960
    },
    {
      "entropy": 5.6228889465332035,
      "epoch": 1.6837038507344184,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004721553460425708,
      "loss": 5.1339,
      "mean_token_accuracy": 0.17912727892398833,
      "num_tokens": 37169817.0,
      "step": 16965
    },
    {
      "entropy": 5.66729245185852,
      "epoch": 1.684200079396586,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000472138270433348,
      "loss": 5.2749,
      "mean_token_accuracy": 0.16259665042161942,
      "num_tokens": 37181009.0,
      "step": 16970
    },
    {
      "entropy": 5.591943740844727,
      "epoch": 1.6846963080587534,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047212118993552097,
      "loss": 5.0844,
      "mean_token_accuracy": 0.17644232660531997,
      "num_tokens": 37192110.0,
      "step": 16975
    },
    {
      "entropy": 5.592531156539917,
      "epoch": 1.6851925367209208,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004721041045495131,
      "loss": 5.1492,
      "mean_token_accuracy": 0.17449676245450974,
      "num_tokens": 37203426.0,
      "step": 16980
    },
    {
      "entropy": 5.623639392852783,
      "epoch": 1.6856887653830885,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047208701427574817,
      "loss": 5.2163,
      "mean_token_accuracy": 0.16688665896654128,
      "num_tokens": 37215025.0,
      "step": 16985
    },
    {
      "entropy": 5.668812799453735,
      "epoch": 1.686184994045256,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047206991911465,
      "loss": 5.1531,
      "mean_token_accuracy": 0.17609402537345886,
      "num_tokens": 37225653.0,
      "step": 16990
    },
    {
      "entropy": 5.604813480377198,
      "epoch": 1.6866812227074237,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004720528190666425,
      "loss": 5.118,
      "mean_token_accuracy": 0.17208495885133743,
      "num_tokens": 37237282.0,
      "step": 16995
    },
    {
      "entropy": 5.519793081283569,
      "epoch": 1.6871774513695912,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047203571413214967,
      "loss": 5.0664,
      "mean_token_accuracy": 0.18504511415958405,
      "num_tokens": 37248476.0,
      "step": 17000
    },
    {
      "entropy": 5.568170690536499,
      "epoch": 1.6876736800317587,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004720186043115957,
      "loss": 5.0745,
      "mean_token_accuracy": 0.174055215716362,
      "num_tokens": 37259354.0,
      "step": 17005
    },
    {
      "entropy": 5.58825798034668,
      "epoch": 1.6881699086939261,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004720014896054048,
      "loss": 5.1439,
      "mean_token_accuracy": 0.16976804435253143,
      "num_tokens": 37269895.0,
      "step": 17010
    },
    {
      "entropy": 5.536388158798218,
      "epoch": 1.6886661373560936,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004719843700140017,
      "loss": 5.1006,
      "mean_token_accuracy": 0.1812942460179329,
      "num_tokens": 37281201.0,
      "step": 17015
    },
    {
      "entropy": 5.573540163040161,
      "epoch": 1.689162366018261,
      "grad_norm": 1.125,
      "learning_rate": 0.0004719672455378105,
      "loss": 5.1115,
      "mean_token_accuracy": 0.1734094053506851,
      "num_tokens": 37293362.0,
      "step": 17020
    },
    {
      "entropy": 5.656302213668823,
      "epoch": 1.6896585946804288,
      "grad_norm": 0.875,
      "learning_rate": 0.000471950116177256,
      "loss": 5.1454,
      "mean_token_accuracy": 0.1715255692601204,
      "num_tokens": 37304291.0,
      "step": 17025
    },
    {
      "entropy": 5.6224040508270265,
      "epoch": 1.6901548233425963,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004719329819327631,
      "loss": 5.1271,
      "mean_token_accuracy": 0.17081913053989412,
      "num_tokens": 37315608.0,
      "step": 17030
    },
    {
      "entropy": 5.583102130889893,
      "epoch": 1.690651052004764,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004719158428047566,
      "loss": 5.0894,
      "mean_token_accuracy": 0.16722065806388856,
      "num_tokens": 37327559.0,
      "step": 17035
    },
    {
      "entropy": 5.56905198097229,
      "epoch": 1.6911472806669314,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004718986987936615,
      "loss": 5.1214,
      "mean_token_accuracy": 0.178587643802166,
      "num_tokens": 37338101.0,
      "step": 17040
    },
    {
      "entropy": 5.594436931610107,
      "epoch": 1.691643509329099,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047188154989990296,
      "loss": 5.2037,
      "mean_token_accuracy": 0.17303823232650756,
      "num_tokens": 37349124.0,
      "step": 17045
    },
    {
      "entropy": 5.5587921142578125,
      "epoch": 1.6921397379912664,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004718643961239062,
      "loss": 5.0507,
      "mean_token_accuracy": 0.17893173098564147,
      "num_tokens": 37360061.0,
      "step": 17050
    },
    {
      "entropy": 5.5921142578125,
      "epoch": 1.6926359666534339,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047184723746609666,
      "loss": 5.142,
      "mean_token_accuracy": 0.17197091728448868,
      "num_tokens": 37371210.0,
      "step": 17055
    },
    {
      "entropy": 5.5502392768859865,
      "epoch": 1.6931321953156013,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047183007392689977,
      "loss": 5.0872,
      "mean_token_accuracy": 0.1693829983472824,
      "num_tokens": 37382547.0,
      "step": 17060
    },
    {
      "entropy": 5.583687829971313,
      "epoch": 1.6936284239777688,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047181290550674115,
      "loss": 5.1368,
      "mean_token_accuracy": 0.17479816526174546,
      "num_tokens": 37392311.0,
      "step": 17065
    },
    {
      "entropy": 5.590213775634766,
      "epoch": 1.6941246526399365,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004717957322060466,
      "loss": 5.082,
      "mean_token_accuracy": 0.1755610451102257,
      "num_tokens": 37402626.0,
      "step": 17070
    },
    {
      "entropy": 5.564432907104492,
      "epoch": 1.694620881302104,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004717785540252419,
      "loss": 5.0881,
      "mean_token_accuracy": 0.17429988384246825,
      "num_tokens": 37412100.0,
      "step": 17075
    },
    {
      "entropy": 5.511949586868286,
      "epoch": 1.6951171099642717,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004717613709647532,
      "loss": 5.0057,
      "mean_token_accuracy": 0.18122729361057283,
      "num_tokens": 37421735.0,
      "step": 17080
    },
    {
      "entropy": 5.621310043334961,
      "epoch": 1.6956133386264391,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047174418302500635,
      "loss": 5.2089,
      "mean_token_accuracy": 0.16157240718603133,
      "num_tokens": 37432179.0,
      "step": 17085
    },
    {
      "entropy": 5.592205476760864,
      "epoch": 1.6961095672886066,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004717269902064278,
      "loss": 5.0575,
      "mean_token_accuracy": 0.17967303991317748,
      "num_tokens": 37443276.0,
      "step": 17090
    },
    {
      "entropy": 5.621029090881348,
      "epoch": 1.696605795950774,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004717097925094438,
      "loss": 5.1771,
      "mean_token_accuracy": 0.17259810864925385,
      "num_tokens": 37454537.0,
      "step": 17095
    },
    {
      "entropy": 5.591892528533935,
      "epoch": 1.6971020246129416,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004716925899344808,
      "loss": 5.0957,
      "mean_token_accuracy": 0.16564809232950212,
      "num_tokens": 37465493.0,
      "step": 17100
    },
    {
      "entropy": 5.523123455047608,
      "epoch": 1.697598253275109,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004716753824819654,
      "loss": 5.0838,
      "mean_token_accuracy": 0.17157469391822816,
      "num_tokens": 37475511.0,
      "step": 17105
    },
    {
      "entropy": 5.499141597747803,
      "epoch": 1.6980944819372767,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004716581701523243,
      "loss": 5.0035,
      "mean_token_accuracy": 0.17352219969034194,
      "num_tokens": 37487046.0,
      "step": 17110
    },
    {
      "entropy": 5.5762416362762455,
      "epoch": 1.6985907105994442,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047164095294598433,
      "loss": 5.1637,
      "mean_token_accuracy": 0.17222529798746108,
      "num_tokens": 37498398.0,
      "step": 17115
    },
    {
      "entropy": 5.590070676803589,
      "epoch": 1.699086939261612,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047162373086337243,
      "loss": 5.1632,
      "mean_token_accuracy": 0.1697535276412964,
      "num_tokens": 37509397.0,
      "step": 17120
    },
    {
      "entropy": 5.583338689804077,
      "epoch": 1.6995831679237794,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047160650390491574,
      "loss": 5.035,
      "mean_token_accuracy": 0.178615602850914,
      "num_tokens": 37519284.0,
      "step": 17125
    },
    {
      "entropy": 5.609713077545166,
      "epoch": 1.7000793965859469,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004715892720710414,
      "loss": 5.1416,
      "mean_token_accuracy": 0.1722246766090393,
      "num_tokens": 37529699.0,
      "step": 17130
    },
    {
      "entropy": 5.643819332122803,
      "epoch": 1.7005756252481143,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004715720353621767,
      "loss": 5.1841,
      "mean_token_accuracy": 0.16601076126098632,
      "num_tokens": 37541756.0,
      "step": 17135
    },
    {
      "entropy": 5.537514734268188,
      "epoch": 1.7010718539102818,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047155479377874917,
      "loss": 5.0758,
      "mean_token_accuracy": 0.17741183191537857,
      "num_tokens": 37552693.0,
      "step": 17140
    },
    {
      "entropy": 5.5561432361602785,
      "epoch": 1.7015680825724493,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047153754732118616,
      "loss": 5.0526,
      "mean_token_accuracy": 0.17462966591119766,
      "num_tokens": 37563826.0,
      "step": 17145
    },
    {
      "entropy": 5.5933976650238035,
      "epoch": 1.7020643112346168,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004715202959899156,
      "loss": 5.0984,
      "mean_token_accuracy": 0.1754734307527542,
      "num_tokens": 37576645.0,
      "step": 17150
    },
    {
      "entropy": 5.450649213790894,
      "epoch": 1.7025605398967845,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004715030397853651,
      "loss": 4.9812,
      "mean_token_accuracy": 0.18178724646568298,
      "num_tokens": 37587467.0,
      "step": 17155
    },
    {
      "entropy": 5.585561895370484,
      "epoch": 1.703056768558952,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047148577870796264,
      "loss": 5.1614,
      "mean_token_accuracy": 0.164557746052742,
      "num_tokens": 37598475.0,
      "step": 17160
    },
    {
      "entropy": 5.653621578216553,
      "epoch": 1.7035529972211196,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004714685127581363,
      "loss": 5.2421,
      "mean_token_accuracy": 0.168009015917778,
      "num_tokens": 37610260.0,
      "step": 17165
    },
    {
      "entropy": 5.603365039825439,
      "epoch": 1.704049225883287,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047145124193631415,
      "loss": 5.1784,
      "mean_token_accuracy": 0.16965125799179076,
      "num_tokens": 37621129.0,
      "step": 17170
    },
    {
      "entropy": 5.581546640396118,
      "epoch": 1.7045454545454546,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047143396624292454,
      "loss": 5.0744,
      "mean_token_accuracy": 0.17354739010334014,
      "num_tokens": 37633130.0,
      "step": 17175
    },
    {
      "entropy": 5.592241859436035,
      "epoch": 1.705041683207622,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047141668567839584,
      "loss": 5.1907,
      "mean_token_accuracy": 0.1708127498626709,
      "num_tokens": 37644247.0,
      "step": 17180
    },
    {
      "entropy": 5.597466468811035,
      "epoch": 1.7055379118697895,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004713994002431566,
      "loss": 5.1193,
      "mean_token_accuracy": 0.17397734224796296,
      "num_tokens": 37655602.0,
      "step": 17185
    },
    {
      "entropy": 5.525761508941651,
      "epoch": 1.706034140531957,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047138210993763534,
      "loss": 5.0398,
      "mean_token_accuracy": 0.17917388528585435,
      "num_tokens": 37667150.0,
      "step": 17190
    },
    {
      "entropy": 5.550930070877075,
      "epoch": 1.7065303691941247,
      "grad_norm": 0.921875,
      "learning_rate": 0.000471364814762261,
      "loss": 5.0594,
      "mean_token_accuracy": 0.17920710146427155,
      "num_tokens": 37677626.0,
      "step": 17195
    },
    {
      "entropy": 5.638219118118286,
      "epoch": 1.7070265978562922,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004713475147174624,
      "loss": 5.1918,
      "mean_token_accuracy": 0.17037307769060134,
      "num_tokens": 37689050.0,
      "step": 17200
    },
    {
      "entropy": 5.628713369369507,
      "epoch": 1.7075228265184597,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004713302098036685,
      "loss": 5.0879,
      "mean_token_accuracy": 0.1852431058883667,
      "num_tokens": 37699190.0,
      "step": 17205
    },
    {
      "entropy": 5.526066684722901,
      "epoch": 1.7080190551806274,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004713129000213085,
      "loss": 5.0634,
      "mean_token_accuracy": 0.1802924633026123,
      "num_tokens": 37710465.0,
      "step": 17210
    },
    {
      "entropy": 5.589704608917236,
      "epoch": 1.7085152838427948,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004712955853708115,
      "loss": 5.1799,
      "mean_token_accuracy": 0.1738348960876465,
      "num_tokens": 37722426.0,
      "step": 17215
    },
    {
      "entropy": 5.563302755355835,
      "epoch": 1.7090115125049623,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000471278265852607,
      "loss": 5.0723,
      "mean_token_accuracy": 0.18183045983314514,
      "num_tokens": 37733104.0,
      "step": 17220
    },
    {
      "entropy": 5.54639892578125,
      "epoch": 1.7095077411671298,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004712609414671245,
      "loss": 5.0083,
      "mean_token_accuracy": 0.17964622080326081,
      "num_tokens": 37742747.0,
      "step": 17225
    },
    {
      "entropy": 5.5648902416229244,
      "epoch": 1.7100039698292973,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004712436122147935,
      "loss": 5.0546,
      "mean_token_accuracy": 0.1747670888900757,
      "num_tokens": 37753338.0,
      "step": 17230
    },
    {
      "entropy": 5.543088960647583,
      "epoch": 1.7105001984914647,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004712262780960438,
      "loss": 5.0537,
      "mean_token_accuracy": 0.18208860009908676,
      "num_tokens": 37763122.0,
      "step": 17235
    },
    {
      "entropy": 5.555970239639282,
      "epoch": 1.7109964271536324,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047120893911130525,
      "loss": 5.0447,
      "mean_token_accuracy": 0.16858939975500106,
      "num_tokens": 37774677.0,
      "step": 17240
    },
    {
      "entropy": 5.560484790802002,
      "epoch": 1.7114926558158,
      "grad_norm": 0.875,
      "learning_rate": 0.00047119159526100773,
      "loss": 5.07,
      "mean_token_accuracy": 0.17217567712068557,
      "num_tokens": 37786186.0,
      "step": 17245
    },
    {
      "entropy": 5.5244334697723385,
      "epoch": 1.7119888844779676,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047117424654558146,
      "loss": 5.0993,
      "mean_token_accuracy": 0.1808667927980423,
      "num_tokens": 37796837.0,
      "step": 17250
    },
    {
      "entropy": 5.596328258514404,
      "epoch": 1.712485113140135,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004711568929654566,
      "loss": 5.1477,
      "mean_token_accuracy": 0.1706267699599266,
      "num_tokens": 37808813.0,
      "step": 17255
    },
    {
      "entropy": 5.569423818588257,
      "epoch": 1.7129813418023025,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004711395345210635,
      "loss": 5.0239,
      "mean_token_accuracy": 0.1794498324394226,
      "num_tokens": 37819770.0,
      "step": 17260
    },
    {
      "entropy": 5.59895429611206,
      "epoch": 1.71347757046447,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004711221712128325,
      "loss": 5.1042,
      "mean_token_accuracy": 0.17340686023235322,
      "num_tokens": 37831831.0,
      "step": 17265
    },
    {
      "entropy": 5.568275356292725,
      "epoch": 1.7139737991266375,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004711048030411943,
      "loss": 5.151,
      "mean_token_accuracy": 0.16977563351392747,
      "num_tokens": 37843092.0,
      "step": 17270
    },
    {
      "entropy": 5.57982702255249,
      "epoch": 1.714470027788805,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047108743000657963,
      "loss": 5.1186,
      "mean_token_accuracy": 0.18013927787542344,
      "num_tokens": 37853192.0,
      "step": 17275
    },
    {
      "entropy": 5.658019876480102,
      "epoch": 1.7149662564509724,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004710700521094192,
      "loss": 5.1443,
      "mean_token_accuracy": 0.17124520391225814,
      "num_tokens": 37863510.0,
      "step": 17280
    },
    {
      "entropy": 5.55731430053711,
      "epoch": 1.7154624851131401,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00047105266935014385,
      "loss": 5.0454,
      "mean_token_accuracy": 0.1851502239704132,
      "num_tokens": 37876156.0,
      "step": 17285
    },
    {
      "entropy": 5.540690469741821,
      "epoch": 1.7159587137753076,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047103528172918487,
      "loss": 5.0863,
      "mean_token_accuracy": 0.18287326544523239,
      "num_tokens": 37887153.0,
      "step": 17290
    },
    {
      "entropy": 5.5813428401947025,
      "epoch": 1.7164549424374753,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004710178892469733,
      "loss": 5.0844,
      "mean_token_accuracy": 0.1767450749874115,
      "num_tokens": 37898820.0,
      "step": 17295
    },
    {
      "entropy": 5.593496704101563,
      "epoch": 1.7169511710996428,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047100049190394044,
      "loss": 5.0732,
      "mean_token_accuracy": 0.17751829475164413,
      "num_tokens": 37910257.0,
      "step": 17300
    },
    {
      "entropy": 5.567697763442993,
      "epoch": 1.7174473997618103,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004709830897005178,
      "loss": 5.1683,
      "mean_token_accuracy": 0.17373329102993013,
      "num_tokens": 37921574.0,
      "step": 17305
    },
    {
      "entropy": 5.632778835296631,
      "epoch": 1.7179436284239777,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004709656826371367,
      "loss": 5.1342,
      "mean_token_accuracy": 0.173052379488945,
      "num_tokens": 37932741.0,
      "step": 17310
    },
    {
      "entropy": 5.56987829208374,
      "epoch": 1.7184398570861452,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004709482707142291,
      "loss": 5.1313,
      "mean_token_accuracy": 0.17625672519207,
      "num_tokens": 37944592.0,
      "step": 17315
    },
    {
      "entropy": 5.562404823303223,
      "epoch": 1.7189360857483127,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047093085393222646,
      "loss": 5.0676,
      "mean_token_accuracy": 0.17690574526786804,
      "num_tokens": 37954983.0,
      "step": 17320
    },
    {
      "entropy": 5.5876294612884525,
      "epoch": 1.7194323144104804,
      "grad_norm": 1.0,
      "learning_rate": 0.00047091343229156094,
      "loss": 5.1442,
      "mean_token_accuracy": 0.1744734987616539,
      "num_tokens": 37964491.0,
      "step": 17325
    },
    {
      "entropy": 5.547980690002442,
      "epoch": 1.7199285430726479,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004708960057926644,
      "loss": 5.0303,
      "mean_token_accuracy": 0.17898047119379043,
      "num_tokens": 37976140.0,
      "step": 17330
    },
    {
      "entropy": 5.546603965759277,
      "epoch": 1.7204247717348156,
      "grad_norm": 1.0,
      "learning_rate": 0.00047087857443596905,
      "loss": 5.0954,
      "mean_token_accuracy": 0.18001052737236023,
      "num_tokens": 37987776.0,
      "step": 17335
    },
    {
      "entropy": 5.579403018951416,
      "epoch": 1.720921000396983,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047086113822190716,
      "loss": 5.1375,
      "mean_token_accuracy": 0.17503750920295716,
      "num_tokens": 37998088.0,
      "step": 17340
    },
    {
      "entropy": 5.557414293289185,
      "epoch": 1.7214172290591505,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047084369715091103,
      "loss": 5.1349,
      "mean_token_accuracy": 0.16945750266313553,
      "num_tokens": 38009802.0,
      "step": 17345
    },
    {
      "entropy": 5.5954272747039795,
      "epoch": 1.721913457721318,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004708262512234133,
      "loss": 5.1575,
      "mean_token_accuracy": 0.17774184346199035,
      "num_tokens": 38020797.0,
      "step": 17350
    },
    {
      "entropy": 5.654356479644775,
      "epoch": 1.7224096863834855,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004708088004398464,
      "loss": 5.167,
      "mean_token_accuracy": 0.1746874451637268,
      "num_tokens": 38031816.0,
      "step": 17355
    },
    {
      "entropy": 5.584352111816406,
      "epoch": 1.722905915045653,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004707913448006433,
      "loss": 5.1528,
      "mean_token_accuracy": 0.16281828433275222,
      "num_tokens": 38043953.0,
      "step": 17360
    },
    {
      "entropy": 5.508972978591919,
      "epoch": 1.7234021437078204,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004707738843062367,
      "loss": 5.018,
      "mean_token_accuracy": 0.184291310608387,
      "num_tokens": 38053929.0,
      "step": 17365
    },
    {
      "entropy": 5.625787687301636,
      "epoch": 1.723898372369988,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004707564189570596,
      "loss": 5.2583,
      "mean_token_accuracy": 0.1726446971297264,
      "num_tokens": 38067091.0,
      "step": 17370
    },
    {
      "entropy": 5.629706239700317,
      "epoch": 1.7243946010321556,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004707389487535451,
      "loss": 5.0862,
      "mean_token_accuracy": 0.17263598889112472,
      "num_tokens": 38079002.0,
      "step": 17375
    },
    {
      "entropy": 5.606559801101684,
      "epoch": 1.7248908296943233,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004707214736961265,
      "loss": 5.1062,
      "mean_token_accuracy": 0.18048200905323028,
      "num_tokens": 38089526.0,
      "step": 17380
    },
    {
      "entropy": 5.591858291625977,
      "epoch": 1.7253870583564908,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004707039937852371,
      "loss": 5.0954,
      "mean_token_accuracy": 0.1746528998017311,
      "num_tokens": 38099902.0,
      "step": 17385
    },
    {
      "entropy": 5.548422908782959,
      "epoch": 1.7258832870186582,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047068650902131034,
      "loss": 5.0771,
      "mean_token_accuracy": 0.1778904005885124,
      "num_tokens": 38111725.0,
      "step": 17390
    },
    {
      "entropy": 5.581420230865478,
      "epoch": 1.7263795156808257,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004706690194047798,
      "loss": 5.0361,
      "mean_token_accuracy": 0.17861855626106263,
      "num_tokens": 38123522.0,
      "step": 17395
    },
    {
      "entropy": 5.581313133239746,
      "epoch": 1.7268757443429932,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004706515249360792,
      "loss": 5.1619,
      "mean_token_accuracy": 0.1691841021180153,
      "num_tokens": 38133968.0,
      "step": 17400
    },
    {
      "entropy": 5.521755170822144,
      "epoch": 1.7273719730051607,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004706340256156425,
      "loss": 5.2134,
      "mean_token_accuracy": 0.16236061751842498,
      "num_tokens": 38145288.0,
      "step": 17405
    },
    {
      "entropy": 5.590142059326172,
      "epoch": 1.7278682016673284,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004706165214439034,
      "loss": 5.0958,
      "mean_token_accuracy": 0.1752851963043213,
      "num_tokens": 38156141.0,
      "step": 17410
    },
    {
      "entropy": 5.605217027664184,
      "epoch": 1.7283644303294958,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004705990124212961,
      "loss": 5.0617,
      "mean_token_accuracy": 0.18354837894439696,
      "num_tokens": 38167211.0,
      "step": 17415
    },
    {
      "entropy": 5.569778871536255,
      "epoch": 1.7288606589916635,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004705814985482548,
      "loss": 5.1352,
      "mean_token_accuracy": 0.17271792143583298,
      "num_tokens": 38177360.0,
      "step": 17420
    },
    {
      "entropy": 5.54235806465149,
      "epoch": 1.729356887653831,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047056397982521366,
      "loss": 5.0267,
      "mean_token_accuracy": 0.18099224120378493,
      "num_tokens": 38187857.0,
      "step": 17425
    },
    {
      "entropy": 5.6416380405426025,
      "epoch": 1.7298531163159985,
      "grad_norm": 1.125,
      "learning_rate": 0.0004705464562526074,
      "loss": 5.1665,
      "mean_token_accuracy": 0.16364758014678954,
      "num_tokens": 38198161.0,
      "step": 17430
    },
    {
      "entropy": 5.580646705627442,
      "epoch": 1.730349344978166,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004705289278308703,
      "loss": 5.0961,
      "mean_token_accuracy": 0.17802167385816575,
      "num_tokens": 38208612.0,
      "step": 17435
    },
    {
      "entropy": 5.581148242950439,
      "epoch": 1.7308455736403334,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047051139456043707,
      "loss": 5.1246,
      "mean_token_accuracy": 0.17107446640729904,
      "num_tokens": 38220153.0,
      "step": 17440
    },
    {
      "entropy": 5.5628162860870365,
      "epoch": 1.731341802302501,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004704938564417427,
      "loss": 5.094,
      "mean_token_accuracy": 0.17820525914430618,
      "num_tokens": 38230790.0,
      "step": 17445
    },
    {
      "entropy": 5.595528841018677,
      "epoch": 1.7318380309646684,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047047631347522175,
      "loss": 5.0479,
      "mean_token_accuracy": 0.18130245208740234,
      "num_tokens": 38242585.0,
      "step": 17450
    },
    {
      "entropy": 5.587844371795654,
      "epoch": 1.732334259626836,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047045876566130957,
      "loss": 5.0943,
      "mean_token_accuracy": 0.17695927023887634,
      "num_tokens": 38253316.0,
      "step": 17455
    },
    {
      "entropy": 5.519856882095337,
      "epoch": 1.7328304882890035,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004704412130004411,
      "loss": 5.0774,
      "mean_token_accuracy": 0.17991492450237273,
      "num_tokens": 38263230.0,
      "step": 17460
    },
    {
      "entropy": 5.613614225387574,
      "epoch": 1.7333267169511712,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047042365549305174,
      "loss": 5.1319,
      "mean_token_accuracy": 0.17395465224981307,
      "num_tokens": 38273416.0,
      "step": 17465
    },
    {
      "entropy": 5.550931024551391,
      "epoch": 1.7338229456133387,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004704060931395768,
      "loss": 5.0626,
      "mean_token_accuracy": 0.18041899800300598,
      "num_tokens": 38285764.0,
      "step": 17470
    },
    {
      "entropy": 5.61083083152771,
      "epoch": 1.7343191742755062,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047038852594045186,
      "loss": 5.177,
      "mean_token_accuracy": 0.17305256724357604,
      "num_tokens": 38296899.0,
      "step": 17475
    },
    {
      "entropy": 5.645906591415406,
      "epoch": 1.7348154029376737,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047037095389611245,
      "loss": 5.1644,
      "mean_token_accuracy": 0.18068629056215285,
      "num_tokens": 38309050.0,
      "step": 17480
    },
    {
      "entropy": 5.56595139503479,
      "epoch": 1.7353116315998411,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004703533770069944,
      "loss": 5.103,
      "mean_token_accuracy": 0.17612729221582413,
      "num_tokens": 38320831.0,
      "step": 17485
    },
    {
      "entropy": 5.56408920288086,
      "epoch": 1.7358078602620086,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047033579527353355,
      "loss": 5.1199,
      "mean_token_accuracy": 0.17394785284996034,
      "num_tokens": 38332096.0,
      "step": 17490
    },
    {
      "entropy": 5.610078620910644,
      "epoch": 1.7363040889241763,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004703182086961659,
      "loss": 5.1343,
      "mean_token_accuracy": 0.17231152802705765,
      "num_tokens": 38343110.0,
      "step": 17495
    },
    {
      "entropy": 5.608712005615234,
      "epoch": 1.7368003175863438,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004703006172753274,
      "loss": 5.1258,
      "mean_token_accuracy": 0.17680202275514603,
      "num_tokens": 38353971.0,
      "step": 17500
    },
    {
      "entropy": 5.5369573593139645,
      "epoch": 1.7372965462485113,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047028302101145467,
      "loss": 5.0639,
      "mean_token_accuracy": 0.1812679275870323,
      "num_tokens": 38365114.0,
      "step": 17505
    },
    {
      "entropy": 5.590271139144898,
      "epoch": 1.737792774910679,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004702654199049836,
      "loss": 5.0941,
      "mean_token_accuracy": 0.1722070664167404,
      "num_tokens": 38375784.0,
      "step": 17510
    },
    {
      "entropy": 5.575156211853027,
      "epoch": 1.7382890035728464,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000470247813956351,
      "loss": 5.0817,
      "mean_token_accuracy": 0.1755967080593109,
      "num_tokens": 38386053.0,
      "step": 17515
    },
    {
      "entropy": 5.510431146621704,
      "epoch": 1.738785232235014,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047023020316599334,
      "loss": 5.0757,
      "mean_token_accuracy": 0.17339733392000198,
      "num_tokens": 38396545.0,
      "step": 17520
    },
    {
      "entropy": 5.543194055557251,
      "epoch": 1.7392814608971814,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047021258753434725,
      "loss": 5.0492,
      "mean_token_accuracy": 0.17810032069683074,
      "num_tokens": 38406344.0,
      "step": 17525
    },
    {
      "entropy": 5.590229749679565,
      "epoch": 1.7397776895593489,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047019496706184966,
      "loss": 5.0339,
      "mean_token_accuracy": 0.18382048606872559,
      "num_tokens": 38416388.0,
      "step": 17530
    },
    {
      "entropy": 5.595349740982056,
      "epoch": 1.7402739182215163,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004701773417489375,
      "loss": 5.1829,
      "mean_token_accuracy": 0.16692702621221542,
      "num_tokens": 38427608.0,
      "step": 17535
    },
    {
      "entropy": 5.568109750747681,
      "epoch": 1.740770146883684,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047015971159604787,
      "loss": 5.0932,
      "mean_token_accuracy": 0.17748434245586395,
      "num_tokens": 38437921.0,
      "step": 17540
    },
    {
      "entropy": 5.560401964187622,
      "epoch": 1.7412663755458515,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000470142076603618,
      "loss": 5.1419,
      "mean_token_accuracy": 0.17027279883623123,
      "num_tokens": 38448767.0,
      "step": 17545
    },
    {
      "entropy": 5.61631965637207,
      "epoch": 1.7417626042080192,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047012443677208496,
      "loss": 5.2189,
      "mean_token_accuracy": 0.16487084031105043,
      "num_tokens": 38460052.0,
      "step": 17550
    },
    {
      "entropy": 5.558698511123657,
      "epoch": 1.7422588328701867,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004701067921018864,
      "loss": 5.0599,
      "mean_token_accuracy": 0.1766940549015999,
      "num_tokens": 38470177.0,
      "step": 17555
    },
    {
      "entropy": 5.583619070053101,
      "epoch": 1.7427550615323542,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047008914259345986,
      "loss": 5.0878,
      "mean_token_accuracy": 0.17292993664741516,
      "num_tokens": 38481608.0,
      "step": 17560
    },
    {
      "entropy": 5.586293172836304,
      "epoch": 1.7432512901945216,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047007148824724295,
      "loss": 5.1522,
      "mean_token_accuracy": 0.16932461261749268,
      "num_tokens": 38492534.0,
      "step": 17565
    },
    {
      "entropy": 5.595059061050415,
      "epoch": 1.743747518856689,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004700538290636734,
      "loss": 5.1516,
      "mean_token_accuracy": 0.17489712536334992,
      "num_tokens": 38503629.0,
      "step": 17570
    },
    {
      "entropy": 5.5303479671478275,
      "epoch": 1.7442437475188566,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047003616504318925,
      "loss": 5.0363,
      "mean_token_accuracy": 0.17710695266723633,
      "num_tokens": 38514428.0,
      "step": 17575
    },
    {
      "entropy": 5.605891704559326,
      "epoch": 1.744739976181024,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047001849618622844,
      "loss": 5.1154,
      "mean_token_accuracy": 0.17448204159736633,
      "num_tokens": 38525468.0,
      "step": 17580
    },
    {
      "entropy": 5.566159915924072,
      "epoch": 1.7452362048431918,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047000082249322916,
      "loss": 5.0787,
      "mean_token_accuracy": 0.1738796979188919,
      "num_tokens": 38535095.0,
      "step": 17585
    },
    {
      "entropy": 5.5847546577453615,
      "epoch": 1.7457324335053592,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004699831439646297,
      "loss": 5.0664,
      "mean_token_accuracy": 0.1764317810535431,
      "num_tokens": 38546867.0,
      "step": 17590
    },
    {
      "entropy": 5.513819885253906,
      "epoch": 1.746228662167527,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004699654606008684,
      "loss": 4.9854,
      "mean_token_accuracy": 0.18180763125419616,
      "num_tokens": 38557231.0,
      "step": 17595
    },
    {
      "entropy": 5.58009295463562,
      "epoch": 1.7467248908296944,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004699477724023838,
      "loss": 5.1295,
      "mean_token_accuracy": 0.1723133385181427,
      "num_tokens": 38568857.0,
      "step": 17600
    },
    {
      "entropy": 5.602437591552734,
      "epoch": 1.7472211194918619,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004699300793696145,
      "loss": 5.0821,
      "mean_token_accuracy": 0.17299691438674927,
      "num_tokens": 38579177.0,
      "step": 17605
    },
    {
      "entropy": 5.577049446105957,
      "epoch": 1.7477173481540293,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004699123815029993,
      "loss": 4.9955,
      "mean_token_accuracy": 0.18278008997440337,
      "num_tokens": 38588683.0,
      "step": 17610
    },
    {
      "entropy": 5.527711296081543,
      "epoch": 1.7482135768161968,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046989467880297704,
      "loss": 5.0682,
      "mean_token_accuracy": 0.17503969222307206,
      "num_tokens": 38599636.0,
      "step": 17615
    },
    {
      "entropy": 5.651794385910034,
      "epoch": 1.7487098054783643,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046987697126998673,
      "loss": 5.1814,
      "mean_token_accuracy": 0.17078714370727538,
      "num_tokens": 38609360.0,
      "step": 17620
    },
    {
      "entropy": 5.613919067382812,
      "epoch": 1.749206034140532,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004698592589044674,
      "loss": 5.1004,
      "mean_token_accuracy": 0.17578664869070054,
      "num_tokens": 38620996.0,
      "step": 17625
    },
    {
      "entropy": 5.613323307037353,
      "epoch": 1.7497022628026995,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046984154170685833,
      "loss": 5.2298,
      "mean_token_accuracy": 0.17495223581790925,
      "num_tokens": 38632935.0,
      "step": 17630
    },
    {
      "entropy": 5.64689884185791,
      "epoch": 1.7501984914648672,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004698238196775989,
      "loss": 5.1905,
      "mean_token_accuracy": 0.170673269033432,
      "num_tokens": 38644299.0,
      "step": 17635
    },
    {
      "entropy": 5.599116945266724,
      "epoch": 1.7506947201270346,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046980609281712857,
      "loss": 5.1304,
      "mean_token_accuracy": 0.17658956348896027,
      "num_tokens": 38654970.0,
      "step": 17640
    },
    {
      "entropy": 5.623294353485107,
      "epoch": 1.7511909487892021,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004697883611258868,
      "loss": 5.1861,
      "mean_token_accuracy": 0.1754039779305458,
      "num_tokens": 38666302.0,
      "step": 17645
    },
    {
      "entropy": 5.602449941635132,
      "epoch": 1.7516871774513696,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004697706246043136,
      "loss": 5.0928,
      "mean_token_accuracy": 0.17453875690698623,
      "num_tokens": 38678346.0,
      "step": 17650
    },
    {
      "entropy": 5.540946435928345,
      "epoch": 1.752183406113537,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046975288325284845,
      "loss": 5.0244,
      "mean_token_accuracy": 0.17798079252243043,
      "num_tokens": 38688242.0,
      "step": 17655
    },
    {
      "entropy": 5.597959518432617,
      "epoch": 1.7526796347757045,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004697351370719315,
      "loss": 5.2171,
      "mean_token_accuracy": 0.1644924744963646,
      "num_tokens": 38698756.0,
      "step": 17660
    },
    {
      "entropy": 5.605363273620606,
      "epoch": 1.753175863437872,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004697173860620027,
      "loss": 5.178,
      "mean_token_accuracy": 0.16965143829584123,
      "num_tokens": 38710410.0,
      "step": 17665
    },
    {
      "entropy": 5.681347322463989,
      "epoch": 1.7536720921000397,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004696996302235024,
      "loss": 5.1261,
      "mean_token_accuracy": 0.1706134036183357,
      "num_tokens": 38721917.0,
      "step": 17670
    },
    {
      "entropy": 5.6153114318847654,
      "epoch": 1.7541683207622072,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046968186955687074,
      "loss": 5.1498,
      "mean_token_accuracy": 0.18041616678237915,
      "num_tokens": 38733085.0,
      "step": 17675
    },
    {
      "entropy": 5.561414909362793,
      "epoch": 1.7546645494243749,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004696641040625483,
      "loss": 5.1707,
      "mean_token_accuracy": 0.17223667353391647,
      "num_tokens": 38744785.0,
      "step": 17680
    },
    {
      "entropy": 5.605169486999512,
      "epoch": 1.7551607780865424,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046964633374097544,
      "loss": 5.1339,
      "mean_token_accuracy": 0.17743027657270433,
      "num_tokens": 38755626.0,
      "step": 17685
    },
    {
      "entropy": 5.601816987991333,
      "epoch": 1.7556570067487098,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046962855859259306,
      "loss": 5.0166,
      "mean_token_accuracy": 0.17540077567100526,
      "num_tokens": 38766360.0,
      "step": 17690
    },
    {
      "entropy": 5.597133445739746,
      "epoch": 1.7561532354108773,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046961077861784167,
      "loss": 5.1079,
      "mean_token_accuracy": 0.17417547553777696,
      "num_tokens": 38776763.0,
      "step": 17695
    },
    {
      "entropy": 5.554697132110595,
      "epoch": 1.7566494640730448,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004695929938171624,
      "loss": 5.1524,
      "mean_token_accuracy": 0.17313022017478943,
      "num_tokens": 38787664.0,
      "step": 17700
    },
    {
      "entropy": 5.592082786560058,
      "epoch": 1.7571456927352123,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004695752041909962,
      "loss": 5.0834,
      "mean_token_accuracy": 0.17332491725683213,
      "num_tokens": 38798018.0,
      "step": 17705
    },
    {
      "entropy": 5.594988393783569,
      "epoch": 1.75764192139738,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004695574097397842,
      "loss": 5.048,
      "mean_token_accuracy": 0.173671892285347,
      "num_tokens": 38810381.0,
      "step": 17710
    },
    {
      "entropy": 5.517021036148071,
      "epoch": 1.7581381500595474,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046953961046396767,
      "loss": 5.0666,
      "mean_token_accuracy": 0.17989593893289565,
      "num_tokens": 38821042.0,
      "step": 17715
    },
    {
      "entropy": 5.533328914642334,
      "epoch": 1.7586343787217151,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046952180636398793,
      "loss": 5.0225,
      "mean_token_accuracy": 0.18065762966871263,
      "num_tokens": 38830890.0,
      "step": 17720
    },
    {
      "entropy": 5.610910177230835,
      "epoch": 1.7591306073838826,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004695039974402866,
      "loss": 5.0669,
      "mean_token_accuracy": 0.18119377344846727,
      "num_tokens": 38842648.0,
      "step": 17725
    },
    {
      "entropy": 5.568660163879395,
      "epoch": 1.75962683604605,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046948618369330533,
      "loss": 4.9973,
      "mean_token_accuracy": 0.1850314885377884,
      "num_tokens": 38852740.0,
      "step": 17730
    },
    {
      "entropy": 5.570417070388794,
      "epoch": 1.7601230647082176,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046946836512348566,
      "loss": 5.1438,
      "mean_token_accuracy": 0.16778323352336882,
      "num_tokens": 38863371.0,
      "step": 17735
    },
    {
      "entropy": 5.610669279098511,
      "epoch": 1.760619293370385,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046945054173126965,
      "loss": 5.12,
      "mean_token_accuracy": 0.17228958755731583,
      "num_tokens": 38873893.0,
      "step": 17740
    },
    {
      "entropy": 5.632155275344848,
      "epoch": 1.7611155220325525,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046943271351709923,
      "loss": 5.1545,
      "mean_token_accuracy": 0.17029665261507035,
      "num_tokens": 38884778.0,
      "step": 17745
    },
    {
      "entropy": 5.537489128112793,
      "epoch": 1.76161175069472,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004694148804814164,
      "loss": 5.1075,
      "mean_token_accuracy": 0.1714395135641098,
      "num_tokens": 38896011.0,
      "step": 17750
    },
    {
      "entropy": 5.602264308929444,
      "epoch": 1.7621079793568877,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004693970426246635,
      "loss": 5.1551,
      "mean_token_accuracy": 0.17127641141414643,
      "num_tokens": 38907686.0,
      "step": 17755
    },
    {
      "entropy": 5.553537607192993,
      "epoch": 1.7626042080190552,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046937919994728287,
      "loss": 5.1029,
      "mean_token_accuracy": 0.17546151131391524,
      "num_tokens": 38919180.0,
      "step": 17760
    },
    {
      "entropy": 5.579212665557861,
      "epoch": 1.7631004366812228,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004693613524497168,
      "loss": 5.076,
      "mean_token_accuracy": 0.17489345818758012,
      "num_tokens": 38929600.0,
      "step": 17765
    },
    {
      "entropy": 5.5715985774993895,
      "epoch": 1.7635966653433903,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046934350013240814,
      "loss": 5.021,
      "mean_token_accuracy": 0.18232827633619308,
      "num_tokens": 38940404.0,
      "step": 17770
    },
    {
      "entropy": 5.5603346824646,
      "epoch": 1.7640928940055578,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004693256429957994,
      "loss": 5.0811,
      "mean_token_accuracy": 0.17613358199596404,
      "num_tokens": 38951929.0,
      "step": 17775
    },
    {
      "entropy": 5.534526729583741,
      "epoch": 1.7645891226677253,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004693077810403335,
      "loss": 5.0734,
      "mean_token_accuracy": 0.17858014702796937,
      "num_tokens": 38961954.0,
      "step": 17780
    },
    {
      "entropy": 5.535514068603516,
      "epoch": 1.7650853513298927,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046928991426645325,
      "loss": 5.0505,
      "mean_token_accuracy": 0.17955663800239563,
      "num_tokens": 38970974.0,
      "step": 17785
    },
    {
      "entropy": 5.584988069534302,
      "epoch": 1.7655815799920602,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046927204267460177,
      "loss": 5.144,
      "mean_token_accuracy": 0.1713603913784027,
      "num_tokens": 38982481.0,
      "step": 17790
    },
    {
      "entropy": 5.633728075027466,
      "epoch": 1.766077808654228,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046925416626522235,
      "loss": 5.0904,
      "mean_token_accuracy": 0.1778990387916565,
      "num_tokens": 38993279.0,
      "step": 17795
    },
    {
      "entropy": 5.548814296722412,
      "epoch": 1.7665740373163954,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004692362850387581,
      "loss": 5.0682,
      "mean_token_accuracy": 0.1794086918234825,
      "num_tokens": 39003748.0,
      "step": 17800
    },
    {
      "entropy": 5.534818410873413,
      "epoch": 1.7670702659785629,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046921839899565263,
      "loss": 5.0935,
      "mean_token_accuracy": 0.16957922130823136,
      "num_tokens": 39014902.0,
      "step": 17805
    },
    {
      "entropy": 5.603226280212402,
      "epoch": 1.7675664946407306,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004692005081363493,
      "loss": 5.1605,
      "mean_token_accuracy": 0.16899473965168,
      "num_tokens": 39026484.0,
      "step": 17810
    },
    {
      "entropy": 5.582795286178589,
      "epoch": 1.768062723302898,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046918261246129176,
      "loss": 5.1471,
      "mean_token_accuracy": 0.17439338117837905,
      "num_tokens": 39036268.0,
      "step": 17815
    },
    {
      "entropy": 5.632704877853394,
      "epoch": 1.7685589519650655,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000469164711970924,
      "loss": 5.1968,
      "mean_token_accuracy": 0.17241257429122925,
      "num_tokens": 39049092.0,
      "step": 17820
    },
    {
      "entropy": 5.6514307975769045,
      "epoch": 1.769055180627233,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004691468066656896,
      "loss": 5.1267,
      "mean_token_accuracy": 0.1702898323535919,
      "num_tokens": 39059255.0,
      "step": 17825
    },
    {
      "entropy": 5.550375080108642,
      "epoch": 1.7695514092894005,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046912889654603286,
      "loss": 5.0341,
      "mean_token_accuracy": 0.1851910710334778,
      "num_tokens": 39069913.0,
      "step": 17830
    },
    {
      "entropy": 5.587201547622681,
      "epoch": 1.770047637951568,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004691109816123977,
      "loss": 5.0603,
      "mean_token_accuracy": 0.18034545630216597,
      "num_tokens": 39080666.0,
      "step": 17835
    },
    {
      "entropy": 5.620095872879029,
      "epoch": 1.7705438666137356,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004690930618652286,
      "loss": 5.1433,
      "mean_token_accuracy": 0.171503709256649,
      "num_tokens": 39091562.0,
      "step": 17840
    },
    {
      "entropy": 5.638610601425171,
      "epoch": 1.7710400952759031,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046907513730496975,
      "loss": 5.1998,
      "mean_token_accuracy": 0.16546449512243272,
      "num_tokens": 39102567.0,
      "step": 17845
    },
    {
      "entropy": 5.6569164276123045,
      "epoch": 1.7715363239380708,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004690572079320657,
      "loss": 5.0975,
      "mean_token_accuracy": 0.17872834503650664,
      "num_tokens": 39113260.0,
      "step": 17850
    },
    {
      "entropy": 5.592140865325928,
      "epoch": 1.7720325526002383,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046903927374696106,
      "loss": 5.1248,
      "mean_token_accuracy": 0.17608948349952697,
      "num_tokens": 39125790.0,
      "step": 17855
    },
    {
      "entropy": 5.58499321937561,
      "epoch": 1.7725287812624058,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004690213347501005,
      "loss": 5.1821,
      "mean_token_accuracy": 0.16590221971273422,
      "num_tokens": 39137220.0,
      "step": 17860
    },
    {
      "entropy": 5.574153995513916,
      "epoch": 1.7730250099245732,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004690033909419289,
      "loss": 5.0412,
      "mean_token_accuracy": 0.17935698330402375,
      "num_tokens": 39146362.0,
      "step": 17865
    },
    {
      "entropy": 5.6109143733978275,
      "epoch": 1.7735212385867407,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046898544232289126,
      "loss": 5.1319,
      "mean_token_accuracy": 0.1734836846590042,
      "num_tokens": 39156822.0,
      "step": 17870
    },
    {
      "entropy": 5.650690364837646,
      "epoch": 1.7740174672489082,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046896748889343267,
      "loss": 5.1081,
      "mean_token_accuracy": 0.16789450347423554,
      "num_tokens": 39166693.0,
      "step": 17875
    },
    {
      "entropy": 5.536057806015014,
      "epoch": 1.7745136959110757,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046894953065399836,
      "loss": 5.0864,
      "mean_token_accuracy": 0.17817210108041764,
      "num_tokens": 39176889.0,
      "step": 17880
    },
    {
      "entropy": 5.579395341873169,
      "epoch": 1.7750099245732434,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046893156760503357,
      "loss": 5.0707,
      "mean_token_accuracy": 0.18512479811906815,
      "num_tokens": 39188077.0,
      "step": 17885
    },
    {
      "entropy": 5.536334562301636,
      "epoch": 1.7755061532354108,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004689135997469838,
      "loss": 5.0736,
      "mean_token_accuracy": 0.17638993114233018,
      "num_tokens": 39198182.0,
      "step": 17890
    },
    {
      "entropy": 5.615090847015381,
      "epoch": 1.7760023818975785,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004688956270802945,
      "loss": 5.0473,
      "mean_token_accuracy": 0.18138537555933,
      "num_tokens": 39208801.0,
      "step": 17895
    },
    {
      "entropy": 5.472872066497803,
      "epoch": 1.776498610559746,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004688776496054117,
      "loss": 4.9308,
      "mean_token_accuracy": 0.18689342588186264,
      "num_tokens": 39220658.0,
      "step": 17900
    },
    {
      "entropy": 5.550466251373291,
      "epoch": 1.7769948392219135,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046885966732278077,
      "loss": 5.1825,
      "mean_token_accuracy": 0.17069241553544998,
      "num_tokens": 39231987.0,
      "step": 17905
    },
    {
      "entropy": 5.599190139770508,
      "epoch": 1.777491067884081,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046884168023284787,
      "loss": 5.0625,
      "mean_token_accuracy": 0.17555293291807175,
      "num_tokens": 39242797.0,
      "step": 17910
    },
    {
      "entropy": 5.618773365020752,
      "epoch": 1.7779872965462484,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046882368833605894,
      "loss": 5.1045,
      "mean_token_accuracy": 0.17858345210552215,
      "num_tokens": 39253621.0,
      "step": 17915
    },
    {
      "entropy": 5.546601104736328,
      "epoch": 1.778483525208416,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046880569163286023,
      "loss": 5.1303,
      "mean_token_accuracy": 0.17442435324192046,
      "num_tokens": 39264126.0,
      "step": 17920
    },
    {
      "entropy": 5.580734443664551,
      "epoch": 1.7789797538705836,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046878769012369807,
      "loss": 5.0656,
      "mean_token_accuracy": 0.18141997009515762,
      "num_tokens": 39275118.0,
      "step": 17925
    },
    {
      "entropy": 5.594839954376221,
      "epoch": 1.779475982532751,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046876968380901864,
      "loss": 5.085,
      "mean_token_accuracy": 0.17664961218833924,
      "num_tokens": 39287237.0,
      "step": 17930
    },
    {
      "entropy": 5.653170347213745,
      "epoch": 1.7799722111949188,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046875167268926865,
      "loss": 5.1354,
      "mean_token_accuracy": 0.17291293740272523,
      "num_tokens": 39296995.0,
      "step": 17935
    },
    {
      "entropy": 5.541932916641235,
      "epoch": 1.7804684398570862,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004687336567648947,
      "loss": 5.0279,
      "mean_token_accuracy": 0.17631857991218566,
      "num_tokens": 39305980.0,
      "step": 17940
    },
    {
      "entropy": 5.597739362716675,
      "epoch": 1.7809646685192537,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046871563603634347,
      "loss": 5.1632,
      "mean_token_accuracy": 0.1635955035686493,
      "num_tokens": 39316494.0,
      "step": 17945
    },
    {
      "entropy": 5.6254558086395265,
      "epoch": 1.7814608971814212,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004686976105040619,
      "loss": 5.1342,
      "mean_token_accuracy": 0.1767813429236412,
      "num_tokens": 39327699.0,
      "step": 17950
    },
    {
      "entropy": 5.6155847072601315,
      "epoch": 1.7819571258435887,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004686795801684969,
      "loss": 5.1476,
      "mean_token_accuracy": 0.17194968163967134,
      "num_tokens": 39338749.0,
      "step": 17955
    },
    {
      "entropy": 5.559019899368286,
      "epoch": 1.7824533545057561,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004686615450300958,
      "loss": 5.1354,
      "mean_token_accuracy": 0.17672545164823533,
      "num_tokens": 39349747.0,
      "step": 17960
    },
    {
      "entropy": 5.52283730506897,
      "epoch": 1.7829495831679236,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004686435050893056,
      "loss": 4.9833,
      "mean_token_accuracy": 0.1863816812634468,
      "num_tokens": 39360068.0,
      "step": 17965
    },
    {
      "entropy": 5.550748205184936,
      "epoch": 1.7834458118300913,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046862546034657376,
      "loss": 5.0167,
      "mean_token_accuracy": 0.17877162396907806,
      "num_tokens": 39369651.0,
      "step": 17970
    },
    {
      "entropy": 5.612965822219849,
      "epoch": 1.7839420404922588,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004686074108023477,
      "loss": 5.2256,
      "mean_token_accuracy": 0.16927145123481752,
      "num_tokens": 39381354.0,
      "step": 17975
    },
    {
      "entropy": 5.614839506149292,
      "epoch": 1.7844382691544265,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046858935645707506,
      "loss": 5.0689,
      "mean_token_accuracy": 0.17666124701499938,
      "num_tokens": 39391793.0,
      "step": 17980
    },
    {
      "entropy": 5.580773162841797,
      "epoch": 1.784934497816594,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004685712973112035,
      "loss": 5.0849,
      "mean_token_accuracy": 0.18061139136552812,
      "num_tokens": 39402194.0,
      "step": 17985
    },
    {
      "entropy": 5.55189528465271,
      "epoch": 1.7854307264787614,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004685532333651809,
      "loss": 5.092,
      "mean_token_accuracy": 0.17617464661598206,
      "num_tokens": 39413193.0,
      "step": 17990
    },
    {
      "entropy": 5.543633079528808,
      "epoch": 1.785926955140929,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046853516461945523,
      "loss": 5.0616,
      "mean_token_accuracy": 0.18054229021072388,
      "num_tokens": 39423072.0,
      "step": 17995
    },
    {
      "entropy": 5.586125230789184,
      "epoch": 1.7864231838030964,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046851709107447446,
      "loss": 5.1833,
      "mean_token_accuracy": 0.16527682691812515,
      "num_tokens": 39435088.0,
      "step": 18000
    },
    {
      "epoch": 1.7864231838030964,
      "eval_entropy": 5.386404513066555,
      "eval_loss": 5.2105607986450195,
      "eval_mean_token_accuracy": 0.17518088955677344,
      "eval_num_tokens": 39435088.0,
      "eval_runtime": 20.7082,
      "eval_samples_per_second": 1625.878,
      "eval_steps_per_second": 203.253,
      "step": 18000
    },
    {
      "entropy": 5.544562768936157,
      "epoch": 1.7869194124652639,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004684990127306868,
      "loss": 5.0472,
      "mean_token_accuracy": 0.1758925125002861,
      "num_tokens": 39446129.0,
      "step": 18005
    },
    {
      "entropy": 5.5608758449554445,
      "epoch": 1.7874156411274316,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046848092958854065,
      "loss": 5.0152,
      "mean_token_accuracy": 0.18972333371639252,
      "num_tokens": 39458127.0,
      "step": 18010
    },
    {
      "entropy": 5.564235877990723,
      "epoch": 1.787911869789599,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004684628416484843,
      "loss": 5.0606,
      "mean_token_accuracy": 0.1836364284157753,
      "num_tokens": 39468204.0,
      "step": 18015
    },
    {
      "entropy": 5.56634750366211,
      "epoch": 1.7884080984517667,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004684447489109663,
      "loss": 5.1261,
      "mean_token_accuracy": 0.17259711325168609,
      "num_tokens": 39479114.0,
      "step": 18020
    },
    {
      "entropy": 5.576540279388428,
      "epoch": 1.7889043271139342,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046842665137643546,
      "loss": 5.1094,
      "mean_token_accuracy": 0.17541175037622453,
      "num_tokens": 39489691.0,
      "step": 18025
    },
    {
      "entropy": 5.613377618789673,
      "epoch": 1.7894005557761017,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046840854904534046,
      "loss": 5.2138,
      "mean_token_accuracy": 0.16347745060920715,
      "num_tokens": 39501879.0,
      "step": 18030
    },
    {
      "entropy": 5.62860598564148,
      "epoch": 1.7898967844382692,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004683904419181302,
      "loss": 5.1327,
      "mean_token_accuracy": 0.17551836967468262,
      "num_tokens": 39513413.0,
      "step": 18035
    },
    {
      "entropy": 5.613643598556519,
      "epoch": 1.7903930131004366,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046837232999525366,
      "loss": 5.07,
      "mean_token_accuracy": 0.1751037210226059,
      "num_tokens": 39522719.0,
      "step": 18040
    },
    {
      "entropy": 5.598378133773804,
      "epoch": 1.7908892417626041,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046835421327716,
      "loss": 5.1326,
      "mean_token_accuracy": 0.17514611035585403,
      "num_tokens": 39532463.0,
      "step": 18045
    },
    {
      "entropy": 5.574982833862305,
      "epoch": 1.7913854704247716,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004683360917642986,
      "loss": 5.1011,
      "mean_token_accuracy": 0.17126934826374055,
      "num_tokens": 39543005.0,
      "step": 18050
    },
    {
      "entropy": 5.645137357711792,
      "epoch": 1.7918816990869393,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046831796545711865,
      "loss": 5.1528,
      "mean_token_accuracy": 0.17088039815425873,
      "num_tokens": 39553468.0,
      "step": 18055
    },
    {
      "entropy": 5.684393453598022,
      "epoch": 1.7923779277491068,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046829983435606973,
      "loss": 5.2145,
      "mean_token_accuracy": 0.1698442742228508,
      "num_tokens": 39563539.0,
      "step": 18060
    },
    {
      "entropy": 5.629439640045166,
      "epoch": 1.7928741564112745,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004682816984616014,
      "loss": 5.1491,
      "mean_token_accuracy": 0.1740817591547966,
      "num_tokens": 39574835.0,
      "step": 18065
    },
    {
      "entropy": 5.626969194412231,
      "epoch": 1.793370385073442,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046826355777416346,
      "loss": 5.1038,
      "mean_token_accuracy": 0.17627165466547012,
      "num_tokens": 39586021.0,
      "step": 18070
    },
    {
      "entropy": 5.616800928115845,
      "epoch": 1.7938666137356094,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046824541229420583,
      "loss": 5.0262,
      "mean_token_accuracy": 0.1750480890274048,
      "num_tokens": 39596333.0,
      "step": 18075
    },
    {
      "entropy": 5.573977327346801,
      "epoch": 1.7943628423977769,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046822726202217825,
      "loss": 5.0737,
      "mean_token_accuracy": 0.17931572049856187,
      "num_tokens": 39607553.0,
      "step": 18080
    },
    {
      "entropy": 5.558435535430908,
      "epoch": 1.7948590710599444,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000468209106958531,
      "loss": 5.1557,
      "mean_token_accuracy": 0.17013917714357377,
      "num_tokens": 39619700.0,
      "step": 18085
    },
    {
      "entropy": 5.585485363006592,
      "epoch": 1.7953552997221118,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004681909471037142,
      "loss": 5.0656,
      "mean_token_accuracy": 0.1767073929309845,
      "num_tokens": 39630623.0,
      "step": 18090
    },
    {
      "entropy": 5.605569648742676,
      "epoch": 1.7958515283842795,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046817278245817815,
      "loss": 5.071,
      "mean_token_accuracy": 0.17898934483528137,
      "num_tokens": 39641089.0,
      "step": 18095
    },
    {
      "entropy": 5.570325088500977,
      "epoch": 1.796347757046447,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004681546130223735,
      "loss": 5.1137,
      "mean_token_accuracy": 0.17296559512615203,
      "num_tokens": 39652381.0,
      "step": 18100
    },
    {
      "entropy": 5.559229326248169,
      "epoch": 1.7968439857086145,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004681364387967505,
      "loss": 5.0439,
      "mean_token_accuracy": 0.17878108620643615,
      "num_tokens": 39662536.0,
      "step": 18105
    },
    {
      "entropy": 5.581470918655396,
      "epoch": 1.7973402143707822,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004681182597817601,
      "loss": 5.0864,
      "mean_token_accuracy": 0.18230085521936418,
      "num_tokens": 39673140.0,
      "step": 18110
    },
    {
      "entropy": 5.54479284286499,
      "epoch": 1.7978364430329496,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004681000759778529,
      "loss": 5.1016,
      "mean_token_accuracy": 0.170714570581913,
      "num_tokens": 39683843.0,
      "step": 18115
    },
    {
      "entropy": 5.5883275985717775,
      "epoch": 1.7983326716951171,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004680818873854799,
      "loss": 5.052,
      "mean_token_accuracy": 0.17596071511507033,
      "num_tokens": 39694178.0,
      "step": 18120
    },
    {
      "entropy": 5.526991939544677,
      "epoch": 1.7988289003572846,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004680636940050922,
      "loss": 5.0461,
      "mean_token_accuracy": 0.17769542783498765,
      "num_tokens": 39704818.0,
      "step": 18125
    },
    {
      "entropy": 5.557652997970581,
      "epoch": 1.799325129019452,
      "grad_norm": 1.0,
      "learning_rate": 0.000468045495837141,
      "loss": 5.0655,
      "mean_token_accuracy": 0.17392921894788743,
      "num_tokens": 39715599.0,
      "step": 18130
    },
    {
      "entropy": 5.611971616744995,
      "epoch": 1.7998213576816195,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046802729288207735,
      "loss": 5.0692,
      "mean_token_accuracy": 0.1791368991136551,
      "num_tokens": 39725736.0,
      "step": 18135
    },
    {
      "entropy": 5.582448720932007,
      "epoch": 1.8003175863437872,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046800908514035285,
      "loss": 5.1163,
      "mean_token_accuracy": 0.17034970968961716,
      "num_tokens": 39735736.0,
      "step": 18140
    },
    {
      "entropy": 5.589652633666992,
      "epoch": 1.8008138150059547,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046799087261241884,
      "loss": 5.1457,
      "mean_token_accuracy": 0.17352526634931564,
      "num_tokens": 39747443.0,
      "step": 18145
    },
    {
      "entropy": 5.602119207382202,
      "epoch": 1.8013100436681224,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004679726552987271,
      "loss": 5.0854,
      "mean_token_accuracy": 0.17739440351724625,
      "num_tokens": 39758128.0,
      "step": 18150
    },
    {
      "entropy": 5.57725567817688,
      "epoch": 1.80180627233029,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004679544331997294,
      "loss": 5.0213,
      "mean_token_accuracy": 0.18451550006866455,
      "num_tokens": 39769153.0,
      "step": 18155
    },
    {
      "entropy": 5.592140245437622,
      "epoch": 1.8023025009924574,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046793620631587743,
      "loss": 5.1228,
      "mean_token_accuracy": 0.16787492334842682,
      "num_tokens": 39779666.0,
      "step": 18160
    },
    {
      "entropy": 5.5849864959716795,
      "epoch": 1.8027987296546248,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046791797464762334,
      "loss": 5.1362,
      "mean_token_accuracy": 0.16851571798324586,
      "num_tokens": 39790329.0,
      "step": 18165
    },
    {
      "entropy": 5.627794218063355,
      "epoch": 1.8032949583167923,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046789973819541915,
      "loss": 5.1115,
      "mean_token_accuracy": 0.1759775310754776,
      "num_tokens": 39801996.0,
      "step": 18170
    },
    {
      "entropy": 5.640277290344239,
      "epoch": 1.8037911869789598,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004678814969597171,
      "loss": 5.0994,
      "mean_token_accuracy": 0.17482159286737442,
      "num_tokens": 39814064.0,
      "step": 18175
    },
    {
      "entropy": 5.570799160003662,
      "epoch": 1.8042874156411273,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004678632509409696,
      "loss": 5.0485,
      "mean_token_accuracy": 0.1783964067697525,
      "num_tokens": 39825369.0,
      "step": 18180
    },
    {
      "entropy": 5.51511812210083,
      "epoch": 1.804783644303295,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000467845000139629,
      "loss": 5.1464,
      "mean_token_accuracy": 0.17741038650274277,
      "num_tokens": 39835218.0,
      "step": 18185
    },
    {
      "entropy": 5.64409646987915,
      "epoch": 1.8052798729654624,
      "grad_norm": 1.078125,
      "learning_rate": 0.000467826744556148,
      "loss": 5.152,
      "mean_token_accuracy": 0.17046817839145662,
      "num_tokens": 39844706.0,
      "step": 18190
    },
    {
      "entropy": 5.534137392044068,
      "epoch": 1.8057761016276301,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004678084841909792,
      "loss": 5.0787,
      "mean_token_accuracy": 0.17841877788305283,
      "num_tokens": 39856348.0,
      "step": 18195
    },
    {
      "entropy": 5.553168678283692,
      "epoch": 1.8062723302897976,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046779021904457555,
      "loss": 5.1314,
      "mean_token_accuracy": 0.1764230966567993,
      "num_tokens": 39866850.0,
      "step": 18200
    },
    {
      "entropy": 5.620355749130249,
      "epoch": 1.806768558951965,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004677719491173898,
      "loss": 5.1494,
      "mean_token_accuracy": 0.1748640462756157,
      "num_tokens": 39877647.0,
      "step": 18205
    },
    {
      "entropy": 5.61283917427063,
      "epoch": 1.8072647876141326,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046775367440987506,
      "loss": 5.1914,
      "mean_token_accuracy": 0.1739193767309189,
      "num_tokens": 39889172.0,
      "step": 18210
    },
    {
      "entropy": 5.610627889633179,
      "epoch": 1.8077610162763,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004677353949224846,
      "loss": 5.0989,
      "mean_token_accuracy": 0.17959760129451752,
      "num_tokens": 39900445.0,
      "step": 18215
    },
    {
      "entropy": 5.601744890213013,
      "epoch": 1.8082572449384675,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046771711065567167,
      "loss": 5.1269,
      "mean_token_accuracy": 0.17249710559844972,
      "num_tokens": 39912825.0,
      "step": 18220
    },
    {
      "entropy": 5.524869203567505,
      "epoch": 1.8087534736006352,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004676988216098896,
      "loss": 5.0641,
      "mean_token_accuracy": 0.18331356942653657,
      "num_tokens": 39923370.0,
      "step": 18225
    },
    {
      "entropy": 5.6290013790130615,
      "epoch": 1.8092497022628027,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004676805277855921,
      "loss": 5.1382,
      "mean_token_accuracy": 0.1756781220436096,
      "num_tokens": 39934724.0,
      "step": 18230
    },
    {
      "entropy": 5.548930931091308,
      "epoch": 1.8097459309249704,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004676622291832326,
      "loss": 5.0704,
      "mean_token_accuracy": 0.17236922234296798,
      "num_tokens": 39947064.0,
      "step": 18235
    },
    {
      "entropy": 5.5184242725372314,
      "epoch": 1.8102421595871379,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046764392580326506,
      "loss": 5.0672,
      "mean_token_accuracy": 0.17813999205827713,
      "num_tokens": 39957649.0,
      "step": 18240
    },
    {
      "entropy": 5.5535906791687015,
      "epoch": 1.8107383882493053,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046762561764614316,
      "loss": 5.0976,
      "mean_token_accuracy": 0.17241887301206588,
      "num_tokens": 39968086.0,
      "step": 18245
    },
    {
      "entropy": 5.582545328140259,
      "epoch": 1.8112346169114728,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000467607304712321,
      "loss": 5.0511,
      "mean_token_accuracy": 0.17834966629743576,
      "num_tokens": 39978794.0,
      "step": 18250
    },
    {
      "entropy": 5.617468595504761,
      "epoch": 1.8117308455736403,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046758898700225283,
      "loss": 5.0753,
      "mean_token_accuracy": 0.1803748846054077,
      "num_tokens": 39990383.0,
      "step": 18255
    },
    {
      "entropy": 5.595078563690185,
      "epoch": 1.8122270742358078,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004675706645163927,
      "loss": 5.1516,
      "mean_token_accuracy": 0.17383835464715958,
      "num_tokens": 40001900.0,
      "step": 18260
    },
    {
      "entropy": 5.592684268951416,
      "epoch": 1.8127233028979752,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000467552337255195,
      "loss": 5.0745,
      "mean_token_accuracy": 0.17622508406639098,
      "num_tokens": 40012482.0,
      "step": 18265
    },
    {
      "entropy": 5.535282278060913,
      "epoch": 1.813219531560143,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004675340052191144,
      "loss": 5.0867,
      "mean_token_accuracy": 0.1756088390946388,
      "num_tokens": 40023264.0,
      "step": 18270
    },
    {
      "entropy": 5.607410144805908,
      "epoch": 1.8137157602223104,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004675156684086052,
      "loss": 5.1448,
      "mean_token_accuracy": 0.17269063293933867,
      "num_tokens": 40034754.0,
      "step": 18275
    },
    {
      "entropy": 5.552013301849366,
      "epoch": 1.814211988884478,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004674973268241224,
      "loss": 5.0621,
      "mean_token_accuracy": 0.18001486212015153,
      "num_tokens": 40045781.0,
      "step": 18280
    },
    {
      "entropy": 5.6179920673370365,
      "epoch": 1.8147082175466456,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004674789804661205,
      "loss": 5.1485,
      "mean_token_accuracy": 0.17380132526159286,
      "num_tokens": 40055700.0,
      "step": 18285
    },
    {
      "entropy": 5.616312074661255,
      "epoch": 1.815204446208813,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046746062933505477,
      "loss": 5.1669,
      "mean_token_accuracy": 0.17142882198095322,
      "num_tokens": 40066702.0,
      "step": 18290
    },
    {
      "entropy": 5.593391799926758,
      "epoch": 1.8157006748709805,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004674422734313801,
      "loss": 5.1609,
      "mean_token_accuracy": 0.17785728722810745,
      "num_tokens": 40077856.0,
      "step": 18295
    },
    {
      "entropy": 5.539958477020264,
      "epoch": 1.816196903533148,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046742391275555176,
      "loss": 5.0528,
      "mean_token_accuracy": 0.17769996225833892,
      "num_tokens": 40089205.0,
      "step": 18300
    },
    {
      "entropy": 5.6347935676574705,
      "epoch": 1.8166931321953155,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000467405547308025,
      "loss": 5.1083,
      "mean_token_accuracy": 0.17130195796489717,
      "num_tokens": 40100028.0,
      "step": 18305
    },
    {
      "entropy": 5.667317581176758,
      "epoch": 1.8171893608574832,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004673871770892553,
      "loss": 5.122,
      "mean_token_accuracy": 0.17141395360231398,
      "num_tokens": 40109635.0,
      "step": 18310
    },
    {
      "entropy": 5.614431476593017,
      "epoch": 1.8176855895196506,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046736880209969814,
      "loss": 5.0958,
      "mean_token_accuracy": 0.1749832510948181,
      "num_tokens": 40120539.0,
      "step": 18315
    },
    {
      "entropy": 5.598215484619141,
      "epoch": 1.8181818181818183,
      "grad_norm": 1.125,
      "learning_rate": 0.00046735042233980923,
      "loss": 5.1485,
      "mean_token_accuracy": 0.16719589829444886,
      "num_tokens": 40131738.0,
      "step": 18320
    },
    {
      "entropy": 5.678369617462158,
      "epoch": 1.8186780468439858,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046733203781004426,
      "loss": 5.2068,
      "mean_token_accuracy": 0.17447738647460936,
      "num_tokens": 40142800.0,
      "step": 18325
    },
    {
      "entropy": 5.661719226837159,
      "epoch": 1.8191742755061533,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046731364851085923,
      "loss": 5.163,
      "mean_token_accuracy": 0.16480324566364288,
      "num_tokens": 40153992.0,
      "step": 18330
    },
    {
      "entropy": 5.588845539093017,
      "epoch": 1.8196705041683208,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004672952544427101,
      "loss": 5.1008,
      "mean_token_accuracy": 0.17264530658721924,
      "num_tokens": 40165213.0,
      "step": 18335
    },
    {
      "entropy": 5.598650074005127,
      "epoch": 1.8201667328304882,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046727685560605305,
      "loss": 5.1607,
      "mean_token_accuracy": 0.1722271367907524,
      "num_tokens": 40176763.0,
      "step": 18340
    },
    {
      "entropy": 5.594415283203125,
      "epoch": 1.8206629614926557,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004672584520013443,
      "loss": 5.0484,
      "mean_token_accuracy": 0.1704975873231888,
      "num_tokens": 40187958.0,
      "step": 18345
    },
    {
      "entropy": 5.5416864395141605,
      "epoch": 1.8211591901548232,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004672400436290402,
      "loss": 5.0644,
      "mean_token_accuracy": 0.17911352962255478,
      "num_tokens": 40197881.0,
      "step": 18350
    },
    {
      "entropy": 5.604315662384034,
      "epoch": 1.821655418816991,
      "grad_norm": 1.0,
      "learning_rate": 0.00046722163048959725,
      "loss": 5.0468,
      "mean_token_accuracy": 0.18278829604387284,
      "num_tokens": 40209322.0,
      "step": 18355
    },
    {
      "entropy": 5.6213000297546385,
      "epoch": 1.8221516474791584,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004672032125834721,
      "loss": 5.1157,
      "mean_token_accuracy": 0.17594925463199615,
      "num_tokens": 40222186.0,
      "step": 18360
    },
    {
      "entropy": 5.648015117645263,
      "epoch": 1.822647876141326,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004671847899111214,
      "loss": 5.133,
      "mean_token_accuracy": 0.17236999273300171,
      "num_tokens": 40232136.0,
      "step": 18365
    },
    {
      "entropy": 5.668455457687378,
      "epoch": 1.8231441048034935,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046716636247300206,
      "loss": 5.1724,
      "mean_token_accuracy": 0.1733849138021469,
      "num_tokens": 40244468.0,
      "step": 18370
    },
    {
      "entropy": 5.630852508544922,
      "epoch": 1.823640333465661,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046714793026957096,
      "loss": 5.1208,
      "mean_token_accuracy": 0.17222301214933394,
      "num_tokens": 40255441.0,
      "step": 18375
    },
    {
      "entropy": 5.546484279632568,
      "epoch": 1.8241365621278285,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004671294933012853,
      "loss": 5.0445,
      "mean_token_accuracy": 0.17297716587781906,
      "num_tokens": 40266109.0,
      "step": 18380
    },
    {
      "entropy": 5.5657001495361325,
      "epoch": 1.824632790789996,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046711105156860223,
      "loss": 5.0579,
      "mean_token_accuracy": 0.1781306579709053,
      "num_tokens": 40277635.0,
      "step": 18385
    },
    {
      "entropy": 5.594200897216797,
      "epoch": 1.8251290194521634,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046709260507197896,
      "loss": 5.1958,
      "mean_token_accuracy": 0.17341551780700684,
      "num_tokens": 40288668.0,
      "step": 18390
    },
    {
      "entropy": 5.579392385482788,
      "epoch": 1.8256252481143311,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046707415381187306,
      "loss": 5.0691,
      "mean_token_accuracy": 0.17841619849205018,
      "num_tokens": 40300262.0,
      "step": 18395
    },
    {
      "entropy": 5.597346830368042,
      "epoch": 1.8261214767764986,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000467055697788742,
      "loss": 5.1323,
      "mean_token_accuracy": 0.1668436661362648,
      "num_tokens": 40310931.0,
      "step": 18400
    },
    {
      "entropy": 5.5746301174163815,
      "epoch": 1.826617705438666,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00046703723700304345,
      "loss": 5.0698,
      "mean_token_accuracy": 0.17608420699834823,
      "num_tokens": 40321623.0,
      "step": 18405
    },
    {
      "entropy": 5.532625293731689,
      "epoch": 1.8271139341008338,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004670187714552353,
      "loss": 5.041,
      "mean_token_accuracy": 0.17891584932804108,
      "num_tokens": 40330951.0,
      "step": 18410
    },
    {
      "entropy": 5.610345888137817,
      "epoch": 1.8276101627630013,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046700030114577533,
      "loss": 5.1255,
      "mean_token_accuracy": 0.17349630892276763,
      "num_tokens": 40342971.0,
      "step": 18415
    },
    {
      "entropy": 5.602973508834839,
      "epoch": 1.8281063914251687,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004669818260751217,
      "loss": 5.0901,
      "mean_token_accuracy": 0.17911479920148848,
      "num_tokens": 40353837.0,
      "step": 18420
    },
    {
      "entropy": 5.575761795043945,
      "epoch": 1.8286026200873362,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004669633462437324,
      "loss": 5.1662,
      "mean_token_accuracy": 0.1747234657406807,
      "num_tokens": 40367389.0,
      "step": 18425
    },
    {
      "entropy": 5.567513561248779,
      "epoch": 1.8290988487495037,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046694486165206573,
      "loss": 4.9878,
      "mean_token_accuracy": 0.18163626044988632,
      "num_tokens": 40376793.0,
      "step": 18430
    },
    {
      "entropy": 5.584813404083252,
      "epoch": 1.8295950774116712,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004669263723005801,
      "loss": 5.0944,
      "mean_token_accuracy": 0.17865925282239914,
      "num_tokens": 40388025.0,
      "step": 18435
    },
    {
      "entropy": 5.586469030380249,
      "epoch": 1.8300913060738389,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046690787818973406,
      "loss": 5.1683,
      "mean_token_accuracy": 0.1715688854455948,
      "num_tokens": 40399731.0,
      "step": 18440
    },
    {
      "entropy": 5.603258275985718,
      "epoch": 1.8305875347360063,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046688937931998613,
      "loss": 5.213,
      "mean_token_accuracy": 0.17325540781021118,
      "num_tokens": 40411078.0,
      "step": 18445
    },
    {
      "entropy": 5.67874846458435,
      "epoch": 1.831083763398174,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004668708756917951,
      "loss": 5.1723,
      "mean_token_accuracy": 0.17386017739772797,
      "num_tokens": 40422806.0,
      "step": 18450
    },
    {
      "entropy": 5.70818018913269,
      "epoch": 1.8315799920603415,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046685236730561975,
      "loss": 5.2219,
      "mean_token_accuracy": 0.16328475251793861,
      "num_tokens": 40434041.0,
      "step": 18455
    },
    {
      "entropy": 5.590016412734985,
      "epoch": 1.832076220722509,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046683385416191914,
      "loss": 5.0656,
      "mean_token_accuracy": 0.182546004652977,
      "num_tokens": 40444658.0,
      "step": 18460
    },
    {
      "entropy": 5.5564549446105955,
      "epoch": 1.8325724493846764,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046681533626115226,
      "loss": 5.0824,
      "mean_token_accuracy": 0.18202535212039947,
      "num_tokens": 40455745.0,
      "step": 18465
    },
    {
      "entropy": 5.5758607387542725,
      "epoch": 1.833068678046844,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004667968136037784,
      "loss": 5.1747,
      "mean_token_accuracy": 0.174778713285923,
      "num_tokens": 40468829.0,
      "step": 18470
    },
    {
      "entropy": 5.573695278167724,
      "epoch": 1.8335649067090114,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046677828619025687,
      "loss": 5.0763,
      "mean_token_accuracy": 0.1742417961359024,
      "num_tokens": 40479817.0,
      "step": 18475
    },
    {
      "entropy": 5.697887134552002,
      "epoch": 1.8340611353711789,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046675975402104706,
      "loss": 5.1802,
      "mean_token_accuracy": 0.1714305803179741,
      "num_tokens": 40490398.0,
      "step": 18480
    },
    {
      "entropy": 5.55704174041748,
      "epoch": 1.8345573640333466,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046674121709660855,
      "loss": 5.027,
      "mean_token_accuracy": 0.17976675629615785,
      "num_tokens": 40500483.0,
      "step": 18485
    },
    {
      "entropy": 5.593721628189087,
      "epoch": 1.835053592695514,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000466722675417401,
      "loss": 5.1362,
      "mean_token_accuracy": 0.16754657328128814,
      "num_tokens": 40511494.0,
      "step": 18490
    },
    {
      "entropy": 5.591253042221069,
      "epoch": 1.8355498213576817,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046670412898388425,
      "loss": 5.1071,
      "mean_token_accuracy": 0.18255719691514968,
      "num_tokens": 40523573.0,
      "step": 18495
    },
    {
      "entropy": 5.566313982009888,
      "epoch": 1.8360460500198492,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046668557779651825,
      "loss": 5.1394,
      "mean_token_accuracy": 0.1674652174115181,
      "num_tokens": 40534904.0,
      "step": 18500
    },
    {
      "entropy": 5.564811944961548,
      "epoch": 1.8365422786820167,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046666702185576295,
      "loss": 5.1004,
      "mean_token_accuracy": 0.17924259155988692,
      "num_tokens": 40545053.0,
      "step": 18505
    },
    {
      "entropy": 5.5799479484558105,
      "epoch": 1.8370385073441842,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004666484611620785,
      "loss": 5.0431,
      "mean_token_accuracy": 0.18056851029396057,
      "num_tokens": 40555305.0,
      "step": 18510
    },
    {
      "entropy": 5.578672552108765,
      "epoch": 1.8375347360063516,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004666298957159251,
      "loss": 5.0716,
      "mean_token_accuracy": 0.172512049973011,
      "num_tokens": 40565611.0,
      "step": 18515
    },
    {
      "entropy": 5.594479417800903,
      "epoch": 1.8380309646685191,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046661132551776327,
      "loss": 5.1503,
      "mean_token_accuracy": 0.17654699683189393,
      "num_tokens": 40578321.0,
      "step": 18520
    },
    {
      "entropy": 5.630540657043457,
      "epoch": 1.8385271933306868,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004665927505680535,
      "loss": 5.1533,
      "mean_token_accuracy": 0.18071224838495253,
      "num_tokens": 40588696.0,
      "step": 18525
    },
    {
      "entropy": 5.60497465133667,
      "epoch": 1.8390234219928543,
      "grad_norm": 1.0,
      "learning_rate": 0.00046657417086725625,
      "loss": 5.102,
      "mean_token_accuracy": 0.17446884363889695,
      "num_tokens": 40599699.0,
      "step": 18530
    },
    {
      "entropy": 5.580640411376953,
      "epoch": 1.839519650655022,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004665555864158324,
      "loss": 5.102,
      "mean_token_accuracy": 0.1813617929816246,
      "num_tokens": 40609706.0,
      "step": 18535
    },
    {
      "entropy": 5.556032800674439,
      "epoch": 1.8400158793171895,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004665369972142428,
      "loss": 5.1456,
      "mean_token_accuracy": 0.17661279439926147,
      "num_tokens": 40619431.0,
      "step": 18540
    },
    {
      "entropy": 5.601679754257202,
      "epoch": 1.840512107979357,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046651840326294845,
      "loss": 5.1062,
      "mean_token_accuracy": 0.18073134422302245,
      "num_tokens": 40629904.0,
      "step": 18545
    },
    {
      "entropy": 5.57069354057312,
      "epoch": 1.8410083366415244,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004664998045624104,
      "loss": 5.157,
      "mean_token_accuracy": 0.17323327660560608,
      "num_tokens": 40641266.0,
      "step": 18550
    },
    {
      "entropy": 5.545009708404541,
      "epoch": 1.8415045653036919,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004664812011130897,
      "loss": 5.0937,
      "mean_token_accuracy": 0.17650105953216552,
      "num_tokens": 40652454.0,
      "step": 18555
    },
    {
      "entropy": 5.605793476104736,
      "epoch": 1.8420007939658594,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046646259291544787,
      "loss": 5.0854,
      "mean_token_accuracy": 0.17539707124233245,
      "num_tokens": 40664311.0,
      "step": 18560
    },
    {
      "entropy": 5.532068681716919,
      "epoch": 1.8424970226280268,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004664439799699463,
      "loss": 5.004,
      "mean_token_accuracy": 0.1828952834010124,
      "num_tokens": 40675241.0,
      "step": 18565
    },
    {
      "entropy": 5.552021169662476,
      "epoch": 1.8429932512901945,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046642536227704665,
      "loss": 5.0896,
      "mean_token_accuracy": 0.17558399885892867,
      "num_tokens": 40685972.0,
      "step": 18570
    },
    {
      "entropy": 5.579754734039307,
      "epoch": 1.843489479952362,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046640673983721037,
      "loss": 5.107,
      "mean_token_accuracy": 0.17592190951108932,
      "num_tokens": 40696967.0,
      "step": 18575
    },
    {
      "entropy": 5.617200708389282,
      "epoch": 1.8439857086145297,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00046638811265089944,
      "loss": 5.1145,
      "mean_token_accuracy": 0.17017566561698913,
      "num_tokens": 40707710.0,
      "step": 18580
    },
    {
      "entropy": 5.6368489265441895,
      "epoch": 1.8444819372766972,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004663694807185758,
      "loss": 5.2389,
      "mean_token_accuracy": 0.17188069224357605,
      "num_tokens": 40718586.0,
      "step": 18585
    },
    {
      "entropy": 5.612079811096192,
      "epoch": 1.8449781659388647,
      "grad_norm": 1.0,
      "learning_rate": 0.00046635084404070135,
      "loss": 5.1121,
      "mean_token_accuracy": 0.17986107170581817,
      "num_tokens": 40729792.0,
      "step": 18590
    },
    {
      "entropy": 5.732817029953003,
      "epoch": 1.8454743946010321,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004663322026177382,
      "loss": 5.1609,
      "mean_token_accuracy": 0.16817573755979537,
      "num_tokens": 40740974.0,
      "step": 18595
    },
    {
      "entropy": 5.567403316497803,
      "epoch": 1.8459706232631996,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046631355645014875,
      "loss": 4.9787,
      "mean_token_accuracy": 0.1824137583374977,
      "num_tokens": 40750208.0,
      "step": 18600
    },
    {
      "entropy": 5.587321853637695,
      "epoch": 1.846466851925367,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046629490553839537,
      "loss": 5.0671,
      "mean_token_accuracy": 0.18070756793022155,
      "num_tokens": 40762163.0,
      "step": 18605
    },
    {
      "entropy": 5.586577987670898,
      "epoch": 1.8469630805875348,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004662762498829405,
      "loss": 5.1843,
      "mean_token_accuracy": 0.1748317524790764,
      "num_tokens": 40773028.0,
      "step": 18610
    },
    {
      "entropy": 5.655228185653686,
      "epoch": 1.8474593092497023,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004662575894842469,
      "loss": 5.1426,
      "mean_token_accuracy": 0.16920928359031678,
      "num_tokens": 40784939.0,
      "step": 18615
    },
    {
      "entropy": 5.637517738342285,
      "epoch": 1.84795553791187,
      "grad_norm": 0.828125,
      "learning_rate": 0.00046623892434277713,
      "loss": 5.1341,
      "mean_token_accuracy": 0.1758325845003128,
      "num_tokens": 40796170.0,
      "step": 18620
    },
    {
      "entropy": 5.612541007995605,
      "epoch": 1.8484517665740374,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004662202544589941,
      "loss": 5.0699,
      "mean_token_accuracy": 0.17797799706459044,
      "num_tokens": 40807184.0,
      "step": 18625
    },
    {
      "entropy": 5.543184471130371,
      "epoch": 1.848947995236205,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046620157983336087,
      "loss": 5.0511,
      "mean_token_accuracy": 0.1740154892206192,
      "num_tokens": 40818474.0,
      "step": 18630
    },
    {
      "entropy": 5.588191509246826,
      "epoch": 1.8494442238983724,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004661829004663404,
      "loss": 5.1035,
      "mean_token_accuracy": 0.17895497977733613,
      "num_tokens": 40828789.0,
      "step": 18635
    },
    {
      "entropy": 5.599267864227295,
      "epoch": 1.8499404525605398,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004661642163583961,
      "loss": 5.0689,
      "mean_token_accuracy": 0.18170301467180253,
      "num_tokens": 40840564.0,
      "step": 18640
    },
    {
      "entropy": 5.561771869659424,
      "epoch": 1.8504366812227073,
      "grad_norm": 0.84375,
      "learning_rate": 0.00046614552750999093,
      "loss": 5.0442,
      "mean_token_accuracy": 0.17436972856521607,
      "num_tokens": 40852528.0,
      "step": 18645
    },
    {
      "entropy": 5.571287345886231,
      "epoch": 1.8509329098848748,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004661268339215887,
      "loss": 5.091,
      "mean_token_accuracy": 0.17794181257486344,
      "num_tokens": 40863242.0,
      "step": 18650
    },
    {
      "entropy": 5.553026628494263,
      "epoch": 1.8514291385470425,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004661081355936528,
      "loss": 5.0458,
      "mean_token_accuracy": 0.1845734491944313,
      "num_tokens": 40873721.0,
      "step": 18655
    },
    {
      "entropy": 5.6354176044464115,
      "epoch": 1.85192536720921,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000466089432526647,
      "loss": 5.1425,
      "mean_token_accuracy": 0.17176969349384308,
      "num_tokens": 40884921.0,
      "step": 18660
    },
    {
      "entropy": 5.625229358673096,
      "epoch": 1.8524215958713777,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046607072472103493,
      "loss": 5.0363,
      "mean_token_accuracy": 0.1734216496348381,
      "num_tokens": 40896243.0,
      "step": 18665
    },
    {
      "entropy": 5.602851104736328,
      "epoch": 1.8529178245335451,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046605201217728067,
      "loss": 5.1705,
      "mean_token_accuracy": 0.17141676396131517,
      "num_tokens": 40907191.0,
      "step": 18670
    },
    {
      "entropy": 5.619650077819824,
      "epoch": 1.8534140531957126,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046603329489584817,
      "loss": 5.1055,
      "mean_token_accuracy": 0.18016466200351716,
      "num_tokens": 40918737.0,
      "step": 18675
    },
    {
      "entropy": 5.661403512954712,
      "epoch": 1.85391028185788,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004660145728772016,
      "loss": 5.1328,
      "mean_token_accuracy": 0.1734810382127762,
      "num_tokens": 40930184.0,
      "step": 18680
    },
    {
      "entropy": 5.589530801773071,
      "epoch": 1.8544065105200476,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004659958461218052,
      "loss": 5.0357,
      "mean_token_accuracy": 0.17786586582660674,
      "num_tokens": 40941483.0,
      "step": 18685
    },
    {
      "entropy": 5.559656953811645,
      "epoch": 1.854902739182215,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046597711463012344,
      "loss": 5.1238,
      "mean_token_accuracy": 0.17495235651731492,
      "num_tokens": 40952392.0,
      "step": 18690
    },
    {
      "entropy": 5.554095697402954,
      "epoch": 1.8553989678443827,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004659583784026207,
      "loss": 5.0616,
      "mean_token_accuracy": 0.17617972195148468,
      "num_tokens": 40962539.0,
      "step": 18695
    },
    {
      "entropy": 5.574426460266113,
      "epoch": 1.8558951965065502,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046593963743976163,
      "loss": 4.983,
      "mean_token_accuracy": 0.18216906487941742,
      "num_tokens": 40972594.0,
      "step": 18700
    },
    {
      "entropy": 5.619575357437133,
      "epoch": 1.8563914251687177,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000465920891742011,
      "loss": 5.1029,
      "mean_token_accuracy": 0.17454908937215804,
      "num_tokens": 40983460.0,
      "step": 18705
    },
    {
      "entropy": 5.570920038223266,
      "epoch": 1.8568876538308854,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004659021413098337,
      "loss": 5.1036,
      "mean_token_accuracy": 0.1689835473895073,
      "num_tokens": 40993945.0,
      "step": 18710
    },
    {
      "entropy": 5.496763038635254,
      "epoch": 1.8573838824930529,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046588338614369465,
      "loss": 5.0957,
      "mean_token_accuracy": 0.1768236815929413,
      "num_tokens": 41005781.0,
      "step": 18715
    },
    {
      "entropy": 5.650920677185058,
      "epoch": 1.8578801111552203,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00046586462624405897,
      "loss": 5.2021,
      "mean_token_accuracy": 0.17417456358671188,
      "num_tokens": 41017650.0,
      "step": 18720
    },
    {
      "entropy": 5.624494934082032,
      "epoch": 1.8583763398173878,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004658458616113918,
      "loss": 5.062,
      "mean_token_accuracy": 0.1745037853717804,
      "num_tokens": 41029196.0,
      "step": 18725
    },
    {
      "entropy": 5.5819993019104,
      "epoch": 1.8588725684795553,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004658270922461585,
      "loss": 5.1284,
      "mean_token_accuracy": 0.17880256921052934,
      "num_tokens": 41040204.0,
      "step": 18730
    },
    {
      "entropy": 5.602885055541992,
      "epoch": 1.8593687971417228,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046580831814882446,
      "loss": 5.1145,
      "mean_token_accuracy": 0.16820195317268372,
      "num_tokens": 41050507.0,
      "step": 18735
    },
    {
      "entropy": 5.552184772491455,
      "epoch": 1.8598650258038905,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046578953931985536,
      "loss": 5.0886,
      "mean_token_accuracy": 0.1727803170681,
      "num_tokens": 41062001.0,
      "step": 18740
    },
    {
      "entropy": 5.602700996398926,
      "epoch": 1.860361254466058,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004657707557597167,
      "loss": 5.0905,
      "mean_token_accuracy": 0.173151370882988,
      "num_tokens": 41073097.0,
      "step": 18745
    },
    {
      "entropy": 5.587178134918213,
      "epoch": 1.8608574831282256,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046575196746887447,
      "loss": 5.1202,
      "mean_token_accuracy": 0.1711211994290352,
      "num_tokens": 41083883.0,
      "step": 18750
    },
    {
      "entropy": 5.5267258644104,
      "epoch": 1.861353711790393,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046573317444779445,
      "loss": 5.0857,
      "mean_token_accuracy": 0.18304546773433686,
      "num_tokens": 41095506.0,
      "step": 18755
    },
    {
      "entropy": 5.601643466949463,
      "epoch": 1.8618499404525606,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004657143766969427,
      "loss": 5.1129,
      "mean_token_accuracy": 0.17489507049322128,
      "num_tokens": 41106568.0,
      "step": 18760
    },
    {
      "entropy": 5.62065110206604,
      "epoch": 1.862346169114728,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004656955742167853,
      "loss": 5.0119,
      "mean_token_accuracy": 0.17956110388040541,
      "num_tokens": 41117661.0,
      "step": 18765
    },
    {
      "entropy": 5.571772193908691,
      "epoch": 1.8628423977768955,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046567676700778865,
      "loss": 5.108,
      "mean_token_accuracy": 0.1785398855805397,
      "num_tokens": 41129209.0,
      "step": 18770
    },
    {
      "entropy": 5.544245290756225,
      "epoch": 1.863338626439063,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046565795507041904,
      "loss": 5.1634,
      "mean_token_accuracy": 0.17192599624395372,
      "num_tokens": 41140715.0,
      "step": 18775
    },
    {
      "entropy": 5.588299036026001,
      "epoch": 1.8638348551012305,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000465639138405143,
      "loss": 5.1277,
      "mean_token_accuracy": 0.17507790476083757,
      "num_tokens": 41151080.0,
      "step": 18780
    },
    {
      "entropy": 5.6462499618530275,
      "epoch": 1.8643310837633982,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046562031701242703,
      "loss": 5.1429,
      "mean_token_accuracy": 0.17553326785564421,
      "num_tokens": 41162816.0,
      "step": 18785
    },
    {
      "entropy": 5.601840305328369,
      "epoch": 1.8648273124255657,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046560149089273803,
      "loss": 5.0456,
      "mean_token_accuracy": 0.18238050639629363,
      "num_tokens": 41173456.0,
      "step": 18790
    },
    {
      "entropy": 5.535126876831055,
      "epoch": 1.8653235410877333,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004655826600465427,
      "loss": 4.9971,
      "mean_token_accuracy": 0.18200599998235703,
      "num_tokens": 41184436.0,
      "step": 18795
    },
    {
      "entropy": 5.515432262420655,
      "epoch": 1.8658197697499008,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046556382447430816,
      "loss": 5.0642,
      "mean_token_accuracy": 0.17848023027181625,
      "num_tokens": 41196417.0,
      "step": 18800
    },
    {
      "entropy": 5.5554639339447025,
      "epoch": 1.8663159984120683,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046554498417650134,
      "loss": 5.0807,
      "mean_token_accuracy": 0.18071357309818267,
      "num_tokens": 41206054.0,
      "step": 18805
    },
    {
      "entropy": 5.558531188964844,
      "epoch": 1.8668122270742358,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004655261391535895,
      "loss": 5.0868,
      "mean_token_accuracy": 0.1764499291777611,
      "num_tokens": 41217002.0,
      "step": 18810
    },
    {
      "entropy": 5.68140606880188,
      "epoch": 1.8673084557364032,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004655072894060399,
      "loss": 5.2405,
      "mean_token_accuracy": 0.17107231318950653,
      "num_tokens": 41228485.0,
      "step": 18815
    },
    {
      "entropy": 5.617528676986694,
      "epoch": 1.8678046843985707,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004654884349343201,
      "loss": 5.1207,
      "mean_token_accuracy": 0.17651147693395614,
      "num_tokens": 41240600.0,
      "step": 18820
    },
    {
      "entropy": 5.666401720046997,
      "epoch": 1.8683009130607384,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046546957573889757,
      "loss": 5.2071,
      "mean_token_accuracy": 0.16461690068244933,
      "num_tokens": 41251870.0,
      "step": 18825
    },
    {
      "entropy": 5.5624865055084225,
      "epoch": 1.868797141722906,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046545071182024,
      "loss": 5.1051,
      "mean_token_accuracy": 0.17211880832910537,
      "num_tokens": 41264679.0,
      "step": 18830
    },
    {
      "entropy": 5.6861940860748295,
      "epoch": 1.8692933703850736,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046543184317881516,
      "loss": 5.2077,
      "mean_token_accuracy": 0.17192999869585038,
      "num_tokens": 41275690.0,
      "step": 18835
    },
    {
      "entropy": 5.584121513366699,
      "epoch": 1.869789599047241,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000465412969815091,
      "loss": 5.0748,
      "mean_token_accuracy": 0.18410394191741944,
      "num_tokens": 41285118.0,
      "step": 18840
    },
    {
      "entropy": 5.539427947998047,
      "epoch": 1.8702858277094085,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004653940917295354,
      "loss": 5.0685,
      "mean_token_accuracy": 0.18072286397218704,
      "num_tokens": 41296330.0,
      "step": 18845
    },
    {
      "entropy": 5.5302225112915036,
      "epoch": 1.870782056371576,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004653752089226166,
      "loss": 5.1096,
      "mean_token_accuracy": 0.1731637641787529,
      "num_tokens": 41307051.0,
      "step": 18850
    },
    {
      "entropy": 5.6016796112060545,
      "epoch": 1.8712782850337435,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004653563213948029,
      "loss": 5.0441,
      "mean_token_accuracy": 0.17943293750286102,
      "num_tokens": 41317111.0,
      "step": 18855
    },
    {
      "entropy": 5.54780535697937,
      "epoch": 1.871774513695911,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046533742914656254,
      "loss": 5.0341,
      "mean_token_accuracy": 0.1795681208372116,
      "num_tokens": 41327509.0,
      "step": 18860
    },
    {
      "entropy": 5.565215110778809,
      "epoch": 1.8722707423580784,
      "grad_norm": 1.0,
      "learning_rate": 0.00046531853217836414,
      "loss": 5.1169,
      "mean_token_accuracy": 0.17437928020954133,
      "num_tokens": 41338721.0,
      "step": 18865
    },
    {
      "entropy": 5.6503376960754395,
      "epoch": 1.8727669710202461,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046529963049067623,
      "loss": 5.1635,
      "mean_token_accuracy": 0.17196326106786727,
      "num_tokens": 41349767.0,
      "step": 18870
    },
    {
      "entropy": 5.707150220870972,
      "epoch": 1.8732631996824136,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046528072408396747,
      "loss": 5.2739,
      "mean_token_accuracy": 0.1686161205172539,
      "num_tokens": 41361151.0,
      "step": 18875
    },
    {
      "entropy": 5.606167602539062,
      "epoch": 1.8737594283445813,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004652618129587069,
      "loss": 5.0572,
      "mean_token_accuracy": 0.17533742338418962,
      "num_tokens": 41372875.0,
      "step": 18880
    },
    {
      "entropy": 5.636981725692749,
      "epoch": 1.8742556570067488,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046524289711536327,
      "loss": 5.1313,
      "mean_token_accuracy": 0.17001581341028213,
      "num_tokens": 41383284.0,
      "step": 18885
    },
    {
      "entropy": 5.56423511505127,
      "epoch": 1.8747518856689163,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004652239765544057,
      "loss": 5.095,
      "mean_token_accuracy": 0.17702211886644365,
      "num_tokens": 41395389.0,
      "step": 18890
    },
    {
      "entropy": 5.532351016998291,
      "epoch": 1.8752481143310837,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046520505127630346,
      "loss": 5.0529,
      "mean_token_accuracy": 0.17678306698799134,
      "num_tokens": 41406740.0,
      "step": 18895
    },
    {
      "entropy": 5.5620441913604735,
      "epoch": 1.8757443429932512,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046518612128152576,
      "loss": 5.0566,
      "mean_token_accuracy": 0.17975561022758485,
      "num_tokens": 41417279.0,
      "step": 18900
    },
    {
      "entropy": 5.533911228179932,
      "epoch": 1.8762405716554187,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004651671865705421,
      "loss": 5.0325,
      "mean_token_accuracy": 0.1778583601117134,
      "num_tokens": 41428574.0,
      "step": 18905
    },
    {
      "entropy": 5.547863912582398,
      "epoch": 1.8767368003175864,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046514824714382204,
      "loss": 5.016,
      "mean_token_accuracy": 0.17704529464244842,
      "num_tokens": 41438784.0,
      "step": 18910
    },
    {
      "entropy": 5.571160078048706,
      "epoch": 1.8772330289797539,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004651293030018351,
      "loss": 5.0563,
      "mean_token_accuracy": 0.17957599312067032,
      "num_tokens": 41448481.0,
      "step": 18915
    },
    {
      "entropy": 5.613469028472901,
      "epoch": 1.8777292576419216,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046511035414505115,
      "loss": 5.1139,
      "mean_token_accuracy": 0.1792847380042076,
      "num_tokens": 41459097.0,
      "step": 18920
    },
    {
      "entropy": 5.624731779098511,
      "epoch": 1.878225486304089,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004650914005739401,
      "loss": 5.1583,
      "mean_token_accuracy": 0.17110537886619567,
      "num_tokens": 41469647.0,
      "step": 18925
    },
    {
      "entropy": 5.61043643951416,
      "epoch": 1.8787217149662565,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004650724422889719,
      "loss": 5.0912,
      "mean_token_accuracy": 0.17152559757232666,
      "num_tokens": 41479826.0,
      "step": 18930
    },
    {
      "entropy": 5.652989768981934,
      "epoch": 1.879217943628424,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004650534792906167,
      "loss": 5.0635,
      "mean_token_accuracy": 0.17691353410482408,
      "num_tokens": 41490551.0,
      "step": 18935
    },
    {
      "entropy": 5.563391828536988,
      "epoch": 1.8797141722905915,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004650345115793447,
      "loss": 5.0976,
      "mean_token_accuracy": 0.1800636127591133,
      "num_tokens": 41500908.0,
      "step": 18940
    },
    {
      "entropy": 5.562153673171997,
      "epoch": 1.880210400952759,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046501553915562644,
      "loss": 5.0547,
      "mean_token_accuracy": 0.17708196938037873,
      "num_tokens": 41511311.0,
      "step": 18945
    },
    {
      "entropy": 5.587686824798584,
      "epoch": 1.8807066296149264,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046499656201993213,
      "loss": 5.0834,
      "mean_token_accuracy": 0.18302844613790512,
      "num_tokens": 41522068.0,
      "step": 18950
    },
    {
      "entropy": 5.5821747303009035,
      "epoch": 1.881202858277094,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046497758017273255,
      "loss": 5.0901,
      "mean_token_accuracy": 0.17068951278924943,
      "num_tokens": 41534099.0,
      "step": 18955
    },
    {
      "entropy": 5.615620422363281,
      "epoch": 1.8816990869392616,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046495859361449834,
      "loss": 5.1553,
      "mean_token_accuracy": 0.16817483454942703,
      "num_tokens": 41545870.0,
      "step": 18960
    },
    {
      "entropy": 5.615835571289063,
      "epoch": 1.8821953156014293,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004649396023457003,
      "loss": 5.1463,
      "mean_token_accuracy": 0.1726374641060829,
      "num_tokens": 41558175.0,
      "step": 18965
    },
    {
      "entropy": 5.565816211700439,
      "epoch": 1.8826915442635968,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004649206063668094,
      "loss": 5.1086,
      "mean_token_accuracy": 0.18199608623981475,
      "num_tokens": 41569326.0,
      "step": 18970
    },
    {
      "entropy": 5.580830478668213,
      "epoch": 1.8831877729257642,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004649016056782968,
      "loss": 5.0585,
      "mean_token_accuracy": 0.18058076351881028,
      "num_tokens": 41580065.0,
      "step": 18975
    },
    {
      "entropy": 5.658510637283325,
      "epoch": 1.8836840015879317,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004648826002806334,
      "loss": 5.1964,
      "mean_token_accuracy": 0.17414325326681138,
      "num_tokens": 41590915.0,
      "step": 18980
    },
    {
      "entropy": 5.571745491027832,
      "epoch": 1.8841802302500992,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004648635901742909,
      "loss": 5.0298,
      "mean_token_accuracy": 0.1777343213558197,
      "num_tokens": 41602096.0,
      "step": 18985
    },
    {
      "entropy": 5.597507524490356,
      "epoch": 1.8846764589122667,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004648445753597403,
      "loss": 5.1539,
      "mean_token_accuracy": 0.17134135514497756,
      "num_tokens": 41612589.0,
      "step": 18990
    },
    {
      "entropy": 5.5516462326049805,
      "epoch": 1.8851726875744343,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00046482555583745336,
      "loss": 5.0398,
      "mean_token_accuracy": 0.17693167328834533,
      "num_tokens": 41624163.0,
      "step": 18995
    },
    {
      "entropy": 5.575661897659302,
      "epoch": 1.8856689162366018,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004648065316079017,
      "loss": 5.0544,
      "mean_token_accuracy": 0.17939051389694213,
      "num_tokens": 41635919.0,
      "step": 19000
    },
    {
      "entropy": 5.5722887992858885,
      "epoch": 1.8861651448987695,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046478750267155703,
      "loss": 5.0788,
      "mean_token_accuracy": 0.17667445987462999,
      "num_tokens": 41644723.0,
      "step": 19005
    },
    {
      "entropy": 5.558045053482056,
      "epoch": 1.886661373560937,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00046476846902889125,
      "loss": 5.1347,
      "mean_token_accuracy": 0.17806994616985322,
      "num_tokens": 41657137.0,
      "step": 19010
    },
    {
      "entropy": 5.602681350708008,
      "epoch": 1.8871576022231045,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004647494306803764,
      "loss": 5.1051,
      "mean_token_accuracy": 0.1714751183986664,
      "num_tokens": 41667773.0,
      "step": 19015
    },
    {
      "entropy": 5.659757137298584,
      "epoch": 1.887653830885272,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046473038762648446,
      "loss": 5.1826,
      "mean_token_accuracy": 0.1747130647301674,
      "num_tokens": 41680549.0,
      "step": 19020
    },
    {
      "entropy": 5.591322517395019,
      "epoch": 1.8881500595474394,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046471133986768784,
      "loss": 5.0709,
      "mean_token_accuracy": 0.1799992576241493,
      "num_tokens": 41689700.0,
      "step": 19025
    },
    {
      "entropy": 5.628189039230347,
      "epoch": 1.888646288209607,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046469228740445874,
      "loss": 5.1183,
      "mean_token_accuracy": 0.1775785967707634,
      "num_tokens": 41699734.0,
      "step": 19030
    },
    {
      "entropy": 5.572957134246826,
      "epoch": 1.8891425168717744,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046467323023726965,
      "loss": 5.1175,
      "mean_token_accuracy": 0.17960592359304428,
      "num_tokens": 41710644.0,
      "step": 19035
    },
    {
      "entropy": 5.510110759735108,
      "epoch": 1.889638745533942,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046465416836659313,
      "loss": 5.0201,
      "mean_token_accuracy": 0.18166979104280473,
      "num_tokens": 41721416.0,
      "step": 19040
    },
    {
      "entropy": 5.611979627609253,
      "epoch": 1.8901349741961095,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000464635101792902,
      "loss": 5.0459,
      "mean_token_accuracy": 0.17501352578401566,
      "num_tokens": 41732139.0,
      "step": 19045
    },
    {
      "entropy": 5.602227354049683,
      "epoch": 1.8906312028582772,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046461603051666883,
      "loss": 5.144,
      "mean_token_accuracy": 0.17550608217716218,
      "num_tokens": 41744233.0,
      "step": 19050
    },
    {
      "entropy": 5.578940296173096,
      "epoch": 1.8911274315204447,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046459695453836676,
      "loss": 5.0658,
      "mean_token_accuracy": 0.18193520456552506,
      "num_tokens": 41755514.0,
      "step": 19055
    },
    {
      "entropy": 5.60691089630127,
      "epoch": 1.8916236601826122,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004645778738584688,
      "loss": 5.0887,
      "mean_token_accuracy": 0.180548731982708,
      "num_tokens": 41765613.0,
      "step": 19060
    },
    {
      "entropy": 5.6429048538208,
      "epoch": 1.8921198888447797,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004645587884774481,
      "loss": 5.1372,
      "mean_token_accuracy": 0.17240234613418579,
      "num_tokens": 41777294.0,
      "step": 19065
    },
    {
      "entropy": 5.539595603942871,
      "epoch": 1.8926161175069471,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004645396983957778,
      "loss": 5.0374,
      "mean_token_accuracy": 0.1821464404463768,
      "num_tokens": 41788278.0,
      "step": 19070
    },
    {
      "entropy": 5.596460676193237,
      "epoch": 1.8931123461691146,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046452060361393147,
      "loss": 5.103,
      "mean_token_accuracy": 0.17253754884004593,
      "num_tokens": 41798663.0,
      "step": 19075
    },
    {
      "entropy": 5.5746985912323,
      "epoch": 1.8936085748312823,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004645015041323825,
      "loss": 5.0552,
      "mean_token_accuracy": 0.18101844638586045,
      "num_tokens": 41808711.0,
      "step": 19080
    },
    {
      "entropy": 5.5667108535766605,
      "epoch": 1.8941048034934498,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046448239995160455,
      "loss": 5.1045,
      "mean_token_accuracy": 0.17138869315385818,
      "num_tokens": 41819679.0,
      "step": 19085
    },
    {
      "entropy": 5.5978536128997805,
      "epoch": 1.8946010321556173,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004644632910720715,
      "loss": 5.0812,
      "mean_token_accuracy": 0.1804081067442894,
      "num_tokens": 41829649.0,
      "step": 19090
    },
    {
      "entropy": 5.60251750946045,
      "epoch": 1.895097260817785,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000464444177494257,
      "loss": 5.1357,
      "mean_token_accuracy": 0.17148305028676986,
      "num_tokens": 41842523.0,
      "step": 19095
    },
    {
      "entropy": 5.59897665977478,
      "epoch": 1.8955934894799524,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004644250592186351,
      "loss": 5.1131,
      "mean_token_accuracy": 0.1739485189318657,
      "num_tokens": 41854514.0,
      "step": 19100
    },
    {
      "entropy": 5.64643669128418,
      "epoch": 1.89608971814212,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046440593624568,
      "loss": 5.077,
      "mean_token_accuracy": 0.17800614833831788,
      "num_tokens": 41865166.0,
      "step": 19105
    },
    {
      "entropy": 5.612811946868897,
      "epoch": 1.8965859468042874,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004643868085758657,
      "loss": 5.0395,
      "mean_token_accuracy": 0.18219392895698547,
      "num_tokens": 41877264.0,
      "step": 19110
    },
    {
      "entropy": 5.578051280975342,
      "epoch": 1.8970821754664549,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046436767620966675,
      "loss": 5.0544,
      "mean_token_accuracy": 0.17759734094142915,
      "num_tokens": 41887345.0,
      "step": 19115
    },
    {
      "entropy": 5.570702743530274,
      "epoch": 1.8975784041286223,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046434853914755744,
      "loss": 5.0995,
      "mean_token_accuracy": 0.17609851509332658,
      "num_tokens": 41898877.0,
      "step": 19120
    },
    {
      "entropy": 5.519587707519531,
      "epoch": 1.89807463279079,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046432939739001237,
      "loss": 5.0433,
      "mean_token_accuracy": 0.18162453174591064,
      "num_tokens": 41910593.0,
      "step": 19125
    },
    {
      "entropy": 5.568113040924072,
      "epoch": 1.8985708614529575,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046431025093750613,
      "loss": 5.0677,
      "mean_token_accuracy": 0.18533090204000474,
      "num_tokens": 41920719.0,
      "step": 19130
    },
    {
      "entropy": 5.631743574142456,
      "epoch": 1.8990670901151252,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046429109979051367,
      "loss": 5.1098,
      "mean_token_accuracy": 0.16751329898834227,
      "num_tokens": 41931630.0,
      "step": 19135
    },
    {
      "entropy": 5.62287392616272,
      "epoch": 1.8995633187772927,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046427194394950985,
      "loss": 5.1433,
      "mean_token_accuracy": 0.17662004381418228,
      "num_tokens": 41943076.0,
      "step": 19140
    },
    {
      "entropy": 5.616760683059693,
      "epoch": 1.9000595474394602,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004642527834149696,
      "loss": 5.0518,
      "mean_token_accuracy": 0.1774890035390854,
      "num_tokens": 41954294.0,
      "step": 19145
    },
    {
      "entropy": 5.550148391723633,
      "epoch": 1.9005557761016276,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046423361818736814,
      "loss": 5.023,
      "mean_token_accuracy": 0.17378406077623368,
      "num_tokens": 41964943.0,
      "step": 19150
    },
    {
      "entropy": 5.519982194900512,
      "epoch": 1.901052004763795,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004642144482671807,
      "loss": 5.0796,
      "mean_token_accuracy": 0.17648021578788758,
      "num_tokens": 41975211.0,
      "step": 19155
    },
    {
      "entropy": 5.596327066421509,
      "epoch": 1.9015482334259626,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046419527365488275,
      "loss": 5.0915,
      "mean_token_accuracy": 0.17651113569736482,
      "num_tokens": 41986492.0,
      "step": 19160
    },
    {
      "entropy": 5.582166337966919,
      "epoch": 1.90204446208813,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004641760943509496,
      "loss": 5.066,
      "mean_token_accuracy": 0.1846793457865715,
      "num_tokens": 41997613.0,
      "step": 19165
    },
    {
      "entropy": 5.564380311965943,
      "epoch": 1.9025406907502977,
      "grad_norm": 1.015625,
      "learning_rate": 0.000464156910355857,
      "loss": 5.1276,
      "mean_token_accuracy": 0.1680097132921219,
      "num_tokens": 42008597.0,
      "step": 19170
    },
    {
      "entropy": 5.538931274414063,
      "epoch": 1.9030369194124652,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046413772167008057,
      "loss": 5.1138,
      "mean_token_accuracy": 0.17547208666801453,
      "num_tokens": 42019698.0,
      "step": 19175
    },
    {
      "entropy": 5.5616978168487545,
      "epoch": 1.903533148074633,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046411852829409626,
      "loss": 5.0133,
      "mean_token_accuracy": 0.1790014386177063,
      "num_tokens": 42029473.0,
      "step": 19180
    },
    {
      "entropy": 5.557767772674561,
      "epoch": 1.9040293767368004,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046409933022837984,
      "loss": 5.1607,
      "mean_token_accuracy": 0.17126497030258178,
      "num_tokens": 42041494.0,
      "step": 19185
    },
    {
      "entropy": 5.598025178909301,
      "epoch": 1.9045256053989679,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004640801274734076,
      "loss": 5.1173,
      "mean_token_accuracy": 0.1730015054345131,
      "num_tokens": 42051798.0,
      "step": 19190
    },
    {
      "entropy": 5.58586573600769,
      "epoch": 1.9050218340611353,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004640609200296557,
      "loss": 5.0391,
      "mean_token_accuracy": 0.17746248543262483,
      "num_tokens": 42061802.0,
      "step": 19195
    },
    {
      "entropy": 5.641064643859863,
      "epoch": 1.9055180627233028,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004640417078976003,
      "loss": 5.1669,
      "mean_token_accuracy": 0.1660948857665062,
      "num_tokens": 42071503.0,
      "step": 19200
    },
    {
      "entropy": 5.486746501922608,
      "epoch": 1.9060142913854703,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004640224910777179,
      "loss": 5.0006,
      "mean_token_accuracy": 0.17608706504106522,
      "num_tokens": 42082104.0,
      "step": 19205
    },
    {
      "entropy": 5.599625205993652,
      "epoch": 1.906510520047638,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000464003269570485,
      "loss": 5.0636,
      "mean_token_accuracy": 0.178280745446682,
      "num_tokens": 42092798.0,
      "step": 19210
    },
    {
      "entropy": 5.615771293640137,
      "epoch": 1.9070067487098055,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004639840433763783,
      "loss": 5.1049,
      "mean_token_accuracy": 0.1755188450217247,
      "num_tokens": 42103062.0,
      "step": 19215
    },
    {
      "entropy": 5.61214542388916,
      "epoch": 1.9075029773719732,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046396481249587454,
      "loss": 5.1537,
      "mean_token_accuracy": 0.1785653531551361,
      "num_tokens": 42115062.0,
      "step": 19220
    },
    {
      "entropy": 5.635091304779053,
      "epoch": 1.9079992060341406,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004639455769294506,
      "loss": 5.133,
      "mean_token_accuracy": 0.17059969156980515,
      "num_tokens": 42125937.0,
      "step": 19225
    },
    {
      "entropy": 5.607561874389648,
      "epoch": 1.9084954346963081,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004639263366775835,
      "loss": 5.0835,
      "mean_token_accuracy": 0.17598254531621932,
      "num_tokens": 42136916.0,
      "step": 19230
    },
    {
      "entropy": 5.612665843963623,
      "epoch": 1.9089916633584756,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046390709174075036,
      "loss": 5.1308,
      "mean_token_accuracy": 0.1735413983464241,
      "num_tokens": 42148348.0,
      "step": 19235
    },
    {
      "entropy": 5.62713623046875,
      "epoch": 1.909487892020643,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004638878421194284,
      "loss": 5.0511,
      "mean_token_accuracy": 0.1818789556622505,
      "num_tokens": 42157586.0,
      "step": 19240
    },
    {
      "entropy": 5.6204204082489015,
      "epoch": 1.9099841206828105,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000463868587814095,
      "loss": 5.1211,
      "mean_token_accuracy": 0.17187168300151826,
      "num_tokens": 42169536.0,
      "step": 19245
    },
    {
      "entropy": 5.539913558959961,
      "epoch": 1.910480349344978,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004638493288252276,
      "loss": 5.0556,
      "mean_token_accuracy": 0.18079178482294084,
      "num_tokens": 42180024.0,
      "step": 19250
    },
    {
      "entropy": 5.5710365772247314,
      "epoch": 1.9109765780071457,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046383006515330374,
      "loss": 5.0548,
      "mean_token_accuracy": 0.17978658229112626,
      "num_tokens": 42191070.0,
      "step": 19255
    },
    {
      "entropy": 5.590140676498413,
      "epoch": 1.9114728066693132,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004638107967988012,
      "loss": 5.1035,
      "mean_token_accuracy": 0.17840124368667604,
      "num_tokens": 42203643.0,
      "step": 19260
    },
    {
      "entropy": 5.626328039169311,
      "epoch": 1.9119690353314809,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046379152376219775,
      "loss": 5.1585,
      "mean_token_accuracy": 0.16778660714626312,
      "num_tokens": 42214979.0,
      "step": 19265
    },
    {
      "entropy": 5.690818643569946,
      "epoch": 1.9124652639936484,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046377224604397127,
      "loss": 5.1353,
      "mean_token_accuracy": 0.16965446323156358,
      "num_tokens": 42225152.0,
      "step": 19270
    },
    {
      "entropy": 5.650046634674072,
      "epoch": 1.9129614926558158,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004637529636445998,
      "loss": 5.0727,
      "mean_token_accuracy": 0.1776326984167099,
      "num_tokens": 42235151.0,
      "step": 19275
    },
    {
      "entropy": 5.590389442443848,
      "epoch": 1.9134577213179833,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004637336765645617,
      "loss": 5.0629,
      "mean_token_accuracy": 0.18053047806024553,
      "num_tokens": 42245419.0,
      "step": 19280
    },
    {
      "entropy": 5.514609050750733,
      "epoch": 1.9139539499801508,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046371438480433504,
      "loss": 5.0422,
      "mean_token_accuracy": 0.1825389474630356,
      "num_tokens": 42255836.0,
      "step": 19285
    },
    {
      "entropy": 5.542360019683838,
      "epoch": 1.9144501786423183,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004636950883643983,
      "loss": 5.1295,
      "mean_token_accuracy": 0.17667202949523925,
      "num_tokens": 42267490.0,
      "step": 19290
    },
    {
      "entropy": 5.638892936706543,
      "epoch": 1.914946407304486,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046367578724523,
      "loss": 5.0959,
      "mean_token_accuracy": 0.17291081845760345,
      "num_tokens": 42276957.0,
      "step": 19295
    },
    {
      "entropy": 5.603362417221069,
      "epoch": 1.9154426359666534,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004636564814473087,
      "loss": 4.9772,
      "mean_token_accuracy": 0.1878822311758995,
      "num_tokens": 42288469.0,
      "step": 19300
    },
    {
      "entropy": 5.547716951370239,
      "epoch": 1.9159388646288211,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004636371709711131,
      "loss": 5.0347,
      "mean_token_accuracy": 0.19002901911735534,
      "num_tokens": 42299141.0,
      "step": 19305
    },
    {
      "entropy": 5.54572286605835,
      "epoch": 1.9164350932909886,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046361785581712224,
      "loss": 5.1264,
      "mean_token_accuracy": 0.17766737192869186,
      "num_tokens": 42310651.0,
      "step": 19310
    },
    {
      "entropy": 5.534989356994629,
      "epoch": 1.916931321953156,
      "grad_norm": 1.0625,
      "learning_rate": 0.000463598535985815,
      "loss": 4.9387,
      "mean_token_accuracy": 0.1915040209889412,
      "num_tokens": 42320529.0,
      "step": 19315
    },
    {
      "entropy": 5.571506595611572,
      "epoch": 1.9174275506153236,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004635792114776704,
      "loss": 5.0015,
      "mean_token_accuracy": 0.18258465677499772,
      "num_tokens": 42331216.0,
      "step": 19320
    },
    {
      "entropy": 5.486332511901855,
      "epoch": 1.917923779277491,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004635598822931677,
      "loss": 5.0099,
      "mean_token_accuracy": 0.1813252389431,
      "num_tokens": 42341730.0,
      "step": 19325
    },
    {
      "entropy": 5.598900413513183,
      "epoch": 1.9184200079396585,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046354054843278624,
      "loss": 5.1192,
      "mean_token_accuracy": 0.171194326877594,
      "num_tokens": 42352643.0,
      "step": 19330
    },
    {
      "entropy": 5.657588863372803,
      "epoch": 1.918916236601826,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046352120989700553,
      "loss": 5.1516,
      "mean_token_accuracy": 0.17405202537775039,
      "num_tokens": 42363979.0,
      "step": 19335
    },
    {
      "entropy": 5.617137384414673,
      "epoch": 1.9194124652639937,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000463501866686305,
      "loss": 5.1028,
      "mean_token_accuracy": 0.17593103647232056,
      "num_tokens": 42374104.0,
      "step": 19340
    },
    {
      "entropy": 5.638616609573364,
      "epoch": 1.9199086939261611,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046348251880116433,
      "loss": 5.0906,
      "mean_token_accuracy": 0.17601355016231537,
      "num_tokens": 42383576.0,
      "step": 19345
    },
    {
      "entropy": 5.596588945388794,
      "epoch": 1.9204049225883288,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046346316624206325,
      "loss": 5.099,
      "mean_token_accuracy": 0.17786043733358384,
      "num_tokens": 42394625.0,
      "step": 19350
    },
    {
      "entropy": 5.595097827911377,
      "epoch": 1.9209011512504963,
      "grad_norm": 1.0,
      "learning_rate": 0.0004634438090094819,
      "loss": 5.045,
      "mean_token_accuracy": 0.1801186740398407,
      "num_tokens": 42404764.0,
      "step": 19355
    },
    {
      "entropy": 5.588011121749878,
      "epoch": 1.9213973799126638,
      "grad_norm": 1.125,
      "learning_rate": 0.00046342444710390006,
      "loss": 5.0222,
      "mean_token_accuracy": 0.18278496861457824,
      "num_tokens": 42414386.0,
      "step": 19360
    },
    {
      "entropy": 5.57353105545044,
      "epoch": 1.9218936085748313,
      "grad_norm": 1.0,
      "learning_rate": 0.000463405080525798,
      "loss": 5.1147,
      "mean_token_accuracy": 0.1710527390241623,
      "num_tokens": 42425671.0,
      "step": 19365
    },
    {
      "entropy": 5.587749290466308,
      "epoch": 1.9223898372369987,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004633857092756559,
      "loss": 5.1468,
      "mean_token_accuracy": 0.17122229635715486,
      "num_tokens": 42436365.0,
      "step": 19370
    },
    {
      "entropy": 5.581427192687988,
      "epoch": 1.9228860658991662,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004633663333539542,
      "loss": 4.999,
      "mean_token_accuracy": 0.17864720076322554,
      "num_tokens": 42447624.0,
      "step": 19375
    },
    {
      "entropy": 5.64170126914978,
      "epoch": 1.923382294561334,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046334695276117324,
      "loss": 5.0975,
      "mean_token_accuracy": 0.1780153974890709,
      "num_tokens": 42458013.0,
      "step": 19380
    },
    {
      "entropy": 5.584343338012696,
      "epoch": 1.9238785232235014,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004633275674977937,
      "loss": 5.1078,
      "mean_token_accuracy": 0.17281357795000077,
      "num_tokens": 42469807.0,
      "step": 19385
    },
    {
      "entropy": 5.5210188865661625,
      "epoch": 1.9243747518856689,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004633081775642964,
      "loss": 5.0423,
      "mean_token_accuracy": 0.17482510656118394,
      "num_tokens": 42479240.0,
      "step": 19390
    },
    {
      "entropy": 5.607713222503662,
      "epoch": 1.9248709805478366,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046328878296116196,
      "loss": 5.1154,
      "mean_token_accuracy": 0.1759380131959915,
      "num_tokens": 42489558.0,
      "step": 19395
    },
    {
      "entropy": 5.636698293685913,
      "epoch": 1.925367209210004,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046326938368887146,
      "loss": 5.1108,
      "mean_token_accuracy": 0.17984693199396135,
      "num_tokens": 42501252.0,
      "step": 19400
    },
    {
      "entropy": 5.647745609283447,
      "epoch": 1.9258634378721715,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046324997974790596,
      "loss": 5.1544,
      "mean_token_accuracy": 0.17736948430538177,
      "num_tokens": 42512330.0,
      "step": 19405
    },
    {
      "entropy": 5.600220775604248,
      "epoch": 1.926359666534339,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004632305711387466,
      "loss": 5.1497,
      "mean_token_accuracy": 0.17401877790689468,
      "num_tokens": 42525216.0,
      "step": 19410
    },
    {
      "entropy": 5.601547622680664,
      "epoch": 1.9268558951965065,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046321115786187466,
      "loss": 5.0827,
      "mean_token_accuracy": 0.17393710166215898,
      "num_tokens": 42537047.0,
      "step": 19415
    },
    {
      "entropy": 5.616623449325561,
      "epoch": 1.927352123858674,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046319173991777153,
      "loss": 5.0781,
      "mean_token_accuracy": 0.17507263123989106,
      "num_tokens": 42549784.0,
      "step": 19420
    },
    {
      "entropy": 5.610589551925659,
      "epoch": 1.9278483525208416,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004631723173069188,
      "loss": 5.0604,
      "mean_token_accuracy": 0.17734439373016359,
      "num_tokens": 42561079.0,
      "step": 19425
    },
    {
      "entropy": 5.6161833763122555,
      "epoch": 1.928344581183009,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004631528900297981,
      "loss": 5.1329,
      "mean_token_accuracy": 0.17262810468673706,
      "num_tokens": 42572531.0,
      "step": 19430
    },
    {
      "entropy": 5.571467161178589,
      "epoch": 1.9288408098451768,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046313345808689113,
      "loss": 5.0423,
      "mean_token_accuracy": 0.18025963306427,
      "num_tokens": 42583836.0,
      "step": 19435
    },
    {
      "entropy": 5.619368839263916,
      "epoch": 1.9293370385073443,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004631140214786798,
      "loss": 5.1637,
      "mean_token_accuracy": 0.17169311195611953,
      "num_tokens": 42595338.0,
      "step": 19440
    },
    {
      "entropy": 5.619530487060547,
      "epoch": 1.9298332671695118,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046309458020564606,
      "loss": 5.1038,
      "mean_token_accuracy": 0.17427470833063125,
      "num_tokens": 42606804.0,
      "step": 19445
    },
    {
      "entropy": 5.573221254348755,
      "epoch": 1.9303294958316792,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046307513426827214,
      "loss": 5.0117,
      "mean_token_accuracy": 0.18551431000232696,
      "num_tokens": 42616987.0,
      "step": 19450
    },
    {
      "entropy": 5.618338918685913,
      "epoch": 1.9308257244938467,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046305568366704006,
      "loss": 5.0898,
      "mean_token_accuracy": 0.17544220238924027,
      "num_tokens": 42627633.0,
      "step": 19455
    },
    {
      "entropy": 5.6062640190124515,
      "epoch": 1.9313219531560142,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004630362284024322,
      "loss": 5.0621,
      "mean_token_accuracy": 0.17597774118185044,
      "num_tokens": 42638802.0,
      "step": 19460
    },
    {
      "entropy": 5.572386407852173,
      "epoch": 1.9318181818181817,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046301676847493116,
      "loss": 5.06,
      "mean_token_accuracy": 0.18015048652887344,
      "num_tokens": 42650802.0,
      "step": 19465
    },
    {
      "entropy": 5.554395389556885,
      "epoch": 1.9323144104803494,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046299730388501935,
      "loss": 5.0283,
      "mean_token_accuracy": 0.1833526000380516,
      "num_tokens": 42660290.0,
      "step": 19470
    },
    {
      "entropy": 5.6117065906524655,
      "epoch": 1.9328106391425168,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046297783463317957,
      "loss": 5.1215,
      "mean_token_accuracy": 0.17703825533390044,
      "num_tokens": 42671182.0,
      "step": 19475
    },
    {
      "entropy": 5.629882526397705,
      "epoch": 1.9333068678046845,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004629583607198945,
      "loss": 5.1112,
      "mean_token_accuracy": 0.17572882026433945,
      "num_tokens": 42682499.0,
      "step": 19480
    },
    {
      "entropy": 5.565653944015503,
      "epoch": 1.933803096466852,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046293888214564704,
      "loss": 5.1156,
      "mean_token_accuracy": 0.17104930728673934,
      "num_tokens": 42693728.0,
      "step": 19485
    },
    {
      "entropy": 5.600042152404785,
      "epoch": 1.9342993251290195,
      "grad_norm": 1.0,
      "learning_rate": 0.00046291939891092036,
      "loss": 5.1106,
      "mean_token_accuracy": 0.17758123576641083,
      "num_tokens": 42704877.0,
      "step": 19490
    },
    {
      "entropy": 5.645932626724243,
      "epoch": 1.934795553791187,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004628999110161975,
      "loss": 5.1447,
      "mean_token_accuracy": 0.16805647760629655,
      "num_tokens": 42715787.0,
      "step": 19495
    },
    {
      "entropy": 5.617727041244507,
      "epoch": 1.9352917824533544,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046288041846196174,
      "loss": 5.0947,
      "mean_token_accuracy": 0.17687833458185195,
      "num_tokens": 42727557.0,
      "step": 19500
    },
    {
      "entropy": 5.619726085662842,
      "epoch": 1.935788011115522,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046286092124869647,
      "loss": 5.1066,
      "mean_token_accuracy": 0.17475387156009675,
      "num_tokens": 42737540.0,
      "step": 19505
    },
    {
      "entropy": 5.6594781398773195,
      "epoch": 1.9362842397776896,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046284141937688513,
      "loss": 5.1315,
      "mean_token_accuracy": 0.17966286689043046,
      "num_tokens": 42748716.0,
      "step": 19510
    },
    {
      "entropy": 5.713207197189331,
      "epoch": 1.936780468439857,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046282191284701123,
      "loss": 5.1694,
      "mean_token_accuracy": 0.17055969685316086,
      "num_tokens": 42758991.0,
      "step": 19515
    },
    {
      "entropy": 5.671346998214721,
      "epoch": 1.9372766971020248,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004628024016595587,
      "loss": 5.1299,
      "mean_token_accuracy": 0.17142341881990433,
      "num_tokens": 42771105.0,
      "step": 19520
    },
    {
      "entropy": 5.60203275680542,
      "epoch": 1.9377729257641922,
      "grad_norm": 0.859375,
      "learning_rate": 0.00046278288581501136,
      "loss": 5.0906,
      "mean_token_accuracy": 0.17757316380739213,
      "num_tokens": 42783317.0,
      "step": 19525
    },
    {
      "entropy": 5.659475708007813,
      "epoch": 1.9382691544263597,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046276336531385305,
      "loss": 5.2089,
      "mean_token_accuracy": 0.17064513266086578,
      "num_tokens": 42794998.0,
      "step": 19530
    },
    {
      "entropy": 5.618157243728637,
      "epoch": 1.9387653830885272,
      "grad_norm": 1.0,
      "learning_rate": 0.00046274384015656785,
      "loss": 5.087,
      "mean_token_accuracy": 0.17916978299617767,
      "num_tokens": 42806197.0,
      "step": 19535
    },
    {
      "entropy": 5.553230237960816,
      "epoch": 1.9392616117506947,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046272431034363997,
      "loss": 4.9985,
      "mean_token_accuracy": 0.17861422449350356,
      "num_tokens": 42816522.0,
      "step": 19540
    },
    {
      "entropy": 5.561730861663818,
      "epoch": 1.9397578404128621,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004627047758755538,
      "loss": 5.015,
      "mean_token_accuracy": 0.18492391258478164,
      "num_tokens": 42826003.0,
      "step": 19545
    },
    {
      "entropy": 5.58969464302063,
      "epoch": 1.9402540690750296,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004626852367527935,
      "loss": 5.0984,
      "mean_token_accuracy": 0.17146885842084886,
      "num_tokens": 42836346.0,
      "step": 19550
    },
    {
      "entropy": 5.5988781452178955,
      "epoch": 1.9407502977371973,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004626656929758438,
      "loss": 5.132,
      "mean_token_accuracy": 0.17709634751081466,
      "num_tokens": 42846296.0,
      "step": 19555
    },
    {
      "entropy": 5.677931928634644,
      "epoch": 1.9412465263993648,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004626461445451894,
      "loss": 5.1508,
      "mean_token_accuracy": 0.176907317340374,
      "num_tokens": 42856323.0,
      "step": 19560
    },
    {
      "entropy": 5.695257472991943,
      "epoch": 1.9417427550615325,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004626265914613149,
      "loss": 5.1559,
      "mean_token_accuracy": 0.17409739941358565,
      "num_tokens": 42866937.0,
      "step": 19565
    },
    {
      "entropy": 5.621702814102173,
      "epoch": 1.9422389837237,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004626070337247052,
      "loss": 5.1146,
      "mean_token_accuracy": 0.1717504620552063,
      "num_tokens": 42877194.0,
      "step": 19570
    },
    {
      "entropy": 5.599123859405518,
      "epoch": 1.9427352123858674,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004625874713358453,
      "loss": 5.119,
      "mean_token_accuracy": 0.17134264409542083,
      "num_tokens": 42887741.0,
      "step": 19575
    },
    {
      "entropy": 5.608036708831787,
      "epoch": 1.943231441048035,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046256790429522035,
      "loss": 5.0891,
      "mean_token_accuracy": 0.1841115564107895,
      "num_tokens": 42898195.0,
      "step": 19580
    },
    {
      "entropy": 5.712985801696777,
      "epoch": 1.9437276697102024,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004625483326033155,
      "loss": 5.1255,
      "mean_token_accuracy": 0.17579784691333772,
      "num_tokens": 42908852.0,
      "step": 19585
    },
    {
      "entropy": 5.5787365436553955,
      "epoch": 1.9442238983723699,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004625287562606162,
      "loss": 5.0473,
      "mean_token_accuracy": 0.17822929471731186,
      "num_tokens": 42920166.0,
      "step": 19590
    },
    {
      "entropy": 5.634564924240112,
      "epoch": 1.9447201270345376,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004625091752676078,
      "loss": 5.1111,
      "mean_token_accuracy": 0.17295468151569365,
      "num_tokens": 42931677.0,
      "step": 19595
    },
    {
      "entropy": 5.59920301437378,
      "epoch": 1.945216355696705,
      "grad_norm": 1.046875,
      "learning_rate": 0.000462489589624776,
      "loss": 5.0936,
      "mean_token_accuracy": 0.17583199888467788,
      "num_tokens": 42942346.0,
      "step": 19600
    },
    {
      "entropy": 5.505224418640137,
      "epoch": 1.9457125843588727,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004624699993326063,
      "loss": 5.0553,
      "mean_token_accuracy": 0.1816968634724617,
      "num_tokens": 42953821.0,
      "step": 19605
    },
    {
      "entropy": 5.533775568008423,
      "epoch": 1.9462088130210402,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046245040439158456,
      "loss": 5.0626,
      "mean_token_accuracy": 0.17836090922355652,
      "num_tokens": 42964358.0,
      "step": 19610
    },
    {
      "entropy": 5.590550565719605,
      "epoch": 1.9467050416832077,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046243080480219673,
      "loss": 5.0602,
      "mean_token_accuracy": 0.17841163128614426,
      "num_tokens": 42976459.0,
      "step": 19615
    },
    {
      "entropy": 5.66695122718811,
      "epoch": 1.9472012703453752,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004624112005649288,
      "loss": 5.2146,
      "mean_token_accuracy": 0.1721201553940773,
      "num_tokens": 42988603.0,
      "step": 19620
    },
    {
      "entropy": 5.591305208206177,
      "epoch": 1.9476974990075426,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046239159168026694,
      "loss": 5.1156,
      "mean_token_accuracy": 0.1745889812707901,
      "num_tokens": 43000262.0,
      "step": 19625
    },
    {
      "entropy": 5.53891806602478,
      "epoch": 1.94819372766971,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004623719781486974,
      "loss": 5.0477,
      "mean_token_accuracy": 0.18263127654790878,
      "num_tokens": 43009363.0,
      "step": 19630
    },
    {
      "entropy": 5.614121150970459,
      "epoch": 1.9486899563318776,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046235235997070664,
      "loss": 5.1104,
      "mean_token_accuracy": 0.17417142242193223,
      "num_tokens": 43019893.0,
      "step": 19635
    },
    {
      "entropy": 5.557586336135865,
      "epoch": 1.9491861849940453,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000462332737146781,
      "loss": 5.0315,
      "mean_token_accuracy": 0.18172991424798965,
      "num_tokens": 43031152.0,
      "step": 19640
    },
    {
      "entropy": 5.5968810558319095,
      "epoch": 1.9496824136562128,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004623131096774071,
      "loss": 5.165,
      "mean_token_accuracy": 0.17095980048179626,
      "num_tokens": 43041247.0,
      "step": 19645
    },
    {
      "entropy": 5.6410276889801025,
      "epoch": 1.9501786423183805,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004622934775630717,
      "loss": 5.0841,
      "mean_token_accuracy": 0.1706358790397644,
      "num_tokens": 43052065.0,
      "step": 19650
    },
    {
      "entropy": 5.61663818359375,
      "epoch": 1.950674870980548,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00046227384080426163,
      "loss": 5.1103,
      "mean_token_accuracy": 0.16961565762758254,
      "num_tokens": 43063217.0,
      "step": 19655
    },
    {
      "entropy": 5.57474684715271,
      "epoch": 1.9511710996427154,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004622541994014639,
      "loss": 5.1157,
      "mean_token_accuracy": 0.17970566004514693,
      "num_tokens": 43074463.0,
      "step": 19660
    },
    {
      "entropy": 5.704758262634277,
      "epoch": 1.9516673283048829,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004622345533551655,
      "loss": 5.1692,
      "mean_token_accuracy": 0.16785794943571092,
      "num_tokens": 43086004.0,
      "step": 19665
    },
    {
      "entropy": 5.655753755569458,
      "epoch": 1.9521635569670504,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004622149026658537,
      "loss": 5.0505,
      "mean_token_accuracy": 0.17980840504169465,
      "num_tokens": 43096973.0,
      "step": 19670
    },
    {
      "entropy": 5.629302549362182,
      "epoch": 1.9526597856292178,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004621952473340156,
      "loss": 5.154,
      "mean_token_accuracy": 0.1756134033203125,
      "num_tokens": 43107337.0,
      "step": 19675
    },
    {
      "entropy": 5.600006103515625,
      "epoch": 1.9531560142913855,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004621755873601388,
      "loss": 5.0698,
      "mean_token_accuracy": 0.18476034104824066,
      "num_tokens": 43118668.0,
      "step": 19680
    },
    {
      "entropy": 5.679725217819214,
      "epoch": 1.953652242953553,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004621559227447108,
      "loss": 5.1749,
      "mean_token_accuracy": 0.17214563637971877,
      "num_tokens": 43129030.0,
      "step": 19685
    },
    {
      "entropy": 5.557803297042847,
      "epoch": 1.9541484716157205,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004621362534882192,
      "loss": 5.0316,
      "mean_token_accuracy": 0.18312050700187682,
      "num_tokens": 43139120.0,
      "step": 19690
    },
    {
      "entropy": 5.600490522384644,
      "epoch": 1.9546447002778882,
      "grad_norm": 0.875,
      "learning_rate": 0.0004621165795911517,
      "loss": 5.1167,
      "mean_token_accuracy": 0.17393818348646164,
      "num_tokens": 43152390.0,
      "step": 19695
    },
    {
      "entropy": 5.574725341796875,
      "epoch": 1.9551409289400556,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046209690105399623,
      "loss": 4.9267,
      "mean_token_accuracy": 0.18806587755680085,
      "num_tokens": 43162230.0,
      "step": 19700
    },
    {
      "entropy": 5.55992021560669,
      "epoch": 1.9556371576022231,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046207721787724085,
      "loss": 5.0551,
      "mean_token_accuracy": 0.1779712587594986,
      "num_tokens": 43172751.0,
      "step": 19705
    },
    {
      "entropy": 5.711976671218872,
      "epoch": 1.9561333862643906,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046205753006137347,
      "loss": 5.1428,
      "mean_token_accuracy": 0.1770087569952011,
      "num_tokens": 43183835.0,
      "step": 19710
    },
    {
      "entropy": 5.652405595779419,
      "epoch": 1.956629614926558,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046203783760688245,
      "loss": 5.0424,
      "mean_token_accuracy": 0.1715905711054802,
      "num_tokens": 43193178.0,
      "step": 19715
    },
    {
      "entropy": 5.629620599746704,
      "epoch": 1.9571258435887255,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004620181405142561,
      "loss": 5.1941,
      "mean_token_accuracy": 0.1669650837779045,
      "num_tokens": 43204968.0,
      "step": 19720
    },
    {
      "entropy": 5.60186276435852,
      "epoch": 1.9576220722508932,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004619984387839829,
      "loss": 5.2149,
      "mean_token_accuracy": 0.17331114709377288,
      "num_tokens": 43214916.0,
      "step": 19725
    },
    {
      "entropy": 5.733885765075684,
      "epoch": 1.9581183009130607,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046197873241655133,
      "loss": 5.202,
      "mean_token_accuracy": 0.17660846710205078,
      "num_tokens": 43226374.0,
      "step": 19730
    },
    {
      "entropy": 5.605831909179687,
      "epoch": 1.9586145295752284,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004619590214124501,
      "loss": 5.07,
      "mean_token_accuracy": 0.1746105208992958,
      "num_tokens": 43237296.0,
      "step": 19735
    },
    {
      "entropy": 5.699593019485474,
      "epoch": 1.9591107582373959,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046193930577216797,
      "loss": 5.2344,
      "mean_token_accuracy": 0.16304323226213455,
      "num_tokens": 43248220.0,
      "step": 19740
    },
    {
      "entropy": 5.599291515350342,
      "epoch": 1.9596069868995634,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004619195854961939,
      "loss": 5.0083,
      "mean_token_accuracy": 0.18184856325387955,
      "num_tokens": 43259913.0,
      "step": 19745
    },
    {
      "entropy": 5.653325128555298,
      "epoch": 1.9601032155617308,
      "grad_norm": 0.890625,
      "learning_rate": 0.00046189986058501687,
      "loss": 5.1334,
      "mean_token_accuracy": 0.17192408740520476,
      "num_tokens": 43272317.0,
      "step": 19750
    },
    {
      "entropy": 5.539155769348144,
      "epoch": 1.9605994442238983,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000461880131039126,
      "loss": 5.0256,
      "mean_token_accuracy": 0.1805944561958313,
      "num_tokens": 43283170.0,
      "step": 19755
    },
    {
      "entropy": 5.5500648498535154,
      "epoch": 1.9610956728860658,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004618603968590107,
      "loss": 5.0987,
      "mean_token_accuracy": 0.1737358123064041,
      "num_tokens": 43293899.0,
      "step": 19760
    },
    {
      "entropy": 5.614160394668579,
      "epoch": 1.9615919015482333,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046184065804516007,
      "loss": 5.1237,
      "mean_token_accuracy": 0.17988063097000123,
      "num_tokens": 43305638.0,
      "step": 19765
    },
    {
      "entropy": 5.69376311302185,
      "epoch": 1.962088130210401,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004618209145980638,
      "loss": 5.2089,
      "mean_token_accuracy": 0.16637139320373534,
      "num_tokens": 43315900.0,
      "step": 19770
    },
    {
      "entropy": 5.716121101379395,
      "epoch": 1.9625843588725684,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004618011665182114,
      "loss": 5.2568,
      "mean_token_accuracy": 0.17059977650642394,
      "num_tokens": 43328464.0,
      "step": 19775
    },
    {
      "entropy": 5.5641309261322025,
      "epoch": 1.9630805875347361,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004617814138060926,
      "loss": 4.9544,
      "mean_token_accuracy": 0.1866731658577919,
      "num_tokens": 43339303.0,
      "step": 19780
    },
    {
      "entropy": 5.665921545028686,
      "epoch": 1.9635768161969036,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004617616564621972,
      "loss": 5.1842,
      "mean_token_accuracy": 0.1721518263220787,
      "num_tokens": 43349269.0,
      "step": 19785
    },
    {
      "entropy": 5.5670370101928714,
      "epoch": 1.964073044859071,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046174189448701516,
      "loss": 5.0504,
      "mean_token_accuracy": 0.17897453010082245,
      "num_tokens": 43361090.0,
      "step": 19790
    },
    {
      "entropy": 5.552548503875732,
      "epoch": 1.9645692735212386,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004617221278810365,
      "loss": 5.0361,
      "mean_token_accuracy": 0.1781848341226578,
      "num_tokens": 43373448.0,
      "step": 19795
    },
    {
      "entropy": 5.712812995910644,
      "epoch": 1.965065502183406,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046170235664475147,
      "loss": 5.185,
      "mean_token_accuracy": 0.17048768997192382,
      "num_tokens": 43386202.0,
      "step": 19800
    },
    {
      "entropy": 5.629647731781006,
      "epoch": 1.9655617308455735,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046168258077865026,
      "loss": 5.0884,
      "mean_token_accuracy": 0.1806666672229767,
      "num_tokens": 43396363.0,
      "step": 19805
    },
    {
      "entropy": 5.575820732116699,
      "epoch": 1.9660579595077412,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004616628002832233,
      "loss": 5.0012,
      "mean_token_accuracy": 0.18337174952030183,
      "num_tokens": 43406076.0,
      "step": 19810
    },
    {
      "entropy": 5.615880727767944,
      "epoch": 1.9665541881699087,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046164301515896124,
      "loss": 5.1488,
      "mean_token_accuracy": 0.1742085710167885,
      "num_tokens": 43416226.0,
      "step": 19815
    },
    {
      "entropy": 5.592219257354737,
      "epoch": 1.9670504168320764,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004616232254063544,
      "loss": 4.9929,
      "mean_token_accuracy": 0.1791839435696602,
      "num_tokens": 43427364.0,
      "step": 19820
    },
    {
      "entropy": 5.628041839599609,
      "epoch": 1.9675466454942439,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046160343102589384,
      "loss": 5.1235,
      "mean_token_accuracy": 0.17873211801052094,
      "num_tokens": 43437854.0,
      "step": 19825
    },
    {
      "entropy": 5.633173036575317,
      "epoch": 1.9680428741564113,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004615836320180702,
      "loss": 5.1433,
      "mean_token_accuracy": 0.17744995951652526,
      "num_tokens": 43449370.0,
      "step": 19830
    },
    {
      "entropy": 5.592914581298828,
      "epoch": 1.9685391028185788,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046156382838337455,
      "loss": 5.089,
      "mean_token_accuracy": 0.18056954592466354,
      "num_tokens": 43461631.0,
      "step": 19835
    },
    {
      "entropy": 5.611041736602783,
      "epoch": 1.9690353314807463,
      "grad_norm": 1.0,
      "learning_rate": 0.00046154402012229807,
      "loss": 5.0646,
      "mean_token_accuracy": 0.1743779346346855,
      "num_tokens": 43472357.0,
      "step": 19840
    },
    {
      "entropy": 5.560678863525391,
      "epoch": 1.9695315601429138,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046152420723533175,
      "loss": 5.0493,
      "mean_token_accuracy": 0.1790838733315468,
      "num_tokens": 43483642.0,
      "step": 19845
    },
    {
      "entropy": 5.662699460983276,
      "epoch": 1.9700277888050812,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046150438972296706,
      "loss": 5.1348,
      "mean_token_accuracy": 0.17695893049240113,
      "num_tokens": 43495136.0,
      "step": 19850
    },
    {
      "entropy": 5.5833886623382565,
      "epoch": 1.970524017467249,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046148456758569536,
      "loss": 5.1043,
      "mean_token_accuracy": 0.17521521896123887,
      "num_tokens": 43506376.0,
      "step": 19855
    },
    {
      "entropy": 5.606155872344971,
      "epoch": 1.9710202461294164,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046146474082400823,
      "loss": 5.066,
      "mean_token_accuracy": 0.18114390224218369,
      "num_tokens": 43516658.0,
      "step": 19860
    },
    {
      "entropy": 5.605082416534424,
      "epoch": 1.971516474791584,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046144490943839735,
      "loss": 5.0388,
      "mean_token_accuracy": 0.18103483468294143,
      "num_tokens": 43526907.0,
      "step": 19865
    },
    {
      "entropy": 5.667767095565796,
      "epoch": 1.9720127034537516,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004614250734293544,
      "loss": 5.1288,
      "mean_token_accuracy": 0.1709216371178627,
      "num_tokens": 43537739.0,
      "step": 19870
    },
    {
      "entropy": 5.549405908584594,
      "epoch": 1.972508932115919,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046140523279737146,
      "loss": 5.093,
      "mean_token_accuracy": 0.17633699476718903,
      "num_tokens": 43549240.0,
      "step": 19875
    },
    {
      "entropy": 5.609155130386353,
      "epoch": 1.9730051607780865,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004613853875429403,
      "loss": 5.0498,
      "mean_token_accuracy": 0.17566222697496414,
      "num_tokens": 43559857.0,
      "step": 19880
    },
    {
      "entropy": 5.581248331069946,
      "epoch": 1.973501389440254,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004613655376665532,
      "loss": 5.0454,
      "mean_token_accuracy": 0.171019084751606,
      "num_tokens": 43570262.0,
      "step": 19885
    },
    {
      "entropy": 5.652416467666626,
      "epoch": 1.9739976181024215,
      "grad_norm": 1.0,
      "learning_rate": 0.0004613456831687024,
      "loss": 5.1014,
      "mean_token_accuracy": 0.1733730211853981,
      "num_tokens": 43581965.0,
      "step": 19890
    },
    {
      "entropy": 5.613769292831421,
      "epoch": 1.9744938467645892,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046132582404988004,
      "loss": 5.066,
      "mean_token_accuracy": 0.17726066708564758,
      "num_tokens": 43592728.0,
      "step": 19895
    },
    {
      "entropy": 5.627223157882691,
      "epoch": 1.9749900754267566,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046130596031057885,
      "loss": 5.0414,
      "mean_token_accuracy": 0.1790455162525177,
      "num_tokens": 43602437.0,
      "step": 19900
    },
    {
      "entropy": 5.600422620773315,
      "epoch": 1.9754863040889243,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046128609195129135,
      "loss": 5.1611,
      "mean_token_accuracy": 0.17228424847126006,
      "num_tokens": 43613115.0,
      "step": 19905
    },
    {
      "entropy": 5.636856317520142,
      "epoch": 1.9759825327510918,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004612662189725101,
      "loss": 5.1344,
      "mean_token_accuracy": 0.17278586328029633,
      "num_tokens": 43623700.0,
      "step": 19910
    },
    {
      "entropy": 5.681130123138428,
      "epoch": 1.9764787614132593,
      "grad_norm": 1.015625,
      "learning_rate": 0.000461246341374728,
      "loss": 5.1454,
      "mean_token_accuracy": 0.1746896579861641,
      "num_tokens": 43633712.0,
      "step": 19915
    },
    {
      "entropy": 5.6480841636657715,
      "epoch": 1.9769749900754268,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046122645915843796,
      "loss": 5.1748,
      "mean_token_accuracy": 0.17486895769834518,
      "num_tokens": 43643705.0,
      "step": 19920
    },
    {
      "entropy": 5.623794937133789,
      "epoch": 1.9774712187375942,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046120657232413305,
      "loss": 5.1416,
      "mean_token_accuracy": 0.17898592501878738,
      "num_tokens": 43655376.0,
      "step": 19925
    },
    {
      "entropy": 5.593525409698486,
      "epoch": 1.9779674473997617,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004611866808723064,
      "loss": 5.0401,
      "mean_token_accuracy": 0.18505800515413284,
      "num_tokens": 43666329.0,
      "step": 19930
    },
    {
      "entropy": 5.565545225143433,
      "epoch": 1.9784636760619292,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046116678480345117,
      "loss": 5.047,
      "mean_token_accuracy": 0.17702313363552094,
      "num_tokens": 43676415.0,
      "step": 19935
    },
    {
      "entropy": 5.538389253616333,
      "epoch": 1.9789599047240969,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00046114688411806083,
      "loss": 5.0672,
      "mean_token_accuracy": 0.18030820190906524,
      "num_tokens": 43686997.0,
      "step": 19940
    },
    {
      "entropy": 5.572056579589844,
      "epoch": 1.9794561333862644,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00046112697881662895,
      "loss": 5.035,
      "mean_token_accuracy": 0.17873485684394835,
      "num_tokens": 43697342.0,
      "step": 19945
    },
    {
      "entropy": 5.654771709442139,
      "epoch": 1.979952362048432,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046110706889964905,
      "loss": 5.1382,
      "mean_token_accuracy": 0.1789373517036438,
      "num_tokens": 43707876.0,
      "step": 19950
    },
    {
      "entropy": 5.594812488555908,
      "epoch": 1.9804485907105995,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004610871543676149,
      "loss": 5.003,
      "mean_token_accuracy": 0.18422971367836,
      "num_tokens": 43718268.0,
      "step": 19955
    },
    {
      "entropy": 5.595753145217896,
      "epoch": 1.980944819372767,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046106723522102024,
      "loss": 5.1433,
      "mean_token_accuracy": 0.16799507886171341,
      "num_tokens": 43729853.0,
      "step": 19960
    },
    {
      "entropy": 5.612300443649292,
      "epoch": 1.9814410480349345,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046104731146035904,
      "loss": 5.1276,
      "mean_token_accuracy": 0.17237960547208786,
      "num_tokens": 43739886.0,
      "step": 19965
    },
    {
      "entropy": 5.637298345565796,
      "epoch": 1.981937276697102,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004610273830861255,
      "loss": 5.1196,
      "mean_token_accuracy": 0.1705731600522995,
      "num_tokens": 43750536.0,
      "step": 19970
    },
    {
      "entropy": 5.655373620986938,
      "epoch": 1.9824335053592694,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004610074500988137,
      "loss": 5.093,
      "mean_token_accuracy": 0.180299611389637,
      "num_tokens": 43763192.0,
      "step": 19975
    },
    {
      "entropy": 5.632276248931885,
      "epoch": 1.9829297340214371,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004609875124989179,
      "loss": 5.1359,
      "mean_token_accuracy": 0.17786213606595994,
      "num_tokens": 43774117.0,
      "step": 19980
    },
    {
      "entropy": 5.578018236160278,
      "epoch": 1.9834259626836046,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004609675702869326,
      "loss": 5.068,
      "mean_token_accuracy": 0.17708240151405336,
      "num_tokens": 43784708.0,
      "step": 19985
    },
    {
      "entropy": 5.605779457092285,
      "epoch": 1.983922191345772,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004609476234633523,
      "loss": 5.1113,
      "mean_token_accuracy": 0.17942753732204436,
      "num_tokens": 43796767.0,
      "step": 19990
    },
    {
      "entropy": 5.669137096405029,
      "epoch": 1.9844184200079398,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046092767202867156,
      "loss": 5.1907,
      "mean_token_accuracy": 0.16681499779224396,
      "num_tokens": 43807203.0,
      "step": 19995
    },
    {
      "entropy": 5.609438228607178,
      "epoch": 1.9849146486701073,
      "grad_norm": 1.328125,
      "learning_rate": 0.00046090771598338525,
      "loss": 5.1117,
      "mean_token_accuracy": 0.175530044734478,
      "num_tokens": 43817123.0,
      "step": 20000
    },
    {
      "entropy": 5.62210259437561,
      "epoch": 1.9854108773322747,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004608877553279881,
      "loss": 5.0748,
      "mean_token_accuracy": 0.17653139680624008,
      "num_tokens": 43828625.0,
      "step": 20005
    },
    {
      "entropy": 5.572797203063965,
      "epoch": 1.9859071059944422,
      "grad_norm": 1.0,
      "learning_rate": 0.0004608677900629753,
      "loss": 5.0529,
      "mean_token_accuracy": 0.1816039502620697,
      "num_tokens": 43840012.0,
      "step": 20010
    },
    {
      "entropy": 5.515506887435913,
      "epoch": 1.9864033346566097,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046084782018884177,
      "loss": 5.0588,
      "mean_token_accuracy": 0.17494238615036012,
      "num_tokens": 43851253.0,
      "step": 20015
    },
    {
      "entropy": 5.6822045803070065,
      "epoch": 1.9868995633187772,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046082784570608266,
      "loss": 5.16,
      "mean_token_accuracy": 0.17430135011672973,
      "num_tokens": 43864389.0,
      "step": 20020
    },
    {
      "entropy": 5.6582173824310305,
      "epoch": 1.9873957919809448,
      "grad_norm": 1.0,
      "learning_rate": 0.00046080786661519354,
      "loss": 5.1539,
      "mean_token_accuracy": 0.16898141652345658,
      "num_tokens": 43876446.0,
      "step": 20025
    },
    {
      "entropy": 5.650602054595947,
      "epoch": 1.9878920206431123,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004607878829166696,
      "loss": 5.104,
      "mean_token_accuracy": 0.1714298814535141,
      "num_tokens": 43888318.0,
      "step": 20030
    },
    {
      "entropy": 5.607369995117187,
      "epoch": 1.98838824930528,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046076789461100653,
      "loss": 5.0516,
      "mean_token_accuracy": 0.18318646997213364,
      "num_tokens": 43898358.0,
      "step": 20035
    },
    {
      "entropy": 5.629449844360352,
      "epoch": 1.9888844779674475,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004607479016987001,
      "loss": 5.1066,
      "mean_token_accuracy": 0.1762085363268852,
      "num_tokens": 43909831.0,
      "step": 20040
    },
    {
      "entropy": 5.631558418273926,
      "epoch": 1.989380706629615,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004607279041802458,
      "loss": 5.0665,
      "mean_token_accuracy": 0.17115451991558076,
      "num_tokens": 43919648.0,
      "step": 20045
    },
    {
      "entropy": 5.599607133865357,
      "epoch": 1.9898769352917824,
      "grad_norm": 0.90625,
      "learning_rate": 0.00046070790205613967,
      "loss": 5.1087,
      "mean_token_accuracy": 0.17320573478937148,
      "num_tokens": 43930563.0,
      "step": 20050
    },
    {
      "entropy": 5.647647380828857,
      "epoch": 1.99037316395395,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046068789532687784,
      "loss": 5.1869,
      "mean_token_accuracy": 0.1690903663635254,
      "num_tokens": 43942490.0,
      "step": 20055
    },
    {
      "entropy": 5.568455553054809,
      "epoch": 1.9908693926161174,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004606678839929563,
      "loss": 4.9558,
      "mean_token_accuracy": 0.19183146506547927,
      "num_tokens": 43952187.0,
      "step": 20060
    },
    {
      "entropy": 5.567154407501221,
      "epoch": 1.9913656212782849,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004606478680548713,
      "loss": 5.0278,
      "mean_token_accuracy": 0.1801212102174759,
      "num_tokens": 43964203.0,
      "step": 20065
    },
    {
      "entropy": 5.5997357845306395,
      "epoch": 1.9918618499404526,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046062784751311917,
      "loss": 5.1079,
      "mean_token_accuracy": 0.17442668229341507,
      "num_tokens": 43974048.0,
      "step": 20070
    },
    {
      "entropy": 5.6581462860107425,
      "epoch": 1.99235807860262,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046060782236819647,
      "loss": 5.0935,
      "mean_token_accuracy": 0.17656888216733932,
      "num_tokens": 43984469.0,
      "step": 20075
    },
    {
      "entropy": 5.580371856689453,
      "epoch": 1.9928543072647877,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004605877926205996,
      "loss": 5.114,
      "mean_token_accuracy": 0.1784413903951645,
      "num_tokens": 43995494.0,
      "step": 20080
    },
    {
      "entropy": 5.609106969833374,
      "epoch": 1.9933505359269552,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004605677582708256,
      "loss": 5.0393,
      "mean_token_accuracy": 0.18157879263162613,
      "num_tokens": 44005869.0,
      "step": 20085
    },
    {
      "entropy": 5.692487573623657,
      "epoch": 1.9938467645891227,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004605477193193709,
      "loss": 5.1092,
      "mean_token_accuracy": 0.1742992639541626,
      "num_tokens": 44016956.0,
      "step": 20090
    },
    {
      "entropy": 5.661140203475952,
      "epoch": 1.9943429932512902,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004605276757667326,
      "loss": 5.1706,
      "mean_token_accuracy": 0.16664988696575164,
      "num_tokens": 44029524.0,
      "step": 20095
    },
    {
      "entropy": 5.644287776947022,
      "epoch": 1.9948392219134576,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004605076276134078,
      "loss": 5.0616,
      "mean_token_accuracy": 0.18139172494411468,
      "num_tokens": 44039523.0,
      "step": 20100
    },
    {
      "entropy": 5.703071546554566,
      "epoch": 1.9953354505756251,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046048757485989345,
      "loss": 5.1663,
      "mean_token_accuracy": 0.1708271697163582,
      "num_tokens": 44050538.0,
      "step": 20105
    },
    {
      "entropy": 5.663781356811524,
      "epoch": 1.9958316792377928,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000460467517506687,
      "loss": 5.1532,
      "mean_token_accuracy": 0.16987523585557937,
      "num_tokens": 44061108.0,
      "step": 20110
    },
    {
      "entropy": 5.58508677482605,
      "epoch": 1.9963279078999603,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004604474555542858,
      "loss": 5.0811,
      "mean_token_accuracy": 0.18187492191791535,
      "num_tokens": 44072706.0,
      "step": 20115
    },
    {
      "entropy": 5.623452520370483,
      "epoch": 1.996824136562128,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004604273890031872,
      "loss": 5.0818,
      "mean_token_accuracy": 0.17422409057617189,
      "num_tokens": 44082697.0,
      "step": 20120
    },
    {
      "entropy": 5.559753942489624,
      "epoch": 1.9973203652242955,
      "grad_norm": 1.0,
      "learning_rate": 0.000460407317853889,
      "loss": 5.0633,
      "mean_token_accuracy": 0.17895758002996445,
      "num_tokens": 44093265.0,
      "step": 20125
    },
    {
      "entropy": 5.587247562408447,
      "epoch": 1.997816593886463,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004603872421068888,
      "loss": 5.0584,
      "mean_token_accuracy": 0.17357878983020783,
      "num_tokens": 44103535.0,
      "step": 20130
    },
    {
      "entropy": 5.602614116668701,
      "epoch": 1.9983128225486304,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004603671617626845,
      "loss": 5.0467,
      "mean_token_accuracy": 0.17467122673988342,
      "num_tokens": 44113903.0,
      "step": 20135
    },
    {
      "entropy": 5.6297746181488035,
      "epoch": 1.9988090512107979,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046034707682177393,
      "loss": 5.0886,
      "mean_token_accuracy": 0.18271566480398177,
      "num_tokens": 44125148.0,
      "step": 20140
    },
    {
      "entropy": 5.6896240234375,
      "epoch": 1.9993052798729654,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004603269872846554,
      "loss": 5.1875,
      "mean_token_accuracy": 0.1657382383942604,
      "num_tokens": 44136563.0,
      "step": 20145
    },
    {
      "entropy": 5.689140558242798,
      "epoch": 1.9998015085351328,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046030689315182676,
      "loss": 5.1462,
      "mean_token_accuracy": 0.17412015050649643,
      "num_tokens": 44147521.0,
      "step": 20150
    },
    {
      "entropy": 5.670950126647949,
      "epoch": 2.0002977371973003,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004602867944237866,
      "loss": 5.1311,
      "mean_token_accuracy": 0.1809731751680374,
      "num_tokens": 44158977.0,
      "step": 20155
    },
    {
      "entropy": 5.6737658977508545,
      "epoch": 2.0007939658594682,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046026669110103315,
      "loss": 5.095,
      "mean_token_accuracy": 0.1784215524792671,
      "num_tokens": 44170986.0,
      "step": 20160
    },
    {
      "entropy": 5.720340061187744,
      "epoch": 2.0012901945216357,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046024658318406493,
      "loss": 5.1454,
      "mean_token_accuracy": 0.17444974780082703,
      "num_tokens": 44180936.0,
      "step": 20165
    },
    {
      "entropy": 5.714664888381958,
      "epoch": 2.001786423183803,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004602264706733807,
      "loss": 5.1416,
      "mean_token_accuracy": 0.1728948548436165,
      "num_tokens": 44191741.0,
      "step": 20170
    },
    {
      "entropy": 5.599285459518432,
      "epoch": 2.0022826518459707,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004602063535694791,
      "loss": 5.0142,
      "mean_token_accuracy": 0.17565245777368546,
      "num_tokens": 44202310.0,
      "step": 20175
    },
    {
      "entropy": 5.621184968948365,
      "epoch": 2.002778880508138,
      "grad_norm": 0.96875,
      "learning_rate": 0.000460186231872859,
      "loss": 5.0752,
      "mean_token_accuracy": 0.1805661901831627,
      "num_tokens": 44213514.0,
      "step": 20180
    },
    {
      "entropy": 5.693288803100586,
      "epoch": 2.0032751091703056,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046016610558401936,
      "loss": 5.1034,
      "mean_token_accuracy": 0.17415388375520707,
      "num_tokens": 44224718.0,
      "step": 20185
    },
    {
      "entropy": 5.667742300033569,
      "epoch": 2.003771337832473,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046014597470345935,
      "loss": 5.1145,
      "mean_token_accuracy": 0.17893408983945847,
      "num_tokens": 44235999.0,
      "step": 20190
    },
    {
      "entropy": 5.612165927886963,
      "epoch": 2.0042675664946406,
      "grad_norm": 1.0,
      "learning_rate": 0.0004601258392316781,
      "loss": 4.9687,
      "mean_token_accuracy": 0.18359076976776123,
      "num_tokens": 44246323.0,
      "step": 20195
    },
    {
      "entropy": 5.593505954742431,
      "epoch": 2.0047637951568085,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046010569916917496,
      "loss": 5.0058,
      "mean_token_accuracy": 0.18122074902057647,
      "num_tokens": 44256323.0,
      "step": 20200
    },
    {
      "entropy": 5.651279449462891,
      "epoch": 2.005260023818976,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004600855545164493,
      "loss": 4.9982,
      "mean_token_accuracy": 0.17736129015684127,
      "num_tokens": 44266845.0,
      "step": 20205
    },
    {
      "entropy": 5.681767082214355,
      "epoch": 2.0057562524811434,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004600654052740008,
      "loss": 5.1003,
      "mean_token_accuracy": 0.180557282269001,
      "num_tokens": 44277861.0,
      "step": 20210
    },
    {
      "entropy": 5.633938026428223,
      "epoch": 2.006252481143311,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004600452514423289,
      "loss": 5.0478,
      "mean_token_accuracy": 0.18581153750419616,
      "num_tokens": 44288511.0,
      "step": 20215
    },
    {
      "entropy": 5.61785101890564,
      "epoch": 2.0067487098054784,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004600250930219336,
      "loss": 5.1459,
      "mean_token_accuracy": 0.16722883135080338,
      "num_tokens": 44299752.0,
      "step": 20220
    },
    {
      "entropy": 5.585857009887695,
      "epoch": 2.007244938467646,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046000493001331465,
      "loss": 4.9925,
      "mean_token_accuracy": 0.18405088186264038,
      "num_tokens": 44309443.0,
      "step": 20225
    },
    {
      "entropy": 5.647888088226319,
      "epoch": 2.0077411671298133,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045998476241697214,
      "loss": 5.0399,
      "mean_token_accuracy": 0.18389932215213775,
      "num_tokens": 44320030.0,
      "step": 20230
    },
    {
      "entropy": 5.585946798324585,
      "epoch": 2.008237395791981,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004599645902334061,
      "loss": 4.9415,
      "mean_token_accuracy": 0.18276743739843368,
      "num_tokens": 44330077.0,
      "step": 20235
    },
    {
      "entropy": 5.533535814285278,
      "epoch": 2.0087336244541483,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045994441346311677,
      "loss": 5.0105,
      "mean_token_accuracy": 0.18256353735923767,
      "num_tokens": 44340336.0,
      "step": 20240
    },
    {
      "entropy": 5.640858125686646,
      "epoch": 2.009229853116316,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004599242321066045,
      "loss": 5.0011,
      "mean_token_accuracy": 0.18143666833639144,
      "num_tokens": 44350704.0,
      "step": 20245
    },
    {
      "entropy": 5.671515464782715,
      "epoch": 2.0097260817784837,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045990404616436976,
      "loss": 5.0423,
      "mean_token_accuracy": 0.1835341766476631,
      "num_tokens": 44360541.0,
      "step": 20250
    },
    {
      "entropy": 5.551262950897216,
      "epoch": 2.010222310440651,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004598838556369131,
      "loss": 4.9559,
      "mean_token_accuracy": 0.18755675703287125,
      "num_tokens": 44372132.0,
      "step": 20255
    },
    {
      "entropy": 5.542077112197876,
      "epoch": 2.0107185391028186,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00045986366052473525,
      "loss": 4.9482,
      "mean_token_accuracy": 0.19246780574321748,
      "num_tokens": 44383594.0,
      "step": 20260
    },
    {
      "entropy": 5.553410959243775,
      "epoch": 2.011214767764986,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004598434608283369,
      "loss": 4.9673,
      "mean_token_accuracy": 0.18205159902572632,
      "num_tokens": 44394101.0,
      "step": 20265
    },
    {
      "entropy": 5.654728651046753,
      "epoch": 2.0117109964271536,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004598232565482191,
      "loss": 5.0891,
      "mean_token_accuracy": 0.17646175771951675,
      "num_tokens": 44405670.0,
      "step": 20270
    },
    {
      "entropy": 5.665910816192627,
      "epoch": 2.012207225089321,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00045980304768488274,
      "loss": 5.0535,
      "mean_token_accuracy": 0.1805179163813591,
      "num_tokens": 44417748.0,
      "step": 20275
    },
    {
      "entropy": 5.675211477279663,
      "epoch": 2.0127034537514885,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045978283423882906,
      "loss": 5.0371,
      "mean_token_accuracy": 0.17577939331531525,
      "num_tokens": 44429648.0,
      "step": 20280
    },
    {
      "entropy": 5.601581668853759,
      "epoch": 2.0131996824136564,
      "grad_norm": 1.28125,
      "learning_rate": 0.00045976261621055915,
      "loss": 5.0314,
      "mean_token_accuracy": 0.1778119221329689,
      "num_tokens": 44440867.0,
      "step": 20285
    },
    {
      "entropy": 5.585496854782105,
      "epoch": 2.013695911075824,
      "grad_norm": 1.0,
      "learning_rate": 0.00045974239360057457,
      "loss": 5.0648,
      "mean_token_accuracy": 0.17971462160348892,
      "num_tokens": 44452981.0,
      "step": 20290
    },
    {
      "entropy": 5.62352876663208,
      "epoch": 2.0141921397379914,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045972216640937676,
      "loss": 4.9889,
      "mean_token_accuracy": 0.18160634785890578,
      "num_tokens": 44464243.0,
      "step": 20295
    },
    {
      "entropy": 5.6410595893859865,
      "epoch": 2.014688368400159,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004597019346374672,
      "loss": 5.0782,
      "mean_token_accuracy": 0.17880146950483322,
      "num_tokens": 44474376.0,
      "step": 20300
    },
    {
      "entropy": 5.5946728706359865,
      "epoch": 2.0151845970623263,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004596816982853477,
      "loss": 4.9589,
      "mean_token_accuracy": 0.19164983183145523,
      "num_tokens": 44485686.0,
      "step": 20305
    },
    {
      "entropy": 5.531017017364502,
      "epoch": 2.015680825724494,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045966145735352,
      "loss": 4.9806,
      "mean_token_accuracy": 0.18666768372058867,
      "num_tokens": 44496096.0,
      "step": 20310
    },
    {
      "entropy": 5.66385850906372,
      "epoch": 2.0161770543866613,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045964121184248604,
      "loss": 5.0946,
      "mean_token_accuracy": 0.17521959096193312,
      "num_tokens": 44507555.0,
      "step": 20315
    },
    {
      "entropy": 5.648081636428833,
      "epoch": 2.0166732830488288,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004596209617527479,
      "loss": 5.0469,
      "mean_token_accuracy": 0.1782549023628235,
      "num_tokens": 44519288.0,
      "step": 20320
    },
    {
      "entropy": 5.684654998779297,
      "epoch": 2.0171695117109962,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004596007070848078,
      "loss": 5.0977,
      "mean_token_accuracy": 0.16740576624870301,
      "num_tokens": 44530146.0,
      "step": 20325
    },
    {
      "entropy": 5.618851137161255,
      "epoch": 2.017665740373164,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045958044783916795,
      "loss": 5.0672,
      "mean_token_accuracy": 0.1778162971138954,
      "num_tokens": 44541411.0,
      "step": 20330
    },
    {
      "entropy": 5.566889047622681,
      "epoch": 2.0181619690353316,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004595601840163306,
      "loss": 5.0331,
      "mean_token_accuracy": 0.17437337189912797,
      "num_tokens": 44552556.0,
      "step": 20335
    },
    {
      "entropy": 5.611506032943725,
      "epoch": 2.018658197697499,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045953991561679855,
      "loss": 5.0531,
      "mean_token_accuracy": 0.18028065264225007,
      "num_tokens": 44563052.0,
      "step": 20340
    },
    {
      "entropy": 5.592464828491211,
      "epoch": 2.0191544263596666,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004595196426410742,
      "loss": 5.0084,
      "mean_token_accuracy": 0.18068556785583495,
      "num_tokens": 44574020.0,
      "step": 20345
    },
    {
      "entropy": 5.552760171890259,
      "epoch": 2.019650655021834,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045949936508966015,
      "loss": 4.9659,
      "mean_token_accuracy": 0.18475698679685593,
      "num_tokens": 44586261.0,
      "step": 20350
    },
    {
      "entropy": 5.641411352157593,
      "epoch": 2.0201468836840015,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004594790829630596,
      "loss": 5.0208,
      "mean_token_accuracy": 0.1739213600754738,
      "num_tokens": 44596391.0,
      "step": 20355
    },
    {
      "entropy": 5.589959526062012,
      "epoch": 2.020643112346169,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004594587962617752,
      "loss": 4.983,
      "mean_token_accuracy": 0.18745748996734618,
      "num_tokens": 44606263.0,
      "step": 20360
    },
    {
      "entropy": 5.616388559341431,
      "epoch": 2.0211393410083365,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004594385049863102,
      "loss": 5.0372,
      "mean_token_accuracy": 0.17638513296842576,
      "num_tokens": 44617332.0,
      "step": 20365
    },
    {
      "entropy": 5.629624891281128,
      "epoch": 2.021635569670504,
      "grad_norm": 1.4375,
      "learning_rate": 0.00045941820913716766,
      "loss": 5.0408,
      "mean_token_accuracy": 0.19338113069534302,
      "num_tokens": 44629784.0,
      "step": 20370
    },
    {
      "entropy": 5.689560842514038,
      "epoch": 2.022131798332672,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045939790871485093,
      "loss": 5.1417,
      "mean_token_accuracy": 0.17355988025665284,
      "num_tokens": 44641398.0,
      "step": 20375
    },
    {
      "entropy": 5.548549699783325,
      "epoch": 2.0226280269948393,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045937760371986334,
      "loss": 4.9072,
      "mean_token_accuracy": 0.19475265592336655,
      "num_tokens": 44652592.0,
      "step": 20380
    },
    {
      "entropy": 5.637330865859985,
      "epoch": 2.023124255657007,
      "grad_norm": 1.71875,
      "learning_rate": 0.00045935729415270857,
      "loss": 4.987,
      "mean_token_accuracy": 0.1953703820705414,
      "num_tokens": 44663896.0,
      "step": 20385
    },
    {
      "entropy": 5.699647426605225,
      "epoch": 2.0236204843191743,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004593369800138901,
      "loss": 5.1459,
      "mean_token_accuracy": 0.17223415672779083,
      "num_tokens": 44675004.0,
      "step": 20390
    },
    {
      "entropy": 5.6865886688232425,
      "epoch": 2.0241167129813418,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004593166613039118,
      "loss": 5.0857,
      "mean_token_accuracy": 0.17750537991523743,
      "num_tokens": 44686778.0,
      "step": 20395
    },
    {
      "entropy": 5.597451019287109,
      "epoch": 2.0246129416435092,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045929633802327743,
      "loss": 5.0111,
      "mean_token_accuracy": 0.18518262207508088,
      "num_tokens": 44697600.0,
      "step": 20400
    },
    {
      "entropy": 5.591700172424316,
      "epoch": 2.0251091703056767,
      "grad_norm": 1.046875,
      "learning_rate": 0.000459276010172491,
      "loss": 5.07,
      "mean_token_accuracy": 0.17683031111955644,
      "num_tokens": 44707453.0,
      "step": 20405
    },
    {
      "entropy": 5.618640804290772,
      "epoch": 2.025605398967844,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045925567775205656,
      "loss": 5.1304,
      "mean_token_accuracy": 0.17077885419130326,
      "num_tokens": 44718351.0,
      "step": 20410
    },
    {
      "entropy": 5.650522327423095,
      "epoch": 2.026101627630012,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045923534076247846,
      "loss": 4.9772,
      "mean_token_accuracy": 0.18641198426485062,
      "num_tokens": 44728884.0,
      "step": 20415
    },
    {
      "entropy": 5.623927593231201,
      "epoch": 2.0265978562921796,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045921499920426083,
      "loss": 5.029,
      "mean_token_accuracy": 0.1782628118991852,
      "num_tokens": 44740769.0,
      "step": 20420
    },
    {
      "entropy": 5.657447338104248,
      "epoch": 2.027094084954347,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045919465307790817,
      "loss": 5.1401,
      "mean_token_accuracy": 0.17781321108341216,
      "num_tokens": 44753094.0,
      "step": 20425
    },
    {
      "entropy": 5.532738876342774,
      "epoch": 2.0275903136165145,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045917430238392505,
      "loss": 4.8955,
      "mean_token_accuracy": 0.19074820280075072,
      "num_tokens": 44762996.0,
      "step": 20430
    },
    {
      "entropy": 5.554237222671508,
      "epoch": 2.028086542278682,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000459153947122816,
      "loss": 5.0828,
      "mean_token_accuracy": 0.18193011581897736,
      "num_tokens": 44775343.0,
      "step": 20435
    },
    {
      "entropy": 5.528848266601562,
      "epoch": 2.0285827709408495,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004591335872950859,
      "loss": 4.9291,
      "mean_token_accuracy": 0.18780219703912734,
      "num_tokens": 44785383.0,
      "step": 20440
    },
    {
      "entropy": 5.595623207092285,
      "epoch": 2.029078999603017,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004591132229012397,
      "loss": 4.9965,
      "mean_token_accuracy": 0.18310811519622802,
      "num_tokens": 44796805.0,
      "step": 20445
    },
    {
      "entropy": 5.580492401123047,
      "epoch": 2.0295752282651844,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045909285394178217,
      "loss": 5.0001,
      "mean_token_accuracy": 0.18979256898164748,
      "num_tokens": 44806800.0,
      "step": 20450
    },
    {
      "entropy": 5.611713504791259,
      "epoch": 2.030071456927352,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004590724804172187,
      "loss": 5.1179,
      "mean_token_accuracy": 0.17173586785793304,
      "num_tokens": 44818745.0,
      "step": 20455
    },
    {
      "entropy": 5.583092403411865,
      "epoch": 2.03056768558952,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004590521023280543,
      "loss": 4.983,
      "mean_token_accuracy": 0.18066620081663132,
      "num_tokens": 44829825.0,
      "step": 20460
    },
    {
      "entropy": 5.583886814117432,
      "epoch": 2.0310639142516873,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045903171967479425,
      "loss": 4.9889,
      "mean_token_accuracy": 0.18530441969633102,
      "num_tokens": 44841742.0,
      "step": 20465
    },
    {
      "entropy": 5.632503080368042,
      "epoch": 2.031560142913855,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045901133245794426,
      "loss": 5.1201,
      "mean_token_accuracy": 0.18114353716373444,
      "num_tokens": 44853019.0,
      "step": 20470
    },
    {
      "entropy": 5.605856037139892,
      "epoch": 2.0320563715760223,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004589909406780096,
      "loss": 4.9647,
      "mean_token_accuracy": 0.1825123906135559,
      "num_tokens": 44863147.0,
      "step": 20475
    },
    {
      "entropy": 5.647651481628418,
      "epoch": 2.0325526002381897,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045897054433549617,
      "loss": 5.0212,
      "mean_token_accuracy": 0.18446905314922332,
      "num_tokens": 44873638.0,
      "step": 20480
    },
    {
      "entropy": 5.56237006187439,
      "epoch": 2.033048828900357,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045895014343090963,
      "loss": 5.0299,
      "mean_token_accuracy": 0.17758500725030898,
      "num_tokens": 44885752.0,
      "step": 20485
    },
    {
      "entropy": 5.554017829895019,
      "epoch": 2.0335450575625247,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00045892973796475594,
      "loss": 4.986,
      "mean_token_accuracy": 0.18601725101470948,
      "num_tokens": 44897471.0,
      "step": 20490
    },
    {
      "entropy": 5.674320650100708,
      "epoch": 2.034041286224692,
      "grad_norm": 1.09375,
      "learning_rate": 0.000458909327937541,
      "loss": 5.0676,
      "mean_token_accuracy": 0.17395548671483993,
      "num_tokens": 44907236.0,
      "step": 20495
    },
    {
      "entropy": 5.621460628509522,
      "epoch": 2.03453751488686,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045888891334977105,
      "loss": 4.993,
      "mean_token_accuracy": 0.1877368465065956,
      "num_tokens": 44918103.0,
      "step": 20500
    },
    {
      "entropy": 5.586451196670533,
      "epoch": 2.0350337435490276,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004588684942019524,
      "loss": 5.0479,
      "mean_token_accuracy": 0.18042974323034286,
      "num_tokens": 44928263.0,
      "step": 20505
    },
    {
      "entropy": 5.572514629364013,
      "epoch": 2.035529972211195,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045884807049459116,
      "loss": 5.0121,
      "mean_token_accuracy": 0.18163361847400666,
      "num_tokens": 44939487.0,
      "step": 20510
    },
    {
      "entropy": 5.623572921752929,
      "epoch": 2.0360262008733625,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000458827642228194,
      "loss": 5.0956,
      "mean_token_accuracy": 0.17479878962039946,
      "num_tokens": 44949749.0,
      "step": 20515
    },
    {
      "entropy": 5.611263084411621,
      "epoch": 2.03652242953553,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045880720940326744,
      "loss": 5.0335,
      "mean_token_accuracy": 0.1824065014719963,
      "num_tokens": 44960347.0,
      "step": 20520
    },
    {
      "entropy": 5.675039529800415,
      "epoch": 2.0370186581976975,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045878677202031804,
      "loss": 5.1042,
      "mean_token_accuracy": 0.17418878078460692,
      "num_tokens": 44970420.0,
      "step": 20525
    },
    {
      "entropy": 5.582677793502808,
      "epoch": 2.037514886859865,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004587663300798529,
      "loss": 5.0175,
      "mean_token_accuracy": 0.18048970997333527,
      "num_tokens": 44980888.0,
      "step": 20530
    },
    {
      "entropy": 5.625572681427002,
      "epoch": 2.0380111155220324,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004587458835823786,
      "loss": 5.0371,
      "mean_token_accuracy": 0.17729188799858092,
      "num_tokens": 44992721.0,
      "step": 20535
    },
    {
      "entropy": 5.600467300415039,
      "epoch": 2.0385073441842,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045872543252840236,
      "loss": 5.0005,
      "mean_token_accuracy": 0.1848856568336487,
      "num_tokens": 45003603.0,
      "step": 20540
    },
    {
      "entropy": 5.600116205215454,
      "epoch": 2.039003572846368,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045870497691843134,
      "loss": 5.0361,
      "mean_token_accuracy": 0.1864559456706047,
      "num_tokens": 45014864.0,
      "step": 20545
    },
    {
      "entropy": 5.650608253479004,
      "epoch": 2.0394998015085353,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004586845167529727,
      "loss": 5.0595,
      "mean_token_accuracy": 0.18074969500303267,
      "num_tokens": 45027766.0,
      "step": 20550
    },
    {
      "entropy": 5.5784200668334964,
      "epoch": 2.0399960301707027,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00045866405203253384,
      "loss": 5.0293,
      "mean_token_accuracy": 0.17954335659742354,
      "num_tokens": 45039524.0,
      "step": 20555
    },
    {
      "entropy": 5.630284547805786,
      "epoch": 2.04049225883287,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004586435827576223,
      "loss": 5.0584,
      "mean_token_accuracy": 0.17280661463737487,
      "num_tokens": 45050848.0,
      "step": 20560
    },
    {
      "entropy": 5.610242128372192,
      "epoch": 2.0409884874950377,
      "grad_norm": 1.0,
      "learning_rate": 0.0004586231089287456,
      "loss": 4.9911,
      "mean_token_accuracy": 0.18106120079755783,
      "num_tokens": 45061229.0,
      "step": 20565
    },
    {
      "entropy": 5.576432132720948,
      "epoch": 2.041484716157205,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004586026305464115,
      "loss": 4.9585,
      "mean_token_accuracy": 0.18340425193309784,
      "num_tokens": 45070466.0,
      "step": 20570
    },
    {
      "entropy": 5.611891460418701,
      "epoch": 2.0419809448193726,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045858214761112777,
      "loss": 5.0443,
      "mean_token_accuracy": 0.17960603833198546,
      "num_tokens": 45081839.0,
      "step": 20575
    },
    {
      "entropy": 5.656205654144287,
      "epoch": 2.04247717348154,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045856166012340235,
      "loss": 5.1078,
      "mean_token_accuracy": 0.18004273474216462,
      "num_tokens": 45092458.0,
      "step": 20580
    },
    {
      "entropy": 5.649285745620728,
      "epoch": 2.0429734021437076,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004585411680837433,
      "loss": 5.0904,
      "mean_token_accuracy": 0.17647339552640914,
      "num_tokens": 45104629.0,
      "step": 20585
    },
    {
      "entropy": 5.634955310821534,
      "epoch": 2.0434696308058755,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045852067149265875,
      "loss": 4.9471,
      "mean_token_accuracy": 0.1830761969089508,
      "num_tokens": 45115556.0,
      "step": 20590
    },
    {
      "entropy": 5.603266286849975,
      "epoch": 2.043965859468043,
      "grad_norm": 0.953125,
      "learning_rate": 0.000458500170350657,
      "loss": 5.0176,
      "mean_token_accuracy": 0.17891958057880403,
      "num_tokens": 45126489.0,
      "step": 20595
    },
    {
      "entropy": 5.631175136566162,
      "epoch": 2.0444620881302105,
      "grad_norm": 1.0,
      "learning_rate": 0.00045847966465824654,
      "loss": 5.1288,
      "mean_token_accuracy": 0.17655792981386184,
      "num_tokens": 45138882.0,
      "step": 20600
    },
    {
      "entropy": 5.669561433792114,
      "epoch": 2.044958316792378,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004584591544159356,
      "loss": 5.0657,
      "mean_token_accuracy": 0.18042346090078354,
      "num_tokens": 45149941.0,
      "step": 20605
    },
    {
      "entropy": 5.71926589012146,
      "epoch": 2.0454545454545454,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004584386396242331,
      "loss": 5.0811,
      "mean_token_accuracy": 0.18266470730304718,
      "num_tokens": 45162267.0,
      "step": 20610
    },
    {
      "entropy": 5.65542688369751,
      "epoch": 2.045950774116713,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045841812028364756,
      "loss": 5.17,
      "mean_token_accuracy": 0.1766157180070877,
      "num_tokens": 45174444.0,
      "step": 20615
    },
    {
      "entropy": 5.593732261657715,
      "epoch": 2.0464470027788804,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045839759639468774,
      "loss": 5.0215,
      "mean_token_accuracy": 0.18275284618139268,
      "num_tokens": 45183957.0,
      "step": 20620
    },
    {
      "entropy": 5.660202932357788,
      "epoch": 2.046943231441048,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004583770679578628,
      "loss": 5.11,
      "mean_token_accuracy": 0.18082353174686433,
      "num_tokens": 45195607.0,
      "step": 20625
    },
    {
      "entropy": 5.593073987960816,
      "epoch": 2.0474394601032158,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004583565349736818,
      "loss": 5.0021,
      "mean_token_accuracy": 0.1832710862159729,
      "num_tokens": 45205627.0,
      "step": 20630
    },
    {
      "entropy": 5.561463737487793,
      "epoch": 2.0479356887653832,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004583359974426538,
      "loss": 5.0537,
      "mean_token_accuracy": 0.17463009804487228,
      "num_tokens": 45216331.0,
      "step": 20635
    },
    {
      "entropy": 5.593317556381225,
      "epoch": 2.0484319174275507,
      "grad_norm": 1.0,
      "learning_rate": 0.00045831545536528806,
      "loss": 5.0748,
      "mean_token_accuracy": 0.172257898747921,
      "num_tokens": 45227826.0,
      "step": 20640
    },
    {
      "entropy": 5.612700128555298,
      "epoch": 2.048928146089718,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045829490874209404,
      "loss": 5.0029,
      "mean_token_accuracy": 0.18185341358184814,
      "num_tokens": 45238093.0,
      "step": 20645
    },
    {
      "entropy": 5.597904539108276,
      "epoch": 2.0494243747518857,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045827435757358136,
      "loss": 5.0345,
      "mean_token_accuracy": 0.18198850452899934,
      "num_tokens": 45248660.0,
      "step": 20650
    },
    {
      "entropy": 5.671583795547486,
      "epoch": 2.049920603414053,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004582538018602594,
      "loss": 5.1078,
      "mean_token_accuracy": 0.17468267679214478,
      "num_tokens": 45261059.0,
      "step": 20655
    },
    {
      "entropy": 5.6259458541870115,
      "epoch": 2.0504168320762206,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004582332416026381,
      "loss": 5.0213,
      "mean_token_accuracy": 0.1784760072827339,
      "num_tokens": 45272210.0,
      "step": 20660
    },
    {
      "entropy": 5.6416059017181395,
      "epoch": 2.050913060738388,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045821267680122725,
      "loss": 5.0977,
      "mean_token_accuracy": 0.18041985481977463,
      "num_tokens": 45283488.0,
      "step": 20665
    },
    {
      "entropy": 5.590239667892456,
      "epoch": 2.0514092894005556,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004581921074565368,
      "loss": 4.9506,
      "mean_token_accuracy": 0.18721193373203276,
      "num_tokens": 45292824.0,
      "step": 20670
    },
    {
      "entropy": 5.601603937149048,
      "epoch": 2.0519055180627235,
      "grad_norm": 1.0,
      "learning_rate": 0.0004581715335690768,
      "loss": 5.0323,
      "mean_token_accuracy": 0.18196602016687394,
      "num_tokens": 45303102.0,
      "step": 20675
    },
    {
      "entropy": 5.577046346664429,
      "epoch": 2.052401746724891,
      "grad_norm": 1.0,
      "learning_rate": 0.00045815095513935747,
      "loss": 4.9544,
      "mean_token_accuracy": 0.17890697419643403,
      "num_tokens": 45313825.0,
      "step": 20680
    },
    {
      "entropy": 5.569059038162232,
      "epoch": 2.0528979753870584,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004581303721678891,
      "loss": 4.9984,
      "mean_token_accuracy": 0.18975316286087035,
      "num_tokens": 45325789.0,
      "step": 20685
    },
    {
      "entropy": 5.616703939437866,
      "epoch": 2.053394204049226,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004581097846551821,
      "loss": 5.0773,
      "mean_token_accuracy": 0.17374017536640168,
      "num_tokens": 45337630.0,
      "step": 20690
    },
    {
      "entropy": 5.666081047058105,
      "epoch": 2.0538904327113934,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000458089192601747,
      "loss": 5.0952,
      "mean_token_accuracy": 0.17444723844528198,
      "num_tokens": 45349815.0,
      "step": 20695
    },
    {
      "entropy": 5.612442493438721,
      "epoch": 2.054386661373561,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045806859600809455,
      "loss": 5.031,
      "mean_token_accuracy": 0.17729373276233673,
      "num_tokens": 45360741.0,
      "step": 20700
    },
    {
      "entropy": 5.529368448257446,
      "epoch": 2.0548828900357283,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004580479948747353,
      "loss": 4.9145,
      "mean_token_accuracy": 0.19023348689079284,
      "num_tokens": 45371258.0,
      "step": 20705
    },
    {
      "entropy": 5.607484006881714,
      "epoch": 2.055379118697896,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045802738920218015,
      "loss": 5.0363,
      "mean_token_accuracy": 0.1823964849114418,
      "num_tokens": 45381914.0,
      "step": 20710
    },
    {
      "entropy": 5.568737459182739,
      "epoch": 2.0558753473600637,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004580067789909403,
      "loss": 5.0544,
      "mean_token_accuracy": 0.18593914210796356,
      "num_tokens": 45393159.0,
      "step": 20715
    },
    {
      "entropy": 5.5792993068695065,
      "epoch": 2.056371576022231,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045798616424152654,
      "loss": 4.9284,
      "mean_token_accuracy": 0.1867551848292351,
      "num_tokens": 45403076.0,
      "step": 20720
    },
    {
      "entropy": 5.593299102783203,
      "epoch": 2.0568678046843987,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045796554495445023,
      "loss": 5.0645,
      "mean_token_accuracy": 0.17648866325616835,
      "num_tokens": 45412962.0,
      "step": 20725
    },
    {
      "entropy": 5.587190866470337,
      "epoch": 2.057364033346566,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004579449211302228,
      "loss": 5.0137,
      "mean_token_accuracy": 0.18243680745363236,
      "num_tokens": 45423620.0,
      "step": 20730
    },
    {
      "entropy": 5.616822147369385,
      "epoch": 2.0578602620087336,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004579242927693554,
      "loss": 5.0349,
      "mean_token_accuracy": 0.17443277537822724,
      "num_tokens": 45435901.0,
      "step": 20735
    },
    {
      "entropy": 5.626122856140137,
      "epoch": 2.058356490670901,
      "grad_norm": 1.0,
      "learning_rate": 0.0004579036598723598,
      "loss": 4.9678,
      "mean_token_accuracy": 0.18667516112327576,
      "num_tokens": 45445409.0,
      "step": 20740
    },
    {
      "entropy": 5.69962329864502,
      "epoch": 2.0588527193330686,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045788302243974745,
      "loss": 5.1238,
      "mean_token_accuracy": 0.1815635770559311,
      "num_tokens": 45455794.0,
      "step": 20745
    },
    {
      "entropy": 5.631127977371216,
      "epoch": 2.059348947995236,
      "grad_norm": 1.0,
      "learning_rate": 0.0004578623804720302,
      "loss": 4.9902,
      "mean_token_accuracy": 0.18181696981191636,
      "num_tokens": 45465922.0,
      "step": 20750
    },
    {
      "entropy": 5.571450042724609,
      "epoch": 2.0598451766574035,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004578417339697201,
      "loss": 5.1028,
      "mean_token_accuracy": 0.17274178564548492,
      "num_tokens": 45476314.0,
      "step": 20755
    },
    {
      "entropy": 5.619325399398804,
      "epoch": 2.0603414053195714,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045782108293332884,
      "loss": 5.0296,
      "mean_token_accuracy": 0.1760486215353012,
      "num_tokens": 45487232.0,
      "step": 20760
    },
    {
      "entropy": 5.707307195663452,
      "epoch": 2.060837633981739,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004578004273633687,
      "loss": 5.0823,
      "mean_token_accuracy": 0.1780001550912857,
      "num_tokens": 45498368.0,
      "step": 20765
    },
    {
      "entropy": 5.627378463745117,
      "epoch": 2.0613338626439064,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045777976726035194,
      "loss": 5.0267,
      "mean_token_accuracy": 0.17547286301851273,
      "num_tokens": 45507600.0,
      "step": 20770
    },
    {
      "entropy": 5.63630223274231,
      "epoch": 2.061830091306074,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004577591026247907,
      "loss": 5.1082,
      "mean_token_accuracy": 0.1742859110236168,
      "num_tokens": 45518594.0,
      "step": 20775
    },
    {
      "entropy": 5.587973546981812,
      "epoch": 2.0623263199682413,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004577384334571976,
      "loss": 4.9938,
      "mean_token_accuracy": 0.1823765054345131,
      "num_tokens": 45529929.0,
      "step": 20780
    },
    {
      "entropy": 5.6025307178497314,
      "epoch": 2.062822548630409,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004577177597580851,
      "loss": 4.9938,
      "mean_token_accuracy": 0.18561710864305497,
      "num_tokens": 45540228.0,
      "step": 20785
    },
    {
      "entropy": 5.54388461112976,
      "epoch": 2.0633187772925763,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045769708152796593,
      "loss": 4.9693,
      "mean_token_accuracy": 0.1841318592429161,
      "num_tokens": 45552026.0,
      "step": 20790
    },
    {
      "entropy": 5.549919652938843,
      "epoch": 2.0638150059547438,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045767639876735276,
      "loss": 5.053,
      "mean_token_accuracy": 0.18321210891008377,
      "num_tokens": 45563579.0,
      "step": 20795
    },
    {
      "entropy": 5.676577711105347,
      "epoch": 2.0643112346169117,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004576557114767585,
      "loss": 5.1306,
      "mean_token_accuracy": 0.17222391963005065,
      "num_tokens": 45573800.0,
      "step": 20800
    },
    {
      "entropy": 5.578466796875,
      "epoch": 2.064807463279079,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045763501965669627,
      "loss": 4.9944,
      "mean_token_accuracy": 0.1815763771533966,
      "num_tokens": 45583758.0,
      "step": 20805
    },
    {
      "entropy": 5.609786033630371,
      "epoch": 2.0653036919412466,
      "grad_norm": 0.96875,
      "learning_rate": 0.000457614323307679,
      "loss": 4.9971,
      "mean_token_accuracy": 0.18011555522680284,
      "num_tokens": 45595154.0,
      "step": 20810
    },
    {
      "entropy": 5.670978975296021,
      "epoch": 2.065799920603414,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045759362243022,
      "loss": 5.1087,
      "mean_token_accuracy": 0.17744120061397553,
      "num_tokens": 45606557.0,
      "step": 20815
    },
    {
      "entropy": 5.547169637680054,
      "epoch": 2.0662961492655816,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004575729170248327,
      "loss": 4.9583,
      "mean_token_accuracy": 0.1880551502108574,
      "num_tokens": 45615531.0,
      "step": 20820
    },
    {
      "entropy": 5.514255857467651,
      "epoch": 2.066792377927749,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004575522070920305,
      "loss": 4.9695,
      "mean_token_accuracy": 0.18175002336502075,
      "num_tokens": 45626512.0,
      "step": 20825
    },
    {
      "entropy": 5.634020709991455,
      "epoch": 2.0672886065899165,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045753149263232687,
      "loss": 5.0846,
      "mean_token_accuracy": 0.17448006421327591,
      "num_tokens": 45638005.0,
      "step": 20830
    },
    {
      "entropy": 5.616469287872315,
      "epoch": 2.067784835252084,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004575107736462355,
      "loss": 5.0759,
      "mean_token_accuracy": 0.17737354040145875,
      "num_tokens": 45647740.0,
      "step": 20835
    },
    {
      "entropy": 5.567018175125122,
      "epoch": 2.0682810639142515,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004574900501342703,
      "loss": 5.0013,
      "mean_token_accuracy": 0.18766775280237197,
      "num_tokens": 45659021.0,
      "step": 20840
    },
    {
      "entropy": 5.5955461978912355,
      "epoch": 2.0687772925764194,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004574693220969451,
      "loss": 5.0808,
      "mean_token_accuracy": 0.17484844624996185,
      "num_tokens": 45669413.0,
      "step": 20845
    },
    {
      "entropy": 5.665853977203369,
      "epoch": 2.069273521238587,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004574485895347738,
      "loss": 5.1414,
      "mean_token_accuracy": 0.1747307762503624,
      "num_tokens": 45680719.0,
      "step": 20850
    },
    {
      "entropy": 5.548875761032105,
      "epoch": 2.0697697499007544,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045742785244827066,
      "loss": 4.9558,
      "mean_token_accuracy": 0.1873426467180252,
      "num_tokens": 45690393.0,
      "step": 20855
    },
    {
      "entropy": 5.593914318084717,
      "epoch": 2.070265978562922,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004574071108379499,
      "loss": 5.0768,
      "mean_token_accuracy": 0.17710587680339812,
      "num_tokens": 45701304.0,
      "step": 20860
    },
    {
      "entropy": 5.603142738342285,
      "epoch": 2.0707622072250893,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004573863647043259,
      "loss": 4.9322,
      "mean_token_accuracy": 0.18677033185958863,
      "num_tokens": 45711892.0,
      "step": 20865
    },
    {
      "entropy": 5.608163690567016,
      "epoch": 2.0712584358872568,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004573656140479129,
      "loss": 5.0134,
      "mean_token_accuracy": 0.17741088569164276,
      "num_tokens": 45723693.0,
      "step": 20870
    },
    {
      "entropy": 5.5411004543304445,
      "epoch": 2.0717546645494243,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004573448588692258,
      "loss": 5.0015,
      "mean_token_accuracy": 0.17854353040456772,
      "num_tokens": 45734392.0,
      "step": 20875
    },
    {
      "entropy": 5.492110013961792,
      "epoch": 2.0722508932115917,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000457324099168779,
      "loss": 4.9306,
      "mean_token_accuracy": 0.19073358923196793,
      "num_tokens": 45745732.0,
      "step": 20880
    },
    {
      "entropy": 5.557785224914551,
      "epoch": 2.0727471218737596,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004573033349470874,
      "loss": 5.0083,
      "mean_token_accuracy": 0.1852720096707344,
      "num_tokens": 45756671.0,
      "step": 20885
    },
    {
      "entropy": 5.667295980453491,
      "epoch": 2.073243350535927,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004572825662046661,
      "loss": 5.0947,
      "mean_token_accuracy": 0.17862008064985274,
      "num_tokens": 45767685.0,
      "step": 20890
    },
    {
      "entropy": 5.633484029769898,
      "epoch": 2.0737395791980946,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004572617929420298,
      "loss": 5.0092,
      "mean_token_accuracy": 0.17923321276903154,
      "num_tokens": 45777251.0,
      "step": 20895
    },
    {
      "entropy": 5.654927349090576,
      "epoch": 2.074235807860262,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045724101515969375,
      "loss": 5.0721,
      "mean_token_accuracy": 0.17735732048749925,
      "num_tokens": 45790037.0,
      "step": 20900
    },
    {
      "entropy": 5.6190110206604,
      "epoch": 2.0747320365224295,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045722023285817327,
      "loss": 5.1199,
      "mean_token_accuracy": 0.1708063319325447,
      "num_tokens": 45800803.0,
      "step": 20905
    },
    {
      "entropy": 5.664886569976806,
      "epoch": 2.075228265184597,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004571994460379837,
      "loss": 5.06,
      "mean_token_accuracy": 0.17411066442728043,
      "num_tokens": 45812447.0,
      "step": 20910
    },
    {
      "entropy": 5.6098064422607425,
      "epoch": 2.0757244938467645,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045717865469964045,
      "loss": 5.0075,
      "mean_token_accuracy": 0.17762482315301895,
      "num_tokens": 45822172.0,
      "step": 20915
    },
    {
      "entropy": 5.66619725227356,
      "epoch": 2.076220722508932,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045715785884365914,
      "loss": 5.1325,
      "mean_token_accuracy": 0.17581140398979186,
      "num_tokens": 45832583.0,
      "step": 20920
    },
    {
      "entropy": 5.65828537940979,
      "epoch": 2.0767169511710994,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045713705847055537,
      "loss": 5.1515,
      "mean_token_accuracy": 0.1759900763630867,
      "num_tokens": 45842640.0,
      "step": 20925
    },
    {
      "entropy": 5.6996887683868405,
      "epoch": 2.0772131798332674,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004571162535808451,
      "loss": 5.1437,
      "mean_token_accuracy": 0.1706842377781868,
      "num_tokens": 45854618.0,
      "step": 20930
    },
    {
      "entropy": 5.610805511474609,
      "epoch": 2.077709408495435,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004570954441750441,
      "loss": 5.0112,
      "mean_token_accuracy": 0.18363475054502487,
      "num_tokens": 45866130.0,
      "step": 20935
    },
    {
      "entropy": 5.653029251098633,
      "epoch": 2.0782056371576023,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004570746302536685,
      "loss": 5.0536,
      "mean_token_accuracy": 0.17630828320980071,
      "num_tokens": 45877187.0,
      "step": 20940
    },
    {
      "entropy": 5.603359746932983,
      "epoch": 2.07870186581977,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045705381181723444,
      "loss": 5.0683,
      "mean_token_accuracy": 0.17510530352592468,
      "num_tokens": 45889557.0,
      "step": 20945
    },
    {
      "entropy": 5.6128908634185795,
      "epoch": 2.0791980944819373,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045703298886625804,
      "loss": 5.0765,
      "mean_token_accuracy": 0.1749125227332115,
      "num_tokens": 45900860.0,
      "step": 20950
    },
    {
      "entropy": 5.620050859451294,
      "epoch": 2.0796943231441047,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004570121614012558,
      "loss": 5.0786,
      "mean_token_accuracy": 0.17758319675922393,
      "num_tokens": 45912231.0,
      "step": 20955
    },
    {
      "entropy": 5.66979660987854,
      "epoch": 2.080190551806272,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045699132942274425,
      "loss": 5.1282,
      "mean_token_accuracy": 0.1743239462375641,
      "num_tokens": 45922722.0,
      "step": 20960
    },
    {
      "entropy": 5.596019983291626,
      "epoch": 2.0806867804684397,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045697049293123984,
      "loss": 5.026,
      "mean_token_accuracy": 0.18102947026491165,
      "num_tokens": 45932444.0,
      "step": 20965
    },
    {
      "entropy": 5.637514925003051,
      "epoch": 2.0811830091306076,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004569496519272593,
      "loss": 5.0545,
      "mean_token_accuracy": 0.17931657731533052,
      "num_tokens": 45942826.0,
      "step": 20970
    },
    {
      "entropy": 5.623694992065429,
      "epoch": 2.081679237792775,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045692880641131945,
      "loss": 5.0479,
      "mean_token_accuracy": 0.17944057285785675,
      "num_tokens": 45953935.0,
      "step": 20975
    },
    {
      "entropy": 5.66606969833374,
      "epoch": 2.0821754664549426,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004569079563839372,
      "loss": 5.1359,
      "mean_token_accuracy": 0.17033987790346145,
      "num_tokens": 45964693.0,
      "step": 20980
    },
    {
      "entropy": 5.66500883102417,
      "epoch": 2.08267169511711,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045688710184562965,
      "loss": 5.0242,
      "mean_token_accuracy": 0.1753959134221077,
      "num_tokens": 45976481.0,
      "step": 20985
    },
    {
      "entropy": 5.64786205291748,
      "epoch": 2.0831679237792775,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045686624279691387,
      "loss": 5.1252,
      "mean_token_accuracy": 0.17682384848594665,
      "num_tokens": 45988879.0,
      "step": 20990
    },
    {
      "entropy": 5.601452779769898,
      "epoch": 2.083664152441445,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004568453792383072,
      "loss": 5.0069,
      "mean_token_accuracy": 0.18703122287988663,
      "num_tokens": 45999332.0,
      "step": 20995
    },
    {
      "entropy": 5.62636432647705,
      "epoch": 2.0841603811036125,
      "grad_norm": 1.0625,
      "learning_rate": 0.000456824511170327,
      "loss": 5.0392,
      "mean_token_accuracy": 0.17932495325803757,
      "num_tokens": 46009841.0,
      "step": 21000
    },
    {
      "epoch": 2.0841603811036125,
      "eval_entropy": 5.391679360028768,
      "eval_loss": 5.1821818351745605,
      "eval_mean_token_accuracy": 0.17811451445805626,
      "eval_num_tokens": 46009841.0,
      "eval_runtime": 20.6224,
      "eval_samples_per_second": 1632.641,
      "eval_steps_per_second": 204.098,
      "step": 21000
    },
    {
      "entropy": 5.565670394897461,
      "epoch": 2.08465660976578,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004568036385934906,
      "loss": 4.9693,
      "mean_token_accuracy": 0.17924870699644088,
      "num_tokens": 46020678.0,
      "step": 21005
    },
    {
      "entropy": 5.613858222961426,
      "epoch": 2.0851528384279474,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045678276150831586,
      "loss": 5.0448,
      "mean_token_accuracy": 0.17825375497341156,
      "num_tokens": 46032036.0,
      "step": 21010
    },
    {
      "entropy": 5.66648416519165,
      "epoch": 2.0856490670901153,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004567618799153202,
      "loss": 5.0541,
      "mean_token_accuracy": 0.17988407015800476,
      "num_tokens": 46042675.0,
      "step": 21015
    },
    {
      "entropy": 5.564108943939209,
      "epoch": 2.086145295752283,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045674099381502167,
      "loss": 5.0582,
      "mean_token_accuracy": 0.18051357865333556,
      "num_tokens": 46053616.0,
      "step": 21020
    },
    {
      "entropy": 5.627999925613404,
      "epoch": 2.0866415244144503,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004567201032079381,
      "loss": 5.0686,
      "mean_token_accuracy": 0.17567255944013596,
      "num_tokens": 46063755.0,
      "step": 21025
    },
    {
      "entropy": 5.64298734664917,
      "epoch": 2.0871377530766178,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045669920809458754,
      "loss": 5.0604,
      "mean_token_accuracy": 0.1815497025847435,
      "num_tokens": 46074459.0,
      "step": 21030
    },
    {
      "entropy": 5.704286050796509,
      "epoch": 2.0876339817387852,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004566783084754881,
      "loss": 5.1236,
      "mean_token_accuracy": 0.18145104199647905,
      "num_tokens": 46085053.0,
      "step": 21035
    },
    {
      "entropy": 5.596402549743653,
      "epoch": 2.0881302104009527,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045665740435115815,
      "loss": 4.9536,
      "mean_token_accuracy": 0.18984670490026473,
      "num_tokens": 46094697.0,
      "step": 21040
    },
    {
      "entropy": 5.6028660297393795,
      "epoch": 2.08862643906312,
      "grad_norm": 1.125,
      "learning_rate": 0.000456636495722116,
      "loss": 5.0848,
      "mean_token_accuracy": 0.17678985595703126,
      "num_tokens": 46105617.0,
      "step": 21045
    },
    {
      "entropy": 5.681221771240234,
      "epoch": 2.0891226677252877,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004566155825888801,
      "loss": 5.0372,
      "mean_token_accuracy": 0.18434203863143922,
      "num_tokens": 46115450.0,
      "step": 21050
    },
    {
      "entropy": 5.606844329833985,
      "epoch": 2.089618896387455,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004565946649519691,
      "loss": 5.1114,
      "mean_token_accuracy": 0.17776641249656677,
      "num_tokens": 46126511.0,
      "step": 21055
    },
    {
      "entropy": 5.581297206878662,
      "epoch": 2.090115125049623,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045657374281190177,
      "loss": 4.9985,
      "mean_token_accuracy": 0.18358195722103118,
      "num_tokens": 46136294.0,
      "step": 21060
    },
    {
      "entropy": 5.5726399421691895,
      "epoch": 2.0906113537117905,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004565528161691969,
      "loss": 5.0757,
      "mean_token_accuracy": 0.17171139866113663,
      "num_tokens": 46148926.0,
      "step": 21065
    },
    {
      "entropy": 5.63823504447937,
      "epoch": 2.091107582373958,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004565318850243733,
      "loss": 5.0799,
      "mean_token_accuracy": 0.18079462498426438,
      "num_tokens": 46160130.0,
      "step": 21070
    },
    {
      "entropy": 5.584488821029663,
      "epoch": 2.0916038110361255,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004565109493779502,
      "loss": 4.9942,
      "mean_token_accuracy": 0.18340159803628922,
      "num_tokens": 46169919.0,
      "step": 21075
    },
    {
      "entropy": 5.625015449523926,
      "epoch": 2.092100039698293,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00045649000923044664,
      "loss": 5.0681,
      "mean_token_accuracy": 0.18181092888116837,
      "num_tokens": 46181186.0,
      "step": 21080
    },
    {
      "entropy": 5.594567108154297,
      "epoch": 2.0925962683604604,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045646906458238196,
      "loss": 5.0761,
      "mean_token_accuracy": 0.1723933547735214,
      "num_tokens": 46192172.0,
      "step": 21085
    },
    {
      "entropy": 5.6209508895874025,
      "epoch": 2.093092497022628,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004564481154342754,
      "loss": 5.0515,
      "mean_token_accuracy": 0.1707015186548233,
      "num_tokens": 46202678.0,
      "step": 21090
    },
    {
      "entropy": 5.632715702056885,
      "epoch": 2.0935887256847954,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004564271617866467,
      "loss": 5.0492,
      "mean_token_accuracy": 0.1765544131398201,
      "num_tokens": 46212213.0,
      "step": 21095
    },
    {
      "entropy": 5.702187967300415,
      "epoch": 2.0940849543469633,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045640620364001524,
      "loss": 5.0876,
      "mean_token_accuracy": 0.17532884180545807,
      "num_tokens": 46223044.0,
      "step": 21100
    },
    {
      "entropy": 5.6903444766998295,
      "epoch": 2.0945811830091308,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045638524099490086,
      "loss": 5.0889,
      "mean_token_accuracy": 0.17716085761785508,
      "num_tokens": 46236008.0,
      "step": 21105
    },
    {
      "entropy": 5.591418027877808,
      "epoch": 2.0950774116712982,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045636427385182334,
      "loss": 5.0448,
      "mean_token_accuracy": 0.1795505553483963,
      "num_tokens": 46247631.0,
      "step": 21110
    },
    {
      "entropy": 5.600918436050415,
      "epoch": 2.0955736403334657,
      "grad_norm": 1.0,
      "learning_rate": 0.00045634330221130265,
      "loss": 5.0038,
      "mean_token_accuracy": 0.1830808624625206,
      "num_tokens": 46258294.0,
      "step": 21115
    },
    {
      "entropy": 5.619723892211914,
      "epoch": 2.096069868995633,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004563223260738589,
      "loss": 5.0166,
      "mean_token_accuracy": 0.17815846651792527,
      "num_tokens": 46269611.0,
      "step": 21120
    },
    {
      "entropy": 5.645808315277099,
      "epoch": 2.0965660976578007,
      "grad_norm": 1.03125,
      "learning_rate": 0.000456301345440012,
      "loss": 5.0945,
      "mean_token_accuracy": 0.18519142121076584,
      "num_tokens": 46280731.0,
      "step": 21125
    },
    {
      "entropy": 5.587784004211426,
      "epoch": 2.097062326319968,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004562803603102825,
      "loss": 5.0268,
      "mean_token_accuracy": 0.1803426444530487,
      "num_tokens": 46290226.0,
      "step": 21130
    },
    {
      "entropy": 5.60252742767334,
      "epoch": 2.0975585549821356,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045625937068519075,
      "loss": 5.0313,
      "mean_token_accuracy": 0.18522249013185502,
      "num_tokens": 46300388.0,
      "step": 21135
    },
    {
      "entropy": 5.602901124954224,
      "epoch": 2.098054783644303,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004562383765652571,
      "loss": 5.075,
      "mean_token_accuracy": 0.1795324370265007,
      "num_tokens": 46312214.0,
      "step": 21140
    },
    {
      "entropy": 5.629798460006714,
      "epoch": 2.098551012306471,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004562173779510023,
      "loss": 5.0891,
      "mean_token_accuracy": 0.17077882587909698,
      "num_tokens": 46322728.0,
      "step": 21145
    },
    {
      "entropy": 5.578860569000244,
      "epoch": 2.0990472409686385,
      "grad_norm": 1.0,
      "learning_rate": 0.000456196374842947,
      "loss": 5.0125,
      "mean_token_accuracy": 0.18184789270162582,
      "num_tokens": 46334169.0,
      "step": 21150
    },
    {
      "entropy": 5.639977169036865,
      "epoch": 2.099543469630806,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045617536724161203,
      "loss": 4.9998,
      "mean_token_accuracy": 0.17842114120721816,
      "num_tokens": 46344601.0,
      "step": 21155
    },
    {
      "entropy": 5.593939590454101,
      "epoch": 2.1000396982929734,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045615435514751834,
      "loss": 5.0804,
      "mean_token_accuracy": 0.17379492074251174,
      "num_tokens": 46355633.0,
      "step": 21160
    },
    {
      "entropy": 5.617863941192627,
      "epoch": 2.100535926955141,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000456133338561187,
      "loss": 5.1316,
      "mean_token_accuracy": 0.1669669568538666,
      "num_tokens": 46365981.0,
      "step": 21165
    },
    {
      "entropy": 5.631189680099487,
      "epoch": 2.1010321556173084,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004561123174831391,
      "loss": 5.0997,
      "mean_token_accuracy": 0.18243973106145858,
      "num_tokens": 46376664.0,
      "step": 21170
    },
    {
      "entropy": 5.57968373298645,
      "epoch": 2.101528384279476,
      "grad_norm": 1.125,
      "learning_rate": 0.0004560912919138962,
      "loss": 5.0058,
      "mean_token_accuracy": 0.18169824481010438,
      "num_tokens": 46387918.0,
      "step": 21175
    },
    {
      "entropy": 5.685410976409912,
      "epoch": 2.1020246129416433,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045607026185397924,
      "loss": 5.0424,
      "mean_token_accuracy": 0.18068176060914992,
      "num_tokens": 46398660.0,
      "step": 21180
    },
    {
      "entropy": 5.612930583953857,
      "epoch": 2.102520841603811,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004560492273039101,
      "loss": 4.9796,
      "mean_token_accuracy": 0.18664183765649794,
      "num_tokens": 46409365.0,
      "step": 21185
    },
    {
      "entropy": 5.587433862686157,
      "epoch": 2.1030170702659787,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045602818826421024,
      "loss": 5.0001,
      "mean_token_accuracy": 0.1856884628534317,
      "num_tokens": 46419457.0,
      "step": 21190
    },
    {
      "entropy": 5.591806173324585,
      "epoch": 2.103513298928146,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004560071447354013,
      "loss": 5.0653,
      "mean_token_accuracy": 0.1782245382666588,
      "num_tokens": 46430189.0,
      "step": 21195
    },
    {
      "entropy": 5.556305885314941,
      "epoch": 2.1040095275903137,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004559860967180053,
      "loss": 5.022,
      "mean_token_accuracy": 0.18152979165315627,
      "num_tokens": 46440555.0,
      "step": 21200
    },
    {
      "entropy": 5.647034788131714,
      "epoch": 2.104505756252481,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004559650442125441,
      "loss": 5.0645,
      "mean_token_accuracy": 0.17478541284799576,
      "num_tokens": 46451582.0,
      "step": 21205
    },
    {
      "entropy": 5.669967985153198,
      "epoch": 2.1050019849146486,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004559439872195397,
      "loss": 5.0691,
      "mean_token_accuracy": 0.17459966093301774,
      "num_tokens": 46463614.0,
      "step": 21210
    },
    {
      "entropy": 5.663447380065918,
      "epoch": 2.105498213576816,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004559229257395144,
      "loss": 5.0877,
      "mean_token_accuracy": 0.1760620042681694,
      "num_tokens": 46474279.0,
      "step": 21215
    },
    {
      "entropy": 5.6717771053314205,
      "epoch": 2.1059944422389836,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045590185977299025,
      "loss": 5.1744,
      "mean_token_accuracy": 0.17289144843816756,
      "num_tokens": 46485531.0,
      "step": 21220
    },
    {
      "entropy": 5.572240972518921,
      "epoch": 2.106490670901151,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004558807893204899,
      "loss": 4.9905,
      "mean_token_accuracy": 0.19041580259799956,
      "num_tokens": 46496197.0,
      "step": 21225
    },
    {
      "entropy": 5.653758430480957,
      "epoch": 2.106986899563319,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004558597143825357,
      "loss": 5.0618,
      "mean_token_accuracy": 0.17553364783525466,
      "num_tokens": 46506759.0,
      "step": 21230
    },
    {
      "entropy": 5.596378564834595,
      "epoch": 2.1074831282254864,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004558386349596503,
      "loss": 5.06,
      "mean_token_accuracy": 0.1802683651447296,
      "num_tokens": 46518956.0,
      "step": 21235
    },
    {
      "entropy": 5.691020536422729,
      "epoch": 2.107979356887654,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004558175510523564,
      "loss": 5.1377,
      "mean_token_accuracy": 0.17486658245325087,
      "num_tokens": 46530109.0,
      "step": 21240
    },
    {
      "entropy": 5.677356958389282,
      "epoch": 2.1084755855498214,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045579646266117696,
      "loss": 4.9894,
      "mean_token_accuracy": 0.1841521069407463,
      "num_tokens": 46542206.0,
      "step": 21245
    },
    {
      "entropy": 5.644448089599609,
      "epoch": 2.108971814211989,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045577536978663475,
      "loss": 5.1587,
      "mean_token_accuracy": 0.17815393656492234,
      "num_tokens": 46554143.0,
      "step": 21250
    },
    {
      "entropy": 5.54333233833313,
      "epoch": 2.1094680428741563,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004557542724292529,
      "loss": 4.9999,
      "mean_token_accuracy": 0.17704057842493057,
      "num_tokens": 46565697.0,
      "step": 21255
    },
    {
      "entropy": 5.657701110839843,
      "epoch": 2.109964271536324,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004557331705895545,
      "loss": 5.0363,
      "mean_token_accuracy": 0.18079574406147003,
      "num_tokens": 46576358.0,
      "step": 21260
    },
    {
      "entropy": 5.6626087665557865,
      "epoch": 2.1104605001984913,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004557120642680631,
      "loss": 5.0796,
      "mean_token_accuracy": 0.173554165661335,
      "num_tokens": 46589422.0,
      "step": 21265
    },
    {
      "entropy": 5.586868000030518,
      "epoch": 2.1109567288606588,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004556909534653018,
      "loss": 5.0519,
      "mean_token_accuracy": 0.1853708878159523,
      "num_tokens": 46599623.0,
      "step": 21270
    },
    {
      "entropy": 5.624314975738526,
      "epoch": 2.1114529575228267,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004556698381817942,
      "loss": 4.9909,
      "mean_token_accuracy": 0.18805098235607148,
      "num_tokens": 46609321.0,
      "step": 21275
    },
    {
      "entropy": 5.588265323638916,
      "epoch": 2.111949186184994,
      "grad_norm": 1.015625,
      "learning_rate": 0.000455648718418064,
      "loss": 5.0685,
      "mean_token_accuracy": 0.17580961883068086,
      "num_tokens": 46620265.0,
      "step": 21280
    },
    {
      "entropy": 5.649675464630127,
      "epoch": 2.1124454148471616,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004556275941746348,
      "loss": 5.0803,
      "mean_token_accuracy": 0.18262410610914231,
      "num_tokens": 46630776.0,
      "step": 21285
    },
    {
      "entropy": 5.55204963684082,
      "epoch": 2.112941643509329,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00045560646545203044,
      "loss": 4.9727,
      "mean_token_accuracy": 0.18477479219436646,
      "num_tokens": 46641382.0,
      "step": 21290
    },
    {
      "entropy": 5.532749319076538,
      "epoch": 2.1134378721714966,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00045558533225077496,
      "loss": 4.9785,
      "mean_token_accuracy": 0.1835481896996498,
      "num_tokens": 46653589.0,
      "step": 21295
    },
    {
      "entropy": 5.565368175506592,
      "epoch": 2.113934100833664,
      "grad_norm": 0.921875,
      "learning_rate": 0.00045556419457139235,
      "loss": 5.0098,
      "mean_token_accuracy": 0.18249086886644364,
      "num_tokens": 46664595.0,
      "step": 21300
    },
    {
      "entropy": 5.628429651260376,
      "epoch": 2.1144303294958315,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045554305241440684,
      "loss": 5.045,
      "mean_token_accuracy": 0.17661100178956984,
      "num_tokens": 46675453.0,
      "step": 21305
    },
    {
      "entropy": 5.549682474136352,
      "epoch": 2.114926558157999,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004555219057803426,
      "loss": 4.9492,
      "mean_token_accuracy": 0.18544286489486694,
      "num_tokens": 46685668.0,
      "step": 21310
    },
    {
      "entropy": 5.613961601257325,
      "epoch": 2.115422786820167,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004555007546697242,
      "loss": 5.0625,
      "mean_token_accuracy": 0.17767696380615233,
      "num_tokens": 46696144.0,
      "step": 21315
    },
    {
      "entropy": 5.642581033706665,
      "epoch": 2.1159190154823344,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045547959908307585,
      "loss": 5.0359,
      "mean_token_accuracy": 0.17636382728815078,
      "num_tokens": 46707892.0,
      "step": 21320
    },
    {
      "entropy": 5.645377159118652,
      "epoch": 2.116415244144502,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004554584390209225,
      "loss": 5.0653,
      "mean_token_accuracy": 0.17896926701068877,
      "num_tokens": 46721099.0,
      "step": 21325
    },
    {
      "entropy": 5.639668798446655,
      "epoch": 2.1169114728066694,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004554372744837886,
      "loss": 5.0945,
      "mean_token_accuracy": 0.17670868784189225,
      "num_tokens": 46732974.0,
      "step": 21330
    },
    {
      "entropy": 5.613441371917725,
      "epoch": 2.117407701468837,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004554161054721992,
      "loss": 5.1169,
      "mean_token_accuracy": 0.17755745500326156,
      "num_tokens": 46743148.0,
      "step": 21335
    },
    {
      "entropy": 5.618980407714844,
      "epoch": 2.1179039301310043,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004553949319866791,
      "loss": 5.0175,
      "mean_token_accuracy": 0.18261358737945557,
      "num_tokens": 46754214.0,
      "step": 21340
    },
    {
      "entropy": 5.6342401027679445,
      "epoch": 2.118400158793172,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00045537375402775334,
      "loss": 5.0321,
      "mean_token_accuracy": 0.17996201664209366,
      "num_tokens": 46764112.0,
      "step": 21345
    },
    {
      "entropy": 5.57990984916687,
      "epoch": 2.1188963874553393,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00045535257159594724,
      "loss": 5.0033,
      "mean_token_accuracy": 0.18520431369543075,
      "num_tokens": 46775202.0,
      "step": 21350
    },
    {
      "entropy": 5.689061164855957,
      "epoch": 2.1193926161175067,
      "grad_norm": 1.078125,
      "learning_rate": 0.000455331384691786,
      "loss": 5.0424,
      "mean_token_accuracy": 0.18202155232429504,
      "num_tokens": 46786557.0,
      "step": 21355
    },
    {
      "entropy": 5.588870191574097,
      "epoch": 2.1198888447796747,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004553101933157949,
      "loss": 5.02,
      "mean_token_accuracy": 0.17553119212388993,
      "num_tokens": 46797082.0,
      "step": 21360
    },
    {
      "entropy": 5.597496271133423,
      "epoch": 2.120385073441842,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045528899746849965,
      "loss": 5.0457,
      "mean_token_accuracy": 0.18675297796726226,
      "num_tokens": 46806679.0,
      "step": 21365
    },
    {
      "entropy": 5.7329161167144775,
      "epoch": 2.1208813021040096,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004552677971504257,
      "loss": 5.1735,
      "mean_token_accuracy": 0.16655195206403733,
      "num_tokens": 46818169.0,
      "step": 21370
    },
    {
      "entropy": 5.771428918838501,
      "epoch": 2.121377530766177,
      "grad_norm": 1.625,
      "learning_rate": 0.0004552465923620988,
      "loss": 5.2248,
      "mean_token_accuracy": 0.17266677618026732,
      "num_tokens": 46829552.0,
      "step": 21375
    },
    {
      "entropy": 5.687802457809449,
      "epoch": 2.1218737594283446,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004552253831040448,
      "loss": 5.1289,
      "mean_token_accuracy": 0.1782443568110466,
      "num_tokens": 46840808.0,
      "step": 21380
    },
    {
      "entropy": 5.684790706634521,
      "epoch": 2.122369988090512,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045520416937678973,
      "loss": 5.0572,
      "mean_token_accuracy": 0.1729547619819641,
      "num_tokens": 46852066.0,
      "step": 21385
    },
    {
      "entropy": 5.610827922821045,
      "epoch": 2.1228662167526795,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045518295118085955,
      "loss": 4.9937,
      "mean_token_accuracy": 0.178229820728302,
      "num_tokens": 46863163.0,
      "step": 21390
    },
    {
      "entropy": 5.641240501403809,
      "epoch": 2.123362445414847,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045516172851678036,
      "loss": 5.0454,
      "mean_token_accuracy": 0.17499405592679979,
      "num_tokens": 46873246.0,
      "step": 21395
    },
    {
      "entropy": 5.602969741821289,
      "epoch": 2.123858674077015,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045514050138507857,
      "loss": 5.0625,
      "mean_token_accuracy": 0.17825856506824495,
      "num_tokens": 46883881.0,
      "step": 21400
    },
    {
      "entropy": 5.642919588088989,
      "epoch": 2.1243549027391824,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004551192697862805,
      "loss": 5.1002,
      "mean_token_accuracy": 0.17308273464441298,
      "num_tokens": 46894907.0,
      "step": 21405
    },
    {
      "entropy": 5.587165260314942,
      "epoch": 2.12485113140135,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004550980337209127,
      "loss": 5.0117,
      "mean_token_accuracy": 0.17711444199085236,
      "num_tokens": 46905220.0,
      "step": 21410
    },
    {
      "entropy": 5.584099197387696,
      "epoch": 2.1253473600635173,
      "grad_norm": 1.34375,
      "learning_rate": 0.00045507679318950166,
      "loss": 5.0137,
      "mean_token_accuracy": 0.18599180281162261,
      "num_tokens": 46916328.0,
      "step": 21415
    },
    {
      "entropy": 5.822368001937866,
      "epoch": 2.125843588725685,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004550555481925743,
      "loss": 5.224,
      "mean_token_accuracy": 0.16209215968847274,
      "num_tokens": 46927450.0,
      "step": 21420
    },
    {
      "entropy": 5.568746042251587,
      "epoch": 2.1263398173878523,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045503429873065723,
      "loss": 4.9614,
      "mean_token_accuracy": 0.18638054877519608,
      "num_tokens": 46936758.0,
      "step": 21425
    },
    {
      "entropy": 5.516079521179199,
      "epoch": 2.1268360460500197,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004550130448042775,
      "loss": 4.9113,
      "mean_token_accuracy": 0.18856871128082275,
      "num_tokens": 46945776.0,
      "step": 21430
    },
    {
      "entropy": 5.554952907562256,
      "epoch": 2.127332274712187,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004549917864139622,
      "loss": 5.0405,
      "mean_token_accuracy": 0.19002811312675477,
      "num_tokens": 46955621.0,
      "step": 21435
    },
    {
      "entropy": 5.6883362293243405,
      "epoch": 2.1278285033743547,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004549705235602384,
      "loss": 5.1371,
      "mean_token_accuracy": 0.18082139641046524,
      "num_tokens": 46966890.0,
      "step": 21440
    },
    {
      "entropy": 5.6678813934326175,
      "epoch": 2.1283247320365226,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004549492562436334,
      "loss": 5.1358,
      "mean_token_accuracy": 0.17637863010168076,
      "num_tokens": 46978380.0,
      "step": 21445
    },
    {
      "entropy": 5.61096453666687,
      "epoch": 2.12882096069869,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045492798446467465,
      "loss": 4.9632,
      "mean_token_accuracy": 0.18350712805986405,
      "num_tokens": 46988286.0,
      "step": 21450
    },
    {
      "entropy": 5.573999261856079,
      "epoch": 2.1293171893608576,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045490670822388957,
      "loss": 4.9776,
      "mean_token_accuracy": 0.18110429644584655,
      "num_tokens": 46997795.0,
      "step": 21455
    },
    {
      "entropy": 5.519832038879395,
      "epoch": 2.129813418023025,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004548854275218059,
      "loss": 4.8784,
      "mean_token_accuracy": 0.19037747979164124,
      "num_tokens": 47007970.0,
      "step": 21460
    },
    {
      "entropy": 5.592421913146973,
      "epoch": 2.1303096466851925,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045486414235895115,
      "loss": 5.0146,
      "mean_token_accuracy": 0.17887501418590546,
      "num_tokens": 47019548.0,
      "step": 21465
    },
    {
      "entropy": 5.630730390548706,
      "epoch": 2.13080587534736,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004548428527358532,
      "loss": 5.1109,
      "mean_token_accuracy": 0.17672485560178758,
      "num_tokens": 47029568.0,
      "step": 21470
    },
    {
      "entropy": 5.675691556930542,
      "epoch": 2.1313021040095275,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004548215586530401,
      "loss": 5.117,
      "mean_token_accuracy": 0.17263758629560472,
      "num_tokens": 47042048.0,
      "step": 21475
    },
    {
      "entropy": 5.623611688613892,
      "epoch": 2.131798332671695,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045480026011103975,
      "loss": 4.9854,
      "mean_token_accuracy": 0.17640349864959717,
      "num_tokens": 47052099.0,
      "step": 21480
    },
    {
      "entropy": 5.629216575622559,
      "epoch": 2.132294561333863,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00045477895711038045,
      "loss": 5.0171,
      "mean_token_accuracy": 0.18179487586021423,
      "num_tokens": 47063310.0,
      "step": 21485
    },
    {
      "entropy": 5.616369390487671,
      "epoch": 2.1327907899960303,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004547576496515904,
      "loss": 5.0352,
      "mean_token_accuracy": 0.17333618700504302,
      "num_tokens": 47073545.0,
      "step": 21490
    },
    {
      "entropy": 5.597084045410156,
      "epoch": 2.133287018658198,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00045473633773519795,
      "loss": 5.079,
      "mean_token_accuracy": 0.18393848240375518,
      "num_tokens": 47085456.0,
      "step": 21495
    },
    {
      "entropy": 5.626670980453492,
      "epoch": 2.1337832473203653,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045471502136173173,
      "loss": 5.0596,
      "mean_token_accuracy": 0.1826125591993332,
      "num_tokens": 47096590.0,
      "step": 21500
    },
    {
      "entropy": 5.645954036712647,
      "epoch": 2.1342794759825328,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004546937005317201,
      "loss": 5.0002,
      "mean_token_accuracy": 0.18948154896497726,
      "num_tokens": 47107876.0,
      "step": 21505
    },
    {
      "entropy": 5.629414987564087,
      "epoch": 2.1347757046447002,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045467237524569193,
      "loss": 5.0302,
      "mean_token_accuracy": 0.183303339779377,
      "num_tokens": 47117662.0,
      "step": 21510
    },
    {
      "entropy": 5.6060834407806395,
      "epoch": 2.1352719333068677,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045465104550417604,
      "loss": 5.0656,
      "mean_token_accuracy": 0.17499312460422517,
      "num_tokens": 47129283.0,
      "step": 21515
    },
    {
      "entropy": 5.701242399215698,
      "epoch": 2.135768161969035,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045462971130770123,
      "loss": 5.1477,
      "mean_token_accuracy": 0.17415452152490615,
      "num_tokens": 47140824.0,
      "step": 21520
    },
    {
      "entropy": 5.62852144241333,
      "epoch": 2.1362643906312027,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004546083726567968,
      "loss": 4.9946,
      "mean_token_accuracy": 0.18879739195108414,
      "num_tokens": 47151733.0,
      "step": 21525
    },
    {
      "entropy": 5.6159289360046385,
      "epoch": 2.1367606192933706,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004545870295519917,
      "loss": 5.1288,
      "mean_token_accuracy": 0.16810811161994935,
      "num_tokens": 47162846.0,
      "step": 21530
    },
    {
      "entropy": 5.667001008987427,
      "epoch": 2.137256847955538,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045456568199381524,
      "loss": 5.0131,
      "mean_token_accuracy": 0.1806864395737648,
      "num_tokens": 47174969.0,
      "step": 21535
    },
    {
      "entropy": 5.731106233596802,
      "epoch": 2.1377530766177055,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045454432998279676,
      "loss": 5.1008,
      "mean_token_accuracy": 0.17472228407859802,
      "num_tokens": 47186903.0,
      "step": 21540
    },
    {
      "entropy": 5.652883577346802,
      "epoch": 2.138249305279873,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004545229735194658,
      "loss": 5.0601,
      "mean_token_accuracy": 0.1828927531838417,
      "num_tokens": 47197716.0,
      "step": 21545
    },
    {
      "entropy": 5.593692874908447,
      "epoch": 2.1387455339420405,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045450161260435196,
      "loss": 5.0081,
      "mean_token_accuracy": 0.185208660364151,
      "num_tokens": 47208623.0,
      "step": 21550
    },
    {
      "entropy": 5.550904750823975,
      "epoch": 2.139241762604208,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004544802472379849,
      "loss": 5.0301,
      "mean_token_accuracy": 0.18052830696105956,
      "num_tokens": 47218248.0,
      "step": 21555
    },
    {
      "entropy": 5.635174989700317,
      "epoch": 2.1397379912663754,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004544588774208945,
      "loss": 5.1425,
      "mean_token_accuracy": 0.1761747658252716,
      "num_tokens": 47232028.0,
      "step": 21560
    },
    {
      "entropy": 5.683732223510742,
      "epoch": 2.140234219928543,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045443750315361055,
      "loss": 5.0155,
      "mean_token_accuracy": 0.1868390291929245,
      "num_tokens": 47241976.0,
      "step": 21565
    },
    {
      "entropy": 5.634995985031128,
      "epoch": 2.140730448590711,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004544161244366632,
      "loss": 5.0216,
      "mean_token_accuracy": 0.17955178320407866,
      "num_tokens": 47253228.0,
      "step": 21570
    },
    {
      "entropy": 5.613738107681274,
      "epoch": 2.1412266772528783,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004543947412705826,
      "loss": 5.0258,
      "mean_token_accuracy": 0.17822352051734924,
      "num_tokens": 47265210.0,
      "step": 21575
    },
    {
      "entropy": 5.578924131393433,
      "epoch": 2.1417229059150458,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045437335365589905,
      "loss": 5.047,
      "mean_token_accuracy": 0.17988192588090895,
      "num_tokens": 47275723.0,
      "step": 21580
    },
    {
      "entropy": 5.647161436080933,
      "epoch": 2.1422191345772132,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045435196159314276,
      "loss": 5.0814,
      "mean_token_accuracy": 0.1767873540520668,
      "num_tokens": 47286469.0,
      "step": 21585
    },
    {
      "entropy": 5.690964698791504,
      "epoch": 2.1427153632393807,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045433056508284425,
      "loss": 5.1064,
      "mean_token_accuracy": 0.17553098648786544,
      "num_tokens": 47296461.0,
      "step": 21590
    },
    {
      "entropy": 5.638466310501099,
      "epoch": 2.143211591901548,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004543091641255342,
      "loss": 5.01,
      "mean_token_accuracy": 0.1841052308678627,
      "num_tokens": 47307217.0,
      "step": 21595
    },
    {
      "entropy": 5.653512763977051,
      "epoch": 2.1437078205637157,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004542877587217432,
      "loss": 5.1377,
      "mean_token_accuracy": 0.17444525957107543,
      "num_tokens": 47317041.0,
      "step": 21600
    },
    {
      "entropy": 5.608870792388916,
      "epoch": 2.144204049225883,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045426634887200224,
      "loss": 5.0594,
      "mean_token_accuracy": 0.17794932872056962,
      "num_tokens": 47328185.0,
      "step": 21605
    },
    {
      "entropy": 5.636754560470581,
      "epoch": 2.1447002778880506,
      "grad_norm": 0.9375,
      "learning_rate": 0.000454244934576842,
      "loss": 5.0939,
      "mean_token_accuracy": 0.1715226173400879,
      "num_tokens": 47339984.0,
      "step": 21610
    },
    {
      "entropy": 5.657262563705444,
      "epoch": 2.1451965065502185,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045422351583679365,
      "loss": 5.0502,
      "mean_token_accuracy": 0.18517396599054337,
      "num_tokens": 47350911.0,
      "step": 21615
    },
    {
      "entropy": 5.671198225021362,
      "epoch": 2.145692735212386,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004542020926523883,
      "loss": 5.0404,
      "mean_token_accuracy": 0.17303962409496307,
      "num_tokens": 47361703.0,
      "step": 21620
    },
    {
      "entropy": 5.622785425186157,
      "epoch": 2.1461889638745535,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004541806650241572,
      "loss": 5.0187,
      "mean_token_accuracy": 0.17401060611009597,
      "num_tokens": 47372878.0,
      "step": 21625
    },
    {
      "entropy": 5.610295343399048,
      "epoch": 2.146685192536721,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004541592329526317,
      "loss": 5.0366,
      "mean_token_accuracy": 0.17514857649803162,
      "num_tokens": 47383056.0,
      "step": 21630
    },
    {
      "entropy": 5.5960956573486325,
      "epoch": 2.1471814211988884,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004541377964383434,
      "loss": 5.0651,
      "mean_token_accuracy": 0.18341880142688752,
      "num_tokens": 47394160.0,
      "step": 21635
    },
    {
      "entropy": 5.658431529998779,
      "epoch": 2.147677649861056,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004541163554818237,
      "loss": 5.0216,
      "mean_token_accuracy": 0.17833720445632933,
      "num_tokens": 47404861.0,
      "step": 21640
    },
    {
      "entropy": 5.668709182739258,
      "epoch": 2.1481738785232234,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045409491008360423,
      "loss": 5.0839,
      "mean_token_accuracy": 0.18184928745031356,
      "num_tokens": 47415236.0,
      "step": 21645
    },
    {
      "entropy": 5.64151291847229,
      "epoch": 2.148670107185391,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004540734602442169,
      "loss": 5.0487,
      "mean_token_accuracy": 0.17825921773910522,
      "num_tokens": 47425789.0,
      "step": 21650
    },
    {
      "entropy": 5.6120617389678955,
      "epoch": 2.149166335847559,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004540520059641937,
      "loss": 4.9228,
      "mean_token_accuracy": 0.18698995262384416,
      "num_tokens": 47436236.0,
      "step": 21655
    },
    {
      "entropy": 5.560188961029053,
      "epoch": 2.1496625645097263,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004540305472440666,
      "loss": 4.9398,
      "mean_token_accuracy": 0.18456323742866515,
      "num_tokens": 47447041.0,
      "step": 21660
    },
    {
      "entropy": 5.603955221176148,
      "epoch": 2.1501587931718937,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004540090840843677,
      "loss": 5.0933,
      "mean_token_accuracy": 0.17158089578151703,
      "num_tokens": 47458001.0,
      "step": 21665
    },
    {
      "entropy": 5.598781967163086,
      "epoch": 2.150655021834061,
      "grad_norm": 1.0,
      "learning_rate": 0.0004539876164856292,
      "loss": 4.959,
      "mean_token_accuracy": 0.18714761584997178,
      "num_tokens": 47468621.0,
      "step": 21670
    },
    {
      "entropy": 5.597853708267212,
      "epoch": 2.1511512504962287,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004539661444483835,
      "loss": 5.0805,
      "mean_token_accuracy": 0.18342088162899017,
      "num_tokens": 47478957.0,
      "step": 21675
    },
    {
      "entropy": 5.574245166778565,
      "epoch": 2.151647479158396,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004539446679731631,
      "loss": 5.0248,
      "mean_token_accuracy": 0.18172956854104996,
      "num_tokens": 47489640.0,
      "step": 21680
    },
    {
      "entropy": 5.670043659210205,
      "epoch": 2.1521437078205636,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004539231870605005,
      "loss": 5.1528,
      "mean_token_accuracy": 0.1741962417960167,
      "num_tokens": 47501581.0,
      "step": 21685
    },
    {
      "entropy": 5.6428345203399655,
      "epoch": 2.152639936482731,
      "grad_norm": 0.921875,
      "learning_rate": 0.00045390170171092845,
      "loss": 5.0761,
      "mean_token_accuracy": 0.17592136710882186,
      "num_tokens": 47513605.0,
      "step": 21690
    },
    {
      "entropy": 5.683101940155029,
      "epoch": 2.1531361651448986,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045388021192497956,
      "loss": 5.0816,
      "mean_token_accuracy": 0.17700353264808655,
      "num_tokens": 47523784.0,
      "step": 21695
    },
    {
      "entropy": 5.5681194305419925,
      "epoch": 2.1536323938070665,
      "grad_norm": 1.0,
      "learning_rate": 0.000453858717703187,
      "loss": 4.9777,
      "mean_token_accuracy": 0.17698473185300828,
      "num_tokens": 47534643.0,
      "step": 21700
    },
    {
      "entropy": 5.600233364105224,
      "epoch": 2.154128622469234,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004538372190460836,
      "loss": 5.0974,
      "mean_token_accuracy": 0.17443242073059081,
      "num_tokens": 47545920.0,
      "step": 21705
    },
    {
      "entropy": 5.590089654922485,
      "epoch": 2.1546248511314015,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004538157159542025,
      "loss": 5.027,
      "mean_token_accuracy": 0.17523930519819259,
      "num_tokens": 47557024.0,
      "step": 21710
    },
    {
      "entropy": 5.629210424423218,
      "epoch": 2.155121079793569,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004537942084280771,
      "loss": 4.9972,
      "mean_token_accuracy": 0.18280754685401918,
      "num_tokens": 47567531.0,
      "step": 21715
    },
    {
      "entropy": 5.606799364089966,
      "epoch": 2.1556173084557364,
      "grad_norm": 0.921875,
      "learning_rate": 0.00045377269646824036,
      "loss": 5.1203,
      "mean_token_accuracy": 0.17742523103952407,
      "num_tokens": 47578642.0,
      "step": 21720
    },
    {
      "entropy": 5.6192946434021,
      "epoch": 2.156113537117904,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045375118007522607,
      "loss": 5.1764,
      "mean_token_accuracy": 0.17301681339740754,
      "num_tokens": 47589346.0,
      "step": 21725
    },
    {
      "entropy": 5.588977718353272,
      "epoch": 2.1566097657800714,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004537296592495678,
      "loss": 4.9925,
      "mean_token_accuracy": 0.1885127380490303,
      "num_tokens": 47599689.0,
      "step": 21730
    },
    {
      "entropy": 5.704806470870972,
      "epoch": 2.157105994442239,
      "grad_norm": 1.109375,
      "learning_rate": 0.000453708133991799,
      "loss": 5.0854,
      "mean_token_accuracy": 0.17406787127256393,
      "num_tokens": 47609214.0,
      "step": 21735
    },
    {
      "entropy": 5.626936054229736,
      "epoch": 2.1576022231044063,
      "grad_norm": 1.0,
      "learning_rate": 0.00045368660430245364,
      "loss": 5.017,
      "mean_token_accuracy": 0.18074029833078384,
      "num_tokens": 47620119.0,
      "step": 21740
    },
    {
      "entropy": 5.565015888214111,
      "epoch": 2.158098451766574,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004536650701820654,
      "loss": 5.0003,
      "mean_token_accuracy": 0.18356887698173524,
      "num_tokens": 47630085.0,
      "step": 21745
    },
    {
      "entropy": 5.586576223373413,
      "epoch": 2.1585946804287417,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004536435316311685,
      "loss": 4.9841,
      "mean_token_accuracy": 0.18409768640995025,
      "num_tokens": 47641457.0,
      "step": 21750
    },
    {
      "entropy": 5.636755847930909,
      "epoch": 2.159090909090909,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045362198865029696,
      "loss": 5.0061,
      "mean_token_accuracy": 0.1770160511136055,
      "num_tokens": 47652269.0,
      "step": 21755
    },
    {
      "entropy": 5.612203454971313,
      "epoch": 2.1595871377530766,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045360044123998495,
      "loss": 5.0388,
      "mean_token_accuracy": 0.18041882663965225,
      "num_tokens": 47664601.0,
      "step": 21760
    },
    {
      "entropy": 5.676976060867309,
      "epoch": 2.160083366415244,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045357888940076685,
      "loss": 5.1611,
      "mean_token_accuracy": 0.17391173988580705,
      "num_tokens": 47676187.0,
      "step": 21765
    },
    {
      "entropy": 5.557441473007202,
      "epoch": 2.1605795950774116,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004535573331331771,
      "loss": 4.9514,
      "mean_token_accuracy": 0.1853115439414978,
      "num_tokens": 47687057.0,
      "step": 21770
    },
    {
      "entropy": 5.62265944480896,
      "epoch": 2.161075823739579,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00045353577243775035,
      "loss": 5.0073,
      "mean_token_accuracy": 0.18049339801073075,
      "num_tokens": 47699074.0,
      "step": 21775
    },
    {
      "entropy": 5.624121046066284,
      "epoch": 2.1615720524017465,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004535142073150211,
      "loss": 4.994,
      "mean_token_accuracy": 0.18511240184307098,
      "num_tokens": 47710400.0,
      "step": 21780
    },
    {
      "entropy": 5.570060062408447,
      "epoch": 2.162068281063914,
      "grad_norm": 1.125,
      "learning_rate": 0.0004534926377655241,
      "loss": 4.8882,
      "mean_token_accuracy": 0.1965904042124748,
      "num_tokens": 47720900.0,
      "step": 21785
    },
    {
      "entropy": 5.5662394046783445,
      "epoch": 2.162564509726082,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045347106378979435,
      "loss": 4.9081,
      "mean_token_accuracy": 0.18683840781450273,
      "num_tokens": 47730702.0,
      "step": 21790
    },
    {
      "entropy": 5.6072712421417235,
      "epoch": 2.1630607383882494,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045344948538836674,
      "loss": 5.0,
      "mean_token_accuracy": 0.18123828917741774,
      "num_tokens": 47740451.0,
      "step": 21795
    },
    {
      "entropy": 5.60313229560852,
      "epoch": 2.163556967050417,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045342790256177646,
      "loss": 5.0613,
      "mean_token_accuracy": 0.1794189006090164,
      "num_tokens": 47752285.0,
      "step": 21800
    },
    {
      "entropy": 5.647682619094849,
      "epoch": 2.1640531957125844,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004534063153105587,
      "loss": 5.0642,
      "mean_token_accuracy": 0.17983706444501876,
      "num_tokens": 47761895.0,
      "step": 21805
    },
    {
      "entropy": 5.602112436294556,
      "epoch": 2.164549424374752,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004533847236352487,
      "loss": 4.9579,
      "mean_token_accuracy": 0.18862951993942262,
      "num_tokens": 47771718.0,
      "step": 21810
    },
    {
      "entropy": 5.543236827850341,
      "epoch": 2.1650456530369193,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000453363127536382,
      "loss": 5.0059,
      "mean_token_accuracy": 0.18348332047462462,
      "num_tokens": 47782183.0,
      "step": 21815
    },
    {
      "entropy": 5.557391357421875,
      "epoch": 2.165541881699087,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045334152701449407,
      "loss": 4.9492,
      "mean_token_accuracy": 0.18478961735963823,
      "num_tokens": 47792190.0,
      "step": 21820
    },
    {
      "entropy": 5.724546957015991,
      "epoch": 2.1660381103612543,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004533199220701205,
      "loss": 5.1664,
      "mean_token_accuracy": 0.17286243587732314,
      "num_tokens": 47803245.0,
      "step": 21825
    },
    {
      "entropy": 5.658423376083374,
      "epoch": 2.166534339023422,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045329831270379714,
      "loss": 5.108,
      "mean_token_accuracy": 0.17177244275808334,
      "num_tokens": 47814605.0,
      "step": 21830
    },
    {
      "entropy": 5.602911138534546,
      "epoch": 2.1670305676855897,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045327669891605975,
      "loss": 5.0678,
      "mean_token_accuracy": 0.17881386131048202,
      "num_tokens": 47825945.0,
      "step": 21835
    },
    {
      "entropy": 5.605758810043335,
      "epoch": 2.167526796347757,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004532550807074445,
      "loss": 5.0625,
      "mean_token_accuracy": 0.17426800727844238,
      "num_tokens": 47838119.0,
      "step": 21840
    },
    {
      "entropy": 5.629299926757812,
      "epoch": 2.1680230250099246,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004532334580784873,
      "loss": 5.067,
      "mean_token_accuracy": 0.17622477263212205,
      "num_tokens": 47849560.0,
      "step": 21845
    },
    {
      "entropy": 5.680968952178955,
      "epoch": 2.168519253672092,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004532118310297244,
      "loss": 5.0737,
      "mean_token_accuracy": 0.17056383341550826,
      "num_tokens": 47860036.0,
      "step": 21850
    },
    {
      "entropy": 5.58285961151123,
      "epoch": 2.1690154823342596,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004531901995616921,
      "loss": 4.9521,
      "mean_token_accuracy": 0.18138885349035264,
      "num_tokens": 47870826.0,
      "step": 21855
    },
    {
      "entropy": 5.5925861358642575,
      "epoch": 2.169511710996427,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004531685636749268,
      "loss": 5.0768,
      "mean_token_accuracy": 0.17293297946453096,
      "num_tokens": 47881415.0,
      "step": 21860
    },
    {
      "entropy": 5.6728369235992435,
      "epoch": 2.1700079396585945,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004531469233699651,
      "loss": 5.0786,
      "mean_token_accuracy": 0.17763193547725678,
      "num_tokens": 47893343.0,
      "step": 21865
    },
    {
      "entropy": 5.604623794555664,
      "epoch": 2.170504168320762,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004531252786473436,
      "loss": 5.0563,
      "mean_token_accuracy": 0.17871999591588975,
      "num_tokens": 47904871.0,
      "step": 21870
    },
    {
      "entropy": 5.617084169387818,
      "epoch": 2.17100039698293,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004531036295075989,
      "loss": 4.9823,
      "mean_token_accuracy": 0.17908366918563842,
      "num_tokens": 47915367.0,
      "step": 21875
    },
    {
      "entropy": 5.623944664001465,
      "epoch": 2.1714966256450974,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045308197595126797,
      "loss": 4.9978,
      "mean_token_accuracy": 0.17927134782075882,
      "num_tokens": 47925987.0,
      "step": 21880
    },
    {
      "entropy": 5.636763381958008,
      "epoch": 2.171992854307265,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004530603179788879,
      "loss": 5.0859,
      "mean_token_accuracy": 0.17933705002069472,
      "num_tokens": 47937688.0,
      "step": 21885
    },
    {
      "entropy": 5.601847171783447,
      "epoch": 2.1724890829694323,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004530386555909955,
      "loss": 5.0361,
      "mean_token_accuracy": 0.1875278100371361,
      "num_tokens": 47948550.0,
      "step": 21890
    },
    {
      "entropy": 5.613557004928589,
      "epoch": 2.1729853116316,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045301698878812813,
      "loss": 5.0328,
      "mean_token_accuracy": 0.18579289466142654,
      "num_tokens": 47957982.0,
      "step": 21895
    },
    {
      "entropy": 5.656593561172485,
      "epoch": 2.1734815402937673,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004529953175708231,
      "loss": 5.09,
      "mean_token_accuracy": 0.17794362604618072,
      "num_tokens": 47969562.0,
      "step": 21900
    },
    {
      "entropy": 5.627099704742432,
      "epoch": 2.1739777689559348,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045297364193961763,
      "loss": 5.0916,
      "mean_token_accuracy": 0.17278537899255753,
      "num_tokens": 47981323.0,
      "step": 21905
    },
    {
      "entropy": 5.620873594284058,
      "epoch": 2.1744739976181022,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045295196189504934,
      "loss": 5.0225,
      "mean_token_accuracy": 0.17571163922548294,
      "num_tokens": 47991454.0,
      "step": 21910
    },
    {
      "entropy": 5.5814167022705075,
      "epoch": 2.17497022628027,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045293027743765595,
      "loss": 4.9072,
      "mean_token_accuracy": 0.18713259547948838,
      "num_tokens": 48001300.0,
      "step": 21915
    },
    {
      "entropy": 5.612323856353759,
      "epoch": 2.1754664549424376,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045290858856797503,
      "loss": 5.0107,
      "mean_token_accuracy": 0.1835276648402214,
      "num_tokens": 48012608.0,
      "step": 21920
    },
    {
      "entropy": 5.654703903198242,
      "epoch": 2.175962683604605,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045288689528654444,
      "loss": 5.0385,
      "mean_token_accuracy": 0.17893052250146865,
      "num_tokens": 48022120.0,
      "step": 21925
    },
    {
      "entropy": 5.6130506038665775,
      "epoch": 2.1764589122667726,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045286519759390217,
      "loss": 5.0426,
      "mean_token_accuracy": 0.17459575533866883,
      "num_tokens": 48032629.0,
      "step": 21930
    },
    {
      "entropy": 5.605486631393433,
      "epoch": 2.17695514092894,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004528434954905863,
      "loss": 4.9703,
      "mean_token_accuracy": 0.18507267087697982,
      "num_tokens": 48042934.0,
      "step": 21935
    },
    {
      "entropy": 5.624318408966064,
      "epoch": 2.1774513695911075,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004528217889771349,
      "loss": 5.0396,
      "mean_token_accuracy": 0.18162720054388046,
      "num_tokens": 48052825.0,
      "step": 21940
    },
    {
      "entropy": 5.652318525314331,
      "epoch": 2.177947598253275,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045280007805408636,
      "loss": 5.1175,
      "mean_token_accuracy": 0.1686982050538063,
      "num_tokens": 48064565.0,
      "step": 21945
    },
    {
      "entropy": 5.664990043640136,
      "epoch": 2.1784438269154425,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045277836272197895,
      "loss": 5.0702,
      "mean_token_accuracy": 0.18747231215238572,
      "num_tokens": 48076006.0,
      "step": 21950
    },
    {
      "entropy": 5.706110525131225,
      "epoch": 2.17894005557761,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045275664298135134,
      "loss": 5.1916,
      "mean_token_accuracy": 0.17168551683425903,
      "num_tokens": 48087110.0,
      "step": 21955
    },
    {
      "entropy": 5.638816881179809,
      "epoch": 2.179436284239778,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004527349188327419,
      "loss": 5.0702,
      "mean_token_accuracy": 0.18229641169309616,
      "num_tokens": 48098894.0,
      "step": 21960
    },
    {
      "entropy": 5.664944839477539,
      "epoch": 2.1799325129019453,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004527131902766895,
      "loss": 5.0826,
      "mean_token_accuracy": 0.1786404386162758,
      "num_tokens": 48109729.0,
      "step": 21965
    },
    {
      "entropy": 5.645889091491699,
      "epoch": 2.180428741564113,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004526914573137329,
      "loss": 5.0728,
      "mean_token_accuracy": 0.17559516280889512,
      "num_tokens": 48120345.0,
      "step": 21970
    },
    {
      "entropy": 5.612215375900268,
      "epoch": 2.1809249702262803,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045266971994441116,
      "loss": 5.0107,
      "mean_token_accuracy": 0.17894042581319808,
      "num_tokens": 48130185.0,
      "step": 21975
    },
    {
      "entropy": 5.550708961486817,
      "epoch": 2.1814211988884478,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000452647978169263,
      "loss": 4.9789,
      "mean_token_accuracy": 0.18415218889713286,
      "num_tokens": 48141020.0,
      "step": 21980
    },
    {
      "entropy": 5.60162239074707,
      "epoch": 2.1819174275506152,
      "grad_norm": 1.0,
      "learning_rate": 0.00045262623198882797,
      "loss": 5.0659,
      "mean_token_accuracy": 0.18369440883398055,
      "num_tokens": 48153361.0,
      "step": 21985
    },
    {
      "entropy": 5.616780853271484,
      "epoch": 2.1824136562127827,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004526044814036451,
      "loss": 5.0321,
      "mean_token_accuracy": 0.18169565498828888,
      "num_tokens": 48163604.0,
      "step": 21990
    },
    {
      "entropy": 5.594942569732666,
      "epoch": 2.18290988487495,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045258272641425377,
      "loss": 5.0317,
      "mean_token_accuracy": 0.1733851581811905,
      "num_tokens": 48174732.0,
      "step": 21995
    },
    {
      "entropy": 5.608852338790894,
      "epoch": 2.183406113537118,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004525609670211934,
      "loss": 5.0062,
      "mean_token_accuracy": 0.18386068493127822,
      "num_tokens": 48187082.0,
      "step": 22000
    },
    {
      "entropy": 5.625245761871338,
      "epoch": 2.1839023421992856,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045253920322500373,
      "loss": 5.0195,
      "mean_token_accuracy": 0.18268399238586425,
      "num_tokens": 48196547.0,
      "step": 22005
    },
    {
      "entropy": 5.625101804733276,
      "epoch": 2.184398570861453,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045251743502622445,
      "loss": 5.0641,
      "mean_token_accuracy": 0.1774401292204857,
      "num_tokens": 48207404.0,
      "step": 22010
    },
    {
      "entropy": 5.545613241195679,
      "epoch": 2.1848947995236205,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004524956624253952,
      "loss": 5.0396,
      "mean_token_accuracy": 0.18724942207336426,
      "num_tokens": 48217093.0,
      "step": 22015
    },
    {
      "entropy": 5.644676733016968,
      "epoch": 2.185391028185788,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045247388542305595,
      "loss": 5.0747,
      "mean_token_accuracy": 0.17717769145965576,
      "num_tokens": 48228093.0,
      "step": 22020
    },
    {
      "entropy": 5.5961919784545895,
      "epoch": 2.1858872568479555,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004524521040197469,
      "loss": 5.0151,
      "mean_token_accuracy": 0.1858178675174713,
      "num_tokens": 48238834.0,
      "step": 22025
    },
    {
      "entropy": 5.651755428314209,
      "epoch": 2.186383485510123,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045243031821600794,
      "loss": 5.0795,
      "mean_token_accuracy": 0.18582922518253325,
      "num_tokens": 48250782.0,
      "step": 22030
    },
    {
      "entropy": 5.501127052307129,
      "epoch": 2.1868797141722904,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004524085280123795,
      "loss": 4.9657,
      "mean_token_accuracy": 0.17718659192323685,
      "num_tokens": 48261711.0,
      "step": 22035
    },
    {
      "entropy": 5.586145639419556,
      "epoch": 2.187375942834458,
      "grad_norm": 1.0,
      "learning_rate": 0.0004523867334094018,
      "loss": 5.0734,
      "mean_token_accuracy": 0.1768350124359131,
      "num_tokens": 48273505.0,
      "step": 22040
    },
    {
      "entropy": 5.574091672897339,
      "epoch": 2.187872171496626,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004523649344076154,
      "loss": 4.9796,
      "mean_token_accuracy": 0.18205586671829224,
      "num_tokens": 48284639.0,
      "step": 22045
    },
    {
      "entropy": 5.619840955734253,
      "epoch": 2.1883684001587933,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045234313100756077,
      "loss": 5.0668,
      "mean_token_accuracy": 0.17670910358428954,
      "num_tokens": 48296275.0,
      "step": 22050
    },
    {
      "entropy": 5.598555326461792,
      "epoch": 2.188864628820961,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045232132320977854,
      "loss": 4.9654,
      "mean_token_accuracy": 0.186260387301445,
      "num_tokens": 48306859.0,
      "step": 22055
    },
    {
      "entropy": 5.60958776473999,
      "epoch": 2.1893608574831283,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004522995110148097,
      "loss": 5.1103,
      "mean_token_accuracy": 0.18081016838550568,
      "num_tokens": 48317561.0,
      "step": 22060
    },
    {
      "entropy": 5.594192171096802,
      "epoch": 2.1898570861452957,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000452277694423195,
      "loss": 5.1254,
      "mean_token_accuracy": 0.16781322211027144,
      "num_tokens": 48327814.0,
      "step": 22065
    },
    {
      "entropy": 5.626345252990722,
      "epoch": 2.190353314807463,
      "grad_norm": 1.0,
      "learning_rate": 0.0004522558734354755,
      "loss": 4.9741,
      "mean_token_accuracy": 0.18527152836322786,
      "num_tokens": 48339658.0,
      "step": 22070
    },
    {
      "entropy": 5.6204540729522705,
      "epoch": 2.1908495434696307,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045223404805219225,
      "loss": 5.0909,
      "mean_token_accuracy": 0.17630693167448044,
      "num_tokens": 48352115.0,
      "step": 22075
    },
    {
      "entropy": 5.64398422241211,
      "epoch": 2.191345772131798,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045221221827388655,
      "loss": 5.0467,
      "mean_token_accuracy": 0.18281607031822206,
      "num_tokens": 48362378.0,
      "step": 22080
    },
    {
      "entropy": 5.5951581478118895,
      "epoch": 2.191842000793966,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004521903841010997,
      "loss": 4.9667,
      "mean_token_accuracy": 0.1891895517706871,
      "num_tokens": 48373035.0,
      "step": 22085
    },
    {
      "entropy": 5.549309253692627,
      "epoch": 2.1923382294561335,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004521685455343731,
      "loss": 5.0255,
      "mean_token_accuracy": 0.1879349783062935,
      "num_tokens": 48383737.0,
      "step": 22090
    },
    {
      "entropy": 5.575103807449341,
      "epoch": 2.192834458118301,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045214670257424835,
      "loss": 5.0223,
      "mean_token_accuracy": 0.18242862373590468,
      "num_tokens": 48394547.0,
      "step": 22095
    },
    {
      "entropy": 5.621955633163452,
      "epoch": 2.1933306867804685,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004521248552212671,
      "loss": 5.0721,
      "mean_token_accuracy": 0.1843912973999977,
      "num_tokens": 48404906.0,
      "step": 22100
    },
    {
      "entropy": 5.609711408615112,
      "epoch": 2.193826915442636,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004521030034759711,
      "loss": 5.0359,
      "mean_token_accuracy": 0.18281267285346986,
      "num_tokens": 48416425.0,
      "step": 22105
    },
    {
      "entropy": 5.607714128494263,
      "epoch": 2.1943231441048034,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004520811473389023,
      "loss": 4.9993,
      "mean_token_accuracy": 0.17996059954166413,
      "num_tokens": 48426967.0,
      "step": 22110
    },
    {
      "entropy": 5.659060764312744,
      "epoch": 2.194819372766971,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045205928681060255,
      "loss": 5.1201,
      "mean_token_accuracy": 0.18495838046073915,
      "num_tokens": 48439720.0,
      "step": 22115
    },
    {
      "entropy": 5.642995166778564,
      "epoch": 2.1953156014291384,
      "grad_norm": 1.03125,
      "learning_rate": 0.000452037421891614,
      "loss": 5.0801,
      "mean_token_accuracy": 0.18232346773147584,
      "num_tokens": 48449949.0,
      "step": 22120
    },
    {
      "entropy": 5.573703241348267,
      "epoch": 2.195811830091306,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045201555258247896,
      "loss": 4.9887,
      "mean_token_accuracy": 0.1854351669549942,
      "num_tokens": 48462515.0,
      "step": 22125
    },
    {
      "entropy": 5.591639184951783,
      "epoch": 2.196308058753474,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004519936788837396,
      "loss": 5.009,
      "mean_token_accuracy": 0.18279222697019576,
      "num_tokens": 48471984.0,
      "step": 22130
    },
    {
      "entropy": 5.704627990722656,
      "epoch": 2.1968042874156413,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00045197180079593836,
      "loss": 5.1077,
      "mean_token_accuracy": 0.17814821898937225,
      "num_tokens": 48482814.0,
      "step": 22135
    },
    {
      "entropy": 5.640182209014893,
      "epoch": 2.1973005160778087,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045194991831961777,
      "loss": 5.0386,
      "mean_token_accuracy": 0.1879052698612213,
      "num_tokens": 48493670.0,
      "step": 22140
    },
    {
      "entropy": 5.594465589523315,
      "epoch": 2.197796744739976,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004519280314553206,
      "loss": 5.0225,
      "mean_token_accuracy": 0.17648000866174698,
      "num_tokens": 48504758.0,
      "step": 22145
    },
    {
      "entropy": 5.657790470123291,
      "epoch": 2.1982929734021437,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004519061402035894,
      "loss": 5.0457,
      "mean_token_accuracy": 0.17046160101890565,
      "num_tokens": 48515846.0,
      "step": 22150
    },
    {
      "entropy": 5.649752950668335,
      "epoch": 2.198789202064311,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045188424456496723,
      "loss": 5.0416,
      "mean_token_accuracy": 0.18515683114528655,
      "num_tokens": 48526136.0,
      "step": 22155
    },
    {
      "entropy": 5.680908155441284,
      "epoch": 2.1992854307264786,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045186234453999693,
      "loss": 5.1483,
      "mean_token_accuracy": 0.174190254509449,
      "num_tokens": 48536031.0,
      "step": 22160
    },
    {
      "entropy": 5.535333585739136,
      "epoch": 2.199781659388646,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045184044012922154,
      "loss": 4.9699,
      "mean_token_accuracy": 0.18365664631128312,
      "num_tokens": 48546965.0,
      "step": 22165
    },
    {
      "entropy": 5.647773933410645,
      "epoch": 2.200277888050814,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045181853133318425,
      "loss": 5.0392,
      "mean_token_accuracy": 0.17437478750944138,
      "num_tokens": 48557637.0,
      "step": 22170
    },
    {
      "entropy": 5.656074714660645,
      "epoch": 2.2007741167129815,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004517966181524285,
      "loss": 5.0781,
      "mean_token_accuracy": 0.1869827002286911,
      "num_tokens": 48568578.0,
      "step": 22175
    },
    {
      "entropy": 5.662961959838867,
      "epoch": 2.201270345375149,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004517747005874975,
      "loss": 5.1416,
      "mean_token_accuracy": 0.178591550886631,
      "num_tokens": 48581782.0,
      "step": 22180
    },
    {
      "entropy": 5.55255446434021,
      "epoch": 2.2017665740373165,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004517527786389349,
      "loss": 4.9893,
      "mean_token_accuracy": 0.19379959404468536,
      "num_tokens": 48592432.0,
      "step": 22185
    },
    {
      "entropy": 5.58923225402832,
      "epoch": 2.202262802699484,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045173085230728417,
      "loss": 5.015,
      "mean_token_accuracy": 0.1816897064447403,
      "num_tokens": 48603251.0,
      "step": 22190
    },
    {
      "entropy": 5.669793510437012,
      "epoch": 2.2027590313616514,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004517089215930892,
      "loss": 5.1318,
      "mean_token_accuracy": 0.17440138012170792,
      "num_tokens": 48616118.0,
      "step": 22195
    },
    {
      "entropy": 5.655147075653076,
      "epoch": 2.203255260023819,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045168698649689375,
      "loss": 5.0839,
      "mean_token_accuracy": 0.17955278605222702,
      "num_tokens": 48626903.0,
      "step": 22200
    },
    {
      "entropy": 5.562200403213501,
      "epoch": 2.2037514886859864,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045166504701924176,
      "loss": 4.9752,
      "mean_token_accuracy": 0.18658990263938904,
      "num_tokens": 48638186.0,
      "step": 22205
    },
    {
      "entropy": 5.644933605194092,
      "epoch": 2.204247717348154,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00045164310316067727,
      "loss": 5.086,
      "mean_token_accuracy": 0.17635336369276047,
      "num_tokens": 48649948.0,
      "step": 22210
    },
    {
      "entropy": 5.637668466567993,
      "epoch": 2.2047439460103218,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004516211549217444,
      "loss": 5.0023,
      "mean_token_accuracy": 0.18441600501537322,
      "num_tokens": 48660869.0,
      "step": 22215
    },
    {
      "entropy": 5.571465110778808,
      "epoch": 2.2052401746724892,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004515992023029875,
      "loss": 5.0289,
      "mean_token_accuracy": 0.17761159986257552,
      "num_tokens": 48673383.0,
      "step": 22220
    },
    {
      "entropy": 5.550521564483643,
      "epoch": 2.2057364033346567,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000451577245304951,
      "loss": 4.9725,
      "mean_token_accuracy": 0.19060962349176408,
      "num_tokens": 48684071.0,
      "step": 22225
    },
    {
      "entropy": 5.620009469985962,
      "epoch": 2.206232631996824,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004515552839281792,
      "loss": 5.0259,
      "mean_token_accuracy": 0.17730650305747986,
      "num_tokens": 48695432.0,
      "step": 22230
    },
    {
      "entropy": 5.606878709793091,
      "epoch": 2.2067288606589917,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004515333181732169,
      "loss": 5.0283,
      "mean_token_accuracy": 0.18352137356996537,
      "num_tokens": 48706102.0,
      "step": 22235
    },
    {
      "entropy": 5.581858587265015,
      "epoch": 2.207225089321159,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045151134804060863,
      "loss": 5.0613,
      "mean_token_accuracy": 0.1753111943602562,
      "num_tokens": 48717748.0,
      "step": 22240
    },
    {
      "entropy": 5.610602855682373,
      "epoch": 2.2077213179833266,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004514893735308993,
      "loss": 4.9756,
      "mean_token_accuracy": 0.18633625358343125,
      "num_tokens": 48728941.0,
      "step": 22245
    },
    {
      "entropy": 5.573286867141723,
      "epoch": 2.208217546645494,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004514673946446338,
      "loss": 5.0373,
      "mean_token_accuracy": 0.17846245765686036,
      "num_tokens": 48738565.0,
      "step": 22250
    },
    {
      "entropy": 5.579632186889649,
      "epoch": 2.208713775307662,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004514454113823572,
      "loss": 5.01,
      "mean_token_accuracy": 0.1794385567307472,
      "num_tokens": 48749537.0,
      "step": 22255
    },
    {
      "entropy": 5.574956893920898,
      "epoch": 2.2092100039698295,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004514234237446146,
      "loss": 5.0172,
      "mean_token_accuracy": 0.1754564970731735,
      "num_tokens": 48760388.0,
      "step": 22260
    },
    {
      "entropy": 5.572825288772583,
      "epoch": 2.209706232631997,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045140143173195117,
      "loss": 5.0406,
      "mean_token_accuracy": 0.17571160644292833,
      "num_tokens": 48770551.0,
      "step": 22265
    },
    {
      "entropy": 5.565246343612671,
      "epoch": 2.2102024612941644,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004513794353449124,
      "loss": 5.0008,
      "mean_token_accuracy": 0.17992022782564163,
      "num_tokens": 48780739.0,
      "step": 22270
    },
    {
      "entropy": 5.6053609371185305,
      "epoch": 2.210698689956332,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004513574345840437,
      "loss": 5.1005,
      "mean_token_accuracy": 0.17709578722715377,
      "num_tokens": 48791478.0,
      "step": 22275
    },
    {
      "entropy": 5.644941663742065,
      "epoch": 2.2111949186184994,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004513354294498907,
      "loss": 5.0888,
      "mean_token_accuracy": 0.17296138703823088,
      "num_tokens": 48802202.0,
      "step": 22280
    },
    {
      "entropy": 5.628314638137818,
      "epoch": 2.211691147280667,
      "grad_norm": 1.0,
      "learning_rate": 0.00045131341994299895,
      "loss": 5.0503,
      "mean_token_accuracy": 0.1794401928782463,
      "num_tokens": 48813619.0,
      "step": 22285
    },
    {
      "entropy": 5.619992399215699,
      "epoch": 2.2121873759428343,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004512914060639143,
      "loss": 4.9982,
      "mean_token_accuracy": 0.18857115209102632,
      "num_tokens": 48824951.0,
      "step": 22290
    },
    {
      "entropy": 5.611372232437134,
      "epoch": 2.212683604605002,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004512693878131827,
      "loss": 5.0268,
      "mean_token_accuracy": 0.18474863469600677,
      "num_tokens": 48834328.0,
      "step": 22295
    },
    {
      "entropy": 5.623195695877075,
      "epoch": 2.2131798332671697,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045124736519135003,
      "loss": 5.0633,
      "mean_token_accuracy": 0.1765931189060211,
      "num_tokens": 48843880.0,
      "step": 22300
    },
    {
      "entropy": 5.618527126312256,
      "epoch": 2.213676061929337,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045122533819896257,
      "loss": 5.0648,
      "mean_token_accuracy": 0.1825164958834648,
      "num_tokens": 48854893.0,
      "step": 22305
    },
    {
      "entropy": 5.634518241882324,
      "epoch": 2.2141722905915047,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004512033068365664,
      "loss": 5.123,
      "mean_token_accuracy": 0.17719939053058625,
      "num_tokens": 48867308.0,
      "step": 22310
    },
    {
      "entropy": 5.671744728088379,
      "epoch": 2.214668519253672,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004511812711047079,
      "loss": 5.105,
      "mean_token_accuracy": 0.18109505921602248,
      "num_tokens": 48880222.0,
      "step": 22315
    },
    {
      "entropy": 5.6507405757904055,
      "epoch": 2.2151647479158396,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045115923100393356,
      "loss": 5.0403,
      "mean_token_accuracy": 0.17849747687578202,
      "num_tokens": 48890345.0,
      "step": 22320
    },
    {
      "entropy": 5.577552366256714,
      "epoch": 2.215660976578007,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045113718653478984,
      "loss": 4.9811,
      "mean_token_accuracy": 0.18120139092206955,
      "num_tokens": 48901267.0,
      "step": 22325
    },
    {
      "entropy": 5.601315832138061,
      "epoch": 2.2161572052401746,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004511151376978234,
      "loss": 5.0582,
      "mean_token_accuracy": 0.18365704417228698,
      "num_tokens": 48911739.0,
      "step": 22330
    },
    {
      "entropy": 5.640086126327515,
      "epoch": 2.216653433902342,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045109308449358114,
      "loss": 5.0047,
      "mean_token_accuracy": 0.18365770131349562,
      "num_tokens": 48922376.0,
      "step": 22335
    },
    {
      "entropy": 5.593026733398437,
      "epoch": 2.21714966256451,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004510710269226098,
      "loss": 4.9302,
      "mean_token_accuracy": 0.1843005508184433,
      "num_tokens": 48934101.0,
      "step": 22340
    },
    {
      "entropy": 5.545723295211792,
      "epoch": 2.2176458912266774,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004510489649854564,
      "loss": 5.0106,
      "mean_token_accuracy": 0.18549758046865464,
      "num_tokens": 48944684.0,
      "step": 22345
    },
    {
      "entropy": 5.614818525314331,
      "epoch": 2.218142119888845,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000451026898682668,
      "loss": 5.0373,
      "mean_token_accuracy": 0.1744355171918869,
      "num_tokens": 48955255.0,
      "step": 22350
    },
    {
      "entropy": 5.578813457489014,
      "epoch": 2.2186383485510124,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004510048280147918,
      "loss": 5.0345,
      "mean_token_accuracy": 0.18148267716169358,
      "num_tokens": 48967034.0,
      "step": 22355
    },
    {
      "entropy": 5.639485883712768,
      "epoch": 2.21913457721318,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004509827529823751,
      "loss": 5.1306,
      "mean_token_accuracy": 0.1719082459807396,
      "num_tokens": 48977903.0,
      "step": 22360
    },
    {
      "entropy": 5.626160144805908,
      "epoch": 2.2196308058753473,
      "grad_norm": 1.0,
      "learning_rate": 0.00045096067358596536,
      "loss": 5.021,
      "mean_token_accuracy": 0.18464754968881608,
      "num_tokens": 48988889.0,
      "step": 22365
    },
    {
      "entropy": 5.6882617473602295,
      "epoch": 2.220127034537515,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045093858982611006,
      "loss": 5.1254,
      "mean_token_accuracy": 0.1660460948944092,
      "num_tokens": 48998946.0,
      "step": 22370
    },
    {
      "entropy": 5.611918258666992,
      "epoch": 2.2206232631996823,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004509165017033569,
      "loss": 5.0206,
      "mean_token_accuracy": 0.18311215788125992,
      "num_tokens": 49008141.0,
      "step": 22375
    },
    {
      "entropy": 5.567286157608033,
      "epoch": 2.2211194918618498,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045089440921825354,
      "loss": 4.9338,
      "mean_token_accuracy": 0.18628780841827391,
      "num_tokens": 49019387.0,
      "step": 22380
    },
    {
      "entropy": 5.5681493282318115,
      "epoch": 2.2216157205240172,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045087231237134787,
      "loss": 4.9857,
      "mean_token_accuracy": 0.18051772564649582,
      "num_tokens": 49030280.0,
      "step": 22385
    },
    {
      "entropy": 5.607708597183228,
      "epoch": 2.222111949186185,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045085021116318783,
      "loss": 4.9826,
      "mean_token_accuracy": 0.18792865574359893,
      "num_tokens": 49041766.0,
      "step": 22390
    },
    {
      "entropy": 5.6917987823486325,
      "epoch": 2.2226081778483526,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004508281055943214,
      "loss": 5.2107,
      "mean_token_accuracy": 0.1673697352409363,
      "num_tokens": 49052931.0,
      "step": 22395
    },
    {
      "entropy": 5.613233280181885,
      "epoch": 2.22310440651052,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004508059956652969,
      "loss": 5.0151,
      "mean_token_accuracy": 0.1801508978009224,
      "num_tokens": 49064153.0,
      "step": 22400
    },
    {
      "entropy": 5.596367645263672,
      "epoch": 2.2236006351726876,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004507838813766625,
      "loss": 5.0272,
      "mean_token_accuracy": 0.18033125549554824,
      "num_tokens": 49075857.0,
      "step": 22405
    },
    {
      "entropy": 5.685031509399414,
      "epoch": 2.224096863834855,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004507617627289666,
      "loss": 5.0655,
      "mean_token_accuracy": 0.18049957752227783,
      "num_tokens": 49086632.0,
      "step": 22410
    },
    {
      "entropy": 5.635158538818359,
      "epoch": 2.2245930924970225,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004507396397227577,
      "loss": 5.0303,
      "mean_token_accuracy": 0.17873288840055465,
      "num_tokens": 49098086.0,
      "step": 22415
    },
    {
      "entropy": 5.531681060791016,
      "epoch": 2.22508932115919,
      "grad_norm": 1.0,
      "learning_rate": 0.00045071751235858454,
      "loss": 4.951,
      "mean_token_accuracy": 0.18719427585601806,
      "num_tokens": 49108857.0,
      "step": 22420
    },
    {
      "entropy": 5.5738760471344,
      "epoch": 2.2255855498213575,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004506953806369956,
      "loss": 4.9785,
      "mean_token_accuracy": 0.1831977993249893,
      "num_tokens": 49119609.0,
      "step": 22425
    },
    {
      "entropy": 5.624512767791748,
      "epoch": 2.2260817784835254,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004506732445585399,
      "loss": 5.0628,
      "mean_token_accuracy": 0.1835842624306679,
      "num_tokens": 49130047.0,
      "step": 22430
    },
    {
      "entropy": 5.575485563278198,
      "epoch": 2.226578007145693,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004506511041237662,
      "loss": 5.0111,
      "mean_token_accuracy": 0.18555817604064942,
      "num_tokens": 49140517.0,
      "step": 22435
    },
    {
      "entropy": 5.629457569122314,
      "epoch": 2.2270742358078603,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045062895933322366,
      "loss": 5.0238,
      "mean_token_accuracy": 0.18141377419233323,
      "num_tokens": 49151497.0,
      "step": 22440
    },
    {
      "entropy": 5.513612890243531,
      "epoch": 2.227570464470028,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004506068101874614,
      "loss": 4.96,
      "mean_token_accuracy": 0.18025163561105728,
      "num_tokens": 49161568.0,
      "step": 22445
    },
    {
      "entropy": 5.6428327560424805,
      "epoch": 2.2280666931321953,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045058465668702856,
      "loss": 5.0742,
      "mean_token_accuracy": 0.18122751265764236,
      "num_tokens": 49173554.0,
      "step": 22450
    },
    {
      "entropy": 5.686952304840088,
      "epoch": 2.2285629217943628,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045056249883247457,
      "loss": 5.0567,
      "mean_token_accuracy": 0.17806104570627213,
      "num_tokens": 49185399.0,
      "step": 22455
    },
    {
      "entropy": 5.627108192443847,
      "epoch": 2.2290591504565302,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045054033662434903,
      "loss": 4.9731,
      "mean_token_accuracy": 0.18383095264434815,
      "num_tokens": 49195489.0,
      "step": 22460
    },
    {
      "entropy": 5.524414348602295,
      "epoch": 2.2295553791186977,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045051817006320127,
      "loss": 4.9733,
      "mean_token_accuracy": 0.1817865714430809,
      "num_tokens": 49204652.0,
      "step": 22465
    },
    {
      "entropy": 5.558700799942017,
      "epoch": 2.230051607780865,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00045049599914958124,
      "loss": 5.0527,
      "mean_token_accuracy": 0.17651440054178238,
      "num_tokens": 49215420.0,
      "step": 22470
    },
    {
      "entropy": 5.615656518936158,
      "epoch": 2.230547836443033,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004504738238840385,
      "loss": 5.0191,
      "mean_token_accuracy": 0.18550100177526474,
      "num_tokens": 49225375.0,
      "step": 22475
    },
    {
      "entropy": 5.64311637878418,
      "epoch": 2.2310440651052006,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004504516442671231,
      "loss": 5.0261,
      "mean_token_accuracy": 0.1768382892012596,
      "num_tokens": 49238399.0,
      "step": 22480
    },
    {
      "entropy": 5.590234279632568,
      "epoch": 2.231540293767368,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00045042946029938495,
      "loss": 5.0007,
      "mean_token_accuracy": 0.17938061207532882,
      "num_tokens": 49251069.0,
      "step": 22485
    },
    {
      "entropy": 5.606921815872193,
      "epoch": 2.2320365224295355,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045040727198137416,
      "loss": 5.0399,
      "mean_token_accuracy": 0.17893531620502473,
      "num_tokens": 49263859.0,
      "step": 22490
    },
    {
      "entropy": 5.670394086837769,
      "epoch": 2.232532751091703,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045038507931364103,
      "loss": 5.0404,
      "mean_token_accuracy": 0.18212466388940812,
      "num_tokens": 49274846.0,
      "step": 22495
    },
    {
      "entropy": 5.5949359893798825,
      "epoch": 2.2330289797538705,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00045036288229673583,
      "loss": 4.9735,
      "mean_token_accuracy": 0.18181572407484053,
      "num_tokens": 49285197.0,
      "step": 22500
    },
    {
      "entropy": 5.534357357025146,
      "epoch": 2.233525208416038,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045034068093120907,
      "loss": 4.9729,
      "mean_token_accuracy": 0.17781236171722412,
      "num_tokens": 49296238.0,
      "step": 22505
    },
    {
      "entropy": 5.708719396591187,
      "epoch": 2.2340214370782054,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004503184752176112,
      "loss": 5.2021,
      "mean_token_accuracy": 0.17302089780569077,
      "num_tokens": 49308262.0,
      "step": 22510
    },
    {
      "entropy": 5.614587450027466,
      "epoch": 2.2345176657403734,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000450296265156493,
      "loss": 4.9817,
      "mean_token_accuracy": 0.18488356918096543,
      "num_tokens": 49318881.0,
      "step": 22515
    },
    {
      "entropy": 5.659103965759277,
      "epoch": 2.235013894402541,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045027405074840506,
      "loss": 5.1286,
      "mean_token_accuracy": 0.1776016414165497,
      "num_tokens": 49329470.0,
      "step": 22520
    },
    {
      "entropy": 5.614867687225342,
      "epoch": 2.2355101230647083,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004502518319938984,
      "loss": 5.0246,
      "mean_token_accuracy": 0.18422296941280364,
      "num_tokens": 49339806.0,
      "step": 22525
    },
    {
      "entropy": 5.735942363739014,
      "epoch": 2.236006351726876,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045022960889352394,
      "loss": 5.1269,
      "mean_token_accuracy": 0.18072216510772704,
      "num_tokens": 49350295.0,
      "step": 22530
    },
    {
      "entropy": 5.642465686798095,
      "epoch": 2.2365025803890433,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045020738144783273,
      "loss": 5.0284,
      "mean_token_accuracy": 0.17840695083141328,
      "num_tokens": 49360735.0,
      "step": 22535
    },
    {
      "entropy": 5.583511829376221,
      "epoch": 2.2369988090512107,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004501851496573761,
      "loss": 4.9919,
      "mean_token_accuracy": 0.18231824785470963,
      "num_tokens": 49372278.0,
      "step": 22540
    },
    {
      "entropy": 5.609589862823486,
      "epoch": 2.237495037713378,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004501629135227051,
      "loss": 5.0838,
      "mean_token_accuracy": 0.17394402474164963,
      "num_tokens": 49382842.0,
      "step": 22545
    },
    {
      "entropy": 5.630756998062134,
      "epoch": 2.2379912663755457,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004501406730443714,
      "loss": 5.0846,
      "mean_token_accuracy": 0.17592091858386993,
      "num_tokens": 49394633.0,
      "step": 22550
    },
    {
      "entropy": 5.651486206054687,
      "epoch": 2.238487495037713,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045011842822292636,
      "loss": 5.0598,
      "mean_token_accuracy": 0.1820286974310875,
      "num_tokens": 49404731.0,
      "step": 22555
    },
    {
      "entropy": 5.669883108139038,
      "epoch": 2.238983723699881,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004500961790589216,
      "loss": 5.0572,
      "mean_token_accuracy": 0.1790739879012108,
      "num_tokens": 49415918.0,
      "step": 22560
    },
    {
      "entropy": 5.596751022338867,
      "epoch": 2.2394799523620486,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000450073925552909,
      "loss": 4.9811,
      "mean_token_accuracy": 0.18394855111837388,
      "num_tokens": 49427458.0,
      "step": 22565
    },
    {
      "entropy": 5.600291633605957,
      "epoch": 2.239976181024216,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004500516677054403,
      "loss": 5.0083,
      "mean_token_accuracy": 0.18762864917516708,
      "num_tokens": 49437696.0,
      "step": 22570
    },
    {
      "entropy": 5.682634019851685,
      "epoch": 2.2404724096863835,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045002940551706726,
      "loss": 5.2167,
      "mean_token_accuracy": 0.17418304830789566,
      "num_tokens": 49449220.0,
      "step": 22575
    },
    {
      "entropy": 5.650023937225342,
      "epoch": 2.240968638348551,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045000713898834225,
      "loss": 5.0601,
      "mean_token_accuracy": 0.17733512371778487,
      "num_tokens": 49459624.0,
      "step": 22580
    },
    {
      "entropy": 5.6336987018585205,
      "epoch": 2.2414648670107185,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004499848681198173,
      "loss": 4.9727,
      "mean_token_accuracy": 0.18369352966547012,
      "num_tokens": 49471813.0,
      "step": 22585
    },
    {
      "entropy": 5.593286323547363,
      "epoch": 2.241961095672886,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004499625929120447,
      "loss": 5.0374,
      "mean_token_accuracy": 0.18345981389284133,
      "num_tokens": 49482624.0,
      "step": 22590
    },
    {
      "entropy": 5.650317621231079,
      "epoch": 2.2424573243350534,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004499403133655768,
      "loss": 5.0676,
      "mean_token_accuracy": 0.17360280007123946,
      "num_tokens": 49493791.0,
      "step": 22595
    },
    {
      "entropy": 5.686618900299072,
      "epoch": 2.2429535529972213,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004499180294809661,
      "loss": 5.0946,
      "mean_token_accuracy": 0.17357805669307708,
      "num_tokens": 49504060.0,
      "step": 22600
    },
    {
      "entropy": 5.5684901714324955,
      "epoch": 2.243449781659389,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044989574125876516,
      "loss": 4.9303,
      "mean_token_accuracy": 0.19110321253538132,
      "num_tokens": 49515962.0,
      "step": 22605
    },
    {
      "entropy": 5.634757232666016,
      "epoch": 2.2439460103215563,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004498734486995267,
      "loss": 5.0581,
      "mean_token_accuracy": 0.17810797095298767,
      "num_tokens": 49526676.0,
      "step": 22610
    },
    {
      "entropy": 5.628345203399658,
      "epoch": 2.2444422389837237,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044985115180380355,
      "loss": 5.0669,
      "mean_token_accuracy": 0.17670186012983322,
      "num_tokens": 49537347.0,
      "step": 22615
    },
    {
      "entropy": 5.613627481460571,
      "epoch": 2.2449384676458912,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044982885057214857,
      "loss": 5.0711,
      "mean_token_accuracy": 0.17337154746055602,
      "num_tokens": 49547422.0,
      "step": 22620
    },
    {
      "entropy": 5.589520883560181,
      "epoch": 2.2454346963080587,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044980654500511485,
      "loss": 4.915,
      "mean_token_accuracy": 0.18988650292158127,
      "num_tokens": 49557250.0,
      "step": 22625
    },
    {
      "entropy": 5.680695629119873,
      "epoch": 2.245930924970226,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044978423510325544,
      "loss": 5.1186,
      "mean_token_accuracy": 0.18005286455154418,
      "num_tokens": 49567106.0,
      "step": 22630
    },
    {
      "entropy": 5.627146005630493,
      "epoch": 2.2464271536323936,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004497619208671237,
      "loss": 5.0588,
      "mean_token_accuracy": 0.1758490100502968,
      "num_tokens": 49578978.0,
      "step": 22635
    },
    {
      "entropy": 5.65406289100647,
      "epoch": 2.246923382294561,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004497396022972728,
      "loss": 5.0958,
      "mean_token_accuracy": 0.1807176500558853,
      "num_tokens": 49589730.0,
      "step": 22640
    },
    {
      "entropy": 5.63521180152893,
      "epoch": 2.247419610956729,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004497172793942564,
      "loss": 5.0102,
      "mean_token_accuracy": 0.18112933933734893,
      "num_tokens": 49600881.0,
      "step": 22645
    },
    {
      "entropy": 5.639083337783814,
      "epoch": 2.2479158396188965,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004496949521586279,
      "loss": 5.0661,
      "mean_token_accuracy": 0.18171762228012084,
      "num_tokens": 49611120.0,
      "step": 22650
    },
    {
      "entropy": 5.6399928569793705,
      "epoch": 2.248412068281064,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044967262059094097,
      "loss": 5.0842,
      "mean_token_accuracy": 0.17527048140764237,
      "num_tokens": 49622978.0,
      "step": 22655
    },
    {
      "entropy": 5.617404174804688,
      "epoch": 2.2489082969432315,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044965028469174956,
      "loss": 5.068,
      "mean_token_accuracy": 0.17994914203882217,
      "num_tokens": 49633678.0,
      "step": 22660
    },
    {
      "entropy": 5.623934125900268,
      "epoch": 2.249404525605399,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044962794446160736,
      "loss": 5.0481,
      "mean_token_accuracy": 0.17750751972198486,
      "num_tokens": 49644886.0,
      "step": 22665
    },
    {
      "entropy": 5.642688655853272,
      "epoch": 2.2499007542675664,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044960559990106844,
      "loss": 5.0045,
      "mean_token_accuracy": 0.18628148287534713,
      "num_tokens": 49656790.0,
      "step": 22670
    },
    {
      "entropy": 5.669056844711304,
      "epoch": 2.250396982929734,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004495832510106869,
      "loss": 5.0547,
      "mean_token_accuracy": 0.175840824842453,
      "num_tokens": 49667943.0,
      "step": 22675
    },
    {
      "entropy": 5.6153035163879395,
      "epoch": 2.2508932115919014,
      "grad_norm": 1.0,
      "learning_rate": 0.0004495608977910169,
      "loss": 5.031,
      "mean_token_accuracy": 0.17484548687934875,
      "num_tokens": 49678652.0,
      "step": 22680
    },
    {
      "entropy": 5.624973249435425,
      "epoch": 2.2513894402540693,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004495385402426128,
      "loss": 4.992,
      "mean_token_accuracy": 0.18833995908498763,
      "num_tokens": 49688951.0,
      "step": 22685
    },
    {
      "entropy": 5.612255477905274,
      "epoch": 2.2518856689162368,
      "grad_norm": 1.0625,
      "learning_rate": 0.000449516178366029,
      "loss": 5.0532,
      "mean_token_accuracy": 0.18091504871845246,
      "num_tokens": 49698750.0,
      "step": 22690
    },
    {
      "entropy": 5.626090431213379,
      "epoch": 2.2523818975784042,
      "grad_norm": 1.0,
      "learning_rate": 0.00044949381216182,
      "loss": 5.1385,
      "mean_token_accuracy": 0.173087278008461,
      "num_tokens": 49709442.0,
      "step": 22695
    },
    {
      "entropy": 5.645272970199585,
      "epoch": 2.2528781262405717,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044947144163054036,
      "loss": 5.0318,
      "mean_token_accuracy": 0.1817680314183235,
      "num_tokens": 49720517.0,
      "step": 22700
    },
    {
      "entropy": 5.663498067855835,
      "epoch": 2.253374354902739,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004494490667727451,
      "loss": 5.0185,
      "mean_token_accuracy": 0.17679052501916886,
      "num_tokens": 49731413.0,
      "step": 22705
    },
    {
      "entropy": 5.705662488937378,
      "epoch": 2.2538705835649067,
      "grad_norm": 0.953125,
      "learning_rate": 0.00044942668758898873,
      "loss": 5.1974,
      "mean_token_accuracy": 0.1696411356329918,
      "num_tokens": 49742492.0,
      "step": 22710
    },
    {
      "entropy": 5.663849878311157,
      "epoch": 2.254366812227074,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004494043040798265,
      "loss": 5.0403,
      "mean_token_accuracy": 0.17619373947381972,
      "num_tokens": 49753981.0,
      "step": 22715
    },
    {
      "entropy": 5.636537027359009,
      "epoch": 2.2548630408892416,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004493819162458133,
      "loss": 5.0542,
      "mean_token_accuracy": 0.17774236798286439,
      "num_tokens": 49764718.0,
      "step": 22720
    },
    {
      "entropy": 5.641672992706299,
      "epoch": 2.255359269551409,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004493595240875043,
      "loss": 5.0235,
      "mean_token_accuracy": 0.18034570217132567,
      "num_tokens": 49775361.0,
      "step": 22725
    },
    {
      "entropy": 5.556173849105835,
      "epoch": 2.255855498213577,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044933712760545486,
      "loss": 5.0139,
      "mean_token_accuracy": 0.17924351543188094,
      "num_tokens": 49786670.0,
      "step": 22730
    },
    {
      "entropy": 5.678067350387574,
      "epoch": 2.2563517268757445,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004493147268002203,
      "loss": 5.1365,
      "mean_token_accuracy": 0.17745867669582366,
      "num_tokens": 49796514.0,
      "step": 22735
    },
    {
      "entropy": 5.601131296157837,
      "epoch": 2.256847955537912,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044929232167235613,
      "loss": 5.0035,
      "mean_token_accuracy": 0.18621800392866134,
      "num_tokens": 49807042.0,
      "step": 22740
    },
    {
      "entropy": 5.637433862686157,
      "epoch": 2.2573441842000794,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004492699122224179,
      "loss": 5.0609,
      "mean_token_accuracy": 0.1797993466258049,
      "num_tokens": 49818779.0,
      "step": 22745
    },
    {
      "entropy": 5.644547367095948,
      "epoch": 2.257840412862247,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044924749845096144,
      "loss": 5.0656,
      "mean_token_accuracy": 0.17352507561445235,
      "num_tokens": 49829901.0,
      "step": 22750
    },
    {
      "entropy": 5.640950727462768,
      "epoch": 2.2583366415244144,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004492250803585425,
      "loss": 5.0545,
      "mean_token_accuracy": 0.181841179728508,
      "num_tokens": 49840837.0,
      "step": 22755
    },
    {
      "entropy": 5.6468099594116214,
      "epoch": 2.258832870186582,
      "grad_norm": 1.0,
      "learning_rate": 0.0004492026579457169,
      "loss": 5.0109,
      "mean_token_accuracy": 0.1758326694369316,
      "num_tokens": 49853052.0,
      "step": 22760
    },
    {
      "entropy": 5.607258272171021,
      "epoch": 2.2593290988487493,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004491802312130408,
      "loss": 5.0348,
      "mean_token_accuracy": 0.17712527066469191,
      "num_tokens": 49864210.0,
      "step": 22765
    },
    {
      "entropy": 5.594322252273559,
      "epoch": 2.2598253275109172,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004491578001610703,
      "loss": 4.9938,
      "mean_token_accuracy": 0.18893788158893585,
      "num_tokens": 49874995.0,
      "step": 22770
    },
    {
      "entropy": 5.548060703277588,
      "epoch": 2.2603215561730847,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044913536479036163,
      "loss": 4.9968,
      "mean_token_accuracy": 0.18296184241771699,
      "num_tokens": 49886240.0,
      "step": 22775
    },
    {
      "entropy": 5.6103928565979,
      "epoch": 2.260817784835252,
      "grad_norm": 0.953125,
      "learning_rate": 0.00044911292510147117,
      "loss": 4.9994,
      "mean_token_accuracy": 0.18265680968761444,
      "num_tokens": 49897580.0,
      "step": 22780
    },
    {
      "entropy": 5.577655172348022,
      "epoch": 2.2613140134974197,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004490904810949553,
      "loss": 5.0489,
      "mean_token_accuracy": 0.17760042250156402,
      "num_tokens": 49907280.0,
      "step": 22785
    },
    {
      "entropy": 5.604145574569702,
      "epoch": 2.261810242159587,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044906803277137064,
      "loss": 5.0749,
      "mean_token_accuracy": 0.18133510649204254,
      "num_tokens": 49917552.0,
      "step": 22790
    },
    {
      "entropy": 5.636494922637939,
      "epoch": 2.2623064708217546,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004490455801312738,
      "loss": 5.0368,
      "mean_token_accuracy": 0.17665129750967026,
      "num_tokens": 49930610.0,
      "step": 22795
    },
    {
      "entropy": 5.634942960739136,
      "epoch": 2.262802699483922,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004490231231752217,
      "loss": 5.0712,
      "mean_token_accuracy": 0.1766537457704544,
      "num_tokens": 49941054.0,
      "step": 22800
    },
    {
      "entropy": 5.55372257232666,
      "epoch": 2.2632989281460896,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004490006619037711,
      "loss": 4.9303,
      "mean_token_accuracy": 0.18579587787389756,
      "num_tokens": 49950885.0,
      "step": 22805
    },
    {
      "entropy": 5.577817344665528,
      "epoch": 2.263795156808257,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044897819631747903,
      "loss": 5.0178,
      "mean_token_accuracy": 0.18237457871437074,
      "num_tokens": 49961723.0,
      "step": 22810
    },
    {
      "entropy": 5.602283763885498,
      "epoch": 2.2642913854704245,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044895572641690256,
      "loss": 4.9987,
      "mean_token_accuracy": 0.18160301446914673,
      "num_tokens": 49974124.0,
      "step": 22815
    },
    {
      "entropy": 5.621014833450317,
      "epoch": 2.2647876141325924,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000448933252202599,
      "loss": 5.0396,
      "mean_token_accuracy": 0.18052243888378144,
      "num_tokens": 49985792.0,
      "step": 22820
    },
    {
      "entropy": 5.567944717407227,
      "epoch": 2.26528384279476,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044891077367512544,
      "loss": 4.9624,
      "mean_token_accuracy": 0.18293532282114028,
      "num_tokens": 49996935.0,
      "step": 22825
    },
    {
      "entropy": 5.711354303359985,
      "epoch": 2.2657800714569274,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004488882908350396,
      "loss": 5.1299,
      "mean_token_accuracy": 0.17742265909910201,
      "num_tokens": 50007545.0,
      "step": 22830
    },
    {
      "entropy": 5.685041046142578,
      "epoch": 2.266276300119095,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004488658036828987,
      "loss": 5.0498,
      "mean_token_accuracy": 0.18104502707719802,
      "num_tokens": 50018797.0,
      "step": 22835
    },
    {
      "entropy": 5.599407911300659,
      "epoch": 2.2667725287812623,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004488433122192606,
      "loss": 5.0199,
      "mean_token_accuracy": 0.18283187597990036,
      "num_tokens": 50030884.0,
      "step": 22840
    },
    {
      "entropy": 5.662004613876343,
      "epoch": 2.26726875744343,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044882081644468286,
      "loss": 5.1656,
      "mean_token_accuracy": 0.1705210641026497,
      "num_tokens": 50042327.0,
      "step": 22845
    },
    {
      "entropy": 5.646599626541137,
      "epoch": 2.2677649861055973,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044879831635972345,
      "loss": 5.0769,
      "mean_token_accuracy": 0.18408871591091155,
      "num_tokens": 50053733.0,
      "step": 22850
    },
    {
      "entropy": 5.618116188049316,
      "epoch": 2.268261214767765,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004487758119649403,
      "loss": 5.0206,
      "mean_token_accuracy": 0.18191457241773606,
      "num_tokens": 50064153.0,
      "step": 22855
    },
    {
      "entropy": 5.520285654067993,
      "epoch": 2.2687574434299327,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044875330326089153,
      "loss": 4.9611,
      "mean_token_accuracy": 0.19104929119348527,
      "num_tokens": 50074397.0,
      "step": 22860
    },
    {
      "entropy": 5.546572637557984,
      "epoch": 2.2692536720921,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004487307902481352,
      "loss": 4.9613,
      "mean_token_accuracy": 0.17853392511606217,
      "num_tokens": 50084074.0,
      "step": 22865
    },
    {
      "entropy": 5.6483618259429935,
      "epoch": 2.2697499007542676,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044870827292722967,
      "loss": 5.0069,
      "mean_token_accuracy": 0.1771961748600006,
      "num_tokens": 50094535.0,
      "step": 22870
    },
    {
      "entropy": 5.636296272277832,
      "epoch": 2.270246129416435,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004486857512987333,
      "loss": 5.0654,
      "mean_token_accuracy": 0.17716252952814102,
      "num_tokens": 50105907.0,
      "step": 22875
    },
    {
      "entropy": 5.598095083236695,
      "epoch": 2.2707423580786026,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044866322536320447,
      "loss": 5.0405,
      "mean_token_accuracy": 0.18135128915309906,
      "num_tokens": 50117361.0,
      "step": 22880
    },
    {
      "entropy": 5.65628023147583,
      "epoch": 2.27123858674077,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044864069512120185,
      "loss": 5.057,
      "mean_token_accuracy": 0.18315195441246032,
      "num_tokens": 50128598.0,
      "step": 22885
    },
    {
      "entropy": 5.655677032470703,
      "epoch": 2.2717348154029375,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044861816057328424,
      "loss": 5.0393,
      "mean_token_accuracy": 0.1795967012643814,
      "num_tokens": 50140470.0,
      "step": 22890
    },
    {
      "entropy": 5.664349699020386,
      "epoch": 2.272231044065105,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044859562172001036,
      "loss": 5.0966,
      "mean_token_accuracy": 0.17870337516069412,
      "num_tokens": 50150950.0,
      "step": 22895
    },
    {
      "entropy": 5.581730699539184,
      "epoch": 2.2727272727272725,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004485730785619391,
      "loss": 4.9516,
      "mean_token_accuracy": 0.1872153550386429,
      "num_tokens": 50161173.0,
      "step": 22900
    },
    {
      "entropy": 5.6546868801116945,
      "epoch": 2.2732235013894404,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004485505310996295,
      "loss": 5.1665,
      "mean_token_accuracy": 0.17529118657112122,
      "num_tokens": 50173874.0,
      "step": 22905
    },
    {
      "entropy": 5.759654569625854,
      "epoch": 2.273719730051608,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004485279793336406,
      "loss": 5.1206,
      "mean_token_accuracy": 0.1709430381655693,
      "num_tokens": 50185123.0,
      "step": 22910
    },
    {
      "entropy": 5.6528442859649655,
      "epoch": 2.2742159587137754,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004485054232645319,
      "loss": 5.0894,
      "mean_token_accuracy": 0.1780424565076828,
      "num_tokens": 50195699.0,
      "step": 22915
    },
    {
      "entropy": 5.673516702651978,
      "epoch": 2.274712187375943,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004484828628928625,
      "loss": 5.0796,
      "mean_token_accuracy": 0.18566827178001405,
      "num_tokens": 50206623.0,
      "step": 22920
    },
    {
      "entropy": 5.675904035568237,
      "epoch": 2.2752084160381103,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044846029821919193,
      "loss": 5.1414,
      "mean_token_accuracy": 0.1753354012966156,
      "num_tokens": 50216086.0,
      "step": 22925
    },
    {
      "entropy": 5.590954065322876,
      "epoch": 2.275704644700278,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004484377292440797,
      "loss": 5.0341,
      "mean_token_accuracy": 0.17475888431072234,
      "num_tokens": 50228472.0,
      "step": 22930
    },
    {
      "entropy": 5.6492925643920895,
      "epoch": 2.2762008733624453,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044841515596808553,
      "loss": 5.0629,
      "mean_token_accuracy": 0.17460928559303285,
      "num_tokens": 50238607.0,
      "step": 22935
    },
    {
      "entropy": 5.633523607254029,
      "epoch": 2.276697102024613,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00044839257839176914,
      "loss": 5.0233,
      "mean_token_accuracy": 0.17985857874155045,
      "num_tokens": 50249890.0,
      "step": 22940
    },
    {
      "entropy": 5.59987850189209,
      "epoch": 2.2771933306867806,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044836999651569055,
      "loss": 5.0183,
      "mean_token_accuracy": 0.18572691231966018,
      "num_tokens": 50260234.0,
      "step": 22945
    },
    {
      "entropy": 5.618515062332153,
      "epoch": 2.277689559348948,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004483474103404095,
      "loss": 5.0749,
      "mean_token_accuracy": 0.18087799847126007,
      "num_tokens": 50271605.0,
      "step": 22950
    },
    {
      "entropy": 5.6484137058258055,
      "epoch": 2.2781857880111156,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044832481986648633,
      "loss": 4.9421,
      "mean_token_accuracy": 0.18496544659137726,
      "num_tokens": 50281837.0,
      "step": 22955
    },
    {
      "entropy": 5.625103425979614,
      "epoch": 2.278682016673283,
      "grad_norm": 1.09375,
      "learning_rate": 0.000448302225094481,
      "loss": 4.9986,
      "mean_token_accuracy": 0.17956970930099486,
      "num_tokens": 50292524.0,
      "step": 22960
    },
    {
      "entropy": 5.637637948989868,
      "epoch": 2.2791782453354505,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044827962602495393,
      "loss": 5.0056,
      "mean_token_accuracy": 0.18071590065956117,
      "num_tokens": 50304009.0,
      "step": 22965
    },
    {
      "entropy": 5.630226516723633,
      "epoch": 2.279674473997618,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044825702265846557,
      "loss": 5.066,
      "mean_token_accuracy": 0.17845210880041124,
      "num_tokens": 50316103.0,
      "step": 22970
    },
    {
      "entropy": 5.653515768051148,
      "epoch": 2.2801707026597855,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004482344149955763,
      "loss": 5.0836,
      "mean_token_accuracy": 0.17831098884344102,
      "num_tokens": 50326127.0,
      "step": 22975
    },
    {
      "entropy": 5.561366558074951,
      "epoch": 2.280666931321953,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004482118030368469,
      "loss": 4.9718,
      "mean_token_accuracy": 0.18455941379070281,
      "num_tokens": 50337044.0,
      "step": 22980
    },
    {
      "entropy": 5.632459115982056,
      "epoch": 2.2811631599841204,
      "grad_norm": 0.984375,
      "learning_rate": 0.000448189186782838,
      "loss": 5.1782,
      "mean_token_accuracy": 0.17138134688138962,
      "num_tokens": 50347796.0,
      "step": 22985
    },
    {
      "entropy": 5.596288108825684,
      "epoch": 2.2816593886462884,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044816656623411044,
      "loss": 5.0589,
      "mean_token_accuracy": 0.18208117336034774,
      "num_tokens": 50358454.0,
      "step": 22990
    },
    {
      "entropy": 5.586537885665893,
      "epoch": 2.282155617308456,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004481439413912252,
      "loss": 4.9667,
      "mean_token_accuracy": 0.1835189625620842,
      "num_tokens": 50369415.0,
      "step": 22995
    },
    {
      "entropy": 5.608301973342895,
      "epoch": 2.2826518459706233,
      "grad_norm": 1.0,
      "learning_rate": 0.00044812131225474333,
      "loss": 5.0512,
      "mean_token_accuracy": 0.1734281525015831,
      "num_tokens": 50379977.0,
      "step": 23000
    },
    {
      "entropy": 5.670031785964966,
      "epoch": 2.283148074632791,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004480986788252259,
      "loss": 5.0535,
      "mean_token_accuracy": 0.17744080126285552,
      "num_tokens": 50390465.0,
      "step": 23005
    },
    {
      "entropy": 5.653299951553345,
      "epoch": 2.2836443032949583,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004480760411032343,
      "loss": 5.0819,
      "mean_token_accuracy": 0.18080077916383744,
      "num_tokens": 50402230.0,
      "step": 23010
    },
    {
      "entropy": 5.5613690376281735,
      "epoch": 2.2841405319571257,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004480533990893298,
      "loss": 4.9683,
      "mean_token_accuracy": 0.18629131317138672,
      "num_tokens": 50413819.0,
      "step": 23015
    },
    {
      "entropy": 5.752253818511963,
      "epoch": 2.284636760619293,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004480307527840739,
      "loss": 5.2494,
      "mean_token_accuracy": 0.1674409791827202,
      "num_tokens": 50424977.0,
      "step": 23020
    },
    {
      "entropy": 5.661560201644898,
      "epoch": 2.285132989281461,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004480081021880282,
      "loss": 5.0501,
      "mean_token_accuracy": 0.18193721771240234,
      "num_tokens": 50438357.0,
      "step": 23025
    },
    {
      "entropy": 5.641304206848145,
      "epoch": 2.2856292179436286,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004479854473017543,
      "loss": 5.0533,
      "mean_token_accuracy": 0.18109430819749833,
      "num_tokens": 50449081.0,
      "step": 23030
    },
    {
      "entropy": 5.551891136169433,
      "epoch": 2.286125446605796,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044796278812581417,
      "loss": 4.9564,
      "mean_token_accuracy": 0.18654400408267974,
      "num_tokens": 50460358.0,
      "step": 23035
    },
    {
      "entropy": 5.635677194595337,
      "epoch": 2.2866216752679636,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004479401246607695,
      "loss": 5.0436,
      "mean_token_accuracy": 0.18149758875370026,
      "num_tokens": 50470955.0,
      "step": 23040
    },
    {
      "entropy": 5.646644544601441,
      "epoch": 2.287117903930131,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004479174569071823,
      "loss": 5.0001,
      "mean_token_accuracy": 0.18407780081033706,
      "num_tokens": 50482250.0,
      "step": 23045
    },
    {
      "entropy": 5.607111644744873,
      "epoch": 2.2876141325922985,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004478947848656149,
      "loss": 5.0396,
      "mean_token_accuracy": 0.18312409669160842,
      "num_tokens": 50493868.0,
      "step": 23050
    },
    {
      "entropy": 5.628189468383789,
      "epoch": 2.288110361254466,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044787210853662943,
      "loss": 5.0805,
      "mean_token_accuracy": 0.1827125370502472,
      "num_tokens": 50504012.0,
      "step": 23055
    },
    {
      "entropy": 5.64683165550232,
      "epoch": 2.2886065899166335,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004478494279207881,
      "loss": 5.0031,
      "mean_token_accuracy": 0.18559710830450057,
      "num_tokens": 50515503.0,
      "step": 23060
    },
    {
      "entropy": 5.603887557983398,
      "epoch": 2.289102818578801,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044782674301865346,
      "loss": 5.0414,
      "mean_token_accuracy": 0.18022053092718124,
      "num_tokens": 50526180.0,
      "step": 23065
    },
    {
      "entropy": 5.63401198387146,
      "epoch": 2.2895990472409684,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004478040538307879,
      "loss": 5.0579,
      "mean_token_accuracy": 0.17227750569581984,
      "num_tokens": 50536711.0,
      "step": 23070
    },
    {
      "entropy": 5.6835980892181395,
      "epoch": 2.2900952759031363,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004477813603577543,
      "loss": 5.12,
      "mean_token_accuracy": 0.1741674080491066,
      "num_tokens": 50547736.0,
      "step": 23075
    },
    {
      "entropy": 5.617640829086303,
      "epoch": 2.290591504565304,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044775866260011516,
      "loss": 5.0073,
      "mean_token_accuracy": 0.18709687888622284,
      "num_tokens": 50558173.0,
      "step": 23080
    },
    {
      "entropy": 5.62267541885376,
      "epoch": 2.2910877332274713,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004477359605584335,
      "loss": 5.0351,
      "mean_token_accuracy": 0.1793284833431244,
      "num_tokens": 50568421.0,
      "step": 23085
    },
    {
      "entropy": 5.612886428833008,
      "epoch": 2.2915839618896388,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044771325423327227,
      "loss": 5.0075,
      "mean_token_accuracy": 0.18390186429023742,
      "num_tokens": 50579182.0,
      "step": 23090
    },
    {
      "entropy": 5.649978828430176,
      "epoch": 2.2920801905518062,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004476905436251945,
      "loss": 5.0392,
      "mean_token_accuracy": 0.18247573226690292,
      "num_tokens": 50590916.0,
      "step": 23095
    },
    {
      "entropy": 5.522751903533935,
      "epoch": 2.2925764192139737,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044766782873476333,
      "loss": 4.9302,
      "mean_token_accuracy": 0.18821359276771546,
      "num_tokens": 50600682.0,
      "step": 23100
    },
    {
      "entropy": 5.682747888565063,
      "epoch": 2.293072647876141,
      "grad_norm": 1.0,
      "learning_rate": 0.000447645109562542,
      "loss": 5.185,
      "mean_token_accuracy": 0.16975466310977935,
      "num_tokens": 50612103.0,
      "step": 23105
    },
    {
      "entropy": 5.714118432998657,
      "epoch": 2.293568876538309,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044762238610909414,
      "loss": 5.1128,
      "mean_token_accuracy": 0.17552594542503358,
      "num_tokens": 50623894.0,
      "step": 23110
    },
    {
      "entropy": 5.725135374069214,
      "epoch": 2.2940651052004766,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004475996583749829,
      "loss": 5.0939,
      "mean_token_accuracy": 0.17670314610004426,
      "num_tokens": 50634495.0,
      "step": 23115
    },
    {
      "entropy": 5.698043537139893,
      "epoch": 2.294561333862644,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044757692636077215,
      "loss": 5.2078,
      "mean_token_accuracy": 0.17463595271110535,
      "num_tokens": 50645921.0,
      "step": 23120
    },
    {
      "entropy": 5.667640542984008,
      "epoch": 2.2950575625248115,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044755419006702545,
      "loss": 5.0549,
      "mean_token_accuracy": 0.1769983857870102,
      "num_tokens": 50657047.0,
      "step": 23125
    },
    {
      "entropy": 5.686962413787842,
      "epoch": 2.295553791186979,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044753144949430667,
      "loss": 5.0312,
      "mean_token_accuracy": 0.17797466069459916,
      "num_tokens": 50668271.0,
      "step": 23130
    },
    {
      "entropy": 5.588381385803222,
      "epoch": 2.2960500198491465,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004475087046431798,
      "loss": 4.9958,
      "mean_token_accuracy": 0.18436052203178405,
      "num_tokens": 50678870.0,
      "step": 23135
    },
    {
      "entropy": 5.603098678588867,
      "epoch": 2.296546248511314,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004474859555142087,
      "loss": 5.0479,
      "mean_token_accuracy": 0.18641940802335738,
      "num_tokens": 50690322.0,
      "step": 23140
    },
    {
      "entropy": 5.59818663597107,
      "epoch": 2.2970424771734814,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044746320210795753,
      "loss": 4.9817,
      "mean_token_accuracy": 0.18174895495176316,
      "num_tokens": 50701958.0,
      "step": 23145
    },
    {
      "entropy": 5.597720003128051,
      "epoch": 2.297538705835649,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004474404444249906,
      "loss": 5.0468,
      "mean_token_accuracy": 0.1810399904847145,
      "num_tokens": 50712962.0,
      "step": 23150
    },
    {
      "entropy": 5.593395519256592,
      "epoch": 2.2980349344978164,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044741768246587226,
      "loss": 5.0861,
      "mean_token_accuracy": 0.18369501680135727,
      "num_tokens": 50723442.0,
      "step": 23155
    },
    {
      "entropy": 5.608185434341431,
      "epoch": 2.2985311631599843,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004473949162311669,
      "loss": 4.9476,
      "mean_token_accuracy": 0.1850915879011154,
      "num_tokens": 50733188.0,
      "step": 23160
    },
    {
      "entropy": 5.64476432800293,
      "epoch": 2.2990273918221518,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004473721457214391,
      "loss": 5.0933,
      "mean_token_accuracy": 0.17341089844703675,
      "num_tokens": 50744023.0,
      "step": 23165
    },
    {
      "entropy": 5.745205450057983,
      "epoch": 2.2995236204843192,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004473493709372535,
      "loss": 5.2715,
      "mean_token_accuracy": 0.1724076673388481,
      "num_tokens": 50756807.0,
      "step": 23170
    },
    {
      "entropy": 5.697967386245727,
      "epoch": 2.3000198491464867,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004473265918791749,
      "loss": 5.1489,
      "mean_token_accuracy": 0.17232249230146407,
      "num_tokens": 50768981.0,
      "step": 23175
    },
    {
      "entropy": 5.67499794960022,
      "epoch": 2.300516077808654,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004473038085477682,
      "loss": 5.0579,
      "mean_token_accuracy": 0.1840350091457367,
      "num_tokens": 50780111.0,
      "step": 23180
    },
    {
      "entropy": 5.674689912796021,
      "epoch": 2.3010123064708217,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044728102094359826,
      "loss": 5.0828,
      "mean_token_accuracy": 0.17615040987730027,
      "num_tokens": 50790318.0,
      "step": 23185
    },
    {
      "entropy": 5.656909227371216,
      "epoch": 2.301508535132989,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004472582290672303,
      "loss": 5.103,
      "mean_token_accuracy": 0.17907795459032058,
      "num_tokens": 50800734.0,
      "step": 23190
    },
    {
      "entropy": 5.558125114440918,
      "epoch": 2.3020047637951566,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044723543291922936,
      "loss": 5.0121,
      "mean_token_accuracy": 0.1849045842885971,
      "num_tokens": 50811236.0,
      "step": 23195
    },
    {
      "entropy": 5.5683611869812015,
      "epoch": 2.3025009924573245,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044721263250016086,
      "loss": 5.0117,
      "mean_token_accuracy": 0.18168788254261017,
      "num_tokens": 50821776.0,
      "step": 23200
    },
    {
      "entropy": 5.651003170013428,
      "epoch": 2.302997221119492,
      "grad_norm": 1.125,
      "learning_rate": 0.00044718982781059017,
      "loss": 5.1543,
      "mean_token_accuracy": 0.1748604118824005,
      "num_tokens": 50834156.0,
      "step": 23205
    },
    {
      "entropy": 5.628415727615357,
      "epoch": 2.3034934497816595,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044716701885108275,
      "loss": 5.0177,
      "mean_token_accuracy": 0.18407442271709443,
      "num_tokens": 50845086.0,
      "step": 23210
    },
    {
      "entropy": 5.686002731323242,
      "epoch": 2.303989678443827,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044714420562220423,
      "loss": 5.036,
      "mean_token_accuracy": 0.18035176396369934,
      "num_tokens": 50858637.0,
      "step": 23215
    },
    {
      "entropy": 5.653166675567627,
      "epoch": 2.3044859071059944,
      "grad_norm": 1.0,
      "learning_rate": 0.0004471213881245204,
      "loss": 5.1483,
      "mean_token_accuracy": 0.17320256531238556,
      "num_tokens": 50871159.0,
      "step": 23220
    },
    {
      "entropy": 5.5747236728668215,
      "epoch": 2.304982135768162,
      "grad_norm": 0.984375,
      "learning_rate": 0.000447098566358597,
      "loss": 5.0219,
      "mean_token_accuracy": 0.18062575459480285,
      "num_tokens": 50884072.0,
      "step": 23225
    },
    {
      "entropy": 5.599307250976563,
      "epoch": 2.3054783644303294,
      "grad_norm": 0.953125,
      "learning_rate": 0.00044707574032500005,
      "loss": 5.0419,
      "mean_token_accuracy": 0.18177380561828613,
      "num_tokens": 50896627.0,
      "step": 23230
    },
    {
      "entropy": 5.55665225982666,
      "epoch": 2.305974593092497,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004470529100242955,
      "loss": 4.9613,
      "mean_token_accuracy": 0.1857459217309952,
      "num_tokens": 50907476.0,
      "step": 23235
    },
    {
      "entropy": 5.58920841217041,
      "epoch": 2.3064708217546643,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004470300754570495,
      "loss": 5.057,
      "mean_token_accuracy": 0.1779991701245308,
      "num_tokens": 50918104.0,
      "step": 23240
    },
    {
      "entropy": 5.5810808658599855,
      "epoch": 2.3069670504168323,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004470072366238284,
      "loss": 5.0062,
      "mean_token_accuracy": 0.18053653836250305,
      "num_tokens": 50930575.0,
      "step": 23245
    },
    {
      "entropy": 5.625973701477051,
      "epoch": 2.3074632790789997,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044698439352519836,
      "loss": 4.9754,
      "mean_token_accuracy": 0.17963041365146637,
      "num_tokens": 50940592.0,
      "step": 23250
    },
    {
      "entropy": 5.627779579162597,
      "epoch": 2.307959507741167,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044696154616172604,
      "loss": 5.0562,
      "mean_token_accuracy": 0.1762169286608696,
      "num_tokens": 50952134.0,
      "step": 23255
    },
    {
      "entropy": 5.621954393386841,
      "epoch": 2.3084557364033347,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004469386945339779,
      "loss": 5.0903,
      "mean_token_accuracy": 0.1758081167936325,
      "num_tokens": 50962721.0,
      "step": 23260
    },
    {
      "entropy": 5.57981595993042,
      "epoch": 2.308951965065502,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004469158386425206,
      "loss": 5.0358,
      "mean_token_accuracy": 0.17933979034423828,
      "num_tokens": 50974893.0,
      "step": 23265
    },
    {
      "entropy": 5.6289268970489506,
      "epoch": 2.3094481937276696,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000446892978487921,
      "loss": 5.0012,
      "mean_token_accuracy": 0.17918888479471207,
      "num_tokens": 50986225.0,
      "step": 23270
    },
    {
      "entropy": 5.630105018615723,
      "epoch": 2.309944422389837,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044687011407074583,
      "loss": 5.0892,
      "mean_token_accuracy": 0.17255181521177293,
      "num_tokens": 50997924.0,
      "step": 23275
    },
    {
      "entropy": 5.611375045776367,
      "epoch": 2.3104406510520046,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004468472453915623,
      "loss": 4.9736,
      "mean_token_accuracy": 0.17822391241788865,
      "num_tokens": 51008179.0,
      "step": 23280
    },
    {
      "entropy": 5.562429428100586,
      "epoch": 2.3109368797141725,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004468243724509373,
      "loss": 5.0019,
      "mean_token_accuracy": 0.18385012894868852,
      "num_tokens": 51018462.0,
      "step": 23285
    },
    {
      "entropy": 5.624410390853882,
      "epoch": 2.31143310837634,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004468014952494382,
      "loss": 5.0784,
      "mean_token_accuracy": 0.1773458018898964,
      "num_tokens": 51029769.0,
      "step": 23290
    },
    {
      "entropy": 5.616210508346557,
      "epoch": 2.3119293370385074,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044677861378763205,
      "loss": 5.0813,
      "mean_token_accuracy": 0.1811317577958107,
      "num_tokens": 51041058.0,
      "step": 23295
    },
    {
      "entropy": 5.597333002090454,
      "epoch": 2.312425565700675,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044675572806608655,
      "loss": 5.0048,
      "mean_token_accuracy": 0.1832798957824707,
      "num_tokens": 51050958.0,
      "step": 23300
    },
    {
      "entropy": 5.636134052276612,
      "epoch": 2.3129217943628424,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044673283808536915,
      "loss": 5.0144,
      "mean_token_accuracy": 0.18979429304599763,
      "num_tokens": 51062459.0,
      "step": 23305
    },
    {
      "entropy": 5.716775846481323,
      "epoch": 2.31341802302501,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004467099438460474,
      "loss": 5.2075,
      "mean_token_accuracy": 0.17235147655010224,
      "num_tokens": 51074596.0,
      "step": 23310
    },
    {
      "entropy": 5.691996288299561,
      "epoch": 2.3139142516871773,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044668704534868886,
      "loss": 5.0648,
      "mean_token_accuracy": 0.17396527230739595,
      "num_tokens": 51085265.0,
      "step": 23315
    },
    {
      "entropy": 5.634092092514038,
      "epoch": 2.314410480349345,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044666414259386173,
      "loss": 5.0224,
      "mean_token_accuracy": 0.17738422751426697,
      "num_tokens": 51095986.0,
      "step": 23320
    },
    {
      "entropy": 5.652958488464355,
      "epoch": 2.3149067090115123,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044664123558213367,
      "loss": 5.0317,
      "mean_token_accuracy": 0.17708924114704133,
      "num_tokens": 51107518.0,
      "step": 23325
    },
    {
      "entropy": 5.599997854232788,
      "epoch": 2.31540293767368,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004466183243140729,
      "loss": 4.9737,
      "mean_token_accuracy": 0.18690697103738785,
      "num_tokens": 51119715.0,
      "step": 23330
    },
    {
      "entropy": 5.531274366378784,
      "epoch": 2.3158991663358477,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044659540879024746,
      "loss": 5.0038,
      "mean_token_accuracy": 0.18382956236600875,
      "num_tokens": 51130801.0,
      "step": 23335
    },
    {
      "entropy": 5.568726253509522,
      "epoch": 2.316395394998015,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004465724890112256,
      "loss": 5.033,
      "mean_token_accuracy": 0.18055622428655624,
      "num_tokens": 51142509.0,
      "step": 23340
    },
    {
      "entropy": 5.699008417129517,
      "epoch": 2.3168916236601826,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044654956497757576,
      "loss": 5.0815,
      "mean_token_accuracy": 0.1800437316298485,
      "num_tokens": 51154839.0,
      "step": 23345
    },
    {
      "entropy": 5.622169303894043,
      "epoch": 2.31738785232235,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044652663668986634,
      "loss": 5.0212,
      "mean_token_accuracy": 0.1883740708231926,
      "num_tokens": 51165293.0,
      "step": 23350
    },
    {
      "entropy": 5.579819631576538,
      "epoch": 2.3178840809845176,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000446503704148666,
      "loss": 5.0945,
      "mean_token_accuracy": 0.17392228841781615,
      "num_tokens": 51176371.0,
      "step": 23355
    },
    {
      "entropy": 5.600088500976563,
      "epoch": 2.318380309646685,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004464807673545433,
      "loss": 5.0215,
      "mean_token_accuracy": 0.18677290081977843,
      "num_tokens": 51187240.0,
      "step": 23360
    },
    {
      "entropy": 5.714385747909546,
      "epoch": 2.3188765383088525,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004464578263080671,
      "loss": 5.1012,
      "mean_token_accuracy": 0.1800432100892067,
      "num_tokens": 51199084.0,
      "step": 23365
    },
    {
      "entropy": 5.627686166763306,
      "epoch": 2.3193727669710205,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044643488100980623,
      "loss": 4.9636,
      "mean_token_accuracy": 0.18528710901737214,
      "num_tokens": 51209885.0,
      "step": 23370
    },
    {
      "entropy": 5.617595195770264,
      "epoch": 2.319868995633188,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004464119314603298,
      "loss": 5.0277,
      "mean_token_accuracy": 0.18587291985750198,
      "num_tokens": 51220644.0,
      "step": 23375
    },
    {
      "entropy": 5.6011340618133545,
      "epoch": 2.3203652242953554,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044638897766020673,
      "loss": 5.0089,
      "mean_token_accuracy": 0.18246026635169982,
      "num_tokens": 51231342.0,
      "step": 23380
    },
    {
      "entropy": 5.5752785205841064,
      "epoch": 2.320861452957523,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004463660196100064,
      "loss": 4.9261,
      "mean_token_accuracy": 0.1885739967226982,
      "num_tokens": 51241368.0,
      "step": 23385
    },
    {
      "entropy": 5.601705408096313,
      "epoch": 2.3213576816196904,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044634305731029803,
      "loss": 5.1272,
      "mean_token_accuracy": 0.17788600474596022,
      "num_tokens": 51252784.0,
      "step": 23390
    },
    {
      "entropy": 5.620908546447754,
      "epoch": 2.321853910281858,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044632009076165096,
      "loss": 4.9987,
      "mean_token_accuracy": 0.18201373517513275,
      "num_tokens": 51264246.0,
      "step": 23395
    },
    {
      "entropy": 5.590815830230713,
      "epoch": 2.3223501389440253,
      "grad_norm": 1.109375,
      "learning_rate": 0.000446297119964635,
      "loss": 4.9727,
      "mean_token_accuracy": 0.18816782236099244,
      "num_tokens": 51274575.0,
      "step": 23400
    },
    {
      "entropy": 5.626755332946777,
      "epoch": 2.322846367606193,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004462741449198194,
      "loss": 4.9959,
      "mean_token_accuracy": 0.18259112536907196,
      "num_tokens": 51285384.0,
      "step": 23405
    },
    {
      "entropy": 5.700454139709473,
      "epoch": 2.3233425962683603,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044625116562777413,
      "loss": 5.0849,
      "mean_token_accuracy": 0.18548366874456407,
      "num_tokens": 51297983.0,
      "step": 23410
    },
    {
      "entropy": 5.552319765090942,
      "epoch": 2.323838824930528,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004462281820890689,
      "loss": 4.9835,
      "mean_token_accuracy": 0.18645429164171218,
      "num_tokens": 51308795.0,
      "step": 23415
    },
    {
      "entropy": 5.564811086654663,
      "epoch": 2.3243350535926957,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044620519430427376,
      "loss": 5.0011,
      "mean_token_accuracy": 0.18694903552532197,
      "num_tokens": 51318597.0,
      "step": 23420
    },
    {
      "entropy": 5.568665266036987,
      "epoch": 2.324831282254863,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004461822022739588,
      "loss": 4.9486,
      "mean_token_accuracy": 0.1891544431447983,
      "num_tokens": 51329238.0,
      "step": 23425
    },
    {
      "entropy": 5.691944932937622,
      "epoch": 2.3253275109170306,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000446159205998694,
      "loss": 5.0469,
      "mean_token_accuracy": 0.18282119184732437,
      "num_tokens": 51340295.0,
      "step": 23430
    },
    {
      "entropy": 5.605220222473145,
      "epoch": 2.325823739579198,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004461362054790497,
      "loss": 5.0111,
      "mean_token_accuracy": 0.18312028795480728,
      "num_tokens": 51350702.0,
      "step": 23435
    },
    {
      "entropy": 5.644420003890991,
      "epoch": 2.3263199682413656,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044611320071559637,
      "loss": 5.0367,
      "mean_token_accuracy": 0.1840574935078621,
      "num_tokens": 51359794.0,
      "step": 23440
    },
    {
      "entropy": 5.639877462387085,
      "epoch": 2.326816196903533,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044609019170890424,
      "loss": 5.0039,
      "mean_token_accuracy": 0.18463980704545974,
      "num_tokens": 51368932.0,
      "step": 23445
    },
    {
      "entropy": 5.6417967796325685,
      "epoch": 2.3273124255657005,
      "grad_norm": 1.0,
      "learning_rate": 0.00044606717845954415,
      "loss": 5.0631,
      "mean_token_accuracy": 0.18312048763036728,
      "num_tokens": 51381111.0,
      "step": 23450
    },
    {
      "entropy": 5.587430763244629,
      "epoch": 2.3278086542278684,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044604416096808656,
      "loss": 4.9764,
      "mean_token_accuracy": 0.19775430858135223,
      "num_tokens": 51392976.0,
      "step": 23455
    },
    {
      "entropy": 5.6342613220214846,
      "epoch": 2.328304882890036,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044602113923510235,
      "loss": 4.9951,
      "mean_token_accuracy": 0.17804827988147737,
      "num_tokens": 51403586.0,
      "step": 23460
    },
    {
      "entropy": 5.6421550750732425,
      "epoch": 2.3288011115522034,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044599811326116246,
      "loss": 5.0558,
      "mean_token_accuracy": 0.17971463203430177,
      "num_tokens": 51414035.0,
      "step": 23465
    },
    {
      "entropy": 5.607219552993774,
      "epoch": 2.329297340214371,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044597508304683784,
      "loss": 5.0555,
      "mean_token_accuracy": 0.17796962410211564,
      "num_tokens": 51424111.0,
      "step": 23470
    },
    {
      "entropy": 5.587814378738403,
      "epoch": 2.3297935688765383,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004459520485926995,
      "loss": 4.975,
      "mean_token_accuracy": 0.18841340839862825,
      "num_tokens": 51434566.0,
      "step": 23475
    },
    {
      "entropy": 5.646214294433594,
      "epoch": 2.330289797538706,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004459290098993188,
      "loss": 5.0444,
      "mean_token_accuracy": 0.18415906131267548,
      "num_tokens": 51445568.0,
      "step": 23480
    },
    {
      "entropy": 5.664250898361206,
      "epoch": 2.3307860262008733,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044590596696726686,
      "loss": 5.0312,
      "mean_token_accuracy": 0.1875285819172859,
      "num_tokens": 51456482.0,
      "step": 23485
    },
    {
      "entropy": 5.6547853469848635,
      "epoch": 2.3312822548630407,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004458829197971153,
      "loss": 5.1047,
      "mean_token_accuracy": 0.17393710613250732,
      "num_tokens": 51467809.0,
      "step": 23490
    },
    {
      "entropy": 5.658076381683349,
      "epoch": 2.3317784835252082,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044585986838943556,
      "loss": 5.013,
      "mean_token_accuracy": 0.17719709873199463,
      "num_tokens": 51477927.0,
      "step": 23495
    },
    {
      "entropy": 5.610753583908081,
      "epoch": 2.3322747121873757,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044583681274479927,
      "loss": 5.0717,
      "mean_token_accuracy": 0.17908694744110107,
      "num_tokens": 51488432.0,
      "step": 23500
    },
    {
      "entropy": 5.648841047286988,
      "epoch": 2.3327709408495436,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004458137528637781,
      "loss": 5.0285,
      "mean_token_accuracy": 0.17959897518157958,
      "num_tokens": 51499149.0,
      "step": 23505
    },
    {
      "entropy": 5.69306116104126,
      "epoch": 2.333267169511711,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044579068874694386,
      "loss": 5.0734,
      "mean_token_accuracy": 0.17763389348983766,
      "num_tokens": 51510672.0,
      "step": 23510
    },
    {
      "entropy": 5.6682701587677,
      "epoch": 2.3337633981738786,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004457676203948687,
      "loss": 5.0668,
      "mean_token_accuracy": 0.17572954446077346,
      "num_tokens": 51521855.0,
      "step": 23515
    },
    {
      "entropy": 5.676049280166626,
      "epoch": 2.334259626836046,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004457445478081244,
      "loss": 5.0433,
      "mean_token_accuracy": 0.17920605093240738,
      "num_tokens": 51531970.0,
      "step": 23520
    },
    {
      "entropy": 5.637546539306641,
      "epoch": 2.3347558554982135,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044572147098728335,
      "loss": 5.1159,
      "mean_token_accuracy": 0.17654835432767868,
      "num_tokens": 51543110.0,
      "step": 23525
    },
    {
      "entropy": 5.6031311511993405,
      "epoch": 2.335252084160381,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044569838993291757,
      "loss": 5.0539,
      "mean_token_accuracy": 0.1784362018108368,
      "num_tokens": 51552991.0,
      "step": 23530
    },
    {
      "entropy": 5.635091257095337,
      "epoch": 2.3357483128225485,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044567530464559963,
      "loss": 5.0705,
      "mean_token_accuracy": 0.183559912443161,
      "num_tokens": 51564472.0,
      "step": 23535
    },
    {
      "entropy": 5.5926045894622805,
      "epoch": 2.3362445414847164,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004456522151259019,
      "loss": 5.0827,
      "mean_token_accuracy": 0.17887851148843764,
      "num_tokens": 51576051.0,
      "step": 23540
    },
    {
      "entropy": 5.698253440856933,
      "epoch": 2.336740770146884,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000445629121374397,
      "loss": 5.0637,
      "mean_token_accuracy": 0.18052899688482285,
      "num_tokens": 51586658.0,
      "step": 23545
    },
    {
      "entropy": 5.650602531433106,
      "epoch": 2.3372369988090513,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044560602339165746,
      "loss": 5.0424,
      "mean_token_accuracy": 0.17425750941038132,
      "num_tokens": 51596614.0,
      "step": 23550
    },
    {
      "entropy": 5.6322003364562985,
      "epoch": 2.337733227471219,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004455829211782562,
      "loss": 5.0643,
      "mean_token_accuracy": 0.1820590540766716,
      "num_tokens": 51607102.0,
      "step": 23555
    },
    {
      "entropy": 5.705343580245971,
      "epoch": 2.3382294561333863,
      "grad_norm": 1.0,
      "learning_rate": 0.0004455598147347661,
      "loss": 5.121,
      "mean_token_accuracy": 0.17273477464914322,
      "num_tokens": 51619079.0,
      "step": 23560
    },
    {
      "entropy": 5.614614343643188,
      "epoch": 2.3387256847955538,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004455367040617601,
      "loss": 4.9687,
      "mean_token_accuracy": 0.1865694686770439,
      "num_tokens": 51630273.0,
      "step": 23565
    },
    {
      "entropy": 5.600826740264893,
      "epoch": 2.3392219134577212,
      "grad_norm": 1.0,
      "learning_rate": 0.00044551358915981135,
      "loss": 5.0395,
      "mean_token_accuracy": 0.18638865649700165,
      "num_tokens": 51640480.0,
      "step": 23570
    },
    {
      "entropy": 5.611730527877808,
      "epoch": 2.3397181421198887,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004454904700294929,
      "loss": 5.0782,
      "mean_token_accuracy": 0.1795220896601677,
      "num_tokens": 51651831.0,
      "step": 23575
    },
    {
      "entropy": 5.647612762451172,
      "epoch": 2.340214370782056,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004454673466713783,
      "loss": 5.073,
      "mean_token_accuracy": 0.1816953256726265,
      "num_tokens": 51664061.0,
      "step": 23580
    },
    {
      "entropy": 5.600452470779419,
      "epoch": 2.3407105994442237,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044544421908604073,
      "loss": 5.0371,
      "mean_token_accuracy": 0.18238913118839264,
      "num_tokens": 51675081.0,
      "step": 23585
    },
    {
      "entropy": 5.59129958152771,
      "epoch": 2.3412068281063916,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044542108727405386,
      "loss": 4.9524,
      "mean_token_accuracy": 0.18965190201997756,
      "num_tokens": 51684477.0,
      "step": 23590
    },
    {
      "entropy": 5.725885248184204,
      "epoch": 2.341703056768559,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044539795123599117,
      "loss": 5.0886,
      "mean_token_accuracy": 0.1819786474108696,
      "num_tokens": 51694420.0,
      "step": 23595
    },
    {
      "entropy": 5.511602067947388,
      "epoch": 2.3421992854307265,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044537481097242654,
      "loss": 4.9221,
      "mean_token_accuracy": 0.19197066277265548,
      "num_tokens": 51704635.0,
      "step": 23600
    },
    {
      "entropy": 5.652921724319458,
      "epoch": 2.342695514092894,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044535166648393377,
      "loss": 5.1382,
      "mean_token_accuracy": 0.17719145864248276,
      "num_tokens": 51716149.0,
      "step": 23605
    },
    {
      "entropy": 5.559132432937622,
      "epoch": 2.3431917427550615,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044532851777108663,
      "loss": 4.9832,
      "mean_token_accuracy": 0.1820719376206398,
      "num_tokens": 51726861.0,
      "step": 23610
    },
    {
      "entropy": 5.667479467391968,
      "epoch": 2.343687971417229,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004453053648344593,
      "loss": 5.2233,
      "mean_token_accuracy": 0.17189372330904007,
      "num_tokens": 51738111.0,
      "step": 23615
    },
    {
      "entropy": 5.6041971206665036,
      "epoch": 2.3441842000793964,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044528220767462595,
      "loss": 5.043,
      "mean_token_accuracy": 0.18069785684347153,
      "num_tokens": 51749004.0,
      "step": 23620
    },
    {
      "entropy": 5.61161618232727,
      "epoch": 2.3446804287415643,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004452590462921607,
      "loss": 5.0337,
      "mean_token_accuracy": 0.175082029402256,
      "num_tokens": 51759557.0,
      "step": 23625
    },
    {
      "entropy": 5.6766650676727295,
      "epoch": 2.345176657403732,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044523588068763804,
      "loss": 5.0966,
      "mean_token_accuracy": 0.1771151140332222,
      "num_tokens": 51769921.0,
      "step": 23630
    },
    {
      "entropy": 5.606097173690796,
      "epoch": 2.3456728860658993,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004452127108616324,
      "loss": 4.9559,
      "mean_token_accuracy": 0.18375281542539595,
      "num_tokens": 51780158.0,
      "step": 23635
    },
    {
      "entropy": 5.673742055892944,
      "epoch": 2.3461691147280668,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044518953681471824,
      "loss": 5.0484,
      "mean_token_accuracy": 0.18278007656335832,
      "num_tokens": 51790502.0,
      "step": 23640
    },
    {
      "entropy": 5.636298179626465,
      "epoch": 2.3466653433902342,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004451663585474703,
      "loss": 5.0782,
      "mean_token_accuracy": 0.1712587594985962,
      "num_tokens": 51801420.0,
      "step": 23645
    },
    {
      "entropy": 5.644163513183594,
      "epoch": 2.3471615720524017,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044514317606046336,
      "loss": 5.0762,
      "mean_token_accuracy": 0.1870712235569954,
      "num_tokens": 51812399.0,
      "step": 23650
    },
    {
      "entropy": 5.677338409423828,
      "epoch": 2.347657800714569,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004451199893542723,
      "loss": 5.1334,
      "mean_token_accuracy": 0.1735447883605957,
      "num_tokens": 51824268.0,
      "step": 23655
    },
    {
      "entropy": 5.598235464096069,
      "epoch": 2.3481540293767367,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004450967984294721,
      "loss": 5.045,
      "mean_token_accuracy": 0.18006807565689087,
      "num_tokens": 51835842.0,
      "step": 23660
    },
    {
      "entropy": 5.628761529922485,
      "epoch": 2.348650258038904,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004450736032866377,
      "loss": 4.9916,
      "mean_token_accuracy": 0.18187623620033264,
      "num_tokens": 51847501.0,
      "step": 23665
    },
    {
      "entropy": 5.616382598876953,
      "epoch": 2.3491464867010716,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044505040392634447,
      "loss": 5.0353,
      "mean_token_accuracy": 0.18370021730661393,
      "num_tokens": 51858056.0,
      "step": 23670
    },
    {
      "entropy": 5.659100866317749,
      "epoch": 2.3496427153632395,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004450272003491677,
      "loss": 5.0114,
      "mean_token_accuracy": 0.1857176274061203,
      "num_tokens": 51869248.0,
      "step": 23675
    },
    {
      "entropy": 5.616753816604614,
      "epoch": 2.350138944025407,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00044500399255568274,
      "loss": 5.1019,
      "mean_token_accuracy": 0.17216083854436875,
      "num_tokens": 51881496.0,
      "step": 23680
    },
    {
      "entropy": 5.607989740371704,
      "epoch": 2.3506351726875745,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044498078054646507,
      "loss": 5.0467,
      "mean_token_accuracy": 0.17704195082187651,
      "num_tokens": 51891851.0,
      "step": 23685
    },
    {
      "entropy": 5.667589473724365,
      "epoch": 2.351131401349742,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044495756432209025,
      "loss": 5.1251,
      "mean_token_accuracy": 0.1776928797364235,
      "num_tokens": 51903581.0,
      "step": 23690
    },
    {
      "entropy": 5.602322769165039,
      "epoch": 2.3516276300119094,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044493434388313423,
      "loss": 4.9714,
      "mean_token_accuracy": 0.18691786080598832,
      "num_tokens": 51913650.0,
      "step": 23695
    },
    {
      "entropy": 5.560722398757934,
      "epoch": 2.352123858674077,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044491111923017254,
      "loss": 4.9801,
      "mean_token_accuracy": 0.18390494883060454,
      "num_tokens": 51924704.0,
      "step": 23700
    },
    {
      "entropy": 5.712212705612183,
      "epoch": 2.3526200873362444,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044488789036378117,
      "loss": 5.1216,
      "mean_token_accuracy": 0.1792377859354019,
      "num_tokens": 51937576.0,
      "step": 23705
    },
    {
      "entropy": 5.646164321899414,
      "epoch": 2.3531163159984123,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004448646572845363,
      "loss": 4.9521,
      "mean_token_accuracy": 0.1847248524427414,
      "num_tokens": 51947359.0,
      "step": 23710
    },
    {
      "entropy": 5.649265193939209,
      "epoch": 2.35361254466058,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004448414199930139,
      "loss": 5.103,
      "mean_token_accuracy": 0.17193868905305862,
      "num_tokens": 51958016.0,
      "step": 23715
    },
    {
      "entropy": 5.679453420639038,
      "epoch": 2.3541087733227473,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044481817848979026,
      "loss": 5.0823,
      "mean_token_accuracy": 0.17544051706790925,
      "num_tokens": 51968770.0,
      "step": 23720
    },
    {
      "entropy": 5.610362243652344,
      "epoch": 2.3546050019849147,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004447949327754418,
      "loss": 4.9642,
      "mean_token_accuracy": 0.18605738282203674,
      "num_tokens": 51980052.0,
      "step": 23725
    },
    {
      "entropy": 5.65678653717041,
      "epoch": 2.355101230647082,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004447716828505447,
      "loss": 5.1359,
      "mean_token_accuracy": 0.17561834901571274,
      "num_tokens": 51992080.0,
      "step": 23730
    },
    {
      "entropy": 5.614643192291259,
      "epoch": 2.3555974593092497,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004447484287156758,
      "loss": 5.0598,
      "mean_token_accuracy": 0.1879371851682663,
      "num_tokens": 52004313.0,
      "step": 23735
    },
    {
      "entropy": 5.681997680664063,
      "epoch": 2.356093687971417,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004447251703714117,
      "loss": 5.0269,
      "mean_token_accuracy": 0.17926384508609772,
      "num_tokens": 52015207.0,
      "step": 23740
    },
    {
      "entropy": 5.660422134399414,
      "epoch": 2.3565899166335846,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000444701907818329,
      "loss": 5.0384,
      "mean_token_accuracy": 0.17645563185214996,
      "num_tokens": 52026565.0,
      "step": 23745
    },
    {
      "entropy": 5.659357786178589,
      "epoch": 2.357086145295752,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004446786410570048,
      "loss": 5.0727,
      "mean_token_accuracy": 0.17633912414312364,
      "num_tokens": 52037654.0,
      "step": 23750
    },
    {
      "entropy": 5.5913492202758786,
      "epoch": 2.3575823739579196,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044465537008801575,
      "loss": 4.9717,
      "mean_token_accuracy": 0.17990387827157975,
      "num_tokens": 52049772.0,
      "step": 23755
    },
    {
      "entropy": 5.619614696502685,
      "epoch": 2.3580786026200875,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004446320949119392,
      "loss": 5.1069,
      "mean_token_accuracy": 0.17624690979719163,
      "num_tokens": 52061363.0,
      "step": 23760
    },
    {
      "entropy": 5.627612161636352,
      "epoch": 2.358574831282255,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044460881552935215,
      "loss": 4.9713,
      "mean_token_accuracy": 0.18468080461025238,
      "num_tokens": 52073153.0,
      "step": 23765
    },
    {
      "entropy": 5.585729122161865,
      "epoch": 2.3590710599444225,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044458553194083204,
      "loss": 5.0105,
      "mean_token_accuracy": 0.18418016880750657,
      "num_tokens": 52083369.0,
      "step": 23770
    },
    {
      "entropy": 5.560564708709717,
      "epoch": 2.35956728860659,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004445622441469562,
      "loss": 4.9597,
      "mean_token_accuracy": 0.18837179690599443,
      "num_tokens": 52093701.0,
      "step": 23775
    },
    {
      "entropy": 5.631415462493896,
      "epoch": 2.3600635172687574,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044453895214830195,
      "loss": 5.0184,
      "mean_token_accuracy": 0.18267978578805924,
      "num_tokens": 52105863.0,
      "step": 23780
    },
    {
      "entropy": 5.5729203701019285,
      "epoch": 2.360559745930925,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004445156559454471,
      "loss": 5.0037,
      "mean_token_accuracy": 0.18487821221351625,
      "num_tokens": 52116476.0,
      "step": 23785
    },
    {
      "entropy": 5.665838098526001,
      "epoch": 2.3610559745930924,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044449235553896914,
      "loss": 5.1244,
      "mean_token_accuracy": 0.17285252064466478,
      "num_tokens": 52127262.0,
      "step": 23790
    },
    {
      "entropy": 5.711639022827148,
      "epoch": 2.36155220325526,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004444690509294461,
      "loss": 5.0813,
      "mean_token_accuracy": 0.1806657522916794,
      "num_tokens": 52138575.0,
      "step": 23795
    },
    {
      "entropy": 5.631810712814331,
      "epoch": 2.3620484319174277,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004444457421174558,
      "loss": 5.0258,
      "mean_token_accuracy": 0.18528532534837722,
      "num_tokens": 52149620.0,
      "step": 23800
    },
    {
      "entropy": 5.662083768844605,
      "epoch": 2.3625446605795952,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004444224291035761,
      "loss": 5.0717,
      "mean_token_accuracy": 0.17842390090227128,
      "num_tokens": 52161929.0,
      "step": 23805
    },
    {
      "entropy": 5.5613486766815186,
      "epoch": 2.3630408892417627,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044439911188838533,
      "loss": 4.9909,
      "mean_token_accuracy": 0.19203223437070846,
      "num_tokens": 52172283.0,
      "step": 23810
    },
    {
      "entropy": 5.609691095352173,
      "epoch": 2.36353711790393,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004443757904724615,
      "loss": 5.0737,
      "mean_token_accuracy": 0.17978875786066056,
      "num_tokens": 52184130.0,
      "step": 23815
    },
    {
      "entropy": 5.616478824615479,
      "epoch": 2.3640333465660976,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00044435246485638317,
      "loss": 5.0749,
      "mean_token_accuracy": 0.18615096360445021,
      "num_tokens": 52195358.0,
      "step": 23820
    },
    {
      "entropy": 5.616151571273804,
      "epoch": 2.364529575228265,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004443291350407285,
      "loss": 4.9823,
      "mean_token_accuracy": 0.18327332735061647,
      "num_tokens": 52207070.0,
      "step": 23825
    },
    {
      "entropy": 5.605486583709717,
      "epoch": 2.3650258038904326,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004443058010260762,
      "loss": 5.1028,
      "mean_token_accuracy": 0.17908568680286407,
      "num_tokens": 52217658.0,
      "step": 23830
    },
    {
      "entropy": 5.591069459915161,
      "epoch": 2.3655220325526,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044428246281300493,
      "loss": 5.0023,
      "mean_token_accuracy": 0.18368549197912215,
      "num_tokens": 52229350.0,
      "step": 23835
    },
    {
      "entropy": 5.694539213180542,
      "epoch": 2.3660182612147675,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044425912040209326,
      "loss": 5.162,
      "mean_token_accuracy": 0.17321085631847383,
      "num_tokens": 52241260.0,
      "step": 23840
    },
    {
      "entropy": 5.710390329360962,
      "epoch": 2.3665144898769355,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044423577379392023,
      "loss": 5.0646,
      "mean_token_accuracy": 0.1779045656323433,
      "num_tokens": 52253049.0,
      "step": 23845
    },
    {
      "entropy": 5.644241714477539,
      "epoch": 2.367010718539103,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004442124229890646,
      "loss": 5.0424,
      "mean_token_accuracy": 0.17962158620357513,
      "num_tokens": 52263405.0,
      "step": 23850
    },
    {
      "entropy": 5.638300132751465,
      "epoch": 2.3675069472012704,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044418906798810554,
      "loss": 5.0529,
      "mean_token_accuracy": 0.18033716678619385,
      "num_tokens": 52273075.0,
      "step": 23855
    },
    {
      "entropy": 5.741637229919434,
      "epoch": 2.368003175863438,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00044416570879162213,
      "loss": 5.2346,
      "mean_token_accuracy": 0.17287510335445405,
      "num_tokens": 52284275.0,
      "step": 23860
    },
    {
      "entropy": 5.609548330307007,
      "epoch": 2.3684994045256054,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004441423454001936,
      "loss": 5.0383,
      "mean_token_accuracy": 0.17191773802042007,
      "num_tokens": 52295084.0,
      "step": 23865
    },
    {
      "entropy": 5.621134996414185,
      "epoch": 2.368995633187773,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044411897781439946,
      "loss": 5.0307,
      "mean_token_accuracy": 0.18354223519563675,
      "num_tokens": 52304172.0,
      "step": 23870
    },
    {
      "entropy": 5.562858533859253,
      "epoch": 2.3694918618499403,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004440956060348191,
      "loss": 4.9739,
      "mean_token_accuracy": 0.18783491104841232,
      "num_tokens": 52315370.0,
      "step": 23875
    },
    {
      "entropy": 5.622923469543457,
      "epoch": 2.369988090512108,
      "grad_norm": 1.0,
      "learning_rate": 0.0004440722300620321,
      "loss": 5.0233,
      "mean_token_accuracy": 0.18388883620500565,
      "num_tokens": 52326517.0,
      "step": 23880
    },
    {
      "entropy": 5.653690910339355,
      "epoch": 2.3704843191742757,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044404884989661804,
      "loss": 5.0809,
      "mean_token_accuracy": 0.18292949497699737,
      "num_tokens": 52337110.0,
      "step": 23885
    },
    {
      "entropy": 5.619952726364136,
      "epoch": 2.370980547836443,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004440254655391568,
      "loss": 5.0067,
      "mean_token_accuracy": 0.1886501207947731,
      "num_tokens": 52347509.0,
      "step": 23890
    },
    {
      "entropy": 5.705073547363281,
      "epoch": 2.3714767764986107,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004440020769902282,
      "loss": 5.1391,
      "mean_token_accuracy": 0.1757641926407814,
      "num_tokens": 52357920.0,
      "step": 23895
    },
    {
      "entropy": 5.6615378856658936,
      "epoch": 2.371973005160778,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044397868425041234,
      "loss": 5.1078,
      "mean_token_accuracy": 0.1766369163990021,
      "num_tokens": 52368138.0,
      "step": 23900
    },
    {
      "entropy": 5.7159546375274655,
      "epoch": 2.3724692338229456,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004439552873202891,
      "loss": 5.0625,
      "mean_token_accuracy": 0.18050333559513093,
      "num_tokens": 52380503.0,
      "step": 23905
    },
    {
      "entropy": 5.645144033432007,
      "epoch": 2.372965462485113,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044393188620043893,
      "loss": 5.0471,
      "mean_token_accuracy": 0.1841754212975502,
      "num_tokens": 52390850.0,
      "step": 23910
    },
    {
      "entropy": 5.565052652359009,
      "epoch": 2.3734616911472806,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004439084808914419,
      "loss": 4.9809,
      "mean_token_accuracy": 0.18963249772787094,
      "num_tokens": 52400829.0,
      "step": 23915
    },
    {
      "entropy": 5.684382963180542,
      "epoch": 2.373957919809448,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044388507139387864,
      "loss": 5.079,
      "mean_token_accuracy": 0.18084112405776978,
      "num_tokens": 52412892.0,
      "step": 23920
    },
    {
      "entropy": 5.648769950866699,
      "epoch": 2.3744541484716155,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004438616577083293,
      "loss": 5.0147,
      "mean_token_accuracy": 0.18247147798538207,
      "num_tokens": 52423662.0,
      "step": 23925
    },
    {
      "entropy": 5.663415956497192,
      "epoch": 2.3749503771337834,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000443838239835375,
      "loss": 5.0575,
      "mean_token_accuracy": 0.1794498607516289,
      "num_tokens": 52433541.0,
      "step": 23930
    },
    {
      "entropy": 5.653537750244141,
      "epoch": 2.375446605795951,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004438148177755959,
      "loss": 5.1035,
      "mean_token_accuracy": 0.17571025788784028,
      "num_tokens": 52445504.0,
      "step": 23935
    },
    {
      "entropy": 5.672600364685058,
      "epoch": 2.3759428344581184,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044379139152957327,
      "loss": 5.0641,
      "mean_token_accuracy": 0.17595665007829667,
      "num_tokens": 52456116.0,
      "step": 23940
    },
    {
      "entropy": 5.6611363887786865,
      "epoch": 2.376439063120286,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044376796109788777,
      "loss": 4.9378,
      "mean_token_accuracy": 0.18862972557544708,
      "num_tokens": 52465462.0,
      "step": 23945
    },
    {
      "entropy": 5.634843397140503,
      "epoch": 2.3769352917824533,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044374452648112046,
      "loss": 5.0889,
      "mean_token_accuracy": 0.17624662965536117,
      "num_tokens": 52477213.0,
      "step": 23950
    },
    {
      "entropy": 5.659652423858643,
      "epoch": 2.377431520444621,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044372108767985253,
      "loss": 5.086,
      "mean_token_accuracy": 0.17638906389474868,
      "num_tokens": 52488095.0,
      "step": 23955
    },
    {
      "entropy": 5.696300792694092,
      "epoch": 2.3779277491067883,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004436976446946652,
      "loss": 5.1444,
      "mean_token_accuracy": 0.1716751828789711,
      "num_tokens": 52499078.0,
      "step": 23960
    },
    {
      "entropy": 5.66860785484314,
      "epoch": 2.3784239777689558,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044367419752613983,
      "loss": 5.0338,
      "mean_token_accuracy": 0.17939472049474717,
      "num_tokens": 52510829.0,
      "step": 23965
    },
    {
      "entropy": 5.683780479431152,
      "epoch": 2.3789202064311237,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004436507461748577,
      "loss": 5.0279,
      "mean_token_accuracy": 0.1852037325501442,
      "num_tokens": 52521466.0,
      "step": 23970
    },
    {
      "entropy": 5.66539568901062,
      "epoch": 2.379416435093291,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004436272906414007,
      "loss": 4.9643,
      "mean_token_accuracy": 0.19241318106651306,
      "num_tokens": 52532617.0,
      "step": 23975
    },
    {
      "entropy": 5.560275077819824,
      "epoch": 2.3799126637554586,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004436038309263501,
      "loss": 5.0277,
      "mean_token_accuracy": 0.18164770007133485,
      "num_tokens": 52543826.0,
      "step": 23980
    },
    {
      "entropy": 5.598338031768799,
      "epoch": 2.380408892417626,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044358036703028773,
      "loss": 5.0539,
      "mean_token_accuracy": 0.18018202781677245,
      "num_tokens": 52556752.0,
      "step": 23985
    },
    {
      "entropy": 5.664846992492675,
      "epoch": 2.3809051210797936,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00044355689895379563,
      "loss": 5.1245,
      "mean_token_accuracy": 0.1730619788169861,
      "num_tokens": 52567268.0,
      "step": 23990
    },
    {
      "entropy": 5.651590061187744,
      "epoch": 2.381401349741961,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044353342669745566,
      "loss": 5.0545,
      "mean_token_accuracy": 0.18406080305576325,
      "num_tokens": 52576894.0,
      "step": 23995
    },
    {
      "entropy": 5.592921209335327,
      "epoch": 2.3818975784041285,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044350995026184976,
      "loss": 5.0396,
      "mean_token_accuracy": 0.18117541074752808,
      "num_tokens": 52587434.0,
      "step": 24000
    },
    {
      "epoch": 2.3818975784041285,
      "eval_entropy": 5.422252529050936,
      "eval_loss": 5.162605285644531,
      "eval_mean_token_accuracy": 0.1806545086863503,
      "eval_num_tokens": 52587434.0,
      "eval_runtime": 20.7243,
      "eval_samples_per_second": 1624.615,
      "eval_steps_per_second": 203.095,
      "step": 24000
    },
    {
      "entropy": 5.687662363052368,
      "epoch": 2.382393807066296,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044348646964756036,
      "loss": 5.1065,
      "mean_token_accuracy": 0.17620053440332412,
      "num_tokens": 52599455.0,
      "step": 24005
    },
    {
      "entropy": 5.688921976089477,
      "epoch": 2.3828900357284635,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004434629848551694,
      "loss": 5.0924,
      "mean_token_accuracy": 0.1784089043736458,
      "num_tokens": 52611020.0,
      "step": 24010
    },
    {
      "entropy": 5.616387891769409,
      "epoch": 2.3833862643906314,
      "grad_norm": 1.0,
      "learning_rate": 0.0004434394958852595,
      "loss": 5.041,
      "mean_token_accuracy": 0.1830443635582924,
      "num_tokens": 52621743.0,
      "step": 24015
    },
    {
      "entropy": 5.637858581542969,
      "epoch": 2.383882493052799,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044341600273841304,
      "loss": 5.0974,
      "mean_token_accuracy": 0.18215125501155854,
      "num_tokens": 52633437.0,
      "step": 24020
    },
    {
      "entropy": 5.607716035842896,
      "epoch": 2.3843787217149663,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044339250541521264,
      "loss": 4.997,
      "mean_token_accuracy": 0.18692731410264968,
      "num_tokens": 52643849.0,
      "step": 24025
    },
    {
      "entropy": 5.771730852127075,
      "epoch": 2.384874950377134,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044336900391624096,
      "loss": 5.1692,
      "mean_token_accuracy": 0.1719026580452919,
      "num_tokens": 52656417.0,
      "step": 24030
    },
    {
      "entropy": 5.614979457855225,
      "epoch": 2.3853711790393013,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004433454982420808,
      "loss": 4.9511,
      "mean_token_accuracy": 0.190848645567894,
      "num_tokens": 52668491.0,
      "step": 24035
    },
    {
      "entropy": 5.620504236221313,
      "epoch": 2.3858674077014688,
      "grad_norm": 1.15625,
      "learning_rate": 0.000443321988393315,
      "loss": 5.0958,
      "mean_token_accuracy": 0.176480633020401,
      "num_tokens": 52679532.0,
      "step": 24040
    },
    {
      "entropy": 5.602590131759643,
      "epoch": 2.3863636363636362,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004432984743705265,
      "loss": 4.9803,
      "mean_token_accuracy": 0.180617618560791,
      "num_tokens": 52689850.0,
      "step": 24045
    },
    {
      "entropy": 5.553384447097779,
      "epoch": 2.3868598650258037,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044327495617429866,
      "loss": 4.964,
      "mean_token_accuracy": 0.18532002866268157,
      "num_tokens": 52699529.0,
      "step": 24050
    },
    {
      "entropy": 5.545459222793579,
      "epoch": 2.3873560936879716,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004432514338052144,
      "loss": 4.9766,
      "mean_token_accuracy": 0.18824486285448075,
      "num_tokens": 52711449.0,
      "step": 24055
    },
    {
      "entropy": 5.706513071060181,
      "epoch": 2.387852322350139,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004432279072638571,
      "loss": 5.1114,
      "mean_token_accuracy": 0.1720067411661148,
      "num_tokens": 52721528.0,
      "step": 24060
    },
    {
      "entropy": 5.649973297119141,
      "epoch": 2.3883485510123066,
      "grad_norm": 1.65625,
      "learning_rate": 0.00044320437655081033,
      "loss": 5.0085,
      "mean_token_accuracy": 0.18264384120702742,
      "num_tokens": 52731367.0,
      "step": 24065
    },
    {
      "entropy": 5.568722915649414,
      "epoch": 2.388844779674474,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044318084166665737,
      "loss": 4.926,
      "mean_token_accuracy": 0.19135257750749587,
      "num_tokens": 52740822.0,
      "step": 24070
    },
    {
      "entropy": 5.540734672546387,
      "epoch": 2.3893410083366415,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000443157302611982,
      "loss": 5.023,
      "mean_token_accuracy": 0.1764022871851921,
      "num_tokens": 52751609.0,
      "step": 24075
    },
    {
      "entropy": 5.65751633644104,
      "epoch": 2.389837236998809,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004431337593873678,
      "loss": 5.0667,
      "mean_token_accuracy": 0.17780413776636123,
      "num_tokens": 52762914.0,
      "step": 24080
    },
    {
      "entropy": 5.749363899230957,
      "epoch": 2.3903334656609765,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004431102119933987,
      "loss": 5.1001,
      "mean_token_accuracy": 0.1793038383126259,
      "num_tokens": 52773894.0,
      "step": 24085
    },
    {
      "entropy": 5.603061723709106,
      "epoch": 2.390829694323144,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004430866604306586,
      "loss": 5.0378,
      "mean_token_accuracy": 0.1826264202594757,
      "num_tokens": 52784606.0,
      "step": 24090
    },
    {
      "entropy": 5.600283288955689,
      "epoch": 2.3913259229853114,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044306310469973147,
      "loss": 5.0199,
      "mean_token_accuracy": 0.17831894010305405,
      "num_tokens": 52795275.0,
      "step": 24095
    },
    {
      "entropy": 5.676590394973755,
      "epoch": 2.391822151647479,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044303954480120155,
      "loss": 5.1002,
      "mean_token_accuracy": 0.18175278306007386,
      "num_tokens": 52805058.0,
      "step": 24100
    },
    {
      "entropy": 5.668203353881836,
      "epoch": 2.392318380309647,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044301598073565295,
      "loss": 5.0189,
      "mean_token_accuracy": 0.17878224551677704,
      "num_tokens": 52815486.0,
      "step": 24105
    },
    {
      "entropy": 5.622556543350219,
      "epoch": 2.3928146089718143,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004429924125036702,
      "loss": 5.0443,
      "mean_token_accuracy": 0.18218316435813903,
      "num_tokens": 52827039.0,
      "step": 24110
    },
    {
      "entropy": 5.601401329040527,
      "epoch": 2.393310837633982,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004429688401058375,
      "loss": 5.0475,
      "mean_token_accuracy": 0.18568631112575532,
      "num_tokens": 52837712.0,
      "step": 24115
    },
    {
      "entropy": 5.5970926761627195,
      "epoch": 2.3938070662961493,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004429452635427394,
      "loss": 4.9403,
      "mean_token_accuracy": 0.18674813359975814,
      "num_tokens": 52846321.0,
      "step": 24120
    },
    {
      "entropy": 5.628134918212891,
      "epoch": 2.3943032949583167,
      "grad_norm": 1.125,
      "learning_rate": 0.00044292168281496086,
      "loss": 4.9956,
      "mean_token_accuracy": 0.18572134971618653,
      "num_tokens": 52856342.0,
      "step": 24125
    },
    {
      "entropy": 5.639142274856567,
      "epoch": 2.394799523620484,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004428980979230863,
      "loss": 5.1337,
      "mean_token_accuracy": 0.17736416012048722,
      "num_tokens": 52866602.0,
      "step": 24130
    },
    {
      "entropy": 5.585164356231689,
      "epoch": 2.3952957522826517,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00044287450886770075,
      "loss": 4.9373,
      "mean_token_accuracy": 0.18733307868242263,
      "num_tokens": 52877066.0,
      "step": 24135
    },
    {
      "entropy": 5.574439430236817,
      "epoch": 2.3957919809448196,
      "grad_norm": 1.125,
      "learning_rate": 0.0004428509156493891,
      "loss": 4.9667,
      "mean_token_accuracy": 0.19109086990356444,
      "num_tokens": 52887907.0,
      "step": 24140
    },
    {
      "entropy": 5.699541091918945,
      "epoch": 2.396288209606987,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004428273182687364,
      "loss": 5.1381,
      "mean_token_accuracy": 0.17955077141523362,
      "num_tokens": 52899308.0,
      "step": 24145
    },
    {
      "entropy": 5.606731176376343,
      "epoch": 2.3967844382691545,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004428037167263279,
      "loss": 4.9568,
      "mean_token_accuracy": 0.18767286092042923,
      "num_tokens": 52909127.0,
      "step": 24150
    },
    {
      "entropy": 5.5957887172698975,
      "epoch": 2.397280666931322,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044278011102274887,
      "loss": 4.9799,
      "mean_token_accuracy": 0.18623005747795104,
      "num_tokens": 52920503.0,
      "step": 24155
    },
    {
      "entropy": 5.609165668487549,
      "epoch": 2.3977768955934895,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044275650115858445,
      "loss": 5.0018,
      "mean_token_accuracy": 0.180817811191082,
      "num_tokens": 52931904.0,
      "step": 24160
    },
    {
      "entropy": 5.662545871734619,
      "epoch": 2.398273124255657,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004427328871344204,
      "loss": 5.1516,
      "mean_token_accuracy": 0.1678496390581131,
      "num_tokens": 52945101.0,
      "step": 24165
    },
    {
      "entropy": 5.624108362197876,
      "epoch": 2.3987693529178244,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004427092689508422,
      "loss": 5.0415,
      "mean_token_accuracy": 0.18428945541381836,
      "num_tokens": 52955999.0,
      "step": 24170
    },
    {
      "entropy": 5.596385431289673,
      "epoch": 2.399265581579992,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004426856466084354,
      "loss": 5.0508,
      "mean_token_accuracy": 0.18043915033340455,
      "num_tokens": 52967273.0,
      "step": 24175
    },
    {
      "entropy": 5.693435382843018,
      "epoch": 2.3997618102421594,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000442662020107786,
      "loss": 5.1023,
      "mean_token_accuracy": 0.18223486244678497,
      "num_tokens": 52977415.0,
      "step": 24180
    },
    {
      "entropy": 5.622218465805053,
      "epoch": 2.400258038904327,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004426383894494797,
      "loss": 5.0076,
      "mean_token_accuracy": 0.18533975630998611,
      "num_tokens": 52990071.0,
      "step": 24185
    },
    {
      "entropy": 5.566663837432861,
      "epoch": 2.400754267566495,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004426147546341027,
      "loss": 4.9661,
      "mean_token_accuracy": 0.1810623288154602,
      "num_tokens": 52999874.0,
      "step": 24190
    },
    {
      "entropy": 5.637349843978882,
      "epoch": 2.4012504962286623,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044259111566224084,
      "loss": 5.0301,
      "mean_token_accuracy": 0.17881642431020736,
      "num_tokens": 53011349.0,
      "step": 24195
    },
    {
      "entropy": 5.590851974487305,
      "epoch": 2.4017467248908297,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044256747253448045,
      "loss": 4.9773,
      "mean_token_accuracy": 0.18969312012195588,
      "num_tokens": 53024256.0,
      "step": 24200
    },
    {
      "entropy": 5.595731973648071,
      "epoch": 2.402242953552997,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044254382525140784,
      "loss": 4.9509,
      "mean_token_accuracy": 0.1940011128783226,
      "num_tokens": 53035617.0,
      "step": 24205
    },
    {
      "entropy": 5.612962532043457,
      "epoch": 2.4027391822151647,
      "grad_norm": 1.125,
      "learning_rate": 0.00044252017381360935,
      "loss": 4.9889,
      "mean_token_accuracy": 0.1886721134185791,
      "num_tokens": 53046219.0,
      "step": 24210
    },
    {
      "entropy": 5.7079229831695555,
      "epoch": 2.403235410877332,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004424965182216716,
      "loss": 5.1082,
      "mean_token_accuracy": 0.17123801857233048,
      "num_tokens": 53057114.0,
      "step": 24215
    },
    {
      "entropy": 5.683101844787598,
      "epoch": 2.4037316395394996,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044247285847618105,
      "loss": 5.0417,
      "mean_token_accuracy": 0.18267682939767838,
      "num_tokens": 53067481.0,
      "step": 24220
    },
    {
      "entropy": 5.6657445430755615,
      "epoch": 2.4042278682016676,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004424491945777245,
      "loss": 5.1318,
      "mean_token_accuracy": 0.17881065607070923,
      "num_tokens": 53078129.0,
      "step": 24225
    },
    {
      "entropy": 5.612861919403076,
      "epoch": 2.404724096863835,
      "grad_norm": 1.125,
      "learning_rate": 0.00044242552652688885,
      "loss": 4.97,
      "mean_token_accuracy": 0.1868362694978714,
      "num_tokens": 53088919.0,
      "step": 24230
    },
    {
      "entropy": 5.647245025634765,
      "epoch": 2.4052203255260025,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004424018543242608,
      "loss": 5.0535,
      "mean_token_accuracy": 0.18590764552354813,
      "num_tokens": 53099716.0,
      "step": 24235
    },
    {
      "entropy": 5.5834312915802,
      "epoch": 2.40571655418817,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004423781779704275,
      "loss": 4.9648,
      "mean_token_accuracy": 0.1903766229748726,
      "num_tokens": 53108798.0,
      "step": 24240
    },
    {
      "entropy": 5.6393499851226805,
      "epoch": 2.4062127828503375,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004423544974659761,
      "loss": 5.0747,
      "mean_token_accuracy": 0.1762581393122673,
      "num_tokens": 53120655.0,
      "step": 24245
    },
    {
      "entropy": 5.603771352767945,
      "epoch": 2.406709011512505,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044233081281149373,
      "loss": 4.9183,
      "mean_token_accuracy": 0.18559200763702394,
      "num_tokens": 53131760.0,
      "step": 24250
    },
    {
      "entropy": 5.641468238830567,
      "epoch": 2.4072052401746724,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044230712400756774,
      "loss": 5.0049,
      "mean_token_accuracy": 0.18340890258550643,
      "num_tokens": 53142347.0,
      "step": 24255
    },
    {
      "entropy": 5.607591342926026,
      "epoch": 2.40770146883684,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044228343105478564,
      "loss": 5.0293,
      "mean_token_accuracy": 0.18346780389547349,
      "num_tokens": 53153678.0,
      "step": 24260
    },
    {
      "entropy": 5.630926036834717,
      "epoch": 2.4081976974990074,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004422597339537349,
      "loss": 5.0625,
      "mean_token_accuracy": 0.18351869732141496,
      "num_tokens": 53165089.0,
      "step": 24265
    },
    {
      "entropy": 5.637145709991455,
      "epoch": 2.408693926161175,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004422360327050032,
      "loss": 5.0015,
      "mean_token_accuracy": 0.1781063571572304,
      "num_tokens": 53175614.0,
      "step": 24270
    },
    {
      "entropy": 5.663777303695679,
      "epoch": 2.4091901548233428,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044221232730917823,
      "loss": 4.9982,
      "mean_token_accuracy": 0.18667846769094468,
      "num_tokens": 53186306.0,
      "step": 24275
    },
    {
      "entropy": 5.612605142593384,
      "epoch": 2.4096863834855102,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044218861776684776,
      "loss": 5.0555,
      "mean_token_accuracy": 0.17905541956424714,
      "num_tokens": 53197730.0,
      "step": 24280
    },
    {
      "entropy": 5.570061874389649,
      "epoch": 2.4101826121476777,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044216490407859995,
      "loss": 4.9811,
      "mean_token_accuracy": 0.18358394652605056,
      "num_tokens": 53208486.0,
      "step": 24285
    },
    {
      "entropy": 5.669879341125489,
      "epoch": 2.410678840809845,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044214118624502267,
      "loss": 5.0562,
      "mean_token_accuracy": 0.18035977184772492,
      "num_tokens": 53220045.0,
      "step": 24290
    },
    {
      "entropy": 5.685708904266358,
      "epoch": 2.4111750694720127,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004421174642667042,
      "loss": 5.1251,
      "mean_token_accuracy": 0.17780056446790696,
      "num_tokens": 53231385.0,
      "step": 24295
    },
    {
      "entropy": 5.684106492996216,
      "epoch": 2.41167129813418,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044209373814423265,
      "loss": 5.097,
      "mean_token_accuracy": 0.1774526506662369,
      "num_tokens": 53243038.0,
      "step": 24300
    },
    {
      "entropy": 5.673605060577392,
      "epoch": 2.4121675267963476,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004420700078781965,
      "loss": 5.0727,
      "mean_token_accuracy": 0.17828863710165024,
      "num_tokens": 53252959.0,
      "step": 24305
    },
    {
      "entropy": 5.553832340240478,
      "epoch": 2.4126637554585155,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044204627346918407,
      "loss": 4.9873,
      "mean_token_accuracy": 0.1849662646651268,
      "num_tokens": 53263542.0,
      "step": 24310
    },
    {
      "entropy": 5.619505834579468,
      "epoch": 2.413159984120683,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044202253491778404,
      "loss": 4.9962,
      "mean_token_accuracy": 0.18811645358800888,
      "num_tokens": 53273511.0,
      "step": 24315
    },
    {
      "entropy": 5.663531732559204,
      "epoch": 2.4136562127828505,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004419987922245852,
      "loss": 5.1141,
      "mean_token_accuracy": 0.1803983271121979,
      "num_tokens": 53286173.0,
      "step": 24320
    },
    {
      "entropy": 5.633026885986328,
      "epoch": 2.414152441445018,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000441975045390176,
      "loss": 5.0098,
      "mean_token_accuracy": 0.18659839481115342,
      "num_tokens": 53296784.0,
      "step": 24325
    },
    {
      "entropy": 5.685882472991944,
      "epoch": 2.4146486701071854,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044195129441514547,
      "loss": 5.0401,
      "mean_token_accuracy": 0.186306868493557,
      "num_tokens": 53308044.0,
      "step": 24330
    },
    {
      "entropy": 5.583441114425659,
      "epoch": 2.415144898769353,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044192753930008267,
      "loss": 4.9867,
      "mean_token_accuracy": 0.18672547489404678,
      "num_tokens": 53319161.0,
      "step": 24335
    },
    {
      "entropy": 5.554642105102539,
      "epoch": 2.4156411274315204,
      "grad_norm": 1.0,
      "learning_rate": 0.0004419037800455765,
      "loss": 4.9563,
      "mean_token_accuracy": 0.1916997402906418,
      "num_tokens": 53329766.0,
      "step": 24340
    },
    {
      "entropy": 5.659880542755127,
      "epoch": 2.416137356093688,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004418800166522163,
      "loss": 5.0562,
      "mean_token_accuracy": 0.182161782681942,
      "num_tokens": 53340329.0,
      "step": 24345
    },
    {
      "entropy": 5.668436670303345,
      "epoch": 2.4166335847558553,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044185624912059125,
      "loss": 5.0622,
      "mean_token_accuracy": 0.18430643379688263,
      "num_tokens": 53352592.0,
      "step": 24350
    },
    {
      "entropy": 5.707031393051148,
      "epoch": 2.417129813418023,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004418324774512908,
      "loss": 5.0293,
      "mean_token_accuracy": 0.177917018532753,
      "num_tokens": 53362877.0,
      "step": 24355
    },
    {
      "entropy": 5.622635555267334,
      "epoch": 2.4176260420801907,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004418087016449043,
      "loss": 5.0371,
      "mean_token_accuracy": 0.1846744731068611,
      "num_tokens": 53374612.0,
      "step": 24360
    },
    {
      "entropy": 5.634442758560181,
      "epoch": 2.418122270742358,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044178492170202153,
      "loss": 5.0635,
      "mean_token_accuracy": 0.17887624651193618,
      "num_tokens": 53385129.0,
      "step": 24365
    },
    {
      "entropy": 5.678120040893555,
      "epoch": 2.4186184994045257,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044176113762323213,
      "loss": 5.1463,
      "mean_token_accuracy": 0.17168127447366716,
      "num_tokens": 53395403.0,
      "step": 24370
    },
    {
      "entropy": 5.694255352020264,
      "epoch": 2.419114728066693,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044173734940912586,
      "loss": 5.1099,
      "mean_token_accuracy": 0.1671777606010437,
      "num_tokens": 53406684.0,
      "step": 24375
    },
    {
      "entropy": 5.607424640655518,
      "epoch": 2.4196109567288606,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044171355706029247,
      "loss": 4.9684,
      "mean_token_accuracy": 0.18753084689378738,
      "num_tokens": 53417256.0,
      "step": 24380
    },
    {
      "entropy": 5.5564226627349855,
      "epoch": 2.420107185391028,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044168976057732225,
      "loss": 4.9763,
      "mean_token_accuracy": 0.18108198344707488,
      "num_tokens": 53428305.0,
      "step": 24385
    },
    {
      "entropy": 5.5676209926605225,
      "epoch": 2.4206034140531956,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004416659599608051,
      "loss": 5.0125,
      "mean_token_accuracy": 0.17957470268011094,
      "num_tokens": 53439301.0,
      "step": 24390
    },
    {
      "entropy": 5.690366268157959,
      "epoch": 2.4210996427153635,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004416421552113312,
      "loss": 5.1325,
      "mean_token_accuracy": 0.17358105331659318,
      "num_tokens": 53450478.0,
      "step": 24395
    },
    {
      "entropy": 5.670365047454834,
      "epoch": 2.421595871377531,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000441618346329491,
      "loss": 5.0694,
      "mean_token_accuracy": 0.17688017189502717,
      "num_tokens": 53461293.0,
      "step": 24400
    },
    {
      "entropy": 5.594392871856689,
      "epoch": 2.4220921000396984,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044159453331587493,
      "loss": 5.0312,
      "mean_token_accuracy": 0.18353135734796525,
      "num_tokens": 53472014.0,
      "step": 24405
    },
    {
      "entropy": 5.631378316879273,
      "epoch": 2.422588328701866,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044157071617107325,
      "loss": 5.0652,
      "mean_token_accuracy": 0.18191105127334595,
      "num_tokens": 53483010.0,
      "step": 24410
    },
    {
      "entropy": 5.657489442825318,
      "epoch": 2.4230845573640334,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044154689489567684,
      "loss": 5.0664,
      "mean_token_accuracy": 0.17824610471725463,
      "num_tokens": 53494151.0,
      "step": 24415
    },
    {
      "entropy": 5.629713726043701,
      "epoch": 2.423580786026201,
      "grad_norm": 1.125,
      "learning_rate": 0.00044152306949027625,
      "loss": 5.0501,
      "mean_token_accuracy": 0.1819209098815918,
      "num_tokens": 53504462.0,
      "step": 24420
    },
    {
      "entropy": 5.6335673332214355,
      "epoch": 2.4240770146883683,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044149923995546244,
      "loss": 5.0313,
      "mean_token_accuracy": 0.18352582752704621,
      "num_tokens": 53514896.0,
      "step": 24425
    },
    {
      "entropy": 5.603960084915161,
      "epoch": 2.424573243350536,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004414754062918262,
      "loss": 4.9884,
      "mean_token_accuracy": 0.18734004348516464,
      "num_tokens": 53524656.0,
      "step": 24430
    },
    {
      "entropy": 5.652584886550903,
      "epoch": 2.4250694720127033,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004414515684999587,
      "loss": 5.0466,
      "mean_token_accuracy": 0.1869993269443512,
      "num_tokens": 53534896.0,
      "step": 24435
    },
    {
      "entropy": 5.587097644805908,
      "epoch": 2.4255657006748708,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004414277265804508,
      "loss": 5.0088,
      "mean_token_accuracy": 0.18401836603879929,
      "num_tokens": 53546108.0,
      "step": 24440
    },
    {
      "entropy": 5.601393604278565,
      "epoch": 2.4260619293370387,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000441403880533894,
      "loss": 5.0021,
      "mean_token_accuracy": 0.17893760949373244,
      "num_tokens": 53557325.0,
      "step": 24445
    },
    {
      "entropy": 5.547101831436157,
      "epoch": 2.426558157999206,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044138003036087956,
      "loss": 4.9395,
      "mean_token_accuracy": 0.18937669992446898,
      "num_tokens": 53568737.0,
      "step": 24450
    },
    {
      "entropy": 5.671169185638428,
      "epoch": 2.4270543866613736,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004413561760619988,
      "loss": 5.0512,
      "mean_token_accuracy": 0.17771768420934678,
      "num_tokens": 53579312.0,
      "step": 24455
    },
    {
      "entropy": 5.593870639801025,
      "epoch": 2.427550615323541,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004413323176378434,
      "loss": 4.9526,
      "mean_token_accuracy": 0.18645201474428177,
      "num_tokens": 53590476.0,
      "step": 24460
    },
    {
      "entropy": 5.614878177642822,
      "epoch": 2.4280468439857086,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004413084550890049,
      "loss": 5.077,
      "mean_token_accuracy": 0.18065236359834672,
      "num_tokens": 53601284.0,
      "step": 24465
    },
    {
      "entropy": 5.62466287612915,
      "epoch": 2.428543072647876,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004412845884160751,
      "loss": 5.0591,
      "mean_token_accuracy": 0.17848360538482666,
      "num_tokens": 53612590.0,
      "step": 24470
    },
    {
      "entropy": 5.661375522613525,
      "epoch": 2.4290393013100435,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004412607176196458,
      "loss": 5.0887,
      "mean_token_accuracy": 0.1732784926891327,
      "num_tokens": 53623049.0,
      "step": 24475
    },
    {
      "entropy": 5.622016286849975,
      "epoch": 2.429535529972211,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044123684270030907,
      "loss": 5.0127,
      "mean_token_accuracy": 0.1863011747598648,
      "num_tokens": 53633961.0,
      "step": 24480
    },
    {
      "entropy": 5.618499612808227,
      "epoch": 2.430031758634379,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044121296365865674,
      "loss": 5.0582,
      "mean_token_accuracy": 0.1805186986923218,
      "num_tokens": 53644634.0,
      "step": 24485
    },
    {
      "entropy": 5.606405973434448,
      "epoch": 2.4305279872965464,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044118908049528105,
      "loss": 4.9942,
      "mean_token_accuracy": 0.18781304061412812,
      "num_tokens": 53655982.0,
      "step": 24490
    },
    {
      "entropy": 5.701815986633301,
      "epoch": 2.431024215958714,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044116519321077435,
      "loss": 5.0939,
      "mean_token_accuracy": 0.17917912006378173,
      "num_tokens": 53667367.0,
      "step": 24495
    },
    {
      "entropy": 5.611343955993652,
      "epoch": 2.4315204446208813,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044114130180572883,
      "loss": 4.9788,
      "mean_token_accuracy": 0.18581620007753372,
      "num_tokens": 53677191.0,
      "step": 24500
    },
    {
      "entropy": 5.6021974086761475,
      "epoch": 2.432016673283049,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004411174062807371,
      "loss": 5.0599,
      "mean_token_accuracy": 0.18191857784986495,
      "num_tokens": 53688636.0,
      "step": 24505
    },
    {
      "entropy": 5.595519924163819,
      "epoch": 2.4325129019452163,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004410935066363916,
      "loss": 5.0124,
      "mean_token_accuracy": 0.1816207841038704,
      "num_tokens": 53699588.0,
      "step": 24510
    },
    {
      "entropy": 5.661431407928466,
      "epoch": 2.4330091306073838,
      "grad_norm": 1.0,
      "learning_rate": 0.00044106960287328503,
      "loss": 5.1322,
      "mean_token_accuracy": 0.16827367544174193,
      "num_tokens": 53710427.0,
      "step": 24515
    },
    {
      "entropy": 5.67356309890747,
      "epoch": 2.4335053592695512,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004410456949920102,
      "loss": 5.0879,
      "mean_token_accuracy": 0.1733045145869255,
      "num_tokens": 53722319.0,
      "step": 24520
    },
    {
      "entropy": 5.645653772354126,
      "epoch": 2.4340015879317187,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044102178299315994,
      "loss": 5.1028,
      "mean_token_accuracy": 0.17569446414709092,
      "num_tokens": 53734059.0,
      "step": 24525
    },
    {
      "entropy": 5.664698553085327,
      "epoch": 2.4344978165938866,
      "grad_norm": 1.125,
      "learning_rate": 0.00044099786687732713,
      "loss": 5.0581,
      "mean_token_accuracy": 0.1822950705885887,
      "num_tokens": 53745795.0,
      "step": 24530
    },
    {
      "entropy": 5.6539629936218265,
      "epoch": 2.434994045256054,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004409739466451049,
      "loss": 5.0755,
      "mean_token_accuracy": 0.17598870992660523,
      "num_tokens": 53756781.0,
      "step": 24535
    },
    {
      "entropy": 5.732645511627197,
      "epoch": 2.4354902739182216,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044095002229708655,
      "loss": 5.1204,
      "mean_token_accuracy": 0.1741567760705948,
      "num_tokens": 53769725.0,
      "step": 24540
    },
    {
      "entropy": 5.6054205894470215,
      "epoch": 2.435986502580389,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044092609383386517,
      "loss": 4.9904,
      "mean_token_accuracy": 0.1889515995979309,
      "num_tokens": 53781217.0,
      "step": 24545
    },
    {
      "entropy": 5.664636135101318,
      "epoch": 2.4364827312425565,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044090216125603426,
      "loss": 5.1035,
      "mean_token_accuracy": 0.18266451805830003,
      "num_tokens": 53791478.0,
      "step": 24550
    },
    {
      "entropy": 5.606493234634399,
      "epoch": 2.436978959904724,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004408782245641872,
      "loss": 4.9963,
      "mean_token_accuracy": 0.18770160377025605,
      "num_tokens": 53802844.0,
      "step": 24555
    },
    {
      "entropy": 5.660340404510498,
      "epoch": 2.4374751885668915,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004408542837589176,
      "loss": 5.0456,
      "mean_token_accuracy": 0.18410823196172715,
      "num_tokens": 53812938.0,
      "step": 24560
    },
    {
      "entropy": 5.65987114906311,
      "epoch": 2.437971417229059,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044083033884081914,
      "loss": 5.0118,
      "mean_token_accuracy": 0.1830706000328064,
      "num_tokens": 53824105.0,
      "step": 24565
    },
    {
      "entropy": 5.70196852684021,
      "epoch": 2.438467645891227,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004408063898104856,
      "loss": 5.151,
      "mean_token_accuracy": 0.1737814098596573,
      "num_tokens": 53834956.0,
      "step": 24570
    },
    {
      "entropy": 5.6442197322845455,
      "epoch": 2.4389638745533944,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044078243666851094,
      "loss": 5.0673,
      "mean_token_accuracy": 0.17556607127189636,
      "num_tokens": 53846257.0,
      "step": 24575
    },
    {
      "entropy": 5.613988780975342,
      "epoch": 2.439460103215562,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044075847941548905,
      "loss": 4.9653,
      "mean_token_accuracy": 0.1883048728108406,
      "num_tokens": 53857107.0,
      "step": 24580
    },
    {
      "entropy": 5.640861463546753,
      "epoch": 2.4399563318777293,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044073451805201396,
      "loss": 5.0458,
      "mean_token_accuracy": 0.17695295661687852,
      "num_tokens": 53867779.0,
      "step": 24585
    },
    {
      "entropy": 5.650749969482422,
      "epoch": 2.440452560539897,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044071055257868,
      "loss": 5.0276,
      "mean_token_accuracy": 0.17850720882415771,
      "num_tokens": 53878661.0,
      "step": 24590
    },
    {
      "entropy": 5.734226274490356,
      "epoch": 2.4409487892020643,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044068658299608145,
      "loss": 5.1218,
      "mean_token_accuracy": 0.17974412739276885,
      "num_tokens": 53891300.0,
      "step": 24595
    },
    {
      "entropy": 5.66215763092041,
      "epoch": 2.4414450178642317,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044066260930481264,
      "loss": 4.9801,
      "mean_token_accuracy": 0.18581278324127198,
      "num_tokens": 53901086.0,
      "step": 24600
    },
    {
      "entropy": 5.633382749557495,
      "epoch": 2.441941246526399,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004406386315054681,
      "loss": 5.0724,
      "mean_token_accuracy": 0.17196900844573976,
      "num_tokens": 53912599.0,
      "step": 24605
    },
    {
      "entropy": 5.568147468566894,
      "epoch": 2.4424374751885667,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044061464959864243,
      "loss": 5.0538,
      "mean_token_accuracy": 0.17932753562927245,
      "num_tokens": 53924293.0,
      "step": 24610
    },
    {
      "entropy": 5.666187429428101,
      "epoch": 2.4429337038507346,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004405906635849302,
      "loss": 5.1084,
      "mean_token_accuracy": 0.18032115399837495,
      "num_tokens": 53935074.0,
      "step": 24615
    },
    {
      "entropy": 5.679695987701416,
      "epoch": 2.443429932512902,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004405666734649265,
      "loss": 5.0425,
      "mean_token_accuracy": 0.18222176879644394,
      "num_tokens": 53946770.0,
      "step": 24620
    },
    {
      "entropy": 5.668135261535644,
      "epoch": 2.4439261611750696,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044054267923922594,
      "loss": 5.0253,
      "mean_token_accuracy": 0.17903458178043366,
      "num_tokens": 53957077.0,
      "step": 24625
    },
    {
      "entropy": 5.646451187133789,
      "epoch": 2.444422389837237,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044051868090842376,
      "loss": 5.0511,
      "mean_token_accuracy": 0.1833239108324051,
      "num_tokens": 53968455.0,
      "step": 24630
    },
    {
      "entropy": 5.541370868682861,
      "epoch": 2.4449186184994045,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044049467847311486,
      "loss": 5.0265,
      "mean_token_accuracy": 0.18293387591838836,
      "num_tokens": 53980429.0,
      "step": 24635
    },
    {
      "entropy": 5.574800252914429,
      "epoch": 2.445414847161572,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004404706719338946,
      "loss": 5.011,
      "mean_token_accuracy": 0.1877894639968872,
      "num_tokens": 53990937.0,
      "step": 24640
    },
    {
      "entropy": 5.641186141967774,
      "epoch": 2.4459110758237395,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044044666129135825,
      "loss": 5.0079,
      "mean_token_accuracy": 0.1844955116510391,
      "num_tokens": 54003526.0,
      "step": 24645
    },
    {
      "entropy": 5.60631160736084,
      "epoch": 2.446407304485907,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044042264654610115,
      "loss": 4.9787,
      "mean_token_accuracy": 0.18989665061235428,
      "num_tokens": 54013875.0,
      "step": 24650
    },
    {
      "entropy": 5.688058757781983,
      "epoch": 2.446903533148075,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044039862769871884,
      "loss": 5.1595,
      "mean_token_accuracy": 0.17500021755695344,
      "num_tokens": 54024103.0,
      "step": 24655
    },
    {
      "entropy": 5.6839619159698485,
      "epoch": 2.4473997618102423,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044037460474980703,
      "loss": 5.105,
      "mean_token_accuracy": 0.17597226947546005,
      "num_tokens": 54035417.0,
      "step": 24660
    },
    {
      "entropy": 5.701973724365234,
      "epoch": 2.44789599047241,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004403505776999614,
      "loss": 5.1014,
      "mean_token_accuracy": 0.17408940494060515,
      "num_tokens": 54045180.0,
      "step": 24665
    },
    {
      "entropy": 5.674214792251587,
      "epoch": 2.4483922191345773,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044032654654977766,
      "loss": 4.997,
      "mean_token_accuracy": 0.1884599432349205,
      "num_tokens": 54056493.0,
      "step": 24670
    },
    {
      "entropy": 5.5966733455657955,
      "epoch": 2.4488884477967447,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004403025112998519,
      "loss": 5.0134,
      "mean_token_accuracy": 0.1836152732372284,
      "num_tokens": 54067445.0,
      "step": 24675
    },
    {
      "entropy": 5.644637393951416,
      "epoch": 2.4493846764589122,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004402784719507799,
      "loss": 5.0111,
      "mean_token_accuracy": 0.18884455859661103,
      "num_tokens": 54076937.0,
      "step": 24680
    },
    {
      "entropy": 5.668718051910401,
      "epoch": 2.4498809051210797,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004402544285031581,
      "loss": 5.0673,
      "mean_token_accuracy": 0.18679813295602798,
      "num_tokens": 54088014.0,
      "step": 24685
    },
    {
      "entropy": 5.684904527664185,
      "epoch": 2.450377133783247,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004402303809575825,
      "loss": 5.0197,
      "mean_token_accuracy": 0.18256447315216065,
      "num_tokens": 54098984.0,
      "step": 24690
    },
    {
      "entropy": 5.698920440673828,
      "epoch": 2.4508733624454146,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004402063293146495,
      "loss": 5.1532,
      "mean_token_accuracy": 0.17757717818021773,
      "num_tokens": 54110490.0,
      "step": 24695
    },
    {
      "entropy": 5.678144264221191,
      "epoch": 2.451369591107582,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004401822735749555,
      "loss": 5.0444,
      "mean_token_accuracy": 0.18121004849672318,
      "num_tokens": 54120785.0,
      "step": 24700
    },
    {
      "entropy": 5.63131628036499,
      "epoch": 2.45186581976975,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044015821373909707,
      "loss": 4.9942,
      "mean_token_accuracy": 0.18264717012643814,
      "num_tokens": 54131339.0,
      "step": 24705
    },
    {
      "entropy": 5.658252954483032,
      "epoch": 2.4523620484319175,
      "grad_norm": 1.125,
      "learning_rate": 0.00044013414980767084,
      "loss": 5.0751,
      "mean_token_accuracy": 0.17629032731056213,
      "num_tokens": 54142270.0,
      "step": 24710
    },
    {
      "entropy": 5.64489107131958,
      "epoch": 2.452858277094085,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004401100817812735,
      "loss": 5.034,
      "mean_token_accuracy": 0.1848447725176811,
      "num_tokens": 54153045.0,
      "step": 24715
    },
    {
      "entropy": 5.596445941925049,
      "epoch": 2.4533545057562525,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044008600966050196,
      "loss": 4.9719,
      "mean_token_accuracy": 0.18657734543085097,
      "num_tokens": 54164069.0,
      "step": 24720
    },
    {
      "entropy": 5.6296271800994875,
      "epoch": 2.45385073441842,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004400619334459531,
      "loss": 5.0654,
      "mean_token_accuracy": 0.17870983928442002,
      "num_tokens": 54176902.0,
      "step": 24725
    },
    {
      "entropy": 5.576212930679321,
      "epoch": 2.4543469630805874,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044003785313822386,
      "loss": 5.0222,
      "mean_token_accuracy": 0.18715306371450424,
      "num_tokens": 54188515.0,
      "step": 24730
    },
    {
      "entropy": 5.706819248199463,
      "epoch": 2.454843191742755,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044001376873791166,
      "loss": 5.0783,
      "mean_token_accuracy": 0.17923934310674666,
      "num_tokens": 54199077.0,
      "step": 24735
    },
    {
      "entropy": 5.748900270462036,
      "epoch": 2.455339420404923,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00043998968024561337,
      "loss": 5.0984,
      "mean_token_accuracy": 0.18280090838670732,
      "num_tokens": 54210216.0,
      "step": 24740
    },
    {
      "entropy": 5.612724924087525,
      "epoch": 2.4558356490670903,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004399655876619267,
      "loss": 5.02,
      "mean_token_accuracy": 0.18568102419376373,
      "num_tokens": 54221143.0,
      "step": 24745
    },
    {
      "entropy": 5.628732204437256,
      "epoch": 2.4563318777292578,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043994149098744884,
      "loss": 5.094,
      "mean_token_accuracy": 0.18287210017442704,
      "num_tokens": 54233070.0,
      "step": 24750
    },
    {
      "entropy": 5.635801887512207,
      "epoch": 2.4568281063914252,
      "grad_norm": 1.0,
      "learning_rate": 0.0004399173902227775,
      "loss": 5.0277,
      "mean_token_accuracy": 0.18385483771562577,
      "num_tokens": 54243793.0,
      "step": 24755
    },
    {
      "entropy": 5.5951799869537355,
      "epoch": 2.4573243350535927,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00043989328536851014,
      "loss": 5.0372,
      "mean_token_accuracy": 0.18198852837085724,
      "num_tokens": 54256983.0,
      "step": 24760
    },
    {
      "entropy": 5.599905776977539,
      "epoch": 2.45782056371576,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043986917642524465,
      "loss": 5.0136,
      "mean_token_accuracy": 0.18366056382656099,
      "num_tokens": 54268789.0,
      "step": 24765
    },
    {
      "entropy": 5.715070915222168,
      "epoch": 2.4583167923779277,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043984506339357887,
      "loss": 5.0588,
      "mean_token_accuracy": 0.17220307141542435,
      "num_tokens": 54279517.0,
      "step": 24770
    },
    {
      "entropy": 5.646164608001709,
      "epoch": 2.458813021040095,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004398209462741107,
      "loss": 5.0388,
      "mean_token_accuracy": 0.1794026955962181,
      "num_tokens": 54290865.0,
      "step": 24775
    },
    {
      "entropy": 5.754239225387574,
      "epoch": 2.4593092497022626,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004397968250674383,
      "loss": 5.0921,
      "mean_token_accuracy": 0.18712487518787385,
      "num_tokens": 54302658.0,
      "step": 24780
    },
    {
      "entropy": 5.644929361343384,
      "epoch": 2.45980547836443,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004397726997741597,
      "loss": 5.0691,
      "mean_token_accuracy": 0.18459361046552658,
      "num_tokens": 54314000.0,
      "step": 24785
    },
    {
      "entropy": 5.579902648925781,
      "epoch": 2.460301707026598,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004397485703948732,
      "loss": 5.0251,
      "mean_token_accuracy": 0.18309285193681718,
      "num_tokens": 54325952.0,
      "step": 24790
    },
    {
      "entropy": 5.677331733703613,
      "epoch": 2.4607979356887655,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004397244369301771,
      "loss": 5.122,
      "mean_token_accuracy": 0.17587903290987014,
      "num_tokens": 54338582.0,
      "step": 24795
    },
    {
      "entropy": 5.68961238861084,
      "epoch": 2.461294164350933,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043970029938066995,
      "loss": 5.0833,
      "mean_token_accuracy": 0.17606962621212005,
      "num_tokens": 54349513.0,
      "step": 24800
    },
    {
      "entropy": 5.65910701751709,
      "epoch": 2.4617903930131004,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004396761577469503,
      "loss": 5.027,
      "mean_token_accuracy": 0.177449594438076,
      "num_tokens": 54360264.0,
      "step": 24805
    },
    {
      "entropy": 5.652011108398438,
      "epoch": 2.462286621675268,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00043965201202961673,
      "loss": 5.0741,
      "mean_token_accuracy": 0.1734939381480217,
      "num_tokens": 54371379.0,
      "step": 24810
    },
    {
      "entropy": 5.657254219055176,
      "epoch": 2.4627828503374354,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043962786222926807,
      "loss": 5.0614,
      "mean_token_accuracy": 0.1822097435593605,
      "num_tokens": 54382844.0,
      "step": 24815
    },
    {
      "entropy": 5.6178652286529545,
      "epoch": 2.463279078999603,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043960370834650325,
      "loss": 5.0202,
      "mean_token_accuracy": 0.18634730726480483,
      "num_tokens": 54392807.0,
      "step": 24820
    },
    {
      "entropy": 5.592738819122315,
      "epoch": 2.4637753076617708,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004395795503819211,
      "loss": 5.0355,
      "mean_token_accuracy": 0.18564964681863785,
      "num_tokens": 54404060.0,
      "step": 24825
    },
    {
      "entropy": 5.620774602890014,
      "epoch": 2.4642715363239382,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004395553883361207,
      "loss": 4.9658,
      "mean_token_accuracy": 0.18100523352622985,
      "num_tokens": 54414636.0,
      "step": 24830
    },
    {
      "entropy": 5.636478567123413,
      "epoch": 2.4647677649861057,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043953122220970127,
      "loss": 5.0131,
      "mean_token_accuracy": 0.1812512159347534,
      "num_tokens": 54426185.0,
      "step": 24835
    },
    {
      "entropy": 5.64766149520874,
      "epoch": 2.465263993648273,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043950705200326205,
      "loss": 5.0229,
      "mean_token_accuracy": 0.17807802557945251,
      "num_tokens": 54436468.0,
      "step": 24840
    },
    {
      "entropy": 5.631024312973023,
      "epoch": 2.4657602223104407,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004394828777174025,
      "loss": 5.0272,
      "mean_token_accuracy": 0.18432718962430955,
      "num_tokens": 54446228.0,
      "step": 24845
    },
    {
      "entropy": 5.619817876815796,
      "epoch": 2.466256450972608,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043945869935272194,
      "loss": 5.0842,
      "mean_token_accuracy": 0.17633639574050902,
      "num_tokens": 54456704.0,
      "step": 24850
    },
    {
      "entropy": 5.653741312026978,
      "epoch": 2.4667526796347756,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043943451690982006,
      "loss": 5.0448,
      "mean_token_accuracy": 0.18923932015895845,
      "num_tokens": 54467780.0,
      "step": 24855
    },
    {
      "entropy": 5.74583101272583,
      "epoch": 2.467248908296943,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004394103303892964,
      "loss": 5.1495,
      "mean_token_accuracy": 0.17261024862527846,
      "num_tokens": 54478428.0,
      "step": 24860
    },
    {
      "entropy": 5.602934837341309,
      "epoch": 2.4677451369591106,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043938613979175087,
      "loss": 4.9349,
      "mean_token_accuracy": 0.1887226715683937,
      "num_tokens": 54489544.0,
      "step": 24865
    },
    {
      "entropy": 5.592215442657471,
      "epoch": 2.468241365621278,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004393619451177834,
      "loss": 5.0131,
      "mean_token_accuracy": 0.1848815307021141,
      "num_tokens": 54501438.0,
      "step": 24870
    },
    {
      "entropy": 5.599107170104981,
      "epoch": 2.468737594283446,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043933774636799365,
      "loss": 4.9663,
      "mean_token_accuracy": 0.18457163721323014,
      "num_tokens": 54511746.0,
      "step": 24875
    },
    {
      "entropy": 5.690106391906738,
      "epoch": 2.4692338229456134,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043931354354298204,
      "loss": 5.0645,
      "mean_token_accuracy": 0.18025142401456834,
      "num_tokens": 54523141.0,
      "step": 24880
    },
    {
      "entropy": 5.764476490020752,
      "epoch": 2.469730051607781,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004392893366433486,
      "loss": 5.1924,
      "mean_token_accuracy": 0.1728057250380516,
      "num_tokens": 54533275.0,
      "step": 24885
    },
    {
      "entropy": 5.715474414825439,
      "epoch": 2.4702262802699484,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043926512566969366,
      "loss": 5.0288,
      "mean_token_accuracy": 0.17949291616678237,
      "num_tokens": 54545309.0,
      "step": 24890
    },
    {
      "entropy": 5.612745523452759,
      "epoch": 2.470722508932116,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043924091062261743,
      "loss": 4.9862,
      "mean_token_accuracy": 0.19408871978521347,
      "num_tokens": 54554905.0,
      "step": 24895
    },
    {
      "entropy": 5.626927375793457,
      "epoch": 2.4712187375942833,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004392166915027206,
      "loss": 5.0472,
      "mean_token_accuracy": 0.1748857840895653,
      "num_tokens": 54566050.0,
      "step": 24900
    },
    {
      "entropy": 5.661313104629516,
      "epoch": 2.471714966256451,
      "grad_norm": 1.0,
      "learning_rate": 0.0004391924683106037,
      "loss": 5.1072,
      "mean_token_accuracy": 0.17874280363321304,
      "num_tokens": 54577791.0,
      "step": 24905
    },
    {
      "entropy": 5.685702610015869,
      "epoch": 2.4722111949186187,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004391682410468674,
      "loss": 5.0348,
      "mean_token_accuracy": 0.18320778608322144,
      "num_tokens": 54589183.0,
      "step": 24910
    },
    {
      "entropy": 5.63431568145752,
      "epoch": 2.472707423580786,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00043914400971211255,
      "loss": 5.0322,
      "mean_token_accuracy": 0.1788840115070343,
      "num_tokens": 54600776.0,
      "step": 24915
    },
    {
      "entropy": 5.627207279205322,
      "epoch": 2.4732036522429537,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004391197743069398,
      "loss": 4.9937,
      "mean_token_accuracy": 0.18963968455791474,
      "num_tokens": 54611167.0,
      "step": 24920
    },
    {
      "entropy": 5.662481784820557,
      "epoch": 2.473699880905121,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004390955348319503,
      "loss": 5.0455,
      "mean_token_accuracy": 0.18355485647916794,
      "num_tokens": 54622811.0,
      "step": 24925
    },
    {
      "entropy": 5.670573043823242,
      "epoch": 2.4741961095672886,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043907129128774525,
      "loss": 5.0624,
      "mean_token_accuracy": 0.1808809757232666,
      "num_tokens": 54635109.0,
      "step": 24930
    },
    {
      "entropy": 5.695790767669678,
      "epoch": 2.474692338229456,
      "grad_norm": 1.25,
      "learning_rate": 0.00043904704367492563,
      "loss": 5.0741,
      "mean_token_accuracy": 0.18457991778850555,
      "num_tokens": 54644260.0,
      "step": 24935
    },
    {
      "entropy": 5.611251497268677,
      "epoch": 2.4751885668916236,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004390227919940928,
      "loss": 4.9637,
      "mean_token_accuracy": 0.18704122602939605,
      "num_tokens": 54654433.0,
      "step": 24940
    },
    {
      "entropy": 5.664799547195434,
      "epoch": 2.475684795553791,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043899853624584825,
      "loss": 5.0951,
      "mean_token_accuracy": 0.1764600917696953,
      "num_tokens": 54666582.0,
      "step": 24945
    },
    {
      "entropy": 5.641666984558105,
      "epoch": 2.4761810242159585,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043897427643079336,
      "loss": 4.9974,
      "mean_token_accuracy": 0.18585244864225386,
      "num_tokens": 54677247.0,
      "step": 24950
    },
    {
      "entropy": 5.650543594360352,
      "epoch": 2.476677252878126,
      "grad_norm": 1.0,
      "learning_rate": 0.00043895001254952963,
      "loss": 5.0772,
      "mean_token_accuracy": 0.17729825675487518,
      "num_tokens": 54688423.0,
      "step": 24955
    },
    {
      "entropy": 5.648799991607666,
      "epoch": 2.477173481540294,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000438925744602659,
      "loss": 5.0549,
      "mean_token_accuracy": 0.1836518943309784,
      "num_tokens": 54698517.0,
      "step": 24960
    },
    {
      "entropy": 5.7065958976745605,
      "epoch": 2.4776697102024614,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043890147259078314,
      "loss": 5.0934,
      "mean_token_accuracy": 0.17384282797574996,
      "num_tokens": 54711635.0,
      "step": 24965
    },
    {
      "entropy": 5.655050754547119,
      "epoch": 2.478165938864629,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043887719651450384,
      "loss": 5.0587,
      "mean_token_accuracy": 0.18345407992601395,
      "num_tokens": 54723051.0,
      "step": 24970
    },
    {
      "entropy": 5.757682800292969,
      "epoch": 2.4786621675267964,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004388529163744233,
      "loss": 5.0499,
      "mean_token_accuracy": 0.18396979868412017,
      "num_tokens": 54734191.0,
      "step": 24975
    },
    {
      "entropy": 5.663413000106812,
      "epoch": 2.479158396188964,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004388286321711435,
      "loss": 5.0231,
      "mean_token_accuracy": 0.18160929828882216,
      "num_tokens": 54745632.0,
      "step": 24980
    },
    {
      "entropy": 5.628558397293091,
      "epoch": 2.4796546248511313,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004388043439052665,
      "loss": 5.0727,
      "mean_token_accuracy": 0.17274345010519027,
      "num_tokens": 54755871.0,
      "step": 24985
    },
    {
      "entropy": 5.577577352523804,
      "epoch": 2.480150853513299,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043878005157739486,
      "loss": 4.9741,
      "mean_token_accuracy": 0.18860432654619216,
      "num_tokens": 54768036.0,
      "step": 24990
    },
    {
      "entropy": 5.6407218933105465,
      "epoch": 2.4806470821754667,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004387557551881308,
      "loss": 5.0064,
      "mean_token_accuracy": 0.18320934772491454,
      "num_tokens": 54778518.0,
      "step": 24995
    },
    {
      "entropy": 5.595664024353027,
      "epoch": 2.481143310837634,
      "grad_norm": 0.9375,
      "learning_rate": 0.000438731454738077,
      "loss": 5.0029,
      "mean_token_accuracy": 0.18817301392555236,
      "num_tokens": 54790786.0,
      "step": 25000
    },
    {
      "entropy": 5.572035503387451,
      "epoch": 2.4816395394998016,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043870715022783575,
      "loss": 4.9906,
      "mean_token_accuracy": 0.18166295289993287,
      "num_tokens": 54800063.0,
      "step": 25005
    },
    {
      "entropy": 5.63029670715332,
      "epoch": 2.482135768161969,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004386828416580101,
      "loss": 5.0817,
      "mean_token_accuracy": 0.1763943076133728,
      "num_tokens": 54810605.0,
      "step": 25010
    },
    {
      "entropy": 5.668052434921265,
      "epoch": 2.4826319968241366,
      "grad_norm": 1.125,
      "learning_rate": 0.0004386585290292026,
      "loss": 4.9964,
      "mean_token_accuracy": 0.18241126090288162,
      "num_tokens": 54820626.0,
      "step": 25015
    },
    {
      "entropy": 5.60145583152771,
      "epoch": 2.483128225486304,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004386342123420161,
      "loss": 4.9822,
      "mean_token_accuracy": 0.18484282493591309,
      "num_tokens": 54831395.0,
      "step": 25020
    },
    {
      "entropy": 5.562696504592895,
      "epoch": 2.4836244541484715,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043860989159705396,
      "loss": 5.0068,
      "mean_token_accuracy": 0.1803899347782135,
      "num_tokens": 54841238.0,
      "step": 25025
    },
    {
      "entropy": 5.675261354446411,
      "epoch": 2.484120682810639,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00043858556679491896,
      "loss": 5.0844,
      "mean_token_accuracy": 0.18092761486768721,
      "num_tokens": 54851301.0,
      "step": 25030
    },
    {
      "entropy": 5.630281162261963,
      "epoch": 2.4846169114728065,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043856123793621437,
      "loss": 5.0138,
      "mean_token_accuracy": 0.17968251407146454,
      "num_tokens": 54863135.0,
      "step": 25035
    },
    {
      "entropy": 5.63353853225708,
      "epoch": 2.485113140134974,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043853690502154355,
      "loss": 5.0674,
      "mean_token_accuracy": 0.17858094424009324,
      "num_tokens": 54874780.0,
      "step": 25040
    },
    {
      "entropy": 5.662199401855469,
      "epoch": 2.485609368797142,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004385125680515099,
      "loss": 5.0686,
      "mean_token_accuracy": 0.18098364919424056,
      "num_tokens": 54885237.0,
      "step": 25045
    },
    {
      "entropy": 5.657610607147217,
      "epoch": 2.4861055974593094,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004384882270267168,
      "loss": 5.066,
      "mean_token_accuracy": 0.18456811159849168,
      "num_tokens": 54896801.0,
      "step": 25050
    },
    {
      "entropy": 5.616772270202636,
      "epoch": 2.486601826121477,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004384638819477681,
      "loss": 5.0272,
      "mean_token_accuracy": 0.17592017948627472,
      "num_tokens": 54908427.0,
      "step": 25055
    },
    {
      "entropy": 5.534160900115967,
      "epoch": 2.4870980547836443,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043843953281526727,
      "loss": 4.9851,
      "mean_token_accuracy": 0.19006904661655427,
      "num_tokens": 54918394.0,
      "step": 25060
    },
    {
      "entropy": 5.6087390899658205,
      "epoch": 2.487594283445812,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043841517962981823,
      "loss": 5.0167,
      "mean_token_accuracy": 0.18577939867973328,
      "num_tokens": 54928613.0,
      "step": 25065
    },
    {
      "entropy": 5.678050374984741,
      "epoch": 2.4880905121079793,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043839082239202496,
      "loss": 4.9721,
      "mean_token_accuracy": 0.18437234312295914,
      "num_tokens": 54938938.0,
      "step": 25070
    },
    {
      "entropy": 5.724816846847534,
      "epoch": 2.4885867407701467,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004383664611024912,
      "loss": 5.1052,
      "mean_token_accuracy": 0.18130384683609008,
      "num_tokens": 54950152.0,
      "step": 25075
    },
    {
      "entropy": 5.674933910369873,
      "epoch": 2.489082969432314,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004383420957618213,
      "loss": 5.1359,
      "mean_token_accuracy": 0.1708926573395729,
      "num_tokens": 54962191.0,
      "step": 25080
    },
    {
      "entropy": 5.61881651878357,
      "epoch": 2.489579198094482,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043831772637061944,
      "loss": 4.9403,
      "mean_token_accuracy": 0.1838528335094452,
      "num_tokens": 54972532.0,
      "step": 25085
    },
    {
      "entropy": 5.698353624343872,
      "epoch": 2.4900754267566496,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043829335292948986,
      "loss": 5.0822,
      "mean_token_accuracy": 0.17695170342922212,
      "num_tokens": 54983212.0,
      "step": 25090
    },
    {
      "entropy": 5.560213994979859,
      "epoch": 2.490571655418817,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043826897543903696,
      "loss": 4.9901,
      "mean_token_accuracy": 0.19021686315536498,
      "num_tokens": 54994094.0,
      "step": 25095
    },
    {
      "entropy": 5.645125150680542,
      "epoch": 2.4910678840809846,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043824459389986534,
      "loss": 5.0698,
      "mean_token_accuracy": 0.18009116351604462,
      "num_tokens": 55004098.0,
      "step": 25100
    },
    {
      "entropy": 5.617304372787475,
      "epoch": 2.491564112743152,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004382202083125795,
      "loss": 4.9627,
      "mean_token_accuracy": 0.18643205910921096,
      "num_tokens": 55013680.0,
      "step": 25105
    },
    {
      "entropy": 5.618179607391357,
      "epoch": 2.4920603414053195,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004381958186777842,
      "loss": 4.9258,
      "mean_token_accuracy": 0.18930024057626724,
      "num_tokens": 55024311.0,
      "step": 25110
    },
    {
      "entropy": 5.611650466918945,
      "epoch": 2.492556570067487,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004381714249960843,
      "loss": 5.1929,
      "mean_token_accuracy": 0.17272419929504396,
      "num_tokens": 55036215.0,
      "step": 25115
    },
    {
      "entropy": 5.686041212081909,
      "epoch": 2.4930527987296545,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004381470272680846,
      "loss": 5.052,
      "mean_token_accuracy": 0.1743343397974968,
      "num_tokens": 55046402.0,
      "step": 25120
    },
    {
      "entropy": 5.66959228515625,
      "epoch": 2.493549027391822,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004381226254943902,
      "loss": 4.9852,
      "mean_token_accuracy": 0.18188990652561188,
      "num_tokens": 55056307.0,
      "step": 25125
    },
    {
      "entropy": 5.641158246994019,
      "epoch": 2.49404525605399,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004380982196756062,
      "loss": 5.0158,
      "mean_token_accuracy": 0.18220286071300507,
      "num_tokens": 55066918.0,
      "step": 25130
    },
    {
      "entropy": 5.621887111663819,
      "epoch": 2.4945414847161573,
      "grad_norm": 0.9375,
      "learning_rate": 0.00043807380981233777,
      "loss": 5.0309,
      "mean_token_accuracy": 0.18041988760232924,
      "num_tokens": 55078166.0,
      "step": 25135
    },
    {
      "entropy": 5.633954620361328,
      "epoch": 2.495037713378325,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004380493959051903,
      "loss": 5.0536,
      "mean_token_accuracy": 0.1818617984652519,
      "num_tokens": 55090919.0,
      "step": 25140
    },
    {
      "entropy": 5.69966197013855,
      "epoch": 2.4955339420404923,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043802497795476916,
      "loss": 5.0805,
      "mean_token_accuracy": 0.18430037796497345,
      "num_tokens": 55100512.0,
      "step": 25145
    },
    {
      "entropy": 5.61855182647705,
      "epoch": 2.4960301707026598,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004380005559616798,
      "loss": 5.0128,
      "mean_token_accuracy": 0.1886113852262497,
      "num_tokens": 55111845.0,
      "step": 25150
    },
    {
      "entropy": 5.684377241134643,
      "epoch": 2.4965263993648272,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004379761299265279,
      "loss": 5.1394,
      "mean_token_accuracy": 0.172038272023201,
      "num_tokens": 55123744.0,
      "step": 25155
    },
    {
      "entropy": 5.713619661331177,
      "epoch": 2.4970226280269947,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004379516998499192,
      "loss": 5.0391,
      "mean_token_accuracy": 0.18914420157670975,
      "num_tokens": 55133789.0,
      "step": 25160
    },
    {
      "entropy": 5.6619892597198485,
      "epoch": 2.497518856689162,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004379272657324594,
      "loss": 5.0541,
      "mean_token_accuracy": 0.1811530336737633,
      "num_tokens": 55145325.0,
      "step": 25165
    },
    {
      "entropy": 5.6695044994354244,
      "epoch": 2.49801508535133,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004379028275747545,
      "loss": 5.0202,
      "mean_token_accuracy": 0.1813921272754669,
      "num_tokens": 55157639.0,
      "step": 25170
    },
    {
      "entropy": 5.628574085235596,
      "epoch": 2.4985113140134976,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043787838537741054,
      "loss": 4.9618,
      "mean_token_accuracy": 0.1822642207145691,
      "num_tokens": 55167796.0,
      "step": 25175
    },
    {
      "entropy": 5.61787748336792,
      "epoch": 2.499007542675665,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004378539391410336,
      "loss": 5.075,
      "mean_token_accuracy": 0.1786922425031662,
      "num_tokens": 55178384.0,
      "step": 25180
    },
    {
      "entropy": 5.707620286941529,
      "epoch": 2.4995037713378325,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043782948886622987,
      "loss": 5.0991,
      "mean_token_accuracy": 0.1769462987780571,
      "num_tokens": 55188363.0,
      "step": 25185
    },
    {
      "entropy": 5.6671336650848385,
      "epoch": 2.5,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004378050345536057,
      "loss": 5.0738,
      "mean_token_accuracy": 0.177908456325531,
      "num_tokens": 55198293.0,
      "step": 25190
    },
    {
      "entropy": 5.601549100875855,
      "epoch": 2.5004962286621675,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043778057620376745,
      "loss": 5.046,
      "mean_token_accuracy": 0.18486909568309784,
      "num_tokens": 55208368.0,
      "step": 25195
    },
    {
      "entropy": 5.649009037017822,
      "epoch": 2.500992457324335,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043775611381732164,
      "loss": 5.0865,
      "mean_token_accuracy": 0.17908266931772232,
      "num_tokens": 55220075.0,
      "step": 25200
    },
    {
      "entropy": 5.649923849105835,
      "epoch": 2.5014886859865024,
      "grad_norm": 1.078125,
      "learning_rate": 0.000437731647394875,
      "loss": 5.0354,
      "mean_token_accuracy": 0.1806858405470848,
      "num_tokens": 55231060.0,
      "step": 25205
    },
    {
      "entropy": 5.678768491744995,
      "epoch": 2.50198491464867,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004377071769370341,
      "loss": 5.0461,
      "mean_token_accuracy": 0.17938314080238343,
      "num_tokens": 55242443.0,
      "step": 25210
    },
    {
      "entropy": 5.524090671539307,
      "epoch": 2.5024811433108374,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043768270244440587,
      "loss": 4.9895,
      "mean_token_accuracy": 0.1892465353012085,
      "num_tokens": 55253285.0,
      "step": 25215
    },
    {
      "entropy": 5.581346035003662,
      "epoch": 2.5029773719730053,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00043765822391759714,
      "loss": 4.9974,
      "mean_token_accuracy": 0.18430763632059097,
      "num_tokens": 55264694.0,
      "step": 25220
    },
    {
      "entropy": 5.619634389877319,
      "epoch": 2.5034736006351728,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004376337413572149,
      "loss": 5.0258,
      "mean_token_accuracy": 0.1803026795387268,
      "num_tokens": 55275429.0,
      "step": 25225
    },
    {
      "entropy": 5.682784557342529,
      "epoch": 2.5039698292973402,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043760925476386635,
      "loss": 5.0815,
      "mean_token_accuracy": 0.18032969981431962,
      "num_tokens": 55286764.0,
      "step": 25230
    },
    {
      "entropy": 5.540089511871338,
      "epoch": 2.5044660579595077,
      "grad_norm": 1.0,
      "learning_rate": 0.0004375847641381587,
      "loss": 4.9294,
      "mean_token_accuracy": 0.19560119062662124,
      "num_tokens": 55296284.0,
      "step": 25235
    },
    {
      "entropy": 5.609358024597168,
      "epoch": 2.504962286621675,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004375602694806993,
      "loss": 5.0344,
      "mean_token_accuracy": 0.1853463500738144,
      "num_tokens": 55307684.0,
      "step": 25240
    },
    {
      "entropy": 5.6401222229003904,
      "epoch": 2.5054585152838427,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043753577079209537,
      "loss": 4.9998,
      "mean_token_accuracy": 0.1812564730644226,
      "num_tokens": 55318202.0,
      "step": 25245
    },
    {
      "entropy": 5.692163276672363,
      "epoch": 2.5059547439460106,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043751126807295464,
      "loss": 5.1018,
      "mean_token_accuracy": 0.1749178186058998,
      "num_tokens": 55328827.0,
      "step": 25250
    },
    {
      "entropy": 5.659640216827393,
      "epoch": 2.506450972608178,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043748676132388464,
      "loss": 5.0903,
      "mean_token_accuracy": 0.17452777326107025,
      "num_tokens": 55338171.0,
      "step": 25255
    },
    {
      "entropy": 5.603626823425293,
      "epoch": 2.5069472012703455,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043746225054549316,
      "loss": 5.0419,
      "mean_token_accuracy": 0.17991005927324294,
      "num_tokens": 55349081.0,
      "step": 25260
    },
    {
      "entropy": 5.61528205871582,
      "epoch": 2.507443429932513,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043743773573838785,
      "loss": 5.0233,
      "mean_token_accuracy": 0.1847813233733177,
      "num_tokens": 55359344.0,
      "step": 25265
    },
    {
      "entropy": 5.61237416267395,
      "epoch": 2.5079396585946805,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043741321690317673,
      "loss": 4.9384,
      "mean_token_accuracy": 0.18879740089178085,
      "num_tokens": 55369643.0,
      "step": 25270
    },
    {
      "entropy": 5.617412900924682,
      "epoch": 2.508435887256848,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004373886940404677,
      "loss": 4.9439,
      "mean_token_accuracy": 0.1929817959666252,
      "num_tokens": 55380812.0,
      "step": 25275
    },
    {
      "entropy": 5.559889650344848,
      "epoch": 2.5089321159190154,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004373641671508691,
      "loss": 5.008,
      "mean_token_accuracy": 0.18839133083820342,
      "num_tokens": 55393136.0,
      "step": 25280
    },
    {
      "entropy": 5.567485570907593,
      "epoch": 2.509428344581183,
      "grad_norm": 0.984375,
      "learning_rate": 0.000437339636234989,
      "loss": 4.9383,
      "mean_token_accuracy": 0.17955061048269272,
      "num_tokens": 55404317.0,
      "step": 25285
    },
    {
      "entropy": 5.632021045684814,
      "epoch": 2.5099245732433504,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043731510129343574,
      "loss": 5.1008,
      "mean_token_accuracy": 0.17649200856685637,
      "num_tokens": 55415812.0,
      "step": 25290
    },
    {
      "entropy": 5.688418912887573,
      "epoch": 2.510420801905518,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043729056232681766,
      "loss": 5.0781,
      "mean_token_accuracy": 0.17885019779205322,
      "num_tokens": 55426402.0,
      "step": 25295
    },
    {
      "entropy": 5.664721298217773,
      "epoch": 2.5109170305676853,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004372660193357434,
      "loss": 5.0755,
      "mean_token_accuracy": 0.1839914157986641,
      "num_tokens": 55437474.0,
      "step": 25300
    },
    {
      "entropy": 5.638381767272949,
      "epoch": 2.5114132592298533,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004372414723208215,
      "loss": 5.0235,
      "mean_token_accuracy": 0.1811476394534111,
      "num_tokens": 55447587.0,
      "step": 25305
    },
    {
      "entropy": 5.738469648361206,
      "epoch": 2.5119094878920207,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004372169212826606,
      "loss": 5.167,
      "mean_token_accuracy": 0.17438976615667343,
      "num_tokens": 55460587.0,
      "step": 25310
    },
    {
      "entropy": 5.638930988311768,
      "epoch": 2.512405716554188,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043719236622186976,
      "loss": 5.0336,
      "mean_token_accuracy": 0.18139548301696778,
      "num_tokens": 55471150.0,
      "step": 25315
    },
    {
      "entropy": 5.605848884582519,
      "epoch": 2.5129019452163557,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004371678071390576,
      "loss": 5.0205,
      "mean_token_accuracy": 0.18585461080074311,
      "num_tokens": 55480368.0,
      "step": 25320
    },
    {
      "entropy": 5.683059978485107,
      "epoch": 2.513398173878523,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043714324403483333,
      "loss": 5.0844,
      "mean_token_accuracy": 0.177450293302536,
      "num_tokens": 55491954.0,
      "step": 25325
    },
    {
      "entropy": 5.661876964569092,
      "epoch": 2.5138944025406906,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043711867690980605,
      "loss": 5.0317,
      "mean_token_accuracy": 0.17539599239826204,
      "num_tokens": 55502560.0,
      "step": 25330
    },
    {
      "entropy": 5.608913946151733,
      "epoch": 2.5143906312028586,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00043709410576458477,
      "loss": 5.0175,
      "mean_token_accuracy": 0.1842965930700302,
      "num_tokens": 55514571.0,
      "step": 25335
    },
    {
      "entropy": 5.653336715698242,
      "epoch": 2.514886859865026,
      "grad_norm": 1.125,
      "learning_rate": 0.00043706953059977907,
      "loss": 5.0369,
      "mean_token_accuracy": 0.1858746275305748,
      "num_tokens": 55525206.0,
      "step": 25340
    },
    {
      "entropy": 5.648258256912231,
      "epoch": 2.5153830885271935,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043704495141599816,
      "loss": 5.0661,
      "mean_token_accuracy": 0.17571464329957961,
      "num_tokens": 55536551.0,
      "step": 25345
    },
    {
      "entropy": 5.638769721984863,
      "epoch": 2.515879317189361,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004370203682138517,
      "loss": 5.0276,
      "mean_token_accuracy": 0.17191925048828124,
      "num_tokens": 55547842.0,
      "step": 25350
    },
    {
      "entropy": 5.70635404586792,
      "epoch": 2.5163755458515285,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004369957809939492,
      "loss": 4.982,
      "mean_token_accuracy": 0.19417157024145126,
      "num_tokens": 55558248.0,
      "step": 25355
    },
    {
      "entropy": 5.619461250305176,
      "epoch": 2.516871774513696,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004369711897569004,
      "loss": 4.9874,
      "mean_token_accuracy": 0.18010385632514953,
      "num_tokens": 55569195.0,
      "step": 25360
    },
    {
      "entropy": 5.644542455673218,
      "epoch": 2.5173680031758634,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004369465945033151,
      "loss": 4.967,
      "mean_token_accuracy": 0.18742364048957824,
      "num_tokens": 55579649.0,
      "step": 25365
    },
    {
      "entropy": 5.567403745651245,
      "epoch": 2.517864231838031,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043692199523380325,
      "loss": 4.8753,
      "mean_token_accuracy": 0.19356495440006255,
      "num_tokens": 55589550.0,
      "step": 25370
    },
    {
      "entropy": 5.622532558441162,
      "epoch": 2.5183604605001984,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004368973919489748,
      "loss": 5.0184,
      "mean_token_accuracy": 0.1831157311797142,
      "num_tokens": 55600673.0,
      "step": 25375
    },
    {
      "entropy": 5.524799537658692,
      "epoch": 2.518856689162366,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043687278464943986,
      "loss": 4.9838,
      "mean_token_accuracy": 0.1821598306298256,
      "num_tokens": 55612045.0,
      "step": 25380
    },
    {
      "entropy": 5.61870756149292,
      "epoch": 2.5193529178245333,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004368481733358088,
      "loss": 5.0386,
      "mean_token_accuracy": 0.18125834316015244,
      "num_tokens": 55623666.0,
      "step": 25385
    },
    {
      "entropy": 5.619310331344605,
      "epoch": 2.519849146486701,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004368235580086917,
      "loss": 5.0158,
      "mean_token_accuracy": 0.18584930747747422,
      "num_tokens": 55634236.0,
      "step": 25390
    },
    {
      "entropy": 5.564135026931763,
      "epoch": 2.5203453751488687,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043679893866869904,
      "loss": 4.9621,
      "mean_token_accuracy": 0.1850930079817772,
      "num_tokens": 55645412.0,
      "step": 25395
    },
    {
      "entropy": 5.582836675643921,
      "epoch": 2.520841603811036,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004367743153164414,
      "loss": 4.9813,
      "mean_token_accuracy": 0.17933174669742585,
      "num_tokens": 55655861.0,
      "step": 25400
    },
    {
      "entropy": 5.603646755218506,
      "epoch": 2.5213378324732036,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043674968795252926,
      "loss": 4.9691,
      "mean_token_accuracy": 0.18434697836637498,
      "num_tokens": 55665227.0,
      "step": 25405
    },
    {
      "entropy": 5.620501565933227,
      "epoch": 2.521834061135371,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004367250565775734,
      "loss": 4.9975,
      "mean_token_accuracy": 0.18768107891082764,
      "num_tokens": 55676104.0,
      "step": 25410
    },
    {
      "entropy": 5.649438571929932,
      "epoch": 2.5223302897975386,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004367004211921847,
      "loss": 5.0898,
      "mean_token_accuracy": 0.17737164050340654,
      "num_tokens": 55688222.0,
      "step": 25415
    },
    {
      "entropy": 5.668508434295655,
      "epoch": 2.522826518459706,
      "grad_norm": 1.0,
      "learning_rate": 0.00043667578179697387,
      "loss": 5.0325,
      "mean_token_accuracy": 0.18523591309785842,
      "num_tokens": 55700138.0,
      "step": 25420
    },
    {
      "entropy": 5.639083051681519,
      "epoch": 2.523322747121874,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004366511383925521,
      "loss": 4.9891,
      "mean_token_accuracy": 0.18719006031751634,
      "num_tokens": 55709444.0,
      "step": 25425
    },
    {
      "entropy": 5.7143035411834715,
      "epoch": 2.5238189757840415,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004366264909795305,
      "loss": 5.1206,
      "mean_token_accuracy": 0.1817706421017647,
      "num_tokens": 55720327.0,
      "step": 25430
    },
    {
      "entropy": 5.5371733665466305,
      "epoch": 2.524315204446209,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004366018395585201,
      "loss": 4.9067,
      "mean_token_accuracy": 0.18627281486988068,
      "num_tokens": 55730727.0,
      "step": 25435
    },
    {
      "entropy": 5.650968742370606,
      "epoch": 2.5248114331083764,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004365771841301323,
      "loss": 5.0567,
      "mean_token_accuracy": 0.18582252711057662,
      "num_tokens": 55741568.0,
      "step": 25440
    },
    {
      "entropy": 5.605025386810302,
      "epoch": 2.525307661770544,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004365525246949786,
      "loss": 5.0098,
      "mean_token_accuracy": 0.1879403606057167,
      "num_tokens": 55752511.0,
      "step": 25445
    },
    {
      "entropy": 5.646641063690185,
      "epoch": 2.5258038904327114,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043652786125367035,
      "loss": 4.9896,
      "mean_token_accuracy": 0.1840684801340103,
      "num_tokens": 55763004.0,
      "step": 25450
    },
    {
      "entropy": 5.634674453735352,
      "epoch": 2.526300119094879,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043650319380681915,
      "loss": 5.0622,
      "mean_token_accuracy": 0.18201306313276291,
      "num_tokens": 55773276.0,
      "step": 25455
    },
    {
      "entropy": 5.666974973678589,
      "epoch": 2.5267963477570463,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004364785223550368,
      "loss": 5.0509,
      "mean_token_accuracy": 0.17824338227510453,
      "num_tokens": 55784607.0,
      "step": 25460
    },
    {
      "entropy": 5.696609973907471,
      "epoch": 2.527292576419214,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00043645384689893504,
      "loss": 5.0788,
      "mean_token_accuracy": 0.17218268811702728,
      "num_tokens": 55796730.0,
      "step": 25465
    },
    {
      "entropy": 5.699306774139404,
      "epoch": 2.5277888050813813,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004364291674391258,
      "loss": 5.0702,
      "mean_token_accuracy": 0.17446513772010802,
      "num_tokens": 55806940.0,
      "step": 25470
    },
    {
      "entropy": 5.626951742172241,
      "epoch": 2.528285033743549,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000436404483976221,
      "loss": 4.9972,
      "mean_token_accuracy": 0.1817326530814171,
      "num_tokens": 55818229.0,
      "step": 25475
    },
    {
      "entropy": 5.6271497249603275,
      "epoch": 2.5287812624057167,
      "grad_norm": 1.0,
      "learning_rate": 0.00043637979651083285,
      "loss": 5.0804,
      "mean_token_accuracy": 0.178194260597229,
      "num_tokens": 55831100.0,
      "step": 25480
    },
    {
      "entropy": 5.653641033172607,
      "epoch": 2.529277491067884,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043635510504357343,
      "loss": 5.0157,
      "mean_token_accuracy": 0.18025284707546235,
      "num_tokens": 55841713.0,
      "step": 25485
    },
    {
      "entropy": 5.613567638397217,
      "epoch": 2.5297737197300516,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043633040957505515,
      "loss": 4.9982,
      "mean_token_accuracy": 0.17254268676042556,
      "num_tokens": 55852737.0,
      "step": 25490
    },
    {
      "entropy": 5.673594522476196,
      "epoch": 2.530269948392219,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004363057101058903,
      "loss": 5.0914,
      "mean_token_accuracy": 0.1823774829506874,
      "num_tokens": 55863471.0,
      "step": 25495
    },
    {
      "entropy": 5.594842958450317,
      "epoch": 2.5307661770543866,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004362810066366915,
      "loss": 4.974,
      "mean_token_accuracy": 0.18353549838066102,
      "num_tokens": 55874305.0,
      "step": 25500
    },
    {
      "entropy": 5.663696432113648,
      "epoch": 2.531262405716554,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00043625629916807117,
      "loss": 5.106,
      "mean_token_accuracy": 0.181083807349205,
      "num_tokens": 55887400.0,
      "step": 25505
    },
    {
      "entropy": 5.591834306716919,
      "epoch": 2.531758634378722,
      "grad_norm": 0.890625,
      "learning_rate": 0.00043623158770064214,
      "loss": 5.0477,
      "mean_token_accuracy": 0.17970513552427292,
      "num_tokens": 55900052.0,
      "step": 25510
    },
    {
      "entropy": 5.7838960647583,
      "epoch": 2.5322548630408894,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004362068722350172,
      "loss": 5.1825,
      "mean_token_accuracy": 0.17349019944667815,
      "num_tokens": 55911609.0,
      "step": 25515
    },
    {
      "entropy": 5.652816104888916,
      "epoch": 2.532751091703057,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043618215277180917,
      "loss": 4.9775,
      "mean_token_accuracy": 0.1900644928216934,
      "num_tokens": 55922873.0,
      "step": 25520
    },
    {
      "entropy": 5.648231554031372,
      "epoch": 2.5332473203652244,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043615742931163106,
      "loss": 5.0354,
      "mean_token_accuracy": 0.17388855367898942,
      "num_tokens": 55932161.0,
      "step": 25525
    },
    {
      "entropy": 5.654283618927002,
      "epoch": 2.533743549027392,
      "grad_norm": 1.171875,
      "learning_rate": 0.000436132701855096,
      "loss": 5.1504,
      "mean_token_accuracy": 0.16971114426851272,
      "num_tokens": 55943601.0,
      "step": 25530
    },
    {
      "entropy": 5.5823814392089846,
      "epoch": 2.5342397776895593,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043610797040281705,
      "loss": 5.0208,
      "mean_token_accuracy": 0.18259500563144684,
      "num_tokens": 55954981.0,
      "step": 25535
    },
    {
      "entropy": 5.653648471832275,
      "epoch": 2.534736006351727,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004360832349554077,
      "loss": 5.0933,
      "mean_token_accuracy": 0.17702866941690446,
      "num_tokens": 55964407.0,
      "step": 25540
    },
    {
      "entropy": 5.624545383453369,
      "epoch": 2.5352322350138943,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004360584955134811,
      "loss": 4.9918,
      "mean_token_accuracy": 0.17877014130353927,
      "num_tokens": 55974634.0,
      "step": 25545
    },
    {
      "entropy": 5.6501233100891115,
      "epoch": 2.5357284636760618,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043603375207765103,
      "loss": 5.0476,
      "mean_token_accuracy": 0.1730469971895218,
      "num_tokens": 55986753.0,
      "step": 25550
    },
    {
      "entropy": 5.626309823989868,
      "epoch": 2.5362246923382292,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004360090046485308,
      "loss": 5.0216,
      "mean_token_accuracy": 0.18296248465776443,
      "num_tokens": 55997839.0,
      "step": 25555
    },
    {
      "entropy": 5.665109252929687,
      "epoch": 2.536720921000397,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004359842532267342,
      "loss": 5.0214,
      "mean_token_accuracy": 0.18698730915784836,
      "num_tokens": 56008765.0,
      "step": 25560
    },
    {
      "entropy": 5.629961824417114,
      "epoch": 2.5372171496625646,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043595949781287505,
      "loss": 5.0278,
      "mean_token_accuracy": 0.18965462148189544,
      "num_tokens": 56020077.0,
      "step": 25565
    },
    {
      "entropy": 5.635515117645264,
      "epoch": 2.537713378324732,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043593473840756715,
      "loss": 5.0188,
      "mean_token_accuracy": 0.18665145486593246,
      "num_tokens": 56031298.0,
      "step": 25570
    },
    {
      "entropy": 5.68096776008606,
      "epoch": 2.5382096069868996,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004359099750114245,
      "loss": 5.0613,
      "mean_token_accuracy": 0.18140997737646103,
      "num_tokens": 56043955.0,
      "step": 25575
    },
    {
      "entropy": 5.623289632797241,
      "epoch": 2.538705835649067,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004358852076250613,
      "loss": 5.0082,
      "mean_token_accuracy": 0.1839645430445671,
      "num_tokens": 56055734.0,
      "step": 25580
    },
    {
      "entropy": 5.693037748336792,
      "epoch": 2.5392020643112345,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043586043624909153,
      "loss": 5.0434,
      "mean_token_accuracy": 0.180618816614151,
      "num_tokens": 56065472.0,
      "step": 25585
    },
    {
      "entropy": 5.736923933029175,
      "epoch": 2.539698292973402,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004358356608841296,
      "loss": 5.0539,
      "mean_token_accuracy": 0.182550285756588,
      "num_tokens": 56075691.0,
      "step": 25590
    },
    {
      "entropy": 5.641342544555664,
      "epoch": 2.54019452163557,
      "grad_norm": 1.28125,
      "learning_rate": 0.00043581088153078984,
      "loss": 5.0363,
      "mean_token_accuracy": 0.18687410652637482,
      "num_tokens": 56085900.0,
      "step": 25595
    },
    {
      "entropy": 5.652346563339234,
      "epoch": 2.5406907502977374,
      "grad_norm": 0.9375,
      "learning_rate": 0.00043578609818968676,
      "loss": 5.0457,
      "mean_token_accuracy": 0.17956445515155792,
      "num_tokens": 56097675.0,
      "step": 25600
    },
    {
      "entropy": 5.69089035987854,
      "epoch": 2.541186978959905,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043576131086143496,
      "loss": 5.0969,
      "mean_token_accuracy": 0.1713161885738373,
      "num_tokens": 56108102.0,
      "step": 25605
    },
    {
      "entropy": 5.576213693618774,
      "epoch": 2.5416832076220723,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043573651954664897,
      "loss": 4.9277,
      "mean_token_accuracy": 0.18964188992977143,
      "num_tokens": 56118489.0,
      "step": 25610
    },
    {
      "entropy": 5.56165041923523,
      "epoch": 2.54217943628424,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043571172424594363,
      "loss": 4.9298,
      "mean_token_accuracy": 0.19313801676034928,
      "num_tokens": 56127949.0,
      "step": 25615
    },
    {
      "entropy": 5.623591184616089,
      "epoch": 2.5426756649464073,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004356869249599339,
      "loss": 4.9818,
      "mean_token_accuracy": 0.1821180060505867,
      "num_tokens": 56138209.0,
      "step": 25620
    },
    {
      "entropy": 5.622080278396607,
      "epoch": 2.5431718936085748,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004356621216892347,
      "loss": 5.0449,
      "mean_token_accuracy": 0.1876795306801796,
      "num_tokens": 56149005.0,
      "step": 25625
    },
    {
      "entropy": 5.618891191482544,
      "epoch": 2.5436681222707422,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043563731443446095,
      "loss": 4.9732,
      "mean_token_accuracy": 0.1786116510629654,
      "num_tokens": 56159406.0,
      "step": 25630
    },
    {
      "entropy": 5.623910427093506,
      "epoch": 2.5441643509329097,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043561250319622806,
      "loss": 5.0409,
      "mean_token_accuracy": 0.1801416754722595,
      "num_tokens": 56169998.0,
      "step": 25635
    },
    {
      "entropy": 5.710141849517822,
      "epoch": 2.544660579595077,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043558768797515116,
      "loss": 5.1177,
      "mean_token_accuracy": 0.17590323686599732,
      "num_tokens": 56181096.0,
      "step": 25640
    },
    {
      "entropy": 5.6459527015686035,
      "epoch": 2.5451568082572447,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043556286877184566,
      "loss": 5.0298,
      "mean_token_accuracy": 0.1806318864226341,
      "num_tokens": 56191909.0,
      "step": 25645
    },
    {
      "entropy": 5.652188730239868,
      "epoch": 2.5456530369194126,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004355380455869269,
      "loss": 5.069,
      "mean_token_accuracy": 0.17939448207616807,
      "num_tokens": 56202725.0,
      "step": 25650
    },
    {
      "entropy": 5.656960296630859,
      "epoch": 2.54614926558158,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004355132184210106,
      "loss": 5.0269,
      "mean_token_accuracy": 0.18643145263195038,
      "num_tokens": 56213596.0,
      "step": 25655
    },
    {
      "entropy": 5.6174335956573485,
      "epoch": 2.5466454942437475,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004354883872747124,
      "loss": 5.0306,
      "mean_token_accuracy": 0.18596246540546418,
      "num_tokens": 56224416.0,
      "step": 25660
    },
    {
      "entropy": 5.610411024093628,
      "epoch": 2.547141722905915,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043546355214864796,
      "loss": 5.0317,
      "mean_token_accuracy": 0.17697339951992036,
      "num_tokens": 56234325.0,
      "step": 25665
    },
    {
      "entropy": 5.5906885147094725,
      "epoch": 2.5476379515680825,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004354387130434332,
      "loss": 5.0084,
      "mean_token_accuracy": 0.18154211938381196,
      "num_tokens": 56244178.0,
      "step": 25670
    },
    {
      "entropy": 5.615332698822021,
      "epoch": 2.54813418023025,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004354138699596841,
      "loss": 5.0591,
      "mean_token_accuracy": 0.18181985467672349,
      "num_tokens": 56256036.0,
      "step": 25675
    },
    {
      "entropy": 5.634104490280151,
      "epoch": 2.548630408892418,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004353890228980167,
      "loss": 5.0171,
      "mean_token_accuracy": 0.17752217650413513,
      "num_tokens": 56266203.0,
      "step": 25680
    },
    {
      "entropy": 5.661534786224365,
      "epoch": 2.5491266375545854,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004353641718590471,
      "loss": 5.0495,
      "mean_token_accuracy": 0.18294423669576645,
      "num_tokens": 56276958.0,
      "step": 25685
    },
    {
      "entropy": 5.653804016113281,
      "epoch": 2.549622866216753,
      "grad_norm": 1.0,
      "learning_rate": 0.00043533931684339156,
      "loss": 5.0472,
      "mean_token_accuracy": 0.18840475380420685,
      "num_tokens": 56287941.0,
      "step": 25690
    },
    {
      "entropy": 5.579885625839234,
      "epoch": 2.5501190948789203,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004353144578516665,
      "loss": 4.9556,
      "mean_token_accuracy": 0.18574873954057694,
      "num_tokens": 56299390.0,
      "step": 25695
    },
    {
      "entropy": 5.606523418426514,
      "epoch": 2.5506153235410878,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043528959488448835,
      "loss": 5.006,
      "mean_token_accuracy": 0.1808902695775032,
      "num_tokens": 56310762.0,
      "step": 25700
    },
    {
      "entropy": 5.667154550552368,
      "epoch": 2.5511115522032553,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043526472794247365,
      "loss": 5.0035,
      "mean_token_accuracy": 0.18425125777721404,
      "num_tokens": 56322522.0,
      "step": 25705
    },
    {
      "entropy": 5.640324354171753,
      "epoch": 2.5516077808654227,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043523985702623896,
      "loss": 5.0775,
      "mean_token_accuracy": 0.1786528080701828,
      "num_tokens": 56334124.0,
      "step": 25710
    },
    {
      "entropy": 5.603016662597656,
      "epoch": 2.55210400952759,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043521498213640113,
      "loss": 5.0156,
      "mean_token_accuracy": 0.18218573033809662,
      "num_tokens": 56344761.0,
      "step": 25715
    },
    {
      "entropy": 5.601904010772705,
      "epoch": 2.5526002381897577,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043519010327357704,
      "loss": 4.9843,
      "mean_token_accuracy": 0.18429724127054214,
      "num_tokens": 56355518.0,
      "step": 25720
    },
    {
      "entropy": 5.640827083587647,
      "epoch": 2.553096466851925,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004351652204383836,
      "loss": 5.0756,
      "mean_token_accuracy": 0.1820999264717102,
      "num_tokens": 56366307.0,
      "step": 25725
    },
    {
      "entropy": 5.654402208328247,
      "epoch": 2.5535926955140926,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004351403336314377,
      "loss": 4.9855,
      "mean_token_accuracy": 0.18472880423069,
      "num_tokens": 56377065.0,
      "step": 25730
    },
    {
      "entropy": 5.67841329574585,
      "epoch": 2.5540889241762605,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004351154428533566,
      "loss": 5.0383,
      "mean_token_accuracy": 0.1824977234005928,
      "num_tokens": 56387741.0,
      "step": 25735
    },
    {
      "entropy": 5.635607099533081,
      "epoch": 2.554585152838428,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004350905481047576,
      "loss": 5.2049,
      "mean_token_accuracy": 0.17501124292612075,
      "num_tokens": 56398927.0,
      "step": 25740
    },
    {
      "entropy": 5.751573705673218,
      "epoch": 2.5550813815005955,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004350656493862579,
      "loss": 5.1731,
      "mean_token_accuracy": 0.1709260195493698,
      "num_tokens": 56408791.0,
      "step": 25745
    },
    {
      "entropy": 5.670055294036866,
      "epoch": 2.555577610162763,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043504074669847517,
      "loss": 4.9695,
      "mean_token_accuracy": 0.18448725938796998,
      "num_tokens": 56419206.0,
      "step": 25750
    },
    {
      "entropy": 5.738000011444091,
      "epoch": 2.5560738388249304,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004350158400420266,
      "loss": 5.1524,
      "mean_token_accuracy": 0.17645785212516785,
      "num_tokens": 56431774.0,
      "step": 25755
    },
    {
      "entropy": 5.702668571472168,
      "epoch": 2.556570067487098,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043499092941753003,
      "loss": 5.0452,
      "mean_token_accuracy": 0.1795080229640007,
      "num_tokens": 56442048.0,
      "step": 25760
    },
    {
      "entropy": 5.638026762008667,
      "epoch": 2.557066296149266,
      "grad_norm": 1.125,
      "learning_rate": 0.00043496601482560327,
      "loss": 5.0266,
      "mean_token_accuracy": 0.17743476927280427,
      "num_tokens": 56453347.0,
      "step": 25765
    },
    {
      "entropy": 5.583108139038086,
      "epoch": 2.5575625248114333,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043494109626686394,
      "loss": 4.9916,
      "mean_token_accuracy": 0.1928045704960823,
      "num_tokens": 56465323.0,
      "step": 25770
    },
    {
      "entropy": 5.652088403701782,
      "epoch": 2.558058753473601,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043491617374193017,
      "loss": 5.0467,
      "mean_token_accuracy": 0.18298674672842025,
      "num_tokens": 56474704.0,
      "step": 25775
    },
    {
      "entropy": 5.654673385620117,
      "epoch": 2.5585549821357683,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004348912472514197,
      "loss": 5.0578,
      "mean_token_accuracy": 0.18629671782255172,
      "num_tokens": 56485028.0,
      "step": 25780
    },
    {
      "entropy": 5.648770761489868,
      "epoch": 2.5590512107979357,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043486631679595097,
      "loss": 4.9577,
      "mean_token_accuracy": 0.19094264805316924,
      "num_tokens": 56496169.0,
      "step": 25785
    },
    {
      "entropy": 5.579991245269776,
      "epoch": 2.559547439460103,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004348413823761421,
      "loss": 4.9572,
      "mean_token_accuracy": 0.18738566637039183,
      "num_tokens": 56506324.0,
      "step": 25790
    },
    {
      "entropy": 5.593776035308838,
      "epoch": 2.5600436681222707,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004348164439926112,
      "loss": 4.9382,
      "mean_token_accuracy": 0.1918426901102066,
      "num_tokens": 56516901.0,
      "step": 25795
    },
    {
      "entropy": 5.56903305053711,
      "epoch": 2.560539896784438,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043479150164597695,
      "loss": 4.9371,
      "mean_token_accuracy": 0.18670898526906968,
      "num_tokens": 56527241.0,
      "step": 25800
    },
    {
      "entropy": 5.670966625213623,
      "epoch": 2.5610361254466056,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043476655533685783,
      "loss": 5.1369,
      "mean_token_accuracy": 0.1806335374712944,
      "num_tokens": 56539078.0,
      "step": 25805
    },
    {
      "entropy": 5.627170848846435,
      "epoch": 2.561532354108773,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043474160506587234,
      "loss": 4.9814,
      "mean_token_accuracy": 0.1858750656247139,
      "num_tokens": 56551141.0,
      "step": 25810
    },
    {
      "entropy": 5.664537715911865,
      "epoch": 2.5620285827709406,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004347166508336393,
      "loss": 5.1037,
      "mean_token_accuracy": 0.1792888358235359,
      "num_tokens": 56561836.0,
      "step": 25815
    },
    {
      "entropy": 5.66357102394104,
      "epoch": 2.5625248114331085,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043469169264077754,
      "loss": 5.0084,
      "mean_token_accuracy": 0.18140707910060883,
      "num_tokens": 56572699.0,
      "step": 25820
    },
    {
      "entropy": 5.63145227432251,
      "epoch": 2.563021040095276,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004346667304879058,
      "loss": 5.091,
      "mean_token_accuracy": 0.18179389983415603,
      "num_tokens": 56582296.0,
      "step": 25825
    },
    {
      "entropy": 5.623054170608521,
      "epoch": 2.5635172687574435,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043464176437564317,
      "loss": 5.0275,
      "mean_token_accuracy": 0.1806883454322815,
      "num_tokens": 56592907.0,
      "step": 25830
    },
    {
      "entropy": 5.615181589126587,
      "epoch": 2.564013497419611,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004346167943046089,
      "loss": 5.0108,
      "mean_token_accuracy": 0.18820158690214156,
      "num_tokens": 56604995.0,
      "step": 25835
    },
    {
      "entropy": 5.6438665866851805,
      "epoch": 2.5645097260817784,
      "grad_norm": 1.015625,
      "learning_rate": 0.000434591820275422,
      "loss": 4.9843,
      "mean_token_accuracy": 0.18526272028684615,
      "num_tokens": 56614809.0,
      "step": 25840
    },
    {
      "entropy": 5.674536037445068,
      "epoch": 2.565005954743946,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004345668422887018,
      "loss": 5.0612,
      "mean_token_accuracy": 0.1752903312444687,
      "num_tokens": 56624409.0,
      "step": 25845
    },
    {
      "entropy": 5.629859733581543,
      "epoch": 2.565502183406114,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004345418603450678,
      "loss": 5.0505,
      "mean_token_accuracy": 0.1800534024834633,
      "num_tokens": 56634596.0,
      "step": 25850
    },
    {
      "entropy": 5.630793142318725,
      "epoch": 2.5659984120682813,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004345168744451395,
      "loss": 5.027,
      "mean_token_accuracy": 0.18169807344675065,
      "num_tokens": 56646710.0,
      "step": 25855
    },
    {
      "entropy": 5.642646312713623,
      "epoch": 2.5664946407304488,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004344918845895364,
      "loss": 5.0156,
      "mean_token_accuracy": 0.1829175278544426,
      "num_tokens": 56658427.0,
      "step": 25860
    },
    {
      "entropy": 5.665085935592652,
      "epoch": 2.5669908693926162,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004344668907788782,
      "loss": 5.0864,
      "mean_token_accuracy": 0.17776597291231155,
      "num_tokens": 56669039.0,
      "step": 25865
    },
    {
      "entropy": 5.610577583312988,
      "epoch": 2.5674870980547837,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004344418930137848,
      "loss": 5.0055,
      "mean_token_accuracy": 0.18383473604917527,
      "num_tokens": 56679063.0,
      "step": 25870
    },
    {
      "entropy": 5.679255867004395,
      "epoch": 2.567983326716951,
      "grad_norm": 1.125,
      "learning_rate": 0.0004344168912948759,
      "loss": 5.0186,
      "mean_token_accuracy": 0.18382539004087448,
      "num_tokens": 56690142.0,
      "step": 25875
    },
    {
      "entropy": 5.650301599502564,
      "epoch": 2.5684795553791187,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043439188562277174,
      "loss": 5.0563,
      "mean_token_accuracy": 0.18120356202125548,
      "num_tokens": 56700918.0,
      "step": 25880
    },
    {
      "entropy": 5.764885187149048,
      "epoch": 2.568975784041286,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043436687599809226,
      "loss": 5.1809,
      "mean_token_accuracy": 0.1719437286257744,
      "num_tokens": 56713520.0,
      "step": 25885
    },
    {
      "entropy": 5.609647750854492,
      "epoch": 2.5694720127034536,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043434186242145763,
      "loss": 5.0782,
      "mean_token_accuracy": 0.17988628149032593,
      "num_tokens": 56723428.0,
      "step": 25890
    },
    {
      "entropy": 5.657195043563843,
      "epoch": 2.569968241365621,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004343168448934881,
      "loss": 5.0074,
      "mean_token_accuracy": 0.18499251157045365,
      "num_tokens": 56734167.0,
      "step": 25895
    },
    {
      "entropy": 5.674086618423462,
      "epoch": 2.5704644700277886,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043429182341480423,
      "loss": 4.9612,
      "mean_token_accuracy": 0.1836756780743599,
      "num_tokens": 56745196.0,
      "step": 25900
    },
    {
      "entropy": 5.6189594745635985,
      "epoch": 2.5709606986899565,
      "grad_norm": 1.125,
      "learning_rate": 0.0004342667979860263,
      "loss": 4.965,
      "mean_token_accuracy": 0.18401169329881667,
      "num_tokens": 56756166.0,
      "step": 25905
    },
    {
      "entropy": 5.575355768203735,
      "epoch": 2.571456927352124,
      "grad_norm": 1.125,
      "learning_rate": 0.00043424176860777495,
      "loss": 4.9921,
      "mean_token_accuracy": 0.18223308622837067,
      "num_tokens": 56766690.0,
      "step": 25910
    },
    {
      "entropy": 5.620118951797485,
      "epoch": 2.5719531560142914,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043421673528067095,
      "loss": 5.0549,
      "mean_token_accuracy": 0.17688563615083694,
      "num_tokens": 56776995.0,
      "step": 25915
    },
    {
      "entropy": 5.7313093662261965,
      "epoch": 2.572449384676459,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043419169800533496,
      "loss": 5.1069,
      "mean_token_accuracy": 0.18173252940177917,
      "num_tokens": 56788509.0,
      "step": 25920
    },
    {
      "entropy": 5.678748655319214,
      "epoch": 2.5729456133386264,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004341666567823879,
      "loss": 5.0947,
      "mean_token_accuracy": 0.17574068009853364,
      "num_tokens": 56799027.0,
      "step": 25925
    },
    {
      "entropy": 5.69102783203125,
      "epoch": 2.573441842000794,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004341416116124508,
      "loss": 5.0591,
      "mean_token_accuracy": 0.18482033163309097,
      "num_tokens": 56809796.0,
      "step": 25930
    },
    {
      "entropy": 5.664019250869751,
      "epoch": 2.5739380706629618,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004341165624961445,
      "loss": 5.0582,
      "mean_token_accuracy": 0.1784698486328125,
      "num_tokens": 56820523.0,
      "step": 25935
    },
    {
      "entropy": 5.638457250595093,
      "epoch": 2.5744342993251292,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043409150943409035,
      "loss": 5.0439,
      "mean_token_accuracy": 0.17876005172729492,
      "num_tokens": 56831668.0,
      "step": 25940
    },
    {
      "entropy": 5.57755560874939,
      "epoch": 2.5749305279872967,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043406645242690967,
      "loss": 4.9372,
      "mean_token_accuracy": 0.1857502728700638,
      "num_tokens": 56841950.0,
      "step": 25945
    },
    {
      "entropy": 5.628582096099853,
      "epoch": 2.575426756649464,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004340413914752237,
      "loss": 5.0586,
      "mean_token_accuracy": 0.17959801554679872,
      "num_tokens": 56852190.0,
      "step": 25950
    },
    {
      "entropy": 5.723267030715943,
      "epoch": 2.5759229853116317,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004340163265796539,
      "loss": 5.0788,
      "mean_token_accuracy": 0.1800284504890442,
      "num_tokens": 56864255.0,
      "step": 25955
    },
    {
      "entropy": 5.674472284317017,
      "epoch": 2.576419213973799,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004339912577408218,
      "loss": 4.9941,
      "mean_token_accuracy": 0.1839619070291519,
      "num_tokens": 56874711.0,
      "step": 25960
    },
    {
      "entropy": 5.5428613185882565,
      "epoch": 2.5769154426359666,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004339661849593492,
      "loss": 4.9739,
      "mean_token_accuracy": 0.1818840906023979,
      "num_tokens": 56885790.0,
      "step": 25965
    },
    {
      "entropy": 5.6394147872924805,
      "epoch": 2.577411671298134,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004339411082358577,
      "loss": 5.0401,
      "mean_token_accuracy": 0.1852944701910019,
      "num_tokens": 56896452.0,
      "step": 25970
    },
    {
      "entropy": 5.671448230743408,
      "epoch": 2.5779078999603016,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00043391602757096923,
      "loss": 5.145,
      "mean_token_accuracy": 0.18198729753494264,
      "num_tokens": 56908216.0,
      "step": 25975
    },
    {
      "entropy": 5.68589677810669,
      "epoch": 2.578404128622469,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043389094296530565,
      "loss": 5.0406,
      "mean_token_accuracy": 0.17234701067209243,
      "num_tokens": 56918346.0,
      "step": 25980
    },
    {
      "entropy": 5.653329610824585,
      "epoch": 2.5789003572846365,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004338658544194891,
      "loss": 5.1565,
      "mean_token_accuracy": 0.1723212942481041,
      "num_tokens": 56930950.0,
      "step": 25985
    },
    {
      "entropy": 5.666849756240845,
      "epoch": 2.5793965859468044,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043384076193414176,
      "loss": 5.0729,
      "mean_token_accuracy": 0.18263812065124513,
      "num_tokens": 56940622.0,
      "step": 25990
    },
    {
      "entropy": 5.657256746292115,
      "epoch": 2.579892814608972,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043381566550988564,
      "loss": 5.0207,
      "mean_token_accuracy": 0.183506079018116,
      "num_tokens": 56952463.0,
      "step": 25995
    },
    {
      "entropy": 5.618536758422851,
      "epoch": 2.5803890432711394,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043379056514734334,
      "loss": 4.9779,
      "mean_token_accuracy": 0.1854998230934143,
      "num_tokens": 56961705.0,
      "step": 26000
    },
    {
      "entropy": 5.640136241912842,
      "epoch": 2.580885271933307,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004337654608471372,
      "loss": 5.0542,
      "mean_token_accuracy": 0.17673154175281525,
      "num_tokens": 56972359.0,
      "step": 26005
    },
    {
      "entropy": 5.676950359344483,
      "epoch": 2.5813815005954743,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043374035260988964,
      "loss": 4.999,
      "mean_token_accuracy": 0.1879095584154129,
      "num_tokens": 56983342.0,
      "step": 26010
    },
    {
      "entropy": 5.6088659286499025,
      "epoch": 2.581877729257642,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043371524043622344,
      "loss": 5.0013,
      "mean_token_accuracy": 0.18785511553287507,
      "num_tokens": 56993859.0,
      "step": 26015
    },
    {
      "entropy": 5.678361558914185,
      "epoch": 2.5823739579198093,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004336901243267614,
      "loss": 5.1211,
      "mean_token_accuracy": 0.18075122982263564,
      "num_tokens": 57006631.0,
      "step": 26020
    },
    {
      "entropy": 5.704943323135376,
      "epoch": 2.582870186581977,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000433665004282126,
      "loss": 5.0564,
      "mean_token_accuracy": 0.17765063643455506,
      "num_tokens": 57018121.0,
      "step": 26025
    },
    {
      "entropy": 5.646619892120361,
      "epoch": 2.5833664152441447,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00043363988030294055,
      "loss": 5.0048,
      "mean_token_accuracy": 0.1818745732307434,
      "num_tokens": 57028127.0,
      "step": 26030
    },
    {
      "entropy": 5.6307765483856205,
      "epoch": 2.583862643906312,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004336147523898279,
      "loss": 4.9846,
      "mean_token_accuracy": 0.1882005140185356,
      "num_tokens": 57038523.0,
      "step": 26035
    },
    {
      "entropy": 5.611447668075561,
      "epoch": 2.5843588725684796,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004335896205434111,
      "loss": 4.9961,
      "mean_token_accuracy": 0.18474855422973632,
      "num_tokens": 57048641.0,
      "step": 26040
    },
    {
      "entropy": 5.575924921035766,
      "epoch": 2.584855101230647,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043356448476431354,
      "loss": 4.9057,
      "mean_token_accuracy": 0.18884123265743255,
      "num_tokens": 57058794.0,
      "step": 26045
    },
    {
      "entropy": 5.64937539100647,
      "epoch": 2.5853513298928146,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004335393450531583,
      "loss": 5.041,
      "mean_token_accuracy": 0.18126943558454514,
      "num_tokens": 57069814.0,
      "step": 26050
    },
    {
      "entropy": 5.741694736480713,
      "epoch": 2.585847558554982,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000433514201410569,
      "loss": 5.1073,
      "mean_token_accuracy": 0.18639199286699296,
      "num_tokens": 57081053.0,
      "step": 26055
    },
    {
      "entropy": 5.617356824874878,
      "epoch": 2.5863437872171495,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043348905383716905,
      "loss": 5.0055,
      "mean_token_accuracy": 0.18497175872325897,
      "num_tokens": 57091386.0,
      "step": 26060
    },
    {
      "entropy": 5.626700687408447,
      "epoch": 2.586840015879317,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043346390233358213,
      "loss": 4.9836,
      "mean_token_accuracy": 0.17986669689416884,
      "num_tokens": 57102659.0,
      "step": 26065
    },
    {
      "entropy": 5.6781408309936525,
      "epoch": 2.5873362445414845,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004334387469004319,
      "loss": 5.0444,
      "mean_token_accuracy": 0.1834562361240387,
      "num_tokens": 57114257.0,
      "step": 26070
    },
    {
      "entropy": 5.619673204421997,
      "epoch": 2.5878324732036524,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004334135875383421,
      "loss": 5.069,
      "mean_token_accuracy": 0.18354780673980714,
      "num_tokens": 57125037.0,
      "step": 26075
    },
    {
      "entropy": 5.605981969833374,
      "epoch": 2.58832870186582,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004333884242479367,
      "loss": 5.0386,
      "mean_token_accuracy": 0.17841214388608934,
      "num_tokens": 57137581.0,
      "step": 26080
    },
    {
      "entropy": 5.63468451499939,
      "epoch": 2.5888249305279873,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043336325702983966,
      "loss": 4.9859,
      "mean_token_accuracy": 0.18693249821662902,
      "num_tokens": 57148507.0,
      "step": 26085
    },
    {
      "entropy": 5.656753253936768,
      "epoch": 2.589321159190155,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043333808588467504,
      "loss": 4.959,
      "mean_token_accuracy": 0.1899676650762558,
      "num_tokens": 57159518.0,
      "step": 26090
    },
    {
      "entropy": 5.610537099838257,
      "epoch": 2.5898173878523223,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004333129108130671,
      "loss": 4.9933,
      "mean_token_accuracy": 0.1868237793445587,
      "num_tokens": 57170600.0,
      "step": 26095
    },
    {
      "entropy": 5.632163143157959,
      "epoch": 2.5903136165144898,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004332877318156402,
      "loss": 5.0196,
      "mean_token_accuracy": 0.18053686171770095,
      "num_tokens": 57181137.0,
      "step": 26100
    },
    {
      "entropy": 5.6217942237854,
      "epoch": 2.5908098451766572,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00043326254889301856,
      "loss": 5.0295,
      "mean_token_accuracy": 0.18082537949085237,
      "num_tokens": 57191975.0,
      "step": 26105
    },
    {
      "entropy": 5.721718835830688,
      "epoch": 2.591306073838825,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004332373620458268,
      "loss": 5.0687,
      "mean_token_accuracy": 0.17921004295349122,
      "num_tokens": 57203139.0,
      "step": 26110
    },
    {
      "entropy": 5.6872968673706055,
      "epoch": 2.5918023025009926,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043321217127468935,
      "loss": 5.1032,
      "mean_token_accuracy": 0.17614084482192993,
      "num_tokens": 57215542.0,
      "step": 26115
    },
    {
      "entropy": 5.661975479125976,
      "epoch": 2.59229853116316,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043318697658023097,
      "loss": 5.0241,
      "mean_token_accuracy": 0.18192309141159058,
      "num_tokens": 57226525.0,
      "step": 26120
    },
    {
      "entropy": 5.625637006759644,
      "epoch": 2.5927947598253276,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004331617779630765,
      "loss": 4.9688,
      "mean_token_accuracy": 0.18832985311746597,
      "num_tokens": 57236120.0,
      "step": 26125
    },
    {
      "entropy": 5.550023412704467,
      "epoch": 2.593290988487495,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004331365754238507,
      "loss": 4.9452,
      "mean_token_accuracy": 0.18387431651353836,
      "num_tokens": 57246947.0,
      "step": 26130
    },
    {
      "entropy": 5.615897560119629,
      "epoch": 2.5937872171496625,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043311136896317855,
      "loss": 4.9722,
      "mean_token_accuracy": 0.1806974783539772,
      "num_tokens": 57258885.0,
      "step": 26135
    },
    {
      "entropy": 5.686323261260986,
      "epoch": 2.59428344581183,
      "grad_norm": 1.125,
      "learning_rate": 0.00043308615858168517,
      "loss": 5.0587,
      "mean_token_accuracy": 0.1854558452963829,
      "num_tokens": 57270047.0,
      "step": 26140
    },
    {
      "entropy": 5.684199094772339,
      "epoch": 2.5947796744739975,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004330609442799957,
      "loss": 5.1219,
      "mean_token_accuracy": 0.18213265687227248,
      "num_tokens": 57281405.0,
      "step": 26145
    },
    {
      "entropy": 5.658795976638794,
      "epoch": 2.595275903136165,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004330357260587355,
      "loss": 5.0364,
      "mean_token_accuracy": 0.18637810200452803,
      "num_tokens": 57292880.0,
      "step": 26150
    },
    {
      "entropy": 5.649238157272339,
      "epoch": 2.5957721317983324,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004330105039185297,
      "loss": 5.0031,
      "mean_token_accuracy": 0.1849469229578972,
      "num_tokens": 57304544.0,
      "step": 26155
    },
    {
      "entropy": 5.635141038894654,
      "epoch": 2.5962683604605004,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043298527786000393,
      "loss": 4.9937,
      "mean_token_accuracy": 0.18569166213274002,
      "num_tokens": 57315869.0,
      "step": 26160
    },
    {
      "entropy": 5.66668848991394,
      "epoch": 2.596764589122668,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004329600478837837,
      "loss": 5.0335,
      "mean_token_accuracy": 0.18169243931770324,
      "num_tokens": 57327933.0,
      "step": 26165
    },
    {
      "entropy": 5.673186492919922,
      "epoch": 2.5972608177848353,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043293481399049474,
      "loss": 5.0221,
      "mean_token_accuracy": 0.1790038153529167,
      "num_tokens": 57338941.0,
      "step": 26170
    },
    {
      "entropy": 5.654667854309082,
      "epoch": 2.597757046447003,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004329095761807626,
      "loss": 5.0449,
      "mean_token_accuracy": 0.18265509754419326,
      "num_tokens": 57350128.0,
      "step": 26175
    },
    {
      "entropy": 5.60224289894104,
      "epoch": 2.5982532751091703,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004328843344552133,
      "loss": 5.0006,
      "mean_token_accuracy": 0.18336853086948396,
      "num_tokens": 57361602.0,
      "step": 26180
    },
    {
      "entropy": 5.753658580780029,
      "epoch": 2.5987495037713377,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004328590888144726,
      "loss": 5.1038,
      "mean_token_accuracy": 0.17271699607372284,
      "num_tokens": 57373093.0,
      "step": 26185
    },
    {
      "entropy": 5.763119125366211,
      "epoch": 2.599245732433505,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004328338392591668,
      "loss": 5.1162,
      "mean_token_accuracy": 0.18011332005262376,
      "num_tokens": 57384799.0,
      "step": 26190
    },
    {
      "entropy": 5.612965202331543,
      "epoch": 2.599741961095673,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043280858578992185,
      "loss": 4.9656,
      "mean_token_accuracy": 0.1893203780055046,
      "num_tokens": 57395703.0,
      "step": 26195
    },
    {
      "entropy": 5.712460851669311,
      "epoch": 2.6002381897578406,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043278332840736397,
      "loss": 5.129,
      "mean_token_accuracy": 0.180124668776989,
      "num_tokens": 57406725.0,
      "step": 26200
    },
    {
      "entropy": 5.610681247711182,
      "epoch": 2.600734418420008,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043275806711211957,
      "loss": 4.9904,
      "mean_token_accuracy": 0.18704276382923127,
      "num_tokens": 57417362.0,
      "step": 26205
    },
    {
      "entropy": 5.629938459396362,
      "epoch": 2.6012306470821756,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00043273280190481503,
      "loss": 5.0249,
      "mean_token_accuracy": 0.18653296530246735,
      "num_tokens": 57429178.0,
      "step": 26210
    },
    {
      "entropy": 5.6206748485565186,
      "epoch": 2.601726875744343,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043270753278607687,
      "loss": 5.0057,
      "mean_token_accuracy": 0.18581515103578566,
      "num_tokens": 57439741.0,
      "step": 26215
    },
    {
      "entropy": 5.6102152347564695,
      "epoch": 2.6022231044065105,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043268225975653176,
      "loss": 5.0855,
      "mean_token_accuracy": 0.18122461289167405,
      "num_tokens": 57451754.0,
      "step": 26220
    },
    {
      "entropy": 5.703652048110962,
      "epoch": 2.602719333068678,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004326569828168064,
      "loss": 5.0506,
      "mean_token_accuracy": 0.17952673584222795,
      "num_tokens": 57463396.0,
      "step": 26225
    },
    {
      "entropy": 5.6464639663696286,
      "epoch": 2.6032155617308455,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004326317019675275,
      "loss": 5.0132,
      "mean_token_accuracy": 0.18401926308870314,
      "num_tokens": 57474023.0,
      "step": 26230
    },
    {
      "entropy": 5.661379051208496,
      "epoch": 2.603711790393013,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043260641720932214,
      "loss": 5.0209,
      "mean_token_accuracy": 0.18727435618638993,
      "num_tokens": 57484751.0,
      "step": 26235
    },
    {
      "entropy": 5.609542655944824,
      "epoch": 2.6042080190551804,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043258112854281726,
      "loss": 4.9981,
      "mean_token_accuracy": 0.17897092849016188,
      "num_tokens": 57496502.0,
      "step": 26240
    },
    {
      "entropy": 5.642799282073975,
      "epoch": 2.6047042477173483,
      "grad_norm": 1.125,
      "learning_rate": 0.0004325558359686399,
      "loss": 5.0221,
      "mean_token_accuracy": 0.186220583319664,
      "num_tokens": 57507111.0,
      "step": 26245
    },
    {
      "entropy": 5.669463682174682,
      "epoch": 2.605200476379516,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043253053948741734,
      "loss": 5.1054,
      "mean_token_accuracy": 0.17440075129270555,
      "num_tokens": 57517844.0,
      "step": 26250
    },
    {
      "entropy": 5.6896155834197994,
      "epoch": 2.6056967050416833,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043250523909977684,
      "loss": 5.0567,
      "mean_token_accuracy": 0.18416428864002227,
      "num_tokens": 57528047.0,
      "step": 26255
    },
    {
      "entropy": 5.706145000457764,
      "epoch": 2.6061929337038507,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004324799348063457,
      "loss": 5.0433,
      "mean_token_accuracy": 0.1807234302163124,
      "num_tokens": 57538489.0,
      "step": 26260
    },
    {
      "entropy": 5.635793209075928,
      "epoch": 2.606689162366018,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004324546266077518,
      "loss": 5.0263,
      "mean_token_accuracy": 0.18605133444070815,
      "num_tokens": 57549741.0,
      "step": 26265
    },
    {
      "entropy": 5.59465799331665,
      "epoch": 2.6071853910281857,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004324293145046222,
      "loss": 4.9642,
      "mean_token_accuracy": 0.18708259463310242,
      "num_tokens": 57559803.0,
      "step": 26270
    },
    {
      "entropy": 5.6628861904144285,
      "epoch": 2.607681619690353,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043240399849758484,
      "loss": 5.0288,
      "mean_token_accuracy": 0.17986540496349335,
      "num_tokens": 57570609.0,
      "step": 26275
    },
    {
      "entropy": 5.748559761047363,
      "epoch": 2.608177848352521,
      "grad_norm": 1.125,
      "learning_rate": 0.0004323786785872675,
      "loss": 5.0517,
      "mean_token_accuracy": 0.18534908443689346,
      "num_tokens": 57580934.0,
      "step": 26280
    },
    {
      "entropy": 5.676252317428589,
      "epoch": 2.6086740770146886,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004323533547742981,
      "loss": 5.1485,
      "mean_token_accuracy": 0.17765509933233262,
      "num_tokens": 57594671.0,
      "step": 26285
    },
    {
      "entropy": 5.679098272323609,
      "epoch": 2.609170305676856,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004323280270593046,
      "loss": 5.0846,
      "mean_token_accuracy": 0.18332365900278091,
      "num_tokens": 57605582.0,
      "step": 26290
    },
    {
      "entropy": 5.668007183074951,
      "epoch": 2.6096665343390235,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043230269544291503,
      "loss": 5.0151,
      "mean_token_accuracy": 0.17989761829376222,
      "num_tokens": 57617185.0,
      "step": 26295
    },
    {
      "entropy": 5.668362236022949,
      "epoch": 2.610162763001191,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004322773599257575,
      "loss": 4.9742,
      "mean_token_accuracy": 0.1865762233734131,
      "num_tokens": 57627205.0,
      "step": 26300
    },
    {
      "entropy": 5.575282192230224,
      "epoch": 2.6106589916633585,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043225202050846043,
      "loss": 4.971,
      "mean_token_accuracy": 0.18572307676076888,
      "num_tokens": 57639496.0,
      "step": 26305
    },
    {
      "entropy": 5.680224084854126,
      "epoch": 2.611155220325526,
      "grad_norm": 1.21875,
      "learning_rate": 0.000432226677191652,
      "loss": 5.0808,
      "mean_token_accuracy": 0.17765178978443147,
      "num_tokens": 57650345.0,
      "step": 26310
    },
    {
      "entropy": 5.638160276412964,
      "epoch": 2.6116514489876934,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043220132997596084,
      "loss": 4.9919,
      "mean_token_accuracy": 0.18384005576372148,
      "num_tokens": 57661409.0,
      "step": 26315
    },
    {
      "entropy": 5.748257303237915,
      "epoch": 2.612147677649861,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004321759788620155,
      "loss": 5.147,
      "mean_token_accuracy": 0.17649771869182587,
      "num_tokens": 57672639.0,
      "step": 26320
    },
    {
      "entropy": 5.662777471542358,
      "epoch": 2.6126439063120284,
      "grad_norm": 1.0,
      "learning_rate": 0.00043215062385044446,
      "loss": 5.1283,
      "mean_token_accuracy": 0.17275869399309157,
      "num_tokens": 57682249.0,
      "step": 26325
    },
    {
      "entropy": 5.58994688987732,
      "epoch": 2.613140134974196,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043212526494187665,
      "loss": 4.9442,
      "mean_token_accuracy": 0.19515361040830612,
      "num_tokens": 57692319.0,
      "step": 26330
    },
    {
      "entropy": 5.685469007492065,
      "epoch": 2.6136363636363638,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004320999021369408,
      "loss": 5.1229,
      "mean_token_accuracy": 0.17923817932605743,
      "num_tokens": 57702334.0,
      "step": 26335
    },
    {
      "entropy": 5.635136985778809,
      "epoch": 2.6141325922985312,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043207453543626587,
      "loss": 5.0134,
      "mean_token_accuracy": 0.1795067384839058,
      "num_tokens": 57712407.0,
      "step": 26340
    },
    {
      "entropy": 5.614953899383545,
      "epoch": 2.6146288209606987,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043204916484048094,
      "loss": 4.9832,
      "mean_token_accuracy": 0.18900940865278243,
      "num_tokens": 57724252.0,
      "step": 26345
    },
    {
      "entropy": 5.631985282897949,
      "epoch": 2.615125049622866,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043202379035021515,
      "loss": 5.0474,
      "mean_token_accuracy": 0.18177716583013534,
      "num_tokens": 57735033.0,
      "step": 26350
    },
    {
      "entropy": 5.724973106384278,
      "epoch": 2.6156212782850337,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004319984119660977,
      "loss": 5.0247,
      "mean_token_accuracy": 0.18257818073034288,
      "num_tokens": 57746182.0,
      "step": 26355
    },
    {
      "entropy": 5.6656866550445555,
      "epoch": 2.616117506947201,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043197302968875795,
      "loss": 4.9744,
      "mean_token_accuracy": 0.18615810126066207,
      "num_tokens": 57756479.0,
      "step": 26360
    },
    {
      "entropy": 5.678879165649414,
      "epoch": 2.616613735609369,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004319476435188254,
      "loss": 5.0922,
      "mean_token_accuracy": 0.17577923387289046,
      "num_tokens": 57767773.0,
      "step": 26365
    },
    {
      "entropy": 5.627799034118652,
      "epoch": 2.6171099642715365,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00043192225345692933,
      "loss": 4.9876,
      "mean_token_accuracy": 0.1884395122528076,
      "num_tokens": 57777331.0,
      "step": 26370
    },
    {
      "entropy": 5.6107423305511475,
      "epoch": 2.617606192933704,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043189685950369963,
      "loss": 5.011,
      "mean_token_accuracy": 0.1855190470814705,
      "num_tokens": 57788538.0,
      "step": 26375
    },
    {
      "entropy": 5.703800058364868,
      "epoch": 2.6181024215958715,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043187146165976577,
      "loss": 5.1541,
      "mean_token_accuracy": 0.178206904232502,
      "num_tokens": 57799983.0,
      "step": 26380
    },
    {
      "entropy": 5.64628529548645,
      "epoch": 2.618598650258039,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004318460599257578,
      "loss": 4.9834,
      "mean_token_accuracy": 0.17794771641492843,
      "num_tokens": 57810195.0,
      "step": 26385
    },
    {
      "entropy": 5.596065616607666,
      "epoch": 2.6190948789202064,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00043182065430230545,
      "loss": 4.9451,
      "mean_token_accuracy": 0.18955641239881516,
      "num_tokens": 57821930.0,
      "step": 26390
    },
    {
      "entropy": 5.6485508441925045,
      "epoch": 2.619591107582374,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004317952447900388,
      "loss": 4.9755,
      "mean_token_accuracy": 0.17947070598602294,
      "num_tokens": 57834272.0,
      "step": 26395
    },
    {
      "entropy": 5.6569465637207035,
      "epoch": 2.6200873362445414,
      "grad_norm": 1.015625,
      "learning_rate": 0.000431769831389588,
      "loss": 5.0111,
      "mean_token_accuracy": 0.18266646713018417,
      "num_tokens": 57845491.0,
      "step": 26400
    },
    {
      "entropy": 5.638046216964722,
      "epoch": 2.620583564906709,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004317444141015831,
      "loss": 5.0887,
      "mean_token_accuracy": 0.17853004038333892,
      "num_tokens": 57857816.0,
      "step": 26405
    },
    {
      "entropy": 5.632267475128174,
      "epoch": 2.6210797935688763,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043171899292665457,
      "loss": 5.0417,
      "mean_token_accuracy": 0.17907481044530868,
      "num_tokens": 57867769.0,
      "step": 26410
    },
    {
      "entropy": 5.702061748504638,
      "epoch": 2.621576022231044,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004316935678654327,
      "loss": 5.1261,
      "mean_token_accuracy": 0.17589090913534164,
      "num_tokens": 57878027.0,
      "step": 26415
    },
    {
      "entropy": 5.635748815536499,
      "epoch": 2.6220722508932117,
      "grad_norm": 1.296875,
      "learning_rate": 0.00043166813891854796,
      "loss": 5.0,
      "mean_token_accuracy": 0.18351231068372725,
      "num_tokens": 57887938.0,
      "step": 26420
    },
    {
      "entropy": 5.673457670211792,
      "epoch": 2.622568479555379,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000431642706086631,
      "loss": 5.0008,
      "mean_token_accuracy": 0.19203165620565416,
      "num_tokens": 57899375.0,
      "step": 26425
    },
    {
      "entropy": 5.683527565002441,
      "epoch": 2.6230647082175467,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043161726937031243,
      "loss": 5.082,
      "mean_token_accuracy": 0.17755960524082184,
      "num_tokens": 57910393.0,
      "step": 26430
    },
    {
      "entropy": 5.651703834533691,
      "epoch": 2.623560936879714,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000431591828770223,
      "loss": 5.0596,
      "mean_token_accuracy": 0.18411429077386857,
      "num_tokens": 57922236.0,
      "step": 26435
    },
    {
      "entropy": 5.661234951019287,
      "epoch": 2.6240571655418816,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004315663842869938,
      "loss": 5.0266,
      "mean_token_accuracy": 0.18553194254636765,
      "num_tokens": 57931804.0,
      "step": 26440
    },
    {
      "entropy": 5.679790687561035,
      "epoch": 2.624553394204049,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043154093592125554,
      "loss": 5.0126,
      "mean_token_accuracy": 0.17866526246070863,
      "num_tokens": 57943259.0,
      "step": 26445
    },
    {
      "entropy": 5.572241687774659,
      "epoch": 2.625049622866217,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043151548367363943,
      "loss": 4.9226,
      "mean_token_accuracy": 0.18880132138729094,
      "num_tokens": 57952945.0,
      "step": 26450
    },
    {
      "entropy": 5.605307006835938,
      "epoch": 2.6255458515283845,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043149002754477655,
      "loss": 5.0154,
      "mean_token_accuracy": 0.18359719216823578,
      "num_tokens": 57964030.0,
      "step": 26455
    },
    {
      "entropy": 5.619307947158814,
      "epoch": 2.626042080190552,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004314645675352982,
      "loss": 4.9783,
      "mean_token_accuracy": 0.19268855601549148,
      "num_tokens": 57974888.0,
      "step": 26460
    },
    {
      "entropy": 5.686883974075317,
      "epoch": 2.6265383088527194,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004314391036458357,
      "loss": 5.0287,
      "mean_token_accuracy": 0.18286088109016418,
      "num_tokens": 57984922.0,
      "step": 26465
    },
    {
      "entropy": 5.602847242355347,
      "epoch": 2.627034537514887,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043141363587702054,
      "loss": 4.9736,
      "mean_token_accuracy": 0.1902470752596855,
      "num_tokens": 57994881.0,
      "step": 26470
    },
    {
      "entropy": 5.660651111602784,
      "epoch": 2.6275307661770544,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004313881642294843,
      "loss": 5.0741,
      "mean_token_accuracy": 0.18624747544527054,
      "num_tokens": 58005335.0,
      "step": 26475
    },
    {
      "entropy": 5.657034683227539,
      "epoch": 2.628026994839222,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004313626887038585,
      "loss": 4.9668,
      "mean_token_accuracy": 0.18769294768571854,
      "num_tokens": 58016692.0,
      "step": 26480
    },
    {
      "entropy": 5.65866436958313,
      "epoch": 2.6285232235013893,
      "grad_norm": 1.046875,
      "learning_rate": 0.000431337209300775,
      "loss": 5.0643,
      "mean_token_accuracy": 0.18816429376602173,
      "num_tokens": 58028282.0,
      "step": 26485
    },
    {
      "entropy": 5.613208246231079,
      "epoch": 2.629019452163557,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004313117260208656,
      "loss": 4.9859,
      "mean_token_accuracy": 0.19126532226800919,
      "num_tokens": 58040108.0,
      "step": 26490
    },
    {
      "entropy": 5.6041069507598875,
      "epoch": 2.6295156808257243,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004312862388647622,
      "loss": 4.9963,
      "mean_token_accuracy": 0.18890976309776306,
      "num_tokens": 58050156.0,
      "step": 26495
    },
    {
      "entropy": 5.626929187774659,
      "epoch": 2.6300119094878918,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043126074783309687,
      "loss": 4.9915,
      "mean_token_accuracy": 0.18219767659902572,
      "num_tokens": 58060929.0,
      "step": 26500
    },
    {
      "entropy": 5.671187210083008,
      "epoch": 2.6305081381500597,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004312352529265016,
      "loss": 5.0507,
      "mean_token_accuracy": 0.17584247142076492,
      "num_tokens": 58071684.0,
      "step": 26505
    },
    {
      "entropy": 5.662590360641479,
      "epoch": 2.631004366812227,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043120975414560886,
      "loss": 5.0701,
      "mean_token_accuracy": 0.1831488773226738,
      "num_tokens": 58082730.0,
      "step": 26510
    },
    {
      "entropy": 5.717787265777588,
      "epoch": 2.6315005954743946,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043118425149105073,
      "loss": 5.1366,
      "mean_token_accuracy": 0.1750657692551613,
      "num_tokens": 58095278.0,
      "step": 26515
    },
    {
      "entropy": 5.770516014099121,
      "epoch": 2.631996824136562,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004311587449634597,
      "loss": 5.183,
      "mean_token_accuracy": 0.1759907566010952,
      "num_tokens": 58107816.0,
      "step": 26520
    },
    {
      "entropy": 5.6997006893157955,
      "epoch": 2.6324930527987296,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004311332345634683,
      "loss": 5.1116,
      "mean_token_accuracy": 0.1729532450437546,
      "num_tokens": 58117779.0,
      "step": 26525
    },
    {
      "entropy": 5.712108993530274,
      "epoch": 2.632989281460897,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043110772029170916,
      "loss": 5.0089,
      "mean_token_accuracy": 0.1905808225274086,
      "num_tokens": 58128665.0,
      "step": 26530
    },
    {
      "entropy": 5.718209314346313,
      "epoch": 2.633485510123065,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043108220214881494,
      "loss": 5.0775,
      "mean_token_accuracy": 0.1787843346595764,
      "num_tokens": 58139658.0,
      "step": 26535
    },
    {
      "entropy": 5.708709716796875,
      "epoch": 2.6339817387852325,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004310566801354184,
      "loss": 5.0752,
      "mean_token_accuracy": 0.1773894116282463,
      "num_tokens": 58149980.0,
      "step": 26540
    },
    {
      "entropy": 5.643571662902832,
      "epoch": 2.6344779674474,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043103115425215246,
      "loss": 4.9503,
      "mean_token_accuracy": 0.18484349250793458,
      "num_tokens": 58159809.0,
      "step": 26545
    },
    {
      "entropy": 5.647557926177979,
      "epoch": 2.6349741961095674,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043100562449965016,
      "loss": 5.0155,
      "mean_token_accuracy": 0.18220870345830917,
      "num_tokens": 58169912.0,
      "step": 26550
    },
    {
      "entropy": 5.675491285324097,
      "epoch": 2.635470424771735,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043098009087854446,
      "loss": 5.0707,
      "mean_token_accuracy": 0.1805168405175209,
      "num_tokens": 58179949.0,
      "step": 26555
    },
    {
      "entropy": 5.632960510253906,
      "epoch": 2.6359666534339024,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043095455338946863,
      "loss": 5.0288,
      "mean_token_accuracy": 0.18482131659984588,
      "num_tokens": 58189847.0,
      "step": 26560
    },
    {
      "entropy": 5.639557504653931,
      "epoch": 2.63646288209607,
      "grad_norm": 1.078125,
      "learning_rate": 0.000430929012033056,
      "loss": 4.9394,
      "mean_token_accuracy": 0.18983572125434875,
      "num_tokens": 58201514.0,
      "step": 26565
    },
    {
      "entropy": 5.641997337341309,
      "epoch": 2.6369591107582373,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043090346680993983,
      "loss": 5.0207,
      "mean_token_accuracy": 0.18936707824468613,
      "num_tokens": 58211940.0,
      "step": 26570
    },
    {
      "entropy": 5.686332559585571,
      "epoch": 2.6374553394204048,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004308779177207536,
      "loss": 5.0557,
      "mean_token_accuracy": 0.18060680031776427,
      "num_tokens": 58222260.0,
      "step": 26575
    },
    {
      "entropy": 5.63807635307312,
      "epoch": 2.6379515680825723,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004308523647661309,
      "loss": 5.0027,
      "mean_token_accuracy": 0.18104435056447982,
      "num_tokens": 58232649.0,
      "step": 26580
    },
    {
      "entropy": 5.617421340942383,
      "epoch": 2.6384477967447397,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043082680794670535,
      "loss": 4.9414,
      "mean_token_accuracy": 0.1907448172569275,
      "num_tokens": 58243373.0,
      "step": 26585
    },
    {
      "entropy": 5.552756690979004,
      "epoch": 2.6389440254069076,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043080124726311084,
      "loss": 5.0141,
      "mean_token_accuracy": 0.18681122213602067,
      "num_tokens": 58253863.0,
      "step": 26590
    },
    {
      "entropy": 5.6046826362609865,
      "epoch": 2.639440254069075,
      "grad_norm": 1.125,
      "learning_rate": 0.0004307756827159811,
      "loss": 5.0963,
      "mean_token_accuracy": 0.18623462468385696,
      "num_tokens": 58265326.0,
      "step": 26595
    },
    {
      "entropy": 5.7173991203308105,
      "epoch": 2.6399364827312426,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043075011430594994,
      "loss": 5.0765,
      "mean_token_accuracy": 0.18527820855379104,
      "num_tokens": 58276006.0,
      "step": 26600
    },
    {
      "entropy": 5.636649465560913,
      "epoch": 2.64043271139341,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043072454203365175,
      "loss": 4.9303,
      "mean_token_accuracy": 0.18744234293699263,
      "num_tokens": 58285932.0,
      "step": 26605
    },
    {
      "entropy": 5.62972207069397,
      "epoch": 2.6409289400555775,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004306989658997204,
      "loss": 5.0093,
      "mean_token_accuracy": 0.18516791462898255,
      "num_tokens": 58297174.0,
      "step": 26610
    },
    {
      "entropy": 5.706906318664551,
      "epoch": 2.641425168717745,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043067338590479023,
      "loss": 5.1439,
      "mean_token_accuracy": 0.1742935761809349,
      "num_tokens": 58308509.0,
      "step": 26615
    },
    {
      "entropy": 5.663527870178223,
      "epoch": 2.641921397379913,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043064780204949543,
      "loss": 5.0469,
      "mean_token_accuracy": 0.1746008187532425,
      "num_tokens": 58318449.0,
      "step": 26620
    },
    {
      "entropy": 5.7033754825592045,
      "epoch": 2.6424176260420804,
      "grad_norm": 1.0,
      "learning_rate": 0.0004306222143344706,
      "loss": 5.0152,
      "mean_token_accuracy": 0.18325702100992203,
      "num_tokens": 58329759.0,
      "step": 26625
    },
    {
      "entropy": 5.633621644973755,
      "epoch": 2.642913854704248,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004305966227603503,
      "loss": 4.9964,
      "mean_token_accuracy": 0.18427804708480836,
      "num_tokens": 58339363.0,
      "step": 26630
    },
    {
      "entropy": 5.602119874954224,
      "epoch": 2.6434100833664154,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004305710273277689,
      "loss": 5.0246,
      "mean_token_accuracy": 0.17671292871236802,
      "num_tokens": 58349482.0,
      "step": 26635
    },
    {
      "entropy": 5.657049512863159,
      "epoch": 2.643906312028583,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043054542803736126,
      "loss": 5.0587,
      "mean_token_accuracy": 0.1827228620648384,
      "num_tokens": 58359879.0,
      "step": 26640
    },
    {
      "entropy": 5.715927314758301,
      "epoch": 2.6444025406907503,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043051982488976216,
      "loss": 5.1082,
      "mean_token_accuracy": 0.17418149411678313,
      "num_tokens": 58370475.0,
      "step": 26645
    },
    {
      "entropy": 5.695374250411987,
      "epoch": 2.644898769352918,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004304942178856066,
      "loss": 5.0297,
      "mean_token_accuracy": 0.1821875587105751,
      "num_tokens": 58381992.0,
      "step": 26650
    },
    {
      "entropy": 5.673343276977539,
      "epoch": 2.6453949980150853,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043046860702552944,
      "loss": 4.9879,
      "mean_token_accuracy": 0.18940699994564056,
      "num_tokens": 58391616.0,
      "step": 26655
    },
    {
      "entropy": 5.696571159362793,
      "epoch": 2.6458912266772527,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004304429923101658,
      "loss": 4.9987,
      "mean_token_accuracy": 0.18624266535043715,
      "num_tokens": 58401800.0,
      "step": 26660
    },
    {
      "entropy": 5.643420934677124,
      "epoch": 2.64638745533942,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000430417373740151,
      "loss": 4.9938,
      "mean_token_accuracy": 0.1875233992934227,
      "num_tokens": 58413376.0,
      "step": 26665
    },
    {
      "entropy": 5.538860654830932,
      "epoch": 2.6468836840015877,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004303917513161201,
      "loss": 4.9566,
      "mean_token_accuracy": 0.19139569252729416,
      "num_tokens": 58423772.0,
      "step": 26670
    },
    {
      "entropy": 5.6369987487792965,
      "epoch": 2.6473799126637556,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004303661250387088,
      "loss": 5.0615,
      "mean_token_accuracy": 0.18004752993583678,
      "num_tokens": 58434838.0,
      "step": 26675
    },
    {
      "entropy": 5.6821331024169925,
      "epoch": 2.647876141325923,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004303404949085522,
      "loss": 5.0869,
      "mean_token_accuracy": 0.17921387255191804,
      "num_tokens": 58445036.0,
      "step": 26680
    },
    {
      "entropy": 5.677316522598266,
      "epoch": 2.6483723699880906,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004303148609262861,
      "loss": 5.0206,
      "mean_token_accuracy": 0.18254247456789016,
      "num_tokens": 58455558.0,
      "step": 26685
    },
    {
      "entropy": 5.653153991699218,
      "epoch": 2.648868598650258,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004302892230925462,
      "loss": 5.0483,
      "mean_token_accuracy": 0.17481958568096162,
      "num_tokens": 58467716.0,
      "step": 26690
    },
    {
      "entropy": 5.611427545547485,
      "epoch": 2.6493648273124255,
      "grad_norm": 0.9375,
      "learning_rate": 0.00043026358140796806,
      "loss": 5.0335,
      "mean_token_accuracy": 0.18112315237522125,
      "num_tokens": 58479421.0,
      "step": 26695
    },
    {
      "entropy": 5.581361246109009,
      "epoch": 2.649861055974593,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004302379358731878,
      "loss": 5.011,
      "mean_token_accuracy": 0.18669814765453338,
      "num_tokens": 58490115.0,
      "step": 26700
    },
    {
      "entropy": 5.6189947605133055,
      "epoch": 2.6503572846367605,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004302122864888411,
      "loss": 4.985,
      "mean_token_accuracy": 0.18563283532857894,
      "num_tokens": 58501711.0,
      "step": 26705
    },
    {
      "entropy": 5.681448268890381,
      "epoch": 2.6508535132989284,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004301866332555643,
      "loss": 5.0028,
      "mean_token_accuracy": 0.19268976300954818,
      "num_tokens": 58513441.0,
      "step": 26710
    },
    {
      "entropy": 5.660407209396363,
      "epoch": 2.651349741961096,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043016097617399327,
      "loss": 4.9839,
      "mean_token_accuracy": 0.1857868865132332,
      "num_tokens": 58523940.0,
      "step": 26715
    },
    {
      "entropy": 5.632970666885376,
      "epoch": 2.6518459706232633,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004301353152447644,
      "loss": 5.0585,
      "mean_token_accuracy": 0.17856046110391616,
      "num_tokens": 58536354.0,
      "step": 26720
    },
    {
      "entropy": 5.671811151504516,
      "epoch": 2.652342199285431,
      "grad_norm": 1.484375,
      "learning_rate": 0.00043010965046851404,
      "loss": 5.0652,
      "mean_token_accuracy": 0.1821175456047058,
      "num_tokens": 58547096.0,
      "step": 26725
    },
    {
      "entropy": 5.66640682220459,
      "epoch": 2.6528384279475983,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043008398184587847,
      "loss": 5.0624,
      "mean_token_accuracy": 0.17939624041318894,
      "num_tokens": 58557951.0,
      "step": 26730
    },
    {
      "entropy": 5.635880041122436,
      "epoch": 2.6533346566097658,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043005830937749447,
      "loss": 4.9812,
      "mean_token_accuracy": 0.18721853941679,
      "num_tokens": 58569384.0,
      "step": 26735
    },
    {
      "entropy": 5.6588475704193115,
      "epoch": 2.6538308852719332,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004300326330639984,
      "loss": 5.0329,
      "mean_token_accuracy": 0.18160244673490525,
      "num_tokens": 58580308.0,
      "step": 26740
    },
    {
      "entropy": 5.64485764503479,
      "epoch": 2.6543271139341007,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004300069529060272,
      "loss": 5.0415,
      "mean_token_accuracy": 0.17842784374952317,
      "num_tokens": 58591780.0,
      "step": 26745
    },
    {
      "entropy": 5.636948537826538,
      "epoch": 2.654823342596268,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042998126890421754,
      "loss": 5.054,
      "mean_token_accuracy": 0.18318145871162414,
      "num_tokens": 58601477.0,
      "step": 26750
    },
    {
      "entropy": 5.6472385883331295,
      "epoch": 2.6553195712584357,
      "grad_norm": 1.125,
      "learning_rate": 0.00042995558105920627,
      "loss": 5.03,
      "mean_token_accuracy": 0.1866440251469612,
      "num_tokens": 58612531.0,
      "step": 26755
    },
    {
      "entropy": 5.662769174575805,
      "epoch": 2.6558157999206036,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042992988937163066,
      "loss": 5.0781,
      "mean_token_accuracy": 0.1759551614522934,
      "num_tokens": 58624951.0,
      "step": 26760
    },
    {
      "entropy": 5.646811056137085,
      "epoch": 2.656312028582771,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004299041938421275,
      "loss": 5.1121,
      "mean_token_accuracy": 0.18012801706790924,
      "num_tokens": 58636484.0,
      "step": 26765
    },
    {
      "entropy": 5.677213382720947,
      "epoch": 2.6568082572449385,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042987849447133423,
      "loss": 5.0304,
      "mean_token_accuracy": 0.18039364963769913,
      "num_tokens": 58647386.0,
      "step": 26770
    },
    {
      "entropy": 5.700357818603516,
      "epoch": 2.657304485907106,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004298527912598879,
      "loss": 5.0974,
      "mean_token_accuracy": 0.1795784279704094,
      "num_tokens": 58658656.0,
      "step": 26775
    },
    {
      "entropy": 5.6745037078857425,
      "epoch": 2.6578007145692735,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004298270842084261,
      "loss": 4.9819,
      "mean_token_accuracy": 0.18868981599807738,
      "num_tokens": 58669901.0,
      "step": 26780
    },
    {
      "entropy": 5.626323938369751,
      "epoch": 2.658296943231441,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042980137331758626,
      "loss": 5.0916,
      "mean_token_accuracy": 0.17607760131359101,
      "num_tokens": 58680849.0,
      "step": 26785
    },
    {
      "entropy": 5.5905022621154785,
      "epoch": 2.6587931718936084,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042977565858800593,
      "loss": 4.9926,
      "mean_token_accuracy": 0.18029944300651551,
      "num_tokens": 58690642.0,
      "step": 26790
    },
    {
      "entropy": 5.60187406539917,
      "epoch": 2.6592894005557763,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042974994002032275,
      "loss": 4.9556,
      "mean_token_accuracy": 0.1910325765609741,
      "num_tokens": 58700480.0,
      "step": 26795
    },
    {
      "entropy": 5.673684024810791,
      "epoch": 2.659785629217944,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042972421761517456,
      "loss": 5.0792,
      "mean_token_accuracy": 0.17593178600072862,
      "num_tokens": 58712443.0,
      "step": 26800
    },
    {
      "entropy": 5.617500257492066,
      "epoch": 2.6602818578801113,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004296984913731991,
      "loss": 4.987,
      "mean_token_accuracy": 0.18259589225053788,
      "num_tokens": 58721172.0,
      "step": 26805
    },
    {
      "entropy": 5.629955816268921,
      "epoch": 2.6607780865422788,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004296727612950344,
      "loss": 5.0252,
      "mean_token_accuracy": 0.18564795851707458,
      "num_tokens": 58731060.0,
      "step": 26810
    },
    {
      "entropy": 5.626841640472412,
      "epoch": 2.6612743152044462,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004296470273813186,
      "loss": 5.0373,
      "mean_token_accuracy": 0.18376408219337464,
      "num_tokens": 58742189.0,
      "step": 26815
    },
    {
      "entropy": 5.712163305282592,
      "epoch": 2.6617705438666137,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004296212896326896,
      "loss": 5.0511,
      "mean_token_accuracy": 0.18378609865903855,
      "num_tokens": 58753617.0,
      "step": 26820
    },
    {
      "entropy": 5.629521369934082,
      "epoch": 2.662266772528781,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042959554804978596,
      "loss": 5.0538,
      "mean_token_accuracy": 0.18863819986581803,
      "num_tokens": 58765299.0,
      "step": 26825
    },
    {
      "entropy": 5.704364347457886,
      "epoch": 2.6627630011909487,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042956980263324576,
      "loss": 4.9735,
      "mean_token_accuracy": 0.18708151578903198,
      "num_tokens": 58775280.0,
      "step": 26830
    },
    {
      "entropy": 5.69110426902771,
      "epoch": 2.663259229853116,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004295440533837075,
      "loss": 4.9852,
      "mean_token_accuracy": 0.18825025707483292,
      "num_tokens": 58786078.0,
      "step": 26835
    },
    {
      "entropy": 5.681525659561157,
      "epoch": 2.6637554585152836,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042951830030180983,
      "loss": 5.0491,
      "mean_token_accuracy": 0.184369096159935,
      "num_tokens": 58796665.0,
      "step": 26840
    },
    {
      "entropy": 5.684406805038452,
      "epoch": 2.6642516871774515,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042949254338819123,
      "loss": 5.0626,
      "mean_token_accuracy": 0.1836922749876976,
      "num_tokens": 58808150.0,
      "step": 26845
    },
    {
      "entropy": 5.693397188186646,
      "epoch": 2.664747915839619,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042946678264349036,
      "loss": 5.0104,
      "mean_token_accuracy": 0.18547731637954712,
      "num_tokens": 58818278.0,
      "step": 26850
    },
    {
      "entropy": 5.637503957748413,
      "epoch": 2.6652441445017865,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042944101806834627,
      "loss": 5.048,
      "mean_token_accuracy": 0.18037547767162324,
      "num_tokens": 58828847.0,
      "step": 26855
    },
    {
      "entropy": 5.6185003280639645,
      "epoch": 2.665740373163954,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004294152496633977,
      "loss": 5.0174,
      "mean_token_accuracy": 0.18187395185232164,
      "num_tokens": 58839698.0,
      "step": 26860
    },
    {
      "entropy": 5.611974287033081,
      "epoch": 2.6662366018261214,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042938947742928357,
      "loss": 4.9911,
      "mean_token_accuracy": 0.18354163318872452,
      "num_tokens": 58850794.0,
      "step": 26865
    },
    {
      "entropy": 5.664192199707031,
      "epoch": 2.666732830488289,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042936370136664314,
      "loss": 5.0693,
      "mean_token_accuracy": 0.1788004994392395,
      "num_tokens": 58861590.0,
      "step": 26870
    },
    {
      "entropy": 5.686606073379517,
      "epoch": 2.6672290591504564,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004293379214761156,
      "loss": 5.0985,
      "mean_token_accuracy": 0.1752436026930809,
      "num_tokens": 58873796.0,
      "step": 26875
    },
    {
      "entropy": 5.6882020950317385,
      "epoch": 2.6677252878126243,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004293121377583401,
      "loss": 5.0488,
      "mean_token_accuracy": 0.18348691910505294,
      "num_tokens": 58885296.0,
      "step": 26880
    },
    {
      "entropy": 5.652047395706177,
      "epoch": 2.6682215164747918,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004292863502139561,
      "loss": 5.0964,
      "mean_token_accuracy": 0.18005965650081635,
      "num_tokens": 58897764.0,
      "step": 26885
    },
    {
      "entropy": 5.681487083435059,
      "epoch": 2.6687177451369593,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004292605588436031,
      "loss": 5.0599,
      "mean_token_accuracy": 0.17666542232036592,
      "num_tokens": 58908964.0,
      "step": 26890
    },
    {
      "entropy": 5.560976648330689,
      "epoch": 2.6692139737991267,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00042923476364792063,
      "loss": 4.9139,
      "mean_token_accuracy": 0.18897676467895508,
      "num_tokens": 58919152.0,
      "step": 26895
    },
    {
      "entropy": 5.568503427505493,
      "epoch": 2.669710202461294,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004292089646275484,
      "loss": 5.0447,
      "mean_token_accuracy": 0.1844946026802063,
      "num_tokens": 58928849.0,
      "step": 26900
    },
    {
      "entropy": 5.7225817203521725,
      "epoch": 2.6702064311234617,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004291831617831261,
      "loss": 5.1733,
      "mean_token_accuracy": 0.1745262160897255,
      "num_tokens": 58941560.0,
      "step": 26905
    },
    {
      "entropy": 5.747252798080444,
      "epoch": 2.670702659785629,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042915735511529356,
      "loss": 5.1105,
      "mean_token_accuracy": 0.17346030324697495,
      "num_tokens": 58953365.0,
      "step": 26910
    },
    {
      "entropy": 5.666268825531006,
      "epoch": 2.6711988884477966,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004291315446246909,
      "loss": 5.0175,
      "mean_token_accuracy": 0.18571991175413133,
      "num_tokens": 58964602.0,
      "step": 26915
    },
    {
      "entropy": 5.645502090454102,
      "epoch": 2.671695117109964,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042910573031195804,
      "loss": 5.0278,
      "mean_token_accuracy": 0.18425729423761367,
      "num_tokens": 58976154.0,
      "step": 26920
    },
    {
      "entropy": 5.643967008590698,
      "epoch": 2.6721913457721316,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004290799121777351,
      "loss": 5.0618,
      "mean_token_accuracy": 0.18664228022098542,
      "num_tokens": 58985989.0,
      "step": 26925
    },
    {
      "entropy": 5.73802433013916,
      "epoch": 2.672687574434299,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004290540902226624,
      "loss": 5.1285,
      "mean_token_accuracy": 0.1725819766521454,
      "num_tokens": 58997080.0,
      "step": 26930
    },
    {
      "entropy": 5.656131219863892,
      "epoch": 2.673183803096467,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004290282644473802,
      "loss": 5.0478,
      "mean_token_accuracy": 0.18098960667848588,
      "num_tokens": 59008683.0,
      "step": 26935
    },
    {
      "entropy": 5.627220392227173,
      "epoch": 2.6736800317586344,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004290024348525289,
      "loss": 4.9613,
      "mean_token_accuracy": 0.191695499420166,
      "num_tokens": 59019019.0,
      "step": 26940
    },
    {
      "entropy": 5.597987079620362,
      "epoch": 2.674176260420802,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004289766014387492,
      "loss": 5.0233,
      "mean_token_accuracy": 0.1839980810880661,
      "num_tokens": 59029706.0,
      "step": 26945
    },
    {
      "entropy": 5.6272059917449955,
      "epoch": 2.6746724890829694,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004289507642066813,
      "loss": 4.9552,
      "mean_token_accuracy": 0.18205113112926483,
      "num_tokens": 59040208.0,
      "step": 26950
    },
    {
      "entropy": 5.59659776687622,
      "epoch": 2.675168717745137,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004289249231569664,
      "loss": 4.9995,
      "mean_token_accuracy": 0.19376865774393082,
      "num_tokens": 59050909.0,
      "step": 26955
    },
    {
      "entropy": 5.566015529632568,
      "epoch": 2.6756649464073043,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042889907829024506,
      "loss": 4.9818,
      "mean_token_accuracy": 0.18260796517133712,
      "num_tokens": 59059834.0,
      "step": 26960
    },
    {
      "entropy": 5.659055566787719,
      "epoch": 2.6761611750694723,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042887322960715813,
      "loss": 5.0658,
      "mean_token_accuracy": 0.17141700237989427,
      "num_tokens": 59071348.0,
      "step": 26965
    },
    {
      "entropy": 5.657482433319092,
      "epoch": 2.6766574037316397,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004288473771083468,
      "loss": 5.0118,
      "mean_token_accuracy": 0.17897712886333467,
      "num_tokens": 59083053.0,
      "step": 26970
    },
    {
      "entropy": 5.700801515579224,
      "epoch": 2.677153632393807,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004288215207944519,
      "loss": 5.0832,
      "mean_token_accuracy": 0.17761906385421752,
      "num_tokens": 59093400.0,
      "step": 26975
    },
    {
      "entropy": 5.636312532424927,
      "epoch": 2.6776498610559747,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042879566066611474,
      "loss": 5.0601,
      "mean_token_accuracy": 0.18215533792972566,
      "num_tokens": 59103653.0,
      "step": 26980
    },
    {
      "entropy": 5.6730588436126705,
      "epoch": 2.678146089718142,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004287697967239767,
      "loss": 5.0689,
      "mean_token_accuracy": 0.17878933548927306,
      "num_tokens": 59116660.0,
      "step": 26985
    },
    {
      "entropy": 5.609055614471435,
      "epoch": 2.6786423183803096,
      "grad_norm": 1.140625,
      "learning_rate": 0.000428743928968679,
      "loss": 4.9512,
      "mean_token_accuracy": 0.1877610996365547,
      "num_tokens": 59127424.0,
      "step": 26990
    },
    {
      "entropy": 5.662371349334717,
      "epoch": 2.679138547042477,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042871805740086317,
      "loss": 5.0634,
      "mean_token_accuracy": 0.1826170340180397,
      "num_tokens": 59137566.0,
      "step": 26995
    },
    {
      "entropy": 5.591405725479126,
      "epoch": 2.6796347757046446,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004286921820211707,
      "loss": 4.9657,
      "mean_token_accuracy": 0.18909405171871185,
      "num_tokens": 59148404.0,
      "step": 27000
    },
    {
      "epoch": 2.6796347757046446,
      "eval_entropy": 5.411024162621474,
      "eval_loss": 5.148655891418457,
      "eval_mean_token_accuracy": 0.18227268531245866,
      "eval_num_tokens": 59148404.0,
      "eval_runtime": 20.7237,
      "eval_samples_per_second": 1624.663,
      "eval_steps_per_second": 203.101,
      "step": 27000
    },
    {
      "entropy": 5.608983039855957,
      "epoch": 2.680131004366812,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042866630283024327,
      "loss": 5.025,
      "mean_token_accuracy": 0.18150854110717773,
      "num_tokens": 59158097.0,
      "step": 27005
    },
    {
      "entropy": 5.568313312530518,
      "epoch": 2.6806272330289795,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004286404198287227,
      "loss": 4.8889,
      "mean_token_accuracy": 0.19849170446395875,
      "num_tokens": 59168940.0,
      "step": 27010
    },
    {
      "entropy": 5.666187763214111,
      "epoch": 2.681123461691147,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004286145330172508,
      "loss": 5.0298,
      "mean_token_accuracy": 0.17867589145898818,
      "num_tokens": 59180231.0,
      "step": 27015
    },
    {
      "entropy": 5.716376161575317,
      "epoch": 2.681619690353315,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004285886423964695,
      "loss": 5.1206,
      "mean_token_accuracy": 0.16998365819454192,
      "num_tokens": 59191565.0,
      "step": 27020
    },
    {
      "entropy": 5.643762826919556,
      "epoch": 2.6821159190154824,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004285627479670208,
      "loss": 5.1078,
      "mean_token_accuracy": 0.17708656936883926,
      "num_tokens": 59203394.0,
      "step": 27025
    },
    {
      "entropy": 5.573599863052368,
      "epoch": 2.68261214767765,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004285368497295468,
      "loss": 4.9619,
      "mean_token_accuracy": 0.1837773397564888,
      "num_tokens": 59213987.0,
      "step": 27030
    },
    {
      "entropy": 5.629443073272705,
      "epoch": 2.6831083763398174,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004285109476846898,
      "loss": 4.998,
      "mean_token_accuracy": 0.18084481805562974,
      "num_tokens": 59225825.0,
      "step": 27035
    },
    {
      "entropy": 5.5928370475769045,
      "epoch": 2.683604605001985,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004284850418330921,
      "loss": 4.9836,
      "mean_token_accuracy": 0.19116006493568422,
      "num_tokens": 59235978.0,
      "step": 27040
    },
    {
      "entropy": 5.7145294666290285,
      "epoch": 2.6841008336641523,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042845913217539613,
      "loss": 5.158,
      "mean_token_accuracy": 0.1769076630473137,
      "num_tokens": 59247505.0,
      "step": 27045
    },
    {
      "entropy": 5.6233823776245115,
      "epoch": 2.6845970623263202,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042843321871224426,
      "loss": 4.953,
      "mean_token_accuracy": 0.18689041286706926,
      "num_tokens": 59258401.0,
      "step": 27050
    },
    {
      "entropy": 5.663552904129029,
      "epoch": 2.6850932909884877,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042840730144427927,
      "loss": 5.0308,
      "mean_token_accuracy": 0.18365978747606276,
      "num_tokens": 59268068.0,
      "step": 27055
    },
    {
      "entropy": 5.648304128646851,
      "epoch": 2.685589519650655,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042838138037214375,
      "loss": 5.0222,
      "mean_token_accuracy": 0.18277447819709777,
      "num_tokens": 59279567.0,
      "step": 27060
    },
    {
      "entropy": 5.645210886001587,
      "epoch": 2.6860857483128227,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042835545549648054,
      "loss": 4.96,
      "mean_token_accuracy": 0.18695928901433945,
      "num_tokens": 59289862.0,
      "step": 27065
    },
    {
      "entropy": 5.662708902359009,
      "epoch": 2.68658197697499,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042832952681793235,
      "loss": 5.0119,
      "mean_token_accuracy": 0.18851834088563918,
      "num_tokens": 59301304.0,
      "step": 27070
    },
    {
      "entropy": 5.635925579071045,
      "epoch": 2.6870782056371576,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042830359433714243,
      "loss": 5.0261,
      "mean_token_accuracy": 0.18466436266899108,
      "num_tokens": 59312184.0,
      "step": 27075
    },
    {
      "entropy": 5.609430742263794,
      "epoch": 2.687574434299325,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042827765805475367,
      "loss": 4.9166,
      "mean_token_accuracy": 0.1867563709616661,
      "num_tokens": 59322688.0,
      "step": 27080
    },
    {
      "entropy": 5.613276672363281,
      "epoch": 2.6880706629614926,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042825171797140926,
      "loss": 4.9786,
      "mean_token_accuracy": 0.1898431360721588,
      "num_tokens": 59333349.0,
      "step": 27085
    },
    {
      "entropy": 5.607793188095092,
      "epoch": 2.68856689162366,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004282257740877525,
      "loss": 4.9719,
      "mean_token_accuracy": 0.1917031690478325,
      "num_tokens": 59344473.0,
      "step": 27090
    },
    {
      "entropy": 5.644851446151733,
      "epoch": 2.6890631202858275,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004281998264044267,
      "loss": 5.0473,
      "mean_token_accuracy": 0.1830822706222534,
      "num_tokens": 59355993.0,
      "step": 27095
    },
    {
      "entropy": 5.725420761108398,
      "epoch": 2.689559348947995,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004281738749220753,
      "loss": 5.1074,
      "mean_token_accuracy": 0.1822989821434021,
      "num_tokens": 59367182.0,
      "step": 27100
    },
    {
      "entropy": 5.736227989196777,
      "epoch": 2.690055577610163,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042814791964134184,
      "loss": 5.0866,
      "mean_token_accuracy": 0.175626739859581,
      "num_tokens": 59379235.0,
      "step": 27105
    },
    {
      "entropy": 5.633092832565308,
      "epoch": 2.6905518062723304,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042812196056287,
      "loss": 5.0128,
      "mean_token_accuracy": 0.18999681770801544,
      "num_tokens": 59388514.0,
      "step": 27110
    },
    {
      "entropy": 5.626195287704467,
      "epoch": 2.691048034934498,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042809599768730357,
      "loss": 5.0434,
      "mean_token_accuracy": 0.18343483358621598,
      "num_tokens": 59399496.0,
      "step": 27115
    },
    {
      "entropy": 5.6358904361724855,
      "epoch": 2.6915442635966653,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004280700310152861,
      "loss": 5.0143,
      "mean_token_accuracy": 0.18254727721214295,
      "num_tokens": 59411986.0,
      "step": 27120
    },
    {
      "entropy": 5.763337898254394,
      "epoch": 2.692040492258833,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004280440605474619,
      "loss": 5.2154,
      "mean_token_accuracy": 0.1718280389904976,
      "num_tokens": 59424232.0,
      "step": 27125
    },
    {
      "entropy": 5.726628923416138,
      "epoch": 2.6925367209210003,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004280180862844747,
      "loss": 5.0746,
      "mean_token_accuracy": 0.17968170940876008,
      "num_tokens": 59435201.0,
      "step": 27130
    },
    {
      "entropy": 5.685031986236572,
      "epoch": 2.693032949583168,
      "grad_norm": 1.125,
      "learning_rate": 0.00042799210822696865,
      "loss": 5.0956,
      "mean_token_accuracy": 0.17668611854314803,
      "num_tokens": 59445552.0,
      "step": 27135
    },
    {
      "entropy": 5.722098636627197,
      "epoch": 2.6935291782453357,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000427966126375588,
      "loss": 5.0888,
      "mean_token_accuracy": 0.17266806066036225,
      "num_tokens": 59455634.0,
      "step": 27140
    },
    {
      "entropy": 5.713706398010254,
      "epoch": 2.694025406907503,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004279401407309771,
      "loss": 5.062,
      "mean_token_accuracy": 0.18118782490491867,
      "num_tokens": 59467955.0,
      "step": 27145
    },
    {
      "entropy": 5.669682788848877,
      "epoch": 2.6945216355696706,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004279141512937802,
      "loss": 5.061,
      "mean_token_accuracy": 0.18603339344263076,
      "num_tokens": 59479566.0,
      "step": 27150
    },
    {
      "entropy": 5.578594493865967,
      "epoch": 2.695017864231838,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004278881580646418,
      "loss": 4.9962,
      "mean_token_accuracy": 0.1863112762570381,
      "num_tokens": 59489967.0,
      "step": 27155
    },
    {
      "entropy": 5.635429763793946,
      "epoch": 2.6955140928940056,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004278621610442066,
      "loss": 5.0558,
      "mean_token_accuracy": 0.18028610050678254,
      "num_tokens": 59501640.0,
      "step": 27160
    },
    {
      "entropy": 5.652790546417236,
      "epoch": 2.696010321556173,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004278361602331192,
      "loss": 4.9823,
      "mean_token_accuracy": 0.1924065724015236,
      "num_tokens": 59512767.0,
      "step": 27165
    },
    {
      "entropy": 5.647517299652099,
      "epoch": 2.6965065502183405,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042781015563202434,
      "loss": 4.9571,
      "mean_token_accuracy": 0.1932586207985878,
      "num_tokens": 59522188.0,
      "step": 27170
    },
    {
      "entropy": 5.654234504699707,
      "epoch": 2.697002778880508,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004277841472415669,
      "loss": 4.9989,
      "mean_token_accuracy": 0.18745788782835007,
      "num_tokens": 59534871.0,
      "step": 27175
    },
    {
      "entropy": 5.746880960464478,
      "epoch": 2.6974990075426755,
      "grad_norm": 1.125,
      "learning_rate": 0.0004277581350623918,
      "loss": 5.1157,
      "mean_token_accuracy": 0.16960226893424987,
      "num_tokens": 59546858.0,
      "step": 27180
    },
    {
      "entropy": 5.622322225570679,
      "epoch": 2.697995236204843,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004277321190951441,
      "loss": 4.9748,
      "mean_token_accuracy": 0.18371348828077316,
      "num_tokens": 59557022.0,
      "step": 27185
    },
    {
      "entropy": 5.587670707702637,
      "epoch": 2.698491464867011,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042770609934046904,
      "loss": 4.9958,
      "mean_token_accuracy": 0.18927717804908753,
      "num_tokens": 59567564.0,
      "step": 27190
    },
    {
      "entropy": 5.758999443054199,
      "epoch": 2.6989876935291783,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042768007579901176,
      "loss": 5.1356,
      "mean_token_accuracy": 0.17351772487163544,
      "num_tokens": 59579122.0,
      "step": 27195
    },
    {
      "entropy": 5.749111557006836,
      "epoch": 2.699483922191346,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042765404847141755,
      "loss": 5.1013,
      "mean_token_accuracy": 0.1823315367102623,
      "num_tokens": 59590477.0,
      "step": 27200
    },
    {
      "entropy": 5.665436935424805,
      "epoch": 2.6999801508535133,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004276280173583319,
      "loss": 4.9964,
      "mean_token_accuracy": 0.18419334441423416,
      "num_tokens": 59601217.0,
      "step": 27205
    },
    {
      "entropy": 5.6176981925964355,
      "epoch": 2.7004763795156808,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042760198246040025,
      "loss": 5.0046,
      "mean_token_accuracy": 0.18624254763126374,
      "num_tokens": 59612271.0,
      "step": 27210
    },
    {
      "entropy": 5.687909460067749,
      "epoch": 2.7009726081778482,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004275759437782683,
      "loss": 5.1031,
      "mean_token_accuracy": 0.17452418059110641,
      "num_tokens": 59623948.0,
      "step": 27215
    },
    {
      "entropy": 5.63287935256958,
      "epoch": 2.701468836840016,
      "grad_norm": 1.0,
      "learning_rate": 0.00042754990131258175,
      "loss": 4.9588,
      "mean_token_accuracy": 0.19044899046421052,
      "num_tokens": 59635720.0,
      "step": 27220
    },
    {
      "entropy": 5.713837623596191,
      "epoch": 2.7019650655021836,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004275238550639863,
      "loss": 5.0644,
      "mean_token_accuracy": 0.18441731482744217,
      "num_tokens": 59645171.0,
      "step": 27225
    },
    {
      "entropy": 5.736519479751587,
      "epoch": 2.702461294164351,
      "grad_norm": 1.109375,
      "learning_rate": 0.000427497805033128,
      "loss": 5.1424,
      "mean_token_accuracy": 0.17513692378997803,
      "num_tokens": 59655266.0,
      "step": 27230
    },
    {
      "entropy": 5.641294956207275,
      "epoch": 2.7029575228265186,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004274717512206527,
      "loss": 5.0302,
      "mean_token_accuracy": 0.18911393284797667,
      "num_tokens": 59667055.0,
      "step": 27235
    },
    {
      "entropy": 5.625741004943848,
      "epoch": 2.703453751488686,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004274456936272065,
      "loss": 4.9863,
      "mean_token_accuracy": 0.1844910979270935,
      "num_tokens": 59678621.0,
      "step": 27240
    },
    {
      "entropy": 5.632012891769409,
      "epoch": 2.7039499801508535,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042741963225343554,
      "loss": 4.9888,
      "mean_token_accuracy": 0.18931201100349426,
      "num_tokens": 59690514.0,
      "step": 27245
    },
    {
      "entropy": 5.728258228302002,
      "epoch": 2.704446208813021,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004273935670999861,
      "loss": 5.061,
      "mean_token_accuracy": 0.1792294517159462,
      "num_tokens": 59701271.0,
      "step": 27250
    },
    {
      "entropy": 5.657385730743409,
      "epoch": 2.7049424374751885,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042736749816750476,
      "loss": 5.1189,
      "mean_token_accuracy": 0.1749150946736336,
      "num_tokens": 59712787.0,
      "step": 27255
    },
    {
      "entropy": 5.587566614151001,
      "epoch": 2.705438666137356,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004273414254566376,
      "loss": 4.9584,
      "mean_token_accuracy": 0.18956088721752168,
      "num_tokens": 59722787.0,
      "step": 27260
    },
    {
      "entropy": 5.664895963668823,
      "epoch": 2.7059348947995234,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004273153489680315,
      "loss": 5.0475,
      "mean_token_accuracy": 0.18656672537326813,
      "num_tokens": 59733302.0,
      "step": 27265
    },
    {
      "entropy": 5.628655004501343,
      "epoch": 2.706431123461691,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042728926870233287,
      "loss": 5.0283,
      "mean_token_accuracy": 0.18300368785858154,
      "num_tokens": 59743290.0,
      "step": 27270
    },
    {
      "entropy": 5.7191245555877686,
      "epoch": 2.706927352123859,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042726318466018855,
      "loss": 5.1021,
      "mean_token_accuracy": 0.17970240414142608,
      "num_tokens": 59754145.0,
      "step": 27275
    },
    {
      "entropy": 5.708598375320435,
      "epoch": 2.7074235807860263,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004272370968422454,
      "loss": 5.0579,
      "mean_token_accuracy": 0.18512318283319473,
      "num_tokens": 59764075.0,
      "step": 27280
    },
    {
      "entropy": 5.641919898986816,
      "epoch": 2.7079198094481938,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042721100524915017,
      "loss": 5.0702,
      "mean_token_accuracy": 0.17787974774837495,
      "num_tokens": 59775495.0,
      "step": 27285
    },
    {
      "entropy": 5.580399036407471,
      "epoch": 2.7084160381103612,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042718490988155014,
      "loss": 4.9069,
      "mean_token_accuracy": 0.20416743606328963,
      "num_tokens": 59787006.0,
      "step": 27290
    },
    {
      "entropy": 5.685935211181641,
      "epoch": 2.7089122667725287,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042715881074009216,
      "loss": 5.0563,
      "mean_token_accuracy": 0.17936498820781707,
      "num_tokens": 59797680.0,
      "step": 27295
    },
    {
      "entropy": 5.6377363204956055,
      "epoch": 2.709408495434696,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042713270782542364,
      "loss": 5.0245,
      "mean_token_accuracy": 0.18179403692483903,
      "num_tokens": 59809666.0,
      "step": 27300
    },
    {
      "entropy": 5.597621107101441,
      "epoch": 2.7099047240968637,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004271066011381917,
      "loss": 4.9932,
      "mean_token_accuracy": 0.1856367215514183,
      "num_tokens": 59819889.0,
      "step": 27305
    },
    {
      "entropy": 5.59941520690918,
      "epoch": 2.7104009527590316,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004270804906790438,
      "loss": 4.9426,
      "mean_token_accuracy": 0.19093361645936965,
      "num_tokens": 59830344.0,
      "step": 27310
    },
    {
      "entropy": 5.618537425994873,
      "epoch": 2.710897181421199,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004270543764486275,
      "loss": 4.9946,
      "mean_token_accuracy": 0.18924577087163924,
      "num_tokens": 59843498.0,
      "step": 27315
    },
    {
      "entropy": 5.691640853881836,
      "epoch": 2.7113934100833665,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004270282584475903,
      "loss": 5.0468,
      "mean_token_accuracy": 0.18050143569707872,
      "num_tokens": 59854633.0,
      "step": 27320
    },
    {
      "entropy": 5.673439645767212,
      "epoch": 2.711889638745534,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042700213667657983,
      "loss": 5.1053,
      "mean_token_accuracy": 0.18222663998603822,
      "num_tokens": 59865925.0,
      "step": 27325
    },
    {
      "entropy": 5.638454294204712,
      "epoch": 2.7123858674077015,
      "grad_norm": 1.25,
      "learning_rate": 0.000426976011136244,
      "loss": 5.0011,
      "mean_token_accuracy": 0.18537127822637559,
      "num_tokens": 59877196.0,
      "step": 27330
    },
    {
      "entropy": 5.699487113952637,
      "epoch": 2.712882096069869,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042694988182723047,
      "loss": 5.0754,
      "mean_token_accuracy": 0.17579150944948196,
      "num_tokens": 59887411.0,
      "step": 27335
    },
    {
      "entropy": 5.620398664474488,
      "epoch": 2.7133783247320364,
      "grad_norm": 0.921875,
      "learning_rate": 0.00042692374875018735,
      "loss": 5.0166,
      "mean_token_accuracy": 0.1932065322995186,
      "num_tokens": 59899123.0,
      "step": 27340
    },
    {
      "entropy": 5.594221305847168,
      "epoch": 2.713874553394204,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042689761190576263,
      "loss": 4.9758,
      "mean_token_accuracy": 0.19109513014554977,
      "num_tokens": 59910326.0,
      "step": 27345
    },
    {
      "entropy": 5.685772275924682,
      "epoch": 2.7143707820563714,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042687147129460447,
      "loss": 5.0698,
      "mean_token_accuracy": 0.1783434733748436,
      "num_tokens": 59920722.0,
      "step": 27350
    },
    {
      "entropy": 5.698673677444458,
      "epoch": 2.714867010718539,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042684532691736104,
      "loss": 5.0458,
      "mean_token_accuracy": 0.17770348191261293,
      "num_tokens": 59932568.0,
      "step": 27355
    },
    {
      "entropy": 5.664941024780274,
      "epoch": 2.715363239380707,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042681917877468076,
      "loss": 5.1001,
      "mean_token_accuracy": 0.17538348883390426,
      "num_tokens": 59943805.0,
      "step": 27360
    },
    {
      "entropy": 5.652146911621093,
      "epoch": 2.7158594680428743,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042679302686721187,
      "loss": 4.9878,
      "mean_token_accuracy": 0.19392020106315613,
      "num_tokens": 59954455.0,
      "step": 27365
    },
    {
      "entropy": 5.665805244445801,
      "epoch": 2.7163556967050417,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004267668711956031,
      "loss": 5.002,
      "mean_token_accuracy": 0.18401228487491608,
      "num_tokens": 59966722.0,
      "step": 27370
    },
    {
      "entropy": 5.654845094680786,
      "epoch": 2.716851925367209,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042674071176050296,
      "loss": 5.0192,
      "mean_token_accuracy": 0.18427737802267075,
      "num_tokens": 59976702.0,
      "step": 27375
    },
    {
      "entropy": 5.604920101165772,
      "epoch": 2.7173481540293767,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042671454856256,
      "loss": 4.983,
      "mean_token_accuracy": 0.185209022462368,
      "num_tokens": 59987038.0,
      "step": 27380
    },
    {
      "entropy": 5.659537506103516,
      "epoch": 2.717844382691544,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004266883816024233,
      "loss": 5.0473,
      "mean_token_accuracy": 0.18123307973146438,
      "num_tokens": 59996755.0,
      "step": 27385
    },
    {
      "entropy": 5.65525975227356,
      "epoch": 2.7183406113537116,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004266622108807415,
      "loss": 4.993,
      "mean_token_accuracy": 0.18034493625164033,
      "num_tokens": 60006419.0,
      "step": 27390
    },
    {
      "entropy": 5.692930459976196,
      "epoch": 2.7188368400158796,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004266360363981637,
      "loss": 5.0888,
      "mean_token_accuracy": 0.18233918845653535,
      "num_tokens": 60017793.0,
      "step": 27395
    },
    {
      "entropy": 5.604616785049439,
      "epoch": 2.719333068678047,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042660985815533905,
      "loss": 4.9616,
      "mean_token_accuracy": 0.18522814512252808,
      "num_tokens": 60029392.0,
      "step": 27400
    },
    {
      "entropy": 5.651560735702515,
      "epoch": 2.7198292973402145,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042658367615291644,
      "loss": 5.0947,
      "mean_token_accuracy": 0.18476058542728424,
      "num_tokens": 60040467.0,
      "step": 27405
    },
    {
      "entropy": 5.635646104812622,
      "epoch": 2.720325526002382,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004265574903915454,
      "loss": 5.021,
      "mean_token_accuracy": 0.18288804292678834,
      "num_tokens": 60051583.0,
      "step": 27410
    },
    {
      "entropy": 5.66191930770874,
      "epoch": 2.7208217546645495,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004265313008718751,
      "loss": 5.037,
      "mean_token_accuracy": 0.17743658721446992,
      "num_tokens": 60062218.0,
      "step": 27415
    },
    {
      "entropy": 5.626641893386841,
      "epoch": 2.721317983326717,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042650510759455516,
      "loss": 5.0138,
      "mean_token_accuracy": 0.18855656087398528,
      "num_tokens": 60074455.0,
      "step": 27420
    },
    {
      "entropy": 5.6532001972198485,
      "epoch": 2.7218142119888844,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042647891056023495,
      "loss": 4.999,
      "mean_token_accuracy": 0.1820360615849495,
      "num_tokens": 60085442.0,
      "step": 27425
    },
    {
      "entropy": 5.700037240982056,
      "epoch": 2.722310440651052,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042645270976956426,
      "loss": 5.0779,
      "mean_token_accuracy": 0.18490006625652314,
      "num_tokens": 60096441.0,
      "step": 27430
    },
    {
      "entropy": 5.736059141159058,
      "epoch": 2.7228066693132194,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042642650522319253,
      "loss": 5.0386,
      "mean_token_accuracy": 0.18442390859127045,
      "num_tokens": 60107399.0,
      "step": 27435
    },
    {
      "entropy": 5.681547975540161,
      "epoch": 2.723302897975387,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004264002969217699,
      "loss": 5.0087,
      "mean_token_accuracy": 0.18628604263067244,
      "num_tokens": 60117568.0,
      "step": 27440
    },
    {
      "entropy": 5.680322170257568,
      "epoch": 2.7237991266375547,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004263740848659461,
      "loss": 5.0713,
      "mean_token_accuracy": 0.1775377482175827,
      "num_tokens": 60127214.0,
      "step": 27445
    },
    {
      "entropy": 5.626444292068482,
      "epoch": 2.724295355299722,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042634786905637123,
      "loss": 5.0012,
      "mean_token_accuracy": 0.18702634871006013,
      "num_tokens": 60138167.0,
      "step": 27450
    },
    {
      "entropy": 5.625717639923096,
      "epoch": 2.7247915839618897,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004263216494936953,
      "loss": 4.9769,
      "mean_token_accuracy": 0.18337800800800325,
      "num_tokens": 60148841.0,
      "step": 27455
    },
    {
      "entropy": 5.621400451660156,
      "epoch": 2.725287812624057,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042629542617856855,
      "loss": 5.0579,
      "mean_token_accuracy": 0.1831665500998497,
      "num_tokens": 60159512.0,
      "step": 27460
    },
    {
      "entropy": 5.612176322937012,
      "epoch": 2.7257840412862246,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004262691991116412,
      "loss": 5.0076,
      "mean_token_accuracy": 0.17921405136585236,
      "num_tokens": 60170463.0,
      "step": 27465
    },
    {
      "entropy": 5.5483348846435545,
      "epoch": 2.726280269948392,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004262429682935637,
      "loss": 4.8879,
      "mean_token_accuracy": 0.19576393365859984,
      "num_tokens": 60180431.0,
      "step": 27470
    },
    {
      "entropy": 5.650901222229004,
      "epoch": 2.7267764986105596,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004262167337249865,
      "loss": 5.0429,
      "mean_token_accuracy": 0.1832689881324768,
      "num_tokens": 60190410.0,
      "step": 27475
    },
    {
      "entropy": 5.662662076950073,
      "epoch": 2.7272727272727275,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004261904954065601,
      "loss": 4.9853,
      "mean_token_accuracy": 0.1852027714252472,
      "num_tokens": 60201909.0,
      "step": 27480
    },
    {
      "entropy": 5.71583514213562,
      "epoch": 2.727768955934895,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004261642533389352,
      "loss": 4.9859,
      "mean_token_accuracy": 0.1856275886297226,
      "num_tokens": 60212577.0,
      "step": 27485
    },
    {
      "entropy": 5.636430644989014,
      "epoch": 2.7282651845970625,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042613800752276264,
      "loss": 5.0803,
      "mean_token_accuracy": 0.17986503392457961,
      "num_tokens": 60222615.0,
      "step": 27490
    },
    {
      "entropy": 5.635940313339233,
      "epoch": 2.72876141325923,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042611175795869306,
      "loss": 5.1038,
      "mean_token_accuracy": 0.1824706181883812,
      "num_tokens": 60234338.0,
      "step": 27495
    },
    {
      "entropy": 5.720212364196778,
      "epoch": 2.7292576419213974,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004260855046473775,
      "loss": 4.9821,
      "mean_token_accuracy": 0.18356282860040665,
      "num_tokens": 60244781.0,
      "step": 27500
    },
    {
      "entropy": 5.6894725322723385,
      "epoch": 2.729753870583565,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042605924758946706,
      "loss": 5.0579,
      "mean_token_accuracy": 0.18275235295295716,
      "num_tokens": 60255150.0,
      "step": 27505
    },
    {
      "entropy": 5.625911521911621,
      "epoch": 2.7302500992457324,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042603298678561277,
      "loss": 5.016,
      "mean_token_accuracy": 0.1837659478187561,
      "num_tokens": 60266073.0,
      "step": 27510
    },
    {
      "entropy": 5.689517021179199,
      "epoch": 2.7307463279079,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004260067222364659,
      "loss": 5.0691,
      "mean_token_accuracy": 0.1794586807489395,
      "num_tokens": 60277520.0,
      "step": 27515
    },
    {
      "entropy": 5.678891324996949,
      "epoch": 2.7312425565700673,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042598045394267757,
      "loss": 5.0371,
      "mean_token_accuracy": 0.18451922982931138,
      "num_tokens": 60288530.0,
      "step": 27520
    },
    {
      "entropy": 5.6718864917755125,
      "epoch": 2.731738785232235,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042595418190489944,
      "loss": 4.9657,
      "mean_token_accuracy": 0.1894846335053444,
      "num_tokens": 60299926.0,
      "step": 27525
    },
    {
      "entropy": 5.713263034820557,
      "epoch": 2.7322350138944023,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042592790612378277,
      "loss": 5.1394,
      "mean_token_accuracy": 0.17471423894166946,
      "num_tokens": 60312129.0,
      "step": 27530
    },
    {
      "entropy": 5.673270988464355,
      "epoch": 2.73273124255657,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042590162659997933,
      "loss": 5.0587,
      "mean_token_accuracy": 0.1791563183069229,
      "num_tokens": 60322405.0,
      "step": 27535
    },
    {
      "entropy": 5.700493097305298,
      "epoch": 2.7332274712187377,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004258753433341407,
      "loss": 5.0544,
      "mean_token_accuracy": 0.17841399610042571,
      "num_tokens": 60333112.0,
      "step": 27540
    },
    {
      "entropy": 5.704744625091553,
      "epoch": 2.733723699880905,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042584905632691867,
      "loss": 5.0144,
      "mean_token_accuracy": 0.1830108568072319,
      "num_tokens": 60343500.0,
      "step": 27545
    },
    {
      "entropy": 5.636524963378906,
      "epoch": 2.7342199285430726,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004258227655789651,
      "loss": 5.0108,
      "mean_token_accuracy": 0.18503265976905822,
      "num_tokens": 60354711.0,
      "step": 27550
    },
    {
      "entropy": 5.642283630371094,
      "epoch": 2.73471615720524,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004257964710909318,
      "loss": 5.0422,
      "mean_token_accuracy": 0.17894639819860458,
      "num_tokens": 60365064.0,
      "step": 27555
    },
    {
      "entropy": 5.64208984375,
      "epoch": 2.7352123858674076,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042577017286347116,
      "loss": 5.0605,
      "mean_token_accuracy": 0.17841062396764756,
      "num_tokens": 60376425.0,
      "step": 27560
    },
    {
      "entropy": 5.663001537322998,
      "epoch": 2.7357086145295755,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042574387089723504,
      "loss": 4.9393,
      "mean_token_accuracy": 0.1878805175423622,
      "num_tokens": 60386745.0,
      "step": 27565
    },
    {
      "entropy": 5.663871955871582,
      "epoch": 2.736204843191743,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042571756519287567,
      "loss": 4.9992,
      "mean_token_accuracy": 0.1819422334432602,
      "num_tokens": 60397038.0,
      "step": 27570
    },
    {
      "entropy": 5.6150798320770265,
      "epoch": 2.7367010718539104,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042569125575104557,
      "loss": 5.013,
      "mean_token_accuracy": 0.18322886973619462,
      "num_tokens": 60406955.0,
      "step": 27575
    },
    {
      "entropy": 5.656726360321045,
      "epoch": 2.737197300516078,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042566494257239695,
      "loss": 5.0279,
      "mean_token_accuracy": 0.18245746493339537,
      "num_tokens": 60418245.0,
      "step": 27580
    },
    {
      "entropy": 5.703398609161377,
      "epoch": 2.7376935291782454,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004256386256575824,
      "loss": 5.0388,
      "mean_token_accuracy": 0.1780583456158638,
      "num_tokens": 60429310.0,
      "step": 27585
    },
    {
      "entropy": 5.6876670837402346,
      "epoch": 2.738189757840413,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004256123050072546,
      "loss": 5.0726,
      "mean_token_accuracy": 0.18292552083730698,
      "num_tokens": 60441355.0,
      "step": 27590
    },
    {
      "entropy": 5.783017826080322,
      "epoch": 2.7386859865025803,
      "grad_norm": 0.953125,
      "learning_rate": 0.00042558598062206616,
      "loss": 5.2149,
      "mean_token_accuracy": 0.17206214517354965,
      "num_tokens": 60452853.0,
      "step": 27595
    },
    {
      "entropy": 5.68136568069458,
      "epoch": 2.739182215164748,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042555965250266984,
      "loss": 5.0187,
      "mean_token_accuracy": 0.1812727317214012,
      "num_tokens": 60462934.0,
      "step": 27600
    },
    {
      "entropy": 5.723084354400635,
      "epoch": 2.7396784438269153,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042553332064971866,
      "loss": 5.1305,
      "mean_token_accuracy": 0.1743614688515663,
      "num_tokens": 60475575.0,
      "step": 27605
    },
    {
      "entropy": 5.561553859710694,
      "epoch": 2.7401746724890828,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042550698506386536,
      "loss": 4.8987,
      "mean_token_accuracy": 0.19488942623138428,
      "num_tokens": 60486495.0,
      "step": 27610
    },
    {
      "entropy": 5.631079292297363,
      "epoch": 2.7406709011512502,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042548064574576323,
      "loss": 5.0069,
      "mean_token_accuracy": 0.19111116379499435,
      "num_tokens": 60496806.0,
      "step": 27615
    },
    {
      "entropy": 5.648279094696045,
      "epoch": 2.741167129813418,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004254543026960654,
      "loss": 5.0102,
      "mean_token_accuracy": 0.18670350015163423,
      "num_tokens": 60508905.0,
      "step": 27620
    },
    {
      "entropy": 5.601529312133789,
      "epoch": 2.7416633584755856,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004254279559154249,
      "loss": 4.9705,
      "mean_token_accuracy": 0.1840098276734352,
      "num_tokens": 60519603.0,
      "step": 27625
    },
    {
      "entropy": 5.647975921630859,
      "epoch": 2.742159587137753,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042540160540449536,
      "loss": 5.0471,
      "mean_token_accuracy": 0.18191648572683333,
      "num_tokens": 60530977.0,
      "step": 27630
    },
    {
      "entropy": 5.667160177230835,
      "epoch": 2.7426558157999206,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004253752511639301,
      "loss": 5.0667,
      "mean_token_accuracy": 0.1778062403202057,
      "num_tokens": 60541354.0,
      "step": 27635
    },
    {
      "entropy": 5.641473054885864,
      "epoch": 2.743152044462088,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004253488931943826,
      "loss": 5.0386,
      "mean_token_accuracy": 0.18428778052330017,
      "num_tokens": 60552531.0,
      "step": 27640
    },
    {
      "entropy": 5.594116497039795,
      "epoch": 2.7436482731242555,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042532253149650653,
      "loss": 5.0048,
      "mean_token_accuracy": 0.18379111140966414,
      "num_tokens": 60562552.0,
      "step": 27645
    },
    {
      "entropy": 5.6441652297973635,
      "epoch": 2.7441445017864234,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004252961660709556,
      "loss": 5.0346,
      "mean_token_accuracy": 0.18078004717826843,
      "num_tokens": 60572849.0,
      "step": 27650
    },
    {
      "entropy": 5.638095140457153,
      "epoch": 2.744640730448591,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042526979691838353,
      "loss": 4.9981,
      "mean_token_accuracy": 0.184382426738739,
      "num_tokens": 60584095.0,
      "step": 27655
    },
    {
      "entropy": 5.659544134140015,
      "epoch": 2.7451369591107584,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042524342403944446,
      "loss": 5.066,
      "mean_token_accuracy": 0.18422713428735732,
      "num_tokens": 60594353.0,
      "step": 27660
    },
    {
      "entropy": 5.678940486907959,
      "epoch": 2.745633187772926,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042521704743479213,
      "loss": 5.0574,
      "mean_token_accuracy": 0.18331601917743684,
      "num_tokens": 60604558.0,
      "step": 27665
    },
    {
      "entropy": 5.630993890762329,
      "epoch": 2.7461294164350933,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042519066710508064,
      "loss": 5.0151,
      "mean_token_accuracy": 0.1809190347790718,
      "num_tokens": 60615670.0,
      "step": 27670
    },
    {
      "entropy": 5.582134342193603,
      "epoch": 2.746625645097261,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042516428305096435,
      "loss": 4.9596,
      "mean_token_accuracy": 0.18373736143112182,
      "num_tokens": 60626452.0,
      "step": 27675
    },
    {
      "entropy": 5.641168689727783,
      "epoch": 2.7471218737594283,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004251378952730973,
      "loss": 5.0313,
      "mean_token_accuracy": 0.1864634335041046,
      "num_tokens": 60637122.0,
      "step": 27680
    },
    {
      "entropy": 5.6693743705749515,
      "epoch": 2.7476181024215958,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004251115037721341,
      "loss": 5.0034,
      "mean_token_accuracy": 0.19158975630998612,
      "num_tokens": 60648835.0,
      "step": 27685
    },
    {
      "entropy": 5.698266124725341,
      "epoch": 2.7481143310837632,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042508510854872905,
      "loss": 5.0598,
      "mean_token_accuracy": 0.17629244029521943,
      "num_tokens": 60658537.0,
      "step": 27690
    },
    {
      "entropy": 5.674579334259033,
      "epoch": 2.7486105597459307,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042505870960353665,
      "loss": 5.0349,
      "mean_token_accuracy": 0.18211353421211243,
      "num_tokens": 60670450.0,
      "step": 27695
    },
    {
      "entropy": 5.721584033966065,
      "epoch": 2.749106788408098,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004250323069372116,
      "loss": 5.0488,
      "mean_token_accuracy": 0.188187637925148,
      "num_tokens": 60680539.0,
      "step": 27700
    },
    {
      "entropy": 5.696606111526489,
      "epoch": 2.749603017070266,
      "grad_norm": 1.125,
      "learning_rate": 0.00042500590055040866,
      "loss": 5.1156,
      "mean_token_accuracy": 0.17777589857578277,
      "num_tokens": 60693106.0,
      "step": 27705
    },
    {
      "entropy": 5.742177534103393,
      "epoch": 2.7500992457324336,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042497949044378267,
      "loss": 5.0973,
      "mean_token_accuracy": 0.1772633373737335,
      "num_tokens": 60703509.0,
      "step": 27710
    },
    {
      "entropy": 5.678531074523926,
      "epoch": 2.750595474394601,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042495307661798837,
      "loss": 5.0819,
      "mean_token_accuracy": 0.1778382822871208,
      "num_tokens": 60714113.0,
      "step": 27715
    },
    {
      "entropy": 5.734507465362549,
      "epoch": 2.7510917030567685,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042492665907368094,
      "loss": 5.1252,
      "mean_token_accuracy": 0.17669990211725234,
      "num_tokens": 60726826.0,
      "step": 27720
    },
    {
      "entropy": 5.666940307617187,
      "epoch": 2.751587931718936,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042490023781151546,
      "loss": 5.0339,
      "mean_token_accuracy": 0.1822245091199875,
      "num_tokens": 60738443.0,
      "step": 27725
    },
    {
      "entropy": 5.632249736785889,
      "epoch": 2.7520841603811035,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042487381283214706,
      "loss": 5.0049,
      "mean_token_accuracy": 0.1837996870279312,
      "num_tokens": 60748865.0,
      "step": 27730
    },
    {
      "entropy": 5.678263998031616,
      "epoch": 2.7525803890432714,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004248473841362311,
      "loss": 5.0441,
      "mean_token_accuracy": 0.18526833057403563,
      "num_tokens": 60760001.0,
      "step": 27735
    },
    {
      "entropy": 5.7283392429351805,
      "epoch": 2.753076617705439,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042482095172442286,
      "loss": 5.0922,
      "mean_token_accuracy": 0.18259043693542482,
      "num_tokens": 60772003.0,
      "step": 27740
    },
    {
      "entropy": 5.62926836013794,
      "epoch": 2.7535728463676064,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004247945155973778,
      "loss": 4.9545,
      "mean_token_accuracy": 0.18947488814592361,
      "num_tokens": 60783167.0,
      "step": 27745
    },
    {
      "entropy": 5.683422183990478,
      "epoch": 2.754069075029774,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004247680757557516,
      "loss": 5.064,
      "mean_token_accuracy": 0.18332672566175462,
      "num_tokens": 60793917.0,
      "step": 27750
    },
    {
      "entropy": 5.74482913017273,
      "epoch": 2.7545653036919413,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042474163220019987,
      "loss": 5.1505,
      "mean_token_accuracy": 0.17734026312828063,
      "num_tokens": 60805166.0,
      "step": 27755
    },
    {
      "entropy": 5.757231664657593,
      "epoch": 2.7550615323541088,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042471518493137836,
      "loss": 5.0791,
      "mean_token_accuracy": 0.18401380479335785,
      "num_tokens": 60816984.0,
      "step": 27760
    },
    {
      "entropy": 5.634587907791138,
      "epoch": 2.7555577610162763,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004246887339499427,
      "loss": 5.0101,
      "mean_token_accuracy": 0.1833496779203415,
      "num_tokens": 60829481.0,
      "step": 27765
    },
    {
      "entropy": 5.6623327255249025,
      "epoch": 2.7560539896784437,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042466227925654917,
      "loss": 5.0003,
      "mean_token_accuracy": 0.18814381659030915,
      "num_tokens": 60839376.0,
      "step": 27770
    },
    {
      "entropy": 5.625752067565918,
      "epoch": 2.756550218340611,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004246358208518535,
      "loss": 5.0378,
      "mean_token_accuracy": 0.18327111154794692,
      "num_tokens": 60850383.0,
      "step": 27775
    },
    {
      "entropy": 5.664331150054932,
      "epoch": 2.7570464470027787,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042460935873651196,
      "loss": 5.0089,
      "mean_token_accuracy": 0.18584911823272704,
      "num_tokens": 60861249.0,
      "step": 27780
    },
    {
      "entropy": 5.684563302993775,
      "epoch": 2.757542675664946,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042458289291118066,
      "loss": 5.0407,
      "mean_token_accuracy": 0.18475522398948668,
      "num_tokens": 60872220.0,
      "step": 27785
    },
    {
      "entropy": 5.649740219116211,
      "epoch": 2.758038904327114,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042455642337651595,
      "loss": 5.0319,
      "mean_token_accuracy": 0.1770552933216095,
      "num_tokens": 60882006.0,
      "step": 27790
    },
    {
      "entropy": 5.660260343551636,
      "epoch": 2.7585351329892815,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004245299501331743,
      "loss": 4.9473,
      "mean_token_accuracy": 0.19464349299669265,
      "num_tokens": 60891732.0,
      "step": 27795
    },
    {
      "entropy": 5.707980060577393,
      "epoch": 2.759031361651449,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042450347318181194,
      "loss": 5.0778,
      "mean_token_accuracy": 0.1844562068581581,
      "num_tokens": 60902990.0,
      "step": 27800
    },
    {
      "entropy": 5.667835712432861,
      "epoch": 2.7595275903136165,
      "grad_norm": 1.0,
      "learning_rate": 0.00042447699252308566,
      "loss": 5.0226,
      "mean_token_accuracy": 0.1772082731127739,
      "num_tokens": 60914673.0,
      "step": 27805
    },
    {
      "entropy": 5.6394952774047855,
      "epoch": 2.760023818975784,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004244505081576521,
      "loss": 4.9788,
      "mean_token_accuracy": 0.18722182810306548,
      "num_tokens": 60925872.0,
      "step": 27810
    },
    {
      "entropy": 5.651049661636352,
      "epoch": 2.7605200476379514,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042442402008616793,
      "loss": 5.0762,
      "mean_token_accuracy": 0.18355192691087724,
      "num_tokens": 60936589.0,
      "step": 27815
    },
    {
      "entropy": 5.651469326019287,
      "epoch": 2.7610162763001194,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042439752830929,
      "loss": 5.0222,
      "mean_token_accuracy": 0.18098794668912888,
      "num_tokens": 60947213.0,
      "step": 27820
    },
    {
      "entropy": 5.711999464035034,
      "epoch": 2.761512504962287,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004243710328276754,
      "loss": 5.0308,
      "mean_token_accuracy": 0.18179891109466553,
      "num_tokens": 60958615.0,
      "step": 27825
    },
    {
      "entropy": 5.652822780609131,
      "epoch": 2.7620087336244543,
      "grad_norm": 1.078125,
      "learning_rate": 0.000424344533641981,
      "loss": 4.9798,
      "mean_token_accuracy": 0.18481871485710144,
      "num_tokens": 60968830.0,
      "step": 27830
    },
    {
      "entropy": 5.636505746841431,
      "epoch": 2.762504962286622,
      "grad_norm": 1.078125,
      "learning_rate": 0.000424318030752864,
      "loss": 4.9574,
      "mean_token_accuracy": 0.18681089580059052,
      "num_tokens": 60979744.0,
      "step": 27835
    },
    {
      "entropy": 5.704662132263183,
      "epoch": 2.7630011909487893,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004242915241609815,
      "loss": 5.0547,
      "mean_token_accuracy": 0.1760562539100647,
      "num_tokens": 60990291.0,
      "step": 27840
    },
    {
      "entropy": 5.691522932052612,
      "epoch": 2.7634974196109567,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000424265013866991,
      "loss": 5.0137,
      "mean_token_accuracy": 0.18511313050985337,
      "num_tokens": 61002170.0,
      "step": 27845
    },
    {
      "entropy": 5.635736417770386,
      "epoch": 2.763993648273124,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004242384998715497,
      "loss": 5.0727,
      "mean_token_accuracy": 0.17982216626405717,
      "num_tokens": 61012639.0,
      "step": 27850
    },
    {
      "entropy": 5.695107460021973,
      "epoch": 2.7644898769352917,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004242119821753153,
      "loss": 5.022,
      "mean_token_accuracy": 0.1816425159573555,
      "num_tokens": 61022657.0,
      "step": 27855
    },
    {
      "entropy": 5.63418025970459,
      "epoch": 2.764986105597459,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004241854607789452,
      "loss": 4.9734,
      "mean_token_accuracy": 0.18961677551269532,
      "num_tokens": 61033108.0,
      "step": 27860
    },
    {
      "entropy": 5.599731016159057,
      "epoch": 2.7654823342596266,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042415893568309714,
      "loss": 5.0092,
      "mean_token_accuracy": 0.18290756195783614,
      "num_tokens": 61043651.0,
      "step": 27865
    },
    {
      "entropy": 5.560627794265747,
      "epoch": 2.765978562921794,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042413240688842884,
      "loss": 4.9426,
      "mean_token_accuracy": 0.18855095654726028,
      "num_tokens": 61053234.0,
      "step": 27870
    },
    {
      "entropy": 5.624645185470581,
      "epoch": 2.766474791583962,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004241058743955982,
      "loss": 5.0085,
      "mean_token_accuracy": 0.18277142494916915,
      "num_tokens": 61065178.0,
      "step": 27875
    },
    {
      "entropy": 5.637347316741943,
      "epoch": 2.7669710202461295,
      "grad_norm": 1.125,
      "learning_rate": 0.0004240793382052633,
      "loss": 4.9719,
      "mean_token_accuracy": 0.19334985911846161,
      "num_tokens": 61075709.0,
      "step": 27880
    },
    {
      "entropy": 5.6982612133026125,
      "epoch": 2.767467248908297,
      "grad_norm": 1.015625,
      "learning_rate": 0.000424052798318082,
      "loss": 5.0763,
      "mean_token_accuracy": 0.17753794193267822,
      "num_tokens": 61086880.0,
      "step": 27885
    },
    {
      "entropy": 5.626554298400879,
      "epoch": 2.7679634775704645,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004240262547347124,
      "loss": 5.0265,
      "mean_token_accuracy": 0.18925384134054185,
      "num_tokens": 61097008.0,
      "step": 27890
    },
    {
      "entropy": 5.648910045623779,
      "epoch": 2.768459706232632,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042399970745581283,
      "loss": 5.0026,
      "mean_token_accuracy": 0.18541458249092102,
      "num_tokens": 61107844.0,
      "step": 27895
    },
    {
      "entropy": 5.599969673156738,
      "epoch": 2.7689559348947994,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042397315648204164,
      "loss": 4.9483,
      "mean_token_accuracy": 0.19646897166967392,
      "num_tokens": 61118652.0,
      "step": 27900
    },
    {
      "entropy": 5.684444570541382,
      "epoch": 2.7694521635569673,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042394660181405716,
      "loss": 5.0942,
      "mean_token_accuracy": 0.1798645794391632,
      "num_tokens": 61129143.0,
      "step": 27905
    },
    {
      "entropy": 5.678222513198852,
      "epoch": 2.769948392219135,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004239200434525179,
      "loss": 5.0303,
      "mean_token_accuracy": 0.18377641439437867,
      "num_tokens": 61140623.0,
      "step": 27910
    },
    {
      "entropy": 5.619979906082153,
      "epoch": 2.7704446208813023,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004238934813980824,
      "loss": 5.0076,
      "mean_token_accuracy": 0.17755461782217025,
      "num_tokens": 61150998.0,
      "step": 27915
    },
    {
      "entropy": 5.667013502120971,
      "epoch": 2.7709408495434698,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042386691565140947,
      "loss": 5.0306,
      "mean_token_accuracy": 0.1835515469312668,
      "num_tokens": 61162086.0,
      "step": 27920
    },
    {
      "entropy": 5.69835057258606,
      "epoch": 2.7714370782056372,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042384034621315773,
      "loss": 5.0051,
      "mean_token_accuracy": 0.18042462319135666,
      "num_tokens": 61171806.0,
      "step": 27925
    },
    {
      "entropy": 5.668163299560547,
      "epoch": 2.7719333068678047,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004238137730839862,
      "loss": 4.9971,
      "mean_token_accuracy": 0.183659228682518,
      "num_tokens": 61182056.0,
      "step": 27930
    },
    {
      "entropy": 5.66475887298584,
      "epoch": 2.772429535529972,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004237871962645538,
      "loss": 5.0229,
      "mean_token_accuracy": 0.18746213912963866,
      "num_tokens": 61193231.0,
      "step": 27935
    },
    {
      "entropy": 5.60828046798706,
      "epoch": 2.7729257641921397,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042376061575551946,
      "loss": 4.9933,
      "mean_token_accuracy": 0.18843139857053756,
      "num_tokens": 61203785.0,
      "step": 27940
    },
    {
      "entropy": 5.638974189758301,
      "epoch": 2.773421992854307,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004237340315575425,
      "loss": 5.0169,
      "mean_token_accuracy": 0.19000605046749114,
      "num_tokens": 61215244.0,
      "step": 27945
    },
    {
      "entropy": 5.6907429695129395,
      "epoch": 2.7739182215164746,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004237074436712819,
      "loss": 4.997,
      "mean_token_accuracy": 0.18459442406892776,
      "num_tokens": 61227147.0,
      "step": 27950
    },
    {
      "entropy": 5.686746978759766,
      "epoch": 2.774414450178642,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042368085209739715,
      "loss": 5.062,
      "mean_token_accuracy": 0.18596989065408706,
      "num_tokens": 61237687.0,
      "step": 27955
    },
    {
      "entropy": 5.580123376846314,
      "epoch": 2.77491067884081,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004236542568365478,
      "loss": 4.9691,
      "mean_token_accuracy": 0.18987889140844344,
      "num_tokens": 61247700.0,
      "step": 27960
    },
    {
      "entropy": 5.680758476257324,
      "epoch": 2.7754069075029775,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00042362765788939307,
      "loss": 4.9749,
      "mean_token_accuracy": 0.1839569866657257,
      "num_tokens": 61258407.0,
      "step": 27965
    },
    {
      "entropy": 5.746888160705566,
      "epoch": 2.775903136165145,
      "grad_norm": 1.25,
      "learning_rate": 0.0004236010552565927,
      "loss": 5.1248,
      "mean_token_accuracy": 0.17564458549022674,
      "num_tokens": 61269065.0,
      "step": 27970
    },
    {
      "entropy": 5.710963392257691,
      "epoch": 2.7763993648273124,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042357444893880645,
      "loss": 5.0339,
      "mean_token_accuracy": 0.17973407208919526,
      "num_tokens": 61280767.0,
      "step": 27975
    },
    {
      "entropy": 5.611778783798218,
      "epoch": 2.77689559348948,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042354783893669385,
      "loss": 4.969,
      "mean_token_accuracy": 0.18546285927295686,
      "num_tokens": 61291967.0,
      "step": 27980
    },
    {
      "entropy": 5.700255250930786,
      "epoch": 2.7773918221516474,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004235212252509151,
      "loss": 5.1421,
      "mean_token_accuracy": 0.17453834265470505,
      "num_tokens": 61303572.0,
      "step": 27985
    },
    {
      "entropy": 5.691414260864258,
      "epoch": 2.777888050813815,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042349460788212995,
      "loss": 5.0502,
      "mean_token_accuracy": 0.18491690307855607,
      "num_tokens": 61315674.0,
      "step": 27990
    },
    {
      "entropy": 5.658632564544678,
      "epoch": 2.7783842794759828,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042346798683099844,
      "loss": 4.9935,
      "mean_token_accuracy": 0.18339915871620177,
      "num_tokens": 61326334.0,
      "step": 27995
    },
    {
      "entropy": 5.657138395309448,
      "epoch": 2.7788805081381502,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004234413620981809,
      "loss": 5.1137,
      "mean_token_accuracy": 0.17573143243789674,
      "num_tokens": 61337505.0,
      "step": 28000
    },
    {
      "entropy": 5.665828132629395,
      "epoch": 2.7793767368003177,
      "grad_norm": 1.390625,
      "learning_rate": 0.00042341473368433735,
      "loss": 5.054,
      "mean_token_accuracy": 0.18909869939088822,
      "num_tokens": 61348451.0,
      "step": 28005
    },
    {
      "entropy": 5.770749568939209,
      "epoch": 2.779872965462485,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004233881015901282,
      "loss": 5.2198,
      "mean_token_accuracy": 0.17041282206773758,
      "num_tokens": 61358356.0,
      "step": 28010
    },
    {
      "entropy": 5.794288921356201,
      "epoch": 2.7803691941246527,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004233614658162139,
      "loss": 5.0627,
      "mean_token_accuracy": 0.18400731384754182,
      "num_tokens": 61370996.0,
      "step": 28015
    },
    {
      "entropy": 5.753354120254516,
      "epoch": 2.78086542278682,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042333482636325504,
      "loss": 5.0831,
      "mean_token_accuracy": 0.1783536732196808,
      "num_tokens": 61383928.0,
      "step": 28020
    },
    {
      "entropy": 5.6646098613739015,
      "epoch": 2.7813616514489876,
      "grad_norm": 1.109375,
      "learning_rate": 0.000423308183231912,
      "loss": 5.0609,
      "mean_token_accuracy": 0.1802104204893112,
      "num_tokens": 61394277.0,
      "step": 28025
    },
    {
      "entropy": 5.608008527755738,
      "epoch": 2.781857880111155,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042328153642284567,
      "loss": 5.0247,
      "mean_token_accuracy": 0.18986217826604843,
      "num_tokens": 61405928.0,
      "step": 28030
    },
    {
      "entropy": 5.722405624389649,
      "epoch": 2.7823541087733226,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004232548859367167,
      "loss": 5.0888,
      "mean_token_accuracy": 0.17905291467905043,
      "num_tokens": 61417893.0,
      "step": 28035
    },
    {
      "entropy": 5.681992387771606,
      "epoch": 2.78285033743549,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000423228231774186,
      "loss": 5.0465,
      "mean_token_accuracy": 0.18080241978168488,
      "num_tokens": 61428099.0,
      "step": 28040
    },
    {
      "entropy": 5.626872920989991,
      "epoch": 2.783346566097658,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004232015739359146,
      "loss": 4.9758,
      "mean_token_accuracy": 0.19101071059703828,
      "num_tokens": 61437636.0,
      "step": 28045
    },
    {
      "entropy": 5.5978922843933105,
      "epoch": 2.7838427947598254,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042317491242256346,
      "loss": 4.9631,
      "mean_token_accuracy": 0.18484254479408263,
      "num_tokens": 61448038.0,
      "step": 28050
    },
    {
      "entropy": 5.647592210769654,
      "epoch": 2.784339023421993,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004231482472347938,
      "loss": 5.018,
      "mean_token_accuracy": 0.1812911331653595,
      "num_tokens": 61460258.0,
      "step": 28055
    },
    {
      "entropy": 5.654427099227905,
      "epoch": 2.7848352520841604,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004231215783732668,
      "loss": 5.0402,
      "mean_token_accuracy": 0.18674118369817733,
      "num_tokens": 61472274.0,
      "step": 28060
    },
    {
      "entropy": 5.6604455471038815,
      "epoch": 2.785331480746328,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042309490583864395,
      "loss": 5.0376,
      "mean_token_accuracy": 0.18891888111829758,
      "num_tokens": 61482975.0,
      "step": 28065
    },
    {
      "entropy": 5.701580572128296,
      "epoch": 2.7858277094084953,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004230682296315865,
      "loss": 5.0448,
      "mean_token_accuracy": 0.179292431473732,
      "num_tokens": 61494223.0,
      "step": 28070
    },
    {
      "entropy": 5.72736873626709,
      "epoch": 2.786323938070663,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000423041549752756,
      "loss": 5.1098,
      "mean_token_accuracy": 0.18131695985794066,
      "num_tokens": 61505981.0,
      "step": 28075
    },
    {
      "entropy": 5.583301401138305,
      "epoch": 2.7868201667328307,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004230148662028141,
      "loss": 4.9878,
      "mean_token_accuracy": 0.18946752548217774,
      "num_tokens": 61516410.0,
      "step": 28080
    },
    {
      "entropy": 5.630557680130005,
      "epoch": 2.787316395394998,
      "grad_norm": 1.125,
      "learning_rate": 0.00042298817898242237,
      "loss": 4.9901,
      "mean_token_accuracy": 0.18124301880598068,
      "num_tokens": 61526551.0,
      "step": 28085
    },
    {
      "entropy": 5.710667085647583,
      "epoch": 2.7878126240571657,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042296148809224275,
      "loss": 5.1167,
      "mean_token_accuracy": 0.17707726657390593,
      "num_tokens": 61538214.0,
      "step": 28090
    },
    {
      "entropy": 5.6909558296203615,
      "epoch": 2.788308852719333,
      "grad_norm": 1.125,
      "learning_rate": 0.0004229347935329371,
      "loss": 5.0806,
      "mean_token_accuracy": 0.1753709003329277,
      "num_tokens": 61549726.0,
      "step": 28095
    },
    {
      "entropy": 5.756187677383423,
      "epoch": 2.7888050813815006,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042290809530516724,
      "loss": 5.1598,
      "mean_token_accuracy": 0.1713593378663063,
      "num_tokens": 61560070.0,
      "step": 28100
    },
    {
      "entropy": 5.772004556655884,
      "epoch": 2.789301310043668,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004228813934095954,
      "loss": 5.0898,
      "mean_token_accuracy": 0.18039618581533431,
      "num_tokens": 61571348.0,
      "step": 28105
    },
    {
      "entropy": 5.649619483947754,
      "epoch": 2.7897975387058356,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042285468784688363,
      "loss": 4.966,
      "mean_token_accuracy": 0.18741677552461625,
      "num_tokens": 61581981.0,
      "step": 28110
    },
    {
      "entropy": 5.673765993118286,
      "epoch": 2.790293767368003,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042282797861769423,
      "loss": 5.0651,
      "mean_token_accuracy": 0.18347305953502654,
      "num_tokens": 61591248.0,
      "step": 28115
    },
    {
      "entropy": 5.6478527069091795,
      "epoch": 2.7907899960301705,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004228012657226895,
      "loss": 5.0705,
      "mean_token_accuracy": 0.18405288755893706,
      "num_tokens": 61602432.0,
      "step": 28120
    },
    {
      "entropy": 5.589995670318603,
      "epoch": 2.791286224692338,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042277454916253186,
      "loss": 5.0261,
      "mean_token_accuracy": 0.18209776729345323,
      "num_tokens": 61613418.0,
      "step": 28125
    },
    {
      "entropy": 5.665133094787597,
      "epoch": 2.791782453354506,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004227478289378838,
      "loss": 5.0237,
      "mean_token_accuracy": 0.17735270857810975,
      "num_tokens": 61623783.0,
      "step": 28130
    },
    {
      "entropy": 5.743353796005249,
      "epoch": 2.7922786820166734,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042272110504940795,
      "loss": 5.103,
      "mean_token_accuracy": 0.17963837087154388,
      "num_tokens": 61633838.0,
      "step": 28135
    },
    {
      "entropy": 5.6544239044189455,
      "epoch": 2.792774910678841,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042269437749776703,
      "loss": 4.9844,
      "mean_token_accuracy": 0.18761261254549028,
      "num_tokens": 61645982.0,
      "step": 28140
    },
    {
      "entropy": 5.63283748626709,
      "epoch": 2.7932711393410083,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004226676462836238,
      "loss": 4.9593,
      "mean_token_accuracy": 0.18781468719244004,
      "num_tokens": 61656030.0,
      "step": 28145
    },
    {
      "entropy": 5.590327453613281,
      "epoch": 2.793767368003176,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004226409114076411,
      "loss": 4.9868,
      "mean_token_accuracy": 0.18320775479078294,
      "num_tokens": 61666746.0,
      "step": 28150
    },
    {
      "entropy": 5.725075435638428,
      "epoch": 2.7942635966653433,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042261417287048196,
      "loss": 5.1511,
      "mean_token_accuracy": 0.17717798352241515,
      "num_tokens": 61677681.0,
      "step": 28155
    },
    {
      "entropy": 5.7714661121368405,
      "epoch": 2.7947598253275108,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004225874306728094,
      "loss": 5.109,
      "mean_token_accuracy": 0.1773577257990837,
      "num_tokens": 61689564.0,
      "step": 28160
    },
    {
      "entropy": 5.654994869232178,
      "epoch": 2.7952560539896787,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042256068481528666,
      "loss": 5.0017,
      "mean_token_accuracy": 0.18242466151714326,
      "num_tokens": 61701074.0,
      "step": 28165
    },
    {
      "entropy": 5.595794248580932,
      "epoch": 2.795752282651846,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004225339352985768,
      "loss": 5.0013,
      "mean_token_accuracy": 0.19210665225982665,
      "num_tokens": 61712203.0,
      "step": 28170
    },
    {
      "entropy": 5.681804847717285,
      "epoch": 2.7962485113140136,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042250718212334334,
      "loss": 4.9726,
      "mean_token_accuracy": 0.18289345800876616,
      "num_tokens": 61723362.0,
      "step": 28175
    },
    {
      "entropy": 5.672980928421021,
      "epoch": 2.796744739976181,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042248042529024956,
      "loss": 4.9795,
      "mean_token_accuracy": 0.18680357336997985,
      "num_tokens": 61733454.0,
      "step": 28180
    },
    {
      "entropy": 5.713920211791992,
      "epoch": 2.7972409686383486,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042245366479995894,
      "loss": 5.1049,
      "mean_token_accuracy": 0.17520588785409927,
      "num_tokens": 61745014.0,
      "step": 28185
    },
    {
      "entropy": 5.647981119155884,
      "epoch": 2.797737197300516,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042242690065313535,
      "loss": 5.0515,
      "mean_token_accuracy": 0.17964137494564056,
      "num_tokens": 61755334.0,
      "step": 28190
    },
    {
      "entropy": 5.603252744674682,
      "epoch": 2.7982334259626835,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042240013285044215,
      "loss": 4.9654,
      "mean_token_accuracy": 0.19349483251571656,
      "num_tokens": 61766024.0,
      "step": 28195
    },
    {
      "entropy": 5.675601863861084,
      "epoch": 2.798729654624851,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00042237336139254336,
      "loss": 5.0329,
      "mean_token_accuracy": 0.1901106134057045,
      "num_tokens": 61778228.0,
      "step": 28200
    },
    {
      "entropy": 5.690108346939087,
      "epoch": 2.7992258832870185,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042234658628010274,
      "loss": 5.0628,
      "mean_token_accuracy": 0.17650038301944732,
      "num_tokens": 61789072.0,
      "step": 28205
    },
    {
      "entropy": 5.696349239349365,
      "epoch": 2.799722111949186,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042231980751378435,
      "loss": 5.0218,
      "mean_token_accuracy": 0.19146714806556703,
      "num_tokens": 61799503.0,
      "step": 28210
    },
    {
      "entropy": 5.675558805465698,
      "epoch": 2.8002183406113534,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042229302509425205,
      "loss": 4.9801,
      "mean_token_accuracy": 0.18714869022369385,
      "num_tokens": 61811063.0,
      "step": 28215
    },
    {
      "entropy": 5.6823896884918215,
      "epoch": 2.8007145692735214,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004222662390221702,
      "loss": 5.0887,
      "mean_token_accuracy": 0.17899414300918579,
      "num_tokens": 61822832.0,
      "step": 28220
    },
    {
      "entropy": 5.678816223144532,
      "epoch": 2.801210797935689,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004222394492982029,
      "loss": 5.0634,
      "mean_token_accuracy": 0.1785065695643425,
      "num_tokens": 61832456.0,
      "step": 28225
    },
    {
      "entropy": 5.684810972213745,
      "epoch": 2.8017070265978563,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004222126559230146,
      "loss": 4.9934,
      "mean_token_accuracy": 0.18938933461904525,
      "num_tokens": 61843043.0,
      "step": 28230
    },
    {
      "entropy": 5.628162288665772,
      "epoch": 2.802203255260024,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004221858588972696,
      "loss": 5.0419,
      "mean_token_accuracy": 0.17900165766477585,
      "num_tokens": 61853249.0,
      "step": 28235
    },
    {
      "entropy": 5.580795335769653,
      "epoch": 2.8026994839221913,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042215905822163245,
      "loss": 4.9643,
      "mean_token_accuracy": 0.18895333260297775,
      "num_tokens": 61863842.0,
      "step": 28240
    },
    {
      "entropy": 5.685923290252686,
      "epoch": 2.8031957125843587,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004221322538967678,
      "loss": 5.0659,
      "mean_token_accuracy": 0.18261933326721191,
      "num_tokens": 61876924.0,
      "step": 28245
    },
    {
      "entropy": 5.737472915649414,
      "epoch": 2.8036919412465267,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004221054459233403,
      "loss": 5.0821,
      "mean_token_accuracy": 0.17682996094226838,
      "num_tokens": 61886766.0,
      "step": 28250
    },
    {
      "entropy": 5.7418200969696045,
      "epoch": 2.804188169908694,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042207863430201466,
      "loss": 5.0461,
      "mean_token_accuracy": 0.1869405150413513,
      "num_tokens": 61896538.0,
      "step": 28255
    },
    {
      "entropy": 5.6336585521698,
      "epoch": 2.8046843985708616,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042205181903345586,
      "loss": 5.0237,
      "mean_token_accuracy": 0.17758622467517854,
      "num_tokens": 61907328.0,
      "step": 28260
    },
    {
      "entropy": 5.663162755966186,
      "epoch": 2.805180627233029,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004220250001183288,
      "loss": 5.0479,
      "mean_token_accuracy": 0.17910438179969787,
      "num_tokens": 61918797.0,
      "step": 28265
    },
    {
      "entropy": 5.689649057388306,
      "epoch": 2.8056768558951966,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004219981775572986,
      "loss": 5.0777,
      "mean_token_accuracy": 0.18622310161590577,
      "num_tokens": 61928755.0,
      "step": 28270
    },
    {
      "entropy": 5.668153429031372,
      "epoch": 2.806173084557364,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004219713513510303,
      "loss": 5.0476,
      "mean_token_accuracy": 0.18034436255693437,
      "num_tokens": 61939993.0,
      "step": 28275
    },
    {
      "entropy": 5.606246757507324,
      "epoch": 2.8066693132195315,
      "grad_norm": 1.125,
      "learning_rate": 0.00042194452150018917,
      "loss": 4.9078,
      "mean_token_accuracy": 0.19675877541303635,
      "num_tokens": 61950982.0,
      "step": 28280
    },
    {
      "entropy": 5.684019088745117,
      "epoch": 2.807165541881699,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042191768800544055,
      "loss": 5.0138,
      "mean_token_accuracy": 0.17976718842983247,
      "num_tokens": 61961786.0,
      "step": 28285
    },
    {
      "entropy": 5.6743631839752195,
      "epoch": 2.8076617705438665,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042189085086744986,
      "loss": 5.0069,
      "mean_token_accuracy": 0.189923095703125,
      "num_tokens": 61972548.0,
      "step": 28290
    },
    {
      "entropy": 5.629531192779541,
      "epoch": 2.808157999206034,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004218640100868826,
      "loss": 4.9892,
      "mean_token_accuracy": 0.18941531032323838,
      "num_tokens": 61981822.0,
      "step": 28295
    },
    {
      "entropy": 5.665629816055298,
      "epoch": 2.8086542278682014,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042183716566440433,
      "loss": 5.1242,
      "mean_token_accuracy": 0.17470804899930953,
      "num_tokens": 61992997.0,
      "step": 28300
    },
    {
      "entropy": 5.6821510791778564,
      "epoch": 2.8091504565303693,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004218103176006808,
      "loss": 5.06,
      "mean_token_accuracy": 0.18029679507017135,
      "num_tokens": 62004391.0,
      "step": 28305
    },
    {
      "entropy": 5.662803697586059,
      "epoch": 2.809646685192537,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004217834658963777,
      "loss": 5.02,
      "mean_token_accuracy": 0.1890384703874588,
      "num_tokens": 62014099.0,
      "step": 28310
    },
    {
      "entropy": 5.679895067214966,
      "epoch": 2.8101429138547043,
      "grad_norm": 1.125,
      "learning_rate": 0.00042175661055216093,
      "loss": 5.0003,
      "mean_token_accuracy": 0.1839670866727829,
      "num_tokens": 62023872.0,
      "step": 28315
    },
    {
      "entropy": 5.681362295150757,
      "epoch": 2.8106391425168717,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004217297515686965,
      "loss": 5.0021,
      "mean_token_accuracy": 0.1809683009982109,
      "num_tokens": 62035566.0,
      "step": 28320
    },
    {
      "entropy": 5.666706514358521,
      "epoch": 2.811135371179039,
      "grad_norm": 1.125,
      "learning_rate": 0.0004217028889466503,
      "loss": 5.0359,
      "mean_token_accuracy": 0.18058500736951827,
      "num_tokens": 62046742.0,
      "step": 28325
    },
    {
      "entropy": 5.6568090438842775,
      "epoch": 2.8116315998412067,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004216760226866887,
      "loss": 4.9912,
      "mean_token_accuracy": 0.18467239290475845,
      "num_tokens": 62056545.0,
      "step": 28330
    },
    {
      "entropy": 5.707092761993408,
      "epoch": 2.8121278285033746,
      "grad_norm": 1.25,
      "learning_rate": 0.00042164915278947775,
      "loss": 5.0848,
      "mean_token_accuracy": 0.18467617332935332,
      "num_tokens": 62066414.0,
      "step": 28335
    },
    {
      "entropy": 5.636412143707275,
      "epoch": 2.812624057165542,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004216222792556838,
      "loss": 4.9774,
      "mean_token_accuracy": 0.18576282411813735,
      "num_tokens": 62078406.0,
      "step": 28340
    },
    {
      "entropy": 5.637642765045166,
      "epoch": 2.8131202858277096,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004215954020859732,
      "loss": 5.0423,
      "mean_token_accuracy": 0.18230371326208114,
      "num_tokens": 62089516.0,
      "step": 28345
    },
    {
      "entropy": 5.645133113861084,
      "epoch": 2.813616514489877,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042156852128101264,
      "loss": 5.011,
      "mean_token_accuracy": 0.184519462287426,
      "num_tokens": 62099287.0,
      "step": 28350
    },
    {
      "entropy": 5.621774578094483,
      "epoch": 2.8141127431520445,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042154163684146855,
      "loss": 4.9793,
      "mean_token_accuracy": 0.1855016365647316,
      "num_tokens": 62111110.0,
      "step": 28355
    },
    {
      "entropy": 5.779215860366821,
      "epoch": 2.814608971814212,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004215147487680076,
      "loss": 5.1513,
      "mean_token_accuracy": 0.1757354184985161,
      "num_tokens": 62123614.0,
      "step": 28360
    },
    {
      "entropy": 5.675618648529053,
      "epoch": 2.8151052004763795,
      "grad_norm": 1.125,
      "learning_rate": 0.00042148785706129657,
      "loss": 5.0965,
      "mean_token_accuracy": 0.17858851402997972,
      "num_tokens": 62133682.0,
      "step": 28365
    },
    {
      "entropy": 5.631322956085205,
      "epoch": 2.815601429138547,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004214609617220024,
      "loss": 5.0112,
      "mean_token_accuracy": 0.18617005944252013,
      "num_tokens": 62144802.0,
      "step": 28370
    },
    {
      "entropy": 5.671516609191895,
      "epoch": 2.8160976578007144,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042143406275079186,
      "loss": 4.9964,
      "mean_token_accuracy": 0.17905711829662324,
      "num_tokens": 62154205.0,
      "step": 28375
    },
    {
      "entropy": 5.604723119735718,
      "epoch": 2.816593886462882,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004214071601483322,
      "loss": 4.9141,
      "mean_token_accuracy": 0.195522503554821,
      "num_tokens": 62164648.0,
      "step": 28380
    },
    {
      "entropy": 5.639846706390381,
      "epoch": 2.8170901151250494,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004213802539152905,
      "loss": 5.0146,
      "mean_token_accuracy": 0.18699899911880494,
      "num_tokens": 62176073.0,
      "step": 28385
    },
    {
      "entropy": 5.714209842681885,
      "epoch": 2.8175863437872173,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042135334405233384,
      "loss": 5.0527,
      "mean_token_accuracy": 0.17635604441165925,
      "num_tokens": 62187350.0,
      "step": 28390
    },
    {
      "entropy": 5.622926378250122,
      "epoch": 2.8180825724493848,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004213264305601296,
      "loss": 4.9699,
      "mean_token_accuracy": 0.18220000118017196,
      "num_tokens": 62197729.0,
      "step": 28395
    },
    {
      "entropy": 5.61753830909729,
      "epoch": 2.8185788011115522,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042129951343934527,
      "loss": 4.965,
      "mean_token_accuracy": 0.19263747483491897,
      "num_tokens": 62209047.0,
      "step": 28400
    },
    {
      "entropy": 5.653810691833496,
      "epoch": 2.8190750297737197,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004212725926906482,
      "loss": 5.0022,
      "mean_token_accuracy": 0.18166322112083436,
      "num_tokens": 62218859.0,
      "step": 28405
    },
    {
      "entropy": 5.623528861999512,
      "epoch": 2.819571258435887,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042124566831470604,
      "loss": 5.0077,
      "mean_token_accuracy": 0.18714044243097305,
      "num_tokens": 62230396.0,
      "step": 28410
    },
    {
      "entropy": 5.627466011047363,
      "epoch": 2.8200674870980547,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004212187403121864,
      "loss": 4.9761,
      "mean_token_accuracy": 0.1889481484889984,
      "num_tokens": 62241790.0,
      "step": 28415
    },
    {
      "entropy": 5.629418802261353,
      "epoch": 2.8205637157602226,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042119180868375707,
      "loss": 4.9519,
      "mean_token_accuracy": 0.18634444773197173,
      "num_tokens": 62253820.0,
      "step": 28420
    },
    {
      "entropy": 5.6284752368927,
      "epoch": 2.82105994442239,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004211648734300859,
      "loss": 5.0174,
      "mean_token_accuracy": 0.18489817827939986,
      "num_tokens": 62263875.0,
      "step": 28425
    },
    {
      "entropy": 5.6528185367584225,
      "epoch": 2.8215561730845575,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042113793455184084,
      "loss": 5.0512,
      "mean_token_accuracy": 0.1777609333395958,
      "num_tokens": 62275390.0,
      "step": 28430
    },
    {
      "entropy": 5.659482431411743,
      "epoch": 2.822052401746725,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004211109920496899,
      "loss": 5.0712,
      "mean_token_accuracy": 0.17206584960222243,
      "num_tokens": 62286969.0,
      "step": 28435
    },
    {
      "entropy": 5.6668007373809814,
      "epoch": 2.8225486304088925,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042108404592430126,
      "loss": 5.021,
      "mean_token_accuracy": 0.18364531695842742,
      "num_tokens": 62297810.0,
      "step": 28440
    },
    {
      "entropy": 5.655645227432251,
      "epoch": 2.82304485907106,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000421057096176343,
      "loss": 5.022,
      "mean_token_accuracy": 0.18361466825008393,
      "num_tokens": 62309474.0,
      "step": 28445
    },
    {
      "entropy": 5.630173826217652,
      "epoch": 2.8235410877332274,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004210301428064835,
      "loss": 5.0435,
      "mean_token_accuracy": 0.18389743864536284,
      "num_tokens": 62320022.0,
      "step": 28450
    },
    {
      "entropy": 5.722822952270508,
      "epoch": 2.824037316395395,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004210031858153911,
      "loss": 5.0291,
      "mean_token_accuracy": 0.189593905210495,
      "num_tokens": 62330450.0,
      "step": 28455
    },
    {
      "entropy": 5.680014896392822,
      "epoch": 2.8245335450575624,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042097622520373434,
      "loss": 5.0447,
      "mean_token_accuracy": 0.18369331061840058,
      "num_tokens": 62341419.0,
      "step": 28460
    },
    {
      "entropy": 5.693494653701782,
      "epoch": 2.82502977371973,
      "grad_norm": 1.125,
      "learning_rate": 0.0004209492609721817,
      "loss": 5.0892,
      "mean_token_accuracy": 0.17600410729646682,
      "num_tokens": 62353312.0,
      "step": 28465
    },
    {
      "entropy": 5.694500207901001,
      "epoch": 2.8255260023818973,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004209222931214019,
      "loss": 5.0307,
      "mean_token_accuracy": 0.18305234909057616,
      "num_tokens": 62364838.0,
      "step": 28470
    },
    {
      "entropy": 5.637123107910156,
      "epoch": 2.8260222310440652,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042089532165206367,
      "loss": 4.9516,
      "mean_token_accuracy": 0.1935519829392433,
      "num_tokens": 62376443.0,
      "step": 28475
    },
    {
      "entropy": 5.710754489898681,
      "epoch": 2.8265184597062327,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004208683465648358,
      "loss": 5.1323,
      "mean_token_accuracy": 0.17581234276294708,
      "num_tokens": 62387263.0,
      "step": 28480
    },
    {
      "entropy": 5.662522268295288,
      "epoch": 2.8270146883684,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042084136786038737,
      "loss": 4.9558,
      "mean_token_accuracy": 0.1868123933672905,
      "num_tokens": 62398929.0,
      "step": 28485
    },
    {
      "entropy": 5.6845917224884035,
      "epoch": 2.8275109170305677,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004208143855393872,
      "loss": 5.078,
      "mean_token_accuracy": 0.1778879225254059,
      "num_tokens": 62409516.0,
      "step": 28490
    },
    {
      "entropy": 5.723330879211426,
      "epoch": 2.828007145692735,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042078739960250446,
      "loss": 5.0331,
      "mean_token_accuracy": 0.18702166974544526,
      "num_tokens": 62421856.0,
      "step": 28495
    },
    {
      "entropy": 5.691647720336914,
      "epoch": 2.8285033743549026,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042076041005040835,
      "loss": 4.9742,
      "mean_token_accuracy": 0.1848184734582901,
      "num_tokens": 62432340.0,
      "step": 28500
    },
    {
      "entropy": 5.650847101211548,
      "epoch": 2.8289996030170705,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004207334168837682,
      "loss": 5.051,
      "mean_token_accuracy": 0.17580082267522812,
      "num_tokens": 62442652.0,
      "step": 28505
    },
    {
      "entropy": 5.685147142410278,
      "epoch": 2.829495831679238,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004207064201032533,
      "loss": 4.9843,
      "mean_token_accuracy": 0.18961671590805054,
      "num_tokens": 62454503.0,
      "step": 28510
    },
    {
      "entropy": 5.64896674156189,
      "epoch": 2.8299920603414055,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042067941970953316,
      "loss": 4.9664,
      "mean_token_accuracy": 0.18370924592018129,
      "num_tokens": 62464858.0,
      "step": 28515
    },
    {
      "entropy": 5.6951006889343265,
      "epoch": 2.830488289003573,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042065241570327734,
      "loss": 5.0496,
      "mean_token_accuracy": 0.18008592128753662,
      "num_tokens": 62475153.0,
      "step": 28520
    },
    {
      "entropy": 5.6690812587738035,
      "epoch": 2.8309845176657404,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004206254080851555,
      "loss": 5.0875,
      "mean_token_accuracy": 0.1764039844274521,
      "num_tokens": 62487883.0,
      "step": 28525
    },
    {
      "entropy": 5.636592721939087,
      "epoch": 2.831480746327908,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042059839685583726,
      "loss": 4.9976,
      "mean_token_accuracy": 0.18572997748851777,
      "num_tokens": 62499905.0,
      "step": 28530
    },
    {
      "entropy": 5.694833707809448,
      "epoch": 2.8319769749900754,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042057138201599247,
      "loss": 5.0686,
      "mean_token_accuracy": 0.178301540017128,
      "num_tokens": 62511403.0,
      "step": 28535
    },
    {
      "entropy": 5.688872671127319,
      "epoch": 2.832473203652243,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042054436356629113,
      "loss": 5.0115,
      "mean_token_accuracy": 0.18995088040828706,
      "num_tokens": 62522543.0,
      "step": 28540
    },
    {
      "entropy": 5.638876867294312,
      "epoch": 2.8329694323144103,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004205173415074032,
      "loss": 4.986,
      "mean_token_accuracy": 0.18605080544948577,
      "num_tokens": 62533053.0,
      "step": 28545
    },
    {
      "entropy": 5.661490964889526,
      "epoch": 2.833465660976578,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004204903158399989,
      "loss": 5.1023,
      "mean_token_accuracy": 0.17402459681034088,
      "num_tokens": 62543699.0,
      "step": 28550
    },
    {
      "entropy": 5.631885957717896,
      "epoch": 2.8339618896387453,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004204632865647481,
      "loss": 5.0426,
      "mean_token_accuracy": 0.19033814072608948,
      "num_tokens": 62554121.0,
      "step": 28555
    },
    {
      "entropy": 5.693836736679077,
      "epoch": 2.834458118300913,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004204362536823213,
      "loss": 5.0004,
      "mean_token_accuracy": 0.18508762419223784,
      "num_tokens": 62564010.0,
      "step": 28560
    },
    {
      "entropy": 5.636257362365723,
      "epoch": 2.8349543469630807,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042040921719338877,
      "loss": 4.9572,
      "mean_token_accuracy": 0.19426522105932237,
      "num_tokens": 62574451.0,
      "step": 28565
    },
    {
      "entropy": 5.644767713546753,
      "epoch": 2.835450575625248,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000420382177098621,
      "loss": 5.0199,
      "mean_token_accuracy": 0.18268250674009323,
      "num_tokens": 62585298.0,
      "step": 28570
    },
    {
      "entropy": 5.669913864135742,
      "epoch": 2.8359468042874156,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042035513339868855,
      "loss": 5.0757,
      "mean_token_accuracy": 0.18502389043569564,
      "num_tokens": 62596248.0,
      "step": 28575
    },
    {
      "entropy": 5.719183111190796,
      "epoch": 2.836443032949583,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004203280860942619,
      "loss": 5.051,
      "mean_token_accuracy": 0.17653724551200867,
      "num_tokens": 62607233.0,
      "step": 28580
    },
    {
      "entropy": 5.677748155593872,
      "epoch": 2.8369392616117506,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004203010351860119,
      "loss": 4.9581,
      "mean_token_accuracy": 0.19314736723899842,
      "num_tokens": 62618564.0,
      "step": 28585
    },
    {
      "entropy": 5.639161777496338,
      "epoch": 2.837435490273918,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004202739806746094,
      "loss": 5.0455,
      "mean_token_accuracy": 0.18123247772455214,
      "num_tokens": 62630049.0,
      "step": 28590
    },
    {
      "entropy": 5.681370830535888,
      "epoch": 2.837931718936086,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004202469225607252,
      "loss": 5.1015,
      "mean_token_accuracy": 0.17373457551002502,
      "num_tokens": 62640887.0,
      "step": 28595
    },
    {
      "entropy": 5.686429119110107,
      "epoch": 2.8384279475982535,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004202198608450302,
      "loss": 5.0542,
      "mean_token_accuracy": 0.17749513387680055,
      "num_tokens": 62651816.0,
      "step": 28600
    },
    {
      "entropy": 5.654358720779419,
      "epoch": 2.838924176260421,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004201927955281957,
      "loss": 4.9968,
      "mean_token_accuracy": 0.1869315430521965,
      "num_tokens": 62662601.0,
      "step": 28605
    },
    {
      "entropy": 5.6344160556793215,
      "epoch": 2.8394204049225884,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004201657266108926,
      "loss": 5.02,
      "mean_token_accuracy": 0.1879321202635765,
      "num_tokens": 62673661.0,
      "step": 28610
    },
    {
      "entropy": 5.598662376403809,
      "epoch": 2.839916633584756,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042013865409379244,
      "loss": 4.971,
      "mean_token_accuracy": 0.1930139034986496,
      "num_tokens": 62683984.0,
      "step": 28615
    },
    {
      "entropy": 5.676898622512818,
      "epoch": 2.8404128622469234,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004201115779775663,
      "loss": 5.0363,
      "mean_token_accuracy": 0.1844118356704712,
      "num_tokens": 62695338.0,
      "step": 28620
    },
    {
      "entropy": 5.7235846519470215,
      "epoch": 2.840909090909091,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004200844982628857,
      "loss": 5.0933,
      "mean_token_accuracy": 0.18139946162700654,
      "num_tokens": 62706341.0,
      "step": 28625
    },
    {
      "entropy": 5.7188379764556885,
      "epoch": 2.8414053195712583,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004200574149504222,
      "loss": 5.076,
      "mean_token_accuracy": 0.18288009017705917,
      "num_tokens": 62717895.0,
      "step": 28630
    },
    {
      "entropy": 5.671785354614258,
      "epoch": 2.841901548233426,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004200303280408474,
      "loss": 5.0424,
      "mean_token_accuracy": 0.18094494044780732,
      "num_tokens": 62729745.0,
      "step": 28635
    },
    {
      "entropy": 5.650658988952637,
      "epoch": 2.8423977768955933,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042000323753483296,
      "loss": 5.0298,
      "mean_token_accuracy": 0.1883445903658867,
      "num_tokens": 62740298.0,
      "step": 28640
    },
    {
      "entropy": 5.622596311569214,
      "epoch": 2.842894005557761,
      "grad_norm": 1.0,
      "learning_rate": 0.00041997614343305067,
      "loss": 5.0039,
      "mean_token_accuracy": 0.18992750197649003,
      "num_tokens": 62751292.0,
      "step": 28645
    },
    {
      "entropy": 5.701124763488769,
      "epoch": 2.8433902342199286,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041994904573617247,
      "loss": 5.1227,
      "mean_token_accuracy": 0.17443469017744065,
      "num_tokens": 62761789.0,
      "step": 28650
    },
    {
      "entropy": 5.65762767791748,
      "epoch": 2.843886462882096,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041992194444487026,
      "loss": 5.014,
      "mean_token_accuracy": 0.18742000162601472,
      "num_tokens": 62773317.0,
      "step": 28655
    },
    {
      "entropy": 5.6544538021087645,
      "epoch": 2.8443826915442636,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041989483955981607,
      "loss": 4.9755,
      "mean_token_accuracy": 0.1935516431927681,
      "num_tokens": 62783780.0,
      "step": 28660
    },
    {
      "entropy": 5.667051219940186,
      "epoch": 2.844878920206431,
      "grad_norm": 1.125,
      "learning_rate": 0.00041986773108168216,
      "loss": 5.0217,
      "mean_token_accuracy": 0.18306943476200105,
      "num_tokens": 62794535.0,
      "step": 28665
    },
    {
      "entropy": 5.6355537414550785,
      "epoch": 2.8453751488685985,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041984061901114057,
      "loss": 5.0147,
      "mean_token_accuracy": 0.17857969403266907,
      "num_tokens": 62804410.0,
      "step": 28670
    },
    {
      "entropy": 5.716057634353637,
      "epoch": 2.845871377530766,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004198135033488638,
      "loss": 4.9972,
      "mean_token_accuracy": 0.19286677688360215,
      "num_tokens": 62814693.0,
      "step": 28675
    },
    {
      "entropy": 5.700736618041992,
      "epoch": 2.846367606192934,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041978638409552417,
      "loss": 5.0052,
      "mean_token_accuracy": 0.18607767075300216,
      "num_tokens": 62826114.0,
      "step": 28680
    },
    {
      "entropy": 5.6232171058654785,
      "epoch": 2.8468638348551014,
      "grad_norm": 1.0,
      "learning_rate": 0.0004197592612517942,
      "loss": 5.0466,
      "mean_token_accuracy": 0.18871924877166749,
      "num_tokens": 62838130.0,
      "step": 28685
    },
    {
      "entropy": 5.750789546966553,
      "epoch": 2.847360063517269,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004197321348183465,
      "loss": 5.0933,
      "mean_token_accuracy": 0.17876583486795425,
      "num_tokens": 62849290.0,
      "step": 28690
    },
    {
      "entropy": 5.755114269256592,
      "epoch": 2.8478562921794364,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041970500479585364,
      "loss": 5.1487,
      "mean_token_accuracy": 0.17861020117998122,
      "num_tokens": 62860888.0,
      "step": 28695
    },
    {
      "entropy": 5.697017192840576,
      "epoch": 2.848352520841604,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041967787118498854,
      "loss": 5.0638,
      "mean_token_accuracy": 0.18736823946237563,
      "num_tokens": 62870231.0,
      "step": 28700
    },
    {
      "entropy": 5.65846471786499,
      "epoch": 2.8488487495037713,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004196507339864241,
      "loss": 4.9755,
      "mean_token_accuracy": 0.185355481505394,
      "num_tokens": 62881147.0,
      "step": 28705
    },
    {
      "entropy": 5.689462089538575,
      "epoch": 2.849344978165939,
      "grad_norm": 1.109375,
      "learning_rate": 0.000419623593200833,
      "loss": 5.0001,
      "mean_token_accuracy": 0.18603104203939438,
      "num_tokens": 62891272.0,
      "step": 28710
    },
    {
      "entropy": 5.625908946990966,
      "epoch": 2.8498412068281063,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041959644882888844,
      "loss": 4.9955,
      "mean_token_accuracy": 0.18194704204797746,
      "num_tokens": 62902376.0,
      "step": 28715
    },
    {
      "entropy": 5.619397878646851,
      "epoch": 2.8503374354902737,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041956930087126357,
      "loss": 4.9874,
      "mean_token_accuracy": 0.19292793571949005,
      "num_tokens": 62913811.0,
      "step": 28720
    },
    {
      "entropy": 5.6445530414581295,
      "epoch": 2.850833664152441,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00041954214932863156,
      "loss": 5.0561,
      "mean_token_accuracy": 0.17890367209911345,
      "num_tokens": 62926001.0,
      "step": 28725
    },
    {
      "entropy": 5.646321535110474,
      "epoch": 2.851329892814609,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004195149942016657,
      "loss": 5.0101,
      "mean_token_accuracy": 0.1857346311211586,
      "num_tokens": 62936183.0,
      "step": 28730
    },
    {
      "entropy": 5.66461877822876,
      "epoch": 2.8518261214767766,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004194878354910394,
      "loss": 4.9736,
      "mean_token_accuracy": 0.1900187462568283,
      "num_tokens": 62947427.0,
      "step": 28735
    },
    {
      "entropy": 5.691399192810058,
      "epoch": 2.852322350138944,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041946067319742605,
      "loss": 5.058,
      "mean_token_accuracy": 0.186115363240242,
      "num_tokens": 62958630.0,
      "step": 28740
    },
    {
      "entropy": 5.684396171569825,
      "epoch": 2.8528185788011116,
      "grad_norm": 1.125,
      "learning_rate": 0.00041943350732149935,
      "loss": 4.9942,
      "mean_token_accuracy": 0.18539201319217682,
      "num_tokens": 62968856.0,
      "step": 28745
    },
    {
      "entropy": 5.650395679473877,
      "epoch": 2.853314807463279,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041940633786393295,
      "loss": 4.926,
      "mean_token_accuracy": 0.1918563053011894,
      "num_tokens": 62978393.0,
      "step": 28750
    },
    {
      "entropy": 5.62968111038208,
      "epoch": 2.8538110361254465,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00041937916482540045,
      "loss": 5.0232,
      "mean_token_accuracy": 0.18741750568151475,
      "num_tokens": 62989786.0,
      "step": 28755
    },
    {
      "entropy": 5.683053207397461,
      "epoch": 2.854307264787614,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041935198820657575,
      "loss": 5.0179,
      "mean_token_accuracy": 0.1869845524430275,
      "num_tokens": 63000021.0,
      "step": 28760
    },
    {
      "entropy": 5.6926967144012455,
      "epoch": 2.854803493449782,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004193248080081329,
      "loss": 5.0225,
      "mean_token_accuracy": 0.18031424880027772,
      "num_tokens": 63011398.0,
      "step": 28765
    },
    {
      "entropy": 5.709293365478516,
      "epoch": 2.8552997221119494,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004192976242307457,
      "loss": 5.0123,
      "mean_token_accuracy": 0.18273424804210664,
      "num_tokens": 63021155.0,
      "step": 28770
    },
    {
      "entropy": 5.714332723617554,
      "epoch": 2.855795950774117,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004192704368750884,
      "loss": 5.0456,
      "mean_token_accuracy": 0.1798575207591057,
      "num_tokens": 63031510.0,
      "step": 28775
    },
    {
      "entropy": 5.727404642105102,
      "epoch": 2.8562921794362843,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004192432459418351,
      "loss": 5.0551,
      "mean_token_accuracy": 0.18110570460557937,
      "num_tokens": 63041978.0,
      "step": 28780
    },
    {
      "entropy": 5.694017744064331,
      "epoch": 2.856788408098452,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004192160514316602,
      "loss": 5.0874,
      "mean_token_accuracy": 0.17707642465829848,
      "num_tokens": 63052194.0,
      "step": 28785
    },
    {
      "entropy": 5.595544528961182,
      "epoch": 2.8572846367606193,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004191888533452379,
      "loss": 4.9404,
      "mean_token_accuracy": 0.18789842128753662,
      "num_tokens": 63062144.0,
      "step": 28790
    },
    {
      "entropy": 5.634398794174194,
      "epoch": 2.8577808654227868,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004191616516832428,
      "loss": 5.05,
      "mean_token_accuracy": 0.18093940913677214,
      "num_tokens": 63072403.0,
      "step": 28795
    },
    {
      "entropy": 5.647373962402344,
      "epoch": 2.8582770940849542,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004191344464463494,
      "loss": 4.987,
      "mean_token_accuracy": 0.18492058366537095,
      "num_tokens": 63084799.0,
      "step": 28800
    },
    {
      "entropy": 5.666636896133423,
      "epoch": 2.8587733227471217,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041910723763523224,
      "loss": 5.0678,
      "mean_token_accuracy": 0.18979636281728746,
      "num_tokens": 63095437.0,
      "step": 28805
    },
    {
      "entropy": 5.672716522216797,
      "epoch": 2.859269551409289,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004190800252505661,
      "loss": 5.0828,
      "mean_token_accuracy": 0.1842593863606453,
      "num_tokens": 63106743.0,
      "step": 28810
    },
    {
      "entropy": 5.622282409667969,
      "epoch": 2.8597657800714567,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004190528092930259,
      "loss": 5.0161,
      "mean_token_accuracy": 0.18073894530534745,
      "num_tokens": 63120525.0,
      "step": 28815
    },
    {
      "entropy": 5.685381364822388,
      "epoch": 2.8602620087336246,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004190255897632863,
      "loss": 5.0222,
      "mean_token_accuracy": 0.18169794231653214,
      "num_tokens": 63131510.0,
      "step": 28820
    },
    {
      "entropy": 5.747107696533203,
      "epoch": 2.860758237395792,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041899836666202254,
      "loss": 5.0762,
      "mean_token_accuracy": 0.1872069790959358,
      "num_tokens": 63143094.0,
      "step": 28825
    },
    {
      "entropy": 5.653612041473389,
      "epoch": 2.8612544660579595,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004189711399899095,
      "loss": 4.9875,
      "mean_token_accuracy": 0.18938495069742203,
      "num_tokens": 63155159.0,
      "step": 28830
    },
    {
      "entropy": 5.631248664855957,
      "epoch": 2.861750694720127,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004189439097476224,
      "loss": 5.0246,
      "mean_token_accuracy": 0.18858589679002763,
      "num_tokens": 63167132.0,
      "step": 28835
    },
    {
      "entropy": 5.652971172332764,
      "epoch": 2.8622469233822945,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004189166759358366,
      "loss": 5.0221,
      "mean_token_accuracy": 0.1896413505077362,
      "num_tokens": 63176864.0,
      "step": 28840
    },
    {
      "entropy": 5.671970653533935,
      "epoch": 2.862743152044462,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041888943855522727,
      "loss": 5.0205,
      "mean_token_accuracy": 0.1808750569820404,
      "num_tokens": 63188003.0,
      "step": 28845
    },
    {
      "entropy": 5.625933599472046,
      "epoch": 2.86323938070663,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041886219760646997,
      "loss": 4.9654,
      "mean_token_accuracy": 0.1822215050458908,
      "num_tokens": 63197634.0,
      "step": 28850
    },
    {
      "entropy": 5.62111644744873,
      "epoch": 2.8637356093687973,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004188349530902402,
      "loss": 4.9352,
      "mean_token_accuracy": 0.1888399228453636,
      "num_tokens": 63207441.0,
      "step": 28855
    },
    {
      "entropy": 5.606763553619385,
      "epoch": 2.864231838030965,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041880770500721346,
      "loss": 4.9883,
      "mean_token_accuracy": 0.18815593421459198,
      "num_tokens": 63217541.0,
      "step": 28860
    },
    {
      "entropy": 5.599077415466309,
      "epoch": 2.8647280666931323,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004187804533580655,
      "loss": 4.9612,
      "mean_token_accuracy": 0.18785772770643233,
      "num_tokens": 63228391.0,
      "step": 28865
    },
    {
      "entropy": 5.65681734085083,
      "epoch": 2.8652242953552998,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004187531981434721,
      "loss": 4.9924,
      "mean_token_accuracy": 0.1921871080994606,
      "num_tokens": 63238462.0,
      "step": 28870
    },
    {
      "entropy": 5.65164155960083,
      "epoch": 2.8657205240174672,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041872593936410923,
      "loss": 5.051,
      "mean_token_accuracy": 0.1798679232597351,
      "num_tokens": 63249305.0,
      "step": 28875
    },
    {
      "entropy": 5.721715021133423,
      "epoch": 2.8662167526796347,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041869867702065266,
      "loss": 5.0563,
      "mean_token_accuracy": 0.1845638707280159,
      "num_tokens": 63260832.0,
      "step": 28880
    },
    {
      "entropy": 5.641789770126342,
      "epoch": 2.866712981341802,
      "grad_norm": 0.875,
      "learning_rate": 0.0004186714111137787,
      "loss": 5.0266,
      "mean_token_accuracy": 0.18260048776865007,
      "num_tokens": 63273689.0,
      "step": 28885
    },
    {
      "entropy": 5.740094470977783,
      "epoch": 2.8672092100039697,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004186441416441632,
      "loss": 5.1605,
      "mean_token_accuracy": 0.17499352842569352,
      "num_tokens": 63284912.0,
      "step": 28890
    },
    {
      "entropy": 5.736366891860962,
      "epoch": 2.867705438666137,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004186168686124825,
      "loss": 5.0683,
      "mean_token_accuracy": 0.17675557136535644,
      "num_tokens": 63296428.0,
      "step": 28895
    },
    {
      "entropy": 5.649750423431397,
      "epoch": 2.8682016673283046,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041858959201941307,
      "loss": 4.9755,
      "mean_token_accuracy": 0.19218114018440247,
      "num_tokens": 63307509.0,
      "step": 28900
    },
    {
      "entropy": 5.692283725738525,
      "epoch": 2.8686978959904725,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041856231186563107,
      "loss": 5.1034,
      "mean_token_accuracy": 0.17435759156942368,
      "num_tokens": 63319194.0,
      "step": 28905
    },
    {
      "entropy": 5.643390655517578,
      "epoch": 2.86919412465264,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041853502815181306,
      "loss": 4.9468,
      "mean_token_accuracy": 0.1845951944589615,
      "num_tokens": 63329122.0,
      "step": 28910
    },
    {
      "entropy": 5.679438066482544,
      "epoch": 2.8696903533148075,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004185077408786357,
      "loss": 5.0279,
      "mean_token_accuracy": 0.18270595371723175,
      "num_tokens": 63340304.0,
      "step": 28915
    },
    {
      "entropy": 5.685539722442627,
      "epoch": 2.870186581976975,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004184804500467756,
      "loss": 5.0289,
      "mean_token_accuracy": 0.1826590418815613,
      "num_tokens": 63351029.0,
      "step": 28920
    },
    {
      "entropy": 5.623853254318237,
      "epoch": 2.8706828106391424,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041845315565690945,
      "loss": 4.9576,
      "mean_token_accuracy": 0.18921599984169007,
      "num_tokens": 63361339.0,
      "step": 28925
    },
    {
      "entropy": 5.705681991577149,
      "epoch": 2.87117903930131,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041842585770971417,
      "loss": 5.018,
      "mean_token_accuracy": 0.1899418205022812,
      "num_tokens": 63373431.0,
      "step": 28930
    },
    {
      "entropy": 5.651039743423462,
      "epoch": 2.871675267963478,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004183985562058667,
      "loss": 4.9928,
      "mean_token_accuracy": 0.18611594438552856,
      "num_tokens": 63384096.0,
      "step": 28935
    },
    {
      "entropy": 5.750449848175049,
      "epoch": 2.8721714966256453,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000418371251146044,
      "loss": 5.0584,
      "mean_token_accuracy": 0.18060240149497986,
      "num_tokens": 63394852.0,
      "step": 28940
    },
    {
      "entropy": 5.714142465591431,
      "epoch": 2.872667725287813,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004183439425309232,
      "loss": 5.0353,
      "mean_token_accuracy": 0.19142523407936096,
      "num_tokens": 63405575.0,
      "step": 28945
    },
    {
      "entropy": 5.645493698120117,
      "epoch": 2.8731639539499803,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004183166303611816,
      "loss": 5.0545,
      "mean_token_accuracy": 0.18561888486146927,
      "num_tokens": 63416677.0,
      "step": 28950
    },
    {
      "entropy": 5.7104710102081295,
      "epoch": 2.8736601826121477,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004182893146374963,
      "loss": 5.1115,
      "mean_token_accuracy": 0.18231411874294282,
      "num_tokens": 63428613.0,
      "step": 28955
    },
    {
      "entropy": 5.712332248687744,
      "epoch": 2.874156411274315,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041826199536054476,
      "loss": 5.0598,
      "mean_token_accuracy": 0.1792968198657036,
      "num_tokens": 63439321.0,
      "step": 28960
    },
    {
      "entropy": 5.67646975517273,
      "epoch": 2.8746526399364827,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004182346725310045,
      "loss": 5.0622,
      "mean_token_accuracy": 0.18246316760778428,
      "num_tokens": 63451279.0,
      "step": 28965
    },
    {
      "entropy": 5.708200216293335,
      "epoch": 2.87514886859865,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041820734614955295,
      "loss": 5.0268,
      "mean_token_accuracy": 0.186353500187397,
      "num_tokens": 63460801.0,
      "step": 28970
    },
    {
      "entropy": 5.710565805435181,
      "epoch": 2.8756450972608176,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004181800162168678,
      "loss": 5.002,
      "mean_token_accuracy": 0.19098189771175383,
      "num_tokens": 63470232.0,
      "step": 28975
    },
    {
      "entropy": 5.653081893920898,
      "epoch": 2.876141325922985,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004181526827336268,
      "loss": 5.0255,
      "mean_token_accuracy": 0.18385687470436096,
      "num_tokens": 63480247.0,
      "step": 28980
    },
    {
      "entropy": 5.681855201721191,
      "epoch": 2.8766375545851526,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004181253457005077,
      "loss": 5.0215,
      "mean_token_accuracy": 0.18108708113431932,
      "num_tokens": 63492104.0,
      "step": 28985
    },
    {
      "entropy": 5.7144664287567135,
      "epoch": 2.8771337832473205,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041809800511818853,
      "loss": 5.0437,
      "mean_token_accuracy": 0.1872404158115387,
      "num_tokens": 63502784.0,
      "step": 28990
    },
    {
      "entropy": 5.723014259338379,
      "epoch": 2.877630011909488,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041807066098734704,
      "loss": 5.1059,
      "mean_token_accuracy": 0.1806424528360367,
      "num_tokens": 63514150.0,
      "step": 28995
    },
    {
      "entropy": 5.688712167739868,
      "epoch": 2.8781262405716554,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004180433133086615,
      "loss": 5.0585,
      "mean_token_accuracy": 0.1841679707169533,
      "num_tokens": 63526682.0,
      "step": 29000
    },
    {
      "entropy": 5.679919624328614,
      "epoch": 2.878622469233823,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004180159620828101,
      "loss": 5.0646,
      "mean_token_accuracy": 0.1805291458964348,
      "num_tokens": 63537448.0,
      "step": 29005
    },
    {
      "entropy": 5.66536512374878,
      "epoch": 2.8791186978959904,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004179886073104709,
      "loss": 4.9861,
      "mean_token_accuracy": 0.1803378477692604,
      "num_tokens": 63547489.0,
      "step": 29010
    },
    {
      "entropy": 5.655525588989258,
      "epoch": 2.879614926558158,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004179612489923224,
      "loss": 4.9837,
      "mean_token_accuracy": 0.19133716374635695,
      "num_tokens": 63558324.0,
      "step": 29015
    },
    {
      "entropy": 5.638166856765747,
      "epoch": 2.880111155220326,
      "grad_norm": 1.078125,
      "learning_rate": 0.000417933887129043,
      "loss": 5.0241,
      "mean_token_accuracy": 0.18508652001619338,
      "num_tokens": 63568630.0,
      "step": 29020
    },
    {
      "entropy": 5.637681007385254,
      "epoch": 2.8806073838824933,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004179065217213111,
      "loss": 4.9577,
      "mean_token_accuracy": 0.1866308331489563,
      "num_tokens": 63578068.0,
      "step": 29025
    },
    {
      "entropy": 5.720435762405396,
      "epoch": 2.8811036125446607,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041787915276980556,
      "loss": 5.07,
      "mean_token_accuracy": 0.178935444355011,
      "num_tokens": 63589103.0,
      "step": 29030
    },
    {
      "entropy": 5.662222242355346,
      "epoch": 2.881599841206828,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041785178027520486,
      "loss": 5.025,
      "mean_token_accuracy": 0.1881968706846237,
      "num_tokens": 63600305.0,
      "step": 29035
    },
    {
      "entropy": 5.75625171661377,
      "epoch": 2.8820960698689957,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041782440423818775,
      "loss": 5.0771,
      "mean_token_accuracy": 0.18473966270685196,
      "num_tokens": 63611910.0,
      "step": 29040
    },
    {
      "entropy": 5.711610221862793,
      "epoch": 2.882592298531163,
      "grad_norm": 1.125,
      "learning_rate": 0.0004177970246594332,
      "loss": 4.9885,
      "mean_token_accuracy": 0.18551447838544846,
      "num_tokens": 63622674.0,
      "step": 29045
    },
    {
      "entropy": 5.680866479873657,
      "epoch": 2.8830885271933306,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041776964153962017,
      "loss": 5.0807,
      "mean_token_accuracy": 0.17463074028491973,
      "num_tokens": 63633274.0,
      "step": 29050
    },
    {
      "entropy": 5.661323165893554,
      "epoch": 2.883584755855498,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041774225487942765,
      "loss": 5.0233,
      "mean_token_accuracy": 0.1879612073302269,
      "num_tokens": 63644738.0,
      "step": 29055
    },
    {
      "entropy": 5.706303215026855,
      "epoch": 2.8840809845176656,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004177148646795348,
      "loss": 5.0867,
      "mean_token_accuracy": 0.17916305512189865,
      "num_tokens": 63655468.0,
      "step": 29060
    },
    {
      "entropy": 5.6394031047821045,
      "epoch": 2.884577213179833,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041768747094062084,
      "loss": 4.9367,
      "mean_token_accuracy": 0.1871682122349739,
      "num_tokens": 63667675.0,
      "step": 29065
    },
    {
      "entropy": 5.684554195404052,
      "epoch": 2.8850734418420005,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00041766007366336494,
      "loss": 5.0647,
      "mean_token_accuracy": 0.18104898929595947,
      "num_tokens": 63679251.0,
      "step": 29070
    },
    {
      "entropy": 5.653007841110229,
      "epoch": 2.8855696705041685,
      "grad_norm": 1.0,
      "learning_rate": 0.00041763267284844675,
      "loss": 4.9826,
      "mean_token_accuracy": 0.18956397473812103,
      "num_tokens": 63690809.0,
      "step": 29075
    },
    {
      "entropy": 5.661154794692993,
      "epoch": 2.886065899166336,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004176052684965456,
      "loss": 4.9907,
      "mean_token_accuracy": 0.1838591992855072,
      "num_tokens": 63702434.0,
      "step": 29080
    },
    {
      "entropy": 5.729284477233887,
      "epoch": 2.8865621278285034,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004175778606083411,
      "loss": 5.1021,
      "mean_token_accuracy": 0.18637105375528334,
      "num_tokens": 63713173.0,
      "step": 29085
    },
    {
      "entropy": 5.724842357635498,
      "epoch": 2.887058356490671,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004175504491845128,
      "loss": 5.0709,
      "mean_token_accuracy": 0.182376666367054,
      "num_tokens": 63723956.0,
      "step": 29090
    },
    {
      "entropy": 5.689053535461426,
      "epoch": 2.8875545851528384,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041752303422574067,
      "loss": 5.0985,
      "mean_token_accuracy": 0.1765409916639328,
      "num_tokens": 63735280.0,
      "step": 29095
    },
    {
      "entropy": 5.744285154342651,
      "epoch": 2.888050813815006,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004174956157327043,
      "loss": 5.0526,
      "mean_token_accuracy": 0.17652880251407624,
      "num_tokens": 63746075.0,
      "step": 29100
    },
    {
      "entropy": 5.68028507232666,
      "epoch": 2.8885470424771738,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004174681937060837,
      "loss": 5.0385,
      "mean_token_accuracy": 0.18972409516572952,
      "num_tokens": 63758152.0,
      "step": 29105
    },
    {
      "entropy": 5.646991634368897,
      "epoch": 2.8890432711393412,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041744076814655893,
      "loss": 5.0444,
      "mean_token_accuracy": 0.1840667814016342,
      "num_tokens": 63768855.0,
      "step": 29110
    },
    {
      "entropy": 5.683372592926025,
      "epoch": 2.8895394998015087,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041741333905481,
      "loss": 5.066,
      "mean_token_accuracy": 0.17522543519735337,
      "num_tokens": 63780258.0,
      "step": 29115
    },
    {
      "entropy": 5.7112708568573,
      "epoch": 2.890035728463676,
      "grad_norm": 1.125,
      "learning_rate": 0.0004173859064315172,
      "loss": 5.0802,
      "mean_token_accuracy": 0.18737806528806686,
      "num_tokens": 63791959.0,
      "step": 29120
    },
    {
      "entropy": 5.708421087265014,
      "epoch": 2.8905319571258437,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041735847027736067,
      "loss": 5.0624,
      "mean_token_accuracy": 0.18103514909744262,
      "num_tokens": 63802859.0,
      "step": 29125
    },
    {
      "entropy": 5.686043834686279,
      "epoch": 2.891028185788011,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004173310305930209,
      "loss": 4.9979,
      "mean_token_accuracy": 0.17763149440288545,
      "num_tokens": 63814275.0,
      "step": 29130
    },
    {
      "entropy": 5.760206794738769,
      "epoch": 2.8915244144501786,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041730358737917823,
      "loss": 5.0419,
      "mean_token_accuracy": 0.1829356372356415,
      "num_tokens": 63824633.0,
      "step": 29135
    },
    {
      "entropy": 5.722727680206299,
      "epoch": 2.892020643112346,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041727614063651313,
      "loss": 5.043,
      "mean_token_accuracy": 0.18369437754154205,
      "num_tokens": 63836478.0,
      "step": 29140
    },
    {
      "entropy": 5.766341590881348,
      "epoch": 2.8925168717745136,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004172486903657065,
      "loss": 5.0524,
      "mean_token_accuracy": 0.17944956868886947,
      "num_tokens": 63847917.0,
      "step": 29145
    },
    {
      "entropy": 5.670679426193237,
      "epoch": 2.893013100436681,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004172212365674388,
      "loss": 5.0636,
      "mean_token_accuracy": 0.1852901592850685,
      "num_tokens": 63857258.0,
      "step": 29150
    },
    {
      "entropy": 5.6676438331604,
      "epoch": 2.8935093290988485,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041719377924239083,
      "loss": 5.1016,
      "mean_token_accuracy": 0.17630672752857207,
      "num_tokens": 63868627.0,
      "step": 29155
    },
    {
      "entropy": 5.725986099243164,
      "epoch": 2.8940055577610164,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041716631839124356,
      "loss": 5.0825,
      "mean_token_accuracy": 0.18576423674821854,
      "num_tokens": 63879954.0,
      "step": 29160
    },
    {
      "entropy": 5.800951528549194,
      "epoch": 2.894501786423184,
      "grad_norm": 1.171875,
      "learning_rate": 0.000417138854014678,
      "loss": 5.1452,
      "mean_token_accuracy": 0.18273855298757552,
      "num_tokens": 63891147.0,
      "step": 29165
    },
    {
      "entropy": 5.676628160476684,
      "epoch": 2.8949980150853514,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004171113861133751,
      "loss": 5.0088,
      "mean_token_accuracy": 0.18046150654554366,
      "num_tokens": 63902380.0,
      "step": 29170
    },
    {
      "entropy": 5.687993335723877,
      "epoch": 2.895494243747519,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004170839146880161,
      "loss": 5.0681,
      "mean_token_accuracy": 0.18421672731637956,
      "num_tokens": 63912557.0,
      "step": 29175
    },
    {
      "entropy": 5.71222186088562,
      "epoch": 2.8959904724096863,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004170564397392821,
      "loss": 5.0896,
      "mean_token_accuracy": 0.18067449778318406,
      "num_tokens": 63923959.0,
      "step": 29180
    },
    {
      "entropy": 5.729063320159912,
      "epoch": 2.896486701071854,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004170289612678545,
      "loss": 4.9927,
      "mean_token_accuracy": 0.18946449160575868,
      "num_tokens": 63935600.0,
      "step": 29185
    },
    {
      "entropy": 5.742109251022339,
      "epoch": 2.8969829297340213,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041700147927441477,
      "loss": 5.1477,
      "mean_token_accuracy": 0.17204033136367797,
      "num_tokens": 63946371.0,
      "step": 29190
    },
    {
      "entropy": 5.702433252334595,
      "epoch": 2.897479158396189,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041697399375964435,
      "loss": 5.0932,
      "mean_token_accuracy": 0.1819624498486519,
      "num_tokens": 63956019.0,
      "step": 29195
    },
    {
      "entropy": 5.7034502029418945,
      "epoch": 2.8979753870583567,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041694650472422467,
      "loss": 5.0444,
      "mean_token_accuracy": 0.180616857111454,
      "num_tokens": 63967330.0,
      "step": 29200
    },
    {
      "entropy": 5.695228719711304,
      "epoch": 2.898471615720524,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041691901216883763,
      "loss": 5.0358,
      "mean_token_accuracy": 0.18674984872341155,
      "num_tokens": 63978952.0,
      "step": 29205
    },
    {
      "entropy": 5.661850118637085,
      "epoch": 2.8989678443826916,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000416891516094165,
      "loss": 5.0585,
      "mean_token_accuracy": 0.1777583658695221,
      "num_tokens": 63989913.0,
      "step": 29210
    },
    {
      "entropy": 5.605976915359497,
      "epoch": 2.899464073044859,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004168640165008884,
      "loss": 4.9641,
      "mean_token_accuracy": 0.1922762781381607,
      "num_tokens": 64000033.0,
      "step": 29215
    },
    {
      "entropy": 5.677633905410767,
      "epoch": 2.8999603017070266,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041683651338968994,
      "loss": 5.0497,
      "mean_token_accuracy": 0.178795425593853,
      "num_tokens": 64010746.0,
      "step": 29220
    },
    {
      "entropy": 5.690529727935791,
      "epoch": 2.900456530369194,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041680900676125147,
      "loss": 5.0123,
      "mean_token_accuracy": 0.18204353302717208,
      "num_tokens": 64021179.0,
      "step": 29225
    },
    {
      "entropy": 5.733297729492188,
      "epoch": 2.9009527590313615,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041678149661625525,
      "loss": 5.0646,
      "mean_token_accuracy": 0.17780177146196366,
      "num_tokens": 64032953.0,
      "step": 29230
    },
    {
      "entropy": 5.639384412765503,
      "epoch": 2.901448987693529,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004167539829553835,
      "loss": 4.9098,
      "mean_token_accuracy": 0.1959208443760872,
      "num_tokens": 64042548.0,
      "step": 29235
    },
    {
      "entropy": 5.643349742889404,
      "epoch": 2.9019452163556965,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041672646577931835,
      "loss": 4.9737,
      "mean_token_accuracy": 0.18412518501281738,
      "num_tokens": 64053254.0,
      "step": 29240
    },
    {
      "entropy": 5.6208594799041744,
      "epoch": 2.9024414450178644,
      "grad_norm": 1.0,
      "learning_rate": 0.00041669894508874225,
      "loss": 5.0281,
      "mean_token_accuracy": 0.1770876631140709,
      "num_tokens": 64063307.0,
      "step": 29245
    },
    {
      "entropy": 5.659515810012818,
      "epoch": 2.902937673680032,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041667142088433766,
      "loss": 5.0241,
      "mean_token_accuracy": 0.186885204911232,
      "num_tokens": 64074274.0,
      "step": 29250
    },
    {
      "entropy": 5.566135311126709,
      "epoch": 2.9034339023421993,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041664389316678713,
      "loss": 4.8849,
      "mean_token_accuracy": 0.20204807817935944,
      "num_tokens": 64084243.0,
      "step": 29255
    },
    {
      "entropy": 5.670801973342895,
      "epoch": 2.903930131004367,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004166163619367733,
      "loss": 4.9897,
      "mean_token_accuracy": 0.18952805995941163,
      "num_tokens": 64095899.0,
      "step": 29260
    },
    {
      "entropy": 5.634149837493896,
      "epoch": 2.9044263596665343,
      "grad_norm": 1.0,
      "learning_rate": 0.00041658882719497876,
      "loss": 4.9277,
      "mean_token_accuracy": 0.18728575706481934,
      "num_tokens": 64107408.0,
      "step": 29265
    },
    {
      "entropy": 5.729316234588623,
      "epoch": 2.9049225883287018,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004165612889420864,
      "loss": 5.0848,
      "mean_token_accuracy": 0.17929170280694962,
      "num_tokens": 64118935.0,
      "step": 29270
    },
    {
      "entropy": 5.698785305023193,
      "epoch": 2.9054188169908692,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004165337471787792,
      "loss": 5.0419,
      "mean_token_accuracy": 0.1823982521891594,
      "num_tokens": 64129085.0,
      "step": 29275
    },
    {
      "entropy": 5.629308223724365,
      "epoch": 2.905915045653037,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041650620190574004,
      "loss": 4.9388,
      "mean_token_accuracy": 0.19082611799240112,
      "num_tokens": 64139605.0,
      "step": 29280
    },
    {
      "entropy": 5.725215768814087,
      "epoch": 2.9064112743152046,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041647865312365186,
      "loss": 5.0742,
      "mean_token_accuracy": 0.1803477242588997,
      "num_tokens": 64151171.0,
      "step": 29285
    },
    {
      "entropy": 5.698294401168823,
      "epoch": 2.906907502977372,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041645110083319805,
      "loss": 5.0549,
      "mean_token_accuracy": 0.17658770233392715,
      "num_tokens": 64162417.0,
      "step": 29290
    },
    {
      "entropy": 5.711456680297852,
      "epoch": 2.9074037316395396,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041642354503506175,
      "loss": 5.038,
      "mean_token_accuracy": 0.17805152982473374,
      "num_tokens": 64172755.0,
      "step": 29295
    },
    {
      "entropy": 5.6910748958587645,
      "epoch": 2.907899960301707,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041639598572992625,
      "loss": 5.0216,
      "mean_token_accuracy": 0.1806837096810341,
      "num_tokens": 64184270.0,
      "step": 29300
    },
    {
      "entropy": 5.685226631164551,
      "epoch": 2.9083961889638745,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041636842291847495,
      "loss": 5.0934,
      "mean_token_accuracy": 0.17828571200370788,
      "num_tokens": 64195400.0,
      "step": 29305
    },
    {
      "entropy": 5.720377349853516,
      "epoch": 2.908892417626042,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041634085660139145,
      "loss": 5.0565,
      "mean_token_accuracy": 0.18087252229452133,
      "num_tokens": 64205101.0,
      "step": 29310
    },
    {
      "entropy": 5.641169309616089,
      "epoch": 2.9093886462882095,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004163132867793592,
      "loss": 4.9395,
      "mean_token_accuracy": 0.18666813373565674,
      "num_tokens": 64216047.0,
      "step": 29315
    },
    {
      "entropy": 5.617502975463867,
      "epoch": 2.909884874950377,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000416285713453062,
      "loss": 5.0333,
      "mean_token_accuracy": 0.18545610457658768,
      "num_tokens": 64226195.0,
      "step": 29320
    },
    {
      "entropy": 5.680472278594971,
      "epoch": 2.9103811036125444,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004162581366231835,
      "loss": 5.0063,
      "mean_token_accuracy": 0.1903948739171028,
      "num_tokens": 64236405.0,
      "step": 29325
    },
    {
      "entropy": 5.678014659881592,
      "epoch": 2.9108773322747123,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041623055629040764,
      "loss": 4.9704,
      "mean_token_accuracy": 0.18703733384609222,
      "num_tokens": 64246989.0,
      "step": 29330
    },
    {
      "entropy": 5.640739679336548,
      "epoch": 2.91137356093688,
      "grad_norm": 1.0,
      "learning_rate": 0.0004162029724554183,
      "loss": 5.007,
      "mean_token_accuracy": 0.18375948816537857,
      "num_tokens": 64257928.0,
      "step": 29335
    },
    {
      "entropy": 5.71713604927063,
      "epoch": 2.9118697895990473,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041617538511889946,
      "loss": 5.0266,
      "mean_token_accuracy": 0.18511397391557693,
      "num_tokens": 64268985.0,
      "step": 29340
    },
    {
      "entropy": 5.793757629394531,
      "epoch": 2.9123660182612148,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004161477942815353,
      "loss": 5.1615,
      "mean_token_accuracy": 0.17983053177595137,
      "num_tokens": 64279907.0,
      "step": 29345
    },
    {
      "entropy": 5.734203863143921,
      "epoch": 2.9128622469233822,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041612019994401,
      "loss": 5.0142,
      "mean_token_accuracy": 0.1884428232908249,
      "num_tokens": 64291496.0,
      "step": 29350
    },
    {
      "entropy": 5.710762453079224,
      "epoch": 2.9133584755855497,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004160926021070078,
      "loss": 5.0234,
      "mean_token_accuracy": 0.18645106106996537,
      "num_tokens": 64302160.0,
      "step": 29355
    },
    {
      "entropy": 5.674535846710205,
      "epoch": 2.913854704247717,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004160650007712132,
      "loss": 5.0847,
      "mean_token_accuracy": 0.18060962706804276,
      "num_tokens": 64313671.0,
      "step": 29360
    },
    {
      "entropy": 5.670097494125367,
      "epoch": 2.914350932909885,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041603739593731046,
      "loss": 5.0649,
      "mean_token_accuracy": 0.17756862193346024,
      "num_tokens": 64325376.0,
      "step": 29365
    },
    {
      "entropy": 5.731511783599854,
      "epoch": 2.9148471615720526,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004160097876059842,
      "loss": 5.0613,
      "mean_token_accuracy": 0.18188437819480896,
      "num_tokens": 64336048.0,
      "step": 29370
    },
    {
      "entropy": 5.781279754638672,
      "epoch": 2.91534339023422,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004159821757779191,
      "loss": 5.1017,
      "mean_token_accuracy": 0.18252268433570862,
      "num_tokens": 64346833.0,
      "step": 29375
    },
    {
      "entropy": 5.635825681686401,
      "epoch": 2.9158396188963875,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041595456045379975,
      "loss": 5.0188,
      "mean_token_accuracy": 0.18165567219257356,
      "num_tokens": 64359656.0,
      "step": 29380
    },
    {
      "entropy": 5.638933897018433,
      "epoch": 2.916335847558555,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004159269416343111,
      "loss": 4.9716,
      "mean_token_accuracy": 0.18325478136539458,
      "num_tokens": 64371130.0,
      "step": 29385
    },
    {
      "entropy": 5.649162673950196,
      "epoch": 2.9168320762207225,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000415899319320138,
      "loss": 5.0167,
      "mean_token_accuracy": 0.18826580047607422,
      "num_tokens": 64381826.0,
      "step": 29390
    },
    {
      "entropy": 5.741985559463501,
      "epoch": 2.91732830488289,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041587169351196536,
      "loss": 5.1281,
      "mean_token_accuracy": 0.17288229912519454,
      "num_tokens": 64392725.0,
      "step": 29395
    },
    {
      "entropy": 5.761675024032593,
      "epoch": 2.9178245335450574,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041584406421047825,
      "loss": 5.0244,
      "mean_token_accuracy": 0.18599190562963486,
      "num_tokens": 64402587.0,
      "step": 29400
    },
    {
      "entropy": 5.676744794845581,
      "epoch": 2.918320762207225,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041581643141636184,
      "loss": 5.0686,
      "mean_token_accuracy": 0.1877944737672806,
      "num_tokens": 64414692.0,
      "step": 29405
    },
    {
      "entropy": 5.655624294281006,
      "epoch": 2.9188169908693924,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004157887951303014,
      "loss": 5.0748,
      "mean_token_accuracy": 0.17872761189937592,
      "num_tokens": 64427278.0,
      "step": 29410
    },
    {
      "entropy": 5.685622262954712,
      "epoch": 2.91931321953156,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004157611553529822,
      "loss": 4.9766,
      "mean_token_accuracy": 0.18406302481889725,
      "num_tokens": 64438221.0,
      "step": 29415
    },
    {
      "entropy": 5.712025833129883,
      "epoch": 2.919809448193728,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041573351208508966,
      "loss": 5.1258,
      "mean_token_accuracy": 0.17376214563846587,
      "num_tokens": 64450139.0,
      "step": 29420
    },
    {
      "entropy": 5.694074296951294,
      "epoch": 2.9203056768558953,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041570586532730926,
      "loss": 5.0305,
      "mean_token_accuracy": 0.18344480246305467,
      "num_tokens": 64461523.0,
      "step": 29425
    },
    {
      "entropy": 5.69264783859253,
      "epoch": 2.9208019055180627,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004156782150803266,
      "loss": 5.1115,
      "mean_token_accuracy": 0.17515574842691423,
      "num_tokens": 64474206.0,
      "step": 29430
    },
    {
      "entropy": 5.6979217529296875,
      "epoch": 2.92129813418023,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041565056134482733,
      "loss": 5.057,
      "mean_token_accuracy": 0.18038298636674882,
      "num_tokens": 64485585.0,
      "step": 29435
    },
    {
      "entropy": 5.758030891418457,
      "epoch": 2.9217943628423977,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004156229041214972,
      "loss": 5.0867,
      "mean_token_accuracy": 0.18211885094642638,
      "num_tokens": 64497237.0,
      "step": 29440
    },
    {
      "entropy": 5.694137096405029,
      "epoch": 2.922290591504565,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004155952434110221,
      "loss": 5.0222,
      "mean_token_accuracy": 0.18303023129701615,
      "num_tokens": 64508235.0,
      "step": 29445
    },
    {
      "entropy": 5.7311766147613525,
      "epoch": 2.922786820166733,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004155675792140879,
      "loss": 5.0371,
      "mean_token_accuracy": 0.18576105684041977,
      "num_tokens": 64520345.0,
      "step": 29450
    },
    {
      "entropy": 5.730460071563721,
      "epoch": 2.9232830488289006,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004155399115313806,
      "loss": 5.0735,
      "mean_token_accuracy": 0.1768611878156662,
      "num_tokens": 64531428.0,
      "step": 29455
    },
    {
      "entropy": 5.745155668258667,
      "epoch": 2.923779277491068,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041551224036358644,
      "loss": 5.0818,
      "mean_token_accuracy": 0.17464009523391724,
      "num_tokens": 64541489.0,
      "step": 29460
    },
    {
      "entropy": 5.652849006652832,
      "epoch": 2.9242755061532355,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041548456571139145,
      "loss": 4.9639,
      "mean_token_accuracy": 0.18946312367916107,
      "num_tokens": 64553026.0,
      "step": 29465
    },
    {
      "entropy": 5.747038412094116,
      "epoch": 2.924771734815403,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041545688757548184,
      "loss": 5.1471,
      "mean_token_accuracy": 0.17680147290229797,
      "num_tokens": 64565988.0,
      "step": 29470
    },
    {
      "entropy": 5.714138221740723,
      "epoch": 2.9252679634775705,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041542920595654423,
      "loss": 5.0218,
      "mean_token_accuracy": 0.1822311371564865,
      "num_tokens": 64578098.0,
      "step": 29475
    },
    {
      "entropy": 5.651210784912109,
      "epoch": 2.925764192139738,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041540152085526476,
      "loss": 4.9308,
      "mean_token_accuracy": 0.1815748855471611,
      "num_tokens": 64588931.0,
      "step": 29480
    },
    {
      "entropy": 5.612627935409546,
      "epoch": 2.9262604208019054,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041537383227233023,
      "loss": 4.9392,
      "mean_token_accuracy": 0.19332271963357925,
      "num_tokens": 64599636.0,
      "step": 29485
    },
    {
      "entropy": 5.690257358551025,
      "epoch": 2.926756649464073,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004153461402084271,
      "loss": 5.0654,
      "mean_token_accuracy": 0.18033911138772965,
      "num_tokens": 64609355.0,
      "step": 29490
    },
    {
      "entropy": 5.6823145866394045,
      "epoch": 2.9272528781262404,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004153184446642421,
      "loss": 5.0163,
      "mean_token_accuracy": 0.1900302767753601,
      "num_tokens": 64619969.0,
      "step": 29495
    },
    {
      "entropy": 5.66778450012207,
      "epoch": 2.927749106788408,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041529074564046204,
      "loss": 4.9827,
      "mean_token_accuracy": 0.18933793306350707,
      "num_tokens": 64629538.0,
      "step": 29500
    },
    {
      "entropy": 5.697700071334839,
      "epoch": 2.9282453354505757,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004152630431377738,
      "loss": 5.0283,
      "mean_token_accuracy": 0.18797388672828674,
      "num_tokens": 64640255.0,
      "step": 29505
    },
    {
      "entropy": 5.641513824462891,
      "epoch": 2.9287415641127432,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004152353371568643,
      "loss": 4.9448,
      "mean_token_accuracy": 0.18846409767866135,
      "num_tokens": 64650051.0,
      "step": 29510
    },
    {
      "entropy": 5.642542171478271,
      "epoch": 2.9292377927749107,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041520762769842065,
      "loss": 4.9944,
      "mean_token_accuracy": 0.19192216992378236,
      "num_tokens": 64660575.0,
      "step": 29515
    },
    {
      "entropy": 5.712416362762451,
      "epoch": 2.929734021437078,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041517991476312994,
      "loss": 5.0593,
      "mean_token_accuracy": 0.18305282294750214,
      "num_tokens": 64671486.0,
      "step": 29520
    },
    {
      "entropy": 5.6858690738677975,
      "epoch": 2.9302302500992456,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004151521983516794,
      "loss": 4.9724,
      "mean_token_accuracy": 0.1831972762942314,
      "num_tokens": 64682315.0,
      "step": 29525
    },
    {
      "entropy": 5.729987716674804,
      "epoch": 2.930726478761413,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004151244784647563,
      "loss": 5.1022,
      "mean_token_accuracy": 0.18301719278097153,
      "num_tokens": 64693940.0,
      "step": 29530
    },
    {
      "entropy": 5.670157861709595,
      "epoch": 2.931222707423581,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004150967551030482,
      "loss": 5.0015,
      "mean_token_accuracy": 0.18746616095304489,
      "num_tokens": 64703347.0,
      "step": 29535
    },
    {
      "entropy": 5.664624977111816,
      "epoch": 2.9317189360857485,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041506902826724234,
      "loss": 5.0715,
      "mean_token_accuracy": 0.17694874107837677,
      "num_tokens": 64714115.0,
      "step": 29540
    },
    {
      "entropy": 5.664454412460327,
      "epoch": 2.932215164747916,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004150412979580264,
      "loss": 4.9695,
      "mean_token_accuracy": 0.1832500234246254,
      "num_tokens": 64724888.0,
      "step": 29545
    },
    {
      "entropy": 5.741007614135742,
      "epoch": 2.9327113934100835,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041501356417608816,
      "loss": 5.0691,
      "mean_token_accuracy": 0.1828896939754486,
      "num_tokens": 64735626.0,
      "step": 29550
    },
    {
      "entropy": 5.685109996795655,
      "epoch": 2.933207622072251,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041498582692211513,
      "loss": 4.9985,
      "mean_token_accuracy": 0.18216715902090072,
      "num_tokens": 64747807.0,
      "step": 29555
    },
    {
      "entropy": 5.64149694442749,
      "epoch": 2.9337038507344184,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041495808619679525,
      "loss": 4.9774,
      "mean_token_accuracy": 0.18948126435279847,
      "num_tokens": 64756845.0,
      "step": 29560
    },
    {
      "entropy": 5.649554872512818,
      "epoch": 2.934200079396586,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041493034200081643,
      "loss": 5.0632,
      "mean_token_accuracy": 0.18243382573127748,
      "num_tokens": 64767316.0,
      "step": 29565
    },
    {
      "entropy": 5.6711358547210695,
      "epoch": 2.9346963080587534,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004149025943348666,
      "loss": 4.9871,
      "mean_token_accuracy": 0.18875080496072769,
      "num_tokens": 64778329.0,
      "step": 29570
    },
    {
      "entropy": 5.643308019638061,
      "epoch": 2.935192536720921,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041487484319963395,
      "loss": 5.0489,
      "mean_token_accuracy": 0.1878663793206215,
      "num_tokens": 64789215.0,
      "step": 29575
    },
    {
      "entropy": 5.652837419509888,
      "epoch": 2.9356887653830883,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004148470885958066,
      "loss": 4.9951,
      "mean_token_accuracy": 0.18214604407548904,
      "num_tokens": 64801878.0,
      "step": 29580
    },
    {
      "entropy": 5.624904489517212,
      "epoch": 2.936184994045256,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004148193305240727,
      "loss": 4.9161,
      "mean_token_accuracy": 0.18760500252246856,
      "num_tokens": 64813442.0,
      "step": 29585
    },
    {
      "entropy": 5.67804102897644,
      "epoch": 2.9366812227074237,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041479156898512076,
      "loss": 4.9896,
      "mean_token_accuracy": 0.1868222713470459,
      "num_tokens": 64824095.0,
      "step": 29590
    },
    {
      "entropy": 5.682849788665772,
      "epoch": 2.937177451369591,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004147638039796391,
      "loss": 5.0252,
      "mean_token_accuracy": 0.18788278698921204,
      "num_tokens": 64833101.0,
      "step": 29595
    },
    {
      "entropy": 5.5912638187408445,
      "epoch": 2.9376736800317587,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041473603550831636,
      "loss": 4.946,
      "mean_token_accuracy": 0.1855123057961464,
      "num_tokens": 64844315.0,
      "step": 29600
    },
    {
      "entropy": 5.685728979110718,
      "epoch": 2.938169908693926,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041470826357184096,
      "loss": 5.053,
      "mean_token_accuracy": 0.18672332763671876,
      "num_tokens": 64855090.0,
      "step": 29605
    },
    {
      "entropy": 5.622514915466309,
      "epoch": 2.9386661373560936,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004146804881709017,
      "loss": 4.998,
      "mean_token_accuracy": 0.18484910130500792,
      "num_tokens": 64865093.0,
      "step": 29610
    },
    {
      "entropy": 5.750680255889892,
      "epoch": 2.939162366018261,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004146527093061873,
      "loss": 5.0999,
      "mean_token_accuracy": 0.17843391448259355,
      "num_tokens": 64876250.0,
      "step": 29615
    },
    {
      "entropy": 5.742556428909301,
      "epoch": 2.939658594680429,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004146249269783867,
      "loss": 5.0323,
      "mean_token_accuracy": 0.18807436227798463,
      "num_tokens": 64886339.0,
      "step": 29620
    },
    {
      "entropy": 5.734189701080322,
      "epoch": 2.9401548233425965,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004145971411881886,
      "loss": 5.117,
      "mean_token_accuracy": 0.180245304107666,
      "num_tokens": 64899473.0,
      "step": 29625
    },
    {
      "entropy": 5.709899663925171,
      "epoch": 2.940651052004764,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041456935193628236,
      "loss": 5.0483,
      "mean_token_accuracy": 0.18614022582769393,
      "num_tokens": 64909919.0,
      "step": 29630
    },
    {
      "entropy": 5.726556873321533,
      "epoch": 2.9411472806669314,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004145415592233569,
      "loss": 5.061,
      "mean_token_accuracy": 0.17999584078788758,
      "num_tokens": 64921069.0,
      "step": 29635
    },
    {
      "entropy": 5.631965065002442,
      "epoch": 2.941643509329099,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041451376305010144,
      "loss": 4.975,
      "mean_token_accuracy": 0.19312846958637236,
      "num_tokens": 64932173.0,
      "step": 29640
    },
    {
      "entropy": 5.657523107528687,
      "epoch": 2.9421397379912664,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004144859634172053,
      "loss": 4.9383,
      "mean_token_accuracy": 0.19112924188375474,
      "num_tokens": 64941650.0,
      "step": 29645
    },
    {
      "entropy": 5.598609256744385,
      "epoch": 2.942635966653434,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041445816032535784,
      "loss": 4.9781,
      "mean_token_accuracy": 0.19210326820611953,
      "num_tokens": 64952520.0,
      "step": 29650
    },
    {
      "entropy": 5.639436388015747,
      "epoch": 2.9431321953156013,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041443035377524844,
      "loss": 5.1243,
      "mean_token_accuracy": 0.1768210157752037,
      "num_tokens": 64965030.0,
      "step": 29655
    },
    {
      "entropy": 5.599715805053711,
      "epoch": 2.943628423977769,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004144025437675668,
      "loss": 4.958,
      "mean_token_accuracy": 0.18966329842805862,
      "num_tokens": 64976260.0,
      "step": 29660
    },
    {
      "entropy": 5.713303756713867,
      "epoch": 2.9441246526399363,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004143747303030025,
      "loss": 5.1219,
      "mean_token_accuracy": 0.18275437653064727,
      "num_tokens": 64988049.0,
      "step": 29665
    },
    {
      "entropy": 5.678568935394287,
      "epoch": 2.9446208813021038,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041434691338224506,
      "loss": 4.985,
      "mean_token_accuracy": 0.1890464574098587,
      "num_tokens": 64998506.0,
      "step": 29670
    },
    {
      "entropy": 5.6771077632904055,
      "epoch": 2.9451171099642717,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004143190930059846,
      "loss": 5.0055,
      "mean_token_accuracy": 0.18224172294139862,
      "num_tokens": 65010533.0,
      "step": 29675
    },
    {
      "entropy": 5.706704759597779,
      "epoch": 2.945613338626439,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004142912691749107,
      "loss": 5.0647,
      "mean_token_accuracy": 0.18316802829504014,
      "num_tokens": 65021238.0,
      "step": 29680
    },
    {
      "entropy": 5.733951902389526,
      "epoch": 2.9461095672886066,
      "grad_norm": 1.125,
      "learning_rate": 0.0004142634418897135,
      "loss": 5.1062,
      "mean_token_accuracy": 0.17584426552057267,
      "num_tokens": 65033251.0,
      "step": 29685
    },
    {
      "entropy": 5.724978637695313,
      "epoch": 2.946605795950774,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004142356111510831,
      "loss": 5.0439,
      "mean_token_accuracy": 0.1833019509911537,
      "num_tokens": 65043845.0,
      "step": 29690
    },
    {
      "entropy": 5.623030948638916,
      "epoch": 2.9471020246129416,
      "grad_norm": 1.0,
      "learning_rate": 0.00041420777695970945,
      "loss": 5.0133,
      "mean_token_accuracy": 0.18845840096473693,
      "num_tokens": 65055395.0,
      "step": 29695
    },
    {
      "entropy": 5.6846967220306395,
      "epoch": 2.947598253275109,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000414179939316283,
      "loss": 5.0448,
      "mean_token_accuracy": 0.18183329105377197,
      "num_tokens": 65064931.0,
      "step": 29700
    },
    {
      "entropy": 5.6932018280029295,
      "epoch": 2.948094481937277,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041415209822149397,
      "loss": 5.0225,
      "mean_token_accuracy": 0.18520502299070357,
      "num_tokens": 65074678.0,
      "step": 29705
    },
    {
      "entropy": 5.6949389457702635,
      "epoch": 2.9485907105994444,
      "grad_norm": 1.125,
      "learning_rate": 0.0004141242536760327,
      "loss": 5.0338,
      "mean_token_accuracy": 0.18668536692857743,
      "num_tokens": 65084562.0,
      "step": 29710
    },
    {
      "entropy": 5.675389432907105,
      "epoch": 2.949086939261612,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004140964056805897,
      "loss": 5.0248,
      "mean_token_accuracy": 0.18454285711050034,
      "num_tokens": 65095924.0,
      "step": 29715
    },
    {
      "entropy": 5.684846687316894,
      "epoch": 2.9495831679237794,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004140685542358556,
      "loss": 5.0595,
      "mean_token_accuracy": 0.1862720400094986,
      "num_tokens": 65106802.0,
      "step": 29720
    },
    {
      "entropy": 5.6983051776885985,
      "epoch": 2.950079396585947,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004140406993425211,
      "loss": 5.0907,
      "mean_token_accuracy": 0.1812806785106659,
      "num_tokens": 65118196.0,
      "step": 29725
    },
    {
      "entropy": 5.66054859161377,
      "epoch": 2.9505756252481143,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004140128410012767,
      "loss": 5.0097,
      "mean_token_accuracy": 0.19303029626607895,
      "num_tokens": 65129924.0,
      "step": 29730
    },
    {
      "entropy": 5.570192480087281,
      "epoch": 2.951071853910282,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041398497921281354,
      "loss": 4.8878,
      "mean_token_accuracy": 0.19425034821033477,
      "num_tokens": 65140994.0,
      "step": 29735
    },
    {
      "entropy": 5.690220689773559,
      "epoch": 2.9515680825724493,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041395711397782236,
      "loss": 5.1018,
      "mean_token_accuracy": 0.18238233923912048,
      "num_tokens": 65153261.0,
      "step": 29740
    },
    {
      "entropy": 5.767699670791626,
      "epoch": 2.9520643112346168,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004139292452969941,
      "loss": 5.1577,
      "mean_token_accuracy": 0.17626996487379074,
      "num_tokens": 65164127.0,
      "step": 29745
    },
    {
      "entropy": 5.643919277191162,
      "epoch": 2.9525605398967842,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041390137317102,
      "loss": 4.9897,
      "mean_token_accuracy": 0.18401235491037368,
      "num_tokens": 65175603.0,
      "step": 29750
    },
    {
      "entropy": 5.649918794631958,
      "epoch": 2.9530567685589517,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004138734976005911,
      "loss": 4.9663,
      "mean_token_accuracy": 0.1807662680745125,
      "num_tokens": 65186780.0,
      "step": 29755
    },
    {
      "entropy": 5.638692331314087,
      "epoch": 2.9535529972211196,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004138456185863986,
      "loss": 4.9403,
      "mean_token_accuracy": 0.19035559445619582,
      "num_tokens": 65198883.0,
      "step": 29760
    },
    {
      "entropy": 5.6387186527252195,
      "epoch": 2.954049225883287,
      "grad_norm": 1.125,
      "learning_rate": 0.00041381773612913413,
      "loss": 5.0199,
      "mean_token_accuracy": 0.18507858663797377,
      "num_tokens": 65209088.0,
      "step": 29765
    },
    {
      "entropy": 5.621559810638428,
      "epoch": 2.9545454545454546,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004137898502294889,
      "loss": 5.0029,
      "mean_token_accuracy": 0.18982888162136077,
      "num_tokens": 65220816.0,
      "step": 29770
    },
    {
      "entropy": 5.691763114929199,
      "epoch": 2.955041683207622,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004137619608881544,
      "loss": 4.9941,
      "mean_token_accuracy": 0.18204448372125626,
      "num_tokens": 65232549.0,
      "step": 29775
    },
    {
      "entropy": 5.693538093566895,
      "epoch": 2.9555379118697895,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004137340681058223,
      "loss": 5.0351,
      "mean_token_accuracy": 0.1803603395819664,
      "num_tokens": 65243691.0,
      "step": 29780
    },
    {
      "entropy": 5.668346309661866,
      "epoch": 2.956034140531957,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041370617188318424,
      "loss": 5.0503,
      "mean_token_accuracy": 0.18601406514644622,
      "num_tokens": 65254651.0,
      "step": 29785
    },
    {
      "entropy": 5.6687438011169435,
      "epoch": 2.956530369194125,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041367827222093195,
      "loss": 5.0122,
      "mean_token_accuracy": 0.18973755240440368,
      "num_tokens": 65264669.0,
      "step": 29790
    },
    {
      "entropy": 5.665396595001221,
      "epoch": 2.9570265978562924,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004136503691197574,
      "loss": 5.0239,
      "mean_token_accuracy": 0.18491706401109695,
      "num_tokens": 65274663.0,
      "step": 29795
    },
    {
      "entropy": 5.68579216003418,
      "epoch": 2.95752282651846,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041362246258035254,
      "loss": 5.0505,
      "mean_token_accuracy": 0.18156137615442275,
      "num_tokens": 65286383.0,
      "step": 29800
    },
    {
      "entropy": 5.648618221282959,
      "epoch": 2.9580190551806274,
      "grad_norm": 1.125,
      "learning_rate": 0.0004135945526034091,
      "loss": 4.885,
      "mean_token_accuracy": 0.18908003270626067,
      "num_tokens": 65297156.0,
      "step": 29805
    },
    {
      "entropy": 5.708209562301636,
      "epoch": 2.958515283842795,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041356663918961956,
      "loss": 5.0718,
      "mean_token_accuracy": 0.182015323638916,
      "num_tokens": 65306836.0,
      "step": 29810
    },
    {
      "entropy": 5.696109771728516,
      "epoch": 2.9590115125049623,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041353872233967595,
      "loss": 5.0851,
      "mean_token_accuracy": 0.1777895376086235,
      "num_tokens": 65318355.0,
      "step": 29815
    },
    {
      "entropy": 5.684665489196777,
      "epoch": 2.95950774116713,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004135108020542705,
      "loss": 5.0609,
      "mean_token_accuracy": 0.18179436773061752,
      "num_tokens": 65330039.0,
      "step": 29820
    },
    {
      "entropy": 5.64874849319458,
      "epoch": 2.9600039698292973,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041348287833409557,
      "loss": 4.9712,
      "mean_token_accuracy": 0.18867619335651398,
      "num_tokens": 65339771.0,
      "step": 29825
    },
    {
      "entropy": 5.661283922195435,
      "epoch": 2.9605001984914647,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004134549511798438,
      "loss": 5.0446,
      "mean_token_accuracy": 0.185467492043972,
      "num_tokens": 65352108.0,
      "step": 29830
    },
    {
      "entropy": 5.669515562057495,
      "epoch": 2.960996427153632,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004134270205922075,
      "loss": 4.9747,
      "mean_token_accuracy": 0.18047998547554017,
      "num_tokens": 65364117.0,
      "step": 29835
    },
    {
      "entropy": 5.674393653869629,
      "epoch": 2.9614926558157997,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041339908657187933,
      "loss": 4.9743,
      "mean_token_accuracy": 0.18701693713665007,
      "num_tokens": 65374168.0,
      "step": 29840
    },
    {
      "entropy": 5.765031623840332,
      "epoch": 2.9619888844779676,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004133711491195522,
      "loss": 5.1715,
      "mean_token_accuracy": 0.17790012508630754,
      "num_tokens": 65385243.0,
      "step": 29845
    },
    {
      "entropy": 5.712383174896241,
      "epoch": 2.962485113140135,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041334320823591857,
      "loss": 5.0258,
      "mean_token_accuracy": 0.17864902317523956,
      "num_tokens": 65396674.0,
      "step": 29850
    },
    {
      "entropy": 5.7398097038269045,
      "epoch": 2.9629813418023025,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041331526392167155,
      "loss": 5.0917,
      "mean_token_accuracy": 0.17234592586755754,
      "num_tokens": 65408121.0,
      "step": 29855
    },
    {
      "entropy": 5.693993425369262,
      "epoch": 2.96347757046447,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000413287316177504,
      "loss": 5.0508,
      "mean_token_accuracy": 0.18227339386940003,
      "num_tokens": 65420093.0,
      "step": 29860
    },
    {
      "entropy": 5.639907741546631,
      "epoch": 2.9639737991266375,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041325936500410905,
      "loss": 4.9725,
      "mean_token_accuracy": 0.18544550538063048,
      "num_tokens": 65431338.0,
      "step": 29865
    },
    {
      "entropy": 5.668152618408203,
      "epoch": 2.964470027788805,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041323141040217976,
      "loss": 4.9912,
      "mean_token_accuracy": 0.18385114818811416,
      "num_tokens": 65442305.0,
      "step": 29870
    },
    {
      "entropy": 5.610756683349609,
      "epoch": 2.9649662564509724,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041320345237240923,
      "loss": 4.9646,
      "mean_token_accuracy": 0.18682418167591094,
      "num_tokens": 65453028.0,
      "step": 29875
    },
    {
      "entropy": 5.673077154159546,
      "epoch": 2.9654624851131404,
      "grad_norm": 1.234375,
      "learning_rate": 0.000413175490915491,
      "loss": 5.0686,
      "mean_token_accuracy": 0.18716893196105958,
      "num_tokens": 65463374.0,
      "step": 29880
    },
    {
      "entropy": 5.687558889389038,
      "epoch": 2.965958713775308,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004131475260321183,
      "loss": 4.9735,
      "mean_token_accuracy": 0.18216246664524077,
      "num_tokens": 65474144.0,
      "step": 29885
    },
    {
      "entropy": 5.64595775604248,
      "epoch": 2.9664549424374753,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041311955772298464,
      "loss": 4.969,
      "mean_token_accuracy": 0.20269307792186736,
      "num_tokens": 65485453.0,
      "step": 29890
    },
    {
      "entropy": 5.679111003875732,
      "epoch": 2.966951171099643,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004130915859887837,
      "loss": 4.9954,
      "mean_token_accuracy": 0.18539386242628098,
      "num_tokens": 65496400.0,
      "step": 29895
    },
    {
      "entropy": 5.737251043319702,
      "epoch": 2.9674473997618103,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004130636108302088,
      "loss": 5.0567,
      "mean_token_accuracy": 0.1812211275100708,
      "num_tokens": 65508524.0,
      "step": 29900
    },
    {
      "entropy": 5.547850561141968,
      "epoch": 2.9679436284239777,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041303563224795393,
      "loss": 4.9204,
      "mean_token_accuracy": 0.19435748606920242,
      "num_tokens": 65519264.0,
      "step": 29905
    },
    {
      "entropy": 5.631857061386109,
      "epoch": 2.968439857086145,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004130076502427128,
      "loss": 4.9129,
      "mean_token_accuracy": 0.19023036062717438,
      "num_tokens": 65531049.0,
      "step": 29910
    },
    {
      "entropy": 5.691957616806031,
      "epoch": 2.9689360857483127,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041297966481517937,
      "loss": 5.061,
      "mean_token_accuracy": 0.18043099194765091,
      "num_tokens": 65543126.0,
      "step": 29915
    },
    {
      "entropy": 5.74925422668457,
      "epoch": 2.96943231441048,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004129516759660475,
      "loss": 5.1026,
      "mean_token_accuracy": 0.17038120627403258,
      "num_tokens": 65555050.0,
      "step": 29920
    },
    {
      "entropy": 5.714630556106568,
      "epoch": 2.9699285430726476,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041292368369601136,
      "loss": 5.0877,
      "mean_token_accuracy": 0.18707094490528106,
      "num_tokens": 65564156.0,
      "step": 29925
    },
    {
      "entropy": 5.7084849834442135,
      "epoch": 2.9704247717348156,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000412895688005765,
      "loss": 5.12,
      "mean_token_accuracy": 0.18239338099956512,
      "num_tokens": 65574639.0,
      "step": 29930
    },
    {
      "entropy": 5.72690258026123,
      "epoch": 2.970921000396983,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004128676888960028,
      "loss": 4.9774,
      "mean_token_accuracy": 0.18757218420505523,
      "num_tokens": 65585628.0,
      "step": 29935
    },
    {
      "entropy": 5.664788389205933,
      "epoch": 2.9714172290591505,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041283968636741903,
      "loss": 4.99,
      "mean_token_accuracy": 0.19050118029117585,
      "num_tokens": 65595544.0,
      "step": 29940
    },
    {
      "entropy": 5.693448495864868,
      "epoch": 2.971913457721318,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004128116804207081,
      "loss": 5.068,
      "mean_token_accuracy": 0.19035024791955948,
      "num_tokens": 65607963.0,
      "step": 29945
    },
    {
      "entropy": 5.711888027191162,
      "epoch": 2.9724096863834855,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004127836710565642,
      "loss": 4.9957,
      "mean_token_accuracy": 0.1956416979432106,
      "num_tokens": 65619881.0,
      "step": 29950
    },
    {
      "entropy": 5.738055324554443,
      "epoch": 2.972905915045653,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041275565827568234,
      "loss": 5.0797,
      "mean_token_accuracy": 0.18088147193193435,
      "num_tokens": 65629967.0,
      "step": 29955
    },
    {
      "entropy": 5.700071001052857,
      "epoch": 2.9734021437078204,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000412727642078757,
      "loss": 5.0193,
      "mean_token_accuracy": 0.1834906443953514,
      "num_tokens": 65640804.0,
      "step": 29960
    },
    {
      "entropy": 5.694824314117431,
      "epoch": 2.9738983723699883,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004126996224664829,
      "loss": 5.1019,
      "mean_token_accuracy": 0.17839098125696182,
      "num_tokens": 65652355.0,
      "step": 29965
    },
    {
      "entropy": 5.6767984390258786,
      "epoch": 2.974394601032156,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041267159943955483,
      "loss": 5.0599,
      "mean_token_accuracy": 0.17426690459251404,
      "num_tokens": 65664047.0,
      "step": 29970
    },
    {
      "entropy": 5.709713363647461,
      "epoch": 2.9748908296943233,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004126435729986678,
      "loss": 5.0225,
      "mean_token_accuracy": 0.18294008672237397,
      "num_tokens": 65674942.0,
      "step": 29975
    },
    {
      "entropy": 5.6961768627166744,
      "epoch": 2.9753870583564908,
      "grad_norm": 1.125,
      "learning_rate": 0.00041261554314451674,
      "loss": 4.9708,
      "mean_token_accuracy": 0.1914064422249794,
      "num_tokens": 65685285.0,
      "step": 29980
    },
    {
      "entropy": 5.624117136001587,
      "epoch": 2.9758832870186582,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004125875098777967,
      "loss": 4.9529,
      "mean_token_accuracy": 0.197966305911541,
      "num_tokens": 65695967.0,
      "step": 29985
    },
    {
      "entropy": 5.661406087875366,
      "epoch": 2.9763795156808257,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041255947319920297,
      "loss": 5.0304,
      "mean_token_accuracy": 0.18372792303562163,
      "num_tokens": 65707434.0,
      "step": 29990
    },
    {
      "entropy": 5.625175714492798,
      "epoch": 2.976875744342993,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041253143310943064,
      "loss": 4.9929,
      "mean_token_accuracy": 0.19387052804231644,
      "num_tokens": 65717836.0,
      "step": 29995
    },
    {
      "entropy": 5.6676288604736325,
      "epoch": 2.9773719730051607,
      "grad_norm": 1.328125,
      "learning_rate": 0.00041250338960917515,
      "loss": 5.0232,
      "mean_token_accuracy": 0.18673622608184814,
      "num_tokens": 65728527.0,
      "step": 30000
    },
    {
      "epoch": 2.9773719730051607,
      "eval_entropy": 5.438166126782324,
      "eval_loss": 5.138014316558838,
      "eval_mean_token_accuracy": 0.1843671524597619,
      "eval_num_tokens": 65728527.0,
      "eval_runtime": 20.6702,
      "eval_samples_per_second": 1628.863,
      "eval_steps_per_second": 203.626,
      "step": 30000
    }
  ],
  "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": 1.635718784311296e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}